Synthesis of Peptide−Oligonucleotide Conjugates for Chromium

JULY/AUGUST 2001 Volume 12, Number 4 © Copyright 2001 by the American Chemical Society

COMMUNICATIONS Synthesis of Peptide-Oligonucleotide Conjugates for Chromium Coordination E. R. Civitello, R. G. Leniek, K. A. Hossler, K. Haebe, and D. M. Stearns* Department of Chemistry, Northern Arizona University, P.O. Box 5698, Flagstaff, Arizona 86011-5698. Received January 31, 2001; Revised Manuscript Received February 28, 2001

The synthesis of the first peptide-oligonucleotide conjugate designed to coordinate chromium(III) is reported. The overall goal of this work is to synthesize di-deoxynucleotides tethered with chromium(III)-coordinating appendages to model chromium-DNA-protein cross-links, which are a type of DNA lesion that may be involved in chromium-induced cancers. The conjugate dGp(NHCH2CH2S-Ac-GlySer-Gly-OH)G was made by coupling the peptide, ClAc-Gly-Ser-Gly-OH, and dinucleotide, dGp(NHCH2CH2SH)G, through a thioether moiety. The conjugate was characterized by HPLC and mass spectrometry. Previously reported methods for small-scale solid-phase synthesis of peptides and dinucleotide were unsuitable; therefore, gram-scale solution-phase methods were developed. We also report the gram-scale syntheses of two other serine-containing peptides, ClAc-βAla-Ser-Gly-OH and ClAc-Ser-Gly-OH, and three histidine-containing peptides, ClAc-Gly-His-Gly-OH, ClAc-βAla-His-GlyOH, and ClAc-His-Gly-OH. The synthesis and characterization of chromium-containing peptideoligonucleotide conjugates will ultimately help us to understand chromium-DNA interactions at a molecular level, which is necessary before we can determine how chromium causes cancer.

INTRODUCTION

Peptide oligonucleotide conjugates have been synthesized for a variety of applications, providing control of cellular uptake and intracellular delivery of oligonucleotides, improvement of DNA or RNA binding, creation and modification of artificial nucleases, and enhancement of nuclease resistance of oligonucleotides (1). We are interested in synthesizing peptide oligonucleotide conjugates to ultimately model chromium-DNA-protein cross-links, which are a type of DNA lesion that may be involved in chromium-induced cancers. Chromium is a carcinogenic metal that may produce DNA damage by two different pathways: oxidative DNA * To whom correspondence should be addressed, Phone: (520) 523-4460; Fax (520) 523-8111; e-mail: [email protected].

damage by high valent chromium intermediates and free radicals, and adduct formation by low valent chromium(III). We are investigating the covalent interaction between Cr(III) and DNA. Amino acid-Cr(III)-DNA adducts and Cr(III)-containing DNA-protein cross-links (Cr-DPC) have been reported in cells (2, 3), animals (4, 5), and humans (6) exposed to Cr(VI) compounds. These Cr-containing DNA lesions are mutagenic (7) and are likely to be involved in Cr-induced cancers because they are more slowly repaired than strand breaks (2). Although Cr(III)-DNA adducts have been detected and quantitated in bulk DNA from cells and animals, there is still very little information available about their structures. Chromium(III) will associate electrostatically with the phosphate backbone. This noncovalent interaction can be disrupted by EDTA or high NaCl (8). There

10.1021/bc0100105 CCC: $20.00 © 2001 American Chemical Society Published on Web 06/07/2001

460 Bioconjugate Chem., Vol. 12, No. 4, 2001

Civitello et al. Scheme 1. Fragment Conjugation

Figure 1. Peptide-dinucleotide conjugate.

