Tumor targeting in a murine tumor xenograft model with the (sFv′)2 divalent form of anti-c-erbB-2 single-chain Fv

�9 1994 by Humana Press Inc. All rights of any nature whatsoever reserved. 0163-4992194/24-25/267-278156.40

Tumor Targeting in a Murine Tumor Xenograft Model with the (sFv'), Divalent

Form of Anti-c-erbB-2 Single-Chain Fv

JAMES S. HClSTON, *'1 GREGORY P. ADAMS, z

JOHN E. MCCARTNEY, 1 MEI-SHENG TAI, 1 ROBERT M. HUDZIAK, 1 HERMANN OPPERMANN, 1

WALTER F. STAFFORD 111, 3 SEN LI(J, 3 IRWIN FAND, 4 GERALD APELL, 5 AXEL I.,AMINET, 5 MICHAEL A. BOOKMAN, 2

L. L. HOUSTON, 5 AriD LO(JIS M. WEINER z

1 Creative BioMolecules, Inc., 45 South Street, Hopkinton, MA O1748; 2Department of Medical Oncology, Fox Chase Cancer Center,

Philadelphia, PA; 3Department of Pathology, Harvard Medical School and Boston Biomedical Research Institute, Boston, MA; 4 Garden State Cancer Center at the Center for MolecuLar Medicine and Immunology,

Newark, N J; and 5 Cetus Oncology Corporation, Emeryville, CA

ABSTRACT

This investigation has utilized novel forms of the single-chain Fv (sFv), wherein a cysteine-containing peptide has been fused to the sFv carboxyl terminus to facilitate disulfide bonding or specific crosslinking of this sFv' to make divalent (sFv')2. The 741F8 anti-c-erbB-2 monoclonal antibody was used as the basis for construction of 741F8 sFv, from which the sFv' and (sFv')2 derivatives were prepared. Re- combinant c-erbB-2 extracellular domain (ECD) was prepared in CHO cells and the bivalency of 741F8 (sFv')2 demonstrated by its complex formation with ECD. The tumor binding properties of 12SI-labeled anti-c-erbB-2 741F8 sFv, sFv', and (sFv')2 were compared with radiolabeled antidigoxin 26-10 sFv' and (sFv')2 controls. Following intravenous administration of radiolabeled species to severe combined

*Author to whom all correspondence and reprint requests should be addressed.

Cell Biophysics 267 Volumes 24/25, 1994

2 6 8 Huston et al.

immune-deficient (SCID) mice bearing SK-OV-3 tumors (which over- express c-erbB-2), blood and organ samples were obtained as a function of time over 24 h. Comparative analysis of biodistribution and tumor-to-organ ratios demonstrated the 741F8 sFv, sFv', and (sFv')2 had excellent specificity for tumors, which improved with time after injection. This contrasted with nonspecific interstitial pooling in tumors observed with the 26-10 sFv, sFv', and (sFv')2, which decreased with time after administration. Tumor localization was significantly better for disulfide or peptide crosslinked 741F8 (sFv')2 having Gly4Cys tails than for monovalent 741F8 sFv' or Fab. The superior properties of the 741F8 (sFv')2 in targeting SK-OV-3 tumors in SCID mice suggests the importance of further investigations of divalent sFv analogs for immunotargeting.

Index Entries: Tumor targeting; murine tumor xenograft model; sFv'2; anti-c-erbB-2.

INTRODUCTION

Successful tumor targeting of antibodies or their conjugates with radio- nuclides, drugs, or toxins received new impetus with the availability of monoclonal antibodies and their proteolytic fragments. However, antigen- specific immunotargeting of relevant tumors has remained problematical, as it has typically suffered from inadequate tumor specificity and insuffi- cient delivery to target sites. One factor contributing to these observa- tions has been that the large size of intact immunoglobulins prolongs their blood residence and impairs their penetration into tumor from the vasculature. Starting in 1988, small engineered antibody fragments became available, and preliminary studies of their in vivo properties sug- gested that they may solve some of these earlier problems. Along these lines, the present investigation involved animal studies with single-chain Fv proteins that may be useful in the diagnosis and therapy of breast cancers that express abnormally high levels of c-erbB-2 on the surface of malignant cells. Overexpression of this oncogene protein has been reported in about 30% of breast cancer biopsies examined by Slamon et al. (1), as well as in ovarian and other human malignancies. The 741F8 monoclonal antibody is directed against the extracellular domain (ECD) of this oncogene and thus provides an antibody binding site that could have clinical application.

