Scientific Report 2000 - inis.iaea.org

PAUL SCHERRER INSTITUT ISSN 1423-7318

March 2001

Scientific Report 2000

Volume II

Life Sciences

ed. by: Rolf Jaussi and Beatrice Gschwend

CH-5232 Villigen PSISwitzerland

Phone: 056/310 21 11Telefax: 056/310 21 99

http://bio.web.psi.ch/Life_Sciences-HP.html

PLEASE BE AWARE THATALL OF THE MISSING PAGES IN THIS DOCUMENT

WERE ORIGINALLY BLANK

Table of Contents

I Overview 1

II Structural Biology 3

III Institute of Medical Radiobioiogy 7

IV Center for Radiopharmaceutical Science 21

V Radiation Medicine 49

VI List of Publications 65

I OVERVIEW

F.K. Winkler

INTRODUCTION

The Life Sciences Department has its major focus inactivities aiming to develop novel approaches in can-cer therapy and diagnostics. For this, it makes use ofthe unique PSI facilities (proton beam, radionuclideproduction) and of the know-how that has been builtup in many associated areas of research and devel-opment and in the treatment of patients. Another im-portant effort is the production and evaluation of newPET (positron emission tomography) radiotracers forvarious applications in neurophysiology and drug de-velopment. In line with PSI's general strategy to fosterin-house research on areas using its large scale facili-ties, a new group in structural biology is being set upto exploit the exciting new possibilities provided by theSLS. The activities of the department thus range frommore basic research over the development of newradiopharmaceuticals or radiotracers to actual patienttreatment by proton irradiation. Given the relativelymodest size of the department, collaborations andcontacts with many outside academic, clinical or in-dustrial partners are essential to provide critical massand know-how to these technically diverse activities.

Organizationally, the department is subdivided intofour areas: Radiation Medicine, including the protontherapy and tumor therapy evaluation groups, theCenter for Radiopharmaceutical Science (ZRP), theInstitute of Medical Radiobiology (IMR) and Struc-tural Biology. In addition, TULOC represents a smallproject aiming to develop a method for fast tracking ofthe location of tumors affected by organ motion. ZRPis a joint activity with ETH Zurich and the University ofZurich, and IMR is a joint venture with the University ofZurich. However, only the activities of the respectivePSI branches are reported here.

In the Proton Therapy project 31 patients could betreated on the PSI spot scanning Gantry between Apriland December 2000, bringing the total number treatedsince 1996 to 72. Treatment performance has beenfurther increased through a number of improvementsin hardware, software and other work procedures. Themastering of the demanding nature of this work to-gether with the continuing implementation of diversetechnical improvements represents a remarkableachievement of this group. The great challenge aheadis the realization of the PROSCAN program aiming toprovide the basis for transferring PSI's unique protonbeam spot scanning technology to major hospitals.The Proton Therapy group will play an essential role inPSI's PROSCAN project, which itself represents a jointeffort between several PSI departments.

The established OPTIS program using a scattered,lower energy proton beam for treating ocular malig-nancies, continues to provide excellent clinical results(98% tumor control) for a large number of patients(236 in the year 2000).

In the Tumor Therapy Evaluation group, first treat-ment planning comparisons between proton and pho-ton treatment of canine nasal tumors have been per-formed. Such comparisons based on real canine pa-tient cases help to demonstrate in which cases thespot scanning proton therapy is advantageous by de-livering lower doses to adjacent healthy areas of thebrain. The close collaboration with the veterinary hos-pital of the University of Zurich, where photon therapywith modern treatment planning can now be carriedout on animal patients is a prerequisite for these com-parative studies in treatment planning and actualtreatment. Such studies are expected to yield highlyinformative data in the coming years and will be indis-pensable for the demonstration of the efficacy of pro-ton therapy.

The ZRP group pursues several projects that couldlead to novel radiopharmaceuticals for tumor diagno-sis and therapy. At the same time it continues to ex-plore more general solutions to recurring problemssuch as unstable chemical coupling of radionuclides totargeting molecules or the undesirable accumulationof carrier proteins or peptides in certain organs. Thetricarbonyl labeling method developed here and ideallysuited to incorporate 99mTc or Re easily and stably intopotential radiopharmaceuticals has been further ex-plored with new tridentate ligands and linkers. Themost advanced study using this type of labeling is theneurotensin project intended to provide specific mole-cules for targeting pancreatic tumors. Neurotensinanalogues with suitable pharmacological propertieshave now been developed to the point that clinicaltrials for diagnostic purposes can probably start earlyin 2001. Other projects aim to incorporate 67Cu asradionuclide because of its more favorable radiationcharacteristics compared to other therapeutically usedradionuclides. 67Cu labeled anti-bladder cancer anti-body C595 will be used in first clinical studies nextyear in collaboration with the Queens Medical Centerin Nottingham. The production of sufficient amounts of67Cu is not trivial and relies heavily on the usability ofour proton irradiation station at the PSI cyclotron. Im-proved production schemes have been successfullyexplored.

In collaboration with pharmaceutical companies, thePET radiotracer group within ZRP is trying to developtracers for glutamatergic receptors by making use ofalready known specific ligands. Such tracers would beenormously helpful for the study of cognitive functionsand neurological disorders. The group is also involvedin a number of collaborative projects, such as usingPET tracers for measuring hypoxia in brain tumors.With the very recent implementation of a PET scannerallowing to study small animals with about 1 mm spa-tial resolution, the group now has enhanced capabili-ties for developing new tracers using animal models.

The IMR group investigates new approaches to tumortherapy by trying to target the tumor vasculature in avariety of ways. The group has entered this competi-tive field just three years ago, but has now acquiredthe necessary know-how such that it has been able toachieve some promising results. Last year it could beshown that the basic peptide derived from the HIV Tatprotein and shown to have agonistic properties onVEGF-R2 (vascular endothelial growth factor receptor2) has a binding site distinct from that of the naturalligand. The potential of this alternative binding sitewhich resides primarily on domain 3 of the extracellu-lar part of the receptor will be further explored includ-ing efforts to determine the structure of the domain 3-basic peptide complex. First progress has also beenachieved in characterizing radiation-inducible promot-ers. These are intended to selectively induce the ex-pression of toxic genes introduced for example intocells of tumor vasculature by modest doses of radia-tion.

The newly established Structural Biology group isstill in the build-up phase as far as personnel and pro-ject work are concerned. The infrastructure to carryout experimental work ranging from the cloning andexpression of gene products to their large scale purifi-cation, crystallization and structure determination byX-ray crystallography has been set up and is func-tional. A number of projects in the general area ofspecific biomolecular recognition are at the stage ofprotein production. A major effort starting next year willbe in the area of membrane protein structure determi-nation, in particular of G-protein coupled receptors.

I would like to express my gratitude to all the people inthe department who have contributed to the achieve-ments summarized in this report and also to the nu-merous collaborators at other academic and clinicalinstitutions, whose support and input is instrumental tomany of our projects.

F. Winkler

II STRUCTURAL BIOLOGY

F.K. Winkler

The planning of new structural projects, the build-up ofthe necessary infrastructure and the start of first ex-perimental work determined the activities of this newlyestablished group. Since August the newly designedand furnished wet laboratory is suitably equipped andfunctional. Similarly the new X-ray data collection sys-tem comprising an X-ray rotating anode generator, a2D image plate detector system and auxiliary equip-ment for cryocooling has become operational duringthe second half of the year. A first small structure, thatof a 17aa peptide designed to form an a-helical coiledcoil (see accompanying report) has been successfullysolved establishing the functional state of the crystal-lographic infrastructure.

By the end of 2000 three scientists and one technicianhad joined the group which will further expand in 2001and engage in several new structure determinationprojects. Funding for PhD students has been obtained

in the areas of DNA repair proteins (collaboration withJ. Jiricny, IMR Zurich) and small ribonucleoproteinsthrough two SNSF grant applications. A major long-term effort is planned in the area of membrane proteinstructure determination where our own commitmentwill be substantially enhanced through funding by theNCCR (National Centres of Competence in Research)project 'Molecular Life Sciences: Three DimensionalStructure, Folding and Interactions'. A number of otherprojects addressing questions of specific biomolecularinteractions have been initiated. An example is the in-house collaboration with the IMR group aiming to de-termine the structure of novel VEGF receptor ligandcomplexes.

DE NOVO DESIGN OF AN a-HELICAL COILED COIL THAT IS PREDISPOSED TOFORM AMYLOID FIBRILS

M.O. Steinmetz, D. Kostrewa, C. Garcfa-Echeverrfa1 R. A. Kammerer2

(1Novartis Pharma AG, Basel, Switzerland;2University of Manchester, Manchester, UK)

We here report the de novo design of a 17-residue peptide that forms a trimeric parallel a-helical coiled coilat low temperatures and switches into p-sheet rich amyloid fibrils at elevated temperatures. The peptideoffers a simple model system to elucidate the mechanisms by which many natural proteins transform intofibrils associated with Alzheimer's, Type II diabetes, and prion diseases.

INTRODUCTION

In recent years it became very clear that several di-verse disorders have the same molecular basis: aswitch in the native protein conformation, which trig-gers protein aggregation into amyloid fibrils. Amyloidshave been proposed as causative agents of the so-called amyloids diseases (amyloidosis), such as Alz-heimer's, Type II diabetes, and prion diseases. Thepathway of protein transformation or misfolding andsubsequent fibril formation involves the transition from,e.g., an a-helix to a (3-sheet structure. The a-to-p tran-sition is most apparent in the conversion of prion pro-teins and for the (3-amyloid peptide (A(3), which is amajor component of the amyloid plaques deposited inthe brains of Alzheimer's disease patients. Althoughunrelated proteins are involved in these diseases, itappears that the amyloid fibrils composed of theseproteins have a remarkably similar basic moleculararchitecture. All amyloid fibrils seem to exhibit thefollowing molecular properties, which are now consid-ered to be diagnostic for amyloid (reviewed in 1): (i)Amyloid aggregates examined under the electron mi-croscope (EM) show uniform fibrils about 10 nm indiameter, (ii) All amyloid fibrils can be stained with thediazo dye Congo red, and display a green birefrin-gence when viewed under polarized light, (iii) Amyloidfibrils give raise to characteristic X-ray fibre diffractionpatterns dominated by a 4.7 A meridional and 10 Aequatorial reflections. These reflections are character-istic of a cross-p structure in which the polypeptidechain is organized in p-sheets arranged parallel to thefibre axis with their constituent p-strands perpendicularto the fibril axis.

To understand the basis of amyloidosis at the molecu-lar level, it is important to assess the molecular deter-minants that predispose a protein with a native fold toform amyloids. Peptides based on de novo designhave provided useful information for constructing andmanipulating peptide conformations and elucidatingcomplex folding mechanisms. Therefore, the design ofa preferably short peptide sequence that on the onehand adopts a stable native-like tertiary fold and onthe other hand experiences conformational intercon-version (e.g., a-helix to p-sheet) and fibril formationshould be advantageous in studying the mechanisticdetails of amyloidosis. Because of its simplicity, the a-helical coiled-coil structural motif has frequently been

used as a model system for de novo protein design.Coiled coils consist of two to five right-handed amphi-pathic a-helices coiled around one another in a left-handed manner. Based on recent studies on coiled-coil 'trigger' sequences (2-7) in combination with avail-able information in the literature on monomeric pep-tides that undergo structural transition, we de novodesigned a 17-residue peptide sequence that is ex-pected to fold into both a stable a-helical coiled coiland p-sheet type of structures. The biophysicalproperties of the peptide were investigated by circulardichroism (CD) spectroscopy, X-ray crystallography,Congo red staining, and electron microscopy (EM).

RESULTS AND DISCUSSION

The secondary structure of the synthetic 17-residuepeptide was analyzed by far-ultraviolet (UV) CD spec-troscopy under physiological conditions. At 4 °C, theCD spectrum recorded from the peptide with well-defined minima at 222 and 208 nm was characteristicof substantial a-helical structure.

At 50 °C the conformation of the peptide completelychanged within one hour as evidenced by a strikingchange in the CD spectrum: A single pronouncedminimum at 216 nm was observed indicating a switchin conformation from a-helix into p-sheet-rich struc-ture. Within 2 hours, the intensity of the CD signalstarted to decrease continuously, probably as a con-sequence of formation of higher molecular weightaggregates.

Fig. 1: Micropgraph of crystals of the de novo de-signed 17-residue peptide obtained at 4 °C.Scale bar, 0.15 mm.

Fig. 2: 2 A resolution crystal structure of the trimericparallel a-helical coiled coil. The side views ofseven trimers as observed in the crystal pack-ing are shown schematically.

The structure of the peptide in its a-helical state wasfurther investigated by X-ray crystallography. Asshown in Fig. 1, crystals were obtained at 4 °C thatwere suitable to record a complete data set to a reso-lution of 2 A. The molecular replacement method wasused for structure determination. As shown in Fig. 2,the structure consists of three parallel a-helices thatwrap around each other in a left-handed manner.Consistent with a coiled-coil structure, the hydrophobicside-chains within the hydrophobic core of the trimericmolecule are packed in a 'knobs-into-holes' like fash-ion.

The (3-sheet state of the peptide was analyzed bynegative stain transmission EM. Long and unbranchedfibrils ~8 nm in diameter that twist around each otherto form large bundles of various diameters were re-

vealed. The fibrillar material appears very similar tothe filaments observed in other amylogenic systems(reviewed in 1). Consistent with this observation, addi-tion of Congo red to the sample produced the charac-teristic green birefringence under cross-polarized light,which is highly diagnostic for the presence of amyloid.Furthermore, preliminary X-ray fibre diffraction pat-terns revealed the characteristic 4.7 A meridional and10 A equatorial reflections confirming the presence ofa cross-p structure.

In conclusion, based on the knowledge on coiled-coiltrigger sites and monomeric peptides that undergostructural transition we successfully designed a 17-residue sequence that folds into a stable native-like a-helical protein at low temperatures and switches intocross-p rich amyloid-like fibrils at elevated tempera-tures. The peptide may offer a simple model system toassess the factors that influence protein transforma-tion.

REFERENCES

1. Sunde, M., Blake, C. The structure of amyloid fibrils byelectron microscopy and X-ray diffraction. Adv. Prot.Chem. 50, 123-159, 1997.

2. Steinmetz, M.O., Stock, A., Schulthess, T., Landwehr,R., Lustig, A., Faix, J., Gerisch, G., Aebi, U. Kam-merer, R.A. A distinct 14-residue site triggers coiled-coil formation in cortexillin I. EMBO J. 17, 1883-1891,1998.

3. Kammerer, R.A., Schulthess, T., Landwehr, R., Lustig,A., Engel, J., Aebi, U., Steinmetz, M.O. An autono-mous folding unit mediates the assembly of two-stranded coiled coils. Proc. Natl. Acad. Sci. USA 95,13419-13424, 1998.

4. Frank S., Lustig A., Schulthess T., Engel J., Kam-merer, R.A. A distinct seven-residue trigger sequenceis indispensable for proper coiled-coil formation of thehuman macrophage scavenger receptor oligomeriza-tion domain. J. Biol. Chem. 275, 11672-11677, 2000.

5. Burkhard, P., Kammerer, R.A., Steinmetz, M.O. Aebi,U. The coiled-coil trigger site of the rod domain of cor-texillin I unveils a distinct network of interhelical and in-trahelical salt bridges. Structure 8, 223-230, 2000.

6. Burkhard, P., Meier, M., Lustig, A. Design of a minimalprotein oligomerization domain by a structural ap-proach. Prot. Sci., (in press).

7. Kammerer, R.A., Jaravine, V.A., Frank, S., Schul-thess, T., Landwehr, R., Lustig, A., Garcfa-Echeverria,C, Alexandrescu, A.T., Engel, J. ,Steinmetz, M.O. Anintrahelical salt bridge within the trigger site stabilizesthe GCN4 leucine zipper. J. Biol. Chem., (in press).

INSTITUTE OF MEDICAL RADIOBIOLOGY

Introduction 9

Signalling properties of a human immunodeficiency virus-encoded angiogenic peptidemimicking vascular endothelial growth factor activity 10

VEGF transiently disrupts gap junctional communication in endothelial cells 13

Potentiation of radiotherapy through inhibition of tumour angiogenesis 16

Inhibition of tumour growth by specific targeting of anti-ED-B fibronectin SCFVantibody-modified liposomes in the F9 mouse 18

INTRODUCTION

J. Jiricny

The long-term strategy of the IMR laboratory at thePSI is to exploit tumour blood vessels as targets forcancer therapy.

Tumours with a diameter larger than ~2mm cannotgrow without access to a blood supply, because thecells in their centre become deprived of oxygen andnutrients. In order to survive, these hypoxic cellsproduce stress signals that lead to the secretion ofseveral growth hormones, which diffuse through thesurrounding tissue. Upon reaching a blood vessel,one of these proteins, Vascular Endothelial GrowthFactor (VEGF), brings about permeabilisation of thevessel and induces sprouting of new capillaries,which grow in the direction of the tumour and eventu-ally ensure its continued growth. The endothelial cellsof the newly-formed blood vessels express VEGFreceptors (VEGFRs) on their surface, which arehighly overexpressed in the tumour vasculature. Wepropose to use these receptors in tumour therapy.Our strategy has two main aspects: (i) to use VEGFreceptors for tumour targeting and (ii) to inhibit an-giogenesis (the process of new blood vessel forma-tion) through antagonists of these receptors. In theformer project, we intend to target the tumour vascu-lature with delivery vehicles (e.g. liposomes) carryingligands for VEGFR. These ligands can be either sin-gle chain antibody fragments to VEGFRs, or variantsof the basic peptide of the HIV-1 TAT protein (seereport by Ballmer and colleagues). Coupling of suchligands to liposomes can bring substantial improve-ments in the delivery of cytotoxic agents to the tu-mours (see report by Schwendener and colleagues).The liposomes can be charged also with biologicalagents, such as expression vectors, capable of pro-ducing proteins in situ. Such an approach is the sub-ject of study of the third IMR group (see report byJaussi and colleagues). This study intends to developvectors that carry desired genes under the control ofinducible promoters. Activation of these promoters bye.g. ionizing radiation should bring about localisedexpression of the proteins encoded by these vectors.The choice of proteins is large; they can be toxinsthat directly kill the targeted endothelial cells, or se-creted molecules that inhibit tumour growth. In this

way, it should be feasible to improve the efficacy ofradiotherapy through the expression of these proteinsin the irradiated tissue. Moreover, thanks to the pos-sibility of confining the biological effect solely to theirradiated zone, we should be able to employ agentsthat cannot be used systemically due to their hightoxicity.

The second program of the IMR group aims to iden-tify novel ligands of the VEGFRs, which would beable to bind their respective cognate receptors withhigh affinity and act either as super-agonists or su-per-antagonists. Our initial effort in this project willconcentrate on the synthesis of variants of the TATpeptide, which has been shown to bind to the VEGFreceptor with high affinity. Using novel technologies,we shall mutagenise this peptide and test the result-ing variants in several biological assays, so as toidentify those with the desired properties.

The success of both the above approaches dependson the knowledge of the ligand/receptor system. Incollaboration with the Structural Biology group at thePSI, we intend to co-crystallise the complex of theVEGF receptor(s) with several ligands, specificallythe basic peptide of the HIV-1 TAT protein. In thisway, we should be able to identify the interactiondomains of the proteins and thus be able to modifythem effectively by site-directed mutagenesis. Therecombinant molecules with promising biologicalactivities will then be tested in different in vivo ex-perimental systems, in collaboration with the Radio-pharmacy, Tumour Therapy Evaluation and ProtonTherapy groups at the PSI.

We are aware of the fact that this field has becomeextremely competitive in recent years. However, ourproject targets a specific niche that is of particularinterest and relevance to the PSI, and which willmake use of technologies available almost exclu-sively here. Moreover, the expertise of the presentscientific staff of the IMR laboratory at PSI, KurtBallmer-Hofer, Rolf Jaussi and Reto Schwendener,covers all areas that are required for this tumour tar-geting program to succeed.

10

SIGNALLING PROPERTIES OF A HUMAN IMMUNODEFICIENCY VIRUS-ENCODEDANGIOGENIC PEPTIDE MIMICKING VASCULAR ENDOTHELIAL GROWTH FAC-

TOR ACTIVITY

P. Scheidegger, W. Weiglhofer, St. Suarez, S. Console, J. Waltenberger1, M.S. Pepper2, A. Demirovic,R. Jaussi, K. Ballmer-Hofer

C University of Ulm, Germany;2University of Geneva, Switzerland)

Human immunodeficiency virus, HIV-1, expresses a multifunctional protein called TAT, whose function invivo is tightly correlated with the incidence of Kaposi's sarcoma in AIDS patients. TAT is angiogenic andapparently binds to receptors specific for vascular endotheiial growth factor, VEGF. Amino acids 46-60have been shown to be responsible for functional interaction with VEGF receptors. To further characterizethe binding properties of this peptide, its coding sequence was fused to the reading frame of bacterialthioredoxin (Trx) allowing the production of high amounts of chimeric poiypeptide in bacteria in biologicallyactive form. Binding of chimeric proteins to VEGF receptors was studied in vitro in endotheiial cell culturesexpressing either of the two receptors. Chimeric Trx proteins carrying the basic domain of TAT bound bothVEGF receptors with affinities similar to HIV-TAT or VEGF. Trx/basic peptide chimera are functional ago-nists mediating VEGF receptor signalling. They stimulate growth of endotheiial cells, together with bFGFthey cause tube formation of endotheiial cells in collagen gels, they induce blood vessel formation on thechicken chorioallantoic membrane, and activate VEGF receptor kinase and MAP kinase activity.

INTRODUCTION

A complex series of biological effects is required forsprouting of blood vessels during angiogenesis. VEGFis among the primary factors modulating the behav-iour of endotheiial cells in this process. VEGF is adimeric poiypeptide growth factor related to plateletderived growth factor (PDGF) and specifically acti-vates endotheiial cells upon binding to either of twotyrosine kinase receptors, VEGFR-1 and VEGFR-2(1,2). VEGF receptors are also targeted by the TATprotein expressed by human immunodeficiency virusinfected cells (3-6). This protein is required for tran-scription of viral and cellular genes. However, uponsecretion via an undefined pathway from infectedcells, TAT is also capable to stimulate angiogenesis.This has for instance been observed in Kapos's sar-coma, a condition often observed in AIDS patients. Ashort basic sequence of 15 amino acids apparentlymediates binding to VEGFR-2. TAT is therefore afunctional peptide mimetic for VEGF, a concept that isalso tentatively supported by clinical data (5,7,8).

The fact that this TAT-derived peptide mimics theeffects of VEGF on its cognate receptors is puzzlingfor several reasons. First, a much smaller bindingepitope on the receptor surface is expected to be rec-ognized by this low Mr mimetic as compared with thehighly complex dimeric VEGF. Second, functionalinteraction of receptors with dimeric ligand moleculeshas been shown to dimerize receptors followed byactivation of the intracellular kinase domain. So farthere is no evidence that the basic peptide formsdimers and it will therefore be interesting to dissectthe molecular mechanisms underlying the activity ofthis VEGF mimetic.

RESULTS

To study the binding and signalling properties of thebasic peptide sequence present in HIV-TAT we engi-neered fusion proteins. The basic peptide was ex-pressed in the context of chimeric proteins derivedfrom either bacterial thioredoxin (Trx) or the lambdapD coat protein. We showed that these chimera spe-cifically interacted with both VEGF receptors (for de-tails see annual report 1999). Binding of this peptideto VEGF receptors was highly specific and involvedinteraction with receptor sites that are, at least par-tially, distinct from VEGF binding sites. The fact thatthe basic peptide and VEGF showed weak competi-tion in binding assays might be indicative of allostericinteraction between binding sites specific for thesetwo ligands.

Trx chimeric proteins triggered cell growth to a similarextent as the full length HIV-TAT protein or VEGFi64,while Trx alone was completely inactive.

The biological activity of Trx/basic peptide chimerawas also tested in vivo using the chorio-allantoicmembrane (CAM) in the chicken egg as a model sys-tem for angiogenesis. Trx/basic peptide stimulatedangiogenesis on the CAM, although to a lesser extentthan VEGF.

To further define the epitopes on VEGFR-2 capable tointeract with HIV-TAT, we expressed the extracellulardomains of this receptor in baculovirus-infected insectcells. As shown in Fig. 1 the basic peptide expressedin the context of the lambda phage pD protein specifi-cally interacted with domain 3 while VEGF requiredboth domains 2 and 3 for specific binding. This tenta-tively identifies domain 3 as the primary site of interac-tion between the basic peptide of HIV-TAT and theVEGFR-2 extracellular domain.

11

5000

4000

3000

2000

1000

pDbp Specific bindingi Unspecific binding

KDR(1) KDR(2) KDR(3) KDR(1-3) KDR(1-7) Gelatine

A HUVEeeMs

1 2 3 4

VEGFR-2

1.0 3.3 1,0 5,0

600

500

400

300

200

100

0

VEGF I Specific bindingii Unspecific binding

1KDR(1) KDR(2) KDR(3) KDR(1-3) KDR(1-7) Gelatine

Fig. 1: Association of pD basic peptide chimera withextracellular VEGFR-2 (KDR) domains.

Top panel shows interaction of radioiodinated basicpeptide chimeric pD protein with VEGFR-2 domains 1,2, 3, 1-3, and 1-7, respectively. Bottom panel showssimilar data for VEGF binding. Unspecific binding wasdetermined in the presence of a 100 fold excess ofunlabelled material.

We finally addressed the question whether Trx/basicpeptide chimera and VEGF activated the same path-ways in the cellular signalling network downstreamfrom VEGF receptors. We measured the activity ofVEGFR-2 and the MAP kinase signalling cascade inPAE cells and HUVE cells. Fig. 2A shows an in vitrokinase assay performed with HUVE cells. In vitroautophosphorylation activity of VEGFR-2 was stimu-lated with both VEGF and Trx/basic peptide but notwith a control Trx protein lacking the sequence of theTAT basic peptide. When lysates of control and ago-nist-stimulated PAE cells were used in MAP kinaseassays, we obtained the result shown in Fig. 2B.

Both Trx/basic peptide and VEGF stimulated the activ-ity of the Erk branch of MAP kinases in PAE cellsexpressing VEGFR-2 (left panel). Control PAE cellslacking VEGFR-2 were not stimulated (right panel).This clearly shows that basic peptide, at least qualita-tively, mimics VEGF-induced signalling via VEGFR-2.The fact that control PAE cells were not activated byTrx/basic peptide supports the notion that signalling isVEGFR specific.

B PAE/KDB PAE

S 8

MBP

1.0 1.7 1.8 0,7 1,0 1.0 1,0 0.9

Fig. 2: In vitro kinase activity of VEGFR-2 (A) andMAP kinase (B).

A: Starved HUVE cells were treated for 15 min with:lane 2, 2 nM canine VEGF164; lane 4, 1 nM Trx-bp-loop; lane 5, 1 nM Trx. Lanes 1 and 3 show resultsfrom unstimulated control cells. A quantification of theautoradiographs is shown below each lane, the con-trol value was arbitrarily set to 1.0.

B: Control PAE and PAE/KDR cells expressingVEGFR-2 were stimulated for 15 min with: lanes 2and 6, 1 nM canine VEGF164; lanes 3 and 7, Trx/basicpeptide; lanes 4 and 8, 1 nM Trx. Lanes 1 and 5 showresults from unstimulated control cells. A quantifica-tion of the autoradiographs is shown below each lane,the control value was arbitrarily set to 1.0.

DISCUSSION

Our data extend earlier studies showing that a HIV-TAT-derived peptide sequence promotes angiogene-sis in a variety of systems through interaction withVEGF receptors (3,9,10). VEGF receptors expressedon endothelial cells bind the basic peptide derivedfrom HIV-TAT. In the case of receptor 2 our data sug-gest interaction with extracellular domain 3. Thiswould imply different molecular interactions than pro-posed for VEGF which apparently requires both do-mains 2 and 3 for tight receptor binding.

This leads to the interesting question of how basicpeptide chimera and the full length TAT protein acti-vate VEGF receptors. Stimulation of VEGFR-2 by fulllength TAT and a TAT-derived short basic peptidehave been described earlier without elucidating theunderlying mechanism (3). Tyrosine kinase receptorsundergo changes in conformation upon ligand-induced dimerization that lead to activation of the in-tracellular kinase domain as described earlier. Several

12

mechanisms may be envisioned for the functionalinteraction of the TAT peptide with VEGF receptors: (i)the basic peptide chimera might form dimers that in-teract with two VEGF receptor monomers. This isunlikely for the basic peptide alone that lacks se-quences required for self association yet might play arole in signalling by basic peptide chimera; (ii) bindingof the basic peptide to the extracellular domain ofVEGF receptors might alter the structure of mono-meric receptor molecules. The acquired conformationmight then be translated into structural changes in theintracellular kinase domain that unleash receptor tyro-sine kinase activity; (iii) structural changes in the ex-tracellular receptor domain upon ligand binding mightindirectly cause receptor dimerization through a struc-tural change in receptor monomers. In all cases re-ceptor activation will be followed by recruitment ofcellular signalling proteins into an active signal trans-duction complex located at the cell membrane.

Mechanism of Receptor Dinrserizatson

by pesattvecharges sr.

domain 4 uponligand binding,neutiaHsation offrgaflve charcjss m

domain 3

Basic Peptide-mectsated Receptor Dtmerization

irsd ifect dimei izalionvia domain 4 uponneutiahzafion of

negative charges indomain 3 by basic

Fig. 3: Model based on structural data for PDGF andFms receptor dimerization. Receptors aredimerized by domain 4 upon neutralization ofnegative charges located in domain 3 ofVEGFR-2 that prevent receptor association inthe absence of ligand. This is the conse-quence of a structural change imposed bybinding of VEGF (top) or by direct interactionof domain 3 with the positively charged pep-tide (bottom).

A model for the mechanism of association betweenreceptor monomers by HIV-TAT is given in Fig. 3.This model is currently being verified by biochemicaland structural studies. We also persue the conceptthat the basic peptide used in these studies can beused as a endothelial cell-specific targeting molecule.

REFERENCES

1. Klagsbrun, M., D'Amore, P.A. Vascular endothelialgrowth factor and its receptors. Cytok. Growth Fact.Rev., 7:259-270, 1996.

2. Mustonen, T., Alitalo, K. Endothelial receptor tyrosinekinases involved in angiogenesis. J. Cell. Biol., 129.895-898, 1995.

3. Albini, A., Soldi, R., Giunciuglio, D., Giraudo, E., Ben-elli, R., Primo, L, Noonan, D., Salio, M., Camussi, G.,Rockl, W., Bussolino, F. The angiogenesis induced byHIV-1 tat protein is mediated by the Flk-1/KDR recep-tor on vascular endothelial cells. Nature Med., 2 :1371-1375, 1996.

4. Albini, A., Benelli, R., Presta, M., Rusnati, M., Ziche,M., Rubartelli, A., Paglialunga, G., Bussolino, F., andNoonan, D. HIV-tat protein is a heparin-binding angio-genic growth factor. Oncogene, 72: 289-297, 1996.

5. Ensoli, B., Gendelman, R., Markham, P., Fiorelli, V.,Colombini, S., Raffeld, M., Cafaro, A., Chang, H.K.,Brady, J.N., and Gallo, R.C. Synergy between basic fi-broblast growth factor and HIV-1 Tat protein in induc-tion of Kaposi's sarcoma. Nature, 371: 674-680, 1994.

6. Albini, A., Barillari, G., Benelli, R., Gallo, R. C , Ensoli,B. Angiogenic properties of human immunodeficiencyvirus type 1 Tat protein. Proc. Natl. Acad. Sci. U.S.A.,92:4838-4842, 1995.

