Skip Navigation

This Article
Right arrow Abstract Freely available
Right arrow FREE Full Text (PDF) Freely available
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrow Search for citing articles in:
ISI Web of Science (82)
Right arrowRequest Permissions
Right arrow Disclaimer
Google Scholar
Right arrow Articles by Korber, B.
Right arrow Articles by Detours, V.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Korber, B.
Right arrow Articles by Detours, V.
Related Collections
Right arrow HIV/AIDS
Right arrow Immunology
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?

British Medical Bulletin 58:19-42 (2001)
© 2001 The British Council

Evolutionary and immunological implications of contemporary HIV-1 variation

Bette Korber*,{dagger}, Brian Gaschen*, Karina Yusim*,{dagger}, Rama Thakallapally*, Can Kesmir{dagger},{ddagger} and Vincent Detours*,{dagger}

*Division of Theoretical Biology and Biophysics, Los Alamos National Laboratory, Los Alamos, New Mexico, USA
{dagger}The Santa Fe Institute, Santa Fe, New Mexico, USA
{ddagger}Center for Biological Sequence Analysis, Department of Biotechnology, Technical University of Denmark, Denmark


   Abstract
Top
 Abstract
Introduction
 The origins of contemporary...
 The extent of contemporary...
 Influenza vaccines: responding...
 Immunological responses to a...
 References
 
Evolutionary modelling studies indicate less than a century has passed since the most recent common ancestor of the HIV-1 pandemic strains and, in that time frame, an extraordinarily diverse viral population has developed. HIV-1 employs a multitude of schemes to generate variants: accumulation of base substitutions, insertions and deletions, addition and loss of glycosylation sites in the envelope protein, and recombination. A comparison between HIV and influenza virus illustrates the extraordinary scale of HIV variation, and underscores the importance of exploring innovative HIV vaccine strategies. Deeper understanding of the implications of variation for both antibody and T-cell responses may help in the effort to rationally design vaccines that stimulate broad cross-reactivity. The impact of HIV-1 variation on host immune response is reviewed in this context.


   Introduction
Top
 Abstract
 Introduction
The origins of contemporary...
 The extent of contemporary...
 Influenza vaccines: responding...
 Immunological responses to a...
 References
 
The natural variability of HIV-1 provides a framework for understanding the complex biology of the virus. As we strive to understand immune evasion and drug resistance, our fundamental knowledge of mutational processes and selection is expanded. Phylogenetic analysis of variable forms of HIV can provide a glimpse into the evolutionary and epidemiological history of the virus. Recent studies have exploited advances both in sequencing technology and in modelling methods. Interpretation of global data sets through the application of improved analysis strategies enables researchers to trace the roots of HIV in terms of primate lentiviral evolution, to estimate divergence rates and the timing of the most recent common ancestor of the epidemic strains, and to model demographic trends in the epidemic. Simultaneously, new experimental methods in immunology1 have rapidly expanded our understanding of the host's response to the virus and the consequences of immune pressure. Scientists are now poised to conduct large scale population studies combining viral sequencing, host genetics, and immunology.

HIV-1's great diversity is seeded by the lack of a proof-reading mechanism in RNA viral polymerase reverse transcriptase, and the consequential high error rate (0.2–2 mutations per genome per cycle)2; but that is not the full story. For example, HTLV-I, like HIV-1, infects CD4+ T-cells and goes though a reverse transcription step, but is far less variable; this may be attributed to differences in the dynamics of the two viruses3. A high replication rate accompanied by rapid viral turn-over4, as well as pressure for change coupled with tolerance of change must be part of the story. Beyond base-substitution, HIV is subject to recombination and relatively large insertions and deletions (indels), which rapidly generate radically divergent forms.

An HIV infection starts out with a homogeneous viral population5,6; over the course of a typical infection, viruses which have mutated to alter more than 10% of their DNA bases in the envelope gene arise6–8. The concept of a viral quasispecies was originally introduced to model within-host viral populations and assumes that the frequency of viral forms in the population will be dictated by their relative fitness; the quasispecies is the realization of the distribution of forms within the sequence space9. However, the term quasispecies has come to be used more loosely in the HIV literature to simply refer to the set of viruses found in an infected individual. Under circumstances of selective pressure, such as therapy10 or immune pressure11–13, the frequency of forms in the viral population can shift. An ‘archive’ of earlier forms of the virus is retained in proviral DNA and these forms on occasion can re-emerge6,7,14, presumably under changing circumstances in the host or through recombination. Combination therapy is effective despite within-host variation because multiple mutations required for resistance to three or more drugs are difficult to achieve unless selected for sequentially15. The correlation between long asymptomatic periods and the breadth of the T-cell immune response16, and the association of HLA-homozygosity with accelerated progression to AIDS17, may be manifestations of this same phenomenon, as simultaneous evasion of multiple immune responses may be difficult for the virus to achieve.

While immune escape clearly can drive positive selection in specific HIV epitopes11,13,18,19, neutral mutations and genetic drift can also contribute to the overall diversity of the virus20,21. One strategy used to try to understand the relative importance of positive selection through immune evasion in HIV evolution is to quantify the force of positive selection within a given gene through analyzing the ratio of synonymous to non-synonymous substitutions22,23. Interpretation of average values is complicated because positive selection can be limited to narrow immunogenic domains within proteins, and average values of these ratios for full genes can be misleading. Many forces concurrently influence the evolutionary pattern in vivo, such as counterbalancing influences of retention of protein structure and function. Hence, there has been a recent interest in estimating these parameters for each site21,24–26. It will be important to establish how mutations which arise in the context of immune evasion within a single host can influence HIV diversity in populations. Such information will help define what is required at a molecular level for stimulating broadly cross-reactive immune responses.

HIV-1 subtypes are distinct lineages of the virus that are defined on the basis of genetic distances and phylogenetic clustering and are labelled A–K (Fig. 1f). It remains unclear whether these genetically defined subtypes will provide a useful way to consider global or regional variation for vaccine design. Neutralizing antibody serotypes do not correlate well with genetically defined subtypes27,28; the antigenic domains that define reactive sites are generally cross-reactive with a subset of viruses, and both cross-reactive and non-reactive viral forms can be found within the same subtype and within different subtypes. Cytotoxic T-cell (CTL) epitopes show a spectrum of cross-reactive responses to epitopes derived from proteins of different subtypes than the stimulation strain29–34.



View larger version (25K):
[in this window]
[in a new window]
  		Fig. 1 A comparison of evolutionary distances of HIV-1 envelope sequences encoding V2-C5 and influenza HA1 domain of the HA gene through phylogenetic analysis. All panels show maximum likelihood trees using a REV model allowing for rate variation at different sites, following the strategy described in Korber et al14. The scale bar is the same in a–f for comparisons. (a) Tree based on 20 HA1 domain sequences of A/Sydney-like viruses circulating in Canada during the first half of the 1997–1998 flu season (Osiowy CK, unpublished observations. Accession numbers: AF087700-AF087708 and AF096306-AF096316). (b) Tree based on 96 HA1 domain sequences of human influenza H3N2 viruses. The tree contains all sequences from the Influenza Sequence Database, Los Alamos National Laboratory131, with an isolation year of 1996. (c) Tree based on 9 V2-C5 sequences from a single asymptomatic individual collected at one time point 73 months post-seroconversion – this was a subtype B infection, and is typical of intrapatient diversity7. (d) Tree based on HIV-1 subtype CRF03_AB V2-C5 sequences from 26 individuals from Kaliningrad, representing a unique situation where a recombinant form of the virus spread explosively through a population of i.v. drug users, and all viruses were very closely related to a single common ancestor65. These samples were collected during 1997–1998, within a year of the introduction of the strain into the population. (e) Tree based on HIV-1 V2-C5 env sequences from a subtype B epidemic, sampled from 23 individuals residing in Amsterdam in 1990–1991117. (f) Tree based on HIV-1 V2-C5 sequences sampled in 1997 from 193 individuals residing in the Democratic Republic of the Congo (DRC), a remarkably diverse set53.

