Chem 261, 11726C11733

Chem 261, 11726C11733. establish that these O-linked glycans can confer a more than 1,000-fold decrease in neutralization sensitivity (IC50) to V3-glycan broadly neutralizing antibodies. These findings uncover a structural modification to the HIV-1 Env and suggest a functional role in promoting viral escape from one category of broadly neutralizing antibodies. In Brief Metallic et al. demonstrate that certain HIV-1 isolates possess O-linked carbohydrate on their Envelope glycoprotein. These sugars allow the computer virus to evade V3-glycan broadly neutralizing antibodies. This Roburic acid work identifies a post-translational modification to the HIV-1 Envelope and sheds light on its role in shielding against the host antibody response. Graphical Abstract INTRODUCTION The addition of O-linked carbohydrate to proteins in the secretory pathway, or mucin-type O-glycosylation, is the most abundant type of O-linked carbohydrate attachment in humans (Tian and Ten Hagen, 2009). Mucin-type O-glycosylation usually begins with the addition of N-acetylgalactosamine to the hydroxyl group of serine or threonine, thus forming the Tn antigen (GalNac-Ser/Thr) (Hart and Akimoto, 2009). The Tn antigen is usually then further altered to form one of eight core mucin-type O-linked carbohydrate structures. Although dense stretches of serines and threonines in a protein sequence can be preferential for O-glycosylation, little is known regarding the specific sequences or structural determinants that guideline the attachment of O-linked carbohydrates (Gooley and Williams, 1994). However, informatic tools such as the NetOGlyc4.0 prediction server may be used to indicate the likelihood that a particular serine or threonine Roburic acid will be O-glycosylated (Steentoft et al., 2013). The Envelope (Env) glycoprotein of HIV-1 consists of a heterotrimeric complex of gp120 and gp41 (Arajo and Almeida, 2013). Although HIV-2 and SIVmac239 are O-glycosylated around the V1 domain name of gp120, the current dogma is usually that HIV-1 lacks such O-glycosylation (Stansell et al., 2011, 2015). Using the same lectin-binding procedures that readily exhibited O-glycosylation around the gp120s of a panel of SIVmac, SIVsm, and HIV-2 strains, Stansell et al. (2011) failed to Mouse monoclonal to CHUK find evidence for O-glycosylation on 11 of 11 strains of HIV-1 and SIVcpz. When HIV-1 gp120 is usually expressed as a secreted product, the C-terminal threonine at position 499 is usually O-glycosylated (Go et al., 2013; Stansell et al., 2015). However, when HIV-1 gp120 is made naturally from its gp160 precursor and analyzed from purified virions, there is no O-linked carbohydrate (Stansell et al., 2015). Furthermore, when HIV-1 is usually produced in a knockout (KO) cell line that is incapable of adding O-linked carbohydrate, the HIV-1 retains full infectivity (Termini et al., 2017a, 2017b). We recently described a primary HIV-1 isolate from an elite controller that displayed an unusually long V1 domain name Roburic acid in Env gp120 (Silver et al., 2019). Careful analysis of this V1 domain name revealed stretches of serines and threonines with positive O-glycosylation prediction scores by the NetOGlyc4.0 algorithm. We therefore hypothesized that, like its SIV and HIV-2 counterparts, some HIV-1 Envs with long V1 domains may be O-glycosylated on V1. In this study, we show that there indeed exists a subset of patient-derived HIV-1 strains whose long V1 domains are altered with O-linked carbohydrate. Furthermore, we show that these O-glycans can result in more than a 1,000-fold decrease in sensitivity to broadly neutralizing antibodies (bNAbs) of the V3-glycan category. Our results uncover a previously unknown structural modification to HIV-1 Env and suggest a role in viral escape from the host antibody response. RESULTS Long V1 Domains Have Increased O-Glycosylation Prediction Scores To determine whether HIV-1 Envs with long V1 domains might be O-glycosylated, we downloaded a curated web alignment of 6,112 Group M (A-K + recombinants) HIV-1 Env sequences from the Los Alamos National Laboratory (LANL) online database (https://www.hiv.lanl.gov). A total of 4,757 sequences remained in the amino acid alignment after removing sequences with character types (X, #, ?, and $) that interfere with the O-glycosylation prediction algorithm. These Env sequences were then systematically uploaded to the NetOGlyc4.0 prediction server to assign an O-glycosylation score between 0.00 and 1.00 to each serine Roburic acid or threonine in a particular Env sequence. A Python-based computer algorithm was devised.