130
Bioconjugate Chem. 1999, 10, 130−136
TECHNICAL NOTES Comparison of the Stability of Technetium-Labeled Peptides to Challenge with Cysteine M. A. Stalteri,† S. Bansal,‡ R. Hider,‡ and S. J. Mather*,†,‡ Imperial Cancer Research Fund, Department of Nuclear Medicine, St. Bartholomew’s Hospital, West Smithfield, London EC1A 7BE, U.K., and Department of Pharmacy, King’s College, London, U.K. Received April 30, 1998; Revised Manuscript Received September 29, 1998
We have labeled a series of short peptides with technetium-99m either by direct labeling at pH 11 or by exchange from [99mTc]technetium glucoheptonate. Typical labeling yields obtained were as follows: N-acetyl-Gly-Cys(S-Acm)-Gly-Cys(S-Acm)-Gly-NH2 (Acm ) acetamidomethyl) 99%, S-benzoylmercaptoacetyltriglycine (S-Benzoyl-MAG3) 95%, mercaptoacetyldiglycine-NH2 (MAG2-NH2) 94%, MAG3 92%, N-acetyl-Aib-Aib-Cys-NH2 (Aib ) aminoisobutyric acid) 91%, Gly-Gly-Gly-Gly 90%, N-acetyl-Gly-Gly-Cys-Gly 87%, cyclo-1,4-(Gly-Gly-Gly-Gln) 40%, S-methyl-MAG2-NH2 0%. In the absence of cysteine, all of the labeled peptides were quite stable in solution, with at least 80-90% of the labeled peptide remaining at 24 h. The order of stability of the labeled peptides to challenge with cysteine was found to be MAG3 > S-benzoyl-MAG3 > N-acetyl-Gly-Cys(S-Acm)-Gly-Cys(S-Acm)-GlyNH2 >N-acetyl-Aib-Aib-Cys-NH2 > MAG2-NH2 > N-acetyl-Gly-Gly-Cys-Gly > cyclo-1,4-(Gly-Gly-GlyGln) > Gly-Gly-Gly-Gly. The peptides without a sulfur donor were least stable to challenge with cysteine.
INTRODUCTION
Radiolabeled peptides have great potential as radiopharmaceuticals (Fischman, 1993; Thakur, 1995; Hom and Katzenellenbogen, 1997; Liu et al., 1997). Naturally occurring peptides are involved in many biological processes, where they may function as hormones, neurotransmitters, cytokines or growth factors. The low molecular weights of peptides, compared to other biomolecules such as antibodies and antibody fragments, result in faster localization and blood clearance. The rapid pharmacokinetics are ideal for labeling peptides with a radioisotope that has a short half-life, such as technetium-99m (Lister-James et al., 1997). In previous work (Mather et al., 1995), we described the use of hybrid peptides which contain both receptorbinding and technetium-chelating amino acid sequences. Such peptides can easily be prepared by solid-phase synthesis techniques, and amino acid sequences designed to chelate technetium can be incorporated at sites where the biological activity will not be affected. This approach has also been used by Lister-James et al. (Lister-James et al., 1995, 1996; Pearson et al., 1996a,b) in the development of technetium-labeled somatostatin receptor-binding peptides and thrombus imaging peptides. The renal function agent MAG3 was the first peptidederived technetium chelator to be developed. Deprotonation of three of the amide nitrogens and deprotection/ deprotonation of the sulfur during the labeling process * To whom correspondence should be addressed. Phone: 44171-601-7153.Fax: 44-171-796-3907.E-mail:
[email protected]. † St. Bartholomew’s Hospital. ‡ King’s College.