is growing evidence that Cr(III) also associates with guanine bases in vitro (9-15), which is consistent with the preference for guanine coordination shown by platinum (16), cobalt and barium (17), and nickel (18). The overall goal of this work is to synthesize dideoxynucleotides tethered with Cr(III)-coordinating appendages to model Cr-DPCs. A tethered Cr(III) will optimize metal DNA interactions by localizing the metal in the coordination environment of the nucleotide and will produce a more stable complex for crystallization and structural characterization. The synthesis of a series of Cr-DNA-peptide conjugates will help us to explain the difference in stabilites between Cr(III) coordination to the nucleotide versus the phosphate backbone. It will also help us to understand the stereochemical and electronic effects of different amino acids when they complete the coordination environment of Cr(III). Characterization of these conjugates will elucidate the mechanisms of Cr(III)induced mutations. An understanding of chromiumDNA interactions at the molecular level is necessary before we can ultimately determine how chromium causes cancer. The design of our initial targets is based on current knowledge of Cr-DNA-protein cross-links. The DNA portion contains guanine dinucleotides because of their known ability to bind metals described above. A series of peptide tethers were made, differing in length, with either serine-glycine or histidine-glycine sequences for the Cr(III)-binding region because these were among the amino acids found in Cr-DPC in cultured cells (3, 19, 20). In these first series of conjugates the tether has been attached to the phosphate bridge of the dinucleotide so that the phosphate oxygen will be blocked from association with Cr(III), thus favoring the coordination of Cr(III) to the guanine bases. Future conjugates will contain the tether attached to the 5′-OH group (21) or the base residue such as the N4-position of cytosine (22). RESULTS AND DISCUSSION

Herein we describe the synthesis of our first peptidedinucleotide conjugate, dGp(NHCH2CH2S-Ac-Gly-SerGly)G (1), specifically designed to bind chromium(III) (Figure 1). We anticipate chromium to bind strongly to the C-terminal peptide sequence, -Ser-Gly-OH. This will serve as a tether holding chromium in proximity to the dinucleotide to allow for the isolation and characterization of the complex. Conjugate 1 was made by the Fragment Conjugation method (1), in that the peptide and dinucleotide were synthesized separately and then coupled through the thioether moiety (Scheme 1). While both the functionalized peptide ClAc-Gly-Ser-Gly-OH (2) and dinucleotide dGp(NHCH2CH2SH)G (3) can be synthesized by solid-phase methods, we found this approach to be unsuitable for our needs. To crystallize and char-

acterize the chromium complex, we envision that gram quantities of 1 will be necessary, and the high cost and small scale of solid-phase methods made this approach uneconomical. Therefore, our first goal was to develop gram-scale solution-phase methods for the synthesis of the peptide and dinucleotide fragments. The synthesis of peptide 2 is outlined in Scheme 2. Diisopropylcarbodiimide (DIC) and N-hydroxysuccinimide (HOSu) were originally used for the coupling of CbzSer(OtBu)-OH (4) to H-Gly-OtBu. However, at multigram scales, the diisopropylurea (DIU) byproduct became problematic. The large amounts of DIU were removed by precipitation from cold CH2Cl2 followed by normalphase column chromatography. However, these methods proved inefficient, and often repeated chromatography was necessary which made this procedure impractical. A superior procedure was found in the mixed anhydride approach. Using isobutyl chloroformate to activate the carboxylic acid moiety of 4, the reaction with H-Gly-OtBu was clean and efficient. The byproducts, isobutanol and N-methylmorpholine (NMM), were easily removed in the workup. Column chromatography was not necessary, and near quantitative yields of dipeptide 5 were obtained. Removal of the benzyloxycarbonyl (Cbz) protecting group by catalytic reduction was straightforward and could be done quantitatively on a multigram scale also without chromatographic purification. These steps were repeated for the subsequent coupling of glycine and chloroacetic acid. The protecting groups were then removed by treatment with TFA, and the final peptide was purified by preparative reversed-phase HPLC.1 Tetrapeptide 2 was isolated in over 90% yield on a 3-g scale. Two other peptides were made by this method, ClAc-βAla-Ser-GlyOH (7) and ClAc-Ser-Gly-OH (8), representing a longer and shorter version of peptide 2, respectively. These peptides were designed to investigate the ideal tether length between the peptide and dinucleotide binding regions. The three corresponding histidine peptides, ClAc-GlyHis-Gly-OH (12), ClAc-βAla-His-Gly-OH (13), and ClAcHis-Gly-OH (14), were made by the method outlined in Scheme 3. Due to known problems with imidazole protecting groups (23), we chose to adapt standard solid1 Reversed-phase HPLC: Varian Dynamax C-18 column, 41.4 × 250 mm, 8 µm, 100 Å; solvent A: 0.1% trifluoroacetic acid (TFA) in water; solvent B: 0.1% TFA in 90/10 acetonitrile/ water; gradient: 0-30% solvent B over 30 min at 50 mL/min; UV detection at 220 nm.