 
In this paper, we will consider the impact of variation in several ways. The first section concerns what can be learned from viral genetic diversity and phylogenetics about the history of the epidemic and rate of evolution. The second section is a summary of the extent of HIV variation as it is known today. The third briefly reviews the influence of variation on design and selection of the annual influenza vaccine (Fig. 2), and contrasts influenza and contemporary HIV-1 variation (Fig. 1). The influenza vaccine is often suggested to be a model system for HIV; however, the variability issues facing those involved with influenza vaccine design, while scientifically challenging and complex, are fundamentally distinct from the issues that need to be faced for an HIV vaccine. Finally, in the last section, the relationship between subtypes and known CTL epitopes is discussed, and the potential impact of variation on epitope processing is modelled.



View larger version (30K):
[in this window]
[in a new window]
  		Fig. 2 Maximum likelihood tree constructed from 68 HA1 domains of the hemagglutinin gene of influenza A subtype H3. The sequences are a subset of the viruses circulating world-wide from 1968 to 2000131. Bold sequence names are the H3N2 strains which have been included sequentially in vaccines for the Northern Hemisphere (summary provided by Catherine Bender).

 

   The origins of contemporary HIV-1 diversity
Top
 Abstract
 Introduction
 The origins of contemporary...
The extent of contemporary...
 Influenza vaccines: responding...
 Immunological responses to a...
 References
 
Primate lentiviruses

Lentiviruses have been found in many African primate species35, and the phylogenetic relationships between the viruses have been used to study the primate origins of HIVs36. HIV-2, the distant cousin of HIV-1 that also causes AIDS in humans, is most closely related to simian immunodeficiency viruses (SIVs) found in sooty mangabeys37,38, while HIV-1 is most closely related to viruses that have been isolated from chimpanzees36,39,40. HIV-1 has three distantly related groups: (i) the main group (M), the group of viruses that cause the global pandemic; (ii) O, a form that is found with a low prevalence in west central Africa and has also been found in Europe41,42; and (iii) N, a very rare form found in Africa43. There are only a small number of viral sequences available from chimpanzees, and they are very diverse36; more information is needed before the precise relationships between chimpanzee viruses and HIV-1 M, N, and O groups can be elucidated.

Modelling the history of the HIV epidemic through phylogenetics

The earliest human sample found to contain HIV-1 was taken in 1959, and found among stored samples from the city now known as Kinshasa in the Democratic Republic of the Congo44. Phylogenetic analysis was used to validate the sample; it behaved like an older sequence would be expected to behave in terms of its position in the tree, branching closer to the root than modern sequences44. So the epidemic had its origins sometime prior to 1959. The lack of hard data prior to 1959 makes this a nebulous period, but important in the human history of HIV. Assuming the evolutionary behaviour of the virus is consistent, one can extrapolate into this pre-1959 period by estimating the rate of diversification using modern isolates with known dates of sampling. This allows a projection based on genetic distances and time, that can be used to estimate the most recent common ancestor of a given viral lineage14,45,46. Current estimates point to an origin of the HIV-1 M group near 1930, with error bars that span roughly a decade or so, depending on the method used14,45. These methods depend upon models of evolution that have inherent assumptions that are imperfectly met by real data47,48. Despite this, the method used in Korber et al14 was validated through accurately estimating the timing of two control time points: the 1959 sample, and the beginning of the epidemic in Thailand. Salemi et al45 used a strategy that excluded the most variable positions, and used a hepatitis sequence set as their control.

These analyses provide information about the rate of evolution, and also the rate of expansion of the epidemic. Dating the most recent ancestral sequence of the pandemic strains does not, however, decisively show when the first M group ancestral virus infected a human. There could have been an ancestral virus in chimpanzee with multiple subsequent cross-over events between species, or a period of time during which the virus was carried in humans prior to the expansion of the M group. If the evolution of the HIV-1 M group took place in humans, then the time of the most recent common ancestor of the M group would be an upper bound on the cross-species transmission event that ultimately gave rise to the epidemic.

Coalescent theory provides a way to extend the boundaries of what can be learned through phylogenetic studies, and model the demographic history of an epidemic through the phylogenetic branching pattern in a tree49–52. But to move beyond the phylogenetic tree into modelling estimations of past epidemic growth, additional assumptions are required. Current coalescent methods are based on the assumptions that: (i) the variance in the reproductive rate of the virus is constant in time (the reproductive rate is related to the number of new infections that arise from a single infection); (ii) the available sequences are sampled randomly; and (iii) the population is panmictic, i.e. the virus is circulating within a population that is not subdividing into isolated lineages and that there is no immigration of new forms into the population. Allowing for these assumptions, again imperfectly met by real data (although clearly some data sets are in better accordance with the assumptions than others), one can model the expansion of the effective population size through the history of the epidemic lineages represented in a tree. Coalescent theory has been applied to the study of HIV at many levels: within a person50, comparisons of subtypes49,51, and to the evolution of the M group virus in the Democratic Republic of the Congo (DRC)52. The DRC set is the most diverse set of HIV-1 M group sequences currently available from any nation, and is considered to be from a region of Africa which harboured the virus for much of its history in humans53. Analysis of this set of central African viruses suggested a period of slow expansion early in the epidemic, with more rapid expansion in recent decades52, consistent with the timing estimates that suggest the virus was in the human population for many decades prior to AIDS being detected and defined14,45.


   The extent of contemporary HIV-1 diversity
Top
 Abstract
 Introduction
 The origins of contemporary...
 The extent of contemporary...
Influenza vaccines: responding...
 Immunological responses to a...
 References
 
HIV subtypes

HIV-1 genetic subtypes have a certain arbitrariness in terms of their definition, and biologists can (and do) reasonably argue for different nomenclature schemes. Nonetheless, a standard nomenclature system is important, as it provides a common language for referring to related lineages and captures a fundamental feature of the virus: the gene and protein sequences within a subtype are more closely related to one another than to the HIV genes and proteins from other subtypes. The associations and groupings of the subtypes can be statistically validated through phylogenetic analysis (Fig. 1f)54–56. For the sake of uniformity in the field and to facilitate communication, a team of experts recently assembled to update the nomenclature system57,58. The genetically distinct subtypes are labelled A–K (with no subtype E or I). Circulating recombinant forms (CRFs) describe viral genomes that contain clearly delineated sections derived from different subtypes, that share a common ancestor, and that are the basis of multiple infections. CRFs are thus epidemic strains, which, like subtypes, are of global importance. There are currently 11 defined CRFs, and more are in the pipeline. Some of the CRFs infect great numbers of people: CRF01(AE) in Asia (originally identified in Thailand and called subtype E59,60), CRF02(AG), with a prototype isolate IbNg, found throughout western and central Africa61–63, a newly defined subtype B and C CRF common in China64, and a subtype A and B recombinant form found in Russia65. Given the exquisite specificity of immune receptors, the genetic relatedness of subtypes and CRFs viruses will certainly have some immunological consequences, but the implications for vaccine design are unclear.

