Introduction

Top Summary Introduction Materials and Methods Results Discussion References

Here we present a study of a marine fungal product, sansalvamide A, which inhibits the topoisomerase enzyme of the pathogenic poxvirus molluscum contagiosum virus (MCV). MCV infection in healthy people causes only small papules that are easily treated. In AIDS patients, however, MCV causes severe lesions that are essentially untreatable. Cells near the surface of lesions become many times larger than normal, forming papules that become filled with a granular mass called "molluscum bodies". Untreated papules in healthy people usually disappear spontaneously within several months, but in AIDS patients dense crops can persist, disfiguring infected patients. In HIV-infected people, rates of infection may be as high as 5 to 18% and as many as 33% of AIDS patients with CD4+ counts of less that 100 cells/mm³ may be infected (Petersen and Gerstoft, 1992; Porter et al., 1992; Schwartz and Myskowski, 1992; Gottlieb and Myskowski, 1994).

Among the MCV genes identified in the recently completed primary sequence was one encoding a putative type I topoisomerase (Senkevich et al., 1996). Type I topoisomerases catalyze the formation of transient nicks in DNA that permit DNA relaxation (Gupta et al., 1995). MCV and the other poxviruses all encode topoisomerases, and in the case of vaccinia, it has been shown that replication requires topoisomerase function (Shuman et al., 1989).

MCV topoisomerase, like that of other poxviruses, is highly sequence-specific (Hwang et al., 1998; Y.H., A. Burgin and F.B., in press). Poxvirus topoisomerases bind to the sequence 5'-(C/T)CCTT-3' and cleave DNA just 3' of the last T, forming a covalent phosphotyrosine linkage. After DNA relaxation, the single strand DNA break is resealed by transesterification with the adjacent 5' hydroxyl, releasing the enzyme from the relaxed DNA product. Topoisomerase activity has been implicated as important for DNA replication, repair, transcription, and other biological processes (Wang, 1996).

We have carried out a survey of crude extracts and purified secondary metabolites from marine bacteria and fungi in an effort to identify useful inhibitors of MCV topoisomerase (Jensen and Fenical, 1994; Davidson, 1995). Here we report that sansalvamide A, a cyclic depsipeptide produced by a marine fungus Fusarium species, inhibits MCV topoisomerase in vitro. Assays of different steps in the topoisomerase catalytic cycle reveal that sansalvamide A inhibits DNA binding, but not strand religation. Sansalvamide A also inhibited the activity of the enzyme catalytic domain, beginning to specify the inhibitor site of action. This represents the first identification of a new inhibitor isolated by primary screening against MCV topoisomerase in vitro.

	Materials and Methods

Top Summary Introduction Materials and Methods Results Discussion References

Purification of MCV Topoisomerase. MCV topoisomerase (MCV-TOP) was purified using nickel-chelating sepharose as described (Hwang et al., 1998). MCV-TOP (82-323), the catalytic domain, was purified using chromatography on nickel-chelating sepharose and carboxy-methyl-sepharose (Y.H., M. Park, W. Fisher and F.D.B., submitted).

MCV Topoisomerase Activity Assays. Standard conditions for assaying relaxation, DNA binding, covalent complex formation, and religation activities were 200 mM potassium glutamate, 20 mM Tris-Cl (pH 8.0), 1 mM dithiothreitol, 0.1% Triton X-100, and 1 mM EDTA. To test inhibition of MCV topoisomerase, sansalvamide A was added to the enzyme and preincubated for 5 min and then the reaction was started with the addition of substrate. The reaction mixtures contained a 10% (v/v) final concentration of dimethyl sulfoxide.

Sansalvamide A. Purification and structure determination for sansalvamide A (Fig. 1), produced by a marine fungus Fusarium species collected off the Bahamas, is described in (G. Belofsky and W. Fenical, submitted). The compound was assayed using NMR and thin-layer chromatography methods.

View larger version (17K): [in this window] [in a new window]   Fig. 1.   Structure of sansalvamide A.

	Results

Top Summary Introduction Materials and Methods Results Discussion References

Screening Compounds from Marine Sources for Inhibitory Activity. To identify inhibitors of MCV topoisomerase, 460 candidate extracts and purified compounds from marine microorganisms were screened. The investigation focused on secondary metabolites from marine fungi and bacteria, because compounds of interest could be readily obtained as needed by culturing the appropriate microorganisms (Jensen and Fenical, 1994). Assays in vitro were carried out initially in the presence of high concentrations of compound or extract in an effort to identify potential inhibitors. Extracts or compounds that selectively inhibited MCV topoisomerase were then titrated to determine the concentration sufficient for 50 percent inhibition (IC₅₀).