Communications

Bioconjugate Chem., Vol. 12, No. 4, 2001 461

Scheme 2. Synthesis of ClAc-GlySerGly-OH (2)

Scheme 3. Synthesis of ClAc-GlyHisGly-OH (12)

phase Fmoc methods for the synthesis of these peptides. Peptide couplings were achieved with the commerically available HBTU reagent (2-(1H-benzotriazol-1-yl 1,1,3,3tetramethyluronium hexafluorophosphate), and the protected peptide intermediates were purified by normalphase column chromatography. The Fmoc protecting group was removed by treatment with diethylamine (DEA) in THF. The free-amine peptides were separated from the dibenzofulvene byproduct by precipitation from cold hexanes. After the chloroacetylation step, the peptides were deprotected with TFA and purified by prep rp-HPLC.1 All three histidine peptides were produced on a 3- to 5-g scale in overall yields ranging from 85 to 92%. In due course, all six peptides will be conjugated to the functionalized dinucleotide 3 to produce a series of peptide-dinucleotide conjugates. These conjugates will then be incubated with chromium(III) to study chromiumDNA binding. Our synthetic approach to the functionalized dinucleotide fragment 3 was inspired by the solid-phase synthesis reported by Fidanza and McLaughlin (24). In our solution-phase synthesis, both guanine nucleoside pieces 16 and 17 were synthesized from DMT-ibu-deoxyguanosine (15) which is commercially available and relatively inexpensive (Scheme 4). Treatment of 15 with PCl3 and triazole afforded the corresponding H-phosphonate 16, which was isolated and purified as the triethylammonium bicarbonate salt (25, 26). H-Phosphonate 16 could be made on a 3-g scale and purified by normal-phase column chromatography using a gradient of 0-10%

Scheme 4. Dinucleotide Synthesis

MeOH in CHCl3 in an overall yield of 85%. Nucleoside 17 was made from 15 by acetylating the 3′-hydroxy group followed by the removal of the 5′-DMT protecting group. Nucleoside 17 was also purified by normal-phase chromatography (0-3% MeOH in CHCl3) and could be

462 Bioconjugate Chem., Vol. 12, No. 4, 2001 Scheme 5. Dinucleotide Functionalization

Civitello et al.

2D NMR (COSY and ROESY). Removal of the isobutyryl and acetyl protecting groups was accomplished on a small scale by heating with concentrated NH4OH at 45 °C for 2 h. The crude product was then treated with 2.5% Cl3COOH/CH2Cl2 to remove the DMT group and then purified by reverse-phase HPLC4 to afford approximately 23 mg of compound 20.5 The disulfide moiety was reduced with dithiothreitol (DTT), and the crude product 36 was then treated with an excess of peptide 2 at pH ∼ 10. Approximately 3 mg of peptide-dinucleotide conjugate 1 was isolated by reverse-phase HPLC7 and characterized by mass-spec.8 CONCLUSIONS

