Dendrimer–protein interactions studied by tryptophan room temperature phosphorescence

Dendrimer–protein interactions studied by tryptophan room

temperature phosphorescence

Edi Gabellieri a, Giovanni B. Strambini

a, Dzmitry Shcharbin

b,

Barbara Klajnert b, Maria Bryszewska

b

a Istituto di Biofisica-CNR, Area della Ricerca, via G. Moruzzi 1, 56124 Pisa,

Italy

b Department of General Biophysics, University of Lodz, 12/16 Banacha St.,

90-237 Lodz, Poland

Abstract

Dendrimers are a relatively new class of materials with unique molecular

architectures, which provide promising opportunities for biological applications as DNA

carriers and drug delivery systems. Progress in these fields, however, requires knowledge of

their potential interactions with biological components at cellular and molecular level. This

study utilizes Trp phosphorescence spectroscopy to examine possible perturbations of the

protein native fold in solution by neutral, positively and negatively charged fifth generation

polyamidoamine (PAMAM) dendrimers. Phosphorescence lifetime measurements, conducted

on model proteins varying in the degree of burial of the triplet probe and in quaternary

structure, show that dendrimers interact with proteins in solutions forming stable complexes

in which the protein structure may be significantly altered, particularly in superficial, flexible

regions of the polypeptide. Both electrostatic and non electrostatic interactions can give rise to

stable complexes, whose affinity and limited number of binding sites distinguish them from

mere aspecific molecular associations. Of direct relevance for the application of these

polymers in the medical field, structural alterations have also been detected in human plasma

proteins such as serum albumin and immunoglobulins (Ig). The above results suggest that Trp

phosphorescence may provide a useful monitor for working out experimental conditions and

protocols that help preserve the structural integrity of proteins in the presence of these

polymers.

2

Introduction

Dendrimers are polymers synthesized in a step-wise manner from branched monomer

units and topologically based on the structure of a tree (1-4) (Figure 1). A typical dendrimer

consists of a multifunctional central core, branched units and surface groups. The branched

units are organized in layers called 'generations', and represent the repeating monomer unit of

these macromolecules [9]. Due to specific synthesis dendrimers possess empty internal

cavities and many superficial functional end groups, which are responsible for high solubility

and reactivity (1-3, 5, 6). Their peculiar structure makes dendrimers potentially suitable for a

variety of biomedical applications. For example, drug molecules can be attached to

dendrimers through covalent bonds. Because of the large number of terminal groups to which

drug molecules can be conjugated, dendrimers are capable to carry drugs at a high density (3,

5-8), a feature used to radiolabel antibodies (5, 7, 8). Another strategy to drug delivery is to

design dendrimers as containers to encapsulate drug molecules(5-8). Drugs entrapped inside

the dendrimer can be released slowly, which is important, for example, in reducing the

toxicity anti-tumor(5). Considering the widespread application in the biomedical field (1-10)

knowledge of their potential interactions with biological structures at the molecular,

membrane and cellular levels, is paramount. While the interaction between dendrimers and

cells has been widely studied (5, 7, 8), many aspects of their interaction with proteins are still

unknown.

Recently, we have studied the interaction between polyamidoamine (PAMAM)

dendrimers of 4th

and 5th

generations and the proteins (HSA, BSA, pure and membrane-bound

acetyl cholinesterase) by fluorescence spectroscopy, differential scanning calorimetry and

fluorescent probes (11-15). It was found that dendrimers did affect enzyme activity (11, 13),

binding properties (15) and protein conformation, their interaction depending on pH and

ionic strength (12, 14).

But there are some questions which still remain unclear: What role can the protein net

charge play in such interactions? In what manner these effects depend on protein quaternary

structure and size? How binding can differ for various proteins?

