Part I: Inter- and Intra-Clade Variation


1.1) Background:

The history of southern African HIV infection dates back to 1982 when the first South African AIDS cases were reported among the homosexual population. Until 1987, HIV-1 diagnosis was almost completely limited to the male portion of the population, but by 1992 the reported cases in women roughly equaled those in men. The rapid evolution of HIV-1 is well documented with clade C being the most prevalent within the region of southern Africa (van Harmelen 1997).

The biological and pathogenic relationships between clades are unknown. (Jeeninga 2000). The current definition of a clade therefore rests solely upon the significant sequence variation that makes it unique. This sequence variation develops from the mutations caused by the low fidelity of the RT enzyme and genetic recombination (Stine 1999). Currently, HIV-1 displays an ever increasing number and combination of subtypes arising from this variation (Jeeninga 2000), especially within the region of southern Africa (Novitsky 1999). Worldwide, 10% of HIV patient isolates are intersubtype combinations (Stine 1999). Large shifts in genetic diversity have already been seen in Thailand (Wasi 1995) and could eventually evolve in southern Africa. Such divergence may affect the transmissibility and pathogenicity of HIV-1. Variations may also have profound impacts upon the formulation of an appropriate vaccine.

It has therefore been imperative that research labs continually monitor the genetic diversity that develops within given regions (van Harmelen 1999-A). This will allow further elucidation into the nature of variation and the resulting biological or clinical differences. Such research will further optimize the development and application of potential vaccines. Ethnic diversity within southern Africa, as well as the abundant presence of emigrants and refugees, contributes to the likelihood that several subtypes may be independently introduced to different regions.


1.2) Methods of Subtype Elucidation:

In 1993, a method was published in the November 19^th issue of Science that provides a rapid way for detecting and estimating genetic diversity among HIV strains. The method is a heteroduplex mobility assay (HMA). It works on the observation that DNA heteroduplexes that arise between related sequences will travel differently in polyacrylamide gels depending on their degree of divergence from one another (Delwart 1993). Peptide enzyme-linked immunosorbent assays (ELISAs) have also been previously used for subtype elucidation (Cheingsong-Popov 1993).

However, these methods are limited to analysis of the exterior envelope (env) region of HIV-1, which compromises less than 10% of the genome (Delwart 1993 and van Harmelen 1999-B). A novel application of a technique called restriction fragment length polymorphism (RFLP), has been used to extend analysis to the group antigen (gag) region. RFLP, already used to differentiate between other viral strains, simply consists in identification of restriction endonucleases (REs) that uniquely act upon a certain clade followed by digestion of unknown samples with these REs. The technique proved useful in screening an HIV epidemic of limited diversity. It would therefore be most productive to use RFLP in conjunction with precise methods such as HMA and peptide ELISAs (van Harmelen 1999). This will allow subtype identification to go beyond the limitations of the env region.


1.3) Monitoring Genetic Divergence:

Researchers monitoring HIV-1 diversity within southern Africa recently phylogenetically compared HIV-1 sequences to published V3 loop env sequences. The comparisons showed a subtype C-restricted epidemic. The sequences, which differed by an average of only 14%, did not cluster according to the country of origin and no specific South African subcluster was identified (van Harmelen 1999-A). However, continued monitoring is considered necessary to keep track of any significant changes in local clade diversity, whether inter- or intra-clade.

One investigation also indicates that subtypes segregate according to their mode of transmission (Figure Four), which implies two epidemics of independent origins within southern Africa (van Harmelen 1997). This shows that novel genotypes can be introduced into a broad population through development in social patterns. Additionally, this genetic divergence is always potentially significant to vaccine efficacy and therefore should be monitored as social patterns change within a given region.

Part II: Cross-clade Reactivity

The profound differences between subtypes that are outlined above continue to develop. This increases the need for a general vaccine that will be effective against all clades of HIV-1 despite any sequence variations that may regionally evolve.
A successful general vaccine should be able to elicit both humoral (virus-neutralization antibodies) and cellular (virus-specific cytotoxic T lymphocyte) immune responses (Chang 1999). Much research into the humoral immune response focuses on the crown of the third variable loop (V3) region of env. The V3 loop has a high degree of sequence variability, but one sequence at the apex of the V3 loop has been shown to elicit cross-reactive neutralizing antibodies (Giara 1997 and Chang 1999). While the scope of this paper will only address the cellular immune response, the humoral response is considered equally important to the development of an adequate vaccine against all clades.