We have synthesized the first of a series of peptidedinucleotide conjugates that will be used to study chromium(III)-DNA binding. The isolation and characterization of conjugate 1 validates our synthetic approach. Our continuing efforts are aimed at the gram-scale synthesis of 3 and its corresponding conjugates with peptides 2, 7, 8, 12, 13, and 14. With sufficient quantities of our peptide-dinucleotide conjugates in hand, we will then begin to explore their interaction with Cr(III). prepared on a gram scale in over 90% yield from 15. The coupling of 16 and 17 was accomplished with adamantanecarbonyl chloride (AdCOCl) in a mixture of acetonitrile and pyridine. The timing for this reaction is critical due to the reactivity of the acylated phosphonate intermediate. H-Phosphonate 16 is treated with AdCOCl and pyridine under anhydrous conditions for 25 s and is then immediately quenched with an anhydrous solution of 17 in acetonitrile. The bis-acyl phosphite (27) and trimetaphosphite (28) byproducts from this reaction continue to make the scale-up difficult. Currently, we have been able to produce dinucleotide 18 in 65% yield on a 500 mg scale. This dinucleotide was characterized by mass-spec2 and 2D NMR (COSY). The functionalization of the H-phosphonate moiety in 18 was accomplished through oxidation with CCl4 in the presence of cystamine (Scheme 5). The terminal amine was subsequently acetylated to facilitate its isolation and purification by normal-phase chromatography (0-3% MeOH in CHCl3). Compound 19 was isolated in 46% yield on a 500 mg scale, and characterized by mass-spec3 and 2 Compound 18 mass-spec: monoisotopic mass M ) 1064.4, calculated for C51H57N12O14P.; Found MH+ at 1065, MNa+ at 1087, and [M - H]- at 1063. 3 Compound 19 mass-spec: monoisotopic mass M ) 1256.4, calculated for C57H69N12O15PS2; Found MH+ at 1257, MNa+ at 1279, and [M - H]- at 1255. 4 Reversed-phase HPLC: Varian Dynamax C-18 column, 41.4 × 250 mm, 8 µm, 100 Å; solvent A: 50 mM HCO2NH4 in water (pH ∼ 6); solvent B: 90/10 acetonitrile/Solvent A.; gradient: 0-20% solvent B over 20 min at 50 mL/min; UV detection at 220 nm. 5 Compound 20 mass-spec: monoisotopic mass M ) 730.2, calculated for C24H35N12O9PS2; Found MH+ at 731, MNa+ at 753, and [M - H]- at 729. 6 Compound 3 mass-spec: monoisotopic mass M ) 655.2, calculated for C22H30N11O9PS.; Found MH+ at 656 and [M H]- at 654. 7 Reversed-phase HPLC: Varian Dynamax C-18 column, 41.4 × 250 mm, 8 µm, 100 Å; solvent A: 50 mM HCO2NH4 in water (pH ∼ 6); solvent B: 90/10 acetonitrile/solvent A.; gradient: 5-15% solvent B over 40 min at 50 mL/min; UV detection at 220 nm. 8 Compound 1 mass-spec: monoisotopic mass M ) 914.2, calculated for C31H43N14O15PS.; Found MNa+ at 937 and [M H]- at 913.

ACKNOWLEDGMENT

This work was supported by NIH Grant Number CA 75298 (D.M.S.), an NIH Minority Supplement (R.G.L.), the NIH Minority Student Development Program of Northern Arizona University (R.G.L.), and the Intramural Grants Program of Northern Arizona University. LITERATURE CITED (1) Tung, C.-H., and Stein, S. (2000) Preparation and application of peptide-oligonucleotide conjugates. Bioconjugate Chem. 11, 605-618. (2) Manning, F. C. R., Xu, J., and Patierno, S. R. (1992) Transcriptional inhibition by carcinogenic chromate: relationship to DNA damage. Mol. Carcinog. 6, 270-279. (3) Zhitkovich, A., Voitkun, V., and Costa, M. (1995) Glutathione and free amino acids form stable complexes with DNA following exposure of intact mammalian cells to chromate. Carcinogenesis 16, 907-913. (4) Cupo, D. Y., and Wetterhahn, K. E. (1985) Binding of chromium to chromatin and DNA from liver and kidney of rats treated with sodium dichromate and chromium(III) chloride in vivo. Cancer Res. 45, 1146-1151. (5) Hamilton, J. W., and Wetterhahn, K. E. (1986) Chromium(VI)-induced DNA damage in chick embryo liver and red blood cells in vivo. Carcinogenesis 7, 2085-2088. (6) Zhitkovich, A., Voitkun, V., Kluz, T., and Costa, M. (1998) Utilization of DNA-protein cross-links as a biomarker of chromium exposure. Environ. Health Perspect. 106 (Suppl 4), 969-974. (7) Voitkun, V., Zhitkovich, A., and Costa, M. (1998) Cr(III)mediated cross-links of glutathione or amino acids to the DNA phospate backbone are mutagenic in human cells. Nucleic Acids Res. 26, 2024-2030. (8) Zhitkovich, A., Shrager, S., and Messer, J. (2000) Reductive metabolism of Cr(VI) by cysteine leads to the formation of binary and ternary Cr-DNA adducts in the absence of oxidative DNA damage. Chem. Res. Toxicol. 13, 1114-24. (9) Tsapakos, M. J., and Wetterhahn, K. E. (1983) The interaction of chromium with nucleic acids. Chem.-Biol. Interact. 46, 265-277. (10) Borges, K. M., and Wetterhahn, K. E. (1989) Chromium cross-links glutathione and cysteine to DNA. Carcinogenesis 10, 2165-2168. (11) Bridgewater, L. C., Manning, F. C. R., Woo, E. S., and Patierno, S. R. (1994) DNA polymerase arrest by adducted trivalent chromium. Mol. Carcinog. 9, 122-133.