We have prepared single-chain Fv (sFv) proteins based on the variable regions of 741F8 IgG and developed novel analogs with carboxyl-terminal cysteine (sFv') (2,3). These sFv' proteins combine the versatility of Fab' fragments for site-specific labeling with the possible advantages of sFv proteins for targeting and penetration of tumors. These proteins may be used as monovalent sFv' or divalent (sFv')2, and the spacing between

Cell Biophysics Volumes 24/25, 1994

Targeting by Anti.c.erbB.2 (sFv')r 269

sFv' subunits can be potentially varied through the use of crosslinking peptides of various lengths (2). Figure 1 schematically highlights the dif- ferences between the various divalent species that have been described in the current literature, all of which are similar in size to an Fab fragment. The (sFv')2, depicted in panel E of Fig. I is smaller than the miniantibody of Pack and Pltickthun (4), which substitutes amphiphilic helices or leucine zippers for the Fc region (Fig. 1, panel G); these helices are sometimes stabilized with C-terminal cysteinyl residues. The D1.3 antilysozyme (Fv- Cys)2 (Fig. 1, panel H) was the first of these minimal divalent binding proteins to be studied in vivo in a normal mouse model (5). The diabody (Fig. 1, panel I) is made up of two tandem Fv fragments (6), and thus it may potentially exhibit instability at low protein concentrations or aggregation with serum proteins, which was observed for the related bis Fv-Cys dimers. This research (2) is the first in vivo examination of any divalent sFv species in the literature.

MATERIALS AND METHODS

The present study, described recently in detail (2), utilized the V genes for 741F8 IgG, which were cloned by polymerase chain reaction (PCR) techniques and assembled into the sFv (VH-VL) as described by McCartney et al. (8) using serine-rich linkers for bridging V regions, and made by expression in Escherichia coli (8); the 741F8 sFv species used the linker -S4G-S4G-S4-, and 26-10 sFv species used -SG-S4G-S4G-S-. All species were made by direct expression, except the 741F8-2 sFv', which had a leader (ADNKFNKDP) fused at its amino terminus to enhance cyto- plasmic inclusion body formation. General refolding methods were developed for these Mr 27,000 sFv' analogs, using a slight modification of our 3M urea/glutathione redox refolding procedure (9). Antidigoxin 2610 sFv' and anti-c-erbB-2 741F8 sFv' species were refolded to yield active species with their C-terminal cysteine in a mixed disulfide with glutathione. These blocked glutathionyl-sFv' species (sFv'-GI), comprising Ser-Cys tails (unless noted otherwise), were purified by affinity chromatography or by anion/cation exchange and size exclusion chromatography (5). The sFv' monomers were deblocked with mild reduction, and (sFv')2 dimers were formed through disulfide bonds by oxidation or through thioether bonds by N~, N~ bis-maleimidocaproyl peptide (MCA) crosslinking, where the amino acid sequence of this MCA crosslinker consists of Gly-Ser3-Gly2-Ser3-Lys (2). The c-erbB-2 extracellular domain with C-terminal His6 sequence was expressed by CHO cells, and purified by immobilized metal affinity chromatography, which specifically bound the His6 chelation sequence (8). The association between ECD and 741F8


2 7 0

A

Fv

B

sFv

Huston et al.

