Vanderbilt Institute of Chemical Biology



Discovery at the VICB







The Hidden Arsenal for Combatting HIV


By: Carol A. Rouzer, VICB Communications
Published:  April 20, 2016



Individuals who were never exposed to HIV carry prototypical broadly neutralizing antibodies that may be the foundation for an effective antiviral vaccine response.


A successful vaccine against the human immunodeficiency virus (HIV) that causes acquired immune deficiency syndrome (AIDS) must be able to induce production of broadly neutralizing antibodies (bnAbs). Effective against multiple strains of the rapidly mutating virus, bnAbs occur naturally during the course of infection in many patients; however they usually do not appear until at least a year after viral exposure, and they peak only after three to four years. One reason for this is that generation of many bnAbs requires a long maturation process involving multiple somatic mutations, clearly not a desirable requirement for a vaccine response. In contrast, the naturally occurring PG9 and PG16 bnAbs are notable for their relatively small number of somatic mutations. These bnAbs neutralize multiple strains of HIV-1 by binding to the V1/V2 epitope of the viral gp120/gp41 envelop glycoprotein (Figure 1). This interaction requires an unusually long complementarity-determining region 3 domain (HCDR3) in the variable portion of the antibody’s heavy chain (Figure 2). The ability of the PG9/PG16 HCDR3 to form a hammerhead-like structure (Figure 3) is key to its interaction with the V1/V2 epitope. The existence of naturally occurring PG9 and PG16 antibodies with this structure led VICB member Jens Meiler and his collaborator, James Crowe of Vanderbilt’s Department of Pathology, Immunology, and Microbiology, to hypothesize that similar antibodies may be present in many HIV-naïve individuals. If so, a vaccine that targets induction of these antibodies could be highly effective against HIV-1. The Meiler and Crowe labs now report that, indeed, such antibodies can be found in the human population [J. R. Willis, et al. (2016) Proc. Natl. Acad. Sci. U.S.A., published online April 4, DOI:10.1073/pnas.1518405113].




FIGURE 1. Schematic diagram of the human immunodeficiency virus. The viral RNA genome and reverse transcriptase are surrounded by capsid proteins, and the capsid is, in turn, surrounded by the membrane. The Env protein, comprising gp41 transmembrane trimers complexed with gp120 trimers, decorates the outer surface of the virus, forming the viral spike that plays a key role in fusion with a target cell membrane. Reprinted from Wikimedia Commons. Figure produced by the U.S. National Institutes of Health, public domain.


FIGURE 2. Diagram of antibody structure. The protein comprises two heavy and two light chains arranged in a Y-shaped complex. Both chains have variable domains at the tips of the arms of the Y where antigen binds. Each variable domain contains three complementarity-determining regions (CDRs), which come into direct contact with the bound antigen. Reproduced from Wikimedia Commons under the Creative Commons Attribution-Share Alike 3.0 Unported license.




FIGURE 3. Crystal structure of the Fab portions of PG9 bound to the V1/V2 domain of pg120. The heavy and light chains of PG9 are in yellow and blue, respectively. Each CDR in the heavy and light chains is labeled. HCDR3 (labeled CDR H3) can be seen projecting into the glycan attached to asparagine-160 (purple) of V1/V2. Figure reproduced by permission from Macmillan Publishers, Ltd. from J. S. McLellan, et al., (2011) Nature, 480, 336. Copyright 2011.09.



The investigators began their research by obtaining white blood cells from 70 individuals that tested negative for HIV. They isolated the RNA from these cells and used next-generation sequencing techniques to identify a total of 2.3 x 107 HCDR3 sequences from that RNA (Figure 4A). Bioinformatic analysis enabled them to select the 26,917 sequences that were of the same length (30 amino acids) as the HCDR3 region of wild-type PG9 (Figure 4B). Of these, they randomly chose 4,000 sequences for greater scrutiny. Using the ROSETTA software platform, the researchers threaded each of the chosen sequences over the hammerhead structure of the wild-type PG9 HCDR3 obtained from the crystal structure of the antibody complexed with the V1/V2 epitope. Energy minimization of these threaded structures revealed three classes of sequences: those that assumed the hammerhead structure with unfavorable energy, those that reached a favorable energy but failed to retain the hammerhead structure, and those that retained the hammerhead structure with favorable energy. Further evaluation of the data enabled the investigators to construct a matrix that revealed the ROSETTA energy of each of the 20 amino acids at every one of the 30 sites of the sequence. Using this matrix, they evaluated all 26,917 of their 30 amino acid-long sequences, identifying 1,000 that exhibited the highest predicted energetic favorability (Figure 4C).