In the present investigation we extend the analysis of potential structuring effects of the

5th

generation PAMAM dendrimers by means of a sensitive spectroscopic technique based on

room temperature Trp phosphorescence. Trp phosphorescence spectroscopy has proven to be

a sensitive structural probe capable of detecting even subtle changes in protein conformation

such as those induced by ligand binding (16, 17), subunit association (18), protein-protein

interactions (19), besides variations of physical and chemical properties of the solution (20-

22). The phosphorescence spectrum reports on the polarity and homogeneity of the protein

structure around the chromophore, whereas the phosphorescence lifetime is essentially a

monitor of the local conformational flexibility, increasing 3-4 orders of magnitude from fluid

to rigid matrices (23, 24). Depending on the local flexibility of the polypeptide Trp residues

buried inside globular proteins may exhibit long-lived room temperature phosphorescence

lifetime (τ0), in the millisecond to second time range (23), which in general is promptly

affected by even minor changes in protein conformation. For this study model systems were

selected among proteins with well characterized phosphorescence properties in a wide range

of experimental conditions, known crystallographic structure and various degrees of

quaternary structure, namely, apoazurin (Az) and ribonuclease T1 (RNase T1) (monomers),

alkaline phosphatase (AP) and liver alcohol dehydrogenase (LADH) (dimers), and aldolase

(ALD) (tetramer). In general, the emission is due entirely to a single Trp residue per subunit,

which has been identified in all the above proteins except ALD. In the latter, the emission is

from either W295 or W313, or both, as the third residue, W147, is totally quenched by C149

in direct contact with it (23). From the known crystallographic structure it is possible to

determine the distance of the phosphorescence probe from the solvent as well as the charge

distribution on the protein surface. Besides these model proteins, the study of protein-

3

dendrimer interactions was extended to the serum proteins human serum albumin (HSA) and

human γ-globulins (Ig), for their direct relevance in the biomedical applications of

dendrimers. Three series of PAMAM macromolecules (G4.5, G5-OH and G5), differing in

the nature of the end groups, carboxyl, hydroxyl and amine, respectively, have permitted to

examine the influence of the chemical properties of the dendrimer surface on the protein-

dendrimer interaction.

The results point out that dendrimers bind to the proteins, also at ionic strength higher

that the physiological value, and that the electrostatic attraction is not the only mechanism

driving the interaction.

Materials and Methods

Chemicals and proteins: All chemicals were of the highest purity grade available from

commercial sources and were used without further purification.

Tris(hydroxymethyl)aminomethane (TRIS), NaCl Suprapur® and sodium acetate (NaAc)

Suprapur® were from Merck (Darmstadt, Germany). PAMAM dendrimers were purchased in

methanol solution from Sigma-Aldrich Co. (St. Louis, MO). Lyophilised ribonuclease T1

(RNase T1) from Aspergillus oryzae was from Calbiochem Corp. (San Diego, CA). Horse

liver alcohol dehydrogenase (LADH) was supplied as crystalline suspension from Fluka

Chemie GmbH (Buchs SG, Switzerland). Rabbit muscle aldolase (ALD), E. coli alkaline

phosphatase (AP), human -globulins (Ig) and human serum albumin (HSA) were from

Sigma-Aldrich Co. (St. Louis, MO). Fatty acids were removed from HSA by charcoal

treatment, following the procedure described by Chen (1967) (25). Apo-azurin (Az) was

prepared by removing copper ion from Pseudomonas aeruginosa azurin, isolated and purified

following published protocols (26). The plasmid carrying the azurin sequence was a generous

gift from Prof. Alessandro Desideri (University of Tor Vergata, Rome, Italy). Water, doubly

distilled over quartz, was purified by a Milli-Q Plus system (Millipore Corporation, Bedford,

MA). All glassware used for sample preparation was conditioned in advance by standing for

24 h in 10% HCl Suprapur® (Merck, Darmstadt, Germany).

Sample preparation: Prior to sample preparation all protein stocks were dialyzed

overnight against 100 mM TRIS buffer, pH 7.5. Different buffers were used for LADH and

RNase T1: LADH was dialyzed against 10 mM TRIS, pH 7.5, whereas 100 mM NaAc/acetic

acid, pH 5.5, was used for RNase T1. The protein concentration in phosphorescence

measurements was about 5 M throughout. Small volumes (100 l) of PAMAM dendrimers

were dialyzed, using disposable dialyzer (Micro DispoDialyzer®, Spectrum Laboratories Inc.,

Rancho Dominguez, CA,) against the same buffer used for the protein. Dendrimers

concentration after dialysis was determined from absorbance to 280 nm, by comparison to

that of stock dendrimer solution.