7. Ensoli, B., Cafaro, A. HIV-1 and Kaposi's sarcoma.(Review) (16 refs). Eur. J. Cancer Prev., 5: 410-412,1996.

8. Gallo, R.C. The Enigmas of Kaposi's Sarcoma. Sci-ence, 282: 1837-1839, 1998.

9. Ganju, R.K., Munshi, N., Nair, B.C., Liu, Z.Y., Gill, P.,Groopman, J.E. Human immunodeficiency virus tatmodulates the Flk-1/KDR receptor, mitogen-activatedprotein kinases, and components of focal adhesion inKaposi's sarcoma cells. J. Virol., 72:6131-6137, 1998.

10. Mitola, S., Sozzani, S., Luini, W., Primo, L, Borsatti,A., Weich, H., Bussolino, F. Tat-human immunodefi-ciency virus-1 induces human monocyte chemotaxisby activation of vascular endothelial growth factor re-ceptor-1. Blood, 90: 1365-1372, 1997.

13

VEGF TRANSIENTLY DISRUPTS GAP JUNCTIONAL COMMUNICATIONIN ENDOTHELIAL CELLS

S. Suarez, K. Ballmer-Hofer

Endothelial cells form multiple types of cell-cell junctions that are required for cellular organization intocomplex networks. These junctions also regulate communication among adjacent cells. Stimulation byvarious growth factors such as Epidermal growth factor (EGF) or Platelet derived growth factor (PDGF)has been shown to disrupt cell-cell junctions, consequently affecting cell-to-cell communication. We areinterested in early signalling events downstream from Vascular endothelial growth factor (VEGF) receptorsthat affect blood vessel homeostasis. We investigated gap junctional communication (GJC) by monitoringthe transfer of a low molecular weight fluorescent tracer molecule between adjacent cells using im-munofluorescence microscopy. VEGF maximally blocked GJC 15 minutes after growth factor administra-tion. The cells resumed communication via gap junctions within 1-2 hours after treatment. This early effectof VEGF on communication correlated with changes in the phosphoryiation state of one of the proteinsinvolved in gap junction formation, connexin 43 (Cx43). The signalling mechanisms involved in this phe-nomenon depend on activation of VEGF receptor 2, impinge on a tyrosine kinase of the Src family andactivate the Erk family of MAP kinases. The function of VEGF-mediated disruption of GJC might be to limitan increase in endothelium permeability to the local environment in injured blood vessels.

INTRODUCTION

Cell-to-cell junctions determine the morphology ofepithelial and endothelial cell layers lining organ cavi-ties and blood vessels. The initial contact sites re-sponsible for maintaining cell junctions in epithelialand endothelial cell monolayers are formed by homo-typic cadherin/cadherin interactions. The formation ofthese so called adherens junctions is a prerequisitefor tight junction assembly. Tight junctions link adja-cent cells and allow only regulated transfer of mole-cules between the apical and the basal side of epithe-lial cell layers. Similarly, tight sealing of endothelialcell monolayers ensures that blood vessels are sealedtowards the interstitial side of vascularized tissues.Cells in contact with each other form also another typeof junctions, called gap junctions. These latter struc-tures are formed by a family of proteins called connex-ins (Cx) (1). Six connexin molecules form eitherhomo- or heteromeric connexons that, upon apposi-tion in adjacent cells, form a pore spanning the mem-branes of the cells in contact with eachother andbridge the extracellular space between the cells. GJsare a means through which cells rapidly exchangeinformation in the form of low Mr molecules such assecond messengers, e.g. Ca2+, or cellular metabolites.In endothelial cells connexins Cx37, Cx40 and Cx43are concomitantly expressed giving rise to connexonsof various composition and specificity (2). Coupling ofcells via GJs is essential for normal cell function andis reduced under many pathological conditions, e.g. intumor tissue (3-5). The exact role of these junctions inphysiological processes such as the regulation of cellmorphology, migration, growth and differentiation is,however, only poorly understood.

We are interested in the role of VEGFs, a subfamily ofthe PDGF family of polypeptide growth factors, inregulating cell-cell communication. Addition of eitherof two members of this ligand family, VEGF-A or pla-centa growth factor (PIGF), causes changes in vesselhomeostasis (6,7). The role of VEGF family proteinshas been studied in tissue culture models, in ex vivo

vessel preparations and in intact vessels in animals(8-13). Here we use endothelial cell monolayers gen-erated from umbilical vein-derived cells to study GJcommunication. We found that VEGF-A, but not PIGF,rapidly and reversibly disrupts GJs. The signallingcascade stimulated by VEGF-A activates VEGF re-ceptor 2 (VEGFR-2) followed by activaton of the c-Srctyrosine kinase and MAP kinases (MAPK) of the Erksubfamily. The activation of these kinases may berequired directly or indirectly for phosphoryiation ofCx43 giving rise to altered connexon function.

RESULTS

To study whether VEGF-stimulated cells still commu-nicated with each other through GJs we microinjectedsingle cells with a solution of lucifer yellow (LY). Wethen quantified the number of cells adjacent to theinjected cell that became LY positive within 15 min-utes using a fluorescence microscope. VEGF treat-ment maximally reduced GJC within 15-30. Commu-nication was reestablished to normal levels within oneto two hours (for details see annual report 1999).

We next investigated which of the two angiogenicreceptors for VEGF family growth factors, VEGFR-1or VEGFR-2, mediates disruption of GJC by usingreceptor-specific VEGF isoforms. VEGF-A activatesboth VEGFR-1 and -2 while a pox family virus-encoded variant, VEGF-E, exclusively activatesVEGFR-2 (14). A more distantly related family mem-ber, PIGF, is specific for VEGFR-1 (15). Fig. 1 showsthat the 164 amino acid isoform of VEGF-A andVEGF-E, but not PIGF, cause disruption of GJC indi-cating that receptor 2 activates the biochemical sig-nals impinging on GJs. These findings were furthercorroborated by the fact that a recently describedVEGFR-2-specific inhibitor, CGP41251 (16) com-pletely blocked disruption of GJC by VEGF-A.

14

oD)C

u

80

60

50

40

30

20

10

Fig. 1:

0 30 120Time after growth factor addition (min)

^ B VEGF1G4 (50ng/ml)— VEGF-E (50ng/ml)gsssi PIGF (SOng/ml)i 1 VEGF164 (50ng/ml) + 100 |jM CGP41251

VEGF-A and VEGF-E, but not PIGF, modulateGJC in Ea.hy926 cells. Cells were treated with50 ng/ml VEGF-A, 50 ng/ml VEGF-E or 50ng/ml PIGF for up to 120 minutes.

VEGFR-2 activates many signalling pathways, themost prominent targets being PLC(, PI-3 kinase, MAPkinases, PKC, and Src family tyrosine kinases (15).The exact role of these signalling intermediates in thevarious responses of endothelial cells to VEGF re-mains unclear. A specific inhibitor of c-Src blockedVEGF-induced disruption of GJC (Fig. 2). Similarly,U0126, an inhibitor of MAP kinase kinase (MEK) thatis specific for the Erk branch of MAPK signalling cas-cades, also blocks disruption of GJC by VEGF (Fig.2). Therefore, both enzymes are necessary for elicit-ing the response of VEGF on cell communication.

Connexins, the transmembrane proteins that form gapjunctions, are susceptible to phosphorylation by avariety of kinases. Phosphorylation at serine,threonine and tyrosine residues is responsible, atleast in part, for the regulation of GJs. Addressing theputative mechanism of disruption of GJC we studiedphosphorylation of connexin 43. Cx43 was transientlyphosphorylated in immunoprecipitates made fromextracts of VEGF-stimulated cells. Phosphorylation ofCx43 was alkali-labile indicative of serine or threonine.This suggests, but does not formally prove, that modi-fication of Cx43 is responsible for the disruption ofGJs.

30 120

Time after VEGF164 addition (min)

30 120

Time after VEGF164 addition (min)

Fig. 2: Effect of c-Src or MAPK inhibition on VEGF-induced changes in GJC in Ea.hy926 cells.Cells were stimulated with 50 ng/ml VEGF-Ain the presence of 1|j,M CGP77675 (top), or1|iMU0126 (bottom).

We finally investigated GJs microscopically usingCx43-specific antibodies (Fig. 3). All antibodies gavehigh cytoplasmic fluorescence that may result fromthe presence of a large pool of soluble material notinvolved in junction formation. A small fraction of Cx43was found in discrete membrane-associated patchesas described before (17). VEGF did not change themorphology or number of these structures indicatingthat Cx43 phosphorylation may directly affect con-nexon function, rather than turnover.

Fig. 3: Immunofluorescence analysis of gap junctionsin Ea.hy926 endothelial cells. Control cells orcells treated with 50 ng/ml VEGF-A were la-belled with a Cx43-specific antibody. Arrow-heads mark patches of membrane-boundCx43 protein presumably representing GJs.

15

DISCUSSION

The data presented here show that VEGF disruptsGJC in endothelial cells. These findings depict a veryrapid and fully reversible biological readout throughVEGFR-2. The characterization of the signallingpathways responsible for disruption of GJs indicatesthat both c-Src and MAPK are required for this newlydiscovered VEGF activity. These kinases are rapidlyand transiently activated upon treatment with VEGF-Aand reach maximal activity after about 15 minutes inagreement with earlier published work (18).

The role of this newly discovered readout fromVEGFR-2 in blood vessel homeostasis is at presentunclear. In a natural situation disruption of GJs byVEGF might be an acute response to growth factorreleased by endothelial cells or pericytes upon vesseldamage. The effect of VEGF on GJC might preventthe propagation of second messenger moleculesthrough cells coupled via GJs in the endothelial layerof the vessel wall. Ca2+, for instance, is known tospread among coupled endothelial cells in blood ves-sels. Blocking GJC will restrict the propagation of sig-nals initiated at a site of local vessel injury into theadjacent healthy part of a vessel presumably prevent-ing excessive vessel leakage that might otherwisecause massive edema. In support of this idea,changes in GJC between endothelial and smoothmuscle cells have been recently reported to modulateendothelium-dependent relaxation of rabbit arteries invessel explants (11).

REFERENCES

1. Goodenough, D.A., Goliger, J.A., Paul, D.L. Connex-ins, connexons, and intercellular communication.Annu. Rev. Biochem., 65: 475-502, 1996.

2. Beyer, E.C., Gemel, J., Seul, K.H., Larson, D.M., Ba-nach, K., Brink, P.R. Modulation of intercellular com-munication by differential regulation and heteromericmixing of co-expressed connexins. Braz. J. Med. Biol.Res., 33:391-397,2000.

3. Simon, A.M. Gap junctions: more roles and new struc-tural data. Trends Cell Biol., 9: 169-170, 1999.

4. Sulkowski, S., Sulkowska, M., Skrzydlewska, E. Gapjunctional intercellular communication and carcino-genesis. Pol. J. Pathol., 50: 227-233, 1999.

5. Simon, A.M. and Goodenough, D.A. Diverse functionsof vertebrate gap junctions. Trends Cell Biol., 8: 477-483, 1998.

6. Bates, D.O., Lodwick, D., Williams, B. Vascular endo-thelial growth factor and microvascular permeability.Microcirculation, 6: 83-96, 1999.

7. Veikkola, T., Alitalo, K. VEGFs, receptors and angio-genesis. Semin. Cancer Biol., 9: 211-220, 1999.

8. Ferrara, N. Alitalo, K. Clinical applications of angio-genic growth factors and their inhibitors. Nat. Med., 5:1359-1364, 1999.

9. Dvorak, H.F. VPF/VEGF and the angiogenic response.Semin. Perinatol., 24: 75-78, 2000.

10. Tallquist, M.D., Soriano, P., Klinghoffer, R.A. Growthfactor signaling pathways in vascular development.Oncogene, 18: 7917-7932, 1999.

11. Chaytor, AT., Evans, W.H., Griffith, T.M. Central roleof heterocellular gap junctional communication in en-dothelium-dependent relaxations of rabbit arteries. J.Physiol., 508:561-573, 1998.

12. Schnittler, H.J. Structural and functional aspects ofintercellular junctions in vascular endothelium. BasicRes. Cardiol., 93 Suppl 3: 30-39, 1998.

13. Brink, P.R., Ricotta, J., Christ, G.J. Biophysical charac-teristics of gap junctions in vascular wall cells: implica-tions for vascular biology and disease. Braz.J.Med.Biol. Res., 33: 415-422, 2000.

14. Wise, L.M., Veikkola, T., Mercer, A.A., Savory, L.J.,Fleming, S.B., Caesar, C, Vitali, A., Makinen, T., Ali-talo, K., Stacker, S.A. Vascular endothelial growth fac-tor (VEGF)-like protein from orf virus NZ2 binds toVEGFR2 and neuropilin-1. Proc.Natl.Acad.Sci.U.S.A.,96:3071-3076, 1999.

15. Petrova, T.V., Makinen, T., Alitalo, K. Signaling viavascular endothelial growth factor receptors. Exp. CellRes., 253: 117-130, 1999.

16. Fabbro, D., Buchdunger, E., Wood, J., Mestan, J.,Hofmann, F., Ferrari, S., Mett- H, Reilly, T., Meyer, T.Inhibitors of protein kinases: CGP 41251, a proteinkinase inhibitor with potential as an anticancer agent.Pharmacol. Ther., 82: 293-301, 1999.

17. Yao, J., Morioka, T., Oite, T. PDGF regulates gapjunction communication and connexin43 phosphoryla-tion by PI 3-kinase in mesangial cells. Kidney Int., 57:1915-1926,2000.

18. Kroll, J., Waltenberger, J. The Vascular EndothelialGrowth Factor Receptor KDR Activates Multiple SignalTransduction Pathways in Porcine Aortic EndothelialCells. J. Biol. Chem., 272: 32521-32527, 1997.

16

POTENTIATION OF RADIOTHERAPY THROUGH INHIBITION OFTUMOUR ANGIOGENESIS

C. Chastel, O. Guillaume-Gentil, K. Ballmer-Hofer, R. Jaussi

As a continuation of our efforts to improve the efficacy of radiotherapy through the inhibition of tumour-specific vasculature, we are constructing radiation-inducible vectors capable of expressing anti-angiogenicfactors. The approach is based on three levels of specificity: I) targeted irradiation of tumour II) selectiveexpression of anti-angiogenic factor(s) in irradiated tissue and III) targeting of the radiation-inducible vec-tors to the vascular endothelial growth factor receptors, which are generally overexpressed in tumourblood vessels.

INTRODUCTION

Rationale of anti-angiogenic cancer therapy

The efficacy of cancer treatment is often limited by thelack of selectivity of the therapeutics for tumour tissueand through the rapid development of resistant tumourcells. The former problem is linked to the fact thatchemotherapeutics in use today generally bring aboutthe destruction of all rapidly-dividing cells. This leadsto the common side effects such as hair loss, immunedeficiency and damage to healthy tissues and organs.The latter problem relates to the heterogeneity of tu-mours, which contain almost without exception ge-netically unstable cell populations that display highermutation rates than normal cells. Treatment of suchtumours encourages the outgrowth of cell populationsthat are resistant to that particular therapy regimenand leads to the emergence of therapy-resistant tu-mours.

Anti-angiogenic treatment of tumours represents anew development, which ought to be free of theabove-mentioned problems. The field was pioneeredby a few research groups, notably those of JudahFolkman at Harvard Medical School in Boston (1) andRobert Kerbel in Toronto. The work of these groupshas sparked many consecutive studies in the field (2).Currently, more than 60 clinical trials employing anti-angiogenic strategies are underway (for detailed in-formation, see http://cancertrials.nci.nih. gov/) reflect-ing the high pace of the developments. By now, threegenerations of anti-angiogenic strategies have arisen(see Table 1). The initial strategy aimed at the directapplication of anti-angiogenic factors (e.g. endostatin),which were efficacious in immuno-compromised micecarrying human tumour xenografts (see ref. 1). Duringthe second stage, the combinations of two differentanti-angiogenic strategies were tried in similar mousemodels (3). The next step in the development of im-proved therapies will involve the combination of clas-sical therapies with anti-angiogenic strategies. Thistype of therapy is the goal of our work, which basesalso on progress of other PSI research projects (K.Ballmer-Hofer, R. Schwendener cf. this Scientific Re-port).

Table 1 : Development of Anti-Angiogenic Therapies.

Firstgeneration

Anti-angiogenesis

Secondgeneration

Anti-angiogenesis

Anti-angiogenesis

A

B

Thirdgeneration

Anti-angiogenesis

Radiotherapy

Tumor targetting

Development of radiation-inducible promoters

Many naturally-existing promoters are induced byionizing radiation. Usually these promoters are in-volved in pleiotropic stress responses and are induc-ible by many toxic treatments. The aim of our currentresearch is the identification and characterization ofseveral existing promoters for their suitability for theproposed clinical application. The ideal promotershould exhibit low basal levels of transcription, butshould be activated many-fold following irradiation (1-2 Gy). Because the response of most known promot-ers has been tested only within the context of the in-tact gene, we have to evaluate the properties of theisolated promoter regions in suitable expression vec-tors. We also need to know how the promoters be-have upon fractionated irradiation. In order to be ableto answer these questions, we are developing a ver-satile test system, employing both the promoter regionand a reporter region as cassettes in small plasmids.Upon transfection into human cells, we can monitorthe expression levels of reporter genes under thecontrol of the chosen promoters (Fig. 1). We usemainly the luciferase reporter, which permits sensitiveluminescence assays. These assays also display abroad range of linearity of the response over severalorders of magnitude.

17

Promoter withBinding sites for:

NF-KappaBEGFMYesst Gal TF

Nco I /

I Hind Ilk**"*"*^ I

Reporter Region with:LwdferaseGFPEndostatin

f ^~5700bp

Eco NI

Fig. 1: Plasmids to test radiation-inducible promoters.

COQ. 30x10

o£ 25x10

<J 20x103 '

SI(0

o

15x10

10x10

>• 5x10

CO O

RESULTS AND DISCUSSION

Expession of anti-angiogenic proteins

We have obtained cDNAs encoding the human anti-angiogenic factors endostatin and angiostatin (InVi-voGen) and Pigment Epithelium-Derived Factor(PEDF, from N. Bouck, Chicago). All cDNAs havebeen subcloned into bacterial, yeast or mammalianexpression vectors and we are currently attempting tooptimise the yields of the expressed proteins. Wehave obtained high levels of intracellular endostatin inbacteria and human cells and we have human cellsthat express secreted PEDF at low level. The effect ofthis medium on irradiated endothelial cells can now betested.

Development of radiation-inducible promoter

The problems associated with high basal levels oftranscription can be avoided by using a promoter ofnon-human origin with no cognate transcription factorin human cells. One such a promoter is the yeast Galpromoter. The yeast Gal transcription factor interactswith this promoter through its DNA binding domainand activates transcription through its signalling do-main. If this latter domain is replaced by a humansignalling domain, for example the signalling domainof Elk (a human transcription factor stimulated in irra-diated cells), the resulting fusion protein (Gal-Elk) willactivate genes with Gal promoter sequences in hu-man cells after irradiation. We have tested this hy-pothesis by transfecting human endothelial cells withtwo plasmids expressing the Gal-Elk transcriptionfactor and harbouring the Gal-Promoter-Luciferasereporter system, respectively. As anticipated, the sys-tem is activated after irradiation in NIH3T3 cells (Fig.2) as well as in endothelial cells (Fig. 3).

These data are very promising and permit us to pro-ceed with the construction of radiation-inducible vec-tors that will express anti-angiogenic factors specifi-cally in irradiated tissues.

Fig. 2: Induction of Gal-Elk system by X-irradiation inNIH3T3 fibroblasts. C, Control (mock) trans-fection; S, 10% serum stimulation, NT, nottransfected.

Fig. 3: Induction of Gal-Elk system by X-irradiation inporcine aorta endothelial cells. C, Control(mock) transfection; S, 10% serum stimula-tion, NT, not transfected.

REFERENCES

1. Boehm, T., Folkman, J., Browder, T., O'Reilly, M.S.Antiangiogenic therapy of experimental cancer doesnot induce acquired drug resistance. Nature, 390: 404-407, 1997.

2. Kerbel, R.S. Tumor angiogenesis: past, present andthe near future. Carcinogenesis, 21: 505-515, 2000.

3. Klement, G., Baruchel, S., Rak, J., Man, S., Clark, K.,Hicklin, D.J., Bohlen, P., Kerbel, R.S. Continuous low-dose therapy with vinblastine and VEGF receptor-2 an-tibody induces sustained tumor regression withoutovert toxicity. J Clin Invest, 105: R15-R24, 2000.

18

INHIBITION OF TUMOUR GROWTH BY SPECIFIC TARGETING OF ANTI-ED-B Fl-BRONECTIN SCFV ANTIBODY-MODIFIED LIPOSOMES IN THE F9 MOUSE

TERATOCARCINOMA MODEL

C. Marty, K. Ballmer-Hofer, D. Neri1 R. Klemenz2, R. A. Schwendener

1ETH Zurich, 2University Hospital Zurich

Small unilamellar immunoliposomes containing single chain antibody fragments (scFv) specific for the ED-B domain of the B-fibronectin isoform which is found only in newly formed blood vessels during tumourangiogenesis were prepared. ScFv fragments were functionalized by introduction of 1-3 cysteines at theC-terminus separated from the binding domains by hydrophilic amino acid spacers. The scFv were linkedvia cysteine thiols to liposomes by formation of S-C-bridges to maleinimide groups located at the terminalends of poly(ethylene glycol) (PEG) modified phospholipids. The binding properties of the liposomes wereanalyzed in vitro on ED-B expressing Caco-2 cells, and in vivo by biodistribution of114mIndium labeled anti-ED-B scFv-PEG-liposomes in mice bearing s.c. F9 tumours. Compared to control PEG-liposomes thescFv-immunoliposomes accumulated in the tumours at 2-3 fold higher concentrations during the first 2 hafter i.v. injection. Later (6-24 h) both liposome types accumulated at comparable amounts (8-10% ID/gtumour). In a therapy experiment mice were treated during 5 days, every 24 h with PEG-liposomes ascontrols and with anti-ED-B-scFv-immunoliposomes, both containing the cytotoxic dimer 5-FdU-NOAC. Areduction of tumour volume by 50-60% was observed with the ED-B specific liposomes, whereas the un-specific control liposomes exerted no significant cytotoxic activity. Immunoliposomes targeted to ED-Bpositive tumours bind specifically to ED-B fibronectin in vitro and in vivo representing a promising drugdelivery system for the anti-angiogenic treatment of solid tumours.

INTRODUCTION

The physiology of solid tumours differs from that ofnormal tissues in a number of important aspects.Compared with the regular ordered vasculature ofnormal tissues, blood vessels in tumours have distin-guished capillaries with leaky walls and sluggish bloodflow (1). Tumour growth requires continuous formationof new blood vessels (angiogenesis). The switch tothe angiogenic phenotype of endothelial cells involvesa change in the local equilibrium between positiveangiogenic regulators (fibroblast growth factors, vas-cular endothelial growth factor (VEGF), placentalgrowth factor, transforming growth factor, etc.) andnegative angiogenic regulators (thrombospondin-1,angiostatin, etc.) (2).Because tumour growth depends on angiogenesis,the selective delivery of toxic agents to specific targetslocated on newly formed blood vessels could offer animportant new therapeutic benefit. VEGF is one of themost important positive angiogenic factors. Therefore,its receptors, VEGFR1 and KDR/VEGFR2 could rep-resent promising markers for angiogenesis (3, 4).Formation of new blood vessels is also strongly linkedto the extracellular matrix (ECM). Molecules in theextracellular matrix such as fibronectins (5), integrins,(6) or selectins (7) are upregulated or induced onstimulated angiogenic vessels. Blocking of ECM pro-teins could therefore represent an additional tool toinhibit angiogenesis. Furthermore, proteases involvedin extracellular matrix degradation are important forendothelial cell migration and angiogenesis. Potentialtargets on tumour blood vessels are molecules thatare expressed on the endothelial cells as shown inFig. 1.

Fig. 1: Targets in the tumour vasculature: A) mole-cules on the endothelial surface exposed tothe blood stream; B) molecules on the endo-thelial surface exposed to the basementmembrane; C) molecules on the basementmembrane; D) molecules on the extracellularmatrix of the luminal side of endothelial cells;E) molecules on the extracellular matrix ofabluminal side of endothelial cells.

We used the ED-B isoform of fibronectin, a compo-nent of the extracellular matrix that is exclusively ex-pressed in tumour tissue, as target for specific cyto-toxic immunoliposomes. Antibody modified liposomesrepresent an interesting system for target cell specificdelivery of cytotoxic molecules (8).

Starting from an anti-ED-B fibronectin single chainantibody fragment (scFv) obtained from Prof. D. Neri(ETH Zurich) 4 different functionalized scFv antibodieswere constructed and produced in large quantities inthe yeast P. pastoris (9). Attachment of proteins to thedistal end of the maleinimide modified PEG chains onliposomes was done by incubation of the functional-ised scFv fragments in presence of the reducing agenttributylphosphine. Modified liposomes and unreacted

19

scFv were separated by metrizamide gradient cen-trifugation.

RESULTS

The binding properties of scFv-PEG-immuno-liposomes were demonstrated in vitro on cell culturesusing ED-B positive Caco-2 and ED-B negative Co-115 cells as shown in Fig. 2.

Fig. 2: Binding of fluorescence labeled scFv-PEG-immuno-liposomes to cells cultured on colla-gen. A,C,E: Phase contrast images of thesections shown in B, D, F; B: a-ED-B scFv-PEG-immunoliposomes on ED-B positiveCaco-2 cells; D: unmodified PEG-liposomeson ED-B positive Caco-2 cells; F: a-ED-BscFv-PEG-immunoliposomes on ED-B nega-tive Co-115 cells.

Based on the specific binding found in the in vitrostudies, the in vivo properties of the a-ED-B scFv-PEG-immunoliposomes were investigated by determi-nation of pharmacokinetic properties, tumour uptakeand finally in a therapy experiment using the murineF9 teratocarcinoma grown subcutaneously in nudemice.

To assess the pharmacokinetic data and tumour up-take, the liposomes were labelled by a remote loadingtechnique with 114mlndium (Fig. 3, Ref. 10).

Fig. 4 summarizes the accumulation in the tumours ofthe specific scFv-PEG-immunoliposomes, as com-pared to the non-specific PEG-liposomes. The ED-Bfibronectin-specific liposomes accumulated in thetumours at higher levels only during the first hoursafter administration. At time points of 6 and 24 h, thePEG-liposomes reached equal values.

To evaluate the cytotoxic activity of ED-B fibronectinspecific liposomes, a therapy experiment was per-formed, in which we used the lipophilic nucleosidedimer 5-FdU-NOAC (11) as the cytotoxic drug. Asshown in Fig. 5, a reduction of tumour growth by 50-60% was obtained with the scFv-PEG-immunoliposomes containing 5-FdU-NOAC, as com-pared to the empty scFv-liposomes.

a" 4 m l n ! *

" 4 m l n "

Cl

114mFig. 3: Scheme of In labelling of liposomes.114mln3+ ions diffuse through an ionophorechannel into the liposome. Inside a highly sta-ble complex with nitrilotriacetic acid (NTA) isformed.

110°19 scFv-PEG-liposomes _ T

Fig. 4: Tumour accumulation of a-ED-B scFv-liposomes. Nude mice bearing s.c. F9 tu-mours were injected i.v. with 11 mln liposomesand killed at different time points.

400-

200-

- • - Controls untreated

- • - empty liposomes

- ¥ - scFv liposomes

-A-scFv-dimers

- * - 5FdU-M3AC liposomes

•HB- scFv-SFdU-NOAC-liposomes

t t tTreatments (i.v.)

Days

Fig. 5: Effect of a-ED-B scFv-liposomes on tumourgrowth in vivo. Nude mice bearing s.c. F9 tu-mours were treated i.v. 5 times every 24h withliposomes containing 600 ja.g 5FdU-NOAC.

20

DISCUSSION

Four new constructs of oc-ED-B fibronectin scFv anti-bodies containing C-terminal cysteines were producedin P. pastoris and successfully attached to the surfaceof cytotoxic liposomes. The oc-ED-B scFv-CM3 anti-body containing 3 cysteines and a spacer between thecysteines and the protein had the best coupling prop-erties.The cytotoxic effect detected with the scFv-5FdU-NOAC-PEG-liposomes are very a promising basis forfurther experiments with immunoliposomes by appli-cation of other target molecules, other therapeuticschedules or other cytotoxic agents (12).

REFERENCES

1. Folkman, J. Angiogenesis in cancer, vascular, rheuma-toid and other disease. Nat. Med., 1: 27-31, 1995.

2. Battegay, E.J. Angiogenesis: mechanistic insights,neovascular diseases, and therapeutic prospects. J.Mol. Med., 73: 333-346, 1995.

3. Neufeld, G., Cohen, T., Gengrinovitch, S. & Poltorak,Z. Vascular endothelial growth factor (VEGF) and itsreceptors. FASEB J., 13: 9-22, 1999.

4. Brown, L.F., Berse, B., Jackman, R.W., Tognazzi, K.,Manseau, E.J., Senger, D.R. & Dvorak, H.F. Expres-sion of vascular permeability factor (vascular endothe-lial growth factor) and its receptors in adenocarcino-mas of the gastrointestinal tract. Cancer Res., 53:4727-4735, 1993.

5. Carnemolla, B., Neri, D., Castellani, P., Leprini, A.,Neri, G., Pini, A., Winter, G., Zardi, L. Phage antibod-ies with pan-species recognition of the oncofoetal an-giogenesis marker fibronectin ED-B domain. Int. J.Cancer, 68: 397-405, 1999.

6. Brooks, P.C., Montgomery, A.M., Rosenfeld, M., Reis-feld, R.A., Hu, T., Klier, G., Cheresh, D.A. Integrin al-pha v beta 3 antagonists promote tumour regressionby inducing apoptosis of angiogenic blood vessels.Cell, 79: 1157-1164, 1994.

7. Nguyen, M., Corless, C.L., Kraling, B.M., Tran, C ,Atha, T., Bischoff, J., Barsky, S.H. Vascular expressionof E-selectin is increased in estrogen-receptor-negative breast cancer: a role for tumour-cell-secretedinterleukin-1 alpha. Am. J. Pathol., 150. 1307-1314,1997.

8. Schwendener, R.A., Trub, T., Schott, H., Langhals, H.,Barth, R.F., Groscurth, P., Hengartner, H. Comparativestudies of the preparation of immunoliposomes withthe use of two bifunctional coupling agents and inves-tigation of in vitro immunoliposome-target cell bindingby cyto-fluorometry and electron microscopy. Biochim.Biophys. Acta, 1026: 69-79, 1990.

9. Marty, C, Scheidegger, P., Ballmer-Hofer, K., D.,Klemenz, R., Schwendener, R.A. Production of func-tionalized single-chain Fv antibody fragments bindingto the ED-B domain of the B-isoform of fibronectin inPichia pastoris. Protein Expr. Purific, 27:156-164,2001.

10. Proffitt, R.T., Williams, L.E., Presant, C.A., Tin, G.W.,Uliana, J.A., Gamble, R.C., Baldeschwieler, J.D. Tu-mour-imaging potential of liposomes loaded with In-111-NTA: biodistribution in mice. J. Nucl. Med., 24: 45-51, 1983.

11. Cattaneo-Pangrazzi, R.M.C., Schott, H., Wunderli-Allenspach, H., Derighetti, M.I., and Schwendener,R.A. The novel hetero-dinucleoside dimer 5-FdU-NOAC is a potent cytotoxic drug and a p53-independent inducer of apoptosis in the androgen-independent human prostate cancer cell lines PC-3and DU-145. The Prostate, 45: 8-18, 2000.

12. Marty, C , Ballmer-Hofer, K., Neri, D., Klemenz, R.,Schott, H., Schwendener, R.A. Inhibition of tumourgrowth by specific targeting of anti-ED-B fibronectinscFv antibody modified liposomes in the F9 mouseteratocarcinoma model. Cancer Res. (in preparation,2001).

