A highly water-soluble platinum(II) complex containing a thiopropyl-1,2-dicarba-closo-dodecaborane(12) ligand functionalised with a pendant glycerol group

Abstract

Reaction of the novel thiopropyl-closo-1,2-carborane ligand bearing a pendant glycerol group HS(CH₂)₃CB₁₀H₁₀CCH₂OCH(CH₂OH)₂ (L) with the labile platinum(II) precursor [Pt(MeCN)(terpy)](OTf)₂ (terpy = 2,2′:6′,2″-terpyridine; OTf = trifluoromethanesulfonate) affords the highly water-soluble platinum(II) complex [PtL(terpy)]OTf, the first example of a metal–carborane complex functionalised with a water-solubilising glycerol group.

---

The effective delivery of boron agents containing dicarba-closo-dodecaboranes(12) (closo-carboranes) to tumour sites for Boron Neutron Capture Therapy (BNCT)^1,2 is highly problematic due to their poor solubility in aqueous media.^3,4 The success of BNCT in part relies on large cellular concentrations of boron which can be difficult to achieve with the hydrophobic closo-carborane structure.⁵ Two main approaches have been explored for solubilising the carborane cage in aqueous solution. These include the conversion of closo-1,2-carborane to the corresponding, hydrophilic nido-7,8-carborane structure by reaction with a strong base or fluoride ion, or the use of polar, hydrophilic functional groups such as cascade polyols which are usually derived from glycerol.^3,5 Previous studies have demonstrated that nido-carborane structures have been found to possess good water solubility characteristics in vivo although their ionic nature can result in non-specific protein binding.^3–5 In contrast, cascade polyols contain polar hydrophilic functional groups which allow for the preparation of non-ionic, water-soluble carborane derivatives,³ and the uptake of boron into cancer cells is usually enhanced.⁴ Furthermore, the number of hydroxyl groups in cascade polyols can be managed at will and thus a systematic variation in the aqueous solubility of the carborane cage can be attained.⁴ Cascade polyols can also be designed to possess no asymmetric centres,^3,6 such that no diastereomers are formed when attached to the 1,2-carborane cage.³ This factor alone facilitates the synthesis and isolation of functionalised carboranes with enhanced aqueous solubility and avoids the many problems associated with the use of natural sugars such as glucose as water-solubilising elements.^6,7

We have demonstrated that platinum(II)–carborane complexes are a novel means of delivering boron atoms close to DNA,^8–10 an important criterion in the development of new classes of BNCT agents.^1,2 Although derivatives such as 1 have demonstrated binding to calf-thymus DNA at high complex: DNA base-pair molar ratios (r_f > 2.5) in vitro, possibly by an intercalative mechanism, unfavourable steric interactions between the carborane moiety and DNA in addition to the poor solubility of the complex in an aqueous environment were likely to account for the relatively high r_f values observed.¹⁰ The synthesis of the glycerol derivative 5 was embarked upon as it directly addresses these two main disadvantages associated with complexes such as 1 by (i) increasing the length of the linker chain between the carborane entity and platinum centre in order to prevent the carborane cage interfering with the DNA-binding of the metallointercalator, and (ii) attaching a pendant glycerol group to the carborane moiety in order to increase the hydrophilicity of the complex and also act a potential scaffold for the preparation of larger cascade-type polyol derivatives.

Malmquist and Sjöberg have investigated a related route to the one described here for the synthesis of water-soluble aminoalkyl-closo-1,2-carborane derivatives,¹¹ but no reports of the metal complexation behaviour of these species have been reported to date. Herein we report the preparation of a novel thiolpropyl-closo-1,2-carborane ligand bearing a glycerol functionality and its 2,2′:6′,2″-terpyridine–platinum(II) derivative, the first example of a metal–carborane complex functionalised with a water-solubilising glycerol group.

The precursor bromopropyl-closo-1,2-carborane derivative bearing a pendant benzyl-protected glycerol group 2 was prepared in five steps by an adaptation of the methods described by Malmquist and Sjöberg¹¹ and Yamamoto and co-workers.³ The S-benzyl derivative 3 was readily prepared from 2 by treatment with sodium benzylthiolate to afford the desired product in an 80% yield. Care was taken in this reaction due to the known partial degradation of the closo-1,2-carborane cage under strongly basic conditions.^12,13 Thus an excess of benzyl mercaptan was first added to the NaOEt solution in order to ensure total consumption of the ethoxide ion, thereby leaving only the less basic but nucleophilic benzylthiolate anion in solution. In the ¹H NMR spectrum the downfield shift of the α-methylene multiplet from the corresponding resonance in 2 (ca. 0.2 ppm) and the presence of the single benzylic methylene peak at δ 3.83 confirmed that the nucleophilic substitution reaction had occurred successfully. The formation of the product was also supported by ¹³C{¹H} NMR spectroscopy which showed the presence of a new benzylic methylene peak at δ 36.2. Resonances at δ −4.1 and −11 in ¹¹B{¹H} NMR spectrum demonstrated that no degradation of the closo-carborane cage had occurred during the reaction.