Phosphorescence spectroscopy: For phosphorescence measurements in fluid solutions,

O2 removal was achieved by the alternative application of moderate vacuum and inlet of ultra

pure N2 (27). The samples were placed in specially designed T-shaped spectrosil quartz

cuvettes (4 mm ID round tubing in the optical section, Hellma, Mullheim/Baden, Germany)

and rocked very gently, for about 10 min, to achieve complete exchange. The cuvette was

connected to the N2/vacuum line by peek tubing (1/16”) and the sample was fully isolated

from the atmosphere by a septum (Hamilton 76003, Alltech, Lancashire, UK) plus O-ring seal

assembly (27). Based on the phosphorescence lifetime of the protein alcohol dehydrogenase

from horse liver, which exhibits one of the highest sensitivities to O2 quenching, this

procedure lowered the O2-level below 2 nM.

Phosphorescence spectra and decays were both measured with pulsed excitation ( ex =

288 nm) on a home made apparatus (27), modified to implement spectral measurements by

4

means of CCD camera. Pulsed excitation was provided by a frequency-doubled Nd/Yag-

pumped dye laser (Quanta Systems, Milan, Italy) with pulse duration of 5 ns and a typical

energy per pulse of 0.5 - 1 mJ. For spectra measurements the emission was collected at 90

from the excitation and dispersed by a 0.3 m focal length triple grating imaging spectrograph

(SpectraPro-2300i, Acton Research Corporation, Acton, MA) with a band pass ranging from

1.0 to 0.2 nm. The emission was monitored by a back-illuminated 1340x400 pixels CCD

camera (Princeton Instruments Spec-10:400B(XTE), Roper Scientific Inc., Trenton, NJ)

cooled to -60°C.

Phosphorescence decays were monitored by collecting the emission at 90° from

vertical excitation through a filter combination with a transmission window of 405-445 nm

(WG405, Lot-Oriel, Milan Italy; plus interference filter DT-Blau, Balzer, Milan, Italy). The

photomultiplier (EMI 9235QA, Middlesex, UK) was protected against fatigue from the strong

excitation/fluorescence pulse by a gating circuit that inverts the polarity of dynodes 1 and 3,

for up to 1.5 ms, after the laser pulse. The photocurrent was amplified by a current-to-voltage

converter (SR570, Stanford Research Systems, Stanford, CA) and digitized by a

computerscope system (ISC-16, RC Electronics, Santa Barbara, CA) capable of averaging

multiple sweeps. All phosphorescence decays were analyzed by a non-linear least squares

fitting algorithm (DAS6, Fluorescence decay analysis software, Horiba Jobin Yvon, Milan,

Italy).

Each spectral and lifetime determination was repeated at least three times.

Results and Discussion

The effect of G5-OH, G5 and G4.5 (100 M) on the phosphorescence spectrum and

lifetime of every protein (typically 5 M) was measured at 20 °C. The buffer is 100 mM

TRIS, pH 7.5 except for LADH (10 mM TRIS, pH 7.5) and RNase T1 (100 mM NaAc, pH

5.5), as τ0 of these proteins depend on the salt concentration (LADH) and on pH (RNase T1).

In no case the phosphorescence spectrum was affected by the dendrimer whereas the

phosphorescence lifetime of some proteins was significantly altered by it.

Dendrimer effects on the phosphorescence lifetime of model proteins: For each protein

table 1 reports the Trp residue responsible for the emission (24), its location relative to the

aqueous phase (rp is minimum distance from the aqueous interface), its intrinsic

phosphorescence lifetime, τ0, and the lifetime ratio, τG/τ0, representing the change in τ induced

by each dendrimer (100 M). In the case of ALD and LADH the phosphorescence decay both

prior and after the addition of the polymer is not exponential (evidence of conformational

heterogeneity) and τ represents the average lifetime τav = α1τ1+α2τ2 obtained by a two

component fit of the data. For convenience the proteins are listed in the order of increasing

burial of the probe within the globular structure (rp). For the most deeply buried Trp residues,

W48 of Az and W109 of AP, the results show that the lifetime ratio τG/τ0 is 1 with every

dendrimer, implying that the structure within the internal rigid cores (large τ0) of these

macromolecules is not affected by the polymers. Thus, the globular fold remains intact and

any perturbation of protein flexibility in the outer layer by transient or stable association

between protein and polymer does not propagate to the interior. The results obtained with the

other proteins do, however, indicate that more superficial protein sites are generally affected

by the polymers. In the case of LADH and RNase T1 every polymer causes a reduction of τ as

if, locally, the native fold became more flexible through its association with the dendrimer.

