key: cord-0722384-9q7uckg9 authors: Balzarini, Jan title: The α(1,2)-mannosidase I inhibitor 1-deoxymannojirimycin potentiates the antiviral activity of carbohydrate-binding agents against wild-type and mutant HIV-1 strains containing glycan deletions in gp120 date: 2007-05-15 journal: FEBS Letters DOI: 10.1016/j.febslet.2007.04.039 sha: e0e0661fc5922366f9b7d22e90d33864f86076fb doc_id: 722384 cord_uid: 9q7uckg9 Abstract Exposure of carbohydrate-binding agents (CBAs) (i.e. the mannose-specific plant lectins Hippeastrum hybrid agglutinin and Galanthus nivalis agglutinin) to HIV-1 progressively select for mutant HIV-1 strains that contain N-glycan deletions in their envelope gp120. This results in resistance of the mutant virus strains to the CBAs. Exposure of such mutant virus strains to the α(1,2)-mannosidase I inhibitor 1-deoxymannojirimycin (DMJ) results in an enhanced suppression of mutant virus-induced cytopathicity in CEM cell cultures. Moreover, when combined with CBAs at concentrations that showed poor if any suppression of mutant virus replication as single drugs, a synergistic antiviral activity of DMJ was observed. These observations argue for a combined exposure of CBAs and glycosylation inhibitors such as DMJ to HIV to afford a more pronounced suppression of virus replication, prior to, or during, CBA resistance development. The majority of enveloped viruses contains multiple glycans on their envelope proteins. In some cases (i.e. human immunodeficiency virus, HIV) [1] , hepatitis C virus (HCV) [2] , coronaviruses (CoV) [3] , influenza virus (INF) [4] ), the envelope is extensively glycosylated. The gp120 envelope of HIV is among the most heavily glycosylated proteins known [5] . Protein glycosylation may serve multiple functions, including proper folding of the nascent peptide, avoiding peptide precipitation due to the presence of lipophylic amino acid domains in the protein, protection against breakdown by proteases, increasing molecular diversity, and last but not least, in some cases, escape of immune surveillance [6] . After the glycan building block (GlcNAc) 2 Man 9 Glc 3 has been added to asparagines of the native peptide that are part of a N-glycosylation motif (NXS/T), the N-glycans are processed by a-glucosidases to remove the terminal three glucoses in the endoplasmatic reticulum (ER). Then, ER and Golgi class I a1,2-mannosidases specifically hydrolyze a1,2-mannose residues, and catalyse the trimming of the high-mannose chains involving four a1,2linked mannose residues, and this process generates Man 5 Glc-NAc 2 . Subsequent action of GlcNAc transferase I initiates complex chain formation and yields the substrate for Golgi a-mannosidase II which trims the terminal a1,3-and a1,6mannose residues [7] . Further processing events in the Golgi apparatus eventually lead to glycans that consist of a wide variety of carbohydrates and combinations thereof [7] [8] [9] [10] . Since mammalian viruses use the host cell glycosylation machinery for glycan synthesis and modification of the glycans that need to be incorporated in their envelope glycoproteins, it has been suggested that it is possible to target the viral envelope glycoproteins by inhibiting certain host-cell glucosidases at low levels that do not affect host-cell viability [5] . The altered glycan structures on the viral envelope proteins may then result in decreased viral infectivity (fitness), virus assembly and/or virus particle release [5] . HIV infectivity has indeed shown to be suppressed in cell culture when the virus was propagated in the presence of the a-glucosidase inhibitor NB-DNJ [11] . The latter drug has been evaluated in phase II clinical trials as an anti-HIV therapeutic [12] . For hepatitis B virus (HBV), it was demonstrated that NN-DNJ (and also to a minor extent NB-DNJ) disrupted the proper folding and efficient release of the viral envelope molecules. It was shown that NB-DNJ could reduce virus levels in a dose-dependent manner [13] . Since the E1 and E2 