C

sFv'

D

Fab

E

(sFv') 2

F

Bi BABS

EVH'VL - VH'VL]

G

Miniantibody

EsFv-amph. helix] 2

H , S S ~ I

(Fv-cys) 2 EWLItVH-cy=II2

C2 > Diabody

E(VH "VH )(VL'VL )~

Fig. 1. Schematic representations of monovalent and divalent Fv and sFv species in the literature. (A) Fv heterodimer of noncovalently associated V domains. (B) The single-chain Fv with linker denoted as a sinusoidal line. (C) The sFv' having free cysteine at its carboxyl terminus. (D) Fab fragment of about 50,000 tool wt. (E) The divalent (sFv')2 of about 53,000 mol wt. (F) The single- chain bispecific or BiBABS (7) of theoretical interest. (G) The miniantibody form of sFv-amphiphilic helix fusion proteins (4). (H) The bis Fv-Cys dimer (5). (I) The tandem Fv dimer or diabody (6).

CeU Biophysics Volumes 24/25, 1994

Targeting by Anti.c-erbB.2 (sFv')2 2 71

sFv' or (sFv')2 was assessed by analytical ultracentrifugation (2) and affinity chromatography (8). In general, about 93% of the ion exchange-purified 741F8 sFv' bound to immobilized c-erbB-2 ECD. The antidigoxin 26-10 sFv and sFv' was used as a negative control in these studies and was prepared by direct expression.

Sedimentation analysis was performed as discussed previously (2, 7-9) and described in the legend to Fig. 1. Association constants were determined using the BIAcore instrument (Pharmacia, Brussels, Belgium) (2). For biodistribution studies, sFv molecules were labeled with radio- iodine using the chloramine-T method (15). Prior to their in vivo administration, all labeled sFv molecules were tested by high-performance liquid chromatography (HPLC), sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE), and a live cell binding assay employing c-erbB-2 expressing cells of the SK-OV 3 line. The labeled preparations used in these studies were found to be > 95% pure by HPLC and SDS-PAGE, and exhibited at least 70% immunoreactivity in live cell binding assays.

Biodistribution studies involved 4-6 wk-old Balb/c or c.B17/ICR severe combined immune-deficient (SCID) mice. Tumor biodistribution analysis utilized SCID mice bearing 100-200 mg c-erbB-2-expressing SK- OV-3 tumors. Labeled protein (20-100 #g) was injected via tail vein into the mice, followed by the sacrifice of groups of 3-6 animals at timed inter- vals. The percent injected dose per gram (%ID/g) in tumor and normal organs, as well as tumor:normal organ ratios (T:O) were determined as previously described (16,17). Specificity indices were derived by dividing the mean injected dose per gram values, or tumor-to-organ ratios of the labeled 741F8 sFv by the corresponding values for the control 26-10 sFv. Whole body and blood pharmacokinetics were evaluated and plotted using the weighted least squares of the percentage of errors (18). Radio- immunoimaging was performed using ~3~Iodine-labeled sFv preparations, with images acquired on a gamma camera using a high-energy col- limator. Whole body autoradiography was performed as described (2).

RESULTS AND DISCUSSION

The 741F8 sFv and (sFv')2 molecules retained the binding characteris- tics of the parental IgG. Analytical ultracentrifugation was employed to demonstrate that 741F8 (sFv')2 associates bivalently with soluble c-erbB-2 ECD (Fig. 2). The kinetics of association and dissociation of 741F8 IgG and sFv with soluble c-erbB-2 extracellular domain were measured on the BIAcore instrument and data were consistent with a single reaction com- ponent with an overall association constant of about 0.5 x 108M -1. The whole body and blood clearances of labeled 741F8 sFv from tumor- bearing SCID mice were extremely rapid and biphasic, with a Tl/2ol of 0.2 h and a Tl/2~ of 3.9 h. The cumulative blood retention, i.e., blood area


2 72 Hus ton et al.