More extensive sampling in regions of sub-Saharan Africa with great viral diversity has resulted in ever greater indications of the potential complexity of HIV diversity. In regions where multiple subtypes are co-circulating with a high prevalence, intersubtype recombination is common61 and recombination between recombinants has also been observed. The large number of novel recombinants found suggests that multiple infections of HIV are not uncommon. Several new isolated examples of strains that are not clearly related to any defined subtype, and that are not obviously recombinants, have also been found. The subtypes themselves are growing ever more diverse with time. The combination of these factors results in a massive pressure on the current nomenclature system, and probably will eventually result in a breakdown of our current HIV classification scheme.

Measures of diversity

There are many ways to compare and contrast the extent of diversity found in HIV-1, and the individual proteins show different levels and patterns of variation. One very basic measure is a tally of sites that are preserved throughout all of the M group, of sites that tolerate insertions and deletions, and of sites that are variable because of nucleotide substitutions. In an alignment of the 132 nearly full-length M group sequences publicly available in the HIV database alignment66, spanning a region from the start of gag to the end of env, there are a total of 7667 bases in HXB2. HXB2 is a sequence derived from the first HIV-1 isolate67 and is a standard reference strain. Roughly 13% (992) of the positions in HXB2 have deletions in one or more of the other 131 HIV-M group sequences, leaving 6675 positions in HXB2 that do not contain any deletions in the alignment. Of these 6675 alignable positions, 2301 bases (30% of the total 7667 bases) are invariant, and 4374 (57%) of the positions have substitutions.

Another way to consider variation is distinctive lengths and glycosylation patterns found in envelope proteins. Among the current database collection (December, 2000) of 566 full-length gp120 protein sequences, protein lengths vary from 484 to 543 amino acids because of the insertions and deletions found in hypervariable regions. The hypervariable regions are typically excluded from phylogenetic analysis, as insertions and deletions do not follow the evolutionary base substitution models used to determine phylogenetic relationships. Thus the branch lengths in the env-based trees are a profound underestimate of the true diversity of the protein, because they capture only one aspect of the variation. Many of the hypervariable sites and insertions and deletions involve glycosylation sequons (the amino acids NX[S or T], sites where N-linked glycosylation occurs). The number of N-linked glycosylation sites in gp120 ranges from 18–33, with a median value of 25. Glycosylation can reduce accessibility to neutralizing antibody epitopes68.

Considering relative HIV variation by protein can also be informative. Using the full-length genome alignment, so a fair comparison between genes can be made, one can model the relative rate of variation between all sites in an alignment using maximum likelihood phylogenies, and then build more accurate phylogenies by including site-specific rate variation69. A gamma distribution was used to assign one of seven evolutionary rates to each site in the nearly full-length HIV-1 M group alignment (the 6675 positions with no insertions or deletions described above). The sites that were evolving most slowly were assigned to category 1, and the fastest were assigned to category 7; in this model, category 7 was estimated to be evolving 20 times faster than category 1. For simplicity, just four HIV-1 genes are shown in Figure 3, representative of the spectrum of evolutionary rates found in HIV-1: gp120, the most rapidly evolving protein; tat; and the gag genes encoding p17, and p24. Gag p24, like pol, is highly conserved. We also used the program SNAP24 to measure the non-synonymous substitution rate for genes included in the full-length genome alignment. In this case, we were not trying to calculate the ratio of non-synonymous to synonymous substitution within genes, but the relative rates of non-synonymous substitution between genes in the 132 full-length sequences. Tat in particular has overlapping coding regions in different frames, and its synonymous substitution rates are low and strongly biased by this. But, as the full-length genome alignment was used to examine the different coding regions, the non-synonymous rates for the different genes can be compared directly. The highest rate of non-synonymous substitutions (dn) was found in Tat, lowest in gp24 (dn values: Tat, 0.170 ± 0.022; gp120, 0.130 ± 0.015; p17, 0.104 ± 0.180; and p24, 0.039 ± 0.010). The high rate of substitution found in Tat is particularly important in the light of recent interest in Tat as a early and potent CTL target13.



View larger version (61K):
[in this window]
[in a new window]
  		Fig. 3 Histogram showing frequency of mutation rate categories assigned to different sites in HIV-1 proteins coding regions, with category one being the slowest evolving sites, category seven the fastest. The rates were calculated based on a full genome maximum likelihood tree, excluding recombinant sequences, using a gamma distribution, and the frequency of assigned rates within gp120, tat, p17 and p24 are shown.

 

   Influenza vaccines: responding to change
Top
 Abstract
 Introduction
 The origins of contemporary...
 The extent of contemporary...
 Influenza vaccines: responding...
Immunological responses to a...
 References
 
Influenza diversity and vaccine design

Influenza A viruses infect many mammalian and avian hosts, including duck, swine, human, geese, quail, whale, seal, and chicken. The viral genome has eight segments and is categorized by the serological and genetic characteristics of its two surface glycoproteins, the neuraminidase gene (NA) and the hemagglutinin gene (HA)70. To date, 15 subtypes of the HA gene and 9 subtypes of the NA gene have been isolated from mammalian and avian hosts; all 15 HA genes occur naturally within the avian population71. Two kinds of viral evolution have been observed in human hosts over the last century, which have been termed antigenic drift and antigenic shift. Antigenic drift is the gradual change in the nucleotide sequence which is due to point mutations accumulating over time. Normally, changes made in the influenza vaccine are made in response to the antigenic drift of the virus. Antigenic shift, on the other hand, is an abrupt change in the serological and genetic properties of the virus due to the introduction of new HA or NA subtypes into previously unexposed populations. Since 1900, four HA genes (H1, H3, H5 and H9) and two NA (N1 and N2) have been isolated from human hosts72–75. Currently, two influenza A subtypes are found co-circulating in the human population, H3N2, a result of an antigenic shift in 1968, and H1N1, a result of a re-introduction of an H1N1 strain in 1977 similar to H1N1 strains circulating around 195076,77.

Because of the annual cost and the threat of particularly dangerous world-wide influenza pandemics, the World Health Organization (WHO) global influenza network was established in 1948. The network comprises over 110 laboratories in 80 countries, and four WHO Collaborative Laboratories78. Both unusual and representative samples of influenza virus are sent by the national laboratories year-round to a Collaborative Laboratory where antigenic and molecular analyses are performed. New viruses are antigenically screened for cross-reactivity with the current vaccine strains by the HA inhibition assay (HI), using antiserum stimulated by the vaccine. Members of the WHO network meet to recommend strains for inclusion in the vaccine for the coming year, currently two influenza A strains (H3N2 and H1N1) and one B strain. Recommendations are made based on the antigenic and molecular analyses of recent isolates, epidemiological data, and post-vaccination serological studies in humans79. In the years spanning 1968–2001, the H3N2 component of the influenza vaccine was changed a total of 17 times (Fig. 2), in one case necessitated by antigenic drift of as little as 3 amino acids substitutions in the 307 amino acid HA1 component of HA (A/Shanghai/16/89 to A/Beijing/353/89).