transmembrane glycoproteins of HCV are important for host cell entry [14] , and since proper folding is calnexin-dependent [15] , glucosidase inhibitors may also be expected to affect HCV entry and infectivity. Recently, we have shown that carbohydrate-binding agents (CBA) are able to force HIV-1 to delete part of the glycans on its gp120 envelope in an attempt to escape drug pressure [16] [17] [18] [19] . Such mutant virus strains display different degrees of phenotypic (in)sensitivity to the CBA's antiviral activity depending the number and the nature of the glycans that were deleted in gp120. In this study, we wanted to investigate whether the concomitant combination of CBAs and the glycosylation inhibitor 1-deoxymannojirimycin (DMJ) against Abbreviations: CBA, carbohydrate-binding agent; HHA, Hippeastrum hybrid agglutinin; GNA, Galanthus nivalis agglutinin; DMJ, 1-deoxymannojirimycin; HIV, human immunodeficiency virus; HBV, hepatitis B virus; HCV, hepatitis C virus; ER, endoplasmatic reticulum wild-type and mutant (glycan-deleted) gp120-containing HIV-1 strains could afford a superior antiviral activity than when added as single drugs. DMJ was used because it selectively inhibits a1,2-mannosidase I resulting in the accumulation of high-mannose glycans on the viral envelope glycoprotein. We found a significantly increased sensitivity of the mutant virus strains to the inhibition by DMJ, and a marked potentiation of the antiviral efficacy of CBAs when co-administered with DMJ, both for wild-type and mutant virus strains. The mannose-specific plant lectins from Galanthus nivalis (GNA) and Hippeastrum hybrid (HHA) were derived and purified from these plants, as described before [20, 21] . DMJ was obtained from Sigma-Aldrich (St. Louis, MO) and from Calbiochem (VWR International, Haasrode, Belgium). Human T-lymphocytic CEM cells were obtained from the American Type Culture Collection (Manassas, VA) and cultivated in RPMI-1640 medium supplemented with 10% fetal bovine serum (FBS) (BioWittaker Europe, Verviers, Belgium), 2 mM L L-glutamine and 0.075 M NaHCO 3 . HIV-1(III B ) was provided by Dr. R.C. Gallo and Dr. M. Popovic (at that time at the National Cancer Institute (NCI), National Institutes of Health (NIH), Bethesda, MD). The mutant virus strains were obtained and characterized as described before [22] . The methodology of the anti-HIV assays has been described previously [16, 17] . Briefly, CEM cells (4.5 · 10 5 cells per ml) were suspended in fresh culture medium and infected with HIV-1 at 100 CCID 50 per ml of cell suspension. Then, 100 ll of the infected cell suspension were transferred to microplate wells, mixed with 100 ll of the appropriate dilutions of the test compounds, and further incubated at 37°C. After 4-5 days, giant cell formation was recorded microscopically in the CEM cell cultures. The 50% effective concentration (EC 50 ) corresponds to the compound concentrations required to prevent syncytium formation by 50% in the virus-infected CEM cell cultures. In the drug combination experiments, DMJ was added to the cell cultures prior to the addition of the CBA and virus infection of the drug-exposed cells. The proper control experiments in which only one of the drugs or none of the drugs were present, were carried out under similar experimental conditions. Data of representative experiment were shown in the figures. CBAs HHA and GNA against wild-type and mutant HIV-1 strains The antiviral activity of the a(1,2)-mannosidase I inhibitor DMJ and the mannose-specific plant lectins HHA and GNA was investigated against wild-type HIV-1(III B ) and three mutant HIV-1(III B ) strains that contain a variety of 7-8 glycan deletions in their envelope gp120 (Table 1) . DMJ did not suppress HIV-1(III B )-induced cytopathicity in CEM cell cultures at a concentration as high as 500 lM. However, when DMJ was evaluated for its antiviral activity against the mutant virus strains, it had gained, as such, measurable antiviral efficacy. DMJ was inhibitory at an