" T

. .Q " 0

>

"7,

E

E v

v

A

0.35

0.30

0.25

0.20

0.15

0.10

0.05

0.00 0.0

---- i q--'- r ] T --~ ----T--: - : ] - . ~ --': T I ..... r - ~ --'F ~ :--7

ECD -(sFv')2- ECD t ECD / ~ 7

2.0 4.0 6.0 8.0 10.0

s* (svedbergs)

B

250.0

200.0

O

E 15o.o

v 100.0

, : , , f , T , : , , T , r " ' i

(sFv')2 + ECD ]

o o o f o ~ ~ i

o o r o

o o o o

OOO~O OOOr

ECD O 0 0 @ ( ~ O O E 3 0 (b 0 0 0 0 (~

(sFv')2 500. ' ~ " ~ ' ~ ~ ....... ~"-'~ . . . . . . . . . . . . . . . . . . ~'~ ~ ...... ~: "

0 . 0 . . . . . . . . . ' . . . . . . . . . . . J . . . . . . . . . . . . " . . . . . . . . . . . . . . . . . .

0.0 0.5 1.0 1.5 2.0

c(r) (mg/ml)


Targeting by Ant i .oerbB.2 (sFu')z 2 73

under the curve, was 51.4 #g-h/mL over the course of the study, and the total body half-life was 4.4 h. These results were similar to those obtained with 125I-labeled 26-10 sFv. The chemically linked sFv dimer, 741F8 MCA (sFv')2, followed a similar clearance pattern, al though the dimer dis- played a greater blood retention at 24 h (0.17% injected dose per mL) than the monomer (0.05% injected dose per mL).

The comparative biodistributions of radioiodinated 741F8 sFv and 26-10 sFv are presented in Table 1. The results of this representative ex- periment show evidence of specific binding of the 741F8 sFv to tumor as early as 1 h following the injection of radiolabeled protein. Tumor

Fig. 2. (opposite page) Physicochemical demonstration that 741F8-3 (sFv')2 associates bivalently with c-erbB-2 ECD. Analytical ultracentrifugation was performed on 741F8-3 (sFv')2 and c-erbB-2 ECD, individually and in combination, in PBS at pH 7.0 (0.05M KH2PO4:0.1M NaCI:0.001M EDTA). (A) Velocity sedimentation analysis at 20~ and 56,000 rpm showing apparent sedimentation coefficient distributions of the 741F8 (sFv')2 and c-erbB-2 ECD alone and in an approx equimolar mixture (each profile is labeled). The X-axis is the apparent sedimentation coefficient in units of Svedbergs, whereas the Y-axis is the apparent dif- ferential sedimentation coefficient distribution function in units of mg/mL per Svedberg. The area under each peak is proportional to the concentration of protein in the corresponding boundary. (B) Equilibrium sedimentation analysis at 4~ showing apparent weight-average molecular weight as a function of protein concentration for 741F8 (sFv')2 dimer (squares), the c-erbB-2 ECD (circles), and an equimolar mixture of both species (diamonds). The (sFv')2 dimer was centri- fuged alone at 22,000 rpm, the c-erbB-2 ECD at 12,000 rpm, and their mixture at 12,000 rpin.

Demonstration of divalent binding of soluble c-erbB-2 ECD by 741F8-3 (sFv')2 was achieved through sedimentation analysis. Sedimentation equilibrium experiments with 741F8 (sFv')2, CHO cell-produced c-erbB-2 ECD, or equimolar amounts of 741F8 (sFv')2 and c-erbB-2 ECD were carried out on an Analytical Ultracentrifuge (Model-E: Beckman Instruments, San Rarnon, CA) equipped with a real-time video-based data acquisition system and Rayleigh optics (10). The video-based system automatically converted each digitized Rayleigh pattern into a computer disk file of fringe displacement vs radius. The camera lens was focused at the 2/3 plane of the cell. The cells were equipped with sapphire win- dows and 12-ram, 6 channel external loading centerpieces (11). Other details and methods of data analysis and curve fitting were described previously (12). All protein samples were dialyzed against their respective buffers for at least 24 h, and optical blanks were done at the same temperature.