2.1) Class I MHC-restricted CD8+ Cytotoxic T Lymphocytes:

Mounting evidence continually indicates that cytotoxic T lymphocyte (CTL) responses will be an important component to an effective HIV vaccine (Riddell 1992 and Cao 1997). T-lymphocytes have membrane receptors that recognize antigen when the antigen is associated with cell-membrane proteins called the major histocompatibility complex (MHC). Monoclonal antibodies that react with a certain molecule are grouped as a cluster of differentiation (CD). CTLs have the CD8 glycoprotein on their surface and function to eliminate any cells from the body that display antigen and are associated with a class I MHC molecule (Kuby 1997).

The priming of CTL activity depends on two important factors: 1) viral sequence variation across subtypes and 2) the MHC class I profile of the infected host or vaccine recipient (Betts 1997). The exact mechanisms of CTL immune protection against HIV-1 or in AIDS are unknown, but research suggests that CTLs have strong defensive and offensive roles. While HIV specific CTLs cannot fully eradicate the virus from infected hosts, they do control disease progression (Riddell 1992 and Buseyne 1998). CTLs are present in the initial control of a viral infection and CTL levels decline as plasma viremia levels increase (Riddell 1992). In addition, powerful CTL responses have been detected in individuals who lack the symptoms of infection and demonstrate low viral loads. HIV-1 specific CTLs have likewise been found in heavily exposed but uninfected individuals. (Cao 1997).


2.2) Cross-clade CTL Reactivity Research:

T lymphocytes are able to recognize specific peptide sequences (epitopes) on the antigen when the antigen is associated with an MHC molecule (Kuby 1997). A single amino acid change within an epitope can lead to CTL recognition failure (Cao 1997). The original anticipation for CTL research was that a vaccine based on a single clade would be unlikely to provide cross-clade protection because clades differ in sequence by about 30% for env and 14% for gag (Dorrell 1999).

However, research conducted thus far repeatedly indicates that cross-clade CTL responses are possible (Cao 1997, Betts 1997, Ferrari 1997, Buseyne 1998, and Dorrell 1999). One study, for example, demonstrated cross-clade recognition in 88% of the Pol, 83% of the Nef (negative regulation factor), 67% of the Gag, and 55% of the Env isolates with positive CTL responses (Buseyne 1998). This cross-reactivity can be stable for several years (Buseyne 1998) and is not limited to one particular HIV protein (Betts 1997).

Recent research has additionally shown that cross-clade reactivity has been able to neutralize X4-, R5- and dualtropic isolates of clade B and C (Verrier 2000). The Verrier study is the first to illustrate that a vaccine constructed from clade B strains can have neutralizing activity upon R5-tropic and clade C isolates. Similar results have been demonstrated in studies of the relative numbers of immunotypes to genotypes. HIV analysis indicates that there are fewer immunotypes than genotypes. In addition, isolate clustering did not depend on the genotype of the strain, the coreceptor used (tropism), or geographic origin (Nyambi 2000). The data from these studies supports the belief that an appropriately formulated vaccine could have broad effects throughout geographic regions with HIV genetic diversity.

Yet, certain research has indicated that while cross-clade recognition can occur, the recognition of specific peptide epitopes can be strongly clade-specific (Dorrell 1999). Studies have previously shown that the recognition of epitopes is not uniform (Ferrari 1997). Definition of novel epitopes through epitope mapping has thus become essential to the formulation of a vaccine. Prior to the Dorrell study, no CTL-binding epitopes had been defined outside of clade B (Dorrell 1999).

Research seems to indicate that cross-clade CTL reactivity occurs despite highly specific epitope recognition because the measured reactivity is within the most highly conserved regions (Cao 1997, Buseyne 1998, and Dorrell 1999). One analysis, for example, found that the proteins most likely to cause cross clade responses were Gag, p24, and RT (Cao 1997), which are in some of the most highly conserved regions of HIV antigen. One study further supports this evidence in that they found that there exist several epitopes shared by all immunotypes (Nyambi 2000).

One drawback of the research done to date is that CTL activity has only been assayed with peripheral blood mononuclear cells (PBMC). Internal lymph node and mucosal CTLs have been virtually ignored in cross-clade CTL activity studies. In addition, it remains uncertain exactly which antigens are essential to induce protection from infection (Girard 2000).


