Journal of General Virology (2003), 84, 353–360
DOI 10.1099/vir.0.18734-0
Glycosylation inhibitors and neuraminidase
enhance human immunodeficiency virus type 1
binding and neutralization by mannose-binding
lectin
Melanie L. Hart, Mohammed Saifuddin and Gregory T. Spear
Correspondence
Gregory Spear
Department of Immunology/Microbiology, Rush-Presbyterian-St Luke’s Medical Center,
1563 W. Congress Parkway, Chicago, IL 60612, USA
[email protected]
Received 29 July 2002
Accepted 10 October 2002
Mannose-binding lectin (MBL), a C-type lectin component of the human innate immune system,
binds to the gp120 envelope glycoprotein of human immunodeficiency virus type 1 (HIV-1). The
objective of this study was to assess the effects of inhibitors of endoplasmic reticulum glucosidases
and Golgi mannosidase as well as neuraminidase (NA) on the interaction between HIV and MBL.
Production of HIV in the presence of the mannosidase I inhibitor deoxymannojirimycin (dMM)
significantly enhanced binding of HIV to MBL and increased MBL neutralization of an M-tropic HIV
primary isolate. In contrast, culturing HIV in the presence of a-glucosidase I and II inhibitors
castanospermine and deoxynojirimycin only slightly affected virus binding and neutralization by
MBL. Removal of sialic acid from HIV by NA also significantly enhanced virus binding and
neutralization by MBL. Treatment of virus grown in the presence of dMM with endoglycosidase F1
substantially reduced binding to MBL, indicating that dMM increased MBL binding by increasing
high-mannose carbohydrates on the virus. In contrast, endoglycosidase F1 did not decrease the
MBL interaction with NA-treated virus, suggesting that NA exposed novel MBL binding sites.
Treatment with dMM increased the immunocapture of HIV by monoclonal antibodies 2F5 and
2G12, indicating that altering the glycosylation of viral glycoproteins increases the accessibility or
reactivity of some epitopes. This study shows that specific alterations of the N-linked carbohydrates
on HIV gp120/gp41 can enhance MBL-mediated neutralization of virus by strengthening the
interaction of HIV-1 with MBL.
INTRODUCTION
The gp120 glycoprotein of human immunodeficiency virus
type 1 (HIV-1) initiates infection by interacting with CD4
on cells. While many studies have explored the feasibility of
inducing antibodies to gp120 to inhibit HIV-1 replication,
the efficacy of this approach in vivo is limited by the rapid
mutation of the variable portions of the peptide backbone of
gp120 and the dense glycosylation of gp120 that blocks
antibody access to conserved regions of gp120 (Wyatt et al.,
1998). In fact, approximately half of the molecular mass of
gp120 consists of N-linked carbohydrates. For example, the
gp120 of HIV-IIIB was found to contain 13 complex-type
glycans and 11 high-mannose or hybrid-type glycans
(Leonard et al., 1990).
Mannose-binding lectin (MBL) is a C-type lectin, present in
human serum, that acts as an effector molecule of the innate
immune system (Jack et al., 2001; Petersen et al., 2001). MBL
binds to carbohydrates on microorganisms that express
repetitive mannose and/or N-acetylglucosamine residues,
such as Candida albicans, Salmonella typhimurium and
0001-8734 G 2003 SGM
Neisseria gonorrhoeae, resulting in opsonization and activation of the lectin complement pathway (van Emmerik et al.,
1994; Neth et al., 2000). Several studies have shown that
MBL also interacts with gp120 of HIV-1. For example,
complement was activated following MBL binding to purified gp120 (Haurum et al., 1993), recombinant MBL bound
to gp120 and gp160 purified from HIV-IIIB (Ohtani et al.,
1999) and preincubation of HIV with MBL inhibited infection of a T cell line (Ezekowitz et al., 1989). Furthermore,
HIV particles lacking gp120/gp41 do not bind MBL, indicating that carbohydrates on gp120 mediate the interaction
between whole virus and MBL (Saifuddin et al., 2000). A
recent study showed that MBL binds to HIV via highmannose carbohydrates on gp120 (M. L. Hart, unpublished).
Several enzyme inhibitors are available that prevent the
formation of complex and hybrid N-linked saccharides during glycoprotein processing in the endoplasmic reticulum
(ER) and Golgi (reviewed in Sears & Wong, 1998). For
example, castanospermine (Csp) and 1-deoxynojirimycin
(dNM) inhibit a-glycosidase I and a-glycosidase II, respectively, in the ER while 1-deoxymannojirimycin (dMM)
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Fri, 16 Jun 2017 22:14:29
Printed in Great Britain
353
M. L. Hart, M. Saifuddin and G. T. Spear
inhibits mannosidase I in the Golgi. Treatment of cells with
these inhibitors results in N-linked carbohydrates that lack
sialic acid but have a relatively high content of mannose
residues. Many of the inhibitors have been shown to decrease the infectious titre of HIV in vitro, possibly by inhibiting the correct folding of gp160 (Gruters et al., 1987;
Jacob, 1995; Karpas et al., 1988; Mehta et al., 1998;
Montefiori et al., 1988; Walker et al., 1987). Several of
these drugs have been evaluated in vivo as treatments for
conditions such as diabetes, cancer metastases and viral
infections including HIV infection (Jacob, 1995).
human serum (NHS). Serum samples from five HIV-seropositive
individuals were heat-inactivated and pooled in equal volumes
(HIVPS). Monoclonal antibodies 2F5, 2G12 and IgG1b12 were obtained
through ARRRP. 2F5 and 2G12 were contributed by Hermann
Katinger (Purtscher et al., 1994; Trkola et al., 1996), and IgG1b12
was contributed by Dennis Burton and Carlos Barbas (Barbas et al.,
1992). Polyclonal sheep anti-HIVgp120 antibody to the conserved
C-terminal sequence (aa 497–511) of HIV-1 gp120 (Moore et al., 1989)
was purchased from International Enzymes.
