Abstract
Boron clusters have been employed successfully as constituents in bioactive substances. In this review, the perspectives of boron clusters for drug design and problems to be solved for a broader application are discussed, and a list of actions is given for overcoming the problems.
Introduction
Boron clusters have properties (including delocalized charge, hydridic hydrogens, isomerism) which are not found in conventional drugs. This changes their interaction potential with biologically relevant molecules. For dodecaborate clusters in water, dihydrogen bonds are formed, and the energy of removing a water molecule from the weakly ordered hydration shell is small [1]. With lipid bilayers, binding of dodecaborate clusters to the surface (resulting in a strong negative zeta potential), changes in the morphology of liposomes, and leakage of liposomal contents is observed [2–4]. The cobaltabisdicarbollide (COSAN) penetrates lipid bilayers and also leads to morphological changes [5]. The interaction of this cluster with amino acids can be used for sensing [6]. When applied to chromatography columns consisting of cross-linked carbohydrates such as superdex and sepharose, strong retention of dodecaborate clusters is observed [7].
Icosahedral boron clusters, such as the three isomeric forms of dicarba-carborane, the monocarbon carborane, and the dodecaborate have about the same size as a rotating phenyl ring or an adamantane (Fig. 1).
Clusters discussed (top); space-filling models of benzene, dodecaborate, and adamantane (bottom). In the structures on the top, gray spheres represent BH units, black ones CH units, while the red one is a metal ion; all H atoms are omitted for clarity.
Several examples of drug-like compounds with boron clusters are known. Some are listed in Table 1.
Literature examples of pharmacologically active boron cluster compounds.
| Target | Effect | Refs. | Remarks |
|---|---|---|---|
| HIV protease | Inhibition | [8–10] | |
| Carbonic anhydrase | Inhibition | [11] | |
| Estrogen receptor | Modulation | [12] | in vivo activity |
| P2X7 receptor | Antagonist | [13] | in vivo activity |
| Nicotinamide phosphoryltransferase | Inhibition | [14] | |
| Cyclooxygenase | Inhibition | [15] | |
| Purinergic receptor A2A | Modulation | [16] | |
| Hypoxia induced factor | Inhibition | [17] | |
| Vitamin D receptor | Agonist | [18] |
Common motifs in cluster interactions
A few common aspects in interactions of boron clusters with biologically relevant molecules can be deduced from the data in the literature.
One aspect is the fact that H atoms on boron clusters (independent of size and possible heteroatoms in the cluster) carry a partial negative charge, and are thus hydridic H atoms [19]. This enables them to form interactions with protic H atoms in organic molecules, bonds which are called dihydrogen bonds [20, 21].
Another aspect is that, although the size of an icosahedral boron cluster is about that of a rotating phenyl ring (see Fig. 1), it extends to the same extent also on the face of the phenyl group where π electrons would normally be available for π–π stacking interactions or interactions with Lewis acids.
Both of these aspects are unique to boron clusters, as compared to organic molecules. Adamantane, which occupies a similar space as a boron cluster (see Fig. 1), has none of these properties.
Despite the similarities between the clusters, there are considerable differences amongst them. Depending on the heteroatoms present in the cluster, they carry a double negative charge (parent compound B12H122-), a single negative charge (CB11H121-), or no charge (C2B10H12). The base-degraded o-carborane C2B9H121- (nido-carborane), a trunctuated icosahedron, carries a single negative charge. The negative charges are distributed on all cluster atoms and the attached H atoms, and thus every atom carries only a fractional charge. The three neutral carborane clusters (o-, m-, p-) also show differences among themselves which translate into different biological acitivy.
Different clusters are different
Drug-like molecules with different clusters have been developed for several specific targets. Among these are nicotinamide phosphoryltransferase (NPT) [14], carbonic anhydrase (CA) [11], and cyclooxygenase (COX) [15].
For NPT, the benzoylpiperidinyl unit of the starting compound FK866 known from the literature [22] was replaced with an adamantyl residue, or with one of the three isomeric carborane clusters. While the starting compound inhibited growth of three different cell lines at 1.6–3.2 nM concentration, the o- and p-carborane derivates showed values between 0.6 and 1.7 nM. The m-carborane derivative was the best inhibitor for cell growth, with EC50 values of 0.3–0.4 nM, while the adamantane derivative required up to 190 nM for 50 % inhibition. Thus, despite a similar space demand of the equally hydrophobic adamantane, it appears not to be a good mimetic of any of the carboranes.