2.3) "Vaccinomics":

Using the information that epitope recognition is clade-specific but cross-clade responses are nonetheless possible, scientists are attempting to develop a cross-clade vaccine by exploiting conserved regions. An important application of the cross-clade CTL reactivity data obtained so far is within the field of vaccinomics. Vaccinomics utilizes genomic and bioinformatic information for vaccine development. Vaccinomic companies are using software to predict epitopes within sequences of amino acids. The software depends on recurring motifs that specify the grooves and positions in which amino acids bind. Once conserved epitopes are located, they can be loaded into the grooves of human leukocyte antigen molecules to provide a cross-clade vaccine against HIV. It is expected that the vaccine will be most useful on individuals who have already been infected with HIV. Animal models already indicate that a vaccination following drug therapy increases the immune response and halts the individual's dependence on further drug intake (Hollon 2000).


2.4) Vaccine Optimizations:

The induction of a T cell response is currently being pursued using ‘prime-boost' combination strategies. This strategy first consists of ‘priming' an immune response through the introduction of a live virus vector into a test subject. The immune response is then ‘boosted' by introducing recombinant viral envelope proteins into the test subject. These proteins stimulate the levels of antibody neutralization necessary for a complete and successful immune response. The ‘prime-boost' strategy still needs optimization. The potency of the immune responses that have been elicited to date is uncertain. Yet, this vaccine strategy is the only current immunization protocol that stimulates cross-clade CTL activity in human volunteers. (Girard 1999).

This fact is important because a great deal of the research into vaccine and pathogenesis conducted so far has utilized the simian immunodeficiency virus (SIV) model in Asian macaques and the chimeric simian/human immunodeficiency virus (SHIV) model (Chen 2000). The quantification of CTL precursors has also been achieved with the SIV model (Kakimoto 1999). The limitation of the SIV and SHIV models are that little is known of how they relate to HIV infection in humans (Girard 1999). Data showing cross-clade CTL activity in humans is therefore extremely significant and needed through the application of optimized ‘prime-boost' methods.

Procedural assays have been developed to detect cross-clade CD8⁺ CTL activity with a good degree of sensitivity (Gorse 2000). Nonetheless, employing different combinations of viral vectors in the prime-boost strategy can optimize the sensitivity of these assay procedures and the activity of vaccine candidates themselves (Girard 2000 and Gorse 2000). In addition to varying viral vectors it is important to compare vaccines through the different antigens expressed in the test subject (Girard 2000). Finally, vaccine optimization cannot be complete without a thorough understanding of the mechanisms of CTL activity and its relations to chemokine receptors, the humoral aspect of immunity, and the MHC class I profile of an individual.

Conclusion

As of the 1999 publication of the Girard et al. review on HIV-1 vaccine development, only one vaccine is in phase 3 clinical trials. This vaccine, AIDSVAX^TM, is a gp120-based vaccine. A prime-boost combination of a recombinant canarypoxvirus and a gp120 subunit vaccine is in phase 2 trials. Furthermore, increasing emphasis has been placed on developing and testing potential vaccines in developing countries. As a result, trial sites have been picked in Kenya, Uganda, South Africa, and Thailand to begin phase 1 trials (Girard 2000).

Although genetic diversity of HIV-1 continues to increase throughout the world, extremely promising data has been gathered from HIV studies. Research continues into elucidating the mechanisms of viral factors and their inhibition. Yet, it is clear that any drug therapy developed from the resulting data will never be an option for the majority of the worldwide HIV-infected population. On the other hand, vaccines formulated towards conserved antigen epitope sequences and optimized through prime-boost regimens can be expected to stimulate cross-clade responses against HIV. Applications to produce such a vaccine in a inexpensive manner are underway and could help to alleviate the tremendous and damaging pressure of AIDS on societies in developing nations.


NOTE: For a more complete and specific discussion of HIV-1 vaccine development please consult the excellent review by Girard et al.. The English title of the review is "New prospects for the development of a vaccine against human immunodeficiency virus type 1: An overview." For further general HIV or immunological information, please consult the texts by Stine and Kuby respectively. An excellent current online resource for HIV information is the University of South Carolina School of Medicine - Microbiology and Immunology - HIV lecture notes by Dr. Richard Hunt. (See the References section).

Acknowledgements

Above all I wish to thank Dr. Lorraine Mulfinger, Dr. Jill Keeney, and Dr. Lawanna Zimmerman for helpful discussions and editorial input. Gratitude goes to the Juniata College Beeghly Library for obtaining several references for me. I also wish to thank Dr. Suzanne Bray in Lille, France and Suzanne Lovett in Gabarone, Botswana for feeding an interest in this subject. Finally, thanks to classmates Mike Acker, Aaron Martin, and Parisha Shah for helpful discussions.