While the high density of oligosaccharides on gp120 is
thought to play a protective role by reducing the immunogenicity of gp120, it is possible that the glycans on gp120
could be used as a target in some antiviral strategies. The
objective of the current study was to explore the effect of
glycosylation inhibitors on the interaction between HIV and
MBL, since alteration of virus carbohydrates in vivo could
increase the amount of MBL-binding carbohydrates on the
virus surface, resulting in more efficient clearance and/or
neutralization of virus. Since one effect of these drugs is to
prevent addition of sialic acid residues to gp120 as it is
processed, we also tested the effect of treatment of virus with
neuraminidase (NA) on binding and neutralization of HIV
by MBL. Our results showed that both NA treatment and
certain glycosylation inhibitors strengthen the interaction of
HIV-1 with MBL and thereby enhance MBL-mediated
neutralization of HIV-1.
Mannose-binding lectin. Recombinant vaccinia virus expressing
Reagents. Csp was purchased from ICN Biomedicals and dNM,
dMM and Clostridium perfringens a2R3,6,8-NA were obtained from
Sigma. Endoglycosidase F1 was from Prozyme.
the cDNA sequence for human MBL was used to produce recombinant MBL (rMBL) (Ma et al., 1997). Briefly, HLF cells were infected
at an m.o.i. of 5 and supernatants were collected 48 h after infection.
MBL was purified by passage over a mannan–Sepharose 4B column,
as previously described (Kawasaki et al., 1983). The resulting purified MBL was .95 % pure by Coomassie blue staining and Western
blots of reducing gels. Some MBL monomers (approximately
30 kDa) were observed on non-reducing gels, but the majority of
the purified material was estimated to contain multimers between 90
and 400 kDa.
Western blotting. Virus (10 ng p24) was lysed in SDS sample
buffer, separated by 7?5 % SDS-PAGE and transferred to nitrocellulose
membranes. Western blotting was performed as previously described
(Takefman et al., 1998) using 2?5 mg sheep anti-gp120 antibody ml21
and horseradish peroxidase-conjugated rabbit anti-sheep immunoglobulin (Biosource). Antibody binding to gp120 was detected with
an enhanced chemiluminescence detection system (Amersham)
followed by exposure to X-ray film (Fuji Photo Film Co.).
Glycosidase treatment. Virus (0?25 ml at 40 ng p24 ml21) was
METHODS
Cells and viruses. T lymphoblastic H9 cells were obtained from
the AIDS Research and Reference Reagent Program (ARRRP),
National Institutes of Health (NIH, Rockville, MD, USA), and
grown in RPMI 1640 medium containing 10 % foetal bovine serum
(FBS) and 50 mg gentamicin ml21 (all from BioWhittaker). Human
liver fibroblast (HLF) cells, obtained from Toshisuke Kawasaki
(Kyoto University, Japan), were grown in DMEM (BioWhittaker)
supplemented with 10 % FBS and gentamicin.
The HIV-1MN (X4) virus obtained from ARRRP (contributed by
Robert Gallo) was produced in H9 cells, and primary isolates of HIV
(HIVGP, HIVTH and HIVBa-L) were grown in peripheral blood
mononuclear cells (PBMCs) obtained from normal healthy donors,
as previously described (Takefman et al., 1998). The HIVGP (X4) and
HIVTH (R5) were previously isolated in our laboratory (Takefman
et al., 1998) and HIVBa-L (R5) was obtained from the ARRRP.
Phytohaemagglutinin (PHA; Sigma)-stimulated PBMCs were infected
with HIV primary isolates in the presence of 30 U recombinant
interleukin-2 ml21 (obtained through the ARRRP from Maurice
Gately, Hoffman LaRoche). For up to 10 days after infection, culture
supernatants were harvested and tested for virus production by p24
ELISA (AIDS Vaccine Program, Frederick, MD, USA). To produce
virus in the presence of inhibitors, PBMCs were infected for 7 days and
were washed and cultured for an additional 48 h in the presence or
absence of 1 mM Csp, dMM or dNM. Similarly, HIV-1MN-infected H9
cells were also washed and cultured for 48 h in the presence or absence
of the drugs.
Antibodies. Serum from an HIV-antibody-negative AB+ donor
was heat-inactivated (56 ˚C for 50 min) as a source of normal
354
treated with 1 U endoglycosidase F1 ml21 for 24 h at 37 ˚C in the
presence of 0?5 mM PMSF (Sigma). Virus samples were then diluted
to 500 ml with veronal-buffered saline (5 mM veronal, pH 7?5,
0?145 M NaCl) containing 10 mM CaCl2 (VBS-Ca). HIVTH was
treated with 0?1 U NA ml21 (Sigma) overnight.