21

IV CENTER FOR RADIOPHARMACEUTICAL SCIENCE

Center for Radiopharmaceutical science of PSI, ETHZ and UNIZ -

Overview of the research activities 23

Targetting renal tumours with anti-L1-CAM mAb chCE7: levels ofL1-CAM expression and phosphorylation of L1-CAM in renal cells 25

In vivo stability of 67Cu-DOTA-F(ab')2 fragments: effect of shorteninga triglycine linker sequence 26

High kidney accumulation of radiometal labelled single chain antibodyfragments: does chemistry make a difference? 28

Chemical modifications on model single chain Fv fragments for improvedin vivo behaviour in tumour targetting and therapy 30

Integrin antagonists for tumour targetting: characterisation of RGD-thioredoxinfusion proteins 31

Neurotensin(8-13) analogues induce Ca2+ mobilisation in human prostate

carcinoma PC-3 cells 32

Neurotensin derivatives: quality control in view of human application 33

Cold rhenium complexes of neurotensin analogues 34Pharmacological evaluation of a 99mTc(l)-labelled bombesin analogue for GRPreceptor-targetted scintigraphy 35

Improved preparation of [M(OH2)3(CO)3]+ M = 99mTc, 188Re) and its applicationfor radiolabelling of various functionalised biomolecules 36

Functionalisation of glucose at position C-2 and C-6 for the labelling with99mTc-tricarbonyl 38

Functionalisation of glucose at position C-3 for the labelling with99mTc(l)-tricarbonyl 40

In-Out-lsomerism of NS3-cage compounds for the in vivo stabilisation of Ag-111 41

Synthesis, radiolabelling and characterisation of potential PET Tracers

for the glutamatergic system 42

Tumour hypoxia measurements by positron emission tomography 45

Small animals PET imaging at PSI 46

RCBF changes associated with the presentation of pleasant stimuli inheroin addicts 47

23

CENTER FOR RADIOPHARMACEUTICAL SCIENCE OF PSI, ETHZ AND UNIZOVERVIEW OF THE RESEARCH ACTIVITIES

PA. Schubiger

TUMOR TARGETING

The use of targeted radiotherapeutics applied locallyor systemically has become a very hot topic as a po-tentially efficient and adjuvant therapy for specificcancer types (e.g. lymphoma, neuroblastoma, bladdercancer). In many cases, targeted radionuclide therapyhas been introduced into the clinical practice. Thecollaboration with Techniclone/Berlex led to the suc-cessful transfer of the know-how for the high dose 1311labelling of anti-lymphoma antibody LYM-1 to thecompany's production site in the USA. A clinical imag-ing study with 1 3 1 I - labelled anti-neuroblastoma anti-body chCE7 could be concluded in collaboration withthe Netherland's Cancer Institute in Amsterdam andthe results look so promising that clinical work will becontinued towards a dose escalation study. After aninitial diagnostic study, in the course of next year, aphase I therapeutic study with a 67Cu- labelled anti-bladder carcinoma antibody is initiated in collaborationwith the Queens Medical Center in Nottingham.

In our research efforts to further enhance the efficacyof targeted radionuclide therapy we focus on the im-provement of existing target molecules, investigationsof new promising target agents, development of suit-able labelling and linker chemistry and characterisa-tion of the radioconjugates in appropriate in vitro andin vivo systems.

Targeting of renal tumours with anti- L1-CAM chCE7antibody is being explored after having found out ear-lier that in addition to neuroblastoma, many humanrenal carcinoma cell lines express high levels of L1-CAM protein. We could show that in renal tumour cellsoverexpression of L1-CAM leads to a loss of control ofits phosphorylation by a growth factor, which is crucialfor the normal development of the kidney.

In order to further improve the in vivo stability and thetumour uptake of 67Cu-labelled F(ab)2 fragmentsvarious peptide- linked DOTA chelates were testedand the length of the linker sequence was found tohave an important effect on the in vivo stability andtumour uptake. A number of studies were aimed at thereduction of the kidney accumulation of " m T c -tricarbonyl labelled antibody single chain (scFv) frag-ments. When the charge of scFv protein backbone orof the polyhistidine tag was modified by chemicalmeans, a significant decrease of unspecific kidneyand liver uptake could be achieved. To further extendthe application of our new 99mTc/*Re-tricarbonyl label-ling technique to larger size proteins and to the use ofmilder reaction conditions, we synthesised tailor madechelators with a cadaverine linker, which enables en-zymatic attachment of the radiolabel to lysine orglutamine residues of the antibody fragments.

In the search for novel tumour seeking agents suitablefor nuclide therapy, integrin antagonists are beingevaluated. Fusion proteins containing the integrin-binding RGD sequence grafted into thioredoxin asbackbone are being produced in e.coli (in collabora-tion with the IMR's PSI group) and the fusion proteinshave been tested on PC3 human prostate carcinomacells.

Within a European BIOMED collaboration on tumourfinding peptides we progressed to the selection ofseveral stable neurotensin (NT) analogues labelledwith 99mTc-tricarbonyl. All compounds have beenbroadly characterised as described earlier and werefurthermore fully and properly characterised by NMRand MS. Recently we found in addition that differ-ences in affinity observed with the various NT deriva-tives are correlated with differences in their agonisticactivity.

One of the stable NT peptides has been selected forfirst imaging studies of pancreatic tumours in patientsat the University Hospital in Lausanne. The carbonyllabelling technology could be applied successfully witha newly introduced chelating group PADA to the neu-ropeptide bombesin, a further very promising tumourseeking peptide. This radioconjugate shows high affin-ity for the GRP receptor and may be selected as afuture agent for the imaging of prostate cancer.

The application of our 99mTc/*Re-tricarbonyl labellingmethod to ever more molecules led us to continueimprovement of carbonyl preparations in terms ofhigher specific activities and lower toxicity during syn-thesis and handling. The CO generation was facili-tated, novel tridentate ligands were developed andapplied to molecules of high interest such as biotin. Inthis case labelling with unprecedented high specificactivity was achieved, which will be of eminent impor-tance in therapeutic applications using multistep pre-targeting approaches involving 188/186Re-labeled bio-tin. Besides 188/I86pje w e continued to focus on 67Cuand 111Ag as therapeutic radionuclides emitting 13-particles. As source for 188Re we have access to aW/Re generator, whereas for 67Cu we rely on ourcyclotron. In order to improve the amount of activityneeded for preparing therapeutic doses of 67Cu-labeled antibodies, we initiated a new productionscheme based on the enriched target material 68Zn.The applicability of 111Ag is still hindered by the lackof suitable chelators. Although highly interesting in-sights on its co-ordination chemistry could be gained,no useful chelator could be found.

We pursued our efforts to synthesise glucose deriva-tives with the 99mTc-tricarbonyl label as a substitutefor 2-18F-deoxyglucose. Whereas the latter needs

24

PET scanning technology, a 99mTc-glucose tracercould be used on the more widely available SPET-scanners. Functionalisation of the glucose must notimpair biological activity, such as the ability to betransported into the cell by the GLUT 1 transporterand to be metabolised (partially) by hexokinase. Sev-eral derivatives with the various linkers and the Tc-tricarbonyl label at the position C-2, C-3 and C-6 havebeen synthesised and characterised. So far noneshowed all desired properties.

PET TRACERS FOR FUNCTIONAL IMAGING

The PET-T racer research program at PSI and Zurichcould profit greatly from the installment of a secondhuman PET-scanner at the university hospital, since50% of its scanning time is reserved for research pur-poses. A further event with great impact is the arrivalof the HIDAC animal PET scanner (resolution 1mm) atthe PSI. It will go into operation next year and greatlyenhance our possibilities in assisting drug develop-ment.

The research efforts concentrated on PET-receptortracers for the glutamate neuroreceptors. This systemis of great importance in cognitive functions and themany neurological disorders associated with them.

However no useful PET tracer has yet been found. Incollaboration with the pharmaceutical industry (Novar-tis, Boehringer) we synthesised several 11C-labeledmolecules. Although the pharmacological in vitrocharacterisation revealed promising properties forsome compounds, the corresponding animal experi-ments did not fulfil the expectations. This was particu-larly true for tracers binding to the PCP-binding site inthe ionotropic NMDA receptor. Our future experimentswill now concentrate on developing tracers for themetabotropic glutamate receptor.

In addition we currently assess the usefulness of 18F-fluoromisonidazole to measure hypoxia of malignantbrain tumours with the goal of a better treatment plan-ning of radiotherapy. This study is performed in col-laboration with the Cantonal Hospital Aarau and theUniversity Veterinary Hospital in Zurich. Finally, thelast PET-study, initiated some years ago byK. Leenders (now in Groningen), has been success-fully completed. The study compared how the brainprocesses pleasant stimuli in young healthy controlswith the processing in a group of former opiate ad-dicts. The results appear to confirm that opiate addic-tion is associated with changes in the processing ofemotionally salient information in the brain.

25

TARGETTING RENAL TUMOURS WITH ANTI L1-CAM MAB CHCE7: LEVELS OFL1-CAM EXPRESSION AND PHOSPHORYLATION OF L1-CAM IN RENAL TUMOR CELLS

/. Novak-Hofer, M. Heiz

In vivo phosphorylation of LI-CAM was measured in human 293 embryonal kidney cells and in renal tu-mour cells. Glial Derived Neurotrophic Factor (GDNF) stimulates L1-CAM phosphorylation in the low ex-pressing 293 cells and in Caki-2 renal carcinoma cells which show a 3 to 5 fold lower expression of LI-CAM than Foehn renal carcinoma cells. No effect of GDNF on the phosphorylation of LI-CAM in the over-expressing Foehn cells was found. The results indicate that overexpression of L1-CAM results in a loss ofcontrol of its phosphorylation by GDNF.

Our earlier results on normal tissue reactivity of anti-neuroblastoma mAb chCE7 obtained via Northern blotand Western blot analysis, indicated that human braintissue is the only normal tissue showing high expres-sion (both at the mRNA and protein level) of L1-CAM,the target antigen of mAb chCE7. In contrast, humankidney tissue showed high levels of L1-CAM mRNAand low levels of L1-CAM protein. When we analyseddirect binding of 131l-mAb chCE7 to tissue sections ofhuman brain and kidney, we found high binding of themAb to brain sections and low, but detectable bindingto kidney sections (Fig. 1).

BRAIN

BRAIN NS

KIDNEY

KIDNEY NS

Fig. 1: Electronic autoradiography of human tissuesections after binding of 131l-mAb chCE7. NS:non-specific binding in the presence of a 100fold excess of mAb chCE7.

We had found earlier that many human renal carci-noma cell lines and a subset of renal tumour tissuesections express high levels of L1-CAM protein, anobservation that serves as the basis for our ongoingevaluation of mAb chCE7 for the targeting of renaltumours.

In this study of L1-CAM phosphorylation in renal car-cinoma cells, we investigated, if the levels of L1-CAMexpression affect the sensitivity of renal cells to GlialDerived Neurotrophic Factor (GDNF), a growth factoressential for renal development.

Low expressing K 293 human embryonal kidney cells,low expressing renal carcinoma Caki-2 cells and highexpressing Foehn cells were chosen. Cell surface

expression of L1-CAM was measured by binding of125l-chCE7 to these cells. Scatchard analysis of bind-ing data gave a Bmax of 6000 sites/ 293 cell, 40 000sites/ Caki-2 cell and a Bmax of 200 000 sites/ Foehncell. In vivo incorporation of 32P into L1-CAM wasmeasured in subconfluent serum-deprived cultures.Cultures were labelled with 0.5 mCi of 32P/ml for 2 hin low serum medium and GDNF was added for thelast 30 min of the labelling period. L1-CAM was iso-lated from cell lysates by immunoprecipitation withmAb chCE7 and immunoprecipitates were separatedby SDS-PAGE and analysed by electronic autoradio-graphy. Fig. 2 shows that GDNF leads to a dose de-pendent increase in phosphate incorporation into the200 000 Mr band corresponding to L1-CAM in thecase of 293 cells (A) and of Caki-2 (C) cells. In con-trast, no effect of GDNF on L1-CAM phosphorylationin Foehn cells (B) is observed. The results indicatethat overexpression of L1 -CAM leads to a loss of con-trol of its phosphorylation by a growth factor which iscrucial for the normal development of the kidney.

A B

Fig. 2: Phosphorylation of L1-CAM in embryonalrenal cells (A) and renal tumor cell lines (B,C)* 5 ng/ml, * * 50 ng/ml GDNF, geni: 50 |aMGenistein. Arrow: 200 kD form of L1-CAM.

26

IN VIVO STABILITY OF 67CU-DOTA-F(ab')2 FRAGMENTS: EFFECT OF SHORTEN-ING A TRIGLYCINE LINKER SEQUENCE

K. Zimmermann, N. Riesen, I. Novak-Hofer

F(ab')2 fragments of anti colon cancer antibody mAb 35 were derivatised with peptide-iinked DOTA che-lates and labelled with 67Cu. Biodistributions in tumour bearing mice as well as in vivo metabolism in theliver was compared. Results demonstrate that in vivo stability and tumour uptake of the 67Cu-F(ab')2 con-jugates are influenced by the length of the linker sequence.

The tumour targeting ability of 67Cu-labeled immuno-conjugates is strongly influenced by the copper che-lates used for labelling. Both the charge of the che-lates and the cleavage patterns of the copper com-plexes from the protein play an important role. In ourongoing effort to improve the biodistributions of 67Cu-labeled antitumour antibody fragments, a number ofmacrocyclic copper chelates (DOTA derivatives R1-R5) were synthetised (Figure 1).

HOOC ' \ / >—COOH

D0TA-R3

D0TA-R4

PA-DOTA

Fig. 1: Peptide-iinked DOTA chelates for 67Cu label-ling.

F(ab')2 fragments prepared from mAb 35 were deriva-tised with R1-R4, labelled with 67Cu, purified by FPLCsize exclusion gel chromatography and evaluated forin vivo stability and biodistributions in mice bearingcolon carcinoma xenografts. We recently found, thattumour uptake and tumour/tissue ratios of 67Cu-DO3A-F(ab')2 fragments can be improved by a trigly-cine linker (R1) and that this effect correlates withenhanced in vivo stability of this conjugate. Bioevalua-tion of R2 and R3 indicated that the amino acid com-position of the linker has a decisive effect on in vivostability and on biodistributions: Replacing the glyresidue in the middle of the sequence by a pro residue(R3) did not affect the stability of the conjugate,whereas a gly-phe-gly linker sequence (R2) resulted

in a high initial uptake of radioactivity in the liver aswell as in lower tumour uptake. We have evidence,that this effect is due to rapid proteolytic cleavage atthe phe residue, resulting in reduced stability of thisconjugate (Annual Report 1999).

Fig. 2: Tumour and blood levels (% injected dose/ gtissue on the y-axis) of 67Cu-F(ab')2 in tumourbearing mice, 48 h post injection.

'A

t, min

R1,0-time

R1.30 min

R4, 30 min

Fig. 3: Size exclusion gel chromatography of liverextracts 30 min post injection of 67Cu-DOTA-R1-F(ab')2 and 67Cu-DOTA-R4-F(ab')2. Thearrow depicts the position of the intact F(ab')2

fragment.

27

We then investigated the role of the length of thelinker in providing the higher stability of the R1-linkedconjugates. The triglycine sequence was shortened toone glycine residue (R4) and the 67Cu-DOTA-R4-F(ab)2 was compared with the R1- and R2 conjugate.

All of the three 67Cu-F(ab')2 were 100% immunoreac-tive in vitro. Figure 2 shows tumour and blood levelsachieved with the conjugates, indicating that a short-ened linker leads to a decrease in tumour uptake.

The in vivo stability of the R4-linked F(ab')2 was com-pared with the R1-linked immunoconjugate by analys-ing liver extracts 30 min post injection (Figure 3).Results show a higher level of degradation productsaccumulating in the liver in the case of the compoundwith the shortened linker.

The initial rationale for using peptide linked chelateswas to provide cleavage sites in the immunoconjugateto achieve rapid clearance of radioactive metabolitesfrom the liver, thus improving tumour/liver ratios. Wethen found however, that the triglycine linker we inves-tigated improved tumour/tissue ratios by a differentmechanism. The results we obtained with the R1-R4peptide-linked DOTA chelates indicate that not in-creased metabolism in the liver but enhanced stabilityof the immunoconjugates towards proteolytic degrada-tion in the liver leads to higher tumour uptake.The effect may be due to the transiently higher bloodlevels that are achieved by the "stabilised" F(ab')2

conjugates, which prolong their availability for tumourbinding.

28

HIGH KIDNEY ACCUMULATION OF RADIOMETAL LABELLED SINGLE CHAIN AN-TIBODY FRAGMENTS: DOES CHEMISTRY MAKE A DIFFERENCE?

R. Waibel, A. Stichelberger, U. Bosshard, Ch. De Pasquale

Charge-modifications of 99mTc-tricarbonyl labeled single chain antibody fragments have been shown tohave an impact on unspecific kidney retention of metallic radionuclides. By modifying the charge of thechelating polyhistidine moiety we could further reduce unspecific renal accumulation as well as liver ac-cumulation.

INTRODUCTION

Single-chain antibody fragments (scFvs) have supe-rior pharmacokinetics and biodistribution characteris-tics compared with larger fragments when used forradioimaging of human tumours. However, persistentlocalisation of radioactivity was observed in nontargettissues such as kidney and liver after administration ofradiometal labelled proteins and peptides, which com-promises the diagnostic accuracy of the radiopharma-ceuticals.

The most likely mechanism for high kidney accumula-tion is that in an effort to salvage amino acids theproximal tubule cells sequester protein fragments andpolypeptides from the glomerular filtrate. It is knownthat glomerular permeability of macromolecules is afunction of their size and charge. Positively chargedmacromolecules cross the glomerular wall more read-ily than neutral molecules, and negatively chargedones are restricted from crossing.

Fig. 1: SDS-gel of kidney extracts of mice show highand persistent accumulation of small MW me-tabolites.

CHEMICAL MODIFICATIONS

We have developed a direct labelling method of re-combinant proteins. 99mTc-tricarbonyl coordinates tosequences of several His (polyHis-tag) in a fast andefficient way. Most recombinant proteins carry at theirC- or N-terminus a polyHis-tag for ease of purificationwith IMAC-chromatography. These 5 to 12 His at theC- or N- terminus are positively charged at neutral pH.Positively charged macromolecules and metabolitesthereof will be retained in the kidney.

By introducing carboxylate groups at the polyHis- tagand at the protein backbone, we investigated theircombined influence on charge modification. Using asan in vivo model a recombinant protein 6His-Thioredoxin, (MW12kDa.) kidney and liver accumula-tion was accessed.lodoacetate can react with a number of functionalgroups within proteins. The relative reactivity toward

protein functionalities is sulfhydryl > imidazoly > thio-ether> amine. At acidic pH carboxymethylhistidinyl ispredominantly formed.

Succinic anhydride is highly reactive toward nucleo-philes and is able to acetylate amine compounds oflysine side chains. On nucleophilic attack, the anhy-dride yields one carboxylic acid for every acetylatedproduct.

Histidine

Lysine Succinic Anhydride

Fig. 2: Chemical modifications to make anionic mole-cules.

RESULTS

These chemical modifications had a profound influ-ence on the isoelectric point of the molecule.

These modifications hade no influence on the labellingefficiency with 99mTc-tricarbonyl. There stability wasalso not impaired. In vivo, these modifications showeda reduced accumulation of radioactivity in the kidney.Succinylation of the lysine side chains and modifyingthe charge of the polyHis-tag had an additive effect onreducing renal retention of metabolites.

29

LIVER UPTAKE

Tis Ts Ti T " ' 5 "

Fig. 3: lEF-gel pH 3.5-8; Markers are T: 6His-thioredoxin, Ti: modified with iodoacetate, Ts:modified with succinic anhydride, Tis: modi-fied with both iodoacetate and succinic anhy-dride.

KIDNEY UPTAKE

200

§ 150

D)OOMOOHa

" 50H

155.35- ±8.98

i i

i i

i i

i i

-

-

68.94±14.0

- r - 49.9±1.0 28.16

±0.48

Xa:

?Xa: xor

24 h after i.v. injection

Fig. 5: Liver uptake of chemically modified 6His-thioredoxin constructs labelled with " m T c -tricarbonyl.

CONCLUSION

The result underscores the need to optimise the poly-His-tag chemistry and the overall charge of the re-combinant targeting-protein carrier by chemical ormolecular engineering means. Important areas forfurther study include characterization of radioactivemetabolites and the design of the polyHis-tag and itsadjunct amino acids which direct the disposition ofmetabolites reducing retention in normal organs.

24 h after i.v. injection

Fig. 4: Kidney uptake of chemically modified 6Histhioredoxin constructs labelled with " m T c -tricarbonyl.

As an added benefit, liver uptake was also drasticallyreduced.

30

CHEMICAL MODIFICATIONS ON MODEL SINGLE CHAIN FV FRAGMENTS FORIMPROVED IN VIVO BEHAVIOUR IN TUMOUR TARGETTING AND THERAPY

A. Stichelberger, R. Waibel

Improving the in vivo behaviour of a 99mTc(l)-labelled scFv by knowledge transfer from behaviour of tri-dentate labelled peptides using the classical approach.

INTRODUCTION

Labelling of IgG's and their corresponding fragments,such as Fab's, Fab'2 and scFv's with the radiometalnuclides 99mTc / 186/188Re(CO)3(H2O)3 is to becomean easy and powerful method for the production ofdiagnostic and therapeutic tools. Studies on peptideslabelled with this method revealed, that tridentatecoordination of these radionuclides, is favourable overbidentate coordination in in vivo behaviour (1). Com-pared to peptides, proteins require milder conditionsfor the usage of this labelling procedure. This meansambient temperature, aqueous solutions within moreor less physiological pH ranges and not more than20% v/v of organic solvent. We report on the testing ofseveral ligandsystems being attached to scFv frag-ments.

DISCUSSION

Coupling of ligands R1 and R2 (Fig 1.) to scFv's wasperformed according to classic DCC (Dicyclohexy-Icarbo-diimid) protocol (2). Furthermore, these ligandswere coupled via active ester formation using TFP(2,3,5,6-Tetralfuorophenol), NHS (N-Hydroxy-succinimide) and S-NHS (N-Hydroxysulfo-succinimide) at variable concentrations and pH ranges(3-4). And finally, the ligand systems R3 and R4 werereacted with squaric acid diethyl ester in order to becoupled to scFv's (5).

R t = -(CH2)4-COOH

R2 = -(CH2)l0-COOH

R3 = -(CH2)6-NH2

R4 = -(CH2CH2O-)2-(CH2)2-NH2

Fig. 1: Different ligand derivatives.

Radiolabelling of the scFv's Ri and R2 conjugates with99mTc(CO)3(H2O)3 showed no significant labellingeffect compared to control. HPLC analysis of R-, andR2 ligands prior to scFv conjugation showed morethan one product. Radiolabelling of these precursorsshowed that several species were present none ofwhich in excess and sufficient yield.Coupling of the ligands R3 and R4 to squaric aciddiethyl ester was not successful whereas coupling ofbenzylamine as model compound worked well.

CONCLUSIONIt is known that the e-nitrogen of the imidazole moietyof histidine is susceptible to many reagents and thuscan be easily modified. We assume that this modifica-tion of the e-nitrogen also takes place in that the acti-vated ligand system may undergo cyclisation and/orpolymerisation because the single proton signal in the1H-NMR of the 2 imidazole protons vanished uponactivation of the ligand system.

OUTLOOK

Considering the labile chemical character of theseligand systems and the versatile residues, which canbe attached to them, we consider to synthesize aligand system bearing a cadaverine residue thus mim-icking a lysine residue. Transglutaminase (TGase) isknown to catalyse an acyl-transfer reaction betweenthe y-car-boxamido group of glutamine and the e-amino group of lysine (Fig. 2).

Methyldiimidazole-cadaverine

H,N H H

Fig. 2: Reaction pathway of TGase.

Either the glutamine or lysine side chain can be mim-icked by small synthetic molecules enabling theTGase reaction to attach small labels selectively atlysine or glutamine side chains of a protein. By this,one should be able to conjugate the desired ligandsystem to scFv's under ambient conditions thus pro-viding a simple and directed method for conjugation ofsmall molecules.

REFERENCES

1. Schibli, R. et al. Influence of the Denticity of LigandSystems on the in Vitro and in Vivo Behaviour of99mTc(l)-Tricarbonly Complexes: A Hint for the FutureFunctionalization of Biomolecules, Bioconjugate Chem.4:345-351, 2000.

2. Visser, G.W.M. et al. Labeling of Mono-clonal Antibod-ies with Rhenium-186 Using the MAG3 Chelate forRadioimmunotherapy of Cancer: A Technical Protocol,J. Nucl. Med. 34:1953-1963, 1993.

3 Tietze, L.F. et al. Squaric Acid Diethyl Ester: A NewCoupling Reagent for the Formation of Drug Bioploy-mer Conjugates. Synthesis of Squaric Acid Ester Am-ides and Diaamides, Chem. Ber. 124, 1215-1221,1991.

31

INTEGRIN ANTAGONISTS FOR TUMOUR TARGETTING: CHARACTERISATIONOF RGD-THIOREDOXIN FUSION PROTEINS

/. Novak-Hofer, U. Bosshard, N. Riesen, K. Zimmermann, K. Ballmer-Hofer1

(1 Institute for Medical Radiobiology, PS I)

An integrin-binding RDG sequence was grafted into a bacterial thioredoxin (trx) protein. Trx-RGD is pro-duced in e.coli and conditions for obtaining monomeric and dimeric forms of trx-RGD were established.PC3 human prostate carcinoma cells are used for investigating in vitro anti-integrin activity and in vivotumour targeting ability of the trx-RGD forms.

av(33-integrin is involved in tumour growth and metas-tasis and is overexpressed in certain tumours. Weinvestigate the use of 67Cu-labelled fusion proteinscontaining RGD sequences to deliver cytotoxic betaparticle radiation to av(33-integrin overexpressingtumours. RGD sequences were engineered into theC-terminal part of a 14 kD thioredoxin (trx) moiety.Trx-RGD fusion proteins were chosen because oftheir ease of expression in e.coli, enhanced in vivostability and slower pharmacokinetics compared withsmaller RGD peptide antagonists. The purificationprocedure was modified to produce either a mono-meric form of trx-RGD or a disulfide-linked trx-RGDdimer. Cells were extracted in either a denaturingbuffer containing 6M guanidinium hydrochloride pH 8or in a "native" buffer consisting of 20 mM Na- phos-phate, 0.5 M NaCI pH 7.8. Proteins recovered in thedenaturing buffer were bound to Ni-NTA and stepwiserenaturated with PBS. Purification was by Ni-NTAaffinity chromatography and elution was with 0.5 Mimidazole in PBS. In the case of purification in "native"buffer 85% of the protein was recovered in themonomeric form, whereas denaturing conditionsyielded 40% of the monomeric form. Overall yieldswere similar. The dimeric form was isolated from themixture by FPLC size exclusion gel chromatography.If necessary the isolated proteins were concentratedby a further Ni-NTA affinity step. In order to character-ize the integrin-binding activity of the fusion proteins,av(33-integrin expressing PC3 human prostate carci-noma cells were chosen. Fig. 1 shows a Western blotof PC3 cell lysates which were immunopreciptatedand then probed with an av-integrin specific antibody.The 116 kD band corresponds to av-integrin, indicat-ing that PC3 cells can be used for targeting the RGD-proteins.

Anti-integrin activity of trx-RGD proteins is assayed byincubating PC3 cells with up to 200 |aM trx-RGD andthe adhesion of cells to vitronectin (VN)- and collagen-1 (C-l) coated tissue culture wells is measured. Inhibi-tion of cell binding by trx-RGD is compared with acontrol peptide (GRGDSP) of known potency. Figure2 shows that the control peptide inhibits attachementto VN but not to C-l. The trx peptide without RGDsequence shows no effect on adhesion of PC3 cells.The results indicate that in contrast to the GRGDSPsequence, the HIV-TAT derived RGD sequence in thetrx-RGD protein inhibits cell attachement to both VNand C-l. The dose response curves for trx-RGDmonomeric and dimeric forms are established at thepresent time.

Fig. 2:

4.0

c1.0

0.0n

PBS RGD trx-RGD

nPBS RGD trx-RGD

Inhibition of cell attachment to vitronectin (A)and collagen-l (B) coated wells. PBS: buffercontrol; RGD: 500 |aM GRGDSP; trx: 17 uJvitrx; trx-RGD:(mixture monomer/dimer): 17 uJvitrx-RGD.

Fig. 1: Western blot of PC3 lysates immunoprecipi-tated with av-integrin antibody. From left toright: reduced sample; non-reduced sample;molecular weight standards. The arrow de-picts the 116kD protein recognized by theav-integrin-specific antibody.

32

NEUROTENSIN(8-13) ANALOGUES INDUCE CA2+ MOBILISATION IN HUMANPROSTATE CARCINOMA PC-3 CELLS

E. Garcia Garayoa, P. Blauenstein, A. Blanc, R. Bugmann (PSI), D. Tourwe1, R. Muff2

(1 Vrije Universiteit Brussel, 2Balgrist Klinik, Zurich)

The effect of some NT(8-13) analogues on Ca2+ mobilisation was analysed in PC-3 cells which show bothNT1 and NT2 receptor subtypes. The affinity for NT1 receptors was previously evaluated in HT-29 cells(expressing only NT1R). Differences in the agonistic activity were found in keeping with the differences inthe affinity.

INTRODUCTION

Intracellular Ca2+ mobilisation has been described asthe major signal transduction pathway for neurotensin(NT) receptors. NT receptors are G-protein coupledreceptors whose activation results in the stimulation ofthe polyphosphoinositide pathway leading to free[Ca2+]| mobilisation. We work with a series of NT(8-

13) derivatives as potential candidates for the in vivotumour imaging and therapy. Three analogues (NT-VIM, NT-XI, NT-XII) showing the best affinities for NTreceptors as well as the highest in vivo tumour uptakehave been selected as promising candidates for clini-cal studies and one of them, NT-XI, will be included ina clinical trial. We evaluated the ability of these threeanalogues of stimulating Ca2+ mobilisation in the hu-man prostate carcinoma PC-3 cell line.

METHODS

PC-3 cells (1 x 106 cells/measurement) were loadedwith Fura2/AM. After extensive washing, the cellswere placed in a Lumin escence spectrometer for[Ca2+]j monitoring (excitation wavelengths: 340 and380 nm; emission wavelength: 510 nm). A dose-response analysis was performed for each analogueand its respective rhenium derivative.

RESULTS

All the derivatives were able to induce a transient risein [Ca2+]| showing a similar response pattern. In previ-ous binding assays we found that the NT(8-13) ana-logues showed a higher affinity for NT1 receptorswhen they were labelled (Table 1). The EC50 valuesfor [Ca2+]j increase were higher for NT-VIM and NT-XIthan for Re-NT-VMI and Re-NT-XI in good agreementwith the affinities (Table 1).

Table 1 : Affinity for human NT1R in HT-29 cells andeffect on Ca2+ mobilisation in PC-3 cells.

Peptide pair Affinity

(IC50, nM)

[Ca2+]i influx

(EC50, nM)

NT(8-13)

NT-VIII / Re-NT

NT-XI / Re~N1

NT-XII / Re-NT

0.9

21.1 / /

135.7/1

26.8 / 3

1.3

31.3 IV

870.5 / £

18.3/1

NT-XI induced a much lower [Ca2+]j influx than NT-VIIIand NT-XII (100 nM, each). The rhenium derivativesshowed a comparable curve. After peptide addition arapid increase in [Ca2+]| was found. Values returned tobaseline after 2-3 min. Fig. 1 illustrates the responseobtained for the three pair of analogues.

700

600

500

400

300

200

100

,£, 700

£ 600

o• ^ 500

(0 400

~ 300

O 200

E 100

O 700

600

500

400

300

200

100

0

NT-VIII

_ Re-NT-VIII

1 \

—JNT-XI

• Re-NT-XI

^ ^ •

. . . . . .