For COX, two boron analogues of indomethacin with an o- or a m-carboranyl and an adamantyl substituent, replacing a chlorophenyl residue, have been tested against the two isoforms of the enzyme, COX-1 and COX-2. Only the o-carborane derivatives showed any inhibition at concentrations below 25 μM, and inhibition of COX-2 was much more pronounced than that of COX-1.
For CA, derivatives of o- and m-carborane with a sulfamidomethyl group attached to one C atom have been compared, and in this case, the m-isomer requires four times the concentration to inhibit CAIX, the isoform which is a target for cancer therapy [11]. While one derivative of the nido-carborane showed the best inhibition of CAIX and the best selectivity for this isoform over CAII, a complete and systematic exploration of these sulfamides has not yet been carried out. Interestingly, while the 1-sulfamidomethyl-2-phenyl-o-carborane is a very poor inhibitor of both CAIX and CAII, its corresponding nido derivative is more potent by about two orders of magnitude.
Mutant inhibition might be more effective for clusters
A number of pathogenic organisms have a very high ability to mutate in essential genes. This applies to bacteria, viruses, parasites, and to some extent also cancer. As a consequence, drugs rapidly lose their ability to prevent, reduce, or cure infections. Boron cluster compounds have shown to be much less influenced in their activity by such mutations. Cobalta-bis-carbollides have been found to be very potent inhibitors of HIV proteases [8–10]. In the inhibitors, two clusters are linked by a long linker (Fig. 2). The linker spans the distance between the two loops responsible for substrate binding in the dimeric enzyme. It is obvious that conventional hydrogen bonds or ionic interactions (observed for inhibitors such as darunavir) cannot play a major role in the binding of the carbollide inhibitors, as conventional H bonds cannot be formed with the H atoms of the cluster.
![Fig. 2
Inhibitors of HIV proteases (top) and their relative inhibitory constants (bottom) (redrawn from [10] with permission). The KI values relative to that of the wild typ HIV protease (PR) are shown. In the top part, black circles represent C, red circles represents Co. All H atoms are omitted for clarity. R1 and R2 for 1: H,H; 2: Et, Et; 4: OHCH2CH2, H; 5: tBu, H. For mutations in PR1 through PR7, see [10]. Abbreviations: SQV: saquinavir; IDB: indinavir; NFV: nelfinavir; LPV: lopinavir; APV: amprenavir; AZV: atazanavir; DRV: darunavir.](/document/doi/10.1515/pac-2014-1007/asset/graphic/j_pac-2014-1007_fig_002.jpg)
Inhibitors of HIV proteases (top) and their relative inhibitory constants (bottom) (redrawn from [10] with permission). The KI values relative to that of the wild typ HIV protease (PR) are shown. In the top part, black circles represent C, red circles represents Co. All H atoms are omitted for clarity. R1 and R2 for 1: H,H; 2: Et, Et; 4: OHCH2CH2, H; 5: tBu, H. For mutations in PR1 through PR7, see [10]. Abbreviations: SQV: saquinavir; IDB: indinavir; NFV: nelfinavir; LPV: lopinavir; APV: amprenavir; AZV: atazanavir; DRV: darunavir.
It remains to be seen whether this promising observation holds also for other targets with high rates of mutations.
Metabolization of clusters
Boron clusters are foreign to the biochemical world. It is therefore expected that compounds containing a cluster will be metabolized differently, and probably less, than corresponding carbon compounds.
Degradation of the cluster moiety itself will probably be relevant only for o-carborane compounds, leading to nido-carboranes. These are known to undergo loss of a boron atom in the presence of base such as an alcoholate [23] or other bases [24]. Compounds with amino groups might undergo intramolecular degradation in aqueous solution [25]. The corresponding nido clusters might eventually hydrolyse to boric acid.
Other clusters appear not to be degraded in vivo. For B12H11SH2-, oxidation of the SH group to dimers has been observed after administration to patients, with the cluster being intact [26].
Gaps of knowledge
For a wider incorporation of boron clusters into drug development, several obstacles can be seen. One very important obstacle is that boron clusters cannot be categorized in the same way as organic moieties are. Ionic boron clusters are not easily comparable to organic ionic groups. Hydrogen bonding potential is fundamentally different. A categorization into hydrophilic/hydrophobic is not readily possible.
Interaction potentials
Water
The interaction of some dodecaborate clusters with water has been studied. Experimental octanol/water distribution coefficients [27] were used to derive appropriate force fields for molecular dynamics (MD) simulations [1]. While the interaction of the parent cluster B12H122- shows some ordering of water around the cluster (mostly through dihydrogen bonds), ammonia-substituted clusters show no defined structure, and based on the MD results, the energy required to remove water from the immediate surrounding of boron is minimal.