Binding of HIV-1 to MBL. Ninety-six-well tissue culture plates
(Costar) were coated with 100 ml 10 mg rMBL ml21 diluted in VBSCa (Saifuddin et al., 2000). After overnight incubation at room temperature, wells were blocked with 3 % BSA for 1 h, washed with
VBS-Ca and then incubated for 4 h with 100 ml (500 pg p24) of
HIVGP, HIVTH or HIVBa-L grown in the presence or absence of glycosylation inhibitors. The plates were washed, bound virus was lysed
with 0?5 % Triton X-100 and p24 was measured by ELISA.
MBL Neutralization of HIV. The TCID50 of viruses produced in
the presence or absence of inhibitors was determined by limiting
dilution. Briefly, PHA-stimulated PBMCs (26105 cells) were infected
for 24 h at 37 ˚C with 0?1 ml of serial tenfold dilutions (1 : 10 to
1 : 100 000; three wells per dilution) of each virus in 96-well tissue
culture plates (Costar). Cells were washed and resuspended in 200 ml
culture medium containing IL-2 (20 U ml21). After 7 days, culture
wells were scored as positive or negative for virus growth by p24
ELISA. To assess HIV neutralization mediated by MBL, 100 TCID50
of virus grown in the presence or absence of glycosylation inhibitors
was treated with 10 mg rMBL ml21 for 2 h at 37C. Treated virus was
then cultured with PHA-stimulated PBMCs for 24 h. Cells were
washed and resuspended in 200 ml culture medium containing IL-2,
and on day 7 supernatants were assayed for virus production by p24
ELISA.
Immunocapture of HIV. Virus immunocapture was performed
as previously described (Takefman et al., 1998). Briefly, virus was
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Fri, 16 Jun 2017 22:14:29
Journal of General Virology 84
Interaction of MBL with altered HIV glycans
adjusted to 100 ng p24 ml21 and incubated on ice for 1 h with NHS
or HIVPS at a 1 : 10 dilution or with 50 mg 2F5, 2G12 or IgG1b12
antibody ml21. The virus–antibody mixtures were then ultracentrifuged over 20 % (w/v) sucrose in 50 mM Tris/HCl (pH 8?0) at
150 000 g for 1 h. Virus was resuspended in RPMI 1640 medium.
Approximately 200 pg p24 of purified virus was incubated overnight
at 4 ˚C with 50 ml Staphylococcus aureus cells expressing protein A
(Pansorbin). Cells were then washed with PBS containing 1 % BSA
and bound virus was lysed by treatment with 0?5 % Triton X-100
and p24 measured by ELISA.
electrophoretic mobility are similar to those observed in
several other studies using these three inhibitors (Dedera
et al., 1990; Gruters et al., 1987; Montefiori et al., 1988;
Walker et al., 1987). Similar changes in mobility were
observed with 0?5, 1, 2 and 4 of each inhibitor for both HIV
primary isolates and HIVMN (not shown). Therefore, 1 mM
of each inhibitor was used in all further experiments.
We hypothesized that culturing virus in the presence of either
a-glucosidase or mannosidase inhibitors would affect binding
RESULTS
The effects of N-linked glycosylation inhibitors
on MBL binding to HIV-1
To determine whether production of virus in the presence of
N-linked glycosylation inhibitors altered the glycosylation of
gp120, virus preparations grown in the presence of 1 mM
Csp, dNM or dMM were analysed by Western blotting. As
shown in Fig. 1(A), gp120 from HIVMN produced in the
presence of either Csp or dNM was slightly decreased in
mobility. In contrast, production of virus in the presence of
the mannosidase I inhibitor dMM markedly increased the
mobility of gp120. Similar changes in gp120 mobility were
observed for HIV primary isolates produced in PHAstimulated PBMCs (not shown). These changes in gp120
of virus to MBL, since it would potentially increase the
N-linked terminal mannose residues and decrease the terminal sialic acid residues exposed on gp120. To test this, virus
preparations were added to wells of microtitre plates coated
with MBL, and HIV captured by MBL was determined by
measuring p24 core protein. In a previous study, we found
that this capture method sensitively detected binding of both
HIV primary isolates and cell-line adapted virus strains to
MBL, and that HIV bound to the carbohydrate-recognition
domain of MBL, since binding did not occur in the absence of
calcium and was blocked by pre-incubation of MBL-coated
wells with mannan (Saifuddin et al., 2000).
Production of HIVTH in the presence of dMM significantly
increased virus binding to MBL while binding to BSAcoated wells was not affected (Fig. 1B). Binding of virus to
Fig. 1. The effect of glycosylation inhibitors on the interaction of MBL with HIV. (A) Immunoblot analysis of gp120 from virus
produced in the presence of inhibitors. H9 cells infected with HIV-1MN were grown in the presence or absence of 1 mM dMM,
Csp or dNM for 48 h. Virus particles were pelleted by ultracentrifugation and analysed by 7?5 % SDS-PAGE under reducing
conditions, followed by Western blot analysis for gp120. Similar results were obtained with HIVGP, HIVTH, and HIVBa-L primary
isolates produced in inhibitors. (B, C) MBL capture of HIV cultured in inhibitors. Microtitre wells were coated with MBL or
BSA and then incubated for 4 h with either HIVTH produced in the presence (%) or absence (&) of 1 mM dMM (B), or HIVGP,
HIVTH or HIVBa-L produced in the presence or absence of 1 mM dMM, Csp or dNM (C). Bound virus was detected by p24
ELISA after detergent lysis. Results are the mean¡SD and are representative of three experiments. *, P,0?05 (t-test)
comparing MBL binding of inhibitor-treated virus with binding of control-treated virus.