. NT-XII

i Re-NT-XII

V\\

mm<m * ~ ~

D 100 200 300

Time (s)

Fig. 1: Time course effect of the analogues and theRe-derivatives (100 nM, each) on [Ca2+]imobilisation in PC-3 cells.

CONCLUSION

[Ca2+]| mobilisation was triggered by the three ana-logues and their rhenium derivatives with some differ-ences in signalling potency which could be related tothe different affinity for NT1 receptors. This could bedue to a distinct conformation of non-labelled ana-logues compared to labelled ones. A consequencewould be a lower probability of side effect incidenceafter treatment with the radiolabelled analogues whencarrier peptide is present.

33

NEUROTENSIN DERIVATIVES:QUALITY CONTROL IN VIEW OF HUMAN APPLICATION

P. Blauenstein, F. Feurer, E. Garcia Garayoa, R. Bugmann, A. Blanc

Neurotensin derivatives are potential vehicles to transport "mTc and 188Re into tumour cells for diagnosticor therapeutic purposes. The search for stabilised neurotensin analogues yielded two analogues whichwere analysed in view of the clinical test, and finally NT-XI was selected as the most promising candidate.

INTRODUCTION

Neurotensin belongs to the group of neuropeptidesand binds to receptors which are present on cells ofhealthy tissues and are over expressed in differenttumours, like the ductal pancreatic carcinoma. Thistype tumour is present in 90 % of the patients with atumour of the pancreas and in about 75 % of thecases neurotensin receptors were found. Pancreastumours are very difficult to treat.

The stability of the NT analogues was found to becrucial and thus, several stabilised NT derivativeswere synthesised by the group of Prof Tourwe (Brus-sels). (In this text "Tc-labelled" means use of 99mTc.)

Table 1: Two out of 15 Neurotensin analogues la-belled with Tc(CO)3 which show highest tu-mour uptake. The natural NT7-13 is shownfor comparison.

NT7-13

NT-VIM

NT-XI

Pro-Arg-Arg-Pro-Tyr-lle-Leu

(NaHis)Ac-(N-CH3)-Arg-Lys-Pro-Tyr-Tle-Leu

(NaHis)Ac-Lys-(YCH2NH)-Arg-Pro-Tyr-Tle-Leu

The histidinyl acetate (NaHis)Ac is a tridentate ligandwhich forms an uncharged and stable complex withthe Tc-tricarbonyl moiety. This complex is only slightlypolar and does not form hydrogen bridges. Thus thecomplex is biologically inert, i.e. it does not influencethe biological properties of the labelled biomoleculeand it is thus an ideal candidate for in vivo application.However, HPLC analysis showed that two differentspecies are formed (Fig. 1). An important point was toshow that there is no pharmacological difference be-tween the two forms. Another goal of this year was theintroduction of a biochemical test which allows a quickanalysis of the finished product.

RESULTS

The binding studies were done according to Lindmoand Scatchard, respectively, and the comparison ofthem (Table 2) shows no difference in the results. Theadvantage of the test according to Lindmo is the lowlevel of activity which is needed, and thus it can beperformed with a small aliquot of a patient dose. Thetest according to Scatchard would need the wholedose.

i_

Trace 1: Reaction mixture15minat75°C,trace 2: re-injected fractioncontaining P1,trace 3: re-injected fractioncontaining P2

Same solutions heatedadditionally for 1 h at 75 °C

Fig. 1: HPLC analysis showing the raw product andboth species of Tc-labelled NT-VIM.

Table 2: Binding of Tc-labelled NT analogues usingdifferent methods of analysis.

NT-VIM

NT-VIM

NT-XI

NT-XI

-log Kd (Scatchard)

-logKd (Lindmo)

-log Kd (Scatchard)

-logKd (Lindmo)

P1

8.7 + 0.6

9.7 + 0.5-

9.4 + 0.4

P2

9.0 + 0.4

9.5 + 0.5

9.3 + 0.6

9.6 + 0.3

CONCLUSION

The results listed in Table 2 show clearly that the mostimportant parameter, namely the receptor affinity ofboth species is the same. Moreover it was noticed thatheating of the separated species at pH 7 (the freepeptide is also absent) did not destroy the complex(this is not the case at high and low pH values).

The test production showed that a sterile product wasobtained and the toxicity study done at the RCC Ltdyielded a value of > 50 ng/kg body weight for NT-XI.All the data indicate that the application to patients willbe safe.

34

COLD RHENIUM COMPLEXES OF NEUROTENSIN ANALOGUES

A.Blanc, P. Blauenstein, R. Schibli, E. Garcia Garayoa, D. Tourwe1 (1Vrije Universiteit Brussel)

The same species of the neurotensin complexes were found after the synthesis with "mTc and cold Rhe-nium tricarbonyl. In contrast to tracer amounts obtained with "mTc it was possible to analyse the Re com-pounds with NMR and MS.

INTRODUCTION

The Technetium tricarbonyl moiety has been shown tohave excellent properties for the labelling of neuro-tensin analogues and the Nahistidinyl acetate to be afavourable ligand. We observed that two differentspecies were formed with Tc-99m (see Neurotensin inview of human application, this report, next page) andtherefore the same complexes were synthesised withnatural Rhenium to enable spectroscopic analysis ofthe products. The first important finding was the simi-lar HPLC trace indicating that the same complexeswere formed. This was confirmed by the similar valuesobtained with Tc and Re for the binding of the Re-NT-XI on HT29 cells and human tumour slices (Prof. J.C.Reubi, University Bern).

METHODS AND RESULTS

Synthesis: 5 (a.mol of the peptide were mixed with aslight excess of 5.25 pmol of (TEA)2Re(CO)3Br3, thepH was adjusted to 7 by addition of NaOH and waterwas added to a total volume of 1 ml. The solution washeated for at least 7 h (sometimes over night) and theproduct analysed with a reversed phase HPLC. TheHPLC runs had to be repeated several times in orderto get sufficient material for further analysis.

The Mass-Spectrum (MALDI-TOF) of both fractionsshowed no difference and the highest peaks werefound at the expected molecular mass of 1240 and1238 Dalton (reflecting the ratio of the natural isotopesof Re). Smaller peaks at 1239 and 1241 Dalton areinterpreted as protonated forms.

The NMR spectrum was recorded using the newnano-probe. Most signals show small differences be-tween the species, and, as it is expected, rather largedifferences of the His-H. If the assignment (which isnot yet unequivocal) is really correct, the signal of theacetate-protons does not change.

Table 1: NMR data: chemical shift (8) of the Histidinesignals of the different species.

Imidazole

His Ha

HisHp

free NT-XI

6.97, 8.5

3.90

3.28

Re-NT-XIFract. 1

6.95, 8.01

3.92

3.18,3.62

Re-NT-XIFract. 2

6.91,8.01

4.05

3.23, 3.92

1 1 2 1Number of H

1 1 1 1Number of H

Fig. 1: Selected part of the NMR spectra of theRe(CO)3-NTXI complexes (left: fraction 1,right fraction 2).

The NMR Spectrum can be divided into 4 differentregions: In addition to the data listed in Table 1 wefound the aromatic protons of Tyr at 6.6 and 7.0 ppm,the alpha-CH of the amino acids between 4 and 4.5ppm and the protons of the methyl- and methylen-groups of the amino acids in the region between 0.7and 3.0 ppm. Two H at about 3.45 and 3.6 ppm aremost probably the H8 of Pro.

CONCLUSION

The analysis of the corresponding Tc-complexesshowed that initially the first peak is formed and afterabout 2 h of heating an equilibrium is observed. Thisleads to the assumption that the first peak representsan only bidentately bound ligand with either a freeamine or carboxylate and 1 H2O still bound to Re. It isclear that such a complex is more hydrophilic becauseof the longer distance between the charged groupsand the bound water molecule (several hydrogenbridges possible). The second peak is thus the triden-tately bound ligand. The NMR data fit rather well withthis assumption. There are two other possible explica-tions: The inversion of the chirality of the a-carbon ofHis or the inversion at the a-amine of His. The MSdata support this assumption, however, they are alsoin agreement with the first assumption, if we assumethat the heating of the molecule with the laser leads tothe elimination of the water bound to Re.

35

PHARMACOLOGICAL EVALUATION OF A 99MTC(I)-LABELLED BOMBESINANALOGUE FOR GRP RECEPTOR-TARGETTED SCINTIGRAPHY

R. La Bella, E. Garcia Garayoa, M. Langer1, P. Blauenstein, A. Blanc, R. Bugmann,A.G. Beck-Sickinger1 (1 University of Leipzig, Germany)

A 99mTc(l)-labelled bombesin analogue was characterised in vitro and in vivo for use in gammascintigraphy.

INTRODUCTION

An overexpression of neuropeptide receptors is ob-served in many cancers what makes them an attrac-tive target for tumour imaging and therapy. Bombesin,a 14 amino acid peptide, exhibits a high affinity for thegastrin releasing peptide receptor which is stronglyoverexpressed by a variety of tumours (1].

METHODS AND RESULTS

The radiolabelled bombesin analogue (Fig. 1) wasobtained by a prelabelling approach and was charac-terized in vitro and in vivo.

NH,

- | - c oCO

VNH S

Fig. 1: Structure of [M(l)-PADA-AVA]bom-besin(7-14) (M = 99mTc, Re).

Binding studies revealed a high affinity of the rhenium-labelled homologue (IC50 = 2.2 nM) for GRP receptorsexpressed on the PC-3 humane prostate cancer cell-line. [99mTc(l)-PADA-AVA]bombesin(7-14) showed arapid and specific internalisation in PC-3 cells. At37°C more than 70% was internalised within the first15 min and remained constant up to 2 hr (Fig. 2).

100

& 75

« § 50

1£ 25

s? o 1MH5 15 30

time [min]

120

Fig. 2: Internalised [125l-Tyr4]bombesin (white bars,control) and [99mTc(l)-PADA-AVA]bombesin(7-14) (black bars) by PC-3 prostate tumourcells incubated at 37°C, after a 2 h preincuba-tion at 4°C.

Biodistribution studies performed in PC-3 tumour-bearing CD-1 immunosuppressed mice showed rapidclearance of the radioactivity from the blood compart-ment. The highest uptake was found in the pancreas.High uptake was also observed in the kidneys and theliver. Tumour uptake was rather low, however, thereceptor blocking study confirmed the specificity of thebioconjugate toward GRP receptors (Fig. 3).

Pancreas Tumor Intestines Blood Liver Kidneys

Fig. 3: [99mTc(l)-PADA-AVA]bombesin(7-14) wasinjected i.v. and the mice were sacrificed 1.5hr p.i. (white bars). For the blockade experi-ment 300 |ug bombesin/mouse were co-injected with the radioligand and the micewere sacrificed 1.5 hr p.i. (black bars).

A significant reduction of the uptake in the pancreasand the tumour, and partly in the intestines, was ob-served with the additionally injected 300 jig of bombe-sin. No reduction was obtained in all other organs(data not shown).

CONCLUSION

The synthesized 99mTc(l)-labelled bombesin analogueshows a high affinity for GRP receptors, is quicklyinternalised by GRP receptor expressing cells and hasin vivo a highly specific uptake into GRP receptor-positive tissues. Therefore, [99mTc(l)-PADA-AVA]bombesin (7-14) might be used successfully forscintigraphy of GRP receptor-expressing tumours.

REFERENCE

1. Breeman, W.A.P. et al., Evaluation of radiolabeledbombesin analogs for receptor-targeted scintigraphyand radiotherapy. Int. J. Cancer 81, 658-665, 1999.

36

IMPROVED PREPARATION OF [M(OH2)3(CO)3]+(M = 9 9 M J C , 1 8 8 R E ) AND ITSAPPLICATION FOR RADIOLABELLING OF VARIOUS

FUNCTIONALISED BIOMOLECULES

R. Schibli, C. Muller, B. Sigrist, M. Netter, J. Stahel, R. Schwarzbach, R. Alberto1,K. Ornter1, I. Zolle2

(1 University of Zurich, 2University of Vienna)

Improved preparations of [M(OH2)3(CO)3l+ (M="mTc, 188Re) have been elaborated in order to achieve

higher specific activities, lower toxicity during synthesis and simplify handling. The last two points could beaddressed by substituting NaBH4 and gaseous CO with K2H3BCO2. Biotin, estradiol, testosterone and

metomidate have been functionalised in order to be able to label them with [M(OH2)3(CO)3]+.

INTRODUCTION

The organometallic technetium and rhenium tricar-bonyl precursor [M(OH2)3(CO)3]

+ (M = 99mTc, 188Re)for labelling of biomolecules has attracted the atten-tion of several research groups(1). Although the cur-rent preparation of the precursor is easy and straight-forward, the fact, that gaseous CO is needed duringpreparation, may hamper its use in a hospital envi-ronment. A stable, water-soluble, non-toxic source ofCO had to be found preferentially with reducing prop-erties.

RESULTS AND DISCUSSION

Together with the group of Prof. Alberto at the Uni-versity of Zurich we have found that potassium boroncarbonates, K2H3BCO2, fulfil all theses prerequisitesideally. K2H3BCO2 releases CO under aqueous con-ditions upon heating (2) (Fig.1). At the same time, ithas the potential to reduce 99mTcO4~ form oxidation

state +VII to +l. Thus, the source of CO and reductionare conveniently hosted in the same molecule.

2-H+ H+

^ H3BCO2H" ^ H3BC(OH)2

B(OH)3 + 3 H2 + CO H3BCO + 3 H2O

Fig. 1: Hydrolytic pathway of K2H3BCO2.

Further studies about the basic reaction mechanismof reduction and carbonylation will be performed inthe future. Initial toxicity studies performed at Mal-linckrodt Inc. (Petten, Holland) have shown that bo-ron carbonates in the applied quantities (5-10 mg)are readily tolerated in rats with no effects when in-jected intravenously. The elegant formulation for[99mTc(OH2)3(CO)3]+ enables now the routine and

harmless preparation of this organometallic fragmentwith high specific activities even in non-chemicallaboratories. Other groups will be able to profit fromthis facile kit preparation in order to readily apply thenovel labelling technique for their purposes.

We also improved the preparation of the Re-188 pre-cursor [188Re(OH2)3(CO)3]+ potentially useful for ther-apy applying the same labelling methodology as with[99mTc(OH2)3(CO)3]+. The new synthesis usingBH3»NH3 enables the preparation of the versatile pre-cursor in better yield and higher specific activities thanthe previous one with NaBH4 (1).

In order to translate the high specific activities of theprecursors [M(OH2)3(CO)3]

+ to the labelling of bio-molecules, we have developed a new versatile andvery effective, tridentate ligand system, enabling theconnection to biomolecules via an amine or a carbox-ylic acid functionality using standard peptide syntheticstrategies. Introducing of this ligand system to an e.g.tumour avid peptide via automated solid phase tech-niques should, thus, be possible. The synthesis of theligand system is simple and the yields generally >85 %(Fig. 2).

H2N-(CH2)n-NHR IN*¥ H/ • ) -N- (CH 2 ) n - lNH2

H2N-(CH2)n-COOR(CH2)n—COOH

• B

Fig. 2: Synthesis of bifunctional, tridentate ligand sys-tems for labelling with [M(OH2)3(CO)3]+.

The X-ray structure (Fig. 3) of a corresponding modelcomplex [Re(LB)(CO)3]Br (R=Me; n=1) has shown, thatthe ligand coordinates indeed tridentately to theM(CO)3 core improving stability as well as in vivoclearance of labelled molecules.

Several novel biomolecules have been exploited in ourgroup for their potential to be functionalised and la-belled with [M(OH2)3(CO)3 ]+ (under retention of bioaf-

finity) during diploma theses and short time scientificmission of the European COST action (B12).

37

c i o C 1 3

C 9

O l

O 3

Fig. 3: ORTEP of complex cation [Re(LB)(CO)3]+(R=Me; n=1).

Namely, biotin (applied e.g. in therapeutic tumour"pre-targeting" approaches), estradiol and testoster-one (diagnosis and therapy of mamma carcinoma orprostate carcinoma) and metomidate (3) (diagnosis ofadrenal gland carcinoma) (Fig. 4).

Fig. 4: Native structures of biomolecules: 1 Biotin, 2Estradiol, 3 Testosterone, 4 Metomidate.

The functionalisation and labelling strategies areoutlined in figure 5.

Fig. 5: Structure of functionalised and labelled bio-molecules.

In case of biotin the above mentioned Ligands LA

(R=H, n=3, 10) were connected via the carboxylicacid functionality and the corresponding bioconjugatesubsequently labelled with [M(OH2)3(CO)3]

+. Specific

activities of 1 TBq/nmol (99mTc) and 400 GBq/|imol(188Re) could be reached, which is significantly

higher (10-100 times) than with other labelling proce-dures. This is mainly due to the excellent labelling andcoordination properties of [M(OH2)3(CO)3]

+ and theapplied chelating systems. The conjugate was stablefor more than 24 h in PBS buffer at 37°C (<95%, de-termined by HPLC) with only minor reformation ofMO4". More important, the labelled biotin derivativerevealed excellent bioaffinity (<90%) comparable tonative biotin. This was verified by standard biotin-streptavidin protocols (Dynabeads®).

The two steroids estradiol and testosterone were suc-cessfully functionalised at position 17a known to onlylittle interfere with the hormone receptor. The bidentatemetal chelate consists of a pyridine ring an imine func-tionality (Schiff base). Log D values (octanol/water) forthe two 99mTc-tricarbonyl labelled compounds werearound 2 and therefore, significantly smaller than forthe native hormones (4.7 and 5.6 respectively) pre-sumably due to the hydrophilic character of the metalcentre. Furthermore, the labelled compounds revealedan unexpected low stability in different buffered media.One reason for the instability is the tendency of Schiffbases to hydrolyse. When incubated in human or bo-vine plasma, fast aggregation of the compounds to theplasma protein was observed, which prohibited furtherin vitro affinity tests. Thus, for future derivatisation ofsteroids more lipophilic, more stable and tridentateligands have to be applied.

During a short time scientific visit the potential tracerfor carcinoma of the adrenal gland, metomidate (3),was modified to be able to label it with99mTc(OH2)3(CO)3]+. From the literature it was pro-posed that for specific receptor binding the ester, imi-dazole and an aromatic moiety is critical. Thus, thelead structure was essentially retained. The aromaticmoiety was replaced by a bipyridine system, whichenables radioactive labelling with the 99mTc-tricarbonyl. Preliminary in vitro binding experimentshave shown significant uptake of the 99mTc-compoundin the corona of the adrenal gland.

CONCLUSIONWe have developed a kit preparation of the or-ganometallic precursor 99mTc(OH2)3(CO)3]+ ready forcommercialisation and have improved the synthesis ofits therapeutic twin [188Re(OH2)3(CO)3]+ in order toobtain higher specific activities. The possibility, limita-tions and future perspectives of labelling biomoleculeswith these precursors could again be exemplified withvarious derivatised biomolecules.

REFERENCES

1. Alberto, R. et al., J. Am. Chem. Soc, 120: 7987-7988,1998.

2. Malone L.J. et al., Inorg. Chem., 6:2260-2262, 1967.

3. Bergstrom M. et al., J. Nuc. Med., 39:982-989, 1997.

38

FUNCTIONALISATION OF GLUCOSE AT POSITION C-2 AND C-6FOR THE LABELLING WITH 99mTc-TRICARBONYL

C. Dumas, R. Schibli, J. Petrig, J. Stahel, D. Kunz, S. Schweri

Cancer cells growth is heavily dependent on increased glucose metabolism. Thus, a glucose derivativethat is transported by GLUT 1 could be useful for the imaging of tumours. Glucoses substituted at positionC-2 and C-6 have been synthesized with the intention to develop of radioactive labelled drugs such as thePET (Positron Emission Tomography) tracer 2-[18F]-2-deoxy-glucose (FDG) for diagnostic purposes. Thestrategy to substitute the glucose at these two positions will be described.

INTRODUCTION

The biological activity of glucose is defined by twofacts: its ability to enter the cells (GLUT transporters)and its metabolism in the cell (hexokinase).

In order to optimise the substitution work on the glu-cose we decided to study the interaction of the glu-cose with its transporter, when it enter the cell, andwith hexokinase, once it is in the cell. As the structureof the GLUT1 is not determined, the modelling of theGLUT1/glucose complex is not yet possible. But, themodelling of the hexokinase/glucose complex (Fig. 1)suggests that glucose can be functionalised at posi-tion C-2 or C-6 and still enter the hexokinase activesite.

Fig. 1: Labelled glucose functionalised at C-2 by aheptene spacer and chelating moiety in theactive site of hexokinase and the availablefree space (1).

Earlier attempts to label glucose directly with thetransition metal " m T c failed due to low in vivo stabil-ity of the compounds and due to the direct coordina-tion of the metal centre to the sugar hydroxy func-tionalities important for substrate recognition (2).

To maintain bio affinity it is necessary to bifunctional-ise glucose with an appropriate metal chelating sys-tem and to separate the sugar moiety and the metalchelating system. We decided to functionalise glu-cose at position C-2 and C-6 with an iminodiaceticacid moiety separated by spacers of different chainlength (Fig. 2) (3).

H2Ck,OH2

.L*co

chelate > _Linkern=2,3...

Glucose

RESULTS AND DISCUSSION

Substitution at C-6

To start with the substitution of glucose, we have de-cided to set up the reactions at position C-6 with alinker of three carbons.

Substitution at position C-6 (Scheme 1) was initiatedby protecting all the hydroxyl functions except the oneat C-6 to form pyranose 1. The addition of A/-protectedaminobromide at C-6 affords the desired glucose ana-logue 2, after acidic treatment to remove the Boc pro-tection. In the later step the amine had be functional-ised by action of methylbromoacetate to form theligand. Finally the derivative 3 was obtained by saponi-fication of the esters and elimination of the acetals.

a)

D-Glucose

N XOoH

Fig. 2: Scheme of novel 99mTctricarbonyl labelledand schematic functionalised glucose.

:O?H

Scheme 1: a) i) p-formaldehyde, AcOH, ii) MeONa,MeOH, 59 %, b) i) MeONa, MeOH, ii)Br(CH2)3NHBoc, NaH, DMF, Hi)CF3CO2H, CH2CI2, 30 %, c) i)BrCH2CO2CH3, NEt3, THF, ii) NaOH,H2O, 78 %.

The work at position C-6 was facilitated by the rapidprotection and deprotection of the alkanol functions inone step. Consequently, the sugar derivative substi-tuted at position C-6 was straightforwardly obtained ina six steps synthesis.

Substitution at C-2

After a broad investigation of the previous work onposition C-2, which is known to be difficult, the synthe-sis of these derivatives was started using three differ-ent strategies. The first one consists in the alkylation ofthe sugar when it is in its pyranose form (Scheme 2).The epoxidation of commercially available tri-

39

benzylglucal took place in presence of m-chloroperbenzoic acid and the methanol directlyopened the epoxide to form the tetraprotected sugar4. Unfortunately all the attempts to alkylate at positionC-2 the free alcohol of sugar 4 failed.

^.Bn

Bn

Tribenzylglucal 4 M

Scheme 2: a) m-CPBA, MeOH, 64%.

A second strategy was used to substitute the sugar atC-2 (Scheme 3). We decided to substitute the C-2 byopening of the 2,3 epoxide of 1,6-anhydroglucose.The commercially available anhydroglucose wastosylated at position C-2 and C-4 to afford 5, and theleaving group, O-tosyl at position C-4, was eliminatedto form the epoxide 6. This epoxide was then openedby addition of benzyl alcohol and the epoxide 7 wasclosed by basic treatment. However, we neverachieved to open the epoxide of 7.

OH OH

Anhydro-D-glucose

Ts'

c)

Scheme 3: a) CITs; Pyridine, 100 %, b) NaOMe,NaOH, 46%, c) i) BnOH, BF3,Et20,Toluol, ii) NaOMe, NaOH, 40 %.

Our last strategy to substitute carbon C-2 was toalkylate the alcohol when the sugar was in its fu-ranose form (Scheme 4). The commercially availableD-glucose was first transformed in its furanoseequivalent and then acetals groups had protected thealcohols at C-1, C-2 and C-5, C-6. Selective depro-tection of the acetal at position C-5, C-6, and thenalkylation of the three free alcohols by benzyl groupsafford, after deprotection of the acetal at position C-1,C-2 the sugar tetraprotected 9. The addition of thelinker and the formation of the ligand proceeded asdescribed for the synthesis of sugar 4 to afford 10.The acetylated sugar 11 was then obtained by hy-drogenolisis of the benzyl groups and transformationof the furanose in pyranose. The final product will beobtained by saponification of the esters.

o a) Bn°Bntf

D-Glucose

OAc

c)

MeO2C

Scheme 4: a) i) Acetone, H+,ii) Resin, MeOH, H2O, 62 %,iii) BnCI, NaH, iv) HCI, MeOH, 57 %,

b) i) Br(CH2)3NHBoc, DMF,ii) CF3CO2H, CH2CI2, 51 %,iii) BrCH2CO2Me, NEt3, THF, 56 %,

c) i) H2, Pd(OH)2, MeOH, ii) H2SO4,Ac2O, 43 %.

After three different strategies we finally found the op-timal one to substitute the glucose at position C-2.Four steps are necessary to protect the alcohols atposition C-1, C-3, C-5 and C-6 and leave the one atposition C-2 free. But it appears to be the safer and themore efficient tactic. Now, that all the protocols are wellestablished we will be able to functionalised this posi-tion with other linker and different ligands.

CONCLUSION

We succeeded in the functionalisation at position C-6and C-2 of glucose and hopefully will soon show theircompatibility with transporter GLUT-1 and hexokinase.This work will go on with the substitution of the sugarby longer linkers, as a height-carbon chain (to fit to thechannel of hexokinase showed in Figure 1), and bychanging the chelating system.

REFERENCES

1. Schibli, R. et al., unpublished results.

2. a) Caner, B.E. et al., Nucl.-Med., 1991, 30, 132-136; b)V.R. Risch et al., Radiology, 1977, 124, 837-838; c)T.S. Wang et al., Nucl.-Med., 16, 89-90, 1977.

3. Petrig, J. et al., 20th International Carbohydrate Sym-posium (Hamburg, Germany).

40

FUNCTIONALISATION OF GLUCOSE AT POSITION C-3 FOR THE LABELLINGWITH 99mTC(l)-TRICARBONYL

J. Petrig, C. Dumas, Ft. Schibli, J. Stahel, D. Kunz

r18rThe aim of this project is to replace the expensive PET tracer [ F]-2-Fluorodeoxyglucose by a inexpen-sive 99mTc labelled analogue for tumour imaging. For this purpose, we synthesised a series of glucosederivatives at position C3 with different ligand systems. We successfully labelled these derivatives withthe organometallic precursor fac-["Tc(OH2)3(00)^.

INTRODUCTION

It is assumed that the high uptake of FDG in tumourcells is due to the strong overexpression of theGLUT1 glucose tranporter in many carcinomas.(1, 2)Recognition sites for the GLUT1 are the positions 1and 3 of the glucose(3). Whereby position 3 seems tobe less important, because 3-Methylglucose has asimilar IC50-value as glucose(4). Since 3-0-

Methylglucose has a similar affinity to the GLUT1 andhigh stability even in rat liver(5), it can serve as leadstructure for designing a glucose derivative for thelabelling with Tc(l)-tricarbonyl.

RESULTS AND DISCUSSION

Presumably the charge of the complex is crucial forthe biological activity of the complexes. Thereforeglucose compounds, derivatised at position 3, withdifferent chelating systems, yielding neutral, positiveand negative charged complexes, were synthesized ina six step synthesis from the commercial Diacetone-D-glucose 1.

HOO<

H O O C — ' \ k .

Scheme 1 : Synthesis of the ligands.

The obtained ligands 2, 3, 4 and 5 were labelled at aligand concetration of 10"4 M at 70° for 30 minuteswith fac-["Tc(OH2)3(CO)3]+, which gave the corres-ponding Tc-complexes 2a, 3a, 4a and 5a.

The labelling afforded in all cases single products inhigh yield (87-97). The radioactive complexes showedhigh stabilities, between 85 and 99 % after 24 hours,in PBS and in human serum. From 1D proton NMR

experiments it can be concluded that the linker is longenough that the sugar part does not interfere with thechelator. For further characterizations the correspond-ing Re-complexes 2b, 3b, 4b, and 5b were synthe-sized and characterized with mass and NMR spec-troscopy. They had similar retention times on theHPLC as the 99mTc-compounds which evidences that

2 ^/3

M = 99mTc, Re [M(CO)3(H2O)3]+

4a, 4bthey have the same structure.

5a, 5b

Scheme 2: Labelling with 99mTc(l)-tricarbonyl andRe(l)-tricarbonyl.

CONCLUSION

We synthesised a series of novel glucose derivativesfor the labelling with low valent organometallic techne-tium precursors. The present strategy allows the sta-ble and specific labelling of glucose with " m T c in highyield. The complexes show also a very high stabilitywhich is important for further investigations as affinitytests on the GLUT1.

REFERENCES

1. Smith, T.A.D. Brit. J. Biomed. Sc. 56:285-292, 1999.

2. Brown, R.S. et al., J. Nuc. Med. 37:1042-47, 1996.

3. Barnett, J.E.G. et al., Biochem. J. 145:417-429, 1975.

Halmos, T. et al., Eur. J. Pharmacol., 318: 477-484,1996.

5. Jay, T. M. et al. J. Neurochem., 55:989-1000, 1990.

41

IN-OUT-ISOMERISM OF NS3-CAGE COMPOUNDS FOR THEIN VIVO STABILISATION OF AG-111

J.M. Baumeister, R. Alberto1, Th.A. Kaden2, K. Ortner1

(1 University of Zurich,2 University of Basel)

Synthesis of NS3-cage compounds yielded in-isomers of the desired ligands. Synthetic strategies leading

to out-isomers as potential111Ag-ligands for radioimmunotherapy are pursued.

Good decay properties make 111Ag one of the mostinteresting radionuclides for applications in radioim-munotherapy. Yet the metal exhibits in its +l (d10)oxidation state a very flexible coordination behaviour.This is mirrored in labile complexes which undergorapid transmetallation with competing ligands.

Especially in biological systems potential therapeuticagents marked with 111Ag will face a wide variety ofe.g. thiolate groups which can coordinate 111Ag. Theradioactive silver ion will then be bound unspecificallyand will be distributed all over the body. Since theconcentration of competing ligands exceeds originaldonors by orders of magnitude, an appropriate ligandhas to form both thermodynamically stable and kineti-cally inert complexes.

Suitable complexes will then be bound to tumorspeci-fic biomolecules (e.g. antibodies) via specific linkermolecules.

We improved the synthesis of one essential buildingblock, 3-(2-Chloroethyl)-1,5-dichloropentane, asshown in Fig. 1: The new synthesis consists ofstraightforward, high-yield reaction steps.

\ — C O

-H2O; -C0 2

COOEt

COOEt

I H2; Pd/C

N — /—COOEt

*—COOEt

OH" > ^ HO

JJX.T SOCI2

JJX.J LiAIH,

1.OH

2. H +

;OOH f/ H + ; EtOH

Fig. 1: Synthetic approach to 3-(2-Chloroethyl)-1,5-dichloropentane.

The next step was to perform a high dilution reactionto synthesise the desired cage compound bearing theNS3-complexating moiety. We reacted Triethanthiola-mine with 3-(2-Chloroethyl)-1,5-dichloropentane inDMF using Cs2CO3 as a template/base. Workup afterone week yielded exclusively in-isomer in 5% yield(Fig.2).

Cs2CO,

i T \ DMF-

Fig. 2: High-dilution reaction to NS3-cage compound.

1H-NMR-studies of this ligand showed extraordinarydownfield-shift of the hydrogen-atom bound to thebridgehead-carbon (Fig. 3), resulting from interactionwith the bridgehead-nitrogen. A crystal structureanalysis of its silver complex confirmed these expec-tations (Fig. 4).

\

Fig. 3: Significant downfield-shift of the hydrogen-atom (arrow) bound to the bridgehead carbon-atom.