FIGURE 4. Experimental design and methodology. (A) Peripheral blood mononuclear cells (PMBCs, a form of white blood cell) from 70 HIV-negative donors were isolated, and RNA was extracted. Next-generation sequencing was then used to acquire a total of 2.3 x 107 HCDR3 sequences from these samples. (B) Bioinformatic analysis of the HCDR3 sequences identified those (26,917) with a length of 30 amino acids. (C) A randomly chosen group of 4,000 of the sequences was threaded over the HCDR3 structure of PG9 to energetically evaluate their ability to assume the hammerhead structure. The results led to the selection of 1,000 most favored sequences for more definitive structural evaluation. Eventually, 100 sequences were chosen for further refinement by mutational maturation. (D) Finally, 84 HDCR3 sequences were chosen for expression within the context of the full PG9 sequence. The resulting antibodies were then evaluated for their ability to bind to the V1/V2 epitope of HIV-1 gp120 and for their ability to neutralize the virus. Image reproduced with permission from J. R. Willis, et al. (2016) Proc. Natl. Acad. Sci. U.S.A., published online April 4, DOI:10.1073/pnas.1518405113. Copyright 2016, J. R. Willis



A more detailed ROSETTA evaluation of the 1,000 selected sequences produced a Z-score that accounted for multiple aspects of the interaction of each sequence with the V1/V2 antigen. From these data, the investigators selected the top 100 sequences for cluster analysis. The results revealed that the selected sequences could be sorted into nine unique groups containing two or more sequences and five independent “groups” containing a single member.


Although considerable effort had been made to select the energetically best sequences, none of them interacted with the V1/V2 antigen as favorably (based on binding energy and structural stability) as did the HCDR3 regions of PG9 or PG16. This was not unexpected, as it is very rare for a person to express an antibody that has high affinity for an antigen to which he/she has never been exposed. To attempt to eliminate the energy gap, the investigators used ROSETTADESIGN to generate mutations of each sequence in search of changes that would improve antigen binding. This process mimics the natural process of antibody maturation through mutation that occurs during the normal immune response. The resulting sequences, which contained mutations in 30% to 60% of the amino acids, displayed binding energies comparable to those of wild-type PG9 and PG16.


Having identified HCDR3 sequences that, on the basis of structure, would be expected to bind to the HIV-1 envelop protein, the investigators next turned to experiments designed to test their effectiveness (Figure 4D). They selected 84 of the sequences, including the top examples from each of the clusters along with a combination of sequences derived from their mutation analyses. They then created expression vectors encoding the full-length PG9 antibody, but replacing the HCDR3 region of the gene with one of the selected sequences. Transfection of each of these vectors into HEK 293F Freestyle cells led to successful protein expression in 70 out of 84 cases. Evaluation of each of the expressed proteins by ELISA using a mixture of the gp120 proteins from eight HIV-1 strains revealed that 32 of these exhibited no binding activity.


Further testing of 30 of the active expressed proteins identified 16 that bound to at least one out of eight gp120 monomers from different HIV-1 strains at concentrations of less than 100 μg/mL. Two of these active proteins were wild-type donor sequences that had not been altered by mutation. The most potent binding was achieved with an antibody bearing five mutations. The antibody with the broadest specificity bound four out of eight of the tested variants, as compared to PG9, which bound to all eight. The investigators also tested the antibodies for their ability to neutralize the HIV-1 virus, using four different viral strains. They found that two of the antibodies bearing native sequences were able to neutralize one of the strains. The most potent of the antibodies contained three mutations. The one with the broadest activity, neutralizing three out of the four strains, also contained three mutations. The investigators also tested the antibodies against a virus containing a mutation in gp120 that eliminates one of the key glycosylation sites required for interaction with PG9. Only two of the 30 antibodies neutralized this virus, indicating that the vast majority of them were interacting with the virus via a PG9-like binding mode.


The results confirm that people who have never encountered HIV-1 naturally express antibodies that are prototypes for bnAbs against the virus. Furthermore, the binding affinities of these antibodies could be substantially improved through selected mutation, suggesting that the normal immune response of selection followed by maturation could lead to expression of high affinity broadly neutralizing antibodies in vivo. The investigators conclude that vaccines designed to selectively favor induction of these antibodies have the potential to be effective against a broad range of HIV-1 strains.



View PNAS article: Long antibody HCDR3s from HIV-naïve donors presented on a PG9 neutralizing antibody background mediate HIV neutralization










The Vanderbilt Institute of Chemical Biology, 896 Preston Building, Nashville, TN 37232-6304, phone 866.303 VICB (8422), fax 615 936 3884
Vanderbilt University is committed to principles of equal opportunity and affirmative action. Copyright © 2014 by Vanderbilt University Medical Center