Influenza and HIV compared

Although influenza, like HIV, is a rapidly evolving virus, a comparison of HIV and influenza A evolution reveals very different patterns. HIV evolution is characterized by a radial spread outward from an ancestral node, while influenza is characterized by bottlenecks and global drift from year to year. Within a single 9 month flu season, little variation was typically found between geographically distinct influenza isolates after the emergence of the epidemic strain. HIV, on the other hand, shows increasing genetic diversity within a population through time. A phylogenetic tree of HA sequences sampled world-wide in 1996 (Fig. 1b) shows much less diversity than a sampling of subtype B HIV-1 envelope sequences from a single city, Amsterdam in 1990–1991 (Fig. 1e). In stark contrast, a sampling of HIV sequences from the Democratic Republic of the Congo in 1997 shows extraordinary diversity, with virtually all HIV subtypes co-circulating in one geographic region (Fig. 1f). The diversity of influenza sequences world-wide in any given year appears to be roughly comparable to the diversity of HIV sequences found within a single infected individual at one time point (Fig. 1c). Thus, while influenza does have a relatively fast rate of mutation when measured over decades (Fig. 2), the vaccine for any given year is targeted towards a relatively homogeneous viral population. Small numbers of changes in the viral amino acid sequence at antigenic sites require a change in the vaccine strain to induce an immunologically appropriate response for currently circulating strains. Thus the diversity which an HIV-1 vaccine must counter through stimulating a broadly reactive immune response is far greater than the diversity countered by the influenza vaccine. If current evolutionary trends continue, the situation will only become worse from an HIV vaccine design standpoint. This is daunting when considered in the context of the small number of amino acid substitutions that result in a loss of antigenic cross-reactivity sufficient to diminish influenza vaccine efficacy (Fig. 2). Thus innovative approaches may be critical for success with HIV-1 vaccines.


   Immunological responses to a rapidly evolving pathogen
Top
 Abstract
 Introduction
 The origins of contemporary...
 The extent of contemporary...
 Influenza vaccines: responding...
 Immunological responses to a...
References
 
Strategies for eliciting cross-reactive B-cell responses

Three broadly cross-reactive neutralizing HIV-1 Env-directed monoclonal antibodies have been found (2F5, IgG1b12, and 2G12)80,81. These bind to conserved parts of Env essential for viral entry, so are promising as vaccine targets. The epitopes tend to be poorly revealed, best exposed during key transitional periods, and the antibodies have high affinity for the native trimer suggesting that they were raised in response to the oligomer on the virion surface rather than dissociated subunits82. This contrasts with the most exposed regions on Env gp120, the V3 and C5; antibodies directed at these sites tend to have weak cross-reactive neutralizing ability83. This could explain the limited neutralizing antibody responses elicited by gp120 vaccines84,85. Novel strategies for designing HIV Env in open conformations, intended to expose conserved critical regions, are being explored, with the hope that such reagents may elicit cross-reactive neutralizing responses81. Examples of such strategies include: (i) disulphide linked gp120–gp41 that mimics a native oligomeric conformation of Env82; (ii) glycosylation-deficient forms of Env that leave neutralizing epitopes exposed and increase their antigenicity68; (iii) Env linked to other immunogenic proteins86; (iv) locking Env into a conformation adopted part way through the fusion process87,88; and (v) deleting variable loops to expose key epitopes89.

CTL responses in HIV vaccines

Several lines of evidence support the notion that CTL contribute to the control of natural infections. HIV CTL have been found in a significant fraction of HIV-1 seronegative high-risk sexual partners of HIV-1 positive individuals90,91. In particular, HIV-specific gamma-interferon secreting cells were detected in cervical mucosa and vaginal washes of exposed seronegative women92,93, suggesting protective immune responses may occur in women chronically exposed to HIV-1, in whom HIV infection cannot be detected virologically or clinically. Broadly cross-reactive CTL were found in highly-exposed seronegative sex-workers in two regions in Africa94, suggesting that broadly cross-reactive CTLs provide protection. Strong CTL responses and T-helper responses are also associated with long-term survival and non-progression in HIV-1 human95,96 and SIV macaque infections97. Some human HLA types, including HLA B5198, are associated with slow disease progression; HLA B51 presents epitopes located in highly conserved regions of HIV-199. Finally, CTL escape mutations have been noted to arise in conjunction with progression to AIDS100.

CTL cross-reactivity and implications for vaccine design

Various degrees of CTL cross-subtype reactivity have been documented, including examples of complete cross-reactivity, diminished cross-subtype responses, as well as cases where the response is completely abrogated30–34,94,101,102. Different target proteins can yield different experimental results in terms of CTL responses to proteins103. CTL experiments often probe the immune response with a single variant, and may favour the detection of responses to conserved epitopes – in other words, a CTL response could be experimentally silent if the probe was sufficiently different from the infecting and stimulating strain in a CTL epitope, and such differences are more likely to occur in highly variable regions. CTL specific for early expressed proteins like Tat may be particularly important for controlling viral replication13,104, but there are few defined HIV-1 Tat specific CTL responses105. Tat responses may be hard to detect106 due to the variability of the Tat protein (see above) and the potential for rapid escape13. It is not clear what the practical implications of viral variation will be for CTL-stimulating vaccines, and how the complexity of host genetics and HLA alleles will factor-in, but people are beginning to attempt to address these issues107–109.

Defining an optimum vaccine to stimulate protective CTL responses against the breadth of circulating viruses in a complex host population is a challenging problem, but DNA vaccines allow great flexibility110,111, and so have great potential for rational design. The biology, however, can be complex. For example, even in the most parallel of infections, identical twins infected with the same batch of factor VIII, profound differences in immune escape and immunodominant CTL responses were observed19. This is a humbling result. If it is characteristic of host-viral interactions, then rational design of a vaccine based on discerning common patterns of response in diverse individuals will be very difficult. One very basic vaccine approach that may be broad enough to over-ride such issues is simply the inclusion of multiple whole HIV-1 protein coding regions to maximize the potential breadth of the response112,113. Another strategy is to concatenate DNA encoding epitopes tailored to maximize potential responses for a population with a particular HLA distribution into a single contiguous DNA-vaccine107. Alternatively, one could use the HIV-1 immunology database to define conserved epitope-rich domains105,108, or to experimentally identify maximally reactive peptide regions109 – such approaches are not exclusive, and can show significant overlap (Fig. 4). Yet another strategy, not directly addressing cross-reactivity but rather potency (which may in turn enhance cross-reactive potential), explores optimization of vaccination with immunomodulatory agents in addition to HIV antigens that both strengthen and tailor the nature of the immune response114–116.