Ag(l)

Fig. 4: In-isomer showing surface coordination ofAg(l). The NS3-coordination site is blockeddue to geometric constraints and the bridge-head-hydrogen. Ag(l) is four-coordinated, twoadditional coordinating NS3-cages are omittedfor clarity. The tosylate-counterion occupiesthe fourth coordination site.

REFERENCES1. Alder, R.W.

1996.et al., Chem. Rev., 2097-2111,

42

SYNTHESIS, RADIOLABELLING AND CHARACTERIZATION OF POTENTIAL PETTRACERS FOR THE GLUTAMATERGIC SYSTEM

M. Honer, M. Kokic, L Kessler, F. Gaspahni1, M. Grauerf, S.M. Ametamey(1Novartis Pharma AG, Basle, 2Boehringer Ingelheim GmbH, Ingelheim, Germany)

The methodology to perform pharmacological in vitro characterizations of potential PET tracers was estab-lished using [11C]Methyl-BIII277CL, a putative ligand for the glutamatergic N-methyl-D-aspartate (NMDA)receptor. M-MPEP has been radiolabelled and biologically evaluated. PET studies in 5 healthy volunteersof [18F]memantine have been completed.

SYNTHESIS AND IN VITRO CHARACTERISATIONOF [11C]METHYL-BIII277CL

INTRODUCTION

A PET tracer has to meet several requirements beforeit can be used in a clinical setting. Apart from a suffi-cient penetration of the blood-brain barrier and a lowmetabolism and toxicity, the most basic prerequisitesfor a PET tracer are a high affinity and selectivity forits target. Furthermore, the pharmacokinetic proper-ties should guarantee that a binding equilibrium canbe quickly reached and that the dissociation of theligand-target-complex is not too slow.

Affinity, selectivity and kinetic properties of a potentialPET tracer can be determined in vitro by radioligandbinding assays. In displacement assays, the potentialPET tracer is used in an unlabelled form and testedagainst various tritiated ligands that specifically inter-act with distinct targets. For saturation assays, thePET tracer is labelled with a PET nuclide and em-ployed in kinetic studies and Scatchard analyses.However, radioligand binding studies with short-livedPET nuclides (e.g. 11C with t1/2 = 20.4 min) are physi-cally limited since the duration of the experimentsshould not exceed five half-lives to maintain adequatesignal-to-noise ratios. We established the methodol-ogy to perform these binding studies and used[11C]Methyl-BIII277CL, a potential PET tracer for theglutamatergic N-methyl-D-aspartate (NMDA) receptor,for an initial test.

RADIOSYNTHESIS

Methyl-BIII277CL is a derivative of the benzomorphanBIII277CL, a specific ligand for the PCP binding site ofthe NMDA receptor with a negligible cross-reactivityfor the G1 -binding site. Methyl-BIII277CL is radio-labelled with carbon-11 at the O-position of the des-methyl precursor (BIII277CL) by heating for 10 min at120°C with 11C-methyl iodide in DMSO (Fig. 1).

HO.

CH3 I .

DMSO.K2CO3 rrc

11C-Methyl-BIII277CL

DISPLACEMENT ASSAYS

The pharmacological profile of Methyl-BIII277CL wasdetermined by an in vitro receptor-screening assay. Ata concentration of 100 nM, Methyl-BIII277CL showeda significant interaction only with the PCP binding siteof the NMDA receptor (79 % inhibition of specific bind-ing) and the cA -binding site (46 % inhibition). Theaffinity of Methyl-BIII277CL for these two binding siteswas further analysed in a displacement assay usingmembranes of mice cortex/hippocampus and [3H]TCPand [3H]Pentazocine as specific radioligands for thePCP- and a1-binding site, respectively. A high affinityinteraction of Methyl-BIII277CL with the PCP bindingsite was identified (Kj = 49 ± 14 nM) whereas bindingto the o1-binding site was 130-fold lower (Kj = 6.35 ±

0.26 (xM) (Fig. 2). Consequently, at nanomolar con-centrations Methyl-BIII277CL can be described as aspecific ligand for the PCP-binding site of the NMDAreceptor.

100-

spec

, bi

ndin

g

40-

2 0 •

0 •

• 3H-TCP; Cx; Ki = 4

$ 3H-Pentazocin; Cx;

• ~ \ X\ X

V V1 1 +^=* -J>=*

3+/-14 nM

Ki = 6350 +/- 260 nM

S 1

Fig. 1: Radiosynthesis of 11C -Methyl-BIII277CL.

log [Methyl-Bill 277 CL]

Fig. 2: Displacement curves of Methyl-BIII277CL(representative experiment).

SATURATION ASSAYS

For the estimation of the kinetic parameter ty2on andthe dissociation constant KD, Methyl-BIII277CL was11C-labelled and employed in saturation studies usingrat whole brain homogenates. Kinetic analysis of[11C]Methyl-BIII277CL binding was analysed by incu-bating the radioligand (70 nM) with the brain homoge-nate (0.5 mg/ml protein) at room temperature for dif-ferent time periods (1 - 85 min) followed by determina-tion of bound radioactivity via filtration. Non-specificbinding was quantified by co-incubation with 100 |j,MPCP and amounted to 45 %. The kinetic saturation

43

curve revealed that saturation of [11C]Methyl-BIII277CL binding was reached after 20 min (Fig. 3).The isotherm was fitted using the software KELL,giving the best fit for a single exponential function anda ty2on-value of 0.02 min.

1,0E-09

.4.0 1000 2000 3000 4000 5000 6000

incubation time [s]

Fig. 3: Time course of [11C]Methyl-BIII277CL bindingto rat whole brain homogenate (representativeexperiment).

Scatchard binding assays were performed in a con-centration range of 5 - 1500 nM of [11C]Methyl-BIII277CL. Increasing concentrations of the radioli-gand were incubated with the homogenate for 40 minat room temperature in absence and presence of 100|iM PCP. Scatchard transformation of [11C]Methyl-BIII277CL saturation binding data resulted in a singlehigh-affinity binding site of [11C]Methyl-BIII277CL witha dissociation constant (KD) of 6 + 1 nM and a maxi-mum number of binding sites (Bmax) of 670 ±154 fmol/mg protein (Fig. 4).

0,06

0,05 --

0,04 --

0,03 --

0,02 --

0,01 --

1.0E-10 2,0E-10 3.0E-10

bound

4,0E-10

Fig. 4: Scatchard plot of [11C]Methyl-BIII277CL bind-ing (representative experiment).

CONCLUSION

This study demonstrates that in vitro determination ofreceptor binding parameters of 11C-labelled receptorligands is feasible despite the physical limitation of ashort half-life. Methyl-BIII277CL interacts with highaffinity with the PCP-binding site of the NMDA recep-tor and shows a more than 100-fold higher bindingpreference to the PCP-binding site compared to theo1-binding site. Furthermore, [11C]MethylBIII277CLbinding equilibrium in vitro is quickly reached within 20min. All these binding parameters warrant furtherevaluation of [11C]Methyl-BIII277CL as a PET tracerfor the NMDA receptor. Currently, in vivo PET studiesin rats and pigs as well as biodistribution studies inrats are in progress.

PET STUDIES OF 18F-MEMANTINE IN HEALTHYVOLUNTEERS

We previously reported on the synthesis, radiolabel-ling, biodistribution studies in mice and PET investiga-tions in a Rhesus monkey of 1-amino-3-[F-18]fluoro-5-methyladamatane ([18F]-memantine). In these stud-ies, the regional brain distribution of [18F]-memantinewas changed by memantine and (+)-MK-801 suggest-ing competition for the same binding sites. To furtherexamine the binding characteristics of[18F]-memantine, we carried out PET investigations in5 healthy volunteers. The time-activity curves of [18F]-memantine uptake indicated a homogeneous distribu-tion in gray matter (cortex and basal ganglia regions)as well as the cerebellum of human brain. Kineticmodelling of the PET data showed that the uptakecurves of [18F]-memantine in receptor-rich regionssuch as striatum and frontal cortex were best de-scribed by a 1-tissue, 2-compartment model. Thedistribution volume (DV") values of all gray matterwere, however, similar and ranged from 15 to 20

ml/ml. The white matter showed a DV" value of 15 +1.4 ml/ml. These results suggest that [18F]-memantinedistribution in human brain does not reflect regionalNMDA receptor concentration, and this radioligandlacks the specificity required for the PET imaging ofthe NMDA receptors.

RADIOLABELLING AND IN VIVO EVALUATION OF11C-M-MPEP FOR PET IMAGING OF THE ME-TABOTROPIC GLUTAMATE RECEPTOR5 (mGLUR5)

Binding of glutamate at the mGluR5, which belongs toclass I of the mGluRs, results in a G-protein mediatedactivation of phospholipase C (PLC) in dendrites andcell body neurons in most of brain areas. An exces-sive activation of these glutamate receptors is thoughtto implicate in a variety of neurological diseases suchas epilepsy, focal and global ischemia, pain and neu-rodegenerative diseases. In order to map the mGluR5in the brain and to evaluate its possible influence inthe above-mentioned disorders by PET, M-MPEP (2-methyl-6-(3-methoxyphenyl) ethynyl-pyridine), a newand highly potent (KD = 3.5 nM) non-competitive an-

44

tagonist for the mGluR5 was radiolabelled with car-bon-11. The radiosynthesis of 11C-M-MPEP was ac-complished by O-methylation of the desmethyl pre-cursor (hydroxy-MPEP) with 11C-methyl iodide (Figure5). 11C-MPEP was purified by reversed phase HPLCusing a semi-preparative |a-Bondapack C-18 columnwith 0.1% H3PO4/MeOH (70:30) as eluent and at aflow rate of 4ml/min. The product was formulated in asolution containing Polysorbatum (0.1%) and 0.9%saline.

DMF,K2CO3

120°C,10min.

Hydroxy-MPEP "C-M-MPEP

Fig. 5: Radiosynthesis of11 C-M-MPEP.

The total synthesis time was between 45-50 minuteswith a radiochemical yield of 10-20% and the finalproduct had a specific activity of 30-60GBq/mmol(800-1600Ci/mmol) at the end of synthesis (EOS).The radiochemical purity of 11C-M-MPEP was >99%.The log P value of M-MPEP was determined using theshake flask method in phosphate buffer(pH=7.4)/octanol and amounted to 1.3.

In vivo Rat-PET studies were performed and a timeactivity curve was generated. The PET-study indi-cated an initial rapid uptake of radioactivity into the ratbrain. Thereafter, a fast clearance from the brain re-gion was observed. Blockade studies by co-injectionof 11C-M-MPEP and non-labelled M-MPEP (1mg/kg)showed contrary to our expectations an increaseduptake of radioactivity in the rat brain (Figure 6).

3 j

2.5 -

2 -|- H I brain control

brain blockade

1000 2000 3000 4000 5000

time (s)

Fig. 6: Time activity curve of 11C-M-MPEP.

11C-M-MPEP, due to its high lipophilicity and non-specific binding may not be a suitable PET-radioligandfor imaging the mGluR5. Synthesis, radiolabelling andfirst biological evaluation of new derivatives of M-MPEP with a lower lipophilicity are currently beingpursued.

45

TUMOUR HYPOXIA MEASUREMENTS BY POSITRON EMISSION TOMOGRAPHY

M. Bruehlmeier, U. Roelcke1, B. Kaser-Hotz2, O. Gardelle, E. Sinnig, C. Vetter, S.M. Ametamey(1Cantonal Hospital Aarau, 2 Veterinary University Clinic Zurich)

Using 18F-fluoromisonidazole and 15O-H2O, we assess hypoxia and perfusion of malignant brain tumoursby PET. The goal is the treatment planning of radiotherapy considering hypoxia.

INTRODUCTION

Positron emission tomography (PET) offers a non-invasive method to determine important parametersfor treatment planning of malignant tumours, e.g. tis-sue perfusion, edema and hypoxia. We currently as-sess the usefulness of 18F-fluoromisonidazole(FMISO) to measure hypoxia of malignant brain tu-mours in humans. Tissue hypoxia leads to a trappingof FMISO by nitroreduction and covalent binding tocellular components, leading to signal increase inPET. However, hypoxia measurements by PET arehampered by limited tracer delivery in low-perfusedtissue and, in the brain, by the influence of the bloodbrain barrier. We correlated these parameters by per-forming multi-tracer PET studies in tumour patients. In4 patients with malignant brain tumours and in addi-tion in one dog with a peripheral fibrosarcoma, wemeasured hypoxia using FMISO and tumour perfusionusing 15O-H2O.

RESULTS

arterial plasmabrain tumournormal brain

2000 4000

Time [s]

6000

Fig. 1: Plasma concentration of FMISO and timeactivity curves (TAC) of tumour and normalbrain tissue in a 51-year old male patient witha glioblastoma multiforme.

In normal brain tissue, radioactivity slowly increasesduring time, indicating that no equilibrium is reached.The TAC of the brain tumour starts at a higher leveland then increases at a steeper slope than normalbrain, the latter indicating tissue hypoxia. However,conventional, static FMISO PET images, typicallyperformed at 2-3 hours after tracer injection, solelydepend on current tissue radioactivity, i.e. the ,,rightend" of the TAC. The steep increase of tumour TAC atthe beginning, compared to a slow tracer uptake into

normal brain, is clearly a consequence of the dis-rupted blood brain barrier in tumour and in part con-founds FMISO accumulation in late images. In otherwords, brain tumours with a disrupted blood brainbarrier would appear more ,,hypoxic" than tumourswith an intact barrier.

14 - IM.

ml]

I m£ 2I

0

1

O venous plasma

—A— low-perfused tumour area

—H— well-perfused tumour area

^ a & — e 0 0 0

2000 4000 6000

Time [s]

Fig. 2: Plasma concentration of FMISO and timeactivity curves (TAC) of a fibrosarcoma in adog, with a low-perfused area in the centre ofthe tumour and well-perfused tumour parts atthe rim.

The TAC from the well-perfused area peaks and pro-gresses virtually parallel with the radioactivity inplasma, indicating FMISO distribution in a single com-partment. The TAC from the low-perfused tissueshows that tracer delivery to tissue is limited at earlyscan times. Both TACs eventually reach the samelevel, i.e. that late FMISO PET images in this caseshow a homogenous FMISO distribution throughoutthe whole tumour, apparently without hypoxia. How-ever, if a kinetic analysis is performed, an irreversibleFMISO trapping in the centre of the tumour can bedemonstrated, indicating hypoxia.

CONCLUSION

When measuring tissue hypoxia by PET, the influenceof perfusion and the brain blood barrier have to beevaluated carefully. FMISO has limitations in braintumours, where BBB alterations may play a role, andin peripheral tumour, where perfusion is very low,especially if hypoxia measurements solely rely on alate, static PET image. These problems can in part beovercome by suitable kinetic analysis, if a dynamicPET data set is available.

46

SMALL ANIMAL PET IMAGING AT PSI

S.M. Ametamey, M. Honer, J. Missimer

A PET camera, the quad HIDAC, designed to image small animals PET was delivered to and installed atPSI in December. This camera, a noninvasive tool to access biological function, will permit us to addressimportant issues in neurobiology and oncology with medical imaging techniques.

Limited until recently to humans and large animalsdue to the spatial resolution of commercial cameras,positron emission tomography (PET) dedicated toimaging small laboratory animals has been developedduring the past five years. The motivation for this de-velopment proceeds from interest in addressing ques-tions in neurobiology and oncology with medical imag-ing techniques (1). With the increasing number ofanimal models of human diseases, e.g. transgenicmice, PET represents a noninvasive tool to accessbiological functions. Clearly, this new tool could bevery useful to the pharmaceutical industry.

Manufactured by Oxford Positron Systems, the quadHIDAC is one of the first commercial PET camerasdesigned to image small animals (2). HIDAC is thepseudonym for High Density Avalanche Chamber,quad refers to the quadratic configuration of 16 planardetector modules, four modules stacked on each side.The distance between opposite modules is 170 mmand the modules are 280 mm deep, providing a cylin-drical field of view with corresponding diameter andlength. This space is adequate for the rats or mice forwhich the camera is designed.

Each detector consists of a multiwire proportionalchamber combined with laminated plates of inter-leaved lead and insulating sheets mechanically drilledwith a dense matrix of small holes. An electric fieldfocuses conversion electrons to the centre of theholes. They are .4 mm in diameter and .5 mm fromcentre to centre, yielding intrinsic submillimeter resolu-tion. In order to assure uniform acquisition, the detec-tor bank rotates continuously 180° backwards andforwards every 6 sec.The camera delivered to PSI in mid December is thesecond commercial model. Before ordering the cam-era, we examined a prototype housed at Hammer-smith Hospital, London. Acceptance tests performedat Hammersmith since delivery of the first commercialmodel last summer and at the factory in Novemberdemonstrated a resolution of 1.2 mm full width at halfmaximum after reconstruction using the filtered backprojection algorithm supplied with the camera. This ishalf the resolution of commercial animal PET camerasusing crystal detectors, justifying our decision to pur-chase the HIDAC. The efficiency of our camera fordetecting coincident photons from a point source isabout 1%.

. • • • • • ,

After initial testing and phantom studies, equipping thecamera with devices to prevent the rodent from mov-ing, to keep it warm, and to monitor and maintain itunder anaesthesia is necessary before the scientificprogram can begin.

The goals of the research program at Centre for Ra-diopharmaceutical Science include the developmentof PET radioligands for the glutamatergic system,monitoring the response to radio-immunotherapy fol-lowing treatment of diseased animals with therapeuticdoses of radiolabelled antibodies and retroviral vec-tors, and imaging receptor-positive tumours with Cu-64 labelled tumour-binding peptides.

REFERENCES

1. Phelps, M.E. PET: The Merging of Biology and Imag-ing into Molecular Imaging, J. Nucl. Med., 41, 661-681,2000.

2. Jeavons, A.P.,Chandler, R.A. Dettmar, C.A.R. A 3DHIDAC-PET Camera with Sub-millimetre Resolution forImaging Small Animals, IEEE Trans. Nucl. Sci., 46,468-473, 1999.

47

RCBF CHANGES ASSOCIATED WITH THE PRESENTATION OFPLEASANT STIMULI IN HEROIN ADDICTS

C. Martin Soelch, A-F. Chevalley1, K.L. Leenders2, J. Missimer, S. Magyar, G. Kuenig,, W. Schultz3

(1University Hospital Geneva, 2University Hospital Groningen, The Netherlands, 3University of Fribourg)

We performed a study comparing how the brain processes pleasant stimuli in young healthy controls withthe processing in a group of former opiate addicts. The results appear to confirm that opiate addiction isassociated with changes in the processing of emotional salient information in the brain.

INTRODUCTION

Substance abuse is one of the major causes of pre-mature and preventable illness, disability and death.Some personality features seem to play a role in de-pendence processes. Thus drug addicts have highervalues in sensation-seeking and in hedonism thancontrols (1,2). Furthermore, in a previous PET-studywe showed that the brains of opiate addicts react in adifferent way to motivational stimuli than controls (3).For this reason, we performed a new study in whichwe compared how pleasant stimuli are processed inthe brain in a group of young healthy controls and in agroup of former opiate addicts. This project is collabo-ration between PSI, the Department of SubstanceAbuse of the University hospital of Geneva, and theInstitute of Physiology of the University of Fribourg.

METHODS AND PATIENTS

We compared a group of 9 right-handed, healthy malevolunteers with a group of 10 right-handed, male for-mer heroin addicts on methadone maintenance. Sub-jects were matched for age. We measured regionalcerebral blood flow (rCBF), an index of neural activity,with H2

15O PET. rCBF was recorded using H215O PET

while subjects watched 3 minute film clips of musicalcolor feature films. The films were selected by sub-jects as producing pleasure. We presented 4 types offilm clips: science-fiction, sea and underwater ani-mals, acrobatic skiing and rebellion. The clips werepresented in a semi-random order. We performed onebaseline scan and four scans with clips selected toproduce pleasure in one session. We recorded sub-jective ratings of pleasure, well being and arousalimmediately after each scan on a visual analoguescale. We presented a 5 minutes visuo-cognitive taskwashout between each clip.

RESULTS AND DISCUSSION

In healthy controls, our results showed rCBF in-creases in the mediodorsal prefrontal cortex, the sup-plementary motor area, the orbitofrontal cortex, thegyrus cinguli as well as in different regions of the pa-rietal, temporal and occipital cortex, including the pri-mary visual area. In opiate addicts, activation wasfound in the same regions as well as in additionalregions like the striatum, the thalamus and the mid-brain.

Thus the orbitofrontal and mediodorsal frontal cortexas well as the occipital cortex seem to be involved inthe processing of emotional information. Interestinglyregions associated with motor behavior were alsoactivated in response to emotional stimulation, al-though no motor response was requested in the ex-periment. The regions that were additionally activatedin opiate addicts are regions influenced by dopamineand involved in reward processing. Furthermore, opi-ate addicts (3) already showed a different pattern ofactivation in these regions in response to reward.These findings appear to confirm that opiate addictionis associated with changes in the processing of emo-tional salient information in the brain.

A. OPIATE ADDICTS B. HEALTHY CONTROLS

Fig. 1: SPM projections of significantly activatedbrain areas in the comparison between pleas-ure and baseline conditions (p=0.001) A. inopiate addicts, B. in controls.

REFERENCES

1. Zuckerman, M. Behavioral expressions and biosocialbases of sensation seeking, Cambridge, New York :Cambridge University Press, 1994.

2. Chevalley, A.F., Mino, A., Jeavons, A. Sensationseeking and hedonism in drug addicts, In preparation.

3. Martin-Soelch, C , Chevalley, A.F., Kuenig, G., Missi-mer, J., Magyar, S., Mino, A., Schultz, W., Leenders,K.L. Changes in reward-induced brain activation inopiate addicts, Submitted to Eur. J. Neurosciences.

49

V RADIATION MEDICINE

Division of Radiation Medicine -

Patient Treatments and Tumor Therapy Evaluation 51

A single low dose of X-rays induces high frequencies of genetic instability and heritable

damage 52

Spontaneous animal tumors 54

Results with a grid chamber for fast proton beam monitoring 55

Routine patient setup at the PSI-gantry using a remote CT scanner 56

Initial experience of using a CCD based dosimetry system 57

Intensity modulated proton therapy: A first clinical example 58

OPTIS: Summary of the results 60

Fourth beam period of proton radiation therapy using the SpotScanning Gantry 62

Tuloc: A novel technique for the tracking of tumor positions in conformal radiotherapy 63

51

DIVISION OF RADIATION MEDICINE -PATIENT TREATMENTS AND TUMOR THERAPY EVALUATION

G. Goitein, H. Blattmann

The beam period 2000 has been the fourth season forpatient treatments on the PSI Spotscanning Gantry,again, like in 1999, with an increase of the number ofpatients. This, of course, is only one positive aspect ofthe Proton Therapy Project, which has two importantactivities: The OPTIS program for irradiation of ocularmalignancies with the scattered 70 MeV proton beamof the pre-injector 1, and the Gantry program fortreatments of deep seated lesions in various anatomi-cal regions with scanned proton beams of variableenergies.

OPTIS is the very established branch with the world-wide largest number of patients per year. The clinicalresults are excellent, as the continuous follow-up ef-forts are showing. OPTIS is at the same time the con-vincing example for the medical benefits one canachieve by using protons, especially for large lesionsin relation to the radiation-sensitive anatomical com-partment, and for the importance of high quality per-formance of the treatments as well as regular, longterm follow-up; the latter being often the most difficultpart of the work.

The Gantry program has become a well regardedmedical contribution to national and international ef-forts to improve cancer treatment. Patients were re-ferred by ten Swiss and four international centers ofRadiation Oncology, Neurosurgery and Surgery.Chordomas and chondrosarcomas were the maingroup of tumors, and lesions outside the base of skullwere more frequent than in previous year. Particularlylarge tumor volumes meant real challenge for treat-ment planning, performance and the whole therapeu-tical concepts. Of the 31 patients treated betweenApril and December 2000, twenty-eight received cura-tive therapy for primary lesions. Combined therapywith protons and photons were on the program as wellas the combination of proton irradiation with chemo-therapy, which is particularly promising for pediatrictumors. The daily treatment performance has beenimproved again, with respect to hardware, software,e.g. implementation of a "PatBase" program and aspeed up of the spot application, and work proceduresin general.

One strong goal of our activities is the development ofproton beam technology, which can be transferred tomajor hospitals in order to add to high quality cancertreatment modalities. Strong national and internationalinterest supports PROSCAN, the program for the im-plementation of a medical accelerator at PSI and thedevelopment of a second generation Gantry, whichcould be commercialized. The decision about the typeof accelerator is about to be taken, and the installationof this machine together with new beam lines for the

existing Gantry, OPTIS and the Gantry 2 will be a jointactivity of many groups at PSI and one of our mainfocuses for the coming year.

Within the TTE program the project for segmentationof abdominal organs in CT scans has been com-pleted. The conclusion from the Ph.D. work of M.Quicken is, that the results of the completely auto-matic, statistical shape model based segmentation forbladder and prostate need manual post-processing toreach sufficient accuracy to be used for quality assur-ance purposes in radiation therapy of the prostate.

At the veterinary hospital of the University of Zurichthe Linac, and with it the treatment planning system,have been taken into full operation this year. Thismade it possible to perform first treatment planningcomparisons between proton and photon treatment forcanine nasal tumors. For patients where the brain hadto be partially included in the treatment volume, thedose volume histograms demonstrate a clear advan-tage for proton irradiation. With a patient throughput ofapprox. 400 patients per year, a good basis for furthercomparative studies is available.

On the cellular level of radiation biology, with theapoptosis assay of leukocytes, the radiation inducedapoptosis rate is being studied in dogs before andduring a treatment course. A second main activity wasthe radiation induced genetic instability, discussed in aseparate contribution (Crompton et al.).

Within the "Microbeam Radiation Therapy Project" atthe ESRF, headed by Prof. J. A. Laissue, Bern, arraysof parallel thin ( =25 - 50 |j,m), closely spaced (5-10/mm) microplanar beams of synchrotron x rayshave been applied on developing brains of rats andpigs uni-directionally or bi-directionally. A 1.5 cm x 1.5cm lateral array of equally spaced 25 |j,m-wide verticalmicrobeams was targeted to the center of the cerebel-lum. The skin-entrance doses ranged from 150 Gy to600 Gy. Absorbed doses were computed by use ofthe PSI version of the Monte Carlo code GEANT. Inpigs, more than one year (first litter), or half a yearafter irradiation (second litter), the irradiated animalsremained indistinguishable from their sham-irradiatedlittermates in terms of body weight, behavior, neuro-logical status and cerebellar imaging criteria. Theobservations in microbeam-treated rats and pigletssuggest that MRT might be useful, where other kindsof radiotherapy for brain tumors of infancy are notsafe. The underlying radiobiological mechanisms willbe studied at the ESRF and possibly later at the SLS,within a collaboration initiated by the Swiss Society forRadiobiology and Medical Physics.

52

A SINGLE LOW DOSE OF X-RAYS INDUCES HIGH FREQUENCIES OF GENETICINSTABILITY AND HERITABLE DAMAGE

N.E.A. Crompton, Y.-Q. Shi, G.C. Emery, F. Wuergler1 H. Blattmann

institute for Toxicology, Swiss Federal Institute of Technology Zurich,CH-8603 Schwerzenbach, Switzerland

Proton therapy reduces the radiation load to healthy tissues during radiotherapy of cancer patients.Genotoxic cancer therapies (radio- and chemotherapies) need to be tailored to the characteristics of apatient, and his or her tumour, in order to keep potential chronic treatment complications associated withinduction of genetic damage in healthy normal tissues to a minimum. We have investigated these issuesusing a flow cytometric assay of radiation-induced genetic instability.

INTRODUCTION

Genetic instability presents a complex dilemma forcancer therapies. Many forms of cancer treatmentinvolve the use of drugs, which are potentially clasto-genic. They can cause second malignancies and drugresistance in non-malignant, healthy cells by inducinggenetic instability (1). Induction of heritable damageleading to genetic instability is an important considera-tion in cancer therapy, which motivates the develop-ment of novel, less genotoxic, treatment designs andtechnologies. In previous studies we reported thatradiation-induced morphologically transformed C3H10T1/2 mouse fibroblasts displayed unstable intracel-lular DNA contents, expressed as aneuploidy and/orpolyploidy (2,3). In the present investigations we ex-amined whether genetic instability is limited to mor-phologically transformed cells, or if it is a commonfeature of both transformed and non-transformed cellsdamaged by ionizing radiation. We examined whetherradiation can induce similar effects in other cell types.We have also reported that C3H 10T1/2 cells displayreduced checkpoint stringency when compared withhamster fibroblasts and human lymphoblastoid cells(3). Therefore, we investigated the role of checkpointstringency and p53 function in the induction of herita-ble damage. In these studies, two parameters weremeasured in the surviving clones of four cell linesexposed to radiation. Intracellular DNA content wasmeasured in two rodent, fibroblast cell lines (C3H10T1/2 and V79). Intracellular DNA content and pre-disposition to apoptosis were measured in two human,lymphoblastoid cell lines (TK6 and WTK1), which dif-fer in functional p53 status.

RESULTS

In total 321 clones were isolated, propagated andanalysed. 102 clones were derived from cells not ex-posed to radiation and 219 clones were derived fromexposed cells. 149 rodent fibroblast clones and 172human lymphoblastoid cell clones were examined. Atotal of 44 C3H T101/2 murine fibroblast clones wereharvested and analysed: 26 clones from non-irradiated cells and 18 clones from irradiated cells.Only one of the 21 non-irradiated clones displayed anabnormal DNA content (3.8%). This control clone

displayed a mixture of diploid and tetraploid DNA con-tent. All eighteen clones grown from cells surviving 8Gy X-rays displayed abnormal DNA contents (100%).14 clones were hypoaneuploid, 2 were hyperane-uploid, and 2 displayed dual hyperaneuploid peaks.None of these clones was isolated from a morphologi-cally transformed focus. Increased intracellular DNAcontent correlated positively with increased cell size.A total of 105 V79 Chinese fibroblast clones wereharvested and analysed: 25 clones from non-irradiated cells and 80 clones from irradiated cells.None of the clones from non-irradiated control cellsdisplayed an abnormal DNA content. Among the 80clones grown from cells surviving 8 Gy X-rays, 9(11.25%) displayed abnormal DNA content: 2 hypo-diploid, 4 hyperdiploid, 1 tetraploid, 1 with both hypo-diploid and diploid peaks, and 1 with both diploid andhyperdiploid peaks.

Because rodent cell lines tend to display less stringentcell cycle checkpoint controls than human cell lines(4), we investigated human B-lymphoblastoid cell linesfor evidence of radiation-induced ploidy change. Weexamined two closely related cell lines: TK6 andWTK1, which have similar cytological and growthcharacteristics but which differ in their functional p53status. TK6 cells have wild-type p53 function butWTK1 cells have mutated p53 function. The lym-phoblastoid cells exposed to radiation received a doseof 1.5 Gy X-rays. A total of 61 TK6 clones were har-vested and analysed: 29 clones from non-irradiatedcells and 32 clones from irradiated cells. All of themdisplayed normal diploid DNA content. A total of 111WTK1 clones were harvested and analysed: 22clones grown from non-irradiated cells and 89 clonesgrown from irradiated cells. A single abnormal clonewas found amongst the control clones, which washyperaneuploid. Three abnormal clones were ob-served in the clones grown from irradiated cells. Twowere hyperaneuploid; the third displayed dual peaks,one diploid and the other hyperaneuploid. The fre-quency of clones with ploidy change grown from non-irradiated WTK1 cells (1/22) was not significantly dif-ferent (p > 0.6, t = 0.273) from the frequency of thosewith ploidy change grown from irradiated WTK1 cells(3/89).

53

Human lymphoblastoid cells are highly sensitive to thepresence of clastogenic damage, which causes themto undergo apoptosis. The apoptosis effectively elimi-nates potentially carcinogenic and mutagenic cellsfrom a population. We examined the TK6 and WTK1cell lines for their ability to undergo radiation-inducedapoptosis. TK6 cells displayed high levels of apop-tosis 48 h after exposure to radiation. However, WTK1cells, which contain two mutated p53 genes, wereunable to mount a radiation-induced apoptotic re-sponse.