EC 50 that ranged between 90 and 155 lM against the mutant virus strains. Thus, the cytopathic activity of the mutant virus strains was invariably suppressed by DMJ (Table 2 ). In contrast, the CBAs HHA and GNA that showed EC 50 values as low as 0.28 and 0.16 lg/ml against wildtype HIV-1(III B ), respectively, markedly lost their pronounced suppressive activity against the three mutant virus strains (EC 50 : 58-500 lg/ml) ( Table 2) . Thus, the deleted glycans in HIV-1 gp120 clearly compromised the antiviral activity of the CBAs. wild-type HIV-1 in CEM cell cultures The effect of 250 and 100 lM DMJ on the inhibitory activity of the CBAs HHA and GNA against wild-type HIV-1 replication in CEM cell cultures was investigated (Fig. 1) . As already mentioned above, DMJ was not inhibitory against HIV-1-induced cytopathicity at the concentrations used (250 and 100 lM). In contrast, HHA (Fig. 1A) and GNA (Fig. 1B) as single drugs completely prevented HIV-1-induced CPE in the CEM cell cultures at concentrations as low as 0.8 lg/ml. At 0.16 lg/ml, HHA and GNA were $25% and 50% inhibitory, respectively. At 0.032 lg/ml, no residual inhibitory effect of the CBAs was observed. Interestingly, co-administration of DMJ to HHA and GNA markedly potentiated the antiviral M + À + a Glycan deletions at the N-glycosylation sites in gp120 (indicated as +) as determined in ref. 22 . The ''À'' notation refers to the presence of the (glycan containing) wild-type sequence. The ''±'' notation refers to the presence of a mixture of the wild-type and mutated sequence in the virus isolate. b C: complex-type glycan, M: high-mannose type glycan [19] . efficacy of these CBAs. The inhibitory activity of 0.16 lg/ml HHA against HIV-1 increased from 25% to 90% in the presence of DMJ, and from 50% to 100% upon co-administration of DMJ with 0.16 lg/ml GNA. At lower HHA and GNA concentrations, no pronounced antiviral activity was noticed for the CBAs, neither in the absence, nor in the presence of DMJ (Fig. 1 ). mutant HIV-1 strains in CEM cell cultures Three different HIV-1 strains that were shown to contain several N-glycan deletions in their gp120 envelope (Table 1) were exposed to GNA and HHA in the presence of a variety of DMJ concentrations. When 50, 20 and 8 lM DMJ was combined with GNA ( Fig. 2A-C) , DMJ acted synergistically in combination with these CBAs. For example, 4 and 20 lg/ml GNA that showed poor, if any, antiviral efficacy against HIV-1/GNA-500(CS) (Fig. 2, panel A) and HIV-1/HHA-500(CS) (Fig. 2, panel B) became 100% protective against the virus-in-duced cytopathic effect in the presence of 50 lM DMJ, and 40-100% protective in the presence of 20 lM DMJ. Such a synergistic effect was also seen for DMJ against HIV-1/HHA-500(SN) when GNA was administered at the higher concentration range (20-500 lM) (Fig. 2, panel C) . A similar synergistic activity of DMJ was noted against the mutant virus strains when combined with HHA (Fig. 3, panels A, B and C) . Surprisingly, when the lower GNA and HHA concentrations (0.032-0.8 lM) were combined with DMJ, rather an antagonistic activity was observed. This phenomenon was consistently seen for both HHA and GNA, in the presence of the different DMJ concentrations (Figs. 2 and 3) . Neither the CBAs (500 lM) nor DMJ (250 lM) proved inhibitory against CEM cell proliferation whether administered to the cell cultures as single drugs or combined (data not shown). The drug combinations also had no inhibitory effect on cell metabolism since radiolabeled thymidine, uridine and leucine incorporation into CEM cell DNA, RNA or proteins was not measurably affected. The glycosylation inhibitor DMJ targets the ER and Golgi a(1,2)-mannosidase I that trims the a1,2-mannose(s) from the Man 9 (GlcNAc) 2 glycan after the ER a-glucosidases I and II have removed the three terminal glucose units from the N-glycan Glc 3 Man 9 (GlcNAc) 2 block [7, 8] . As a result, the amount of high-mannose type glycan structures on the glycoprotein markedly increases in the presence