The benzyl group is commonly used to protect hydroxyl groups,¹⁴ and we have found that thiols are also readily protected by this group,¹⁵ although the usual deprotection conditions employing Pd catalysts and H₂ are not feasible owing to the presence of the thiol group. BBr₃ and AlCl₃ were both found to cleave the O-benzyl groups in a facile manner, but BBr₃ was unable to remove the S-benzyl group. Therefore, AlCl₃ was selected as the reagent of choice. All the benzyl-protecting groups in 3 were successfully removed in one step using this reagent to afford the hygroscopic, thiol ligand 4 (Scheme 1). The characterisation of 4 was confirmed by means of ESI-MS, with a molecular ion peak detected at m/z 320 ([M − H]⁻). The ¹H NMR spectrum showed the presence of a triplet signal at δ 1.43 (³J_HH = 8 Hz) due to the SH group and a singlet resonance at δ 1.63 due to the two OH groups which immediately disappeared upon the addition of a drop of D₂O to the NMR solution. The structure of 4 was further supported by its ¹H and ¹³C{¹H} NMR spectra with no peaks observed in the aromatic regions, indicating that benzyl deprotection of both the thiol and alcohol groups had been completed successfully. The ¹¹B{¹H} NMR spectrum showed peaks at δ −3.9 and −10.9, once again demonstrating that the closo-carborane cage remained intact.

Scheme 1 Reagents: (i) NaOEt, BnSH; (ii) AlCl₃/C₆H₆, NaK-tartrate; (iii) [Pt(MeCN)(terpy)](OTf)₂.

The use of [Pt(MeCN)(terpy)](OTf)₂ (terpy = 2,2′:6′,2″-terpyridine; OTf = trifluoromethanesulfonate) in the synthesis of the target complex 5 allowed for the highly labile acetonitrile ligand to be displaced easily when treated with the carborane ligand.^15,16 Thus, the addition of 4 to [Pt(MeCN)(terpy)](OTf)₂ in acetone solution afforded 5 in a reaction that was accompanied by an immediate colour change from pale-yellow to orange–red (Scheme 1). Complex 5 is stable at room temperature but it is quite hygroscopic due to the hydrophilic nature of the glycerol moiety. The preparation of 5 was confirmed by means of multinuclear (¹H, ¹³C, ¹¹B, and ¹⁹⁵Pt) NMR spectroscopy, ESI-MS, and microanalysis. The appearance of a signal at δ −3072 in the ¹⁹⁵Pt{¹H} NMR spectrum confirms the presence of a PtN₃S core.^10,17 Despite repeated attempts, suitable crystals of 5 could not be obtained for X-ray diffraction owing to its highly hygroscopic nature. In contrast to 1, which displays negligible aqueous solubility, the solubility of 5 in deionised water was found to be approximately 7 g L⁻¹ at room temperature. This figure is significantly greater than the glycerol derivative of aminopropyl-1,2-carborane (0.7 g L⁻¹);¹¹ the substantial increase in the water solubility of 5 is probably related to its cationic nature in addition to the presence of the glycerol group.

In conclusion, the functionalisation of a metal–carborane complex with a glycerol moiety is a feasible strategy for increasing its aqueous solubility. It is now possible to investigate the further functionalisation of the carborane cage using higher-order, cascade-type polyols such as tetraols,³ which can be prepared in a similar manner to the parent glycerol ligand 4, thus leading to metal–carborane complexes with even greater aqueous solubility than 5. We are also in the process of determining the DNA-binding, tumour cell uptake characteristics, and the efficacy of 5 in the presence of thermal neutrons. The results of this work will be reported in due course.

Acknowledgements

We thank Dr I. Luck (University of Sydney) for recording the ¹¹B{¹H} and ¹⁹⁵Pt{¹H} NMR spectra, Dr K. Fisher (University of Sydney) for recording the ESI-MS, and Dr M. Coster (University of Sydney) for useful advice. We are grateful to Johnson Matthey for the generous loan of platinum salts. Finally, we thank the Australian Research Council (ARC) for financial support.

References

1. M. F. Hawthorne, Angew. Chem., Int. Ed., 1993, 32, 950.
2. A. H. Soloway, W. Tjarks, B. A. Barnum, F. G. Rong, R. F. Barth, I. M. Codogni and J. G. Wilson, Chem. Rev., 1998, 98, 1515.
3. H. Nemoto, J. G. Wilson, H. Nakamura and Y. Yamamoto, J. Org. Chem., 1992, 57, 435.
4. Y. Yamamoto and N. Sadayori, Bioorg. Med. Chem. Lett., 1996, 6, 9.
5. B. C. Das, G. W. Kabalka, R. R. Srivastava, W. Bao, S. Das and G. Li, J. Organomet. Chem., 2000, 614–615, 255.
6. H. Nemoto, J. Cai, S. Iwamoto and Y. Yamamoto, J. Med. Chem., 1995, 38, 1673.
7. G. Livesey, Nutr. Res. Rev., 2003, 16, 163.
8. S. L. Woodhouse and L. M. Rendina, Chem. Commun., 2001, 2464.
9. S. L. Woodhouse and L. M. Rendina, Dalton. Trans., 2004, 3669.
10. J. A. Todd and L. M. Rendina, Inorg. Chem., 2002, 41, 3331.
11. J. Malmquist and S. Sjöberg, Acta Chem. Scand., 1994, 48, 886.
12. V. I. Bregadze, Chem. Rev., 1992, 92, 209.
13. R. N. Grimes, Carboranes, Academic Press, New York, 1970.
14. T. W. Greene and P. G. M. Wuts, Protective Groups In Organic Synthesis, John Wiley & Sons, New York, 3rd edn., 1999, p. 17ff.
15. J. A. Todd, P. Turner, E. J. Ziolkowski and L. M. Rendina, Inorg. Chem., in press.
16. R. Büchner, J. S. Field, R. J. Haines, C. T. Cunningham and D. R. McMillin, Inorg. Chem., 1997, 36, 3952.
17. A. K. Fazlur-Rahman and J. G. Verkade, Inorg. Chem., 1992, 31, 2064.

---

Footnotes

† Electronic supplementary information (ESI) available: Experimental details of the synthesis and characterisation of 2–5. See http://dx.doi.org/10.1039/b508020g

‡ Present address: The Heart Research Institute, Camperdown, Sydney, NSW, 2050, Australia.

---