http://vir.sgmjournals.org
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Fri, 16 Jun 2017 22:14:29
355
M. L. Hart, M. Saifuddin and G. T. Spear
MBL was significantly enhanced for all three primary
isolates in multiple experiments (X4 HIVGP, R5 HIVTH,
P,0?0005, t-test; R5 HIVBa-L, P,0?05, t-test) when grown
in the presence of dMM (Fig. 1C). In contrast, virus binding
only slightly increased when grown in the presence of dNM,
while Csp treatment did not significantly increase binding
(Fig. 1C). None of the glycosylation inhibitors increased
binding to BSA-coated wells (not shown). The inhibitor
dMM also significantly increased the binding of HIVMN
produced in H9 cells to MBL, while Csp and dNM did not
(not shown).
N-linked high-mannose and hybrid oligosaccharides. MBL
binding to HIVTH grown in the presence of dMM was substantially reduced when virus was treated with eF1, providing evidence that the increase in virus binding to MBL
due to dMM was the result of an increase in the number
of N-linked high-mannose or hybrid glycans on the virus
surface (Fig. 2). Since the major difference between the structure of N-linked carbohydrates produced in the presence of
either a-glucosidase and mannosidase I inhibitors is two to
three terminal glucose residues, these data suggest that
terminal glucose residues substantially inhibit the interaction of MBL with N-linked high-mannose glycans.
To determine whether virus binding to MBL was mediated
by high-mannose-type carbohydrates on the virus, HIVTH
produced in the presence of dMM was treated with endoglycosidase F1 (eF1), an enzyme that preferentially cleaves
The influence of glucosidase and mannosidase
inhibitors on neutralization mediated by MBL
Virus-infected cells were grown in the presence of glycosylation inhibitors, and supernatants containing virus were
harvested. Viable cell numbers were not affected by the
inhibitors and no significant inhibition of virus production
was observed, since p24 values were similar in controltreated cultures and cultures with drug concentrations from
0?5 to 4 mM (not shown). However, HIVTH produced in
PBMCs treated with 1 mM of each of the inhibitors showed
an approximate 1 log reduction in infectious virus compared with virus produced from untreated cells (Fig. 3A).
Similar reductions in virus titres from all three inhibitors
were also observed for HIVGP and HIVBa-L (not shown).
Fig. 2. Effect of endoglycosidase F1 treatment on HIV binding to
MBL. HIVTH produced in PHA-stimulated PBMCs in the presence
(dMM) or absence (Control) of dMM was untreated or treated
with endoglycosidase F1 (Endo F1) and added to MBL-coated
wells. Bound virus was detected as described in Fig. 1. Results
are the mean¡SD and are representative of two experiments.
The ability of MBL to neutralize HIV primary isolates
produced in the presence of glycosylation inhibitors was
assessed. MBL mediated low levels of neutralization of the
three primary isolates (2–10 %) when produced in the
absence of glycosylation inhibitors (Fig. 3B). Production of
virus in the presence of dMM resulted in a trend towards
increased neutralization by MBL, since neutralization
increased 2?5- to 6-fold depending on the virus strain.
However, only the increase in neutralization of HIVTH
Fig. 3. Neutralization by MBL of HIV cultured in the presence of glucosidase and mannosidase inhibitors. (A) TCID50 of HIVTH
cultured in the presence or absence of dMM, Csp or dNM. Similar results were obtained with HIVGP and HIVBa-L. (B)
Neutralization by MBL of HIVGP, HIVTH, and HIVBa-L cultured in the presence or absence of Csp, dNM, or dMM. HIV (100
TCID50) produced in the presence or absence of inhibitors was treated with 20 mg MBL ml21 or medium for 2 h. Treated virus
was added to PHA-stimulated PBMCs and cultured for 7 days. The data shown are the mean of three experiments¡SD. *,
P,0?05 (ANOVA Fischer PLSD) comparing neutralization by MBL of drug-treated virus with neutralization by MBL of
untreated virus.
356
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Fri, 16 Jun 2017 22:14:29
Journal of General Virology 84
Interaction of MBL with altered HIV glycans
was significant at the P,0?05 level, while HIVBa-L showed
near significant neutralization (P,0?10). In contrast, virus
production in the presence of Csp or dNM did not significantly increase neutralization by MBL.
Since one effect of these three inhibitors is to prevent
addition of sialic acid to N-linked carbohydrates, we also
assessed the effect of NA removal of sialic acid on MBLmediated neutralization and binding. To confirm that NA
treatment removed sialic acid from gp120, a gp120 Western
blot was performed. Treatment of virus with NA removed
a substantial amount of sialic acid residues from gp120,
as evidenced by a change in mobility to approximately
105 kDa (Fig. 4A). NA treatment significantly increased
neutralization by MBL (P,0?0005, t-test) (Fig. 4B). The
increase in MBL-mediated neutralization due to NA treatment corresponded to an approximate fourfold increase
in binding of virus to MBL (Fig. 4C). Interestingly, eF1 did
not significantly decrease the amount of NA-treated virus
binding to MBL, suggesting that NA exposed MBL binding
sites on the virus that were not N-linked high-mannose
glycans (Fig. 4C).