of DMJ since further trimming/ processing of the high-mannose glycans to hybrid-or complex-type glycans has been largely prevented by the DMJ-mediated blockade of the a1,2-mannosidases I. Since the CBAs GNA and HHA are known to specifically bind to a(1,3)-and/or a(1,6)-mannose oligomer structures [23] , it could be reasoned that a higher amount of high-mannose type glycans on gp120 may allow these CBAs to concomittantly bind to a higher amount of glycans on the HIV-1 envelope. Consequently, they may exert a more pronounced antiviral activity in DMJtreated virus-infected cells. We observed indeed a potentiation of the anti-HIV-1 activity of HHA and GNA in the presence of DMJ concentrations that exerted themselves no antiviral activity when used as a single drug. Thus, concomittant administration of glycosylation inhibitors (such as DMJ) and CBAs in HIV-1-infected cell cultures may further potentiate the antiviral activity of the mannose-specific CBAs. Interestingly, whereas wild-type virus infection and replication efficiently proceed in the presence of high (i.e. 250 and 100 lM) DMJ concentrations (EC 50 > 500 lM), the mutant HIV-1 strains that contain multiple deletions of N-glycans in gp120 gained sensitivity to the inhibitory activity of the a1,2mannosidase I inhibitor DMJ in the CEM cell cultures. Whereas DMJ was not effective at all at 500 lM against parent wild-type virus it could indeed inhibit mutant virus infection at an EC 50 that ranged between 90 and 150 lM, that is at an at least more than 5-10-fold lower DMJ concentration. These mutant virus strains showed 7 or 8 glycan deletions at putative N-glycosylation sites in gp120 [22] , and the deletions preferentially occurred at high-mannose type glycan sites (Table 1) . Such glycan deletions in the mutant HIV-1 strains resulted in a marked phenotypic resistance to the HHA and GNA CBAs. It could be assumed that DMJ converts at least part of the remaining glycans of the mutant gp120 into high-mannose-type glycan structures, making them more vulnerable to interaction with the CBAs. Consequently, an increased antiviral activity would then be expected upon co-administration with DMJ, a phenomenon that we indeed observed to occur in the HIV-infected cell cultures. Interestingly, a similar phenomenon has been observed to occur when Pradimicin A, a high mannose-type glycan-binding antibiotic, was exposed to mammalian U937 cells that had been pretreated with DMJ [24] . Under these experimental conditions, the cells express high levels of high mannose-type oligosaccharides and become sensitive to PRM-A (resulting in apoptosis induction). No such apoptosis induction was observed in PRM-A-exposed cell cultures that were not pretreated with DMJ [24, 25] . Thus, the combined use of CBAs and the a1,2-mannosidase-inhibitor DMJ enabled partial restoration of the phenotypic sensitivity of the mutant HIV-1 strains against the CBAs. Our results argue for combined administration of CBAs and DMJ to wild-type virus, because phenotypic resistance development may be expected to slow down when DMJ is present during CBA treatment of HIV-1. The slight but consistently observed antagonistic activity that has been observed for DMJ when combined with the lowest CBA concentrations is rather puzzling and the molecular basis of this phenomenon is yet unclear. There is, in general, a concern for the therapeutic application of inhibitors that target cellular enzymes such as the a1,2-mannosidase I inhibitor DMJ. Indeed, inhibition of cellular glycosylation enzymes in virus-infected cells may not only compromise proper viral glycopeptide formation, but may also have deleterious effects on glycoproteins of non-infected cells. However, therapy with drugs, in casu glycosylation inhibitors, should not necessary aim to ablate enzyme activity but should rather be used to modulate enzyme activities involved in glycosylation [26] . Since the envelope gp120 glycoprotein of HIV is among the highest glycosylated glycoproteins currently