Normal (< 14% apoptosis) WTK1 clones, grown fromnon-irradiated cells, displayed more backgroundapoptosis (8.4%) than normal TK6 clones (5.7%),reflecting the skewed character of the WTK1 fre-quency distribution. Even in the absence of radiation,predisposition to apoptosis caused by mutated p53function was observed in 5 of 22 WTK1 clones(22.7%). The average apoptotic fraction in these 5clones was 23.9%. Predisposition to apoptosis in-duced by exposure to 1.5 Gy X-rays was observed in5 of 32 TK6 clones (15.6%). The average apoptoticfraction in these 5 clones was 40.0%. Although fewerclones expressed predisposition to apoptosis after X-ray exposure, in those that did, the predisposition wassignificantly greater. Predisposition to apoptosis in-duced by exposure to 1.5 Gy X-rays was observed in53 of 89 WTK1 clones (59.6%). The average apop-totic fraction in these 53 clones was 37.9%. No signifi-cant differences between the apoptotically predis-posed TK6 and WTK1 clones were observed, either inthe predisposition to apoptosis (p > 0.8, t = 0.217) orin the time required till harvest (p > 0.8, t = 0.208).

Our data on predisposition to apoptosis suggest thatheritable damage is frequently induced by radiationexposure. Under suitable conditions it can lead togenetic instability and tumorigenesis or normal tissuedamage. Patients displaying autonomous geneticinstability due to specific gene deficiencies, such asAtaxia telangiectasia or Nijmegen breakage syn-drome, show both enhanced radiosensitivity (5,6) andan enhanced predisposition to cancer (7). Diagnosisof patients, who are hypersensitive to radiation-induced toxicity, is of significant clinical relevance (8).More targeted or conformal treatments, such as pro-ton therapy, should be used during radiotherapy toreduce dose to healthy tissues received by patients,particularly to patients with extensive post-therapy life-expectancy. Our data clearly demonstrate the induc-tion of heritable damage in both rodent and humancell lines. A single dose of 1.5 Gy X-rays induced pre-disposition to apoptosis in at least 15.6% of the TK6cells and 36.9% of the WTK1 cells. In order to induceheritable damage in these cells it was not necessarythat they be exposed to multiple rounds of treatment,as suggested by recent studies of genetic instabilityinduced by chemotherapy (1). A single low dose was

sufficient indicating that mammalian cells are exqui-sitely sensitive to its induction. In the different celltypes studied, radiation induced a different spectrumof genetic instabilities, as has been reported previ-ously (9). We also observed that various mechanismsto suppress expression of radiation-induced heritabledamage exist in cells, depending on cell type and p53function. Nevertheless, we conclude that genotoxiccancer therapies (radio- and chemotherapies) shouldbe tailored to the characteristics of the patient, and hisor her tumour, in order to keep chronic treatmentcomplications associated with induction of heritabledamage in healthy normal tissues to a minimum. Notabene: Dr. med. Dr. rer. nat. Yu-Quan Shi was heavilyinvolved in these studies. We are pleased to be ableto congratulate him on the successful completion ofhis PhD studies at the ETH Zurich. He has begun apost doc in Philadelphia at the University of Pennsyl-vania and we wish him the best for the future.

REFERENCES

1. Finette, B.A., Homans, A.C., Albertini, R.J. Emergenceof genetic instability in children treated for leukemia.Science. 288: 514-517, 2000.

2. Crompton, N.E.A., Sigg, M., Jaussi, R. Genome labilityin radiation-induced transformants of C3H 10T1/2mouse fibroblasts. Radiat. Res. 138: S105-108, 1994.

3. Crompton, N.E.A., Emery, G.C., Shi, Y., Sigg, M.,Blattmann, H. Radiation-induced genetic instability: noassociation with changes in radiosensitivity or cell cy-cle checkpoints in C3H 10T1/2 mouse fibroblasts. Ra-diat. Environ. Biophys. 36: 255-259, 1998.

4. Schimke, R.T., Kung, A.L., Rush, D.F., Sherwood,S.W. Differences in mitotic control among mammaliancells. In The Cell Cycle, 417-425. Cold Spring HarbourLaboratory Press 1991.

5. Carr, A.M., Piecing together the p53 puzzle. Science287: 1765-1766,2000.

6. Shiloh, Y. Ataxia-telangiectasia and Nijmegen break-age syndrome: related disorders but genes apart.Annu. Rev. Genet. 31, 635-662, 1997.

7. Difilippantonio, M.J., Ahu, J., Chen, H.T., Meffre, E.,Nussenzweig, M.C., Max, E.E., Ried, T., Nussenzweig,A. DNA repair protein Ku80 suppresses chromosomalaberrations and malignant transformation. Nature 404:510-514,2000.

8. Crompton, N.E.A., Miralbell, R., Rutz, H.-P., Ersoy, F.,Sanal, O., Wellmann, D., Bieri, S., Coucke, P.A., Em-ery G.C., Ozsahin, M. Altered apoptotic profiles in irra-diated patients with increased toxicity. Int. J. Radiat.Oncol. Biol. Phys. 45: 707-714, 1999.

9. Miyazaki, M., Furuya, T., Shiraki, A., Sato, T., Oga, A.,Sasaki, K. The relationship of DNA ploidy to chromo-somal instability in primary human colorectal cancers.Cancer Res. 59: 5283-5285, 1999.

54

SPONTANEOUS ANIMAL TUMORS

B. Kaser Hotz, A. Sumova, J. Fidel, C. Rohrer, S. Stankeova, P. Theiler, R. Achermann,O. Gardelle, H. Blattmann

The installation of a linear accelerator at theveterinary school was the highlight of the year2000. This finally enables us to perform directcomparisons between photon and proton therapyplanning for the same patient. We continue toexploit the NTCP model of Lyman and Burmannto study normal tissue side effects. Our prelimi-nary data on canine nasal tumors indicates, thatproton therapy is most advantageous when apartial brain volume of a considerable size has tobe irradiated. Once the volume is very large,such that almost the entire brain is involved, theadvantages of proton irradiation is less pro-nounced.

Fig. 1: Dose distribution for a dog with a nasaltumor, left for protons right for photons.The arrows mark the area of major dif-ference.

0 20 40 60 80 100Relative target dose %

35 40 45 50 55 6'

Dose Gy (3.5 Gy/fx)

Fig. 2: Dose volume histograms and complica-tion probabilities for protons and photonsof the same dog.

There were 4 new dogs entered into the doseescalation program. The highest selected dose isnow being used and it seems tolerated, at leastclinically. We are continuing to collect histopa-thologic data.

Last fall we started along a new research ave-nue: molecular biology-based radiation therapyplanning. There is renewed interest in tumorhypoxia, for several reasons: hypoxic tumors arenot only more radioresistant but they also tendto behave biologically more aggressively. Spe-cifically, hypoxic conditions promote the upregu-lation of hypoxia induced factor-1 (HIF-1) whichinduces vascular endothelial growth (VEGF).VEGF enhances tumor angiogenesis and thuspromotes tumor growth. In vivo determination ofintratumoral oxygen tension with the Eppendorfneedle electrode suggests that the level of hy-poxia varies considerably between and withintumors. If representative images of the three-dimensional tumor hypoxia levels were available,this information could be directly incorporatedinto conformal proton therapy or intensity modu-lated radiation therapy (IMRT) planning. Conse-quently, the regions of more pronounced tumorhypoxia could then be targeted with higher radia-tion doses.

Fig. 3: With the Eppendorf needle profiles, ofoxygen partial pressures are measuredin vivo.

j 10 15 20 25 30 35 40 45 50 55 60 65 70 75 SO 85 90

Tissue oxygen pressure (mmHg)

Fig. 4: Example of a distribution of partial pO2pressure measurements in a myosar-coma of a dog. The high frequencies ofoxygen pressure measurements below20 mm Hg. are evidence for hypoxia inthis tumor.

55

RESULTS WITH A GRID CHAMBER FOR FAST PROTON BEAM MONITORING

S. Lin, A. Coray, T. Bohringer, E. Pedroni

INTRODUCTION

A grid ionization chamber, based on a design ofa detector traditionally used for measuring alpharadioactivity (1), has been developed andchecked for four year along with patient treat-ments at PSI for use as a fast proton beam fluxmonitor. The results have proved that it is stableand reliable for our applications. Comparing itwith our parallel plane ionization chamber (PIC),monitor 1 (2), it has the advantages that it isfaster and is not sensitive to microphonics ef-fects (3). The grid chamber will be used to re-place monitor 1 as the dose steering monitor.

OVERVIEW OF THE GRID CHAMBER SYS-TEM

monitor 2cathode

guard ring i

ionization current

nomnalpreset 2 I'Wstch dcg1

of next spot<£bes=rncff»

> interlock

Fig. 1: Schematic diagram showing the principleof the grid chamber system, where d =1.0mm, r =0.05mm, p = 10.0mm, q =30.0mm, V c = - 4500V, Vg =-1500V;

d 2 = 1.0 cm, V 2 = +2000V.

Figure 1 illustrates the principle of the system.The dose check monitor 2 (PIC) (2) is locateddownstream of the grid chamber. The anode ofthe grid chamber collects the ionization elec-trons. The grid shields the anode from the induc-tion of the positive ionization ions which are mov-ing toward the cathode. The grid chamber isfilled with pure nitrogen gas instead of ambientair, in order to avoid electron attachment by the

electronegative impurities. The gas pressure iskept at about 1 % higher than the ambient pres-sure by adjusting the gas flow with a gas pres-sure regulator. The theoretical and experimentalstudies of the chamber have been described inref (3).

RESULTS FROM THE LOG FILES OF THEPATIENT TREAMENTS

J 0.99 -

0 2000 4000 6000 8000 10000 12000

Nominal spot preset

Fig. 2: The ratio of the measured single spotcounts (gridmonitor/monitor 1) againstthe nominal spot presets in a period ofone patient treatment. The spot dosewas defined by monitor 1. (Monitor 1 islocated between the grid chamber andmonitor 2 and not shown in Figure 1.

1 .06 -

^ 1 .03-

0 1 .02-

| 1.01 -

1 1 .00-

1 ° "

0.97 -

" " m'%

t-

x

P00011

j , " -•>

»cx ,(>

• • + * •

160 MeV

t - x*8p »Sois * S< j f x

,_ _ x 2 *< » a

+ * % * , • !

**+ *+ + + • ••

x• -• j j ^ X '

• • • *

£ | f l

+ ' + +

"+ • + •

+SD = 0.0037

Fig. 3: Comparison of the measured ratio(gridmonitor/monitor 1) of one patienttreatment against the whole treatmentperiod (3 months, 96 treatments and 32fractions). The spot dose was defined bymonitor 1.

In Figure 2, when the nominal spot preset islower than about 2000, the spot length will beless than 1 millisecond. Therefore the precisionof the ratio measurement is limited and the re-sults show more deviations. Figure 3 illustratesthat the grid chamber remained stable during thewhole patient treatment period of three months.

1. Bunemann, O. et. Al. Canadian J. of Research,Vol. 27:191-206, 1949.

2. Pedroni, E. et al. Med.Phys. 22(1), January1995.

3. Lin, S. et. al. 16-17, Annual Report, Annex II,1995.

56

ROUTINE PATIENT SETUP AT THE PSI-GANTRY USING A REMOTE CT SCANNER

W. Roser, T. Bohringer, E. Egger, G. Goitein, M. Grossmann, A. Lomax, L. Wisser

Conformal radiotherapy requires a highly precise patient positioning procedure for the daily repositioningof the patients. At the PSI gantry, this procedure has been simplified in the year 2000 using table offsets,leading to a more accurate patient positioning, which is easier to achieve than in previous year.

INTRODUCTION

Due to the highly conformal nature of the treatmentsprovided by the PSI gantry, an accurate and repro-ducible patient setup procedure is required. In order toaccurately assess a patient's position, a remotelysituated computer tomography (CT) system has beenadopted. This decision was also taken in order to in-crease the patient throughput and thus to reduce theeffective costs of the whole treatment facility. With thisarrangement, most of the preparatory work with thepatients is performed outside the treatment room,leading to a more efficient utilization of the treatmentgantry.

METHOD

We are using a General Electric CT system, (GEMedical Systems, Milwaukee, Wl, U.S.A.) which wasinstalled in the winter shutdown period 1998/1999. Forthe determination of the correct patient position, wefirst acquire a few pre-selected CT slices. From thesepictures, which are directly compared to those fromthe planning CT, it is decided whether the patientneeds to be repositioned in his mould. Afterwards twoorthogonal CT topograms are acquired to confirm thecorrect positioning of the patient in the spatial direc-tions. In-house developed software allows a directcomparison of the position of predefined anatomicallandmarks in the topograms in and around the tumorregion to an accuracy of ± 0.5 mm.In summer 2000, our treatment control system hasbeen modified in such a way that a proton treatmentcan be applied with user-defined gantry table offsetsin order to correct translational errors in a patient'spositioning. This now enables us to skip most of thepatient's repositionings in the CT, provided that thepatient is not rotated and that the delivered fields areall coplanar, which is the case for about one half ofour patients.

RESULTS

The first patient with whom this possibility has rou-tinely been used was a patient being treated for acervical tumor. Fig. 1 shows an enlarged section ofthe lateral CT topogram used for daily positioning. Thetwo arrows indicate the positions of two anatomicalpoints, which were used for the determination of thetable offsets, i.e., the edges of a metal fixation plate atthe anterior aspect of the target region.

Fig. 1: Metal fixation plate in the patient's spine.

Table 1 summarizes the distribution of the daily CTpositioning offsets of the metal plate shown in Fig. 1,the applied gantry table offsets, and the resulting pa-tient offsets in the 37 treatment fractions. From thesedata it is clear that it was possible to obtain a position-ing accuracy of this patient, which was below 1 mm inall directions.

Table 1: Mean positioning errors of the metal plateshown in Fig. 1 over all 37 treatment frac-tions.

X

y

z

CT topogramoffset

2.0 ± 1.0 mm

1.8 ± 1.8 mm

-0.9 ± 1.5 mm

Applied gantrytable offset

-1.7 ± 1.2 mm

-1.6 ± 1.7 mm

0.7 ± 1.4 mm

Resulting treat-ment offset

0.3 ±0.7 mm

0.2 ±0.8 mm

-0.2 ± 0.6 mm

CONCLUSIONS

By the use of orthogonal topograms with our remoteCT system we are able to prepare a patient for treat-ment with a high degree of positional accuracy whileanother patient is still being treated on the proton gan-try. With the introduction of automatic corrections ofmeasured offsets from our CT topogram analysis, wehave shown that in head and neck cases a positioningaccuracy of 1 mm or less over the whole treatmentperiod is achievable. This option offers an increasedpatient throughput together with an improved preci-sion of the preparatory process for each single treat-ment fraction.

57

INITIAL EXPERIENCE OF USING A CCD BASED DOSIMETRY SYSTEM

E. Pedroni, S. Lin, H.U. Stauble, O. Stadelmann

INTRODUCTION

The use of a scintillating screen viewed through amirror by a CCD camera has been firstly proposed fordosimetry in proton beams by our colleagues in Gron-ingen. Results obtained with scanning beams at PSIby this group have been recently published (1). In thisarticle we report on the initial experience of using ourown equipment developed specifically for routine do-simetry on the PSI gantry.

THE NEW EQUIPMENT

Fig. 1 : Photograph of the CCD equipment.

The new apparatus can be mounted on the PSI gantryin a reproducible way and provides the capability tomeasure, under remote control, dose distributions atdifferent gantry angles and at different depths below astack of Plexiglas plates.

SYSTEM PERFORMANCE

Fig. 2: Dose shadows of a thimble ionisation cham-ber in the homogenous portion of a dose field(see text).

To give an impression of the level of sensitivity, whichcan be achieved with the CCD system, we show inFig. 2 a dose distribution measured in the middle ofthe Spread Out Bragg Peak of a perfectly homogene-ous square box of dose. An Exradin ionisation cham-ber was mounted in a tight hole in the lowest of the

Plexiglas plates placed in front of the screen. For theimage of Figure 2 we have adapted the grey scalecontrast to see details in the very last few % of thehomogenous portion of the dose. One can easily rec-ognise the shadows of the thimble ionisation chamber.These effects arise from small changes in proton flu-ence, due to multiple scattering at the density inter-faces of the chamber. That such effects are depictedat the level of only some percent of the dose can beseen from picture 2, where we show the measureddose profiles with and without chamber.

FIRST RESULTS FOR ROUTINE DOSIMETRY

The CCD system combines at the same time a goodposition resolution with an excellent stability of themeasured signal, making it an ideal instrument forroutine verification dosimetry with scanning protonbeams. Possible disadvantages are a moderate de-pendence of the response with LET and the fact thatthe system is suitable only for relative dosimetry. Weshow in Fig. 3 (see Appendix 1) as an example thecomparison of measured and calculated dose distribu-tions for the central field of the first intensity modu-lated proton therapy treated at PSI in 1999 (2). Thecalculation of the dose contains a small dE/dx de-pendent correction for the quenching of the scintillatorresponse with LET, but the comparison is performedwith the same constant adjustment factor for alldepths. In order to give a more quantitative idea onthe agreement of the results, we show the dose com-parison for an arbitrary profile (position of the yellowband in the first image of Figure 2). The red curve (themeasurement) differs very little from the calculation (inblue) (the systematic discrepancy at the first depth isprobably due to a build-up effect of the dose depos-ited by nuclear interactions). The quick availability of alarge number of data points permits to rapidly obtain acomplete and precise overview on 3d shaped dosefields, otherwise very difficult to obtain using ionisationchambers alone. These good results show the highaccuracy of the physical dose model used in treat-ment planning and the reliability of the beam deliverysystem developed for the PSI gantry. Beside routinedosimetry, the CCD system is also well suited forcharacterising the details of the proton pencil beamused for scanning, for investigating multiple Coulombscattering effects in complex anatomical structures (inantropomorphic phantoms) and to measure the pe-numbra of collimators used in addition to scanning.

REFERENCES

1. Boon S.N. et al, "Performance of a fluorescent screenand CCD camera as two-dimensional dosimetry sys-tem ...", Med. Phys. 27(10), 2000.

2. Lomax, A. in this newsletter issue.

58

INTENSITY MODULATED PROTON THERAPY: A FIRST CLINICAL EXAMPLE

A.J Lomax, T. Boehringer, A. Coray, E.. Egger, G. Goitein, M. Grossmann, P. Juelke, S. Lin,E. Pedroni, B. Rohrer, W. Roser, B. Rossi, B. Siegenthaler, O. Stadelmann, H. Stauble,

C. Vetter, L. Wisser

INTRODUCTION

Here we report on the first treatment of a patient usingfully automated 3D intensity modulated proton therapy(IMPT). By 3D IMPT, we refer to the calculation anddelivery of multiple ports for which the optimisationprocess has been applied to all three dimensionallydistributed proton Bragg peaks from all fields simulta-neously. This results in a set of individually in-homogenous fields which nevertheless combine todeliver a homogenous resultant dose across the tar-get volume.

MATERIALS AND METHODS

A 34 year old man was referred for proton therapywith a Chondrosarcoma in the posterior portion of theseventh and eighth thoracic vertebrae. The definedtarget volume was prescribed a dose of 72Gy andwas found to wrap partially around the spinal cordcreating a 'classic' treatment planning problem (fig-ure 1). The spinal cord was assigned a maximumtolerance dose of 60Gy. The planning problem in thiscase was accentuated by two additional physical con-straints. Firstly, beams through the heart and lungwere to be avoided due to errors in the delivered dosethat could result from motion of these organs. Sec-ondly, the delivery of highly weighted Bragg peaksdeposited directly against the spinal cord were to beavoided, due to the risk of dose 'overshoot' into thespinal cord and possible RBE effects in the distaledge of the Bragg peaks.

In order to avoid these problems, a three field,patched technique was adopted, with Bragg peaksfrom all three fields being optimised using the 3D

IMPT method described by Lomax (1). Three inde-pendent assessments of the sensitivity of the calcu-lated plan to potential delivery and calculation errorswere used; 1) range error analysis, 2) patient set-uperror analysis and 3) Monte Carlo verification of theanalytic dose calculation.

RESULTS

Figure 1 shows the optimised dose distributions of theindividual fields planned for this case. Three fieldswere defined; right posterior oblique (figure 1a), poste-rior (figure 1b) and left posterior oblique (figure 1c). Inorder to avoid directing proton pencil beams directly atthe spinal cord, both the RPO and LPO fields weredesigned such as to only cover the left and right lobesof the target volume respectively (figure 1a and 1c).

With the RPO and LPO beams alone, it is impossibleto achieve a homogenous dose across the full targetvolume due to the combined plateau doses in themiddle-posterior portion of the target volume (the por-tion immediately posterior to the spinal cord). Thethird, posterior beam was therefore introduced. As aresult of the raised dose in the RPO-LPO overlapregion, and in order to obtain a homogenous resultantdose, the multiple field optimisation process necessar-ily reduces beam weights in the central portion of theposterior beam (see figure 1b). Due to the geometryof the patched field arrangement, these beams arealso the ones that are incident on the spinal cord.Thus, after optimisation, the posterior field deliversonly Bragg peaks of a relatively low weight against thespinal cord.

A CFig. 1: RPO (A), posterior (B) and LPO (C) fields for the optimised 3D IMPT plan. Note the clearly reduced dose in

the central portion of the posterior field (the portion incident on the spinal cord) and the patched arrange-ment of the RPO and LPO fields.

59

A CFig. 2: The nominal (A), 10% overshoot (B) and worst case (C ) 3 field IMPT plans. Note degree of conformation

achieved by the nominal plan (A), the sparing of the spinal cord even when considering large range uncer-tainties (B) and the potential cold areas (< 95% of PTV dose) along the patch lines that could result frompositioning errors of 5mm in the AP direction.

The resultant dose distribution for all fields together isshown in figure 2a. Through the use of three patchedfields (none of which pass through the thorax) and 3DIMPT, a high degree of conformity could be achievedwith almost complete sparing of both the lungs andheart and with no high weighted Bragg peaks beingdirected at the spinal cord.

A recalculated dose distribution with 10% reduced CTvalues is shown in figure 2b. Although an overshoot ofdose from the RPO and LPO beams into the lungs isclearly seen, no similar effect is evident in the regionof the spinal cord, indicating that the plan is relativelyinsensitive to range errors in this region. Indeed, com-parison of the nominal and overshoot spinal cordDVHs showed that although the mean dose to thecord would be increased in the case of such an errorin range, the maximum dose would remain the same(58Gy in both cases).

Figure 2c shows a composite, or 'worst case', dosedistribution derived from the six error dose distribu-tions resulting from the set-up error analysis (one foreach of the offsets along the major anatomical axes).The 'worst case' distribution clearly shows the possi-bility of cold regions in the target volume as the resultof positioning errors (figure 2c). Further analysis re-vealed that these areas would result from offsets ofthe patients positioning in the AP direction in particu-lar. Post-treatment analysis of the patients positioningshowed that, although day-to-day variations of theorder +/- 4mm (2SD) could be observed, the meanoffset over all fractions in the AP direction was foundto be only 0.1mm.

Due to the site of the lesion and the dynamic nature ofthe treatment method, it was also considered neces-sary to assess the magnitude of the organ motion inthe region of the target volume. To this purpose, rep-resentative inhalation and exhalation CT slices wereacquired twice through the course of the treatment.Both checks indicated that motion in the region of thetarget was minimal (of the order of 1 mm), even thoughmotion in the thorax (away from the target volume)could be quite considerable.

DISCUSSION

With only three fields, field patching and 3D IMPT, ahigh degree of conformity has been achieved for thischallenging case within a relatively short overalltreatment time (typical delivery times per field of 1-2minutes). In addition, by choosing three fields that allcome from the posterior aspect of the patient, minimaldose has been delivered to the heart and lungs andoverall integral dose has also been significantly re-duced.The main advantage of this approach however, is theinherently safer set of Bragg peak positions andweights that result. That this is the case has beenshown by the error analysis applied to this case. Therange error analysis presented here has been calcu-lated assuming a 10% underestimate of the CT (pro-ton stopping power) values in treatment planning. Infact, this figure is very much a worst-case value. Di-rect measurements of proton stopping powers forbiological tissues indicate that these can be estimatedfrom our CT with an accuracy of about 1 % in soft tis-sues and 2% in bone (2). However, as this patientwas relatively large, we took a value of 10% for thiscase to also allow for day-to-day variations in theamount of soft tissue that may in the posterior aspectof the target volume. Even with this relatively largeerror, the spinal cord has been found to be well pro-tected using this treatment technique.

CONCLUSIONS

3D IMPT has been clinically applied for the first time.Detailed error analysis of the plan has demonstratedthat the treatment can be applied safely (indeed thiswas the main impetus to apply IMPT methods in thiscase) and that highly conformal plans can be con-structed using small numbers of fields in a challengingtreatment site.

REFERENCES

1. Lomax, A. 1999 Phys. Med. Biol. 44 185-205.

2. Schaffner, B., Pedroni, E. 1998 Phys. Med. Biol.43: 1579-1592.

60

OPTIS: SUMMARY OF THE RESULTS

E. Egger, G. Goitein, L Zografos , D. Beati*, T. Bohringer, L Chamot*, W. Roser, A. Schalenbourg*

*H6pital Ophtalmique Jules Gonin, Avenue de France 15, CH-1004 Lausanne

The OPTIS program continued successfully during the year 2000, in which we treated 239 patients withocular tumors. By the end of December 2000, the total number of patients treated at our facility is 3259.Our main effort during the past year was concentrated on the analysis of our treatment and follow-up dataand the presentation of our results. A brief summary of these results is presented here.

INTRODUCTION

Proton therapy has become the treatment of choicefor large uveal melanomas and tumors located closeto sensitive structures like optic disc and nerve, andmacula. Since proton therapy is not available every-where, other treatment techniques like irradiation withthe gamma knife have been tried. The results pub-lished in (1), while reporting only on 7 years of follow-up, show that gamma knife irradiation of large uvealmelanoma is followed by significant morbidity. Thegamma knife was developed for radiosurgery and isnot indicated for fractionated treatments. Zehet-meyer, who has pioneered gamma knife treatmentsfor uveal melanoma, has abandoned this treatmentbecause he concluded that a larger number of frac-tions is required to reduce the complication rates (1).Proton therapy remains the only effective conserva-tive treatment for large uveal melanoma.

PATIENTS AND METHODS

Between March 1984 and December 1999, we havetreated 3014 patients presenting with ocular tumors.Among them, 2638 were uveal melanomas. They arethe subject of this report.

Follow up data available through August 2000 wereused in this analysis. The endpoint events assessedwere death from any cause, ocular tumor relateddeath, recurrence, enucleation, and loss of visualacuity to less than 20/200.

RESULTS

By August 2000, 225 patients had been lost to followup. 21 patients refused any follow up examinations.426 patients had died. The death was related to theocular tumor in 333 cases, 27 patients died fromother tumors. Four patients died from other causeswith known metastases from the ocular tumor. 35patients died from other causes without metastases.In 27 cases the cause of death could not be found byus. 76 patients had to receive a second treatmentbecause of tumor recurrence. 38 of them received asecond conservative treatment, 38 eyes were enu-cleated. In total, 218 eyes had to be enucleatedmainly because of uncontrolled neovascular glau-coma and functional loss.Our main concern was local tumor control, knowingthat recurrences have an important effect on survival

(2). We identified different causes for local tumorcontrol failures, such as reduction of safety margin,incorrect estimation of the thickness of the uppereyelid within the irradiation field, and invisible exten-sion of the tumor within the ciliary body. Identifyingthese causes, we were able to find remedies to theseproblems and thus, to increase progressively the rateof local tumor control. The rate of local tumor controlat 5 year passed from 90.8 % for patients treatedbefore 1988 to 99.1 % for patients treated after 1994.The Kaplan-Meier local tumor control rates areshown in figure 1.

Recurrence free survival1,00

o

24 48 72 96 120 144 168 192

Time [months]

Fig. 1: Local tumor control in function of treatmentyear.1. Patients treated before December 882. Patients treated between January 89 and

December 933. Patients treated after January 94

The improvement of local tumor control had also aninfluence on survival, as shown in figure 2. However,in order to be sure that the improved survival wasdue to improved local tumor control rate and not to amore favorable selection of patients, we analyzed thedata of the three groups of patients. This analysisshowed that tumor diameter, patient's age and thenumber of patients with extrascleral extension be-comes less favorable in the second and the thirdgroup, while tumor thickness becomes somewhatmore favorable. Tumor diameter being the most im-

61

portant parameter, we conclude that the improvementof the survival rate is related to the better local tumorcontrol. The 5-year survival rate was improved from82.7% for patients treated before 1988 to 91.3% forpatients treated after 1994.

Survival Functions

Eye Retention

EO

24 48 72 96 120 144 168 192

Time [months]

Fig. 2: Survival in function of treatment year.1. Patients treated before December 882. Patients treated between January 89 and

December 933. Patients treated after January 94

Half of the patients with recurrent tumors weretreated by enucleation. Therefore, we expected im-provement of local tumor control rate to have also abenefic effect on the eye retention rate, as shown infigure 3. While 81.4% of the patients treated before1988 still had conserved their eye after 5 year, thisnumber increased to 94.5% for patients treated after1994.The proportion of patients with useful vision is shownin figure 4. There are no changes at 5 year betweenthe 3 groups of patients analyzed.

CONCLUSION

Proton therapy of uveal melanoma has been demon-strated to be excellent in terms of local tumor controlwith a local tumor control rate greater than 99% at 5year. As compared to enucleation, survival is notreduced by application of such an efficient conserva-tive treatment. The eye retention rate is greater than90% at 5 year with no increased risk of death frommetastases as compared to enucleation. 40% of thetreated patients still have a useful vision 5 year aftertreatment. This is the only outcome which has not yetbeen improved. It will be subject for special care inthe future.

2.LLJ

E

d

0 24 48 72 96 120 144 168 192

Time [months]

Fig. 3: Eye retention in function of treatment year.1. Patients treated before December 882. Patients treated between January 89 and

December 933. Patients treated after January 94

Visual acuity

o

24 48 72 96 120 144 168 192

Time [months]

Fig. 4: Proportion of patients with useful vision infunction of treatment year.1. Patients treated before December 882. Patients treated between January 89 and

December 933. Patients treated after January 94

REFERENCES

1. Zehetmeyer, M. et al. Local tumor control and morbid-ity after one to three fractions of stereotactic externalbeam irradiation for uveal melanoma. Radiotherapyand Oncology, 55:135-144, 2000.

2. Suit, H.D. Potential for improving survival rates for thecancer patient by increasing the efficacy of treatmentof the primary lesion. Cancer; 50:1227-1234, 1982.

62

FOURTH BEAM PERIOD OF PROTON RADIATION THERAPYUSING THE SPOTSCANNING GANTRY

G. Goitein, L. Wisser, A. Lomax, and Team Radiation Medicine

NEW PATIENTS

During the beam period of the year 2000 thirty-onenew patients underwent proton radiation therapy onthe PSI Spotscanning Gantry. 16 patients were male,15 female. Age ranged from 7 to 79 year with a meanof 46.9 and a median age of 50 year.

Five patients, (4 females, 1 male) presented with thediagnose of benign meningioma. Two of these lesionswere affecting the optic nerve and showed typicalcharacteristics of benign meningiomas in radiologicalstudies as well as in clinical course. Due to the local-ization, no histological verification had been per-formed. The remaining three meningiomas were lo-cated in the sphenoid region, the cavernous sinus andthe orbit and one postoperatively along the falx cere-bri. The latter case was complicated by previousgamma knife irradiation in 1994 and extensive lefthemispherical surgery in 1990. We applied 54 CGE(Cobalt Gray Equivalent) to all five meningiomas, fourfractions per week, and 2 CGE single dose. The le-sions required each careful design of beam anglesand spot weights in order to achieve appropriate pro-tection of the adjacent optic apparatus and sensitivebrain structures. The pre-irradiated patient wastreated with compromised dose in the region of thegamma knife radiation - according to the existingdocumentation, which resulted in a substantial doseinhomogeneity in parts of the target volume.