For other clusters, such data are not available. This pertains to the three uncharged isomers of carborane, to the nido-carborane, and to the monocarbon carborane. Derivatives of the latter have been used in halogenated forms as weakly coordinating ion [28].
Organic molecules
For the interaction of boron clusters with organic molecules, such as proteins, lipids, carbohydrates, only experimental observations are available. These show that interaction does not follow expectations. For instance, dodecaborate clusters were found to be strongly retained on carbohydrate-containing chromatography matrices [7], while their interaction with silica in the same aqueous buffer was minimal [29]. Liposomes are influenced in morphology and tightness [2, 4], and metalla-bis(carbollides) were shown to penetrate lipid bilayers [5].
For proper modelling of boron clusters in contact with proteins, further knowledge and a deeper understanding of these phenomena are required.
Chemistry of clusters
Dicarba-carboranes
For o-carborane, a well-developed chemistry is available for functionalization at the C atoms, and at any of the B atoms. This is not so for the two other isomers, m- and p-carborane. Here, the available reactions described are limited, which might hinder considerably their incorporation into drugs.
Monocarba-Carboranes
For the CB11 cluster, methods for C-functionalization are well-known [30]. Functionalization at the B atom in para-position to the C atom and a few other modification reactions are described [31], but the possibilities for functionalization are still limited.
Dodecaborates
B12H122- is readily mono-substituted with oxygen (especially through oxonium ions), nitrogen (hydroxylamine-O-sulfonic acid) and halogens (through elemental halogens). Methods for direct introduction of carbon substituents are not readily available, but the cross-coupling reaction of monoiodo-dodecaborate with Grignard reagents is a useful reaction [32]. Microwave mediation increases the yields of such reactions (including those with oxonium derivatives) considerably [33].
There is little systematic investigation of the disubstitution of this cluster. The introduction of a second substituent appears to be more difficult than that of a first substituent, and it is introduced preferentially in the 7-position (corresponding to the meta-position in benzene).
A few reports on trisubstituted clusters are available [34]. Such derivatives allow to compensate one or both of the negative charges of the dodecaborate.
Drug discovery tools
Docking programs
Docking programs are widely used for drug discovery. Different approaches exist, where interactions between drug and protein are dealt with in different levels of precision. Only two groups have applied such programs to boron clusters [35, 36].
In both approaches, the lack of appropriate descriptors for interaction potentials of boron (and the attached hydrogen) has led the authors of both articles to substitute a BH group for a CH group. With this, a boron cluster is almost treated as if it were an adamantane (see Fig. 1). This is most probably not appropriate, as seen by the big difference in activity between an adamantane and a boron cluster inhibitor of NPT [14], and the strong differentiation between o- and m-carborane for COX-2 inhibitors [37].
Libraries
Libraries of compounds are a well-used tool for drug discovery. For boron cluster compounds, a complete library of compounds will probably never be available, smaller libraries such as the ones for adamantanes [38] might be achievable. It might be a more practical approach to use a library for fragment-based drug discovery with less complexity, and with only a limited number of individual entries [39]. The complexity score for boron clusters should, however, reflect the nature of the clusters, with 12 heavy atoms.
Membrane permeation
As noted above, some boron clusters interact with membranes, and appear to be able to penetrate membranes [3, 5, 40]. Oral availability is crucially dependent on membrane permeation. No systematic study of this property of boron clusters has been published so far, although it is the first step in the ADMET or ADMETox (absorption, distribution, metabolism, excretion and toxicity) concept of drug action.
Conclusions
For the acceptance of boron-based pharmacophores, a number of tools and technologies are required in synthetic chemistry, computational chemistry, biochemistry, drug screening, pharmacology, and medicinal chemistry. With this, it will eventually be possible to produce a wider range of new molecules with unique biological properties and clinical applications. A coordinated effort toward this goal appears necessary, and offers rewarding prospects.
Article Note
A collection of invited papers based on presentations at the 15th International Meeting on Boron Chemistry (IMEBORON-XV), Prague, Czech Republic, 24–28 August 2014.