Effect of dMM on interaction with anti-HIV
antibodies
The results shown above indicate that changes in the
glycosylation of HIV affect virus interaction with MBL.
Experiments were also performed to determine whether
the dMM-induced changes in carbohydrates on intact virus
affected the immunoreactivity of anti-gp120 or anti-gp41
antibodies. Virus was grown in the presence or absence of
dMM and assessed for binding to either several monoclonal
antibodies or HIVPS. Production of HIV in the presence
of dMM had no effect on the amount of virus bound by
IgG1b12 anti-CD4 binding site monoclonal antibody or
HIVPS (Fig. 5). However, treatment with dMM significantly increased virus immunocapture by the anti-gp41 2F5
monoclonal antibody (Muster et al., 1993). A significant
increase in immunocapture due to dMM was also seen for
the 2G12 antibody, which recognizes a conformational,
carbohydrate-dependent epitope in the C3–C4 region of
gp120 (Trkola et al., 1996). These results indicated that while
production of virus in the presence of dMM increases virus
interaction with MBL, alteration of carbohydrates on gp120/
gp41 by dMM can also enhance the immunoreactivity of
HIV with some antiviral antibodies but not others.
DISCUSSION
Fig. 4. Effects of neuraminidase treatment on MBL-mediated
neutralization and binding of HIV. (A) Western blot analysis of
HIVTH gp120 treated (NA) or untreated (Control) with NA. (B)
Neutralization by MBL of untreated (Control) and NA-treated
HIVTH. Each virus (100 TCID50) was treated with MBL or medium
for 2 h. Treated virus was added to PHA-stimulated PBMCs and
cultured for 7 days. The data shown are the mean¡SD of triplicate cultures and are representative of three experiments. *,
P,0?0005 (t-test). (C) Effect of endoglycosidase F1 on virus
binding. Virus was added to microtitre wells coated with MBL.
The amount of bound virus was detected as described in Fig. 1.
Results are the mean¡SD and are representative of three experiments. *, P,0?05, comparing MBL binding of NA-treated virus or
eF1/NA-treated virus with binding of untreated virus.
This study has shown that alteration of carbohydrates on
HIV can increase binding and neutralization of virus by
MBL. Binding and neutralization by MBL were significantly
increased when virus was produced in the presence of the
mannosidase I inhibitor dMM or when virus was treated
with NA. In contrast, MBL binding and neutralization were
relatively unchanged when virus was produced in the presence of Csp and dNM, inhibitors of a-glucosidase I and II,
respectively. Treatment with the mannosidase I inhibitor
results in N-linked glycans containing approximately nine
mannose residues terminated with mannose while treatment with the glucosidase inhibitors yields similar glycans
terminated with two or three glucose residues (Sears &
Wong, 1998). Therefore, this study has shown that increasing
http://vir.sgmjournals.org
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Fri, 16 Jun 2017 22:14:29
357
M. L. Hart, M. Saifuddin and G. T. Spear
Fig. 5. Effect of dMM on immunocapture of HIV. HIVTH cultured in the presence or absence of dMM was incubated with
monoclonal antibodies 2F5, 2G12 or IgG1b12, or HIV-positive pooled sera (HIVPS). Virus was then purified by
ultracentrifugation through a sucrose layer and incubated overnight with protein A–S. aureus. Bound virus was lysed and
detected by p24 ELISA. Results are the mean¡SD of three experiments. *, P,0?05, comparing control and dMM.
the high-mannose glycans on HIV increases virus neutralization by MBL. Furthermore, the data suggest that terminal glucose residues inhibit the interaction of MBL with
N-linked glycans.
Treatment with NA also substantially increased virus neutralization and binding by MBL. Removal of sialic acid by
NA has several potential effects on MBL binding, including
reducing negative charge on the virus, reducing the size of
the complex glycan and revealing galactose as a terminal
sugar residue. However, the increase in virus neutralization
by MBL was not likely to be due to a reduction of the negative charge or a decrease in the size of the glycans, since the
a-glucosidase inhibitors would have had these same effects
on the virus but had little effect on neutralization.
Furthermore, in contrast to virus produced in the presence
of dMM, eF1 treatment did not reduce the amount of NAtreated virus binding to MBL. Together, these data indicate
that NA treatment increased neutralization by exposing new
MBL binding sites on the virus that were distinct from
N-linked high-mannose carbohydrates, although it is possible
that the identity of the binding sites is unclear since it is reported that MBL binds well to terminal N-acetylglucosamine,
mannose, glucose and fucose residues, but binding is reduced
or blocked by terminal galactose (Childs et al., 1989). However,
although sparse, some terminal N-acetylglucosamine residues are normally present on N-linked complex structures
including gp120 (Malhotra et al., 1995; Scanlan et al., 2002),
and it is possible that removing sialic acid provides better
access of MBL to these residues.