View larger version (47K):
[in this window]
[in a new window]
  		Fig. 4 HXB2 based maps of experimentally-defined CTL epitopes in p17 and Nef from the HIV immunology compendium (http://hiv-web.lanl.gov/). p17 and Nef have very distinctive epitope clustering patterns. The clustering patterns coincide with peptides that were found to serve as a stimulating antigen for CTL from many individuals (although the frequency depended on the HLA types in the population); the maximally reactive peptides are shown in grey109. HLA presenting molecules are shown above delineated epitopes from the database. Alignments show amino acid variation between the HXB2 reference strain and consensus sequences of different subtypes; subtype C and CRF01 were selected as they are important vaccine candidates. Predicted C-terminal proteasome cleavage sites are marked as bold letters, and differ between subtypes. Note that subtype consensus sequences are more closely related to each other than would be modern circulating strains.

 
Consensus sequences are more central than modern sequences

Consensus, or most recent common ancestor sequences, might provide a better reference strains than modern viral sequences for designing peptides for immunological testing, and for vaccine preparations. A consensus is ‘central’, and thus more similar to circulating strains than they are to each other, and can remain relatively stable over time117. To provide a sense of the magnitude of the overall differences within and between subtypes, and of the virtues of consensus sequences, diversity was estimated through tallying the number of amino acid substitutions in pairwise comparisons of subtype A and subtype B protein sequence alignments from the year 2000 HIV Sequence Database66. Subtypes A and B were selected because numerous well-characterized sequences were available (17 subtype A and 95 B Env sequences; 12 A and 37 B Tat sequences; and 7 A and 35 B Gag sequences are included in the following summary). Positions with insertions and deletions were excluded.

Comparisons of actual sequences to their subtype consensus
 Median percentage of amino acid differences are 10% (range 7–14%) in Env, 9% (4–14%) in Tat, and 5% in Gag (1–10%).

Pairwise comparisons of within subtype protein sequences
 Median percentage of amino acid differences are 17% (4–30%) in Env, 15% (3–30%) in Tat, and 8% (2–15%) in Gag. (Note that sequences within a subtype are generally much more distant from one another than they are from the consensus of their subtype.)

Inter-subtype comparisons between A and B subtype sequences
 Median number of amino acid changes is 25% (20–36%) in Env, 35% (27–42%) in Tat, and 17% (15–22%) in Gag.

Variable regions in all HIV proteins are not evenly distributed, but clustered with patterns that probably reflect conservation of functional domains, regions that tolerate change, and regions that are subject to immune pressure. Still, 11% amino acid substitution roughly translates to an average of about one change per epitope. A detailed summary of the number of substitutions found for actual experimentally defined epitopes between and within subtypes can be found in Korber et al118.

Epitope processing and variation

Viral proteins are not recognized by CTL in their native form, but are cleaved into peptides in the cytosol of infected cells119,120, transported to the endoplasmic reticulum120,121, and loaded into a groove on class I HLA molecules120,122. Peptide-loaded class I molecules are then expressed on the surface of infected cells, accessible to antigen-specific receptors of CTLs. The class I peptide binding groove accommodates peptides of length 8–11 amino acids, and peptide binding to a particular class I molecule depends most critically upon amino acids known as anchor residues122.

Virus variability may influence any step in epitope processing123. Immune escape driven by mutations in epitope flanking regions has been documented in vivo124,125, showing that escape through processing abrogation can be of immunological significance. Protein cleavage is usually carried out by the proteasome119 and can be sequence-dependent119,126; thus, viral variation might alter cleavage points, and positions with potential to serve as epitope boundaries may differ among viral strains. There are also alternatives to proteasomal cleavage pathways127, and proteasomal cleavage products can require additional trimming by cytosolic proteases for epitope generation128. After cleavage and trimming, the transporter associated with antigen processing (TAP) translocates peptides into the endoplasmic reticulum for loading onto HLA class I molecules121. TAP affinity for peptides also can contribute to epitope selection129.

Epitope clustering and the potential importance of processing

As noted before, there are regions in proteins in which experimentally defined CTL epitopes cluster (Fig. 4), and the biology underlying this non-uniform distribution is not yet understood. One possible contributing factor is the presence of conserved proteasomal cleavage sites at the boundaries of the epitope rich regions. We explored this possibility through use of an artificial neural network program that predicts C-terminal epitope boundaries (http://www.cbs.dtu.dk/services/NetChop/)130. This program correctly predicts 65% of cleavage sites and 85% of non-cleavage sites, using a training set of human peptides eluted from class I proteins, and a window of 19 residues surrounding the putative cleavage site as a basis for the prediction130. A value is assigned to each position in a protein, such that higher values indicate sites that are most likely to be cleaved, and to be the C-terminus of an epitope. Predicted cleavage sites for HXB2 (Fig. 4) often coincided with, or are within one or two amino acids of, experimentally-defined C-terminal amino acids of epitope or epitope rich regions. Sites with the highest cleavage site prediction scores vary between HIV-1 sequences and subtypes (Fig. 4).

Detailed examples of the influence of variation on cleavage site predictions are shown in Figure 5. The predicted cleavage scores for positions in the B subtype HXB2 p17 protein sequence are highly correlated with predicted cleavage scores for same position in the B subtype consensus sequence (R2 = 0.95). But for the more variable Nef protein, the correlation between cleavage scores breaks down somewhat (R2 = 0.63), even for these two very closely related sequences. For both p17 and Nef, HXB2 and consensus sequences from other subtypes have very different predicted C-terminal cleavage site scores at most positions. This suggests that cleavage may be subtype-specific and processing might be important. It is worth noting that this effect might go undetected in many of the cross-reactivity experiments that have been conducted to date, as often the optimal peptide is defined, the sequence variants within the optimal peptide are synthesized and tested for recognition, but processing is not tested.



View larger version (35K):
[in this window]
[in a new window]
  		Fig. 5 Distinctions between proteasomal cleavage sites within and between subtypes. This scatter diagram shows the correlation between the predicted probability of a site being a proteasome cleavage site based on a program developed by Can Kesmir et al (www.cbs.dtu.dk/services/NetChop/). The HXB2 sequence (x-axis) is compared to the subtype B, C, and AE consensus sequences as well as consensus of subtype consensus sequences (y-axis). The consensus sequences were aligned to HXB2 (see Fig. 4). Each point on the graph corresponds to an amino acid position and has two co-ordinates: cleavage prediction for this position in HXB2 and the cleavage prediction for the appropriate consensus sequence. Points representing the B consensus sequence cleavage prediction as ordinate are denoted by filled diamonds; C consensus, empty triangles; CRF01_AE consensus sequence, empty circles; the main group consensus, grey rectangles. p17a and Nefa: cleavage predictions for B consensus versus B.HXB2 in p17 and Nef, respectively. R2 values for best fit lines are 0.95 for p17 and for 0.63 for Nef. p17b and Nefb: cleavage predictions for other consensus sequences versus HXB2 in p17 and Nef, respectively. R2 values are 0.52 for p17 and 0.55 in Nef for the C consensus; 0.54 in p17 and 0.38 in Nef for the CRF01_AE consensus; and 0.77 for p17 and 0.62 for Nef for the M group consensus.