This effect is markedly larger about the most superficial W59 of RNase T1. The opposite

trend, an increase of τ by dendrimers, is observed with ALD and only by the charged

polymers G4.5 and G5. Here, the probe reports a tightening of the structure about

W295/W313, sites that, judging from the small τ0, are both rather flexible.

5

It should be pointed out that a decrease of τ may also be caused by quenching

reactions with the functional groups of the dendrimer itself or with impurities associated to

the dendrimer stock solution. However, the functional groups on the dendrimer surface, OH

for G5-OH, NH3+

for G5 and COO- for G4.5, are equivalent to the side chains of Ser, Lys and

Glu, which in proteins have been shown to be inert with respect to the phosphorescence

lifetime of Trp in direct contact with them (23, 24). As to quenching by trace impurities a non

linear dependence of the decay rate on the dendrimer concentration, with early saturation

plateaus (see below), is a strong indication that impurity quenching is not important under

these conditions.

The above perturbations of the intrinsic lifetime of buried Trp residues by dendrimers

provide direct evidence that these polymers may interact sufficiently strongly with protein

molecules to alter their conformation in peripheral regions and at the aqueous interface.

Although the number of proteins examined is too limited to establish any firm correlation

with either their chemical composition or 3D structure it appears that, beyond the above noted

correlation with rp, other factors such quaternary structure and net surface charge are not

crucial for their association with the polymers. Indeed, a significant perturbation is observed

with monomeric RNase T1 as well as with dimeric LADH and tetrameric ALD. Likewise, the

results with RNase T1 indicate that the net protein charge is not the dominant steering force of

the interaction as neutral G5-OH and oppositely charged polymers are all quite effective. For

LADH and ALD the largest effects are observed with positively charged G5 in spite the net

protein charge, at pH 7.5, is also positive (table1).

Effects of G5 concentration on of LADH and ALD: To shed light on the nature of

the interaction between proteins and dendrimers we studied the effect of increasing

concentration of G5 on τ of LADH and ALD. The systems LADH/G5 and ALD/G5 were

selected for the abundant protein supply and because G5 is among the three dendrimers that

causes the largest variation of τ, decreasing the lifetime of LADH and increasing that of ALD.

As indicated in figure 2, A and B, the addition of micro-molar quantities of G5 is sufficient to

produce the large changes of τ observed in these proteins. A plateau is reached just above 5

and 40 µM G5 for LADH and ALD, respectively. In both cases the 1/ vs. G5 concentration

profile is roughly hyperbolic, reminiscent of a binding curve. This finding suggests that

collisions between LADH/ALD and G5 change the protein conformation, presumably by

producing a bimolecular complex.

A simple scheme to account for the change of protein P upon transient complex

formation with a non-phosphorescent ligand L is as follows:

kon

P* + L ↔ P

*L

↓ koff ↓

kP kPL

where kP and kPL are the phosphorescence decay rate constant of free protein (P*) and of the

protein complex (P*L), respectively. Under rapid equilibrium conditions (kP, kPL < kon, koff)

the overall decay rate, 1/τ, is governed by

1/τ = fP kP + (1-fP) kPL

where fP is the fraction of free protein, which is related to the dissociation constant, KD =

koff/kon = fP[L]/(1- fP). Alternatively, when the equilibrium is slow the phosphorescence decay

is non exponential with lifetimes τ0 = 1/kP and τPL = 1/kPL and amplitudes equal to fP and (1-

fP), respectively. Because the decay of both proteins is throughout heterogeneous it is not

possible to distinguish rapid from slow equilibrium regimes and apply a rigorous analysis. In

either case, however, the average decay rate is governed by the dissociation constant and

6

therefore the profiles of figure 2 provide a rough estimate of KD. The best fit of lifetime data,

assuming rapid equilibrium conditions, yields dissociation constants of 1.5 μM for the

complex LADH/G5 and 10 μM for the complex ALD/G5, respectively. Theoretical curves

corresponding to these values of KD are drawn in figure 2. Compared to affinities of 10-6

-10-9

M observed for specific protein-protein complexes (28), these estimates suggest that the

affinity of G5 for LADH and ALD is comparable to or just weaker than the subunit affinity in

some multimeric proteins. These complexes are characterized by a dissociation free energy

ΔGD (ΔGD = -RTlnKD) in the range of 7 to 8 kcal mol-1

.