Sedimentation velocity patterns were acquired every 20 s with the on-line Rayleigh optical system. The concentration vs radius data were converted into apparent (i.e., uncorrected for the effects of diffusion) sedimentation coefficient patterns, g*(s) curve according to published procedures (13,14). Twelve-milli- meter charcoal-filled Epon double-sector centerpieces were used. Protein solu- tions were dialyzed against their respective buffers, and the dialysate was used as the reference solution. [Reproduced with permission from Adams et al., Cancer Res. 53, 4026-4034, 1993.]


274 H u s t o n et al.

Table 1 Comparison of Biodistributions of ~2sI-741F8 sFv and ~2sI-26-10 sFv a

12sI-741F8 sFv 12sI-26-10 sFv

Organ 1 h 4 h 24 h 1 h 4 h 24 h

Tumor 5.64 2.89 0.79 3.63 0.89 d 0.06 Liver 4.05 1.05 0.06 3.20 0.68 r 0.05 Spleen 4.61 1.44 0.08 6.49 0.71 c 0.07 Kidney 27.48 2.05 0.29 15.48 0.91 0.14 Lung 7.20 1.40 0.09 5.94 1.18 e 0.07 Muscle 1.87 0.51 c 0.01 1.96 0.29 d 0.01 Heart 4.06 1.01 0.04 c 3.36 0.80 d 0.04 Stomach 18.98 4.09 0.09 14.93 2.16 0.07 Intestine 3.28 1.53 0.03 2.83 1.10 0.04 Bone 2.17 0.45 0.03 1.55 0.24 0.02 Blood b 7.80 1.10 0.05 4.00 0.60 d 0.04

a One hundred micrograms of either 12sI-741F8 sFv (specific for c-erbB-2) or the control, 12sI-26-10 sFv (specific for Digoxin), was injected iv into SCID mice bearing sc SK-OV-3 tumors expressing c-erbB-2. The average %ID/g and the SEM (_< 20% of the value, unless indicated) were determined. [Reproduced with permission from Adams et al., Cancer Res. 53, 4026-4034, 1993.]

bRepresented as %ID/mL. c SEM _< 30% of the value shown. a SEM _< 40% of the value shown. e SEM ___ 50% of the value shown.

localization of the anti-c-erbB-2 741F8 sFv was far superior to that of the 26-10 sFv at 24 h following injection, wi th 0.79% injected dose of 12sI-741F8 sFv retained per g of tumor, in contrast to only 0.06% injected dose of 12si-26-10 sFv. As shown in Table 2, this advantage for tumor retent ion translated to substantial increases in specificity of delivery, with specificity indices for a broad panel of normal organs rang- ing from 7.0 to 16.7. The advantages of antigen-specific sFv binding to tumor were further confirmed by gamma camera imaging of 131Iodine labeled proteins at 2 and 24 h following their intravenous administration. Al though tumors were imaged by both the 741F8 and 26-10 radiolabeled sFv proteins at 2 h following administration, only the 741F8 sFv was capable of detecting tumor at the 24 h time point. These data provide compelling evidence that the retention of the 741F8 sFv monomer in tumor is regulated by antigen expression, and precludes a significant role for passive interstitial or blood pooling effects.

We hypothesized that divalent single-chain Fv species might retain the advantages of small size with the benefit of increased avidity. Accord- ingly, several divalent 741F8 (sFv')2 proteins were prepared, which dif- fered by having either Ser-Cys or Gly4-Cys tails on the sFv' and by crosslinking with disulfides or the MCA linker. Table 3 summarizes the results