Communications (12) Bridgewater, L. C., Manning, F. C. R., and Patierno, S. R. (1994) Base-specific arrest of in vitro DNA replication by carcinogenic chromium: relationship to DNA interstrand cross-linking. Carcinogenesis 15, 2421-2427. (13) Xu, J., Bubley, G. J., Detrick, B., Blankenship, L. J., and Patierno, S. R. (1996) Chromium(VI) treatment of normal human lung cells results in guanine-specific DNA polymerase arrest, DNA-DNA cross-links and S-phase blockade of cell cycle. Carcinogenesis 17, 1511-1517. (14) Bridgewater, L. C., Manning, F. C., and Patierno, S. R. (1998) Arrest of replication by mammalian DNA polymerases alpha and beta caused by chromium-DNA lesions. Mol. Carcinog. 23, 201-206. (15) Arakawa, H., Ahmad, R., Naoui, M., Tajmir-Riahi, H. A. (2000) A comparative study of calf thymus DNA binding to Cr(III) and Cr(VI) ions. Evidence for the guanine N-7chromium-phosphate chelate formation. J. Biol. Chem. 275, 10150-10153. (16) Sherman, S. E., Gibson, D., Wang, A. H.-J., and Lippard, S. J. (1988) Crystal and molecular structure of cis-[Pt(NH3)2{d(pGpG)}], the principal adduct formed by cis-diamminedichloro-platinum(II) with DNA. J. Am. Chem. Soc. 110, 7368-7381. (17) Gao, Y.-G., Sriram, M., and Wang, A. H.-J. (1993) Crystallographic studies of metal ion-DNA interactions: different binding modes of cobalt(II), copper(II) and barium(II) to N7 of guanines in Z-DNA and a drug-DNA complex. Nucl. Acids Res. 21, 4093-4101. (18) Shih, H. C., Kassahun, H., Burrows, C. J., and Rokita, S. E. (1999) Selective association between a macrocyclic nickel complex and extrahelical guanine residues. Biochemistry 38, 15034-15042. (19) Voitkun, V., Zhitkovich, A., and Costa, M. (1994) Complexing of amino acids to DNA by chromate in intact cells. Environ. Health Perspect. 102 (Suppl 3) 251-255.

Bioconjugate Chem., Vol. 12, No. 4, 2001 463 (20) Zhitkovich, A., Voitkun, V., and Costa, M. (1996) Formation of the amino acid-DNA complexes by hexavalent and trivalent chromium in vitro: importance of trivalent chromium and the phosphate group. Biochemistry 35, 7285-7282. (21) Wachter, L., Jablonski, J.-A., and Ramachandran, K. L. (1986) A simple and efficient procedure for the synthesis of 5′-aminoalkyl oligodeoxynucleotides. Nucl. Acids Res. 14, 7985-7994. (22) MacMillan, A. M., and Verdine, G. L. (1991) Engineering tethered DNA molecules by the convertible nucleoside approach. Tetrahedron Lett. 47, 2603-2616. (23) General review: Rolf, G., and Koenig, W. (1981) Amine protecting groups in peptide synthesis. Peptides 3, 70-81. (24) Fidanza, J. A., and McLaughlin, L. W. (1992) Use of a thiol tether for the site-specific attachment of reporter groups to DNA. J. Org. Chem. 57, 2340-2346. (25) Misra, H. K., Naseema, N. K., Agrawal, S., and Wright, G. E. (1992) Chemical and enzymatic incorporation of N2-(pn-butylphenyl)-2′-deoxyguanosine into an oligodeoxyribonucleotide. Nucleic Acids Res. 20, 4547-4551. (26) Froehler, B. C., and Ng, P. C. (1986) Synthesis of DNA via deoxynucleoside H-phosphonate intermediates. Nucleic Acids Res. 14, 5399-5407. (27) de Vroom, E., Spierenburg, M. L., Dreef, C. E., van der Marel, G. A., and van Boom, J. H. (1987) A general procedure to convert H-phosphonate mono- or diesters of nucleic acids into valuable phosphate di- or triesters. Recl. Trav. Chim. Pays-Bas 106, 65-66. (28) Garegg, P. J., Stawinski, J., and Stro¨mberg, R. (1987) Nucleoside H-phosphonates. 8. Activation of hydrogen phosphonate monoesters by chlorophosphates and arenesulfonyl derivatives. J. Org. Chem. 52, 284-287.

BC0100105