Inhibition of DNA Relaxation by Sansalvamide A. Type I topoisomerases carry out relaxation of DNA by first binding to duplex DNA, cleaving one strand to generate a enzyme-DNA covalent complex, relaxing the DNA, and subsequently resealing the nick by religation of the cleaved strand (Fig. 2).

View larger version (11K): [in this window] [in a new window]   Fig. 2.   Diagram of the mechanism of type I topoisomerases such as that encoded by MCV. Rotation as drawn in the "isomerization" step is hypothetical.

View larger version (62K): [in this window] [in a new window]   Fig. 3.   Inhibition of DNA relaxation by sansalvamide A. Control reactions were performed without enzyme (lane 1) or without sansalvamide A (lane 2). Concentrations of sansalvamide A added were 10 µM (lane 3), 30 µM (lane 4), 100 µM (lane 5), 200 µM (lane 6), and 300 µM (lane 7).

Specificity of Inhibition by Sansalvamide A. As a counterscreen, inhibition of HIV-1 integrase was tested. HIV-1 integrase directs the cleavage of the termini of the HIV-1 cDNA and the subsequent covalent integration of the cleaved DNA ends into target DNA (for review, see Hansen et al., 1998). No inhibition of integrase was detectable (IC₅₀ > 850 µM, data not shown).

Effect of Sansalvamide A on DNA Binding. To investigate the mechanism of inhibition of MCV topoisomerase by sansalvamide A, we assayed inhibition of DNA binding, covalent complex formation, and religation separately.

View larger version (83K): [in this window] [in a new window]   Fig. 4.   Inhibition of DNA binding by sansalvamide A. Control reactions were performed without enzyme (lane 1) or without sansalvamide A (lane 2). Concentrations of sansalvamide A added were 10 µM (lane 3), 30 µM (lane 4), 100 µM (lane 5), 200 µM (lane 6), and 300 µM (lane 7). Enzyme-DNA complexes are indicated by the arrow.

Effect of Sansalvamide A on Covalent Complex Formation. The effect of sansalvamide A on covalent complex formation was tested using sub a as substrate. Sub a was end-labeled on the 5'-end of the scissile DNA strand, then mixed with MCV topoisomerase and incubated at 37°C for 5 min. Covalent protein-DNA complex formation was assayed by SDS-polyacrylamide gel electrophoresis and autoradiography. The covalent complex was visualized as a labeled species migrating more slowly than the substrate DNA that exhibited the molecular weight expected of the topoisomerase linked to the covalently bound DNA. Exposure of topoisomerase to sansalvamide A before addition of the sub a substrate caused a concentration-dependent decrease in the extent of covalent complex formation (Fig. 5, lanes 3-7) with an IC₅₀ of approximately 110 µM.

View larger version (70K): [in this window] [in a new window]   Fig. 5.   Inhibition of covalent complex formation by sansalvamide A. Control reactions were performed without enzyme (lane 1) or without sansalvamide A (lane 2). Concentrations of sansalvamide A added were 10 µM (lane 3), 30 µM (lane 4), 100 µM (lane 5), 200 µM (lane 6), and 300 µM (lane 7).

Effect of Sansalvamide A on Religation of the Covalent Complex. The effect of sansalvamide A on the religation activity of MCV topoisomerase was also tested. The covalent complex was trapped using a suicide substrate, which contains a short (5 base pairs) DNA segment 3' of the 5'-CCCTT-3' sequence. Upon covalent complex formation, the resulting 5-base strand will be released and lost by diffusion (Fig. 6A). To monitor religation, a labeled 15-base sequence was added that is complementary to the single stranded DNA in the covalent complex. Religation of the labeled input DNA was detected by the formation of a radioactive 29-mer product (Fig. 6B).

View larger version (24K): [in this window] [in a new window]   Fig. 6.   Effect of sansalvamide A on DNA strand religation. A, diagram of the religation reaction. The suicide substrate labeled on 5' end of the top strand (indicated by the ball) was incubated with MCV topoisomerase (step 1). Covalent complex formation resulted in releasing the short duplex extension 3' of the 5'-CCCTT-3' sequence (step 2), thereby trapping the covalent complex. This allowed the religation reaction to be carried out after addition of the 15-mer donor DNA to covalently linked topoisomerase. Religation generates a product containing a radiolabeled 29-mer DNA strand (step 3). MCV topoisomerase protein is indicated by the shaded circle. The number of bases in single stranded DNAs are indicated in parentheses. B, lack of Inhibition of DNA strand religation by sansalvamide A. Control religation reactions were performed without enzyme (lane 1), with enzyme (lane 2), or with enzyme and 15-mer single strand DNA (lane 3). Concentrations of sansalvamide A added were 10 µM (lane 4), 30 µM (lane 5), 100 µM (lane 6), 200 µM (lane 7), and 300 µM (lane 8).