Targeting by Anti.c.erbB.2 (sFv')2 2 75

Table 2 Comparison of Tumor-to-Organ Ratios of 1251-741F8 sFv and 12sI-2610 sFv Injected

into C.B17/lcr-SCID Mice Bearing SK-OV-3 Ovarian Carcinoma Xenografts a

12sI-741F8 sFv 1251-26-10 sFv Specificity index

Organ I h 4 h 24 h 1 h 4 h 24 h 1 h 4 h 24 h

Liver 1.4 2.8 14.2 1.4 1.2 1.1 1.0 2.3 12.9 Spleen 1.2 2.0 10.3 0.5 1.2 0.9 2.4 1.7 11.4 Kidney 0.2 1.4 2.8 0.3 0.9 0.4 0.7 1.6 7.0 Lung 0.8 2.1 9.4 0.9 0.9 0.9 0.9 2.3 10.4 Muscle 3.0 6.0 78.8 1.8 2.9 5.2 1.7 2.1 15.2 Heart 1.4 2.9 22.0 1.1 1.2 1.6 1.3 2.4 13.8 Stomach 0.3 0.7 8.9 0.3 0.4 0.9 1.0 1.8 9.9 Intestine 1.7 1.9 25.0 1.3 0.8 1.5 1.3 2.4 16.7 Bone 2.6 6.9 30.0 2.3 3.3 3.2 1.1 2.1 9.4 Blood b 0.7 2.6 14.7 0.9 1.4 1.4 0.8 1.9 10.5

a Groups of 4-6 mice were sacrificed at the indicated times, the %ID/g of tumor and normal organs were determined for each radiopharmaceutical and the tumor-to-organ ratios (T:O; %ID/g tumor divided by the %ID/g organ) were calculated. The %ID[g values for the 741F8 sFv are found in Table 1. The SEM of each listed values was less than 25%. The specificity index was derived by determining the ratios T:O of 741F8 sFv/T:O of 26-10 sFv. [Adapted with permission from Adams et al., Cancer Res. 53, 4026-4034, 1993.]

bTumor-to-blood ratios were determined by dividing the %ID/g tumor by the %ID/mL blood.

of their biodistribution properties. The MCA was used to crosslink 741F8 (sFv')2 with a Ser-Cys tail, which together with the peptide bridge (-Gly- Ser3-Gly2-Ser3-Lys-) be tween maleimidocaproyl groups produced the longest spacer be tween sFv' subunits. Compared to other forms of monovalent or divalent species, this MCA-linked form of the (sFv')2 demonstrated the best tumor localization, with 1.6-2.1%ID/g retained in tumor at 24 h postinjection, which overlapped with the Gly4-Cys version of 741F8 (sFv')2 at 1.6%ID/g. Furthermore, this increased tumor localization did not result in a loss of specificity, as evidenced by essentially unchanged tumor-to-normal-organ ratios. More recent experiments have shown that the 741F8 (sFv')2 prepared with a Gly4-Cys spacer is the best overall choice w h e n considering the ease of production, extent of quantitative delivery to tumor, and level of tumor specificity. These results suggest that this form of the 741F8 (sFv')2 molecule is appropriate for initial clinical evaluations designed to determine biodistribution and tumor localization capabilities of this novel class of immunotarget ing molecules.

The remarkable specifity of localization achieved with these molecules has been best illustrated by serial sagittal section autoradiography studies in mice bearing c-erbB-2 positive SK-OV-3 tumors. Cross-sectional tumor diameters as small as 1 m m can be detected by both the divalent and monovalent forms of 741F8 sFv, with minimal background interference.


2 76 Huston et al.

r

0

o C o

o

~L~ b o ('xl ~ ,,D o'3 o'x.,D tr) Lr) b,, b., t~r3, . - .. ~ 6 ~ 6 o < ~ c , i ~ 6 c 5 ~ 6 ~ - o

~ . ~ . . m, ~ c ~ o ~ m " ~ ' ~ c

i & ~ d d d d d d d d d d d F ' i~ x

m,~ .

r...~ m

~ ~ , - - 4 d d d d d d d d d d

~ 1 ~ ; d , - ; c s N

L~

m,m, :

t-. ~ M d d d d d d o d o d .~ ~ E

u<

~ , ~ o . ~ . ~.

---. C ~ ..~

Cell Biophysics Volumes 2 4 / 2 5 , 1994

Targeting by Anti-c.erbB-2 (sFv')2 2 7 7

To identify conditions that optimize quantitative delivery and tumor specificity, our recent studies have examined the impact of continuous infu- sions, multiple dosing, and extremely high doses of radiolabeled (sFv')2. Such data will provide important directions for the clinical development of these novel molecules, particularly in clarifying their potential for radioimmunodiagnosis and radioimmunotherapy.