gives an N3S triamide thiol ligand which forms a stable complex with a technetium(V) oxo core. MAG2 forms a similar type of complex with technetium-99m, acting as a tetradentate S,N,N,O donor ligand, with the terminal carboxyl group acting as an O donor (Johannsen et al., 1993). N4 donor tetrapeptides, such as tetraglycine, also form stable complexes with technetium-99m, with three deprotonated amide nitrogens and one amine nitrogen coordinated to technetium (Vanbilloen et al., 1995). In previous work (Mather et al., 1995), we investigated the use of the peptide sequence Gly-Gly-Cys as a technetium chelator. Like MAG3, the Gly-Gly-Cys sequence can act as a triamide thiol ligand. Lister-James and coworkers have developed peptides with the technetiumchelating sequences Gly-Gly-Cys, Cys-Gly-Cys and LysGly-Cys, which provide triamide thiol, diamide dithiol, and diamide amine thiol chelating ligands, respectively (Lister-James et al., 1996). Peers et al. are developing peptides for imaging inflammation using the N-dimethylGly-Ser-Cys sequence as a technetium chelator (Peers et al., 1997). Peptide sequences therefore show great promise as chelators for technetium-99m. Factors such as the overall charge around the metal center or the side chains on the chelating amino acids may affect the biodistribution and receptor-binding affinity of technetium-labeled peptides, but the performance of a particular chelator will also depend on its ability to retain the technetium in vivo. Trans-chelation to other sulfur-containing ligands has been shown to be an important route of radiopharmaceutical instability (Pearson et al., 1996a). Thus, it would be useful to explore the technetium chelating abilities of different peptide sequences and the stability of their complexes. In the present work, we have investigated the labeling,
10.1021/bc9800466 CCC: $18.00 © 1999 American Chemical Society Published on Web 12/16/1998
Technetium-Labeled Peptides
Bioconjugate Chem., Vol. 10, No. 1, 1999 131
Figure 1. Structures of the peptide derivatives (a) S-benzoyl-MAG3, (b) MAG3, (c) MAG2-NH2, (d) S-methyl-MAG2-NH2, (e) N-acetylGly-Gly-Cys-Gly, (f) N-acetyl-Aib-Aib-Cys-NH2, (g) N-acetyl-Gly-Cys(S-Acm)-Gly-Cys(S-Acm)-Gly-NH2, (h) Gly-Gly-Gly-Gly, and (i) cyclo-1,4-(Gly-Gly-Gly-Gln).
solution stability, and stability to cysteine challenge of a series of peptides (Figure 1), including the sequences -Gly-Gly-Cys-, -Cys-Gly-Cys-, MAG3, and tetraglycine in order to compare their suitability as technetium chelators. MATERIALS AND METHODS
Radioisotope. Sodium [99mTc]pertechnetate was obtained from an Amertec II technetium-99m generator (Amersham, UK). Reagents for Peptide Synthesis. Tetraglycine was obtained from Sigma (Poole, U.K.). S-Benzoyl-MAG3 was obtained as a commercial kit (TechneScan MAG3, Mallinckrodt Medical, Petten, Holland). p-(9-Fluorenylmethoxycarbonyl-amino)-2,4-dimethoxybenzyl)phenoxyacetic acid (Rink amide linker), Tentagel resin, suitably protected NR-9-fluorenylmethyloxycarbonyl (Fmoc) amino acids, and [2-(1H-benzotriazole-1-yl]-1,1,3,3-tetramethyluronium tetrafluoroborate (TBTU) were obtained from NovaBiochem (U.K.). Dimethylformamide (DMF) was freshly distilled by fractional distillation at reduced pressure. All organic solvents were of analytical reagent grade quality or better. 2-[(Triphenylmethyl)thio]acetic acid was prepared by reaction of mercaptoacetic acid with triphenylmethanol and boron trifluoride etherate in acetic acid according to published methods (Brenner et al., 1984). Peptide Synthesis. Peptides were synthesized by solid-phase methods using a Milligen Biosearch 9050