known and has a high (functional) requirement of high-mannose type glycans, it may be assumed that a moderate attenuation of a1,2-mannosidase I activity may have a more pronounced deleterious impact on the synthesis of the viral envelope glycoprotein than on the cellular glycoproteins, allowing for a certain degree of selectivity of such inhibitors. Proper in vivo experiments should reveal the therapeutic efficacy and feasibility of such drugs. In conclusion, the a(1,2)-mannosidase I inhibitor DMJ was found to potentiate the inhibitory activity of CBAs against wild-type HIV-1. Administration of DMJ to cell cultures infected with mutant HIV-1 strains that contain N-glycan deletions in the gp120 envelope render the mutant virus susceptible to the inhibitory activity of DMJ. Moreover, DMJ can partially reverse the phenotypic resistance of CBAs to the mutant virus strains. These three phenomena may argue for further investigation of glycosidase inhibitors such as, but not limited to, DMJ to be used in combination with CBAs with the aim to further potentiate the antiviral activity of the CBAs and to delay resistance development that may develop under CBA drug pressure. Carbohydrate structures of the human immunodeficiency virus (HIV) recombinant envelope glycoprotein gp120 produced in Chinese-hamster ovary cells Functional hepatitis C virus envelope glycoproteins Mass spectrometric characterization of proteins from the SARS virus: a preliminary report Effect of the addition of oligosaccharides on the biological activities and antigenicity of influenza A/H3N2 virus hemagglutinin Glycobiology against viruses Antibody neutralization and escape by HIV-1 Structural basis for catalysis and inhibition of N-glycan processing class I a1,2-mannosidases Substrate specificities of recombinant murine Golgi a1,2-mannosidases IA and IB and comparison with endoplasmic reticulum and Golgi processing a1,2-mannosidases Glycosylation and the immune system The a-glucosidase inhibitor N-butyldeoxynojirimycin inhibits human immunodeficiency virus entry at the level of post-CD4 binding The safety and efficacy of combination Nbutyldeoxynojirimycin (SC-48334) and zidovudine in patients with HIV-1 infection and 200-500 CD4 cells/mm 3 Treatment of chronic hepadnavirus infection in a woodchuck animal model with an inhibitor of protein folding and trafficking Infectious hepatitis C virus pseudo-particles containing functional E1-E2 envelope protein complexes Hepatitis C virus glycoprotein folding: disulfide bond formation and association with calnexin Profile of resistance of human immunodeficiency virus to mannose-specific plant lectins Carbohydrate-binding agents cause deletions of highly conserved glycosylation sites in HIV gp120: a new therapeutic concept to hit the Achilles heel of HIV Mutational pathways, resistance profile, and side effects of cyanovirin relative to human immunodeficiency virus type 1 strains with Nglycan deletions in their gp120 envelopes Pradimicin: a carbohydrate-binding non-peptidic lead compound for treatment of virus infections with a highly glycosylated envelope such as the human immunodeficiency virus Leaves of the orchid twayblade (Listera ovata) contain a mannose-specific lectin Related mannose-specific lectins from different species of the family Amaryllidaceae Marked depletion of glycosylation sites in HIV-1 gp120 under selection pressure by the mannose-specific plant lectins of Hippeastrum hybrid and Galanthus nivalis Handbook of Plant Lectins: Properties and Biomedical Applications Mannose-binding quinone glycoside, MBQ: potential utility and action mechanism Studies on the mode of antifungal action of pradimicin antibiotics. III. Spectrophotometric sequence analysis of the ternary complex formation of BMY-28864 with D D-mannopyranoside and calcium Targeting glycosylation as a therapeutic approach We are grateful to Ann Absillis and Yoeri Schrooten for excellent technical assistance and Christiane Callebaut for dedicated editorial assistance. The