While MBL is a serum protein, it reacts weakly with host
high-mannose N-linked carbohydrates due to the relatively
low density of these glycans and the low affinity of MBL
monomers for sugars. In contrast, the multimeric nature of
both native MBL and terminal MBL-binding sugar residues
on microbes results in a high-avidity interaction. Also, the
conformation of MBL suggests that it does not react with
multiple terminal mannose residues within one N-linked
358
high-mannose carbohydrate structure and therefore reactivity of MBL with one high-mannose carbohydrate structure is likely to be of low avidity (Weis et al., 1998). In the
case of HIV gp120, there is a very high density of N-linked
carbohydrates (about half being the high-mannose type),
which most likely accounts for the interaction of HIV with
MBL. Thus, MBL reacts strongly with HIV but does not bind
to virus particles lacking gp120 (Saifuddin et al., 2000) and
we observed in the current study that treatment of HIV with
eF1 substantially reduced binding of virus to MBL. While
treatment with the mannosidase I inhibitor would change
host-cell protein complex glycosylation to the high-mannose
type and could thus contribute to MBL binding to virus, we
propose that the majority of the dMM-mediated increase in
binding between HIV and MBL is most likely to be due to an
increase in high-mannose carbohydrates on gp120, since
dMM treatment would most likely double the density of
MBL binding sites on gp120. Since the complex carbohydrates on gp120 are thought to be relatively exposed, their
change to the high-mannose type may have an especially
large effect on MBL binding (Moore et al., 1994; Wyatt et al.,
1998).
Two glycosylation inhibitor drugs have been tested in HIVinfected people since early studies showed that some of
these inhibitors had antiviral activity in vitro (Jacob, 1995).
However, both in vivo trials used a-glucosidase inhibitors.
According to our studies, this alteration of carbohydrates
would probably not have substantially affected the interaction of virus with MBL in vivo. Mannosidase inhibitors have
been used in vivo in trials to inhibit tumour metastasis
(Jacob, 1995). Our data suggest that administration of
mannosidase inhibitors during HIV infection could increase
MBL-mediated antiviral effects such as neutralization, complement activation and virus clearance. These enhanced MBL
effects could also change the way that virus is presented to
the adaptive immune system by affecting virus uptake by
antigen-presenting cells.
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Fri, 16 Jun 2017 22:14:29
Journal of General Virology 84
Interaction of MBL with altered HIV glycans
While the binding of MBL to virus was increased by dMM,
virus binding to monoclonal antibodies 2F5 and 2G12 was
also increased, although the virus interaction with pooled
serum and monoclonal antibody IgG1b12 was unchanged.
These results could be partially explained by studies showing
that complex-type glycans have a more pronounced shielding effect on antibody binding and are located on more
exposed regions of gp120 than high-mannose glycans (Back
et al., 1994; Moore et al., 1994; Wyatt et al., 1998). Therefore,
by inhibiting production of bulkier complex carbohydrates
with dMM, virus is produced that binds at higher levels to
some neutralizing antibodies. Alternatively, recent studies
have shown that the 2G12 epitope is not only dependent on
carbohydrates but also appears to be at least partially composed of mannose residues (Sanders et al., 2002; Scanlan
et al., 2002). Therefore, it is possible that treatment with
dMM increases the density of mannose residues on gp120
causing the increased binding of 2G12. Thus, alteration of
carbohydrates can simultaneously affect the interaction of
HIV with arms of both the innate and adaptive immune
responses.
A study by Means & Desrosiers (2000) assessed the effect of
glycosidases and glycosylation inhibitors on neutralization
of simian immunodeficiency virus (SIV) by pooled sera
from infected macaques. Treatment of two SIV strains with
NA decreased neutralization sensitivity while other glycosidases did not have a consistent effect on sensitivity. Treatment of virus-producing cells with either the mannosidase
inhibitor swainsonine or DANA, an inhibitor of sialic acid,
also slightly decreased neutralization sensitivity. These results
suggest that removal of sialic acid somehow increases the
resistance of virus to antibody neutralization by the pooled
sera, a somewhat unexpected result given that the bulk of complex glycans is thought to shield the envelope protein from
antibody binding (Wyatt et al., 1998). The effect of NA and
swainsonine in the SIV system are also interesting in light of
our results with MBL neutralization of HIV, since we found
that NA treatment of virus or dMM treatment of virusproducing cells increased MBL-mediated neutralization.
This is the first study to assess MBL-mediated effects on HIV
after treatment of virus-producing cells with N-linked glycosylation inhibitors or treatment of virus with NA. The results
show that, by altering the glycosylation of the envelope
protein of HIV with dMM or NA, neutralization by and
binding of MBL is increased. It is not known whether dMM
or related compounds offer a realistic prospect of therapy
for those infected with HIV, but this study suggests that
further exploration of this possibility is warranted. Important issues to be considered before therapies are attempted
include whether effective in vivo drug concentrations can be
achieved and the resultant side effects. No clinical trials of
dMM have yet been reported. A trial in HIV-infected people
with the a-glucosidase inhibitor N-butyl-deoxynojirimycin
resulted in significant side effects including diarrhoea,
flatulence, leukopenia and neutropenia (Tierney et al., 1995)
and the concentration achieved in serum was slightly lower
http://vir.sgmjournals.org
than the antiviral concentration predicted in vitro. Many
of the glycosylation inhibitor drugs cause similar side
effects (Jacob, 1995). N-nonyl-deoxynojirimycin, another
a-glucosidase inhibitor, was shown to substantially reduce
viraemia in hepadnavirus-infected woodchuck (Block et al.,
1998). Also, the a-mannosidase I inhibitor swainsonine was
given to cancer patients over 5 days and serum levels of the
drug were achieved that were much higher than the 50 %
in vitro inhibitory concentrations (Baptista et al., 1994). The
above studies suggest that, while inhibitory concentrations
of some of the glycosylation inhibitors could be achieved
in vivo for treatments of short duration, long-term treatment
with these types of drug, which is likely to be needed in the
case of HIV-infected people, may be difficult due to their
toxicity.