Pepsynthesizer. Concentrations were performed under reduced pressure at temperatures of S-benzoyl-MAG3 > N-acetylGly-Cys(S-Acm)-Gly-Cys(S-Acm)-Gly-NH2 > N-acetylAib-Aib-Cys-NH2 > MAG2-NH2 > N-acetyl-Gly-Gly-CysGly > cyclo-1,4-(Gly-Gly-Gly-Gln) > Gly-Gly-Gly-Gly. The peptides without a sulfur ligand were least stable to challenge with cysteine. DISCUSSION
Five of the peptides we chose to study were triamide thiol ligands (Table 1). The other four peptides we studied included a triamide thioether, a diamide dithiol, a triamide amine and a tetraamide ligand, respectively. The peptides with a sulfur donor were labeled by transchelation from 99mTc-glucoheptonate or from 99mTc-tartrate in the case of S-benzoyl-MAG3. The peptides cyclo1,4-(Gly-Gly-Gly-Gln) and Gly-Gly-Gly-Gly, which have a tetraamide and a triamide amine set of donors, respectively, were labeled by direct labeling at pH 11. High labeling yields were obtained for all the peptides except cyclo-1,4-(Gly-Gly-Gly-Gln) and S-methyl-MAG2NH2. All of the labeled peptides were stable in solution at pH 7. In the case of the peptides N-acetyl-Aib-Aib-Cys-NH2, N-acetyl-Gly-Cys(S-Acm)-Gly-Cys(S-Acm)-Gly-NH2, and cyclo-1,4-(Gly-Gly-Gly-Gln), multiple unresolved peaks were present in the HPLC radiochromatograms. Square
134 Bioconjugate Chem., Vol. 10, No. 1, 1999
Stalteri et al.
Figure 3. Reversed-phase HPLC radiochromatograms showing the time course of the reaction of 99mTc-labeled N-acetyl-Gly-GlyCys-Gly with cysteine. Beckman Ultrasphere 5 µ ODS column; mobile phase 0.01 M sodium phosphate pH 6.0 (A), methanol (B) gradient; flow rate 1.0 mL/min. (a) t ) 0, (b) 10/1 cysteine/peptide, t ) 0.5 h; retention times 99mTc-cysteine 5.2 min; [99mTcO(Nacetyl-Gly-Gly-Cys-Gly)]2- 12.2 min, (c) 10/1 cysteine/peptide, t ) 2 h; retention times 99mTc-cysteine 4.4 min; [99mTcO(N-acetyl-GlyGly-Cys-Gly)]2- 11.6 min.
Figure 4. Percent of 99mTc remaining bound to peptide versus time after incubation with cysteine at 37 °C for cysteine-to-peptide ratios of 2/1 (left panel) and 10/1 (right panel).
pyramidal oxo-technetium complexes containing chiral centers on the ligand backbone may form diastereomers. (Liu et al., 1997). In the peptide cyclo-1,4-(Gly-Gly-GlyGln), the R carbon of the Gln residue is chiral. The Gln side chain can be either syn or anti with respect to the technetium-oxo core, thus diastereomers can be formed. N-Acetyl-Gly-Cys(S-Acm)-Gly-Cys(S-Acm)-Gly-NH2 is a pentapeptide, and so more than one mode of coordination is possible. Diastereomers can be separated by HPLC, but the gradient we used was possibly too steep to resolve the various components. Syn and anti isomers of square pyramidal complexes with a Tc-oxo core may show differences in stability and biodistribution (Vanbilloen et al., 1991), and it is therefore also possible that different isomers will show different stabilities to challenge with cysteine (Sharma, 1998). However, in the case of the peptides N-acetyl-Aib-AibCys-NH2, N-acetyl-Gly-Cys(S-Acm)-Gly-Cys(S-Acm)-GlyNH2, and cyclo-1,4-(Gly-Gly-Gly-Gln), which showed multiple unresolved components on HPLC analysis, we found no evidence that one of the components was more or less stable to cysteine than the others. The low labeling efficiency obtained for cyclo-1,4-(GlyGly-Gly-Gln) may be due to the slow kinetics of labeling observed with macrocyclic ligands (Simon et al., 1981), or it may be due to the relative sizes of the technetium oxo core and the macrocyclic ring formed by the cyclic peptide. The 14 membered ring formed by the peptide may be either too small or too large for the technetium to be bound strongly in the plane of the four amide nitrogens. When the reaction mixture was monitored for several hours at room temperature after the initial heating step, the percent of labeled peptide was found to increase with time, from 40% at 30 min to 77% at 3 h.
