Biological Applications of Fullerene Derivatives: A Brief Overview
Tatiana Da Ros,* Giampiero Spalluto, and Maurizio Prato
Dipartimento di Scienze Farmaceutiche, Università di Trieste, Piazzale Europa 1, 34127 Trieste, Italy
Received February 8, 2001; revised April 27, 2001; accepted May 4, 2001
Starting soon after the production of fullerenes in 1990, many ef¬forts have been devoted to the application of C60 and its deriva¬tives. In fact, [60]fullerene possesses a variety of interesting biolog¬ical properties, such as HIV-P inhibition, DNA photocleavage, neuroprotection, apoptosis, etc. Unfortunately, the low solubility in biological fluids limits the use of these compounds as new pharma-cophores for structure-activity relationship studies in medicinal chemistry. This article briefly summarizes recent studies on the functionalization of C60 aimed at increasing water solubility as well as the preliminary studies performed on biological targets. In particular, the HIV-P inhibition, DNA photocleavage and antibacte¬rial activity are discussed.
In the last few years, fullerene C60 (1) and its derivatives have started to be investigated due to their promising preliminary biological activities, such as DNA photocleavage, HIV-Protease (HIV-P) inhibition, neuroprotection and apoptosis.1–3
Our group has been involved in the synthesis and study of the properties of C60 derivatives. In particular, a novel functionalization of C60 has been developed via cycloaddition of azomethine ylides to fullerene.4,5 The azome-
thine ylides are generated in situ by condensation of a-amino acids and al¬dehydes or ketones. In this way fulleropyrrolidines are obtained with the 5-membered ring fused to a 6,6 bond on the fullerene (Scheme 1).
This reaction is very useful because it is possible to introduce different substituents on nitrogen and on carbons 2 and 5 using different substituted reagents. This class of fullerene derivatives retain the main properties of the parent molecule, such as the ground state absorptions, which extend throughout the visible region up to 700 nm, and the excited state properties.6
In medicinal chemistry, the potential applications of C60 derivatives in¬clude inhibition of HIV-P,7 antibacterial activity,8–10 and photocytotoxicity.11,12 Being a good radical scavenger, this all-carbon molecule might be also used as an anti-apoptotic and/or anti aging agent.13–15 All these interesting possi¬bilities of utilizing fullerenes in biology and medicinal chemistry face a sig¬nificant problem: the natural repulsion of fullerenes to water. Encapsulation of C60 in cyclodextrins16 or in calixarenes17 or water suspension prepara-tions18 are useful methodologies for overcoming this limitation, but the most versatile technique is modifying the solubility properties by covalent attach¬ment of water-soluble appendages, such as dendrimers,19 cyclodextrins20 or calixarenes.21,22
In this review, we briefly summarize the biological profile of some C60 derivatives prepared in our laboratory, especially focusing our attention on the synthetic aspects.
INHIBITION OF HIV-P
Earlier experiments performed by Wudl and coworkers demonstrated that there is inhibition of the HIV-P in the presence of C60.7,23,24 This activ¬ity has been supported by molecular modeling studies, which proved that the fullerene can be accommodated inside the hydrophobic cavity present in the enzyme and its location might prevent the interaction between the cata¬lytic portions of the HIV-P and the virus substrates.
The binding constant found experimentally for a »first generation« inhibitor7 was not significant in terms of affinity (Kd 10–6/10–9 M) but repre¬sented a starting point for further experiments, which would require a structural optimization of C60 derivatives for HIV-P interaction. The cata¬lytic site of HIV-P contains two aspartic residues. A stable interaction with the aspartates could increase the efficiency of the potential inhibitions. On this basis, the same authors proposed an ideal inhibitor 2, in which two am¬monium groups at 5.5 Å distance are directly linked to C60 (Figure 2).
