AIDSWEEKLY Plus, Monday, 18 November 1996
Daniel J. DeNoon, Senior Editor

After reviewing the history and present status of HIV vaccine research, Duke University AIDS Center researcher Barton F. Haynes concluded that a successful AIDS vaccine will have to account for extensive genetic variability not only in the virus, but also in human populations.

The best way to do this, he argued, is to create HIV vaccines based on human leukocyte antigens (HLA).

Haynes's review-cum-manifesto appeared in the journal The Lancet ("HIV Vaccines: Where We Are and Where We Are Going," Lancet, 1996;348:933-37).

"In view of the complex biology of HIV/host interactions, the most fruitful avenue may be development of multivalent HIV immunogens tailored to HIV isolates in specific geographical locations," he wrote.

Haynes began his review by considering the body of research on the correlates of immunity to HIV - research that admittedly is still in its infancy. He noted that rapid progressors (the 10 percent of people who progress to AIDS in the first few years after infection) generally have low or absent antibodies capable of neutralizing the HIV isolates with which they were infected. They also have a much higher level of HIV viremia than non-progressors (the 10 to 15 percent of individuals who do not develop AIDS after 10 years of HIV infection), however.

"In conjunction with data showing that the fall in plasma viremia in primary HIV infection correlates with the presence of CD8 anti-HIV CTL [cytotoxic lymphocyte] activity, these results suggest that anti-HIV CD8 CTL that kill HIV infected cells and antibodies that broadly neutralize HIV primary isolates might be protective anti-HIV immune responses in uninfected individuals who are subsequently exposed to HIV," Haynes wrote.

He stated that the very definition of "protection" against HIV is controversial. "Protective anti-HIV immune responses may prevent HIV infection completely; may allow only transient infection, leading to clearance of virus; or may merely limit the extent of HIV infection, but in so doing prevent the development of AIDS," he wrote.

If protection against HIV is indeed a function of cellular immune responses to the virus, what is the nature of these responses?

Several research groups have shown that ineffective anti-HIV CD8(+) CTL responses are based on oligoclonal expansion of CTL V(beta) T-cell receptor (TCR) populations, resulting in CTL that target non-immunodominant HIV epitopes. Better CTL responses result from polyclonal expansion of the V(beta) TCR repertoire, leading to CTL that target fewer but immunodominant virus epitopes.

"These observations, together with data suggesting that the inheritance of certain HLA [human leukocyte antigen]-encoded or other host genes may be associated with either rapid progression with either rapid progression or non-progression to AIDS, indicate that host gene expression may determine the quality and/or quality of host anti-HIV immune responses," Haynes proposed.

He suggested that HLA or HLA-associated gene expression could:

• Determine the HIV epitopes to which CTL and/or T-helper cells respond.
• Determine which TCR regions are used in anti-HIV T-cell responses.
• Regulate anti-HIV cytokine responses.

"For induction of CD8 'HIV suppressor' cells, CD8 CTL, and CD4 T helper cells by an HIV immunogen, one probably needs immunogens that induce these anti-HIV responses to a sufficient number of HIV variants so that most variants in a geographical area will be recognized," Haynes wrote.

"The large number of HIV variants available for transmission and the possible immunodominant nature of what may be protective anti-HIV T-cell responses have suggested the need for HLA-based HIV subunit vaccines."

In order to create such vaccines, researchers will have to consider several variables, including the HLA molecules expressed by the vaccine's target population; the T-cell epitopes of the vaccine immunogen; the HLA-restricting elements of the immunogen's T-cell epitopes; and the HIV variants endemic in the area where the vaccine will be used.

To account for all of these variables, Haynes argued, HLA-based vaccines will have to contain a mixture of immunogens that not only represent the relevant regional HIV variants, but also contain enough T-cell epitopes that can bind to the HLA molecules expressed on the target population's antigen-presenting cells (APC).

"Such a mixture of immunogens could range from a mixture of non-HIV vectors expressing HIV proteins to mixtures of HIV recombinant proteins and/or synthetic peptides," he proposed.

Such HLA-based vaccines would have several consequences, according to Haynes:

• Not all members of a vaccine cohort will have the particular HLA antigens needed to present the HIV epitope in the immunogen.
• Not all HIV variants can be represented in the immunogen.
• Vaccines would have to be designed individually for each of the geographic regions - and possibly even for specific intraregion ethnic groups - in which they are deployed.
• The HLA-based vaccines are unlikely to be 100 percent effective.

The good news is that much of the data needed to develop HLA-based vaccines are becoming available (see the related story on the EpiMatrix algorithm in this issue).

Developmental steps include analysis of the HLA types present in the cohort to be vaccinated, analysis of the HLA class I CTL epitopes restricted by HLA types in the cohort, and analysis of the HIV types in the area where the vaccine cohort is located.

"There are considerable limitations and concerns of such an analysis," Haynes noted. It remains unknown, he said, whether the degree of HLA variability seen in the data collected to date accurately reflects the degree of HLA variability in a population; whether all the variants of each HLA type are equally efficient at presentation of HIV peptides to T cells; and whether all of the variants of HIV CTL epitopes will trigger T-cell responses.

Haynes also noted that for the immunogen to elicit neutralizing antibodies as well as T-cell responses, it will have to include B-cell immunogens, as well. This may be important as Letvin and colleagues (unpublished data, according to Haynes' review) have shown that rhesus monkeys with anti-SIV CTL, but not anti-SIV antibodies, are not protected against intravenous SIV challenge.

"We also need immunogens and immunization strategies for inducing genital and gastrointestinal mucosal T- and B-cell immunity to HIV," he argued. "Studies targeted at defining the presence and effectiveness of mucosal antiretroviral T- and B-cell responses following infection or vaccination with retroviral immunogens are critical to successful vaccine development."

The corresponding author for this study is Barton F. Haynes, Duke University AIDS Center, Box 3258, Duke University Medical Center, Durham, North Carolina 27710.

This work was supported by NIH grant AI35351, and the Department of Defense grant DAM D17-94-4467.

961118
AW961105

Published by Charles Henderson, Publisher. Editorial & Publishing Office: P.O. Box 5528, Atlanta, GA 30307-0528 / Telephone: (800) 633-4931; Subscription Office: P.O. Box 830409, Birmingham, AL 35283-0409 / FAX: (205) 995-1588 http://www.newsrx.net

Copyright © 1996 - Charles Henderson, Publisher. All rights Reserved. Permission to reproduce granted to AEGIS by Charles W. Henderson. Authorization to reproduce for personal use granted granted by C. W. Henderson, Publisher, provided that the fee of US$4.50 per copy, per page is paid directly to the Copyright Clearance Center, 27 Congress Street, Salem, Massachusetts 01970, USA.