It is also possible that heating for 10 min is not enough to deprotonate the amide nitrogens. We obtained somewhat higher labeling yields at pH 12, but cleavage of peptide bonds is also more likely to occur at a pH greater than 11 (Bormans et al., 1995). We chose to study the peptide S-methyl-MAG2-NH2 in order to determine what effect, if any, the S-Me group would have on labeling and stability by comparison with MAG2-NH2. Unprotected thiol groups are generally not stable and may oxidize to form disulfides, so it is more convenient to block the thiol with a protecting group (Bodanszky, 1988). Protecting groups such as S-benzoyl and S-acetamidomethyl are easily removed on heating to 100 °C for a short time (Liu and Edwards, 1995). However, some protecting groups, such as S-benzyl and S-benzamidomethyl, are more difficult to remove under the conditions used for radiolabeling (Bormans et al., 1995). We did not expect the S-methyl group of S-methylMAG2-NH2 to be cleaved during the labeling, however the total failure of S-methyl-MAG2-NH2 to label was somewhat unexpected. The N2S2 ligand N,2-methylthiobenzyl-N′-(1-oxo-2-mercapto-2-methyl)propyl ethylenediamine, which has one thiol sulfur donor, and one thioether sulfur donor forms a neutral complex with technetium (Oya et al., 1995). Thioether ligands coordinated to technetium are also found in technetium complexes of sulfur macrocycles (White et al., 1992). Instability of radiolabeled peptides in vivo can occur through two principal routes; through proteolytic degradation of the peptide backbone by serum proteases or by transchelation of the technetium to other high-affinity binding sites. The first route of breakdown can be minimized by the use of unnatural amino acids such as D-optical isomers and through cyclization of the peptide.
Technetium-Labeled Peptides
The second route can be overcome by the use of suitable high-affinity chelators which prevent transfer of the radiolabel to other competing ligands in vivo. Since technetium shows a high affinity for sulfur-containing molecules, its tendency to transchelate in vivo can be modeled in vitro by challenge with the sulfur containing amino acid cysteine. The peptides examined in this study fall into two clear categories as far as their stability to cysteine challenge is concerned. The thiol-containing ligands show much greater stability that those relying entirely on coordination by amide nitrogens. This suggests that sulfurcontaining sequences would provide the chelators of choice for use in receptor-binding peptides. However, in addition to this requirement, this study points to a number of other desirable properties in a peptide technetium chelator which would help in the design of the optimal peptide sequence. MAG3 shows greater stability than MAG2-NH2. Since the only difference between the two molecules is the terminal carboxylate, this group must assist the stability of the complex, either by unexpected involvement in coordination of the technetium or by steric hindrance of competing ligands attacking from the base of the pyramidal complex. This effect is also observed if the terminal amide nitrogen is acetylated as in the N-acetyl-gly-gly cys-gly sequence. However, the presence of bulky substituents on the peptide backbone produces a marked increase in stability as can be observed in a comparison of N-acetyl-gly-gly cys-gly and N-acetyl-aib-aib-cys-gly. Such substituents will tend to promote the folding of the peptide in a manner which tends to favor coordination of the technetium with consequent increase in the stability of the complex. The