Six patients, one of whom 7 years, one 19 years old(2 females, 4 males), were treated for sarcomas out-side the skull base or spine. One myxoid liposarcomain the area of the apex of the left lung, infiltrating thebrachial plexus, the subclavian artery and vessels ofthe neck by extending into the paravertebral region,had been incompletely resected prior to radiation.Radiotherapy was performed as combination of pho-tons up to 40 Gy and protons up to 30 CGE, addingup to a total target dose of 70 CGE. The combinationof the two radiation modalities was planned from thebeginning in order to compensate for the motion of thetarget and normal tissues in the area of the upperthorax. Three other adult patients received combinedphoton and proton therapy: One large pleomorphicleiomyosarcoma of the left iliac muscle had beentreated with surgery and chemotherapy prior to radia-tion with 50.4 Gy photons. Additional 19.8 CGE pro-tons have been applied with excellent sparing of theintestinal and pelvic normal tissues. One patient suf-fered from a giant chondrosarcoma extending fromthe cervical region into the thorax. After incompletesurgery prior to radiation therapy, we could only apply

16 CGE protons due to the onset of the yearly shut-down at PSI. A third patient received irradiation afterincomplete surgical excision of a pelvic chondrosar-coma. The intervention had included extensive osteo-

synthetic procedures to stabilize the pelvic girdle,using titanium screws and plates. The postoperativeradiotherapy was again a combined approach usingprotons and photons. The completion of radiationtherapy has been planned and calculated - as in thecases described before - in close collaboration withthe referring and co-operating radiation oncologists,members of the Swiss PROTON USERS GROUP SPUG.One 19-year-old patient diagnosed with rhabdomyo-sarcoma of the ethnocide and right orbital region re-ceived 8 x 2 CGE protons in the framework of com-bined chemo-radiation therapy for juvenile patients.The radiation part was again completed with photons.The youngest patient, a 7-year-old boy, suffered froman extensive rhabdomyosarcoma in the left infra-temporal region destroying part of the facial bones.The child was treated according to the MMT-95 Studywith combined chemo-radiotherapy. After some firstfractions with photons, the boy was referred to protontherapy in order to take advantage of the superiorsparing of normal tissues - the very important aspectfor the future life of the patient. The treatment had tobe interrupted twice for two and one weeks respec-tively due to severe mucositis and febrile infection inconnection with the cycles of chemotherapy. At theend of proton irradiation, the patient had only mildintraoral mucositis in the left part of the maxilla, whichwas part off the target volume, and slight grade 1 skinreaction with hyperemia on the left cheek.

Three patients presented with lesions of the head andneck: One 15-year-old girl with a cancer of the naso-pharynx with involvement of the nuchal lymph nodes.The patient was treated with combined chemo-radiotherapy, the radiation part administered with 50Gy photons to a volume including the neck nodes and14 CGE protons to the primary lesion. One patientreceived a combination of 40 Gy photons and 26 CGEprotons for a carcinoma of the nasal cavity. The thirdpatient in this group suffered from an esthesioneuro-blastoma, involving the right ethmoid region and orbit.After several surgical interventions since 1992, thetumor had only slightly been debulked. According tothe recommendations of the group at the Mass. Gen-eral Hospital in Boston, where these tumors are partof the proton program, the patient received chemo-therapy prior to proton therapy with 68 CGE.

Chordomas and chondrosarcomas of the base ofskull, spine and sacrum formed the largest group oftumors in the year 2000. Six patients were irradiatedfor lesions in the base of skull, one chondrosarcoma,receiving 70 CGE, and five chordomas, which weretreated up to 74 CGE. In all cases, the prescribeddose was administered using a shrinking volumetechnique in order to respect the - for these tumorsestablished - normal tissue constraints of the surface

63

and center of the brain stem and spinal cord of 65 and55 CGE respectively as well as of the optic chiasmand nerves of maximally 62 CGE. This technique of-ten results in a certain underdosage in part of thetarget volume adjacent to the critical structure, whichrequires the reduction of the total dose. Four patientshad chordomas of the spine, all of which had beenoperated on one or several times with tumor extirpa-tion and stabilization of the spinal column. This stabili-zation was done in all but one cases using titaniumimplants plus Palacos® or autologous bone implant.Our dose calculation takes these materials into ac-count, though the strong differences in density aredemanding with respect to beam angles, spot weightsand the aspects of safety of the dose deposition.Again, shrinking volume technique and one or morereplannings are standard in such situations. The fifthpatient, an 11-year-old girl, had no metal implant,instead only an autologous tibia graft to stabilize thecervical spine after resection of parts of C3 and C4.The vitality of the graft was known to be at risk for infuture, but careful follow up may lead to another inter-vention to stabilize the cervical spine in case of ne-cessity or risk of instability. Four patients receivedproton therapy for sacral chordomas. In two cases, R1resection had been performed; a proton dose of 74CGE could be applied while protecting the largest partof the rectal circumference. One patient suffered froma local relapse after surgery two years ago, combinedproton - photon radiation was given with a protondose of 24 CGE. The fourth patient had been judgedinoperable at the occasion of open biopsy. This gen-tleman had had previous photon radiation up to 54 Gyto shrinking volumes, based on only radiological find-ings, before the histological verification had been un-dertaken. The proton dose had to be accordingly lowwith 19.8 CGE.

Three patients received palliative proton beam ther-apy: In one case of a large local recurrence of a rectalcancer, involving and in part destroying the sacralbones, even a low dose photon irradiation would havebeen substantially risky due to several accompanyingdiseases and postoperative complications like perito-nitis and sepsis. After 55.8 CGE and chemotherapywith 5-FU, the patient was mobilized again. One pa-tient with a singular metastasis of a thyroid cancer inthe right occipital bone and base of skull region wasirradiated with 70 CGE. The treatment was plannedwithin the overall concept with radioiodine treatments,assuming that the survival time and the excellentphysical status of this young mother of two childrenshould be maintained over a long period. Ophthal-mologists, who referred a patient with an orbital me-tastasis of a malignant melanoma, initiated the thirdpalliative irradiation. After incomplete resection of thislesion, consolidation of the local manifestation wasaimed at, though the total dose had to be limited at 60CGE due to the tolerance of the intraorbital structures.In addition, a lesion in the ipsilateral cavernous sinuswas treated with 54 CGE, assuming the histology of ameningioma.

All proton treatments were performed using our 3-dtreatment-planning program. Normal tissue con-straints were taken care of by applying multiple beamangles, variable spot weights, individual adaptation ofselected volumes ("technical volumes") and shrinkingvolume techniques with two or more individual plans.In case of strong tissue inhomogeneities, we usedadditional Monte Carlo calculations in order to avoidinaccuracies in the application of high therapeuticaldoses. All patients were immobilized with individualwhole body molds, vacuum bite blocks for treatmentsin the whole head and neck region or hard body-conform covers, made out of thermo-elastic material,in cases of irradiation of truncal lesions. Daily position-ing controls were performed using the CT- scoutviews and reference slices as well as X-rays on thegantry on selected days. Except one (recurrence ofrectal cancer), all patients had ambulatory treatments.We saw no unexpected toxicity during the courses ofproton therapy. Normal, dose-adequate reactions likee.g. dryness of oral mucosa, skin reaction, are notjudged as complication and were treated symptomati-cally, if necessary.

FOLLOW UP

The follow up reports of the patients treated until theend of 1999 with protons have been collected. Of the41 patients, 32 had received curative proton irradia-tion: 10 meningiomas, 7 lesions of the brain tissues, 8chordomas and chondrosarcomas of the skull baseand spine, 4 sarcomas of the trunk and extremities, 1prostate cancer, 1 carcinoma of the epipharynx, 1esthesioneuroblastoma (re-irradiation). One of themeningioma cases had been special with four lesionsof atypical histology. After a well tolerated conformaltreatment, one of the manifestations showed earlycentral tumor necrosis with extensive surroundingedema. This was the reason for necrosectomy. Unfor-tunately, the patient suffered postoperatively frominternal medical complications, which finally causedfatal outcome. One patient with an olfactory men-ingioma, causing severe reduction in visual acuity,reports 15 months after therapy almost normalizationon one eye and improvement of visual acuity on theother side. Of the 7 patients with brain tumors, threehad presented with focal necroses within the highdose region outside the lesion. Two were sympto-matic, requiring long term steroid medication in onecase, short term treatment in the other. The third pa-tient was and is obviously asymptomatic. One patientsuffers from a histologically verified local relapse 18months after treatment for an astrocytoma II. Notreatment related complication or failure has beenreported to date from the other groups of tumors. Atthe end of 2000, thirty-one of the 32 curatively irradi-ated patients are alive, 1 2 - 4 2 months after protontherapy, with local tumor control in 31 cases.

We would like to mention the excellent co-operationwith the Swiss PROTON USERS GROUP SPUG, whocontributes to the success of the proton program.

64

TULOC: A NOVEL TECHNIQUE FOR THE TRACKING OF TUMOR POSITIONS INCONFORMAL RADIOTHERAPY

THE TULOC METHOD WHOSE NAME STANDS FOR "TUMOR LOCATION" IS FILED FOR PATENT UNDERPCT/CH97/00132 AND WO97/36192 AND IS BEING ASSIGNED TO THE MEDNETIX AG,

A SPIN-OFF COMPANY OF PSI

R.K. Munch, P.G. Seiler, J. Verwey

Tumor movements induced by organ motion (respiration, circulation and peristalsis) can impede the fullexploitation of precision radiotherapy. We are developing a novel technique for the magnetic tracking oftumor positions. This technique shall allow the real time adaptation of the irradiation pattern to the positionand shape of the tumor during conformal radiotherapy. Eddy currents in near by electrically conducting objectsdistort our position measurements. We have developed a method to correct for that effect.

The TULOC method is a novel magnetic trackingtechnique described elsewhere (1,2). For the trackingof miniaturized implantable sensors, we usealternating magnetic fields that are slowly modulatedwith a frequency of about 10 kHz. Such fields induceeddy currents in near by electrically conductingobjects. The eddy currents, leading to distortions ofthe original field, falsify the position measurement. Wehave developed a method to correct for those eddycurrent effects and demonstrated the correctness ofthe method on various example setups in thelaboratory. The method to compute the eddy currentshas been filed for patent.

Fig. 1 shows an experimental setup chosen to test theeddy current calculations for a copper disc. The pointsmarked 1 to 125 are the positions in which the sensorwas placed sequentially.

I f' •

Held generator

Measure volume'

Copper disc •

14

12-

— 10-EE s-l

with field correction

20 40 60 80 100

Sensor position number120

Fig. 2 shows the positions determined by themagnetic tracking system without and with taking theeffect of eddy currents into account. Note theconsiderable reduction of the systematic error!

We are now adapting the correction calculations to thereal geometry of selected irradiation facilities.

REFERENCES

1. Seiler, P.G. et al TULOC: A NOVEL TECHNIQUE, PSIScientific Report 1999 Volume II, Life Sciences.

2. Seiler, P.G. et al., A novel tracking technique for thecontinuos precise measurement of tumor positions inconformal radiotherapy, Phys. Med. Biol. 45 N103-N110,2000.

Fig. 1: Schematic setup to compare eddy currentcalculations with measurements for a copperdisc. The sensor is positioned sequentiallyinto the position 1 through to 125.

65

VI LIST OF PUBLICATIONS

STRUCTURAL BIOLOGY

H.-J. Boehm, M. Boehringer, D. Bur, H. Gmuender,W. Huber, W. Klaus, D. Kostrewa, H. Kuehne, T. Luebbers,N. Meunier-Keller, F. Mueller "Novel inhibitors of DNA gy-rase: 3D structure based biased needle screening, hit vali-dation by biophysical methods, and 3D guided optimization.A promising alternative to random screening." J. Med.Chem., 43, 2664-2674 (2000).

P. Burkhard, R.A. Kammerer, M.O. Steinmetz, U. Aebi. "Thecoiled-coil trigger site of the rod domain of cortexillin I un-veils a distinct network of interhelical and intrahelical saltbridges". Structure, 8, 223-230 (2000).

E.M. De La Cruz, A. Mandinova, M.O. Steinmetz,D. Stoffler, U. Aebi, T.D. Pollard "Nucleotide dependence ofactin polymerization, structure and dynamics". J. Mol. Biol.,295,517-526(2000).

R. Grisshammer, C. Kambach, C.G. Tate "Expression sys-tems". In: DNA binding proteins - a practical approach, eds:A. Travers and M. Buckle, IRL press, Oxford, pp.1-24(2000).

M. Lehmann, D. Kostrewa, M. Wyss, R. Brugger, A. D'Arcy,L. Pasamontes, A.P.G.M. van Loon. "From DNA sequenceto improved functionality: using protein sequence compari-sons to rapidly design a thermostable consensus phytase".Protein Engineering, 13, 49-57 (2000).

B. Masjost, P. Ballmer, E. Borroni, G. Zurcher, F.K. Winkler,R. Jakob-Roetne, F. Diederich "Structure based design,synthesis, and in vitro evaluation of bisubstrate inhibitors forcatechol O-methyltransferase (COMT)". Chemistry,. 6, 971-82 (2000).

C. Oefner, A. D'Arcy., M. Hennig, F.K. Winkler, G.E. Dale"Structure of human neutral endopeptidase (neprilysin)complexed with phosphoramidon". J. Mol. Biol., 296, 341-349 (2000).

M.O. Steinmetz, A. Hoenger, D. Stoffler, A.A. Noegel,U. Aebi, C.-A. Schoenenberger "Polymerization, Three-dimensional structure and mechanical properties of dictyos-telium versus rabbit muscle actin filaments". J. Mol. Biol.,303, 171-184(2000).

M.O. Steinmetz, R.A. Kammerer, W. Jahnke, K.N. Goldie,A. Lustig, J. van Oostrum "Stathmin/Op18 caps a kinkedprotofilament-like tubulin tetramer". EMBO J., 19, 572-580(2000).

D.J. Thiel, M.-H. le Du, R.L. Walter, A. D'Arcy, C. Chene,M. Fountoulakis, G. Garotta, F.K. Winkler, S.E. Ealick "Ob-servation of an unexpected third receptor molecule in thecrystal structure of human interferony-receptor complex".Structure with Folding and Design, 8, 927-936 (2000).

A. Tomschy, M. Wyss, D. Kostrewa, K. Vogel, M. Tessier,S. Hbfer, H. Burgin, A. Kronenberger, R. Remy, A.P.G.M.van Loon, L. Pasamontes "Active site residue 297 of Asper-gillus niger phytase critically affects the catalytic properties"FEBS Lett., 472, 169-172 (2000).

F. Wohnsland, A.A.P. Schmitz, M.O. Steinmetz U. Aebi,G. Vergeres "Influence of the effector peptide of MARCKS-related protein on actin polymerization: a kinetic analysis"Biophys. Chem., 85, 169-177 (2000).

F. Wohnsland, A.A.P. Schmitz, M.O. Steinmetz, U. Aebi,G. Vergeres "Interaction between actin and the effectorpeptide of MARCKS-related protein (MRP): Identification offunctional amino acid segments" J. Biol. Chem., 275,20873-20879 (2000).

F. Wohnsland, M.O. Steinmetz, U, Aebi, G. Vergeres."MARCKS-related protein (MRP) binds to actin withoutsignificantly affecting actin polymerization or network struc-ture" J. Struct. Biol., 131, 217-224 (2000).

INSTITUTE OF MEDICAL RADIOBIOLOGY

R.M.C. Cattaneo-Pangrazzi, H. Schott, H. Wunderli-Allenspach, M.I. Derighetti, R.A. Schwendener "New am-phiphilic heterodinucleoside phosphate dimers of 5-Fluorodeoxyuridine (5FdUrd): Cell cycle dependent cytotox-icity and induction of apoptosis in PC-3 prostate tumorcells". Biochem. Pharmacol, (in press, 2000).

R.M.C. Cattaneo-Pangrazzi, H. Schott, H. Wunderli-Allenspach, B. Rothen-Rutishauser, M. Guenthert, R.A.Schwendener "Cell cycle arrest and p53-independent induc-tion of apoptosis by the new anticancer drugs 5-FdUrd-5-FdC18 and dCpam-5-FdUrd in DU-145 human prostatecancer cells". J. Cancer Res. Clin. Oncol. 126: 247-256(2000).

R.M.C. Cattaneo-Pangrazzi, H. Schott, H. Wunderli-Allenspach, M.I. Derighetti, R.A. Schwendener "The novelheterodinucleoside dimer 5-FdU-NOAC is a potent cytotoxicdrug and a p53-independent inducer of apoptosis in theandrogen-independent human prostate cancer cell linesPC-3 and DU-145". The Prostate 45:8-18 (2000).

J.A. Fresno Vara, M.V. Carretero, H. Geronimo, K. Ballmer-Hofer, J. Martin Perez "Stimulation of c-Src by prolactin isindependent of Jak2". Biochem J. 345, 17-24 (2000).

S. Hor, A. Ensser, C. Reiss, K. Ballmer-Hofer, C. Biesinger"Herpesvirus saimiri protein StpB associates with cellularSrc". J. Gen. Virol. (In press).

D.H. Horber, R.M.C. Cattaneo-Pangrazzi, P. von Ballmoos,H. Schott, P.S. Ludwig, S. Eriksson, I. Fichtner, R.A.Schwendener "Cytotoxicity, cell cycle perturbations andapoptosis in human tumor cells by lipophilic N4-alkyl-1- -D-arabinofuranosylcytosine derivatives and the new heteronu-cleoside phosphate dimer arabinocytidylyl-(5'Y5')-N4-octadecyl-1- y-D-ara-C". J. Cancer Res. Clin Oncol. 126:311-319(2000).

B. Ludewig, F. Barchiesi, M. Pericin, R.M. Zinkernagel,H. Hengartner, R.A. Schwendener "In vivo antigen loadingand activation of dendritic cells via a liposomal peptidevaccine mediated protective antiviral and antitumor immu-nity". Vaccine 19: 23-32 (2000).

C. Marty, K. Ballmer-Hofer, D. Neri, R. Klemenz, H. Schott,R.A. Schwendener "Inhibition of tumor growth by specifictargeting of anti-ED-B fibronectin scFv antibody modifiedliposomes in the F9 mouse teratocarcinoma model". Can-cer. Res. (in preparation, 2000).

C. Marty, P. Scheidegger, K. Ballmer-Hofer, R. Klemenz,R.A. Schwendener "Production of functionalized single-chain Fv antibody fragments binding to the EDB domain ofthe B-isoform of fibronectin" in Pichia pastoris.

66

B. Roscic-Mrkic, R.A. Schwendener, B. Odermatt,A. Zuniga, J. Pavlovic, M.A. Billeter, R. Cattaneo "The roleof macrophages in measles virus infection of geneticallymodified mice". J. Virol, (submitted, 2000).

P. Scheidegger, W. Weiglhofer, S. Suarez, S. Console,J. Waltenberger, M. Pepper, R. Jaussi, K. Ballmer-Hofer"Signalling properties of a human immunodeficiency virus-encoded angiogenic peptide mimicking vascular endothelialgrowth factor activity". Biochem. J. (In press).

R.A. Schwendener, K. Feil, H. Depenbrock, H. Schott,A.-R. Hanauske "In vitro activity of liposomal N4-octadecyl-i-y-D-arabinofuranosyl-cytosine(NOAC), a new lipophilicderivative of 1-y-D-arabinofuranocylcvtosine on biopsizedclonogenic human tumour cells and haematopoietic precur-sor cells". Invest. New Drugs (in press).

S. Suarez, K. Ballmer-Hofer "VEGF transiently disrupts gapjunctional communication in endothelial cells". J. Cell. Sci.114:1229-1235 (2001).

CENTER FOR RADIOPHARMACEUTICALSCIENCE

S.M. Ametamey, M. Kokic, N. Carrel-Remy, P. Blauenstein,M. Willmann, S. Bischoff, M. Schmutz, P.A. Schubiger,Y.P. Auberson "Synthesis, Radiolabelling and BiologicalCharacterization of (D)-7-lodo-A/-(1-phosphonoethyl)-5-aminomethylquinoxaline-2,3-dione, a Glycine-Binding SiteAntagonist of NMDA Receptors" Bio. Organic and MedicinalChem. Letters, 10, 75-78 (2000).

S.M. Ametamey, G. Westera, P.M Gucker,R. Schonbachler,M. Honer, J.E. Spang, P.A. Schubiger "Functional BrainReceptor Imaging with Positron Emission Tomography"Chimia 54, 622-626 (2000).

Blasberg RG, Roelcke U, Beattie B, von Ammon K, Yone-kawa Y, Landolt H, Guenther I, Crompton NEA, Vontobel P,Weinreich R, Maguire RP, Missimer J, Knust E, Finn RD,Leenders KL "Imaging Brain Tumor Proliferative Activitywith [124l]-lododeoxyuridine" Cancer Research 60, 624-635(2000).

R. Boni, P. Blauenstein; R. Dummer, G.K. von Schulthess,P.A. Schubiger, H.C. Steinert "Non-invasive assessment oftumour cell proliferation with positron emission tomographyand [76Br]bromodeoxyuridine" Melanoma-Res. 9, 569-73(1999).

M. Bruehlmeier, K.L. Leenders, P. Vontobel, C. Calonder,A. Antonini, A. Weindl "Increased cerebral iron uptake inWilson's disease: A 52Fe-citrate PET study" J Nucl Med;47,781-787(2000).

A. Buck, P.M. Gucker, R. Schonbachler, M. Arigoni,S. Kneifel, F.X. Vollenweider, S.M. Ametamey, C. Burger"Evaluation of Serotonergic Transporters using PET and [11C]-(+)McN-5652: Assessment of methods" J. Cer. BloodFlow Metabol. 20, 253-62 (2000).

L. Curtis, F. Chiodini, J.E. Spang, S. Bertrand, J.T. Patt,,G. Westera, D. Bertrand "A new look at the nicotinic acetyl-choline receptor phamacophore" Eur. J. Pharmacology 393,155-163(2000).

O.D.M. Hughes, M.C. Bishop, A.C. Perkins, M.L. Wastie,G. Denton, M.R. Price, M. Frier, H. Denley., R. Rutherford,P.A. Schubiger "Targeting Superficial Bladder Cancer bythe Intravesical Administration of Copper-67-Labeled Anti-MUC1 Mucin Monoclonal Antibody C595" J of Clinical On-cology, 18(2), 363-370 (2000).

G. Kiinig, K.L. Leenders, C. Martin-Soelch, J. Missimer,S. Magyar, W. Schultz "Reward processing in the brains ofParkinsonian patients". Neuro Report 11 3681-3687 (2000).

J.T. Patt, J.E. Spang, G. Westera, P.A. Schubiger "[C-11]N-methylhomoepibatidine: Radiolabelling and biodistributionstudies in mice" J. Labelled Cpd. Radiopharm. 43, 127-136(2000).

H.J. Pietzsch, A. Gupta, M. Reisgys, A. Drews, S. Seifert,R. Syhre, H. Spess, R. Alberto, U. Abram, P.A. Schubiger,B. Johannsen "Chemical and Biological Characterization ofTechnetium (I) and Rhenium (I) Tricarbonyl Complexes withDithioether Ligands Serving as Linkers for Coupling theTc(CO)3 and Re(CO)3 Moieties to Biologically Active Mol-

cules" Bioconj Chem 11(3), 414-424 (2000).

R. Schibli, R. La Bella, R. Alberto, E. Garcia Garayoa,O. Kirstin, U. Abram, P.A. Schubiger "Influence of the Den-ticity of Ligand Systems on the in Vitro and in Vivo Behav-iour of 99mTc(l)-Tricarbonyl Complexes: A Hint for the Fu-ture Functionalization of Biomolecules" Bioconjugate Chem-istry, 11(3), 345-451 (2000).

J.E. Spang, J.T. Patt, G. Westera, P.A. Schubiger, "Com-

parision of N-[^C]Methyl-Norchloroepibatidine and N-

[11C]Methyl-2(2-pyridyl)-7-azabicyclo [2.2.1]heptane with N-

[11C]Methyl-epibatidine in Small Animals PET Studies"Nucl. Med. Biol., 27, 239-247 (2000).

J.E. Spang, S. Bertrand, G. Westera, J.T. Patt,P.A. Schubiger, D. Bertrand "Chemical modification of epi-batidine causes a switch from agonist to antagonist andmodifies its selectivity for neuronal nicotinic acetylcholinereceptors" Chemistry & Biology, 7(7), 545-555 (2000).

R. Waibel, I. Novak-Hofer, R. Schibli, P. Blauenstein,E. Garcia Garayoa, R. Schwarzbach, K. Zimmermann,R. Pelllikka, O. Gasser, A. Blanc, M. Bruhlmeier,P.A. Schubiger "Radiopharmaceuticals for Targeted Tumor-diagnosis and Therapy" Chimia 54, 683-689 (2000).

G. Westera, P.A. Schubiger "Clinical Radiopharmacy inClinical Positron Emission Tomography (PET). Correlationwith Morphological Cross-Sectional Imaging" G.K. vonSchulthess Editor, by Lippincott, Wlliams and Wilkins(Wolters Kluwer), Philadelphia (2000).

R. Zurbriggen, I. Novak-Hofer, A. Seelig, R. Gluck "IRIV-adjuvanted hepatitis A vaccine: in vivo absorption and bio-physical characterization" Prog Lipid Res 39, 3-18 (2000).

RADIATION MEDICINE

R.G. Blasberg, U. Roelcke, B. Beattie, K. von Ammon,Y. Yonekawa, H. Landolt, I. Guenther, N.E.A. Crompton, P.Vontobel, R. Weinreich, R.P. Maguire, E.J. Knust, R.D.Finn, K.L. Leenders "Imaging brain tumor proliferative activ-ity with 124-l-iododeoxyuridine". Cancer Res., 60, 624-635(2000).

S.N. Boon, P. van Luijk, T. Bohringer, A. Coray, A.J. Lomax,E. Pedroni, B. Schaffner J.M. Schippers, "Performance of afluorescent screen and CCD camera as a 2D dosimetry-system for dynamic treatment techniques" Med Phys., 27:2198-2208(2000).

M. Bucciolini, G. Cuttone, A. Guasti, S. Lo Nigro, S. Maz-zocchi, L. Raffaele, M.G. Sabini, A. Kacperek, E. Egger,"Check on the use of thermoluminescence detectors forproton dose distribution measurements", Physica MedicaXVI(3) 117-120(2000).

67

N.E.A. Crompton "Das wie und warum unseres Alterns, mitStreiflichtern zur Strahlenforschung". Schweiz. Krebs Bulle-tin, 20, 92-95 (2000).

N.E.A. Crompton, G.C. Emery, M. Ozsahin "Individual pa-tient radiosensitivity". Proc. of the Journées Scientifiques dela Société Suisse de Radiobiologie et Physique Médicale,Suisse, 1996, pp. 87-92. Ed.: J. Roth, H.W. Nemec, H.W.Roser, and B. Schneckenburger. Verlag A. Schudel Co. AG,Riehen, Switzerland (2000). ISBN 3-85895-005-X.

N.E.A., Crompton, Y.Q. Shi, G.C. Emery, H. Blattmann"Very high frequencies of heritable damage result fromexposure to even low doses of X-rays". Proc. of the Jour-nées Scientifiques de la Société Suisse de Radiobiologie etPhysique Médicale, Suisse, 2000, pp. 131-137. Ed.: J.Roth, H.W. Nemec, H.W. Roser, and B. Schneckenburger.Verlag A. Schudel Co. AG, Riehen, Switzerland (2000).ISBN 3-85895-004-1.

N.E.A. Crompton, J. Gobrecht, C. Higgs, G. Kuehne,E. Lehmann, H. Reist, J. Stepanek, T. Suda, H. Blattmann,B. Larsson "CASE-MATE: Computer-aided cell seeding,microcloning, analysis and telemetric evaluation applied toneutron capture therapy". In: Frontiers in Neutron CaptureTherapy, Eds: M.F. Hawthorne, R.J. Wiersma and K. Shel-ley. Plenum Publ. (2000).

N.E.A. Crompton, H. Walt, S.B. Kahl, S.D. Klein, Y.Q. Shi,H. Blattmann, U. Haller, B. Larsson "Neutron capture ther-apy (BNCT) and photodynamic therapy (PDT) can both beaccomplished by a single compound: boronated protopor-phyrin (BOPP)". In: Frontiers in Neutron Capture Therapy,Eds: M.F. Hawthorne, R.J. Wiersma and K. Shelley. Ple-num Publ. (2000).

E. Egger, L. Zografos, A. Schalenbourg, G. Goitein "Varioustherapeutic regimens for proton beam irradiation of neovas-cular membranes in age-related macular degeneration" inMEDICAL RADIOLOGY - Radiation Oncology, Volume'Age-related macular degeneration, current treatment con-cepts', Springer Verlag, 181-186 (2001).

G. Kühne, N.E.A Crompton, C Higgs, H. Reist, E. Lehmann,J. Stepanek, T Suda, H. Blattmann, B. Larsson "A CompactCell Culture Disc for Neutron Capture Radiography". In:Frontiers in Neutron Capture Therapy, Eds: M.F. Haw-thorne, R.J. Wiersma and K. Shelley. Plenum Publ. (2000).

B. Larsson, N.E.A. Crompton "Monochromatic X-rays, slowneutrons and radionuclides for targetted radiotherapy -present developments and prospects". In: Frontiers in Neu-tron Capture Therapy, Eds: M.F. Hawthorne, R.J. Wiersmaand K. Shelley. Plenum Publ. (2000).

A.J. Lomax Comment on "Intensity-modulated conformairadiation therapy and three-dimensional treatment planningwill significantly reduce the need for therapeutic approacheswith particles such as protons" , Med. Phys., 27: 622-623(2000).

A.J. Lomax, M. Grossmann, L. Cozzi, P.A. Tercier, T. Boe-hringer, U. Schneider, M. Logean, W. Volken, O. Ratib, R.Miralbell "The exchange of radiotherapy data as part of anelectronic patient-referral system". Int. J. Radiât. Oncol.,Biol., Phys. 47: 1449-1456 (2000).

I. Mader, W. Roser, L. Kappos, G. Hagberg, J. Seelig,E.W. Radue, W. Steinbrich, "Serial proton MR spectroscopyof contrast-enhancing multiple sclerosis plaques: absolutemetabolic values over 2 years during a clinical pharmacol-ogical study". AJNR Am. J. Neuroradiol. 21, 1220-1227(2000).

R. Miralbell, L. Celia, D. Webe, A.J. Lomax "Optimizingradiotherapy of orbital and paraorbital tumors: intensitymodulated X-ray beams versus intensity modulated protonbeams". Int. J. Radiât. Oncol., Biol., Phys., 47: 1111-1119(2000).

F. de Pasquale, G. Sebastiani, E. Egger, L. Guidoni,A.M. Luciani, P. Marzola, R. Manfredi, M. Pacilio, A. Pier-mattei, V. Viti, P. Barone, "Bayesian estimation of relaxationtime Ti in MR images of irradiated Fricke-agarose gels",Magnetic Resonance Imaging 18: 721-731 (2000).

S.N. Boon et. al, Groningen and PSI, "Performance of afluorescent screen and CCD camera as a two-dimensionaldosimetry system for dynamic treatment techniques", Med.Phys., 27: No 10, October 2000.

U. Schneider, A.J. Lomax, N. Lombriser "Calculation ofsecondary cancer incidence following photon and protonradiation treatment of Hodgkin's disease." Rad. Res., 154:382 (2000).

PM. Schweizer, P. Spanne, M. Di Michiel, U. Jauch,H. Blattmann, JA. Laissue "Tissue lesions caused by mi-croplanar beams of synchrotron-generated X-rays in Droso-phila melanogaster". International Journal of RadiationBiology. 76: 567-574, 2000.

