Faculty Bibliography

The anti-HIV agent MAP30 (Momordica anti-HIV protein, 30 kDa) inhibits the proliferation of BC-2, an AIDS-related primary effusion lymphoma (PEL) cell line derived from an AIDS patient. BC-2 cells are latently infected with Kaposi's sarcoma-associated herpes virus (KSHV), also known as human herpes virus 8 (HHV8). We examined the effect of MAP30 on the expression of viral and cellular genes in BC-2 during latent and lytic states of the viral life cycle. By Northern analysis and RT-PCR, we found that MAP30 downregulates the expression of viral cyclin D (vCD), viral interleukin-6 (vIL-6), and viral FLIP (vFLIP), genes involved in cell cycle regulation, viral pathogenesis, and apoptosis. By pathway-specific cDNA microarray analysis, we found that BC-2 cells express high levels of egr-1, ATF-2, hsp27, hsp90, IkappaB, mdm2, skp1, and IL-2, cellular genes involved in mitogenesis, tumorigenesis, and inhibition of apoptosis in NFkappaB and p53 signaling pathways. These results define for the first time the specific cellular pathways involved in AIDS-related tumorigenesis and suggest specific novel targets for the treatment. Furthermore, we found that MAP30 downregulates the expression of egr-1, ATF-2, hsp27, hsp90, IkappaB, mdm2, and Skp1, while it upregulates the pro-apoptotic-related genes Bax, CRADD, and caspase-3. Thus, MAP30 modulates the expression of both viral and cellular genes involved in KS pathogenesis. These results provide valuable insight into the molecular mechanisms of MAP30 anti-KS action and suggest its utility as a therapeutic agent against AIDS-related tumors

For the past several years, we have been searching for novel antiviral and anti-tumor agents from nature products. From hundreds of samples investigated, we identified, purified to homogeneity, charac- terized and cloned a new class of anti-HIV agents with high potency and low toxicity from distinct and unrelated sources. The first group consists of anti-HIV proteins MAP30 (Momordica Anti-HIV Protein 30 kD) and GAP31 (Gelonium Anti-HIV Protein 31 kD) from medicinal plants and the second group consists of AVL (anti-viral lysozyme) and AVR (anti-viral RNase) from urine of pregnant women. MAP30 and GAP31 are isolated from medicinal plants Momordica charantia and Gelonium multiflorum, also known as bitter melon and Himalayan fruit, respectively. These compounds are unique in that they not only inhibit de novo infection by HIV-l but also block the replication of the virus in already infected cells. We found that they affect HIV-1-infected cells with EC50s (effective concentration at 50% inhibition) in the subnanomolar range (0.2-0.3 nM). They show no apparent cytotoxic or cytostatic effects on normal human cells even at 1,000-fold higher dose levels. MAP30 and GAP3l possess multiple therapeutic targets at different stages of the HIV-l life cycle. We have characterized at least three biological activities that may be relevant to their therapeutic use. The first is an RNA N-glycosidase activity that cleaves the link between a ribose and adenine A4324 of 28S ribosomal rRNA. This inactivates the 60S ribosomal subunit and inhibits polypeptide chain elongation. The second is a DNA topological inactivation activity that renders HIV-LTR topologically inactive as substrates for DNA gyrase. This topoinactivation is similar to the effect of cellular topoisomerases in the presence of topoisomerase inhibitors. The third is inhibition of each of the three reactions catalyzed by HIV-l integrase: 3' processing of the viral DNA. strand transfer, and cleavage at the viral/target junction. It is thus important to define the extent to which each of these mechanisms contributes to desired antiviral and antitumor actions or to undesired cytotoxicity. We carried out structural and activity mapping of MAP30 and GAP31 by X-ray diffraction of crystals and by limited proteolysis. We identified and isolated proteolytic fragments of MAP30 and GAP31 that are fully active against HIV-1 but not in ribosome inactivation. These peptides are as active as their parent molecules in HIV-l inhibition with EC50 in the range of 0.2-0.4 nM. They inhibit HIV-integrase activity and HIV-LTR topological interconversion, but they do not inhibit ribosome activity. These results demonstrate that the antiviral activity of MAP30 and GAP31 is independent from their ribosome-inactivating protein activity. This is of great significance and may provide useful insights in the design and development of antiviral and anti-tumor agents with specific therapeutic targets toward viral-infected and/or tumor cells, without cytotoxicity towards cellular targets. The second group of antiviral agents consists of AVL and AVR. To our knowledge, this is the first report that lysozymes and ribonucleases possess anti-HIV activity, and the first identification of these proteins as components present in crude b-core preparations that contribute to its anti-HIV effects. They represent a totally new class of therapeutic agents because they are naturally occurring human proteins that modulate viral infection. Details of these studies are presented in the following abstract