presence of two thiols in the sequence does not necessarily produce an advantage over one, since both the MAG3 complexes and N-acetyl-gly-cys-gly-cys-glyNH2 show little or no instability in this assay. However, the dithiol sequence achieves this without the presence of the protective terminal carboxyl group. It seems likely therefore that in the absence of the carboxylate, two sulfurs are better than one. The sequence showing the poorest stability to cysteine challenge was the open-chain tetraglycine sequence; however, cyclization of the peptide showed little if any increase in stability, perhaps, as suggested above, because the technetium is not fully located within the cage of the macrocycle. The ultimate performance of a radiolabeled receptorbinding sequence will depend on a number of factors, in particular its affinity for the receptor, its pattern of biodistribution, and its in vivo stability. An advantage of the hybrid peptide approach is that the chelation sequence and the sequence responsible for receptor binding can be optimized independently. However, changes in both of these regions of the peptide are likely to result in an alteration in biodistribution since they will alter the hydrophilicity and possibly the resulting charge of the molecule. The rules which govern the influence of peptide structure on its biodistribution are not yet clearly understood and further studies on the factors which control this relationship are required before an optimal design for these radiopharmaceuticals can be realized. CONCLUSION
Short peptide sequences can be incorporated into receptor-binding peptides to provide a radiolabeling site
Bioconjugate Chem., Vol. 10, No. 1, 1999 135
for technetium-99m. These peptides can be radiolabeled with high efficiency and high solution stability. Increased protection against possible loss of the radiolabel in vivo can be provided by the presence in the chelation sequence of either two sulfur atoms or one sulfur atom and a pendant carboxyl group on the terminal nitrogen. A further increase in stability is provided by the presence of bulky peptide side chains. This information will be of value in the design of optimized peptide chelators for technetium-99m. ACKNOWLEDGMENT
We gratefully acknowledge the financial support of the Imperial Cancer Research Fund and the use the facilities of the Dominion House Centre for Clinical Research. We thank Angela Glenn for technical assistance with the peptide synthesis. LITERATURE CITED (1) Bodanszky, M. (1988) Peptide Chemistry, Springer-Verlag, Berlin. (2) Bormans, G., Cleynhens, B., Adriaens, P., Vanbilloen, H., De Roo, M., and Verbruggen, A. (1995) Investigation of the labeling characteristics of Tc-99m-mercaptoacetyltriglycine. Nucl. Med. Biol. 22, 339-349. (3) Brenner, D., Davison, A., Lister-James, J., and Jones, A. G. (1984) Inorg. Chem. 23, 3793-3797. (4) Fischman, A. J., Babich, J. W., and Strauss, H. W. (1993) A ticket to ride: peptide radiopharmaceuticals. J. Nucl. Med. 34, 2253-2263. (5) Hom, R. K., and Katzenellenbogen, J. A. (1997). Technetium99m-labeled receptor-specific small-molecule radiopharmaceuticals: recent developments and encouraging results. Nucl. Med. Biol. 24, 485-498. (6) Johannsen, B., Noll, B., Leibnitz, P., Reck, G., Noll, S., and Spies, H. (1993) Technetium and rhenium complexes of mercapto-containing peptides. 