Inhibition of DNA Binding by the Catalytic Domain of MCV Topoisomerase. Residues 82 to 323 of MCV topoisomerase comprise a flexible linker and the carboxyl terminal catalytic domain of the enzyme (Y.H., M. Park, W. Fisher and F.D.B., submitted). To examine the part of the enzyme affected by sansalvamide A, MCV-TOP (82-323), the purified linker and catalytic domain, was tested in reactions containing sansalvamide A. Titration of sansalvamide A into reactions containing 100 nM MCV-TOP (82-323) and labeled sub a DNA revealed inhibition of DNA binding (Fig. 7). The IC₅₀ was indistinguishable from that seen with the full length enzyme. Note that covalent complex formation by MCV-TOP (82-323) is slower than with the full enzyme, so most of the complex seen in this experiment is the noncovalently bound form. Separate analysis of inhibition of covalent complex formation revealed inhibition paralleling the inhibition of DNA binding (data not shown). Evidently sansalvamide A is also capable of inhibiting function of the catalytic domain by blocking DNA binding, specifying the site of action of the compound.

View larger version (58K): [in this window] [in a new window]   Fig. 7.   Inhibition of DNA binding of the catalytic domain (MCV-TOP 32-323) by sansalvamide A. Control reactions were performed without enzyme (lane 1) or without sansalvamide A (lane 2). Concentrations of sansalvamide A added were 10 µM (lane 3), 30 µM (lane 4), 100 µM (lane 5), 200 µM (lane 6), and 300 µM (lane 7). DNA binding complexes are indicated by the arrows.

	Discussion

Top Summary Introduction Materials and Methods Results Discussion References

Here we describe the identification of the marine depsipeptide sansalvamide A as an inhibitor of MCV topoisomerase. Sansalvamide A inhibited the initial DNA binding by MCV topoisomerase and, consequently, formation of covalent protein-DNA complexes, but did not inhibit religation by the covalent protein-DNA intermediate. This work provides a starting point for possible development of depsipeptide inhibitors for treating MCV infection.

Cyclic peptides such as sansalvamide A are a potent class of naturally occurring bioactive molecules. The immunosuppressant drug cyclosporin is a cyclic peptide secondary metabolite produced by the fungus Cylindrocarpon lucidum (Borel et al., 1976). Bacitracin and polymyxin, also cyclic peptides of microbial origin, are in use as topical antibiotic agents (Strohl, 1997). Marine invertebrates are also prolific producers of bioactive cyclic peptides, including the antiviral and cytotoxic molecule didemnin B (Rinehart et al., 1988), the thrombin inhibitor cyclotheonamide A (Fusetani et al., 1990), and patellamide and lissoclinamide cytotoxins (Ireland et al., 1982). Recently, several potent bioactive molecules have also been isolated from combinatorial libraries of cyclic peptides (Eichler et al., 1995; Giebel et al., 1995).

Cyclic depsipeptides have several advantages as inhibitors. Cyclic depsipeptides by definition contain one or more amino acids replaced by a hydroxy acid, forming at least one ester bond in the core ring structure. These compounds lack charges at the peptide amino and carboxyl termini and lacking zwitterionic character are more lipophilic and membrane-permeable. Oral bioavailability is increased by faster membrane absorption in the digestive tract (Amidon and Lee, 1994) and cyclic peptides have greater half-lives in vivo than the cognate linear peptides (Blackburn and Van Breemen, 1993; Pauletti et al., 1996). The cyclic nature of these compounds also restricts bond rotation, creating more rigid three dimensional structures. This conformational constraint can result in greater binding affinity and selectivity for protein ligands (Giebel et al., 1995). Even slight changes in the core ring structure of molecules such as cyclosporin and didemnin B can greatly reduce their biological activities, emphasizing the specificity of binding (Wenger, 1986; Sakai et al., 1996).

The mechanism by which sansalvamide A inhibits DNA binding has not been fully clarified. It seems unlikely that sansalvamide A binds indiscriminately to the substrate DNA, because it did not inhibit HIV-1 integrase or HindIII. Sansalvamide A did inhibit catalysis by an isolated domain of MCV topoisomerase containing the catalytic center, implying action at least in part against this protein domain and potentially the active site. It has not been possible to study the target of sansalvamide A in vivo due to the toxicity of the compound to cells. If more potent or less toxic derivatives of sansalvamide A can be identified, it may be possible to isolate a poxvirus insensitive to sansalvamide A and map the target of action by identifying the location of viral drug escape mutants.