 
Epitope clustering was apparent from the first compilation of experimentally defined HIV-1 CTL epitopes in the 1995 HIV Immunology Database105, and has become increasingly evident with each new release of the HIV Immunology Database. This, we believe, might have broad practical implications. One could preferentially exclude from vaccines regions that do not encode known epitopes, and focus instead on epitope dense regions. Such constructs may produce safer vaccines than coding regions of full proteins, while preserving essentially full immunological potential for CTL. Shorter gene regions might also allow for inclusion of a cocktail of variants. Furthermore, epitope dense regions may make possible natural processing of epitopes that polyepitope vaccines could not. Thus it may be worth considering incorporating clustering information in vaccine design strategies.


   Footnotes
 
Correspondence to: Dr Bette Korber, Division of Theoretical Biology and Biophysics, Los Alamos National, Laboratory, Los Alamos, New Mexico 98545, USA


   References
Top
 Abstract
 Introduction
 The origins of contemporary...
 The extent of contemporary...
 Influenza vaccines: responding...
 Immunological responses to a...
 References
 

  1. Kaul R, Rowland-Jones S. Methods of detection of HIV-specific CTL and their role in protection against HIV infection. In: Korber BTM, Moore JP, Brander BFHC, Walker BD, Koup R. (eds) HIV Molecular Immunology Database, part IV. Los Alamos, NM: Theoretical Biology and Biophysics, Los Alamos National Laboratory, 1999 (http://hiv-web.lanl.gov)
  2. Drake JW. Rates of spontaneous mutation among RNA viruses. Proc Natl Acad Sci USA 1993; 90: 4171–5[Abstract/Free Full Text]
  3. Wodarz D, Bangham C. Evolutionary dynamics of HTLV-I. Mol Evol 2000; 50: 448–55[ISI][Medline]
  4. Perelson AS, Essunger P, Cao Y et al. Decay characteristics of HIV-1-infected compartments during combination therapy. Nature 1997; 387: 188–91[Medline]
  5. Zhang LQ, MacKenzie P, Cleland A, Holmes EC, Leigh-Brown AJ, Simmonds P. Selection for specific sequences in the external envelope protein of human immunodeficiency virus type 1 upon primary infection. J Virol 1993; 67: 3345–3356[Abstract/Free Full Text]
  6. Wolinsky S, Korber B, Neumann A et al. Adaptive evolution of human immunodeficiency virus type 1 during the natural course of infection. Science 1996; 277: 537–42
  7. Shankarappa R, Margolick J, Gange S et al. Consistent viral evolutionary changes associated with the progression of human immunodeficiency virus type 1 infection. J Virol 1999; 73: 10489–502[Abstract/Free Full Text]
  8. Lukashov VV, Kuiken CL, Goudsmit J. Intrahost human immunodeficiency virus type 1 evolution is related to length of the immunocompetent period. J Virol 1995; 69: 6911–6[Abstract]
  9. Eigen M. On the nature of virus quasispecies. Trends Microbiol 1996; 4: 216–8[ISI][Medline]
  10. Condra JH. Resistance to HIV protease inhibitors. Haemophilia 1998; 4: 610–5[ISI][Medline]
  11. Price DA, Goulder PJ, Klenerman P et al. Positive selection of HIV-1 cytotoxic T lymphocyte escape variants during primary infection. Proc Natl Acad Sci USA 1997; 94: 1890–5[Abstract/Free Full Text]
  12. Poignard P, Sabbe R, Picchio GR et al. Neutralizing antibodies have limited effects on the control of established HIV-1 infection in vivo. Immunity 1999; 10: 431–8[ISI][Medline]
  13. Allen TM, O'Connor DH, Jing P et al. Tat-specific cytotoxic T lymphocytes select for SIV escape variants during resolution of primary viraemia. Nature 2000; 407: 386–90[Medline]
  14. Korber B, Muldoon M, Theiler J et al. Timing the ancestor of the HIV-1 pandemic strains. Science 2000; 288: 1789–96[Abstract/Free Full Text]
  15. Colgrove R. A combinatorial ledge: reverse transcriptase fidelity, total body viral burden, and the implications of multiple-drug HIV therapy for the evolution of antiviral resistance. Antiviral Res 1999; 41: 45–56[ISI][Medline]
  16. Lubaki NM, Ray SC, Dhruva B, Quinn TC, Siliciano RF, Bollinger RC. Characterization of a polyclonal cytolytic T lymphocyte response to human immunodeficiency virus in persons without clinical progression. J Infect Dis 1997; 6: 1360–7
  17. Tang J, Costello C, Keet IP et al. HLA class I homozygosity accelerates disease progression in human. AIDS Res Hum Retroviruses 1999; 15: 317–24[ISI][Medline]
  18. Borrow P, Lewicki H, Wei X et al. Anti-viral pressure exerted by HIV-1-specific cytotoxic T lymphocytes (CTLs) during primary infection demonstrated by rapid selection of CTL escape virus. Nat Med 1997; 3: 205–11[ISI][Medline]
  19. Goulder P, Sewell A, Lalloo D et al. Patterns of immunodominance in HIV-1-specific cytotoxic T lymphocyte responses in two human histocompatibility leukocyte antigens (HLA)-identical siblings with HLA-A*0201 are influenced by epitope mutation. J Exp Med 1997; 8: 1423–33
  20. Sala M, Wain-Hobson S. Are RNA viruses adapting or merely changing? Mol Evol 2000; 51: 12–20[ISI][Medline]
  21. Yang Z, Nielsen R, Goldman N, Pedersen AM. Codon-substitution models for heterogeneous selection pressure at amino acid sites. Genetics 2000; 155: 431–49[Abstract/Free Full Text]
  22. Gojobori T, Yamaguchi Y, Ikeo K, Mizokami M. Evolution of pathogenic viruses with special reference to the rates of synonymous and non-synonymous substitutions. Jpn J Genet 2000; 69: 481–8
  23. Simmonds P, Balfe P, Ludlam CA, Holmes EC, Leigh-Brown AJ. Analysis of sequence diversity in hypervariable regions of the external glycoprotein of human immunodeficiency virus type 1. J Virol 1990; 64: 5840–50[Abstract/Free Full Text]
  24. Korber B. HIV signature and sequence variation analysis. In: Rodrigo AG, Learn GH. (eds) Computational Analysis of HIV Molecular Sequences. Dordrecht, The Netherlands: Kluwer, 2000
  25. Yamaguchi-Kabata Y, Gojobori T. Re-evaluation of amino acid variability of the human immunodeficiency virus type 1 gp120 envelope glycoprotein and prediction of new discontinuous epitopes. J Virol 2000; 74: 4335–50[Abstract/Free Full Text]
  26. Zanotto PM, Kallas EG, de Souza RF, Holmes EC. Genealogical evidence for positive selection in the nef gene of HIV-1. Genetics 1999; 153: 1077–89[Abstract/Free Full Text]