References
[1] K. Karki, D. Gabel, D. Roccatano. Inorg. Chem. 51, 4894 (2012).Search in Google Scholar
[2] D. Gabel, D. Awad, T. Schaffran, D. Radovan, D. Daraban, L. Damian, M. Winterhalter, G. Karlsson, K. Edwards. ChemMedChem. 2, 51 (2007).Search in Google Scholar
[3] D. Awad, L. Damian, M. Winterhalter, G. Karlsson, K. Edwards, D. Gabel. Chem. Phys. Lipids 157, 78 (2009).10.1016/j.chemphyslip.2008.11.006Search in Google Scholar
[4] T. Schaffran, J. Li, G. Karlsson, K. Edwards, M. Winterhalter, D. Gabel. Chem. Phys. Lipids 163, 64 (2010).10.1016/j.chemphyslip.2009.09.004Search in Google Scholar
[5] C. Verdia-Baguena, A. Alcaraz, V. M. Aguilella, A. M. Cioran, S. Tachikawa, H. Nakamura, F. Teixidor, C. Vinas. Chem. Commun. 50, 6700 (2014).Search in Google Scholar
[6] A.-I. Stoica, C. Viñas, F. Teixidor. Chem. Commun. 4988 (2009).10.1039/b910645fSearch in Google Scholar
[7] P. Fan, J. Neumann, S. Stolte, J. Arning, D. Ferreira, K. Edwards, D. Gabel. J. Chromatogr. A 1256, 98 (2012).10.1016/j.chroma.2012.07.055Search in Google Scholar
[8] P. Cigler, M. Kozisek, P. Rezacova, J. Brynda, Z. Otwinowski, J. Pokorna, J. Plesek, B. Gruner, L. Doleckova-Maresova, M. Masa, J. Sedlacek, J. Bodem, H. G. Krausslich, V. Kral, J. Konvalinka. Proc. Natl. Acad. Sci. USA 102, 15394 (2005).Search in Google Scholar
[9] M. Kozisek, P. Cigler, M. Lepsik, J. Fanfrlik, P. Rezacova, J. Brynda, J. Pokorna, J. Plesek, B. Gruner, K. Grantz Saskova, J. Vaclavikova, V. Kral, J. Konvalinka. J. Med. Chem. 51, 4839 (2008).Search in Google Scholar
[10] P. Rezacova, J. Pokorna, J. Brynda, M. Kozisek, P. Cigler, M. Lepsik, J. Fanfrlik, J. Rezac, K. Grantz Saskova, I. Sieglova, J. Plesek, V. Sicha, B. Gruner, H. Oberwinkler, J. Sedlacek, H.G. Krausslich, P. Hobza, V. Kral, J. Konvalinka. J. Med. Chem. 52, 7132 (2009).Search in Google Scholar
[11] J. Brynda, P. Mader, V. Šícha, M. Fábry, K. Poncová, M. Bakardiev, B. Grüner, P. Cígler, P. Řezáčová. Angew. Chem. Int. Ed. 52, 13760 (2013).Search in Google Scholar
[12] M. Hirata, M. Inada, C. Matsumoto, M. Takita, T. Ogawa, Y. Endo, C. Miyaura. Biochem. Biophys. Res. Commun. 380, 218 (2009).Search in Google Scholar
[13] S. M. Wilkinson, H. Gunosewoyo, M. L. Barron, A. Boucher, M. McDonnell, P. Turner, D. E. Morrison, M. R. Bennett, I. S. McGregor, L. M. Rendina, M. Kassiou. ACS Chem. Neurosci. 5, 335 (2014).Search in Google Scholar
[14] M. W. Lee, Y. V. Sevryugina, A. Khan, S. Q. Ye. J. Med. Chem. 55, 7290 (2012).Search in Google Scholar
[15] M. Scholz, A. L. Blobaum, L. J. Marnett, E. Hey-Hawkins. Bioorg. Med. Chem. 19, 3242 (2011).Search in Google Scholar
[16] K. Bednarska, A. B. Olejniczak, A. Piskala, M. Klink, Z. Sulowska, Z. J. Lesnikowski. Bioorg. Med. Chem. 20, 6621 (2012).Search in Google Scholar
[17] H. S. Ban, K. Shimizu, H. Minegishi, H. Nakamura. J. Am. Chem. Soc 132, 11870 (2010).10.1021/ja104739tSearch in Google Scholar
[18] S. Fujii, H. Masuno, Y. Taoda, A. Kano, A. Wongmayura, M. Nakabayashi, N. Ito, M. Shimizu, E. Kawachi, T. Hirano, Y. Endo, A. Tanatani, H. Kagechika. J. Am. Chem. Soc. 133, 20933 (2011).Search in Google Scholar