ACKNOWLEDGMENTS

The authors wish to thank Dr. Craig Reynolds and Dr. George Johnson of the National Cancer Institute for their assistance and helpful discus- sions; Leah Conroy and David Lowe for providing sf9 culture supernatants containing c-erbB-2; Robert Valerio of Chiron Mimotopes for synthesis of the MCA crosslinker; Frank LaCreta for assisting with the phar- macokinetic analysis; Samuel Litwin for biostatistical support; Josephine Schultz for the in vitro characterization of the sFv' species; Philip Moldof- sky and the Fox Chase Cancer Center Nuclear Medicine Department for assisting with the gamma camera images; and Zhiyun Tang for assistance with the sagittal section autoradiography.

This investigation was supported by National Cooperative Drug Dis- covery Group grant U01 CA51880, and was funded by the National Cancer Institute of the US Public Health Service.

REFERENCES

1. Slamon, D. J., Clark, G. M., Wong, S. G., Levin, W. J., Ullrich, A., McGuire, W. I., et al. (1987) Science 235, 177-182.

2. Adams, G. P., McCartney, J. E., Tai, M.-S., Oppermann, H., Huston, J. S., Stafford III, W. F., et al. (1993) Cancer Res. 53, 4026-4034.

3. McCartney, J. E., Tai, M.-S., Oppermann, H., Jin, D., Warren, F. D., Weiner, L. M., et al. (1993) Miami Short Reports 3, 91.

4. Pack, P. and Pltickthun, A. (1992) Biochemistry 31, 1579-1584. 5. Cumber, A. J., Ward, E. S., Winter, G., Parnell, G. D., and Wawrzynczak,

E. J. (1992) J. Immunol. 149, 120-126. 6. Holliger, P., Prospero, T., and Winter, G. (1993) Proc. Natl. Acad. Sci. USA

90, 6444-6448. 7. Huston, J. S., Mudget-Hunter, M., Tai, M.-S., McCartney, J., Warren, F.,

Haber, E., et aI. (1991) Meth. Enzymol. 203, 46-88. 8. McCartney, J. E., Tai, M.-S., Hudziak, R. M., Adams, G. P., Weiner, L. M.,

Jin, D., et al. (1994) In review. 9. Tai, M.-S., Mudgett-Hunter, M., Levinson, D., Wu, G. M., and Haber, E.

(1990) Biochemistry 29, 8024-8030. 10. Liu, S. and Stafford, W. F. (1992) Biophys J. 61, A476.


2 78 Huston et al.

11. Ansevin, A. T., Roark, D. E., and Yphantis, D. A. (1970) Anal. Biochem. 34, 237-261.

12. Brenner, S. L., Zlotnick, A., and Stafford, W. F. (1990)J. MoI. Biol. 216, 949-969.

13. Stafford, W. F. (1994) Methods Enzymol. in press. 14. Stafford, W. F. (1992) Anal. Biochem. 203, 295-301. 15. DeNardo, S. J., Peng, J. S., DeNardo, G. L., Mills, S. M., and Epstein, A. L.

(1986) Nucl. Med. Biol. 13, 303-310. 16. DeNardo, G. L., Krohn, K. A., and DeNardo, S. J. (1977) Cancer (Phila.) 40,

2923-2929. 17. Adams, G. P., DeNardo, S. J., Amin, A., Kroger, L. A., DeNardo, G. L.,

Hellstrom, I., et al. (1992) Antibody Immunoconj. Radiopharma. 5, 81-95. 18. Draper, N. R. and Smith, H. (1966) AppI. Reg. Anal. Wiley, New York, pp.

77-81.


Tumor targeting in a murine tumor xenograft model with the (sFv′)2 divalent form of anti-c-erbB-2 single-chain Fv

Documents