J. Stepanek, H. Blattmann, JA. Laissue, N. Lyubimova,M. Di Michiel, DN. Slatkin "Physics study of microbeamradiation therapy with PSI version of Monte Carlo codeGEANT as a new computational tool". Medical Physics. 27:1664-1675,2000.

W. Thomlinson, P. Berkvens, G. Berruyer, B. Bertrand,H. Blattmann, E. Brauer-Krisch, T. Brochard, AM. Charvet,S. Corde, M. Dimichiel, H. Elleaume, F. Esteve; S. Fiedler;JA. Laissue, JF. Le-Bas, G. Le-Duc, N. Lyubimova, C. Ne-moz, M. Renier, DN. Slatkin, P. Spanne, P. Suortti "Re-search at the European Synchrotron Radiation Facilitymedical beamline". Cellular and Molecular Biology 46.1053-1063(2000).

S.O. Troja, E. Egger, P. Francescon, A.M. Gueli,A. Kacperek, M. Coco, R. Musmeci, A. Pedalino "2D and 3Ddose distribution determination in proton beam radiotherapywith GafChromic film detectors", Technol Health Care 8(2)155-64(2000).

A. Zurlo, A.J. Lomax, A. Hoess, T. Bortfeld, M. Russo,G. Goitein, V. Valentini, L. Marucci, R. Capparella, A.Loasses "The role of proton therapy in the treatment oflarge irradiation volumes: A comparative planning study ofpancreatic and biliary tumours". Int. J. Radiât. Oncol., Biol.,Phys., 48: 277-288 (2000).

TULOC

PG. Seiler, H. Blattmann, S. Kirsch, RK. Münch,Ch. Schilling "A novel tracking technique for the continuousprecise measurement of tumor positions in conformairadiotherapy", Phys. Med. Biol. 45(2000) N103-N110.

68

INVITED TALKS, CONTRIBUTIONS TO CONFERENCES AND WORKSHOPS

STRUCTURAL BIOLOGY

C. Kambach "Socialise your proteins and make themhappy!- Coexpression as a tool to improve and study pro-tein behaviour." 19th European Crystallographic Meeting,Nancy, France, August 27, 2000.

D. Kostrewa, Co-chairman in the Microsymposium "Phasingand Structure Solution" 19th European CrystallographicMeeting in Nancy, August 25-31, 2000.

M.O. Steinmetz "Mechanism of action of Op18/stathmin ontubulin and Microtubules" INSERM, Institut du Fer a Moulin,Paris, France, January 21, 2000.

M.O. Steinmetz "Mechanism of action of Op18/stathmin ontubulin and Microtubules" UC Berkeley, Molecular and CellBiology Department, Berkeley, CA, USA, February 22,2000.

M.O. Steinmetz "Mechanism of action of Op18/stathmin ontubulin and Microtubules" European Molecular BiologyLaboratory, Structural and Computational Biology Pro-gramme, Heidelberg, Germany, May 23, 2000.

M.O. Steinmetz "Mechanism of action of Op18/stathmin ontubulin and Microtubules" 15th meeting of the EuropeanCytoskeleton Forum, Blankenberge, Belgium, August 27,2000.

M.O. Steinmetz "Mechanism of action of Op18/stathmin ontubulin and Microtubules" 14th Regio Meeting in StructuralBiology, Braunwald, Switzerland, September 21, 2000.

F.K. Winkler "The use of SR to investigate biological mole-cules of industrial interest" Hercules X Euroconference,Grenoble, France, April 7, 2000.

F.K. Winkler "Structural studies on p-lactam sensitive andresistant bacterial transpeptidases", Institute of Cell Biology,ETH Zurich, May 26, 2000.

F.K. Winkler "Structure-based approaches in modern drugdiscovery research", Ernst Schering Research FoundationWorkshop. 'Data mining in Structural Biology: Signal Trans-duction and beyond', Berlin, Germany, June 30, 2000.

INSTITUTE OF MEDICAL RADIOBIOLOGY

K. Ballmer-Hofer "Binding of HIV Tat-derived angiogenicpeptide to VEGF receptors 1 and 2; implications for recep-tor function" Novartis Pharma, June 8, 2000.

R. Schwendener "Liposomes as peptide and DNA vaccines:In vivo antigen delivery and activation of dendritic cellsprovides protective antiviral immunity" Klinik fur Rheuma-tologie und Klinische Immunologie und Allergologie. No-vember 2, 2000.

R. Jaussi "The basic region of the HIV-1 Tat protein andtumor targeting" Dermatologische Klinik UNI-Spital Zurich,November 29, 2000.

C. Chastel, S. Suarez, A. Maeder, K. Ballmer-Hofer "Inhibi-tion of radiation-induced apoptosis by Glutathione-S-Transferase or PI 3-kinase-activated serine/threoninekinases" Wolfsberg meeting, Wolfsberg 2000.

K. Ballmer-Hofer, P. Scheideger, W. Weiglhofer, R. Jaussi"Binding of an HIV Tat-derived angiogenic peptide to VEGFreceptors 1 and 2: identification of new receptor bindingsites" Meeting on vascular biology and medicine, Numberg1999.

K. Ballmer-Hofer, P. Scheideger, A. Demirovic,W. Weiglhofer, R. Jaussi "Binding of an HIV Tat-derivedangiogenic peptide to VEGF receptors 1 and 2" EMBLmeeting on cell signalling, Heidelberg 2000.

K. Ballmer-Hofer, P. Scheideger, A. Demirovic,W. Weiglhofer, R. Jaussi "Binding of an HIV Tat-derivedangiogenic peptide to VEGF receptors 1 and 2"ELSO/USGEB meeting, Geneva 2000.

P. Scheidegger, A. Demirovic, W. Weiglhofer, S. Suarez,R. Jaussi, K. Ballmer-Hofer "HIV-1 TAT-derived AngiogenicPeptides and VEGF Bind to Different Sites on VEGF Re-ceptors" Gordon Conference on Vascular Cell Biology,Plymouth State College 2000.

CENTER FOR RADIOPHARMACEUTICALSCIENCE

S.M. Ametamey "PET-radioligands: From development tohuman studies" invited lecture, University Irchel, Zurich,March 28, 2000.

S.M. Ametamey "PET activities at the Center for Radio-pharmaceutical Science" invited lecture, Philadelphia, USA,August 25, 2000.

S.M. Ametamey "PET in drug development" invited lecture,Novartis, Basel, Switzerland, December 15, 2000.

S.M. Ametamey "Effects of MDMA on McN-5652 binding to5-HT binding sites in humans" Gemeinsame Jahrestagungder Deutschen. Osterreichischen und SchweizerischenGesellschaft fur Nuklearmedizin, Munich, Germany, March29 - April 1, 2000.

P. Blauenstein, E. Garcia Garayoa, M. Bruhlmeier,R. Bugmann, N. Carrel-Remy, M. Willmann, R. La Bella,P.A. Schubiger, D. Tourve "Pharmacological Evaluation OFmetabolically stabilised Neurotensin Analogues labelledwith technetium-99m" 24th Iternational Symposium onadioactive Isotopes in Clinical Medicine and Research, BadGastein, Germany, January 11-14, 2000.

M. Bruehlmeier, P. Blauenstein, E. Garcia-Garaoya,

R. Bugman, P.A. Schubiger "99mTC-markierte Peptide:Fortschritte bei Neurotensinderivaten" 2. Radio-pharmazeutische Gesprachsrunde, Baden, Switzerland,January 21, 2000.

M. Bruehlmeier, P. Vontobel, T.H. Locher, K.L. Leenders."Changes in the cerebral iron metabolism in Wilson's dis-ease. A 52Fe-citrate PET study" 3-Lander Tagung derDGN, SGNM und OGN, Munich, Germany, March 29 - April1, 2000.

F. Feurer, E. Garcfa Garayoa, A. Blanc, R. Bugmann, R. LaBella, P. Blauenstein "Neurotensin Markierung ausgehend

vom 99mTc-Tricarbonyl-Triaqua Komplex" 8. Arbeitstreffender Arbeitsgemeinschaft Radiochemie, Radiopharmazie,Erne, Germany, October 5-7, 2000.

69

E.Garcia Garayoa, P. Bläuenstein, M. Brühlmeier,R. Bugmann, D. Tourwé, P.A. Schubiger "PharmacologicalProperties of new 99mTc-labelled neurotensin analogues"Nuc. Med. Comm. 21, 571, 2000.

M. Kokic, M. Honer, S.M. Ametamey, F. Gasparini,H. Andres, S. Bischoff, P.J. Flor, M. Heinrich, I. Vranesia,W. Spooren, R. Kuhn, P.A. Schubiger "Radiosynthesis andPET evaluation of [11C]M-MPEP in rats" ZNZ Symposium,Zurich, Switzerland, October 20, 2000.

M. Kokic "Synthesis, Radiolabelling and Evaluation of Glu-tamate Receptor Ligands for Positron Emission Tomogra-phy (PET)", Basel, October 24, 2000.

R. La Bella "Tumor targeting with 99mTc(l)-tricarbonyl la-beled bombesin analogues", ETH Zürich, Switzerland,March 31,2000.

1. Novak-Hofer, C.A. Hoefnagel, J. DeKraker, M. Rutgers,M. Meli, P.A. Schubiger "Complementarity of targeting ofneuroblastoma with MIBG and anti-L1-CAM mAb chCE7"Nuklearmedizin 2000, Munich, Germany, March 29—April 1,2000.

J. Petrig "Functionalization of Glucose for the labelling with99mTc(l)-tricarbonyl", ETH Zürich, Switzerland, March 31,2000.

R. Schönbächler, P.M. Gucker, S. Ametamey, A. Buck,M. Arigoni, S. Kneifel, C. Burger, T. Berthold,F.X. Vollenweider, P.A. Schubiger "Evaluation von [C-11]-ß-CPPIT als PET Radioligand für den Dopamin-Transporter"2. Radiopharmazeutisch-Nuklearmedizinische Gesprächs-runde, Baden, Switzerland, January 21, 2000.

R. Schönbächler, P.M. Gucker, S. Ametamey, A. Buck,M. Arigoni, S. Kneifel, C. Burger, T. Berthold,F.X. Vollenweider, P.A. Schubiger "Evaluation von [C-11]-ß-CPPIT als PET Radioligand für den Dopamin-Transporter"Nuklearmedizin 2000, Munich, Germany, March 29 - April1, 2000.

P.A. Schubiger "Vehicles, Chelators, and Radionuclides:choosing the "Building Blocks" of an Effective TherapeuticRadioimmunoconjugate" plenary lecture, Tumor TargetingSymposium ETH, Zurich, Switzerland, February 25, 2000.

P.A. Schubiger "Peptide zur Tumordiagnostik" Invited Lectu-re, Nuklearmedizin 2000 - Jahrestagung, Munich, Germa-ny, March 29 - April 1, 2000.

P.A. Schubiger "Glutamaterge Rezeptorliganden" InvitedSeminar, Forschungszentrum Julich, Germany, April 17,2000.

P.A. Schubiger, E. Garcia Garayoa, P. Bläuenstein "Designand radiolabelling of stable neurotensin dérivâtes" invitedlecture, 7th International Symposium on the Synthesis andApplication of Isotopes, Dresden, Germany, June 18-22,2000.

P.A. Schubiger "Targetted Tumortherapy with RadionuclideConjugates" invited seminar, Frédéric Joliot - Curie Insti-tute, Budapest, June 23, 2000.

P.A. Schubiger "Recent Progress with Stable labelling ofBiomolecules with Tc/Re-Carbonyles" invited lecture, 5th

International Conference on Nuclear and Radiochemistry,September 3-8, 2000.

A. Stichelberger "Modifications to Reduce Renal Uptake of99mjc Tricarbonyl Labeled Single Chain Antibody Frag-ments", Doktorandentag, Basel, Switzerland, October 24,2000.

R. Zurbriggen, I. Novak-Hofer, R. Glueck "Immunopoten-tiating reconstituted influenza virosomes (IRIV)-An efficientantigen carrier system for efficacious vaccination" 9th Inter-national Congress on Infections Diseases, Buenos Aires,Argentina, April, 2000

RADIATION MEDICINE

M. Biaggi, F. Ballarini, W. Burkard, E. Egger, A. Ferrari,A. Ottolenghi, D. Scannicchio "Applications and possiblegeneralisations of a method tested at the OPTIS facility, foranalysing physical and radiobiological properties of thera-peutic proton beams", International conference on ocularpathologies therapy with proton beams, Catania, Italy, Oc-tober 12-13, 2000,

N.E.A. Crompton "The Leukocyte Apoptosis Assay" For-schungsseminar, Paul Scherrer Institut, Villigen, Switzer-land, January 12, 2000.

N.E.A. Crompton "Warum und Wie wir Altern" Naturfor-schende Gesellschaft Graubünden, Chur, Switzerland Ja-nuary 13, 2000.

N.E.A. Crompton, R. Miralbell, H.-P. Rutz, F. Ersoy,Ö. Sanal, D. Wellmann, S. Bieri, G. Emery, Y.-Q. Shi, H.Blattmann, M. Ozsahin "Patients With Increased ToxicityTo Radiotherapy display An Altered Apoptotic Prolfile."Lecture, Int. Cong. Translational Res. Lugano, Switzerland,March 6, 2000.

N.E.A. Crompton, Y.-Q. Shi, G.C. Emery, H. Blattmann"Wortmannin enhances radiation-induced apoptosis in acell-type dependent manner." 4th Ann. Meet. SASRO. St.Gallen, Switzerland, April 7, 2000.

N.E.A. Crompton "The biological component of radiationresponse". Invited lecture. National Institute of nuclearPhysics INFN. Genoa, Italy, September 29, 2000.

N.E.A. Crompton, G. Kühne, J. Crawford, S. Gay, T. Pap,"Boron Neutron Capture Synovectomy at SINQ in Switzer-land." 9th Int. Symp. Neutron Capture for Therapy, OsakaJapan, October 4, 2000.

N.E.A. Crompton, Y.-Q. Shi, G.C. Emery, H. Blattmann,"Very High Frequencies Of Heritable Damage Result FromExposure To Even Low Doses Of X-rays". SGSMP, Basel,Switzerland, October 20, 2000.

N.E.A. Crompton "DNA damage and cell death" Vorlesung,University Hospital, Zürich, Switzerland, November 11,2000.

N.E.A. Crompton, Y.Q, Shi, G.C. Emery, H. Blattmann "HighFrequencies of Genetic Instability (Aneuploidy) and Herita-ble Damage (Apoptosis) at Low X-Ray Doses SuggestMammalian Cells Mistake Radiation Damage for Physio-logical Cell Signals". Ann. Meet. Swiss Soc. Oncol. Lau-sanne, Switzerland, November 10, 2000.

N.E.A. Crompton "Radiation response of the cell and theorganism" Vorlesung, University Hospital, Zürich, Switzer-land, November 21, 2000.

N.E.A. Crompton "A highly reproducible flow cytometricleukocyte apoptosis assay" research Seminar, Faculty ofHealth Science, Linköping, Sweden, December 6, 2000.

N.E.A. Crompton "Boron Neutron Capture using BOPP andK2B12H12" Lecture, Symposium, Tumor therapy andsynovectomy based on boron Neutron Capture, Ryttargar-den, Linköping, Sweden, December 6, 2000.

70

N.E.A. Crompton, G.C.Emery, E.M. Ozsahin "Radiationsensitivity Testing using the leukocyte apoptosis assay" 1st

Int. Cong. S. African Radiobiology Soc, University of Stel-lenbosch, Stellenbosch, South Africa, December 11, 2000.

E. Egger, T. Boehringer, A. Coray, G. Goitein,M. Grossmann, P. Jülke, S. Lin, A. Lomax, E. Pedroni,W. Roser, O. Stadelmann, L. Wisser "Quality assurance atthe PSI spot scanning gantry" ICTR2000, First InternationalConference on Translational Research and Pre-ClinicalStrategies in Clinical Radio-Oncology, Lugano, Switzerland,March 5-8, 2000.

E. Egger, L. Zografos, G. Goitein, T. Boehringer, W. Roser,L. Bercher, L. Chamot, A. Schalenbourg "Proton beamradiotherapy of uveal melanoma: results after 15 year"ICTR2000, First International Conference on TranslationalResearch and Pre-Clinical Strategies in Clinical Radio-Oncology, Lugano, Switzerland, March 5-8, 2000.

E. Egger, L. Zografos, G. Goitein "Proton beam radiother-apy of choroidal hemangiomas", 4th annual meeting of theScientific Association of Swiss Radiation Oncology SASRO,St.Gallen, Switzerland, April 6-8, 2000.

E. Egger "Dose prescription and dose reporting" Lecture atthe teaching course on planning radiation treatments: theinput and output phase Paul Scherrer Institute, Villigen, May12-13,2000.

E. Egger "Treatment planning for 3000+ ocular tumors: ourexperience with EYEPLAN", PTCOG XXXIII meeting, Berlin,Germany, September 25-27, 2000,

E. Egger "The OPTIS facility at PSI: Experience and re-sults", International conference on ocular pathologies ther-apy with proton beams, Catania, Italy, October 12-13, 2000

G. Goitein "Strahlentherapie mit Protonen" Invited lecture,Kolleg St. Blasien, St. Blasien, Germany, January 2000

G. Goitein "Drawing Target Volumes" Invited lecture, PSISpring School - Planning radiation treatments: the inputand output phases, Paul Scherrer Institute, Villigen, Swit-zerland, May 2000

G. Goitein "Klinische Ergebnisse der Protonentherapie"Invited talk, Deutsche Gesellschaft f. Radio-OnkologieDEGRO, Annual Meeting, München, Germany, October2000

G. Goitein "Krebstherapie mit Protonen am PSI" Invitedlecture, Forum Wissenschaft und Energie FWE, AnnualSymposium, Zürich, Switzerland, October 2000

M. Grossmann, A. Lomax, M Goitein "Network-wide applica-tion sharing as part of an electronic patient referral sys-tem.", Proc XIIIth ICCR, Heidelberg, Germany, May 2000.

H. Kooy, U. Oelfke, A. Lomax, H. Paganetti, W. Newhauser,T. Bortfeld, M. Goitein, "Design Considerations for IntensityModulated Proton Therapy Treatment Planning", Proc XIIIICCR, Heidelberg, Germany, May 2000.

A.J. Lomax, L. Celia , D.C. Weber, J.M. Kurtz, R. Miralbell,"Potential role of IMRT and protons in the treatment of thebreast and regional nodes", Proc ASTRO 2000, Int. J.Radiât. Oncol., Biol., 48, (2000) S2065.

A. Lomax "Therapy planning for proton therapy" Invitedseminar, Kernphysik Verschnellungs Institut, Groningen,Holland, February 2000.

A. Lomax "Conformai therapy with protons" Invited seminar,Academisch Ziekenhuis Groningen, Holland, February2000.

A. Lomax "Heavy charged particles: their practical advan-tages in radiation therapy" Invited talk, First InternationalConference on Translational Research and Pre-clinicalStrategies in Radio-Oncology, Lugano, Switzerland, March2000.

A. Lomax "The display and analysis of dose distributions",Lecture, PSI spring school - Planning radiation treatments:the input and output phases, Paul Scherrer Institute, Villi-gen, Switzerland, May 2000.

A. Lomax "The prescription and scoring of plans", Lecture,PSI spring school - Planning radiation treatments: the inputand output phases, Paul Scherrer Institute, Villigen, May2000.

A. Lomax "Treatment planning for proton therapy", Lecture,Nachdiplom fur Medizinische Physik, ETH, Zurich, Switzer-land, June 2000.

A. Lomax "Practical aspects of proton therapy with spotscanning" Invited talk, Massachusetts General Hospital,Boston, USA, August 2000.

A. Lomax, H Paganetti "Uncertainties in proton therapy: Theeffect of RBE and set-up errors" Invited talk, 7th Workshopon Heavy Charged Particles in Biology and Medicine,Darmstadt, Germany, September 2000.

A. Lomax "Estimating uncertainties in proton therapy" In-vited seminar, DKFZ, Heidelberg, Germany, November2000.

A. Lomax "Therapy planning for proton therapy" Invitedseminar, University of Pennsylvania, USA, December 2000.

A. Lomax "Conformai therapy with protons: How does itcompare?" Invited seminar, University of Pennsylvania,USA, December 2000.

A. Lomax, "3D IMPT: A first clinical example", SASRO IV,St. Gallen, Switzerland, April 2000.

A. Lomax, T. Boehringer, A. Coray, E. Egger, G. Goitein,M. Grossmann, P. Juelke, S. Lin, E. Pedroni, B. Rohrer, W.Roser, B. Rossi, B. Siegenthaler, O. Stadelmann, H.Stauble, C Vetter, L Wisser. "Intensity modulated protontherapy: A first clinical example.", Proc XIIIth ICCR, Heidel-berg, Germany, May 2000.

R. Miralbell, A Lomax, L Celia, "Potential role of intensitymodulated proton beams in prostate cancer radiotherapy",Proc ASTRO 2000, Int. J. Radiât. Oncol., Biol., 48, (2000)S2090.

E. Pedroni "Studies for a new improved commercial versionof the PSI compact proton gantry", annual conference of theSASRO, St. Gallen, Switzerland, April 7, 2000.

E. Pedroni "Medical Accelerators", Seminar at the CERNAccelerator School (CAS), Lufthansa Bildungszentrum,Seeheim, Germany, May 8, 2000

E. Pedroni "Latest Developments in Proton Therapy" invitedtalk at 7. European Accelerator Conference (EPAC 2000)Austria Center Vienna Vienna, Austria, June 26-30, 2000

E. Pedroni "Proton Therapy at PSI-gantry and active scan-ning" invited talk at the International Meeting on MedicalAccelerators (IMMAC 2000) Vienna University of TechnolgyVienna, Austria, July 1, 2000

71

E. Pedroni "The PSI compact gantry system dedicated tobeam scanning: acquired experience and possible furtherdevelopments", invited talk at the "Workshop Cyclotrons inProton Therapy". PTCOG XXXIII, Berlin Germany, Septem-ber 24, 2000

E. Pedroni "Dose Homogeneity errors due to Movementsduring Beam Scanning: the Simulation of Single FractionIrradiation of Mice as a Warning Example", PTCOG XXXIIIBerlin, Germany, September 26, 2000

E. Pedroni "Physik+Biologie der Protonentherapie" eingela-dener Vortrag - Symposium Hadronen-Therapie Jahreskon-gress 2000 der Degro, Ogro und DGMP,Munchen, Germa-ny, Oktober 7, 2000

E. Pedroni "Optimisation of a Compact Gantry System forProton Therapy", invited talk at the 16. International Confer-ence on the Application of Accelerators in Research andIndustry (CAARI 2000), Denton Texas, USA, November 1-4,2000

W. Roser MRI image distortions. Teaching Course "Plan-ning radiation treatments", Villigen, Switzerland, May 12,2000.

W. Roser, E. Egger, G. Goitein, A. Lomax, E. Pedroni, L.Wisser "Use of a Remote CT Scanner for Routine PatientSetup in Precision Radiotherapy". Proceedings, AnnualMeeting of the Swiss Society for Radiation Biology andMedical Physics, 115-120, (2000).

W. Roser et al. "Use of a Remote CT Scanner for RoutinePatient Setup in Precision Radiotherapy". Annual Meeting ofthe Swiss Society for Radiation Biology and Medical Phys-ics, Basel, Switzerland, October 20, 2000.

W. Roser "Nutzung der Imaging Plates fur die Protonenthe-rapie". 2nd Imaging Plate Meeting, Villigen, Switzerland,March 9, 2000.

W. Roser "Use of a Remote CT Scanner for Routine PatientSetup in Precision Radiotherapy", PTCOG XXXII, Uppsala,Sweden, April 19, 2000.

Y.-Q. Shi, Emery, G., Blattmann, H., Crompton, N.E.A."Different types of apoptosis are induced by the p53 andceramide signal-transduction pathways in two human lym-phoblastoid cell lines." 4th Ann. Meet. SASRO. St. GallenSwitzerland, April 7, 2000.

Y.-Q. Shi, Emery, G., Blattmann, H., Crompton, N.E.A."Wortmannin Selectively Enhances Radiation-InducedApoptosis in Proliferative but not Quiescent Cells." Lecture,Int. Cong. Translational Res. Lugano, Switzerland, March 7,2000.

Y.-Q. Shi, L. Li, O., Sanal, H. Blattmann, N.E.A. Crompton"High Levels of Delayed Radiation-Induced Apoptosis Ob-seved in Lymphoblastoid Cell Lines from Ataxia Telangiec-tasia Patients." Lecture, Int. Cong. Translational Res.Lugano Switzerland, March 6, 2000.

A.R. Smith, J.S. Loeffler, J.A. Adams, A.J. Lomax,A. Niemierko "The potential for proton therapy to improveclinical outcome: Comparisons of proton and x-ray treat-ment plans for the purpose of tumour dose escalationand/or reduction of treatment modality" Proc ASTRO 2000,Int. J. Radiat. Oncol., Biol., 48, (2000) S2150.

TULOC

R.K. Munch, H. Blattmann, S. Kirsch, Ch. Schilling,P.G. Seiler, J. Verwey "A novel tracking technique for thecontinuous precise measurement of tumor positions inconformal radiotherapy" Wissenschaftliche Jahrestagungder Schweizerischen Gesellschaft fur Strahlenbiologie undMedizinische Physik SGSMP Basel, Switzerland, Oktober19-20,2000.

P.G. Seiler "Prazisionsstrahlentherapie von Krebs"Physikalisches Kolloquium, Universitat Bern, Switzerland,May 5, 2000.

P.G. Seiler, H. Blattmann, S. Kirsch, R.K. Munch,Ch. Schilling, J. Verwey "TULOC - a novel method for realtime tumor tracking" Conference gatedradiotherapy 2000,Hokkaido University Hospital, Sapporo, Hokkaido, Japan,July 28, 2000.

P.G. Seiler, H. Blattmann, S. Kirsch, R.K. Munch,Ch. Schilling, J. Verwey "Echtzeitortung vonTumoren zurVerbesserung der konformalen Radiotherapie" BMT-Kongress 2000, , Lubeck, Germany, September 28-302000.

P.G. Seiler "Prazisionsstrahlentherapie von Krebs" Physik-Seminar, Universitat Zurich, Switzerland, December 30,2000.

J. Verwey, P.G. Seiler, H. Blattmann, S. Kirsch,R.K. Munch, Ch. Schilling "Developments of the TULOC(TUmor LOCation) Project" PTCOG XXXIII, Berlin,September 25-27, 2000.

72

TEACHING ACTIVITIES

STRUCTURAL BIOLOGY

F.K. Winkler "Molekularbiologie und Biophysik V: Proteine:Struktur, Funktion und Engineering" (with others) ETH Zü-rich, WS 2000/2001.

INSTITUTE OF MEDICAL RADIOBIOLOGY

R. Jaussi, K. Ballmer-Hofer WS 1999/2000, PostgraduateKurs für experimentelle Medizin und Biologie (Leitung J.Zapf) Vorlesungsverzeichnis #858 UNIZH.

R. Jaussi WS 1999/2000, Biochemie III (Koordinator:A. Plückthun) Vorlesungsverzeichnis #2247 UNIZH.

R. Jaussi, K. Ballmer-Hofer SS 2000, Postgraduate Kurs fürexperimentelle Medizin und Biologie (Leitung J. Zapf R.Jaussi, K. Ballmer-Hofer) Vorlesungsverzeichnis #86UNIZH.

R. Jaussi SS 2000, Zellzyklus und Umweltnoxen Vorle-sungsverzeichnis #2249 UNIZH.

R. Jaussi WS 2000/2001, Molekulare Zellbiologie (Koordi-nator A. Plückthun) Vorlesungsverzeichnis #2279 UNIZH.

K. Ballmer-Hofer, Vorlesung D3/1 und D3/2, Cellular signal-ling, Biozentrum Basel.

K. Ballmer-Hofer, Vorlesung 11/1 und II/2, Molecular Virolo-gy, Biozentrum Basel.

K. Ballmer-Hofer Schülerinnen der Herbstschule, 4 KlassenOktober 2000.

CENTER FOR RADIOPHARMACEUTICALSCIENCE

P.A. Schubiger "Tumor-Targeting" Vorlesung ND5, Medizin-physik ETH, Zurich, May 18, 2000.

P.A. Schubiger "Radiopharmazeutische und bioanorgani-sche Chemie" Vorlesung WS99/00 ETH, Zurich, zusammenmit R. Alberto.

P.A. Schubiger "Einführung in die pharmazeutischen Wis-senschaften" Vorlesung WS00/01 ETH, Zurich, 3 Stundenmit S. Ametamey, E. Garcia Garayoa und B. Kaser-Hotzüber Tiermodelle.

RADIATION MEDICINE

H. Blattmann „Bestrahlungstechniken bei perkutanen Be-strahlungen". XII. Winterschule der DGMP, ÖGMP,SGSMP, Medizinische Physik in der Strahlentherapie, Pichl,Österreich, January 17-21, 2000.

N.E.A. Crompton "Krebs, Altern und Zelltod" Vorlesung No.611, Universität Zürich.

N.E.A. Crompton "Einführung in der Durchflusszytometrie"Vorlesung No. 471, Universität Zürich.

N.E.A. Crompton "Zellen und Noxen" Vorlesung No. 496,Universität Zürich.

N.E.A. Crompton "Durchflusszytometrie für Forschungs-zwecke" Vorlesung No. 485, Universität Zürich.

K. Ballmer-Hofer Schnuppertage von Schülerinnen am IMROktober und Juli 2000.

73

HIGHER DEGREES AWARDED

INSTITUTE OF MEDICAL RADIOBIOLOGY

P. Scheidegger: PhD. thesis:

Targeting of human endothelial cells by VEGF and VEGF-like peptide mimetics.

Dissertation ETH Nr. 13667, Winkler, F., Klemenz, R. , Jaussi, R., Ballmer-Hofer, K.

W. Weiglhofer: PhD thesis:

Targeting of the VEGF receptors by a human immunodeficiency virus-encoded angiogenic peptide.

Dissertation ETH Nr. 13981, Winkler, F., Neri, D., Jaussi, R., Ballmer-Hofer, K.

CENTER FOR RADIOPHARMACEUTICAL SCIENCE

P.M. Gucker: Title of thesis:

Effects of ecstasy (MDMA) on the brain uptake of [C-11](+)McN-5652 studied by positron emissiontomography

Dissertation ETH Nr. 13650, ETH Zurich, Switzerland, 1-163 (2000).

U. Kessler: Title of thesis:

Random Synthesis and Biological Characterization of Nucleoside Analogs - New Perspectives forDrug Discovery

Dissertation ETH Nr. 13592, ETH Zurich, Switzerland, 1-109 (2000)

RADIATION MEDICINE

Yu-Quan Shi Doktor der Naturwissenschaften

Studies of programmed cell death and genetic instability induced by X-rays.

ETH Zurich

Referent: Prof. Dr. F. E.Wiirgler

Korreferent: Prof. Dr. Ch. Glanzmann

Korreferent: PD Dr. N. E. A. Crompton

Michael Quicken Doktor der technischen Wissenschaften

Generation and application of statistical shape models for the segmentation of abdominal organs inradiotherapy planning.

ETH Zurich

Referent: Prof. Dr. Gabor Szekely

Korreferent: Prof. Dr. Lawrence H. Staib

Korreferent: Dr. Hans Blattmann

Appendix 1

W=7.82 cm W=8.98 cm

Fig. 3: Example of a dose verification with CCD dosimetry. The 8 groups of images correspond todifferent depths of the dose field (W = water equivalent range). The upper image in eachgroup is the CCD measurement and the lower picture is the corresponding dosecalculation. In order to give a quantitative impression of the agreement, a dose profile at anarbitrarily chosen position (the yellow band in the first image) is shown in the lower part ofeach picture (in red the measurement, in blue the calculation). The true size of the imagesis about 14 cm.