Starting from these experimental and theoretical observations, but us¬ing a different approach, we synthesized C60 derivative 3 in which the dis¬tance between the two ammonium residues was 5.1 Å.25 The synthetic pro¬cedure is based on the cycloaddition of N,N'-Boc-1,3-diamino-2-propanone (4) with sarcosine (5) or N-(3,6,9-trioxadecyl)glycine (6) (Scheme 2). These
products (7 and 8), after deprotection by TFA, afford the final compounds 3 and 9, which are still under investigation for their biological profiles.
Molecular modeling studies on derivatives 3 and 9 showed that they could fit very well inside the HIV-P cavity and the electrostatic interactions can take place between the carboxylic residues of aspartates and the ammo-
nium groups. Figure 3 shows the accommodation of the synthesized com¬pound 3 inside the HIV-P cavity. This effect could contribute to stabilizing the complexation of the enzyme by derivative 3.25
DNA-PHOTOCLEAVAGE
Another potential biological application of C60 is related to the easy photoexcitation of fullerenes. In fact, from the ground state, the fullerene can be excited to 1C60 by photoirradiation. This short-lived species is readily converted to the long-lived 3C60 via intersystem crossing. In the presence of molecular oxygen, the fullerene can decay from its triplet to the ground state, transferring its energy to O2, generating 1O2, known to be a highly cytotoxic species. In addition, the high-energy species 1C60 and 3C60 are excellent ac¬ceptors and, in the presence of a donor, can undergo a different process, be¬ing easily reduced to C60.– by electron transfer. Again, in the presence of ox¬ygen, the fullerene radical anion can transfer one electron, producing O2.–. The excited fullerene can be reduced in the presence of the guanosine resi¬due present into DNA. Hydrolysis of oxidized guanosines followed by DNA cleavage is a consequence of the electron transfer from G to C60*.26
On the other hand, singlet oxygen and superoxide radical anion are well known reactive species towards DNA. The 1O2 in fact modifies G by cycloaddition to the imidazole portion and the resulting modified base is subject to a rapid alkaline hydrolysis of the phosphate bond. Also in this case, the effect is DNA cleavage.
In this field, many fullerene conjugates with different units possessing biological affinity to nucleic acids or proteins might be particularly interest¬ing. In particular, conjugates between C60 and specific agents that interact with nucleic acid, such as acridine,27 netropsin11 or complementary oligonu-cleotides,26,28 have been synthesized with the aim to understand the mecha¬nism of action of this class of conjugates and to increase both cytotoxicity and sequence selectivity. Many fullerene derivatives linked to an inter-calator or a minor groove binder have been reported,11 however, DNA cleav¬age occurs at guanine residues without significant sequence selectivity.29 Only when C60 was conjugate to an oligonucleotide, a good selectivity was observed.28
In this context, with the aim to obtain higher sequence-selectivity, we undertook to prepare a derivative of C60 (10) bearing a minor groove binder and an oligonucleotide sequence. The rational design of derivative 10 is based on a reinforced effect due to the simultaneous presence of two differ¬ent agents able to confer sequence selectivity, such as trimethoxyindole
(TMI) and oligonucleotide. The TMI nucleus is characteristic of a class of natural compounds named duocarmycins (Figure 4), possessing high cytotoxicity (pM range, 72 h of incubation for Leukemia Cells L 1210), and high selectivity for AT rich regions of DNA (Figure 5).30,31
On the other hand, the oligonucleotide chain could increase both the se¬quence-selectivity and water solubility, the biggest problem of C60 deriva¬tives for their biological use.
The synthesis of this derivative started from the ethyl diamine N-Boc-protected (11), which was reacted with benzyl bromoacetate (12) to afford compound 13,32 which after deprotection of the amino group (14), and subsequent condensation with trimethoxy indol 2-carboxilic acid (15) in the presence of EDC, gave derivative 16 (Scheme 4).
The latter (16), after catalytic hydrogenation, afforded the correspond¬ing amino acid 17, which was allowed to react with fullerene (1) and the N-Boc 6-aminohexanal (18) (Scheme 5). Removal of the N-Boc protection led to compound 20, whose amino group could be used for the oligonucleotide coupling reaction.