ACKNOWLEDGEMENT
This work was supported by National Institutes of Health Grant
# AI46963.
REFERENCES
Back, N. K., Smit, L., De Jong, J. J., Keulen, W., Schutten, M.,
Goudsmit, J. & Tersmette, M. (1994). An N-glycan within the
human immunodeficiency virus type 1 gp120 V3 loop affects virus
neutralization. Virology 199, 431–438.
Baptista, J. A., Goss, P., Nghiem, M., Krepinsky, J. J., Baker, M. &
Dennis, J. W. (1994). Measuring swainsonine in serum of cancer
patients: phase I clinical trial. Clin Chem 40, 426–430.
Barbas, C. F., III Bjorling, E., Chiodi, F. & 8 others. (1992).
Recombinant human Fab fragments neutralize human type 1
immunodeficiency virus in vitro. Proc Natl Acad Sci U S A 89,
9339–9343.
Block, T. M., Lu, X., Mehta, A. S. & 7 other authors (1998).
Treatment of chronic hepadnavirus infection in a woodchuck animal
model with an inhibitor of protein folding and trafficking. Nat Med
4, 610–614.
Childs, R. A., Drickamer, K., Kawasaki, T., Thiel, S., Mizuochi, T. &
Feizi, T. (1989). Neoglycolipids as probes of oligosaccharide
recognition by recombinant and natural mannose-binding proteins
of the rat and man. Biochem J 262, 131–138.
Dedera, D., Vander Heyden, N. & Ratner, L. (1990). Attenuation of
HIV-1 infectivity by an inhibitor of oligosaccharide processing. AIDS
Res Hum Retrov 6, 785–794.
Ezekowitz, R. A., Kuhlman, M., Groopman, J. E. & Byrn, R. A. (1989).
A human serum mannose-binding protein inhibits in vitro infection
by the human immunodeficiency virus. J Exp Med 169, 185–196.
Gruters, R. A., Neefjes, J. J., Tersmette, M., de Goede, R. E., Tulp, A.,
Huisman, H. G., Miedema, F. & Ploegh, H. L. (1987). Interference
with HIV-induced syncytium formation and viral infectivity by
inhibitors of trimming glucosidase. Nature 330, 74–77.
Haurum, J. S., Thiel, S., Jones, I. M., Fischer, P. B., Laursen, S. B. &
Jensenius, J. C. (1993). Complement activation upon binding of
mannan-binding protein to HIV envelope glycoproteins. AIDS 7,
1307–1313.
Jack, D. L., Klein, N. J. & Turner, M. W. (2001). Mannose-binding
lectin: targeting the microbial world for complement attack and
opsonophagocytosis. Immunol Rev 180, 86–99.
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Fri, 16 Jun 2017 22:14:29
359
M. L. Hart, M. Saifuddin and G. T. Spear
Jacob, G. S. (1995). Glycosylation inhibitors in biology and
medicine. Curr Opin Struct Biol 5, 605–611.
Karpas, A., Fleet, G. W., Dwek, R. A., Petursson, S., Namgoong, S. K.,
Ramsden, N. G., Jacob, G. S. & Rademacher, T. W. (1988).
Aminosugar derivatives as potential anti-human immunodeficiency
virus agents. Proc Natl Acad Sci U S A 85, 9229–9233.
Kawasaki, N., Kawasaki, T. & Yamashina, I. (1983). Isolation and
characterization of a mannan-binding protein from human serum.
J Biochem 94, 937–947.
Leonard, C. K., Spellman, M. W., Riddle, L., Harris, R. J., Thomas,
J. N. & Gregory, T. J. (1990). Assignment of intrachain disulfide
bonds and characterization of potential glycosylation sites of the type
1 recombinant human immunodeficiency virus envelope glycoprotein (gp120) expressed in Chinese hamster ovary cells. J Biol Chem
265, 10373–10382.
Ma, Y., Shida, H. & Kawasaki, T. (1997). Functional expression of
human mannan-binding proteins (MBPs) in human hepatoma cell
lines infected by recombinant vaccinia virus: post-translational
modification, molecular assembly, and differentiation of serum and
liver MBP. J Biochem 122, 810–818.
Malhotra, R., Wormald, M. R., Rudd, P. M., Fischer, P. B., Dwek, R. A.
& Sim, R. B. (1995). Glycosylation changes of IgG associated with
rheumatoid arthritis can activate complement via the mannosebinding protein. Nat Med 1, 237–243.
Means, R. E. & Desrosiers, R. C. (2000). Resistance of native,
oligomeric envelope on simian immunodeficiency virus to digestion
by glycosidases. J Virol 74, 11181–11190.
Mehta, A., Zitzmann, N., Rudd, P. M., Block, T. M. & Dwek, R. A.