Another feature of these complexes is the relative small number of potential binding

sites. From the profiles of figure 2 we note that saturation is reached at a protein:G5 ratio of

about 1:1 for LADH and 1:4 for ALD. These correspond to one binding site per dimer for the

former and one binding site per subunit for the latter. Considering a Mw of 80,000 and

160,000 for the dimer and tetramer, respectively, it is clear that in either case the protein

surface could accommodate many more G5 molecules (Mw of 28,000). Hence, binding of G5

to these proteins, far from being limited by steric hindrance, appears to be directed to specific

regions of the protein surface.

Effect of ionic strength on G5 binding to LADH and ALD:. Electrostatic forces are

expected to play a preeminent role in the interaction between charged macromolecules, a

contribution that becomes strongly attenuated at high ionic strength (µ). Both ALD (pI = 8.4-

9.15) and LADH (pI = 8.7) are positively charged at pH 7.5, and the net charge calculated

from the amino acid sequence is +12 for ALD and +16 for LADH. Because the G5 surface is

also positively charged, having the primary amino groups of PAMAM dendrimers a pKa = 9

(29), we may expect that by increasing µ the dendrimer would form even stronger complexes

with the above proteins. To assess the importance of electrostatic “binding” in the protein/G5

interaction, the phosphorescence lifetime vs. G5 concentration profiles were monitored also at

higher ionic strength, through the addition of NaCl.

We first point out that the intrinsic lifetime of both LADH and ALD vary with ionic

strength, increasing slightly in the case of ALD, decreasing for LADH. For this reason τ0

refers to the intrinsic lifetime of the protein at the value of µ under examination. Figure 3A

compares the τG5/τ0 profiles of LADH at values of equal to 0.004 and 0.1 M. Rather

surprising, in the presence of 100 mM salt the dendrimer effect on the lifetime is totally

abolished. The response to increased µ is opposite to that expected from the net protein

charge. As pointed out above both protein and dendrimer are positively charged, so that

increasing µ would reduce the charge repulsion between macromolecules and favor their

association. Evidently, the sharp effect of µ implies that G5 binds to LADH predominantly

through an electrostatic attraction and that must involve a specific, negatively charged patch

of the protein surface.

The effect of µ is totally different with ALD where the change in lifetime induced by

G5 is maintained up to 0.5 M NaCl. In this case the main effect of the salt is towards

increasing the affinity of the complex (figure 3B), presumably a result of diminished charge

repulsion between like charged macromolecules. The invariance to ionic strength of the

ALD/G5 complex points out that forces other than electrostatic attraction (H-bonding, van der

Waals and hydrophobic) are responsible for the association of G5 to the ALD. This is a

significant, if surprising, result for it demonstrates that charged dendrimers can make relative

strong associations to proteins also through non ionic interactions, namely without the

formation of salt bridges at the polymer-protein interface. Because burial of charged groups at

the protein-dendrimer interface is costly in free energy terms, it probably requires that the

ionization state of the functional groups in the contact region changes to neutrality. After all

the pKa of the amino group in G5 is considerably lower than that of an isolated primary amine

(~ 9 vs. 12-13) and, further, only a few terminal groups of the dendrimer need to be involved

in the complex. Energetic considerations (30) conducted on protein-protein associations

7

indicate that to make a stable complex, each molecule involved needs to bury a surface area of

about 600 Å2. Considering that on G5 the surface area available per terminal group is

estimated to be 276 Å2 (31) 2-3 functional groups comprise an area comparable in size to

those found in aspecific protein-protein associations (32).