The transmission of HIV-1 from mother to fetus is rare during the first trimester of pregnancy when the secretion of hCG is high in the placenta. Consequently, hCG preparations were considered to have a role in the inhibition of HIV-1 transmission. Indeed, many hCG and its beta-subunit (hCGbeta) in particular were found to contain the anti-HIV activity both in vivo and in vitro studies. However, there has been controversy about whether some biological activities of hCGbeta preparations are due to the beta-subunit itself, or to other proteins present in the preparations. To determine whether proteins other than hCGbeta itself might contribute to the anti-HIV activity of hCGbeta preparations, we fractionated commercial preparations derived from the urine of pregnant women. We found that a significant portion of the anti-HIV-1 activity is associated with the beta-core fraction. The beta-core is a dimer of two peptide fragments of hCGbeta linked by disulfide bridges. When a beta-core fraction was purified by reverse-phase HPLC, the pure beta-core molecules identified by N-terminal amino acid sequencing and SDS-PAGE were completely devoid of anti-HIV activity assayed by p24 production in chronically HIV-1 infected ACH2 lymphocytes and U1 monocytes. The bulk of anti-HIV activity was eluted behind the beta-core fractions. Further purification of the fractions containing the anti-HIV activity by SDS-PAGE followed by Sephadex G-25 superfine, 18- and 18.5-kD fractions was identified by N-terminal amino acid sequencing as ribunuclease (RNase) U and 14 kD as urinary lysozyme C. Both purified enzymes exhibited not only respective authentic enzymatic activities but also anti-HIV activity. As expected, RNase U effectively degraded total RNA isolated from HIV-infected ACH2 lymphocytes. Similarly, ribonuclease A was found as 23 kD on SDS-PAGE in an extensively purified beta-core preparation. We therefore designate the ribunuclease and lysozyme as anti-viral ribonucleases (AVR) and anti-viral lysozyme (AVL), respectively. Furthermore, commercially available lysozyme from chicken egg white, human milk, and human neutrophils and RNase A from bovine pancreas were demonstrated in these studies to possess activity against HIV-1. Since lysozyme is elevated in the urine of pregnant women and known to reduce the absorption of ectromelia virus, it may play important protective roles during pregnancy. This may explain why HIV infection from mother to fetus is rare. Collectively, these findings may offer new strategies for the treatment of HIV-1 infection

GAP31 (Gelonium protein of 31 kDa) and MAP30 (Momordica protein of 30 kDa) are agents isolated from the medicinal plants Gelonium multiflorum and Momordica charantia, respectively. The current study was conducted to investigate the efficacy of GAP31 and MAP30 on estrogen-independent and highly metastatic human breast tumor MDA-MB-231 both in vitro and in vivo. The effect of these agents on the expression of breast tumor antigen HER2 (also known as neu or as c-erbB 2) was also examined. Treatment of MDA-MB-231 breast cancer cells with GAP31 and MAP30 resulted in inhibition of cancer cell proliferation as well as inhibition of the expression of HER2 gene in vitro. When MDA-MB-231 human breast cancer cells were transferred into SCID mice, the mice developed extensive metastases and all mice succumbed to tumor by day 46. Treatment of the human breast cancer bearing SCID mice with GAP31 or MAP30 at 10 micrograms/injection EOD for 10 injections resulted in significant increases in survival, with 20-25% of the mice remaining tumor free for 96 days. Thus, anti-tumor agents GAP31 and MAP30 are effective against human breast cancer MDA-MB-231 in vitro and in vivo. These agents may therefore be a potential therapeutic use against breast carcinomas

MAP30 is a 30 kDa single-stranded, type-I ribosome inactivating protein (RIP) possessing anti-tumor and anti-HIV activities. It binds both ribosomal RNA and the HIV-1 long-terminal repeat DNA. To understand the structural basis for MAP30 activities, we undertook the study of MAP30 by solution NMR spectroscopy. We report nearly complete 1H, 13C, and 15N chemical shift assignments of its 263 amino acids. Based upon an analysis of secondary 13C chemical shifts, 3J(HNHA) coupling constants, hydrogen exchange data, and nuclear Overhauser effect patterns, we find that the secondary structure and beta-sheet topology of MAP30 are very similar to those of the ricin A chain, a subunit of the well-known type-II RIP, even though two proteins display distinct activities. We therefore suggest that MAP30 and ricin A chain share a similar three-dimensional fold, and that the reported functional differences between two proteins arise primarily from differences in local three-dimensional structure and other structural properties such as surface electrostatic potentials