1. Tc(V) and Re(V) complexes with mercaptoacetyl diglycine (MAG2) and X-ray structure of AsPh4[TcO(MAG2)]‚C2H5OH. Inorg. Chim. Acta 210, 209214. (7) Kates, S. A., Sole´, N. A., Johnson, C. R., Hudson, D., Barany G., and Albericio, F. (1993) Tetrahedron Lett. 34, 1549. (8) Lister-James, J., McBride, W. J., Buttram, S., Civitello, E., Martel, L. J., Pearson, D. A., Wilson, D. M., and Dean, R. T. (1995) Technetium-99m chelate-containing receptor-binding peptides. In Technetium and rhenium in chemistry and nuclear medicine. (M. Nicolini, G. Bandoli, and U. Mazzi, Eds.), Vol. 4, pp 269-274, SG Editoriali, Padova, Italy. (9) Lister-James, J., Moyer, B. R., and Dean, T. (1996) Small peptides radiolabeled with 99mTc. Q. J. Nucl. Med. 40, 221233. (10) Lister-James, J., Moyer, B. R., and Dean, R. T. (1997) Pharmacokinetic considerations in the development of peptide-based imaging agents. Q. J. Nucl. Med. 41, 111-118. (11) Liu, S., and Edwards, D. S. (1995) New N2S2 diamidedithiol and N3S triamidethiols as bifunctional chelating agents for labeling small peptides with technetium-99m. In Technetium and rhenium in chemistry and nuclear medicine. (M. Nicolini, G. Bandoli, U. Mazzi, Eds.) Vol. 4, pp 383-393, SG Editoriali, Padova, Italy. (12) Liu, S., Edwards, D. S., and Barrett, J. A. (1997) 99mTc labeling of highly potent small peptides. Bioconjugate Chem. 8, 621-636. (13) Mather, S. J., Ellison, D., and Bard, D. S. (1995) Technetium-99m labeled hybrid receptor-binding peptides. In Technetium and rhenium in chemistry and nuclear medicine. (M. Nicolini, G. Bandoli, U. Mazzi, Eds.) Vol. 4, pp 491-497, SG Editoriali, Padova, Italy. (14) Oya, S., Kung, M. P., Frederick, D., and Kung, H. F. (1995) New bisaminoethanethiol (BAT) ligands which form two interconvertible Tc-99m complexes. Nucl. Med. Biol. 22, 749757.
136 Bioconjugate Chem., Vol. 10, No. 1, 1999 (15) Pearson, D. A., Lister-James, J., McBride, W. J., Wilson, D. M., Martel, L. J., Civitello, E. R., Taylor, J. E., Moyer, B. R., and Dean, R. T. (1996a) Somatostatin receptor-binding peptides labeled with technetium-99m: Chemistry and initial biological studies. J. Med. Chem. 39, 1361-1371. (16) Pearson, D. A., Lister-James, J., McBride, W. J., Wilson, D. M., Martel, L. J., Civitello, E. R., and Dean, R. T. (1996b) Thrombus imaging using technetium-99m-labeled highpotency GPIIb/IIIa receptor antagonists.Chemistry and initial biological studies. J. Med. Chem. 39, 1372-1382. (17) Peers, S. H., Paul, C., Woodruff, L., Thornback, J. R., Goodbody, A. E., and Bolton, C. (1997) The study of Tc-99m RP128 in experimental allergic encephalomyelitis (EAE), an animal model of multiple sclerosis. J. Nucl. Med. 38, 187P188P (abstract). (18) Sharma, S. (1998) Differential stability of syn and anti isomers in Tc-99m complexes-evidence from cysteine challenge studies. Q. J. Nucl. Med. 42 (3) (Suppl. 1), 39. (19) Simon, J., Troutner, D. E., and Volkert, W. A. (1981) Radiochemical characterisation of Tc-cyclam. J. Radiochem. Radioanal. Lett. 47, 111.
Stalteri et al. (20) Thakur, M. L. (1995) Radiolabeled peptides: Now and the future. Nucl. Med. Commun. 16, 724-732. (21) VanBilloen, H., Bormans, G., Heylen, J., Hoogmartens, M., DeRoo, M., and Verbruggen, A. (1991), Complexes of technetium-99m with tetraglycine and tetra-L-alanine and their biodistribution in mice. J. Labeled Compd. Radiopharm. 30, 42-44. (22) Vanbilloen, H. P., Bormans, G. M., De Roo, M. J., and Verbruggen, A. M. (1995) Complexes of technetium-99m with tetrapeptides, a new class of 99mTc-labeled agents. Nucl. Med. Biol. 22, 325-338. (23) White, D. J., Kuppers, H. J., Edwards, A. J., Watkin, D. J., and Cooper, R. S. (1992). Crown thioether chemistry. The first homoleptic thioether complex of technetium, and its potential application in tumor imaging. Inorg. Chem. 31, 5351-5352.
BC9800466