Interaction of dendrimers with plasma proteins HSA and -globulins: HSA has a sole

Trp residue (W214) buried within a deep cleft of the globular structure with the edge of the

benzene ring exposed to the solvent (33). Its phosphorescence emission is characterized by a

short-lived non exponential decay, the lifetime heterogeneity reflecting the presence of

multiple Trp environments/protein conformations in solution. Fig 4A shows the effect of each

dendrimer (100 µM) on the decay of HSA in 100 mM Tris, pH 7.5. Significant changes in the

decay rate are found in the presence of G5, whose effect is to widen the distribution to

shorter and longer-lived components, relative to HSA alone or in the presence of G5-OH and

G4.5. Lifetimes of sub milliseconds or shorter are due to quenching reactions with

neighboring Cys, His or Tyr side-chains (23). The crystal structure of HSA shows that H242

can come in direct contact with the indole ring whereas C200 is about 11 Ǻ from it, just

outside of its interaction range (33). Evidently, binding of G5 to HSA induces subtle

alterations in the protein conformation bringing the quenching side-chains either closer or

farther to the indole ring.

The human Ig has 11 Trp residues distributed among light and heavy chains and, further, no

assignment has been made yet of the room temperature phosphorescence emission to

individual residues. As shown in figure 4B the phosphorescence decay of Ig in 100 mM Tris,

pH 7.5, is highly heterogeneous with lifetime components ranging from sub milliseconds to

40 ms. Whereas neutral G5-OH does not affect the phosphorescence decay, positively and

negatively charged dendrimers alter considerably the decay profile and in a surprisingly

similar manner. As a result the lifetime of short-lived components increased 3-4 folds while

the long components have been shortened. In either case the change of τ is a clear

manifestation that the structure about some Trp residues has been altered, presumably

increasing the separation from quenching side-chains as well by enhancing the flexibility of

structured regions harboring the residues exhibiting the longer phosphorescence lifetimes.

Conclusions

The sensitivity of Trp phosphorescence to perturbations of the native fold has pointed out that

dendrimers tend to interact with proteins in solution forming complexes in which the protein

structure may be significantly altered. The results obtained with model proteins show that

structural changes are probably confined to superficial, flexible regions of the polypeptide.

Both electrostatic and non electrostatic interactions can give rise to stable complexes, whose

affinity and limited number of binding sites distinguish them from mere, aspecific molecular

associations. Of direct relevance for the application of these polymers in the medical field,

structural alterations have also been detected in two plasma proteins. The above results

suggest that Trp phosphorescence may also provide a useful monitor for working out

experimental conditions and protocols that help preserve the structural integrity of proteins in

the presence of these polymers.

Acknowledgement

Dr. Dzmitry Shcharbin is a beneficiary of a Marie Curie International Incoming Fellowship

within the 6th EU Framework Programme (grant 510018).

8

REFERENCES

(1) Fischer, M., and Vogtle, F. (1999) Dendrimers: From design to application - A

progress report. Angewandte Chemie-International Edition 38, 885-905.

(2) Frechet, J. M. J. (1994) Functional Polymers and Dendrimers - Reactivity, Molecular

Architecture, and Interfacial Energy. Science 263, 1710-1715.

(3) Hawker, C. J., and Frechet, J. M. J. (1992) Unusual Macromolecular Architectures -

the Convergent Growth Approach to Dendritic Polyesters and Novel Block

Copolymers. Journal of the American Chemical Society 114, 8405-8413.

(4) Tomalia, D. A., Naylor, A. M., and Goddard, W. A. (1990) Starburst Dendrimers -

Molecular-Level Control of Size, Shape, Surface-Chemistry, Topology, and

Flexibility from Atoms to Macroscopic Matter. Angewandte Chemie-International

Edition in English 29, 138-175.

(5) Cloninger, M. J. (2002) Biological applications of dendrimers. Current Opinion in

Chemical Biology 6, 742-748.

(6) Matthews, O. A., Shipway, A. N., and Stoddart, J. F. (1998) Dendrimers - Branching

out from curiosities into new technologies. Progress in Polymer Science 23, 1-56.

(7) Aulenta, F., Hayes, W., and Rannard, S. (2003) Dendrimers: a new class of

nanoscopic containers and delivery devices. European Polymer Journal 39, 1741-

1771.

(8) Esfand, R., and Tomalia, D. A. (2001) Poly(amidoamine) (PAMAM) dendrimers:

from biomimicry to drug delivery and biomedical applications. Drug Discovery Today

6, 427-436.

(9) Diederich, F., and Felber, B. (2002) Supramolecular chemistry of dendrimers with

functional cores. Proceedings of the National Academy of Sciences of the United

States of America 99, 4778-4781.