We present the solution structure of MAP30, a plant protein with anti-HIV and anti-tumor activities. Structural analysis and subsequent biochemical assays lead to several novel discoveries. First, MAP30 acts like a DNA glycosylase/apurinic (ap) lyase, an additional activity distinct from its known RNA N-glycosidase activity toward the 28S rRNA. Glycosylase/ap lyase activity explains MAP30's apparent inhibition of the HIV-1 integrase, MAP30's ability to irreversibly relax supercoiled DNA, and may be an alternative cytotoxic pathway that contributes to MAP30's anti-HIV/anti-tumor activities. Second, two distinct, but contiguous, subsites are responsible for MAP30's glycosylase/ap lyase activity. Third, Mn2+ and Zn2+ interact with negatively charged surfaces next to the catalytic sites, facilitating DNA substrate binding instead of directly participating in catalysis

OBJECTIVE: To investigate the effects of two virucidal compounds, MAP30 (Momordica anti-human immunodeficiency virus [HIV] protein; molecular weight, 30 kd) and GAP31 (Gelonium anti-HIV protein; molecular weight, 31 kd), obtained from Momordica charantia and Gelonium multiflorum, respectively, on the motility and vitality of human spermatocytes. DESIGN: Prospective, controlled study. SETTING: New York University School of Medicine. PATIENT(S): Ten healthy men undergoing evaluation for infertility provided 10 semen specimens. INTERVENTION(S): Human sperm were treated with the anti-HIV agents, MAP30 and GAP3 1. Nonoxynol-9, a commonly used spermicide, and phosphate-buffered saline were used as the positive and negative controls, respectively. MAIN OUTCOME MEASURE(S): The motility and vitality of human spermatocytes treated with MAP30 and GAP31 at doses that inhibit HIV-1 and herpes simplex virus. RESULT(S): MAP30 and GAP31 did not inhibit the motility or vitality of human sperm cells over a dose range of 100-0.1 microg/mL, whereas nonoxynol-9 demonstrated spermicidal action on all 10 samples over the same dose range. CONCLUSION(S): The antiviral agents, MAP30 and GAP31, were not toxic to human sperm cells at the doses at which they inhibit HIV-1 and herpes simplex virus. They had no effect on the motility of spermatozoa, even at a dose of 1,000 times the maximum effective concentration. These results indicate that MAP30 and GAP31 may be useful as nonspermicidal protection against sexually transmitted diseases

We analyzed the structural and functional organization of anti-HIV and anti-tumor proteins MAP30 and GAP31 by limited proteolysis with endopeptidases Lys-C and Glu-C (V8). MAP30 and GAP31 are resistant to proteolytic digestion under conditions of as much as 5% (w/w) proteases. In the presence of 10% (w/w) protease, the central regions of the proteins are still resistant to proteolysis, whereas the N- and C-termini are accessible. Peptide fragments were purified by FPLC on Superdex 75 columns, characterized by gel electrophoresis, identified by amino acid sequencing, and analyzed for anti-HIV, anti-tumor, and other biochemical activities. We report here that limited proteolysis yields biologically active fragments of both MAP30 and GAP31. These fragments are active against HIV-1 and tumor cells with EC(50)s in the sub-nanomolar ranges, 0.2-0.4 nM. At the dose levels used in the assays, little cytotoxicity to normal cells was observed. In addition, these fragments remain fully active in HIV-integrase inhibition and HIV-LTR topological inactivation, but not ribosome inactivation. These results demonstrate that the antiviral and anti-tumor activities of MAP30 and GAP31 are independent of ribosome inactivation activity. In addition, we demonstrate that portions of the N- and C-termini are not essential for antiviral and anti-tumor activities, but do appear to be required for ribosome inactivation. These results may provide novel strategies for rational design and targeted development of mimetic antiviral and anti-tumor therapeutics.

A method is described which permits detection of 3hJNC' scalar couplings across hydrogen bonds in larger, perdeuterated proteins. The experiment is demonstrated for the uniformly 2H/13C/15N-enriched 30 kDa ribosome inactivating protein MAP30. The 3hJNC' interactions are smaller than 1 Hz, but their detection in an HNCO experiment is made possible through the use of constructive interference between the 15N chemical shift anisotropy and 1H-15N dipole-dipole relaxation mechanisms in a manner similar to that of recently proposed TROSY schemes. Sensitivity of the HNCO experiment depends strongly on the 15N transverse relaxation rate of the downfield 15N multiplet component and on the amide proton T1. In perdeuterated MAP30 at 40 degrees C, the average TROSY T2 was 169 ms at 750 MHz 1H frequency, and a wide range of longitudinal relaxation rates was observed for the amide protons