  27. Moore JP, Cao Y, Leu J, Qin L, Korber B, Ho DD. Inter- and intra-subtype neutralization of human immunodeficiency virus type 1: the genetic subtypes do not correspond to neutralization serotypes but partially correspond to gp120 antigenic serotypes. J Virol 1995; 70: 427–444[Abstract]
  28. Nyambi PN, Nkengasong J, Lewi P et al. Multivariate analysis of human immunodeficiency virus type 1 neutralization data. J Virol 1996; 70: 6235–43[Abstract]
  29. Menendez-Arias L, Mas A, Domingo E. Cytotoxic T-lymphocyte responses to HIV-1 reverse transcriptase. Viral Immunol 1998; 11: 167–81[ISI][Medline]
  30. Dorrell L, Dong T, Ogg GS et al. Distinct recognition of non-clade B human immunodeficiency virus type 1 epitopes by cytotoxic T lymphocytes generated from donors infected in Africa. J Virol 1999; 73: 1708–14[Abstract/Free Full Text]
  31. Cao H, Kanki P, Sankale JL et al. CTL cross-reactivity among different HIV-1 clades: implications for vaccine development. J Virol 1997; 71: 8615–23[Abstract]
  32. Durali D, Morvan J, Letourneur F et al. Cross-reactions between the cytotoxic T-lymphocyte responses of human immunodeficiency virus-infected African and European patients. J Virol 1998; 72: 3547–53[Abstract/Free Full Text]
  33. Rowland-Jones SL, Dong T, Fowke KR et al. Cytotoxic T cell responses to multiple conserved HIV epitopes in HIV-resistant prostitutes in Nairobi. J Clin Invest 1988; 102: 1758–65[ISI][Medline]
  34. Buseyne F, Chaix ML, Fleury B et al. Cross-clade-specific cytotoxic T lymphocytes in HIV-1-infected children. Virology 1998; 250: 316–24[ISI][Medline]
  35. Beer BE, Bailes E, Sharp PM, Hirsch VM. Diversity and evolution of primate lentiviruses. In: Kuiken CL, B. Foley B, B. Hahn B et al. (eds) AIDS and Human Retroviruses HIV Sequence Database, part III. Los Alamos, NM: Theoretical Biology and Biophysics, Los Alamos National Laboratory, 1999 (http://hiv-web.lanl.gov)
  36. Hahn B, Shaw G, De Cock K, Sharp P. AIDS as a zoonosis: scientific and public health implications. Science 2000; 287: 607–14[Abstract/Free Full Text]
  37. Chen Z, Telfier P, Gettie A et al. Genetic characterization of new West African simian immunodeficiency virus SIVsm: geographic clustering of household-derived SIV strains with human immunodeficiency virus type 2 subtypes and genetically diverse viruses from a single feral sooty mangabey troop. J Virol 1996; 70: 3617–-27[Abstract]
  38. Chen A, Luckay A, Sodora DL et al. Human immunodeficiency virus type 2 (HIV-2) seroprevalence and characterization of a distinct HIV-2 genetic subtype from the natural range of simian immunodeficiency virus-infected sooty mangabeys. J Virol 1997; 71: 3953–60[Abstract]
  39. Peeters M, Honore C, Huet T et al. Isolation and partial characterization of an HIV-related virus occurring naturally in chimpanzees in Gabon. AIDS 1989; 10: 625–30
  40. Gao F, Bailes E, Robertson DL et al. Origin of HIV-1 in the chimpanzee Pan troglodytes. Nature 1999; 397: 436–41[Medline]
  41. Mauclere P, Loussert-Ajaka I, Damond F et al. Serological and virological characterization of HIV-1 group O infection in Cameroon. AIDS 1997; 11: 445–53[ISI][Medline]
  42. Peeters M, Gueye A, M'Boup S et al. Geographical distribution of HIV-1 group O viruses in Africa. AIDS 1997; 11: 493–8[ISI][Medline]
  43. Simon F, Mauclere P, Roques P et al. Identification of a new human immunodeficiency virus type 1 distinct from group M and group O. Nat Med 1998; 4: 1032–7[ISI][Medline]
  44. Zhu T, Korber BT, Nahmias AJ, Hooper E, Sharp PM, Ho DD. An African HIV-1 sequence from 1959 and implications for the origin of the epidemic. Nature 1998; 391: 594–6[Medline]
  45. Salemi M, Strimmer K, Hall WW et al. Dating the common ancestor of SIVcpz and HIV-1 group M and the origin of HIV-1 subtypes by using a new method to uncover clock-like molecular evolution. FASEB J Express 2000; On line, Dec 8: Article 10.1096/fj.00–0449fje
  46. Sharp PM, Bailes E, Gao F, Beer BE, Hirsch VM, Hahn BH. Origins and evolution of AIDS viruses: estimating the time-scale. Biochem Soc Trans 2000; 28: 275–82[ISI][Medline]
  47. Hillis DM, Mable BK, Moritz C. Application of molecular systematics: the state of the field and a look to the future. In: Hillis DM, Moritz C, Mable BK. (eds) Molecular Systematics, 2nd edn. Sunderland, MA: Sinauer, 1996; 515–43
  48. Korber B, Theiler J, Wolinsky S. Limitations of a molecular clock applied to considerations of the origins of HIV-1. Science 1998; 280: 1868–71[Abstract/Free Full Text]
  49. Grassly NC, Harvey PH, Holmes EC. Population dynamics of HIV-1 inferred from gene sequences. Genetics 1999; 151: 427–38[Abstract/Free Full Text]
  50. Rodrigo AG, Shpaer EG, Delwart EL et al. Coalescent estimates of HIV-1 generation time in vivo. Proc Natl Acad Sci USA 1999; 96: 2187–91[Abstract/Free Full Text]
  51. Pybus OG, Rambaut A, Harvey PH. An integrated framework for the inference of viral population history from reconstructed genealogies. Genetics 2000; 155: 1429–37[Abstract/Free Full Text]
  52. Yusim K, Peeters M, Pybus OG, Bhattacharya T, Delaporte E, Mulanga C et al. Using HIV-1 sequences to infer historical features of the AIDS epidemic and HIV evolution. Philos Trans R Soc Lond B Biol Sci 2001; 356: 855–65[ISI][Medline]
  53. Vidal N, Peeters M, Mulanga-Kabeya C, Nzilambi N, Robertson D, Ilunga W et al. Unprecedented degree of human immunodeficiency virus type 1 (HIV-1) group M genetic diversity in the Democratic Republic of Congo suggests that the HIV-1 pandemic originated in Central Africa. J Virol 2000; 74: 10498–507[Abstract/Free Full Text]
  54. Louwagie J, McCutchan FE, Peeters M et al. Phylogenetic analysis of gag genes from 70 international HIV-1 isolates provides evidence for multiple genotypes. AIDS 1993; 7: 769–80[ISI][Medline]
  55. Gao F, Robertson DL, Carruthers CD et al. A comprehensive panel of near-full-length clones and reference sequences for non-subtype B isolates of human immunodeficiency virus type 1. J Virol 1998; 72: 5680–98[Abstract/Free Full Text]
  56. The European Commission and the Joint United Nations Programme on HIV/AIDS. HIV-1 subtypes: implications for epidemiology, pathogenicity, vaccines and diagnostics. AIDS 1997; 11: 17–36[ISI][Medline]
  57. Robertson D, Anderson J, Bradac J et al. HIV-1 nomenclature proposal. In: Kuiken CL, Foley B, Hahn B et al. (eds) AIDS and Human Retroviruses HIV Sequence Database, part III. Los Alamos, NM: Theoretical Biology and Biophysics, Los Alamos National Laboratory, 1999 (http://hiv-web.lanl.gov)