[19] P. Farras, N. Vankova, L. L. Zeonjuk, J. Warneke, T. Dülcks, T. Heine, C. Viñas, F. Teixidor, D. Gabel. Chemistry 18, 13208 (2012).10.1002/chem.201200828Search in Google Scholar
[20] P. E. M. Siegbahn, O. Eisenstein, A. L. Rheingold, T. F. Koetzle. Acc. Chem. Res. 29, 348 (1996).Search in Google Scholar
[21] J. G. Planas, C. Viñas, F. Teixidor, A. Comas-Vives, G. Ujaque, A. Lledós, M. E. Light, M. B. Hursthouse. J. Am. Chem. Soc. 127, 15976 (2005).Search in Google Scholar
[22] M. Hasmann, I. Schemainda. Cancer Res. 63, 7436 (2003).Search in Google Scholar
[23] M. F. Hawthorne, D. C. Young, P. M. Garrett, D. A. Owen, S. G. Schwerin, F. N. Tebbe, P. A. Wegner. J. Am. Chem. Soc. 90, 862 (1968).Search in Google Scholar
[24] M. A. Fox, W. R. Gill, P. L. Herbertson, J. A. H. MacBride, K. Wade, H. M. Colquhoun. Polyhedron 15, 565 (1996).10.1016/0277-5387(95)00297-6Search in Google Scholar
[25] E. Svantesson, J. Pettersson, Å. Olin, K. Markides, S. Sjöberg. Acta Chem. Scand. 53, 731 (1999).Search in Google Scholar
[26] P. L. Mauri, F. Basilico, P. G. Pietta, E. Pasini, D. Monti, W. Sauerwein. J. Chromatogr. B 788, 9 (2003).10.1016/S1570-0232(02)01006-1Search in Google Scholar
[27] T. Schaffran, E. Justus, M. Elfert, T. Chen, D. Gabel. Green Chem. 11, 1458 (2009).Search in Google Scholar
[28] C. A. Reed, K. C. Kim, R. D. Bolskar, L. J. Mueller. Science 289, 101 (2000).10.1126/science.289.5476.101Search in Google Scholar PubMed
[29] P. Fan, S. Stolte, D. Gabel. Anal. Methods 6, 3045 (2014).10.1039/C4AY00119BSearch in Google Scholar
[30] C. Douvris, J. Michl. Chem. Rev. 113, PR179 (2013).10.1021/cr400059kSearch in Google Scholar PubMed
[31] M. Hailmann, L. Herkert, A. Himmelspach, M. Finze. Chemistry Eur. J. 19, 15745 (2013).Search in Google Scholar
[32] T. Peymann, C. B. Knobler, M. F. Hawthorne. Inorg. Chem. 37, 1544 (1998).Search in Google Scholar
[33] A. Himmelspach, G. J. Reiss, M. Finze. Inorg. Chem. 51, 2679 (2012).Search in Google Scholar
[34] E. J. M. Hamilton, H. T. Leung, R. G. Kultyshev, X. Chen, E. A. Meyers, S. G. Shore. Inorg. Chem. 51, 2374 (2012).Search in Google Scholar
[35] R. Tiwari, K. Mahasenan, R. Pavlovicz, C. Li, W. Tjarks. J. Chem. Inf. Model 49, 1581 (2009).10.1021/ci900031ySearch in Google Scholar PubMed PubMed Central
[36] M. Calvaresi, F. Zerbetto. J. Chem. Inf. Model 51, 1882 (2011).10.1021/ci200216zSearch in Google Scholar PubMed
[37] M. Scholz, A. L. Blobaum, L. J. Marnett, E. Hey-Hawkins. Bioorg. Med. Chem. 20, 4830 (2012).Search in Google Scholar
[38] N. Basarić, M. Sohora, N. Cindro, K. Mlinarić-Majerski, E. De Clercq, J. Balzarini. Arch. Pharm. 347, 334 (2014).Search in Google Scholar
[39] A. R. Blaazer, K. M. Orrling, A. Shanmugham, C. Jansen, L. Maes, E. Edink, G. J. Sterk, M. Siderius, P. England, D. Bailey, I. J. P. de Esch, R. Leurs. J. Biomolecul. Screen. (2014).Search in Google Scholar
[40] T. I. Rokitskaya, A. V. Zaitsev, V. A. Ol’shevskaya, V. N. Kalinin, M. M. Moisenovich, I. I. Agapov, Y. N. Antonenko. Biochemistry Moscow 77, 975 (2012).10.1134/S0006297912090039Search in Google Scholar PubMed
©2014 IUPAC & De Gruyter