(1998). Alpha-glucosidase inhibitors as potential broad based anti-
viral agents. FEBS Lett 430, 17–22.
Montefiori, D. C., Robinson, W. E., Jr. & Mitchell, W. M. (1988). Role
of protein N-glycosylation in pathogenesis of human immunodeficiency virus type 1. Proc Natl Acad Sci U S A 85, 9248–9252.
Moore, J. P., Wallace, L. A., Follett, E. A. & McKeating, J. A. (1989).
An enzyme-linked immunosorbent assay for antibodies to the envelope glycoproteins of divergent strains of HIV-1. AIDS 3, 155–163.
Moore, J. P., Sattentau, Q. J., Wyatt, R. & Sodroski, J. (1994).
Probing the structure of the human immunodeficiency virus surface
glycoprotein gp120 with a panel of monoclonal antibodies. J Virol
68, 469–484.
Muster, T., Steindl, F., Purtscher, M., Trkola, A., Klima, A., Himmler, G.,
Ruker, F. & Katinger, H. (1993). A conserved neutralizing epitope on
gp41 of human immunodeficiency virus type 1. J Virol 67, 6642–6647.
Neth, O., Jack, D. L., Dodds, A. W., Holzel, H., Klein, N. J. & Turner,
M. W. (2000). Mannose-binding lectin binds to a range of clinically
relevant microorganisms and promotes complement deposition.
Infect Immun 68, 688–693.
Petersen, S. V., Thiel, S. & Jensenius, J. C. (2001). The mannan-
binding lectin pathway of complement activation: biology and
disease association. Mol Immunol 38, 133–149.
Purtscher, M., Trkola, A., Gruber, G., Buchacher, A., Predl, R.,
Steindl, F., Tauer, C., Berger, R., Barrett, N., Jungbauer, A. and
others (1994). A broadly neutralizing human monoclonal antibody
against gp41 of human immunodeficiency virus type 1. AIDS Res
Hum Retroviruses 10, 1651–1658.
Saifuddin, M., Hart, M. L., Gewurz, H., Zhang, Y. & Spear, G. T.
(2000). Interaction of mannose-binding lectin with primary
isolates of human immunodeficiency virus type 1. J Gen Virol 81,
949–955.
Sanders, R. W., Venturi, M., Schiffner, L., Kalyanaraman, R.,
Katinger, H., Lloyd, K. O., Kwong, P. D. & Moore, J. P. (2002).
The mannose-dependent epitope for neutralizing antibody 2G12 on
human immunodeficiency virus type 1 glycoprotein gp120. J Virol
76, 7293–7305.
Scanlan, C. N., Pantophlet, R., Wormald, M. R. & 7 other authors
(2002). The broadly neutralizing anti-human immunodeficiency
virus type 1 antibody 2G12 recognizes a cluster of a1R2 mannose
residues on the outer face of gp120. J Virol 76, 7306–7321.
Sears, P. & Wong, C. H. (1998). Enzyme action in glycoprotein
synthesis. Cell Mol Life Sci 54, 223–252.
Takefman, D. M., Sullivan, B. L., Sha, B. E. & Spear, G. T. (1998).
Mechanisms of resistance of HIV-1 primary isolates to complementmediated lysis. Virology 246, 370–378.
Tierney, M., Pottage, J., Kessler, H. & other authors (1995). The
tolerability and pharmacokinetics of N-butyl-deoxynojirimycin in
patients with advanced HIV disease (ACTG 100). The AIDS Clinical
Trials Group (ACTG) of the National Institute of Allergy and
infectious Diseases. J Acquir Immune Defic Syndr Hum Retrovirol
10, 549–553.
Trkola, A., Purtscher, M., Muster, T. & 7 other authors (1996).
Human monoclonal antibody 2G12 defines a distinctive neutralization epitope on the gp120 glycoprotein of human immunodeficiency
virus type 1. J Virol 70, 1100–1108.
van Emmerik, L. C., Kuijper, E. J., Fijen, C. A., Dankert, J. & Thiel, S.
(1994). Binding of mannan-binding protein to various bacterial
pathogens of meningitis. Clin Exp Immunol 97, 411–416.
Walker, B. D., Kowalski, M., Goh, W. C., Kozarsky, K., Krieger, M.,
Rosen, C., Rohrschneider, L., Haseltine, W. A. & Sodroski, J. (1987).
Inhibition of human immunodeficiency virus syncytium formation
and virus replication by castanospermine. Proc Natl Acad Sci U S A
84, 8120–8124.
Weis, W. I., Taylor, M. E. & Drickamer, K. (1998). The C-type lectin
superfamily in the immune system. Immunol Rev 163, 19–34.
Ohtani, K., Suzuki, Y., Eda, S. & 7 other authors (1999). High-level
Wyatt, R., Kwong, P. D., Desjardins, E., Sweet, R. W., Robinson, J.,
Hendrickson, W. A. & Sodroski, J. G. (1998). The antigenic
and effective production of human mannan-binding lectin (MBL) in
Chinese hamster ovary (CHO) cells. J Immunol Methods 222, 135–144.
structure of the HIV gp120 envelope glycoprotein. Nature 393,
705–711.
360
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Fri, 16 Jun 2017 22:14:29
Journal of General Virology 84