  58. Robertson D, Anderson J, Bradac J et al. HIV-1 nomenclature proposal. Science 2000; 288: 55–6
  59. Gao F, Robertson DL, Morrison SG et al. The heterosexual human immunodeficiency virus type 1 epidemic in Thailand is caused by an intersubtype (A/E) recombinant of African origin. J Virol 1996; 70: 7013–29[Abstract/Free Full Text]
  60. Carr JK, Salminen MO, Koch C et al. Full-length sequence and mosaic structure of a human immunodeficiency virus type 1 isolate from Thailand. J Virol 1996; 70: 5935–43[Abstract]
  61. McCutchan FE, Carr JK, Bajani M et al. Subtype G and multiple forms of A/G intersubtype recombinant human immunodeficiency virus type 1 in Nigeria. Virology 1999; 254: 226–34[ISI][Medline]
  62. Cornelissen M, van Den Burg R, Zorgdrager F, Goudsmit J. Spread of distinct human immunodeficiency virus type 1 AG recombinant lineages in Africa. J Gen Virol 2000; 81: 515–23[Abstract/Free Full Text]
  63. Carr JK, Laukkanen T, Salminen MO et al. Characterization of subtype A HIV-1 from Africa by full genome sequencing. AIDS 1999; 13: 1819–26[ISI][Medline]
  64. Su L, Graf M, Zhang Y et al. Characterization of a virtually full-length human immunodeficiency virus type 1 genome of a prevalent intersubtype (C/B') recombinant strain in China. J Virol 2000; 74: 11367–76[Abstract/Free Full Text]
  65. Liitsola K, Tashkinova I, Laukkanen T et al. HIV-1 genetic subtype A/B recombinant strain causing an explosive epidemic in injecting drug users in Kaliningrad. AIDS 1998; 12: 1907–19[ISI][Medline]
  66. Kuiken C. HIV-1 sequence alignments. In: Kuiken CL, Foley B, Hahn B et al. (eds) AIDS and Human Retroviruses HIV Sequence Database, part I. Los Alamos, NM: Theoretical Biology and Biophysics, Los Alamos National Laboratory, 1999 (http://hiv-web.lanl.gov)
  67. Hahn BH, Shaw GM, Arya SK, Popovic M, Gallo RC, Wong-Staal F. Molecular cloning and characterization of the HTLV-III virus associated with AIDS. Nature 1984; 312: 166–9[Medline]
  68. Reitter J, Means R, Desrosiers R. A role for carbohydrates in immune evasion in AIDS. Nat Med 1998; 4: 679–84[ISI][Medline]
  69. Yang Z. Maximum likelihood phylogenetic estimation from DNA sequences with variable rates over sites: approximate methods. J Mol Evol 1994; 39: 306–14[ISI][Medline]
  70. Lamb RA. Genes and proteins of the influenza viruses. In: Krug RM. (ed) The Influenza Viruses. New York: Plenum, 1989; 1–87
  71. Alexander DJ. A review of avian influenza in different bird species. Vet Microbiol 2000; 74: 3–13[ISI][Medline]
  72. Smith FI, Palese P. Variation in influenza virus genes. In: Krug RM. (ed) The Influenza Viruses. New York: Plenum, 1989; 319–59
  73. Subbarao K, Klimov A, Katz J et al. Characterization of an avian influenza A (H5N1) virus isolated from a child with a fatal respiratory illness [see comments]. Science 1998; 279: 393–6[Abstract/Free Full Text]
  74. Guo Y, Li J, Cheng X et al. Discovery of men infected by avian influenza A (H9N2) virus. Chin J Exp Cline Virol 1999; 13: 105–8
  75. Peiris M, Yuen KY, Leung CW et al. Human infection with influenza H9N2 [letter]. Lancet 1999; 354: 916–7[ISI][Medline]
  76. Nakajima K, Desselberger U, Palese P. Recent human influenza A (H1N1) viruses are closely related genetically to strains isolated in 1950. Nature 1978; 274: 334–9[Medline]
  77. Kendal AP, Noble GR, Skehel JJ, Dowdle WR. Antigenic similarity of influenza A (H1N1) viruses from epidemics in 1977–1978 to ‘Scandinavian’ strains isolated in epidemics of 1950–1951. Virology 1978; 89: 632–6[ISI][Medline]
  78. Stamboulian D, Bonvehi PE, Nacinovich FM, Cox N. Influenza. Infect Dis Cline North Am 2000; 14: 141–66
  79. Hay A, Gust I, Hampson A et al. Update: influenza activity – United States and worldwide, 1999–2000 season, and composition of the 2000–01 influenza vaccine. MMWR 2000; 78: 1913–8
  80. D'Souza MP, Livnat D, Bradac JA, Bridges SH. Evaluation of monoclonal antibodies to human immunodeficiency virus type 1 primary isolates by neutralization assays: performance criteria for selecting candidate antibodies for clinical trials. Infect Dis 1997; 175: 1056–62
  81. Parren PW, Moore JP, Burton DR, Sattentau QJ. The neutralizing antibody response to HIV-1: viral evasion and escape from humoral immunity. AIDS 1999; 13 (Suppl A): S137–62
  82. Binley J, Sanders R, Clas B et al. A recombinant human immunodeficiency virus type I envelope glycoprotein complex stabilized by an intermolecular disulfide bond between the gp120 and gp41 subunits is an antigenic mimic of the trimeric virion-associated structure. J Virol 2000; 74: 627–43[Abstract/Free Full Text]
  83. Nyambi P, Mbah H, Burda S, Gorny M, Nadas A, Zolla-Pazner S. Conserved and exposed epitopes on intact, native, primary human immunodeficiency virus type I virions of group M. J Virol 2000; 74: 7096–107[Abstract/Free Full Text]
  84. Connor R, Korber BTM, Cao Y et al. Immunological and virological analyses of persons infected by human immunodeficiency virus type 1 while participating in trials of recombinant gp120 subunit vaccines. J Virol 1998; 72: 1552–76[Abstract/Free Full Text]
  85. Locher CP, Grant RM, Collisson EA et al. Antibody and cellular immune responses in breakthrough infection subjects after HIV type 1 glycoprotein 120 vaccination. AIDS Res Hum Retroviruses 1999; 15: 1685–9[ISI][Medline]
  86. Collado M, Rodriguez D, Rodriguez J, Vazquez I, Gonzalo R, Esteban M. Chimeras between the human immunodeficiency virus (HIV-1) Env and vaccinia virus immunogenic proteins p14 and p39 generate in mice broadly reactive antibodies and specific activation of CD8+ T cell responses to Env. Vaccine 2000; 18: 3123–33[ISI][Medline]
  87. LaCasse R, Follis K, Trahey M, Scarborough JD, Littman DR, Nunberg J. Fusion-competent vaccines: broad neutralization of primary isolates of HIV. Science 1999; 283: 357–62[Abstract/Free Full Text]
  88. Salzwedel K, Smith ED, Dey B, Berger EA. Sequential CD4-coreceptor interactions in human immunodeficiency virus type 1 Env function: soluble CD4 activates Env for coreceptor-dependent fusion and reveals blocking activities of antibodies against cryptic conserved epitopes on gp120. J Virol 2000; 74: 326–33