(10) Ozaki, S., DeWald, D. B., Shope, J. C., Chen, J., and Prestwich, G. D. (2000)

Intracellular delivery of phosphoinositides and inositol phosphates using polyamine

carriers. Proceedings of the National Academy of Sciences of the United States of

America 97, 11286-11291.

(11) Klajnert, B., Sadowska, M., and Bryszewska, M. (2004) The effect of

polyamidoamine dendrimers on human erythrocyte membrane acetylcholine sterase

activity. Bioelectrochemistry 65, 23-26.

(12) Klajnert, B., Stanislawska, L., Bryszewska, M., and Palecz, B. (2003) Interactions

between PAMAM dendrimers and bovine serum albumin. Biochimica Et Biophysica

Acta-Proteins and Proteomics 1648, 115-126.

(13) Shcharbin, D., Jokiel, M., Klajnert, B., and Bryszewska, M. (2006) Effect of

dendrimers on pure acetylcholinesterase activity and structure. Bioelectrochemistry

68, 56-59.

(14) Shcharbin, D., Klajnert, B., and Bryszewska, M. (2005) The effect of PAMAM

dendrimers on human and bovine serum albumin at different pH and NaCl

concentrations. Journal of Biomaterials Science-Polymer Edition 16, 1081-1093.

(15) Shcharbin, D., Klajnert, B., Mazhul, V., and Bryszewska, M. (2005) Dendrimer

interactions with hydrophobic fluorescent probes and human serum albumin. Journal

of Fluorescence 15, 21-28.

(16) Cioni, P., Piras, L., and Strambini, G. B. (1989) Tryptophan phosphorescence as a

monitor of the structural role of metal ions in alkaline phosphatase. Eur J Biochem

185, 573-9.

(17) Strambini, G. B., and Gonnelli, M. (1990) Tryptophan luminescence from liver

alcohol dehydrogenase in its complexes with coenzyme. A comparative study of

protein conformation in solution. Biochemistry 29, 196-203.

9

(18) Strambini, G. B., Cioni, P., and Puntoni, A. (1989) Relationship between the

conformation of glutamate dehydrogenase, the state of association of its subunit, and

catalytic function. Biochemistry 28, 3808-14.

(19) Gabellieri, E., and Strambini, G. B. (1994) Conformational changes in proteins

induced by dynamic associations. A tryptophan phosphorescence study. Eur J

Biochem 221, 77-85.

(20) Cioni, P., and Strambini, G. B. (2002) Tryptophan phosphorescence and pressure

effects on protein structure. Biochimica Et Biophysica Acta-Protein Structure and

Molecular Enzymology 1595, 116-130.

(21) Gabellieri, E., and Strambini, G. B. (2001) Structural perturbations of azurin deposited

on solid matrices as revealed by trp phosphorescence. Biophys J 80, 2431-8.

(22) Strambini, G. B., and Gabellieri, E. (1996) Proteins in frozen solutions: Evidence of

ice-induced partial unfolding. Biophysical Journal 70, 971-976.

(23) Gonnelli, M., and Strambini, G. B. (2005) Intramolecular quenching of tryptophan

phosphorescence in short peptides and proteins. Photochemistry and Photobiology 81,

614-622.

(24) Gonnelli, M., and Strambini, G. B. (1995) Phosphorescence Lifetime of Tryptophan in

Proteins. Biochemistry 34, 13847-13857.

(25) Chen, R. F. (1967) Removal of Fatty Acids from Serum Albumin by Charcoal

Treatment. J. Biol. Chem. 242, 173-181.

(26) van de Kamp, M., Silvestrini, M. C., Brunori, M., van Beeumen, J., Hali, F. C., and

Canters, G. W. (1990) Involvement of the Hydrophobic Patch of Azurin in the

Electron-Transfer Reactions with Cytochrome-C551 and Nitrite Reductase. European

Journal of Biochemistry 194, 109-118.

(27) Strambini, G. B., Kerwin, B. A., Mason, B. D., and Gonnelli, M. (2004) The triplet-

state lifetime of indole derivatives in aqueous solution. Photochemistry and

Photobiology 80, 462-470.

(28) Nooren, I. M. A., and Thornton, J. M. (2003) Structural characterisation and

functional significance of transient protein-protein interactions. Journal of Molecular

Biology 325, 991-1018.