The designed product 10 was synthesized following a general synthetic strategy for the preparation of oligonucleotide conjugates reported previou-sly33–35 and summarized in Scheme 6, but this new procedure involved sev¬eral modifications. Activation of the oligonucleotide terminal phosphate was achieved by the Mukaiyama reagents, triphenylphosphine-dipyridyl-2,2'-di-sulfide in the presence of DMAP,36 but the conjugation yield was not satis¬factory. Thus, 6-aminocaproic acid (ACA) was utilized as a spacer between the two moieties.
After activation of phosphorylated oligonucleotide (16-mer) at its termi¬nal phosphate and purification,33 coupling to the s-amino group of 6-amino-caproic acid in water in the presence of triethylamine was performed, giving the carboxylic acid derivative of oligonucleotide in quantitative yield. The latter was coupled with fullerene derivative 20 in a similar way by activa¬tion of carboxylic group with Mukaiyama reagents in organic media, giving the desired conjugate 10. Purification of 10 was performed by electrophore-sis in 1% agarose/0.1% triton X-100 gel using trisacetate buffer, as previ¬ously described.26,28
ANTIBACTERIAL ACTIVITY
This aspect has been investigated on the hypothesis that C60 could pro¬duce membrane disruptions by insertion into phospholipidic bilayers. The consequent membrane-disorder could lead to the discharge of metabolites and cell death.
To obtain water-soluble derivatives, we employed different aldehydes (paraformaldehyde or 3,6,9- trioxadecane aldehyde, Scheme 7) to synthesize N-mTEG (mTEG = monomethoxytriethylene glycol) substituted fulleropy-rrolidines 21–23. The latter were also alkylated with methyl iodide to afford the corresponding ammonium salts 24–26.8,10
As shown in Table I and as expected, quaternization ammonium salts 24–26 show increased water solubility (DMSO/Water 1/9) compared to neu¬tral compounds (21–23).
This allowed us to examine the microbiological profile of these com¬pounds on different microorganisms. The most interesting result was ob¬tained on Mycobacterium avium (complete inhibition by compound 23, 260 |ig mL–1) and on Mycobacterium tubercolosis. In this case, there is complete growth inhibition by 24–26 at a concentration of 50 |ig mL–1 for 24 and 5 |ig mL–1 for 25 and 26. The activity mechanism is under investigation but a plausible explanation could be that the presence of the carbon cage destabilizes the cell wall by intercalation in the hydrophobic part. In any case, the fullerene spheroid must be responsible for the activity, since the same experiments performed using non-fullerenic analogues gave negative results.
Besides, the in vivo behavior of compound 26 after m£ra-peritoneal in¬jection at different concentrations (0, 0.5, 1.0 and 1.5 g kg-1) was studied on mice.37 It was found that the LD50 is higher than 1.2 g kg-1 but also that this compound is not very well absorbed and produces deposits on different organs, without altering their appearance. Only at very high doses, the liv¬ers of the mice show an alteration typical of fibrogenosis, but their micro¬scopic examination, surprisingly, shows normal tissues. It is interesting to note that there is an important deposition in Tyson's glands. This could be explained by the fact that this tissue is rich in ceramides, the same hydro-phobic constituents of Mycobacterium cell wall.
CONCLUSIONS AND PERSPECTIVES
Although they pose serious manipulation and dosage problems, the pre¬liminary biological investigations of fullerene derivatives have given encour¬aging results. The hydrophobic spheroid and the radical sponge character of fullerene are responsible for the activity in different fields.
These preliminary findings, along with the low toxicity detected so far in fullerenes, are sufficiently promising to stimulate researchers in chemistry and in biology to unite their efforts and systematically investigate the bio¬logical properties of these fascinating molecules.
Acknowledgements. – Part of the work reviewed here was financed by MURST (cofin. ex 40%, prot. no. 9803194198_005) and Regione Friuli Venezia Giulia (Fondo 1998).
