April 30-May 1, 2012, Masur Auditorium, NIH
The Vaccine Research Program of the Division of AIDS of the National Institute of Health (DAIDS) co-sponsored a 1.5 day workshop entitled “Functional Glycomics in HIV Vaccine Design.” The objective of the workshop was to bring together glycobiologists, virologists and immunologists to promote synergistic multi-disciplinary HIV-glycomics research and by doing so explore new avenues for HIV vaccine design research. The workshop reviewed the role and function of HIV envelope glycans and their interaction with neutralizing antibodies, current tools to characterize glycosylation patterns and the role and function of immunoglobulin glycosylation.
Glycans are ubiquitous and most proteins are subject to post-translational modifications, of which glycosylation is the most common form. These glycan modifications follow a series of processing steps taking place at the endoplasmic reticulum and the Golgi apparatus. A diverse array of enzymes, glycosyltransferases and glycosidases, are involved, resulting in multiple alternative glycosylation motifs for each glycosylation site. Structure and degree of glycosylation vary significantly across different cell types, cell states and tissues which increases the difficulty to characterize these patterns.
HIV-1 envelope glycans play key roles in transmission, antigenicity and immunogenicity. HIV gp120 has between 25-30 N-linked glycosylation sites and the carbohydrate structures on gp120 comprise half of its mass. Although N-linked glycans are the most common sugars on the envelope, O-linked glycans can also be found. N-linked glycans are linked to the protein via an aspargine residue in an Asn-X-Ser/Thr motif, where X can be any amino acid, except Pro. O-linked glycans are linked to either a serine or a threonine residue. Understanding the function of N- and O-linked glycans on the trimer beyond merely shielding the virus from the immune system will be critical. Even though the glycan profiles are extremely heterogeneous, cryo-electron tomography density maps suggest that some glycan domains are quite conserved. It seems that on the native trimer, gp41 is also surrounded by glycans that are parallel to the viral membrane (J. Sodroski).
Over the course of the infection the number of potential gly-sites on gp120 increases and is associated with viral escape. However, the lower number of N-linked gly sites in the transmitted virus is a trend, not a rule. J. Arthos showed data on the role of a leader peptide found in transmitted variants in modulating glycan patterns. Inserting the transmitted leader sequence into an HIV sequence from chronic stage resulted in altered gp120 glycan processing, increased DC-SIGN-binding activity and modified the mature gp120 subunit structure.
Several broadly neutralizing antibodies (Abs) to HIV are glycan dependent. 2G12 recognizes a high-mannose glycan cluster on gp120, PG9/16 recognize glycans on the V1/V2 loop, Abs from the PGT family recognize Man8 or Man9GlcNAc2 glycans on position 332 or 301 (PGT 125-128, 130 and 131) and PGT121-123 are N332 dependent (P. Kwong, I. Wilson).
D. Burton presented their work aimed to elicit 2G12-like antibodies. A key feature of 2G12 is the domain exchange of the heavy-chain (VH) with the VH’ of the adjacent Fab of the same immunoglobulin, resulting in the loss of the typical “Y” antibody structure. A small number of mutations are crucial for domain exchange: I19 at the VH/VH' interface, P113 in the elbow region, A14 and E75. Generation of a non-domain exchange 2G12 antibody is possible by introducing a single mutation on I19R.
P. Berman characterized glycosylation patterns of the recombinant gp120 proteins, derived from MN and A244 viral strains, used in the RV144 trial when produced in CHO or 293 cells. Production in 293 cells both MN and A244 gp120 proteins have more glycoforms but less complex sugars compared to production in CHO cells. Interestingly, PG9 does not bind to MN produced in 293 or CHO cells, but the binding was restored when it was produced in cells that promote addition of high mannose glycans.
Glycans also modulate the immune response, playing an essential role in Fc-receptor and complement activity as well as innate immune responses. Glycosylation patterns of immunoglobulins (Ig) are critical for pro- and anti-inflammatory activity. For example, de-fucosylation of the Fc domain increases ADCC activity, whereas higher levels of sialylation reduce the affinity of IgG for Fcγ-receptors and increases the affinity of IgG for SIGN-R1/DC-SIGN (F. Nimmerjahn). Over 90% of the innate immune receptors sense sugars on glycoproteins. G. Alter looked at glycosylation patterns of HIV-specific antibodies (looking at the number of galactose and fucose molecules) and compared those to glycosylation patterns of total antibodies (bulk). HIV-specific antibodies seem to be less fucosylated and less sialylated than bulk antibodies.
Current tools to characterize glycosylation patterns were also reviewed, including: mass spectrometry, nuclear magnetic resonance spectroscopy (NMR), lectin and antibody arrays, glycoarrays and others (R. Cummings, J. Prestegard and H. Desaire).
In summary, the workshop discussed the role of glycosylation on the HIV envelope and HIV-specific antibodies and the need to better understand glycan-dependent immune responses and utilize this knowledge for vaccine design research. Tools to characterize carbohydrate profiles exist, however the sheer numbers of possible glycan structures is a big impediment to better characterize their role and map glycans to specific functions.
The workshop was co-sponsored by several NIAID divisions (DAIT and DAIDS), other ICs (NIGMS, NCI), the Foundation for NIH (FNIH), the Global HIV Vaccine Enterprise, GlycoMimetics, IAVI, the Society for Glycobiology, Merck and Novartis.