(29) Cakara, D., Kleimann, J., and Borkovec, M. (2003) Microscopic protonation equilibria

of poly(amidoamine) dendrimers from macroscopic titrations. Macromolecules 36,

4201-4207.

(30) Cyrus Chothia, J. J. (1975) Principles of protein-protein recognition. Nature 256, 705-

708.

(31) Meltzer, A. D., Tirrell, D. A., Jones, A. A., Inglefield, P. T., Hedstrand, D. M., and

Tomalia, D. A. (1992) Chain Dynamics in Poly(Amido Amine) Dendrimers - a Study

of C-13 Nmr Relaxation Parameters. Macromolecules 25, 4541-4548.

(32) Bahadur, R. P., Chakrabarti, P., Rodier, F., and Janin, J. (2004) A dissection of

specific and non-specific protein - Protein interfaces. Journal of Molecular Biology

336, 943-955.

(33) Bhattacharya, A. A., Curry, S., and Franks, N. P. (2000) Binding of the general

anesthetics propofol and halothane to human serum albumin - High resolution crystal

structures. Journal of Biological Chemistry 275, 38731-38738.

10

Table 1: Change in phosphorescence lifetimes ( ) of proteins after addition of 100 M G5-

OH ( G5-OH), G4.5 ( G4.5) or G5 ( G5) at 20°C.

Protein Buffer, pH Net

charge‡)

# Trp rp (Å) 0 (ms) G5-

OH/ 0 G4.5/ 0 G5/ 0

RNase T1 0.1 M NaAc, 5.5 W59 1.5-2 25.7 2 0.26 0.23 0.42

ALD 0.1 M TRIS, 7.5 +12 W313/

W295

3-4 16*)

1 0.94 1.25 1.94

LADH 0.01 M TRIS, 7.5 + 16 W314 4.5 550*)

30 0.91 .84 0.53

Az 0.1 M TRIS, 7.5 - 5 W48 8 590 30 1 0.96 0.96

AP 0.1 M TRIS, 7.5 - 18 W109 11 1870 60 1 0.96 0.97

*)

Average lifetime: = 1 1+ 2 2 ‡)

Calculated from the aminoacid composition.

FIGURE LEGENDS

Figure 1: Structure of the fifth generation PAMAM dendrimers: A) G5; B) G5-OH; C) G4.5.

Figure 2: Effect of G5 concentration on the average phosphorescence lifetime of LADH and

ALD, at 20°C. A) 4.2 μM LADH in 10 mM TRIS pH 7.5; B) 5μM ALD in 100 mM TRIS pH

7.5. Continuous lines represent the best fit of experimental data to the theoretical binding

curve (see text): 1/τ = 1/τ0 +([G5]/(KD+[G5])*(1/τPL-1/τ0)). The reproducibility of τ

determinations is 7 % or better.

Figure 3: Comparative effect of high ionic strength on the phosphorescence lifetime vs. G5

concentration profiles of LADH and ALD. A) 4.2 μM LADH in (■) 10 mM TRIS pH 7.5 and

in (●) 10 mM TRIS plus 100 mM NaCl pH 7.5. B) 5 μM ALD in (■) in 100 mM TRIS, pH

7.5, and in (●) 100 mM TRIS plus 500 mM NaCl, pH 7.5. The temperature is 20°C.

Figure 4: Representative phosphorescence decays of A) 5 μM HSA and B) human Ig in 100

mM TRIS, pH 7.5, before and after the addition of 100 μM PAMAM dendrimers (G5-OH,

G4.5 and G5), at 20 °C.

11

Branch

Branch

Branch

R:

Figure 1.

0 10 20 30 401.6

2.0

2.4

2.8

3.2

3.6

0 40 80 120 160

30

40

50

60

1 /

av

[G5] M

A

LADH

[G5]

B

ALDOLASE

Figure 2.

12

0 50 1000.4

0.6

0.8

1.0

0 50 100 150

1.0

1.2

1.4

1.6

1.8

2.0

[G5] M

[G5] M

B

A

LADH

/ 0

/ 0

ALDOLASE

Figure 3

0 4 8

0.1

1

0 2 4

0.01

0.1

1

ms

B

Ig

+ G5

+ G4.5+ GOH

Ph

osp

ho

resc

ence

In

ten

sity

ms

A

+ G5+ GOH

+ G4.5

HSA

Figure 4