Recombinants and the Future of the Epidemic:
A Roundtable Discussion

Richard Marlink

Marlink: Perhaps we can start our discussion by laying some groundwork concerning HIV itself. What is HIV and what are the subtypes of HIV that are now being described so often?

Essex: HIV is a retrovirus. Retroviruses have three basic characteristics that set them in a group apart from other viruses, for example, those that cause influenza or measles. For one thing, retroviruses use dual strands of RNA rather than DNA for their genetic material. They also replicate in what we consider to be a reverse or retro manner, that is, their RNA is translated into DNA rather than the other way around. Finally, they use the enzyme reverse transcriptase to accomplish this DNA from RNA switch.

Early on, when HIV was being isolated and described, researchers recognized that the vast majority of samples of the virus organized themselves into four or five subtypes, based largely on their genetic similarities in a region of the genome known to code for the virus’s outer coating or envelope. Based on this rather arbitrary way of grouping viral strains, most of the existing viruses in the world were organized into subtypes. The nomenclature system of A, B, C, and the like, evolved from that point.

Sankalé: Yes, and we’re seeing now that those subtype definitions were indeed arbitrary. For instance, subtype B and D were considered different when they were first discovered, but lately people have seen that they are not that different, that on further analysis, these subtypes are from the same branch of HIV’s “family tree” and should have been considered as sub-subtypes.

Recently, there has been an effort to clear up the nomenclature of HIV. The Los Alamos HIV Sequence Database group convened a meeting about a year ago during which they acknowledged past imperfections of naming conventions and came up with specific definitions and guidelines for nomenclature that reflects what we now know about subtypes.

Marlink: Why do you think the subtypes of the virus exist?

Renjifo: Subtypes are a result of the selective accumulation of mutations produced during reverse transcription. Like other retroviruses, HIV produces mutations as a natural part of its replication process. However, HIV produces mutations more frequently that most retroviruses. If you drop the same virus in Alaska and Argentina and come back twenty years later you will find that they will have different particularities, depending upon the environmental factors or constraints they have confronted. With HIV, we have considerable mixing of the viruses because of the large movement of infected people worldwide. If you have enough people being infected and exposed to viruses, this process will lead to the generation of intersubtype recombinants.

Max Essex

Essex: The easiest frame of reference we have to understanding how easily viruses can change is the flu. With the flu virus in particular, a sort of recombination or reassortment within its genetic makeup creates new viruses that are recombinants. These occur as a result of mixing of viruses found in the different species, say, plucking one gene from a virus that had been in pigs and mixing that one gene with a virus that had been in humans for the last forty years. So that virus, which is mainly composed of human genes, is a new recombinant virus and easy to categorize with that one porcine gene totally different from the others.

I think the most attractive hypothesis for the origin of HIVs is that they came independently from nonhuman primate species. It’s easier to make this case for some subtypes than others. But I think it is at least a reasonable possibility that other HIVs, like C or E, came in separately from some nonhuman primate contact. The evidence is there that this happened with HIV-2, and it obviously happened with some subtypes outside subtypes of the main group M, such as N and O viruses.

Renjifo: To add to that, another example is the case of simian immunodeficiency viruses, the SIVs, and HIV-2. There are some HIV-2 viruses that group together with distinct SIV viruses suggesting that each of them were introduced in the human population as separate SIVs. The genetic tools we use today for classification allows us to discern where they came from.
Marlink: So we have these different subtypes defined broadly by the degree of diversity of their envelope genes and more recently redefined by criteria from the working group from Los Alamos. But what exactly is a recombinant HIV virus?

Renjifo: Recombination per se is part of the retrovirus replication cycle. This process occurs normally during the initial steps of the reverse transcription when the virus copies its RNA genome to form a DNA molecule. It has been known for years that the virus uses both copies of its genome to make the DNA, the reverse transcriptase enzyme jumps from one RNA copy to the other RNA copy to produce the DNA—that is the recombination process.

Sankalé: The diploid structure is the origin of recombination, especially in HIV, which replicates a large number of copies. Because most recombinations happen between homologous genomes, genomes of viruses of the same subtype, we don’t recognize them as recombinants. What we generally recognize as, and identify as, recombinants are intersubtype, genomic hybrids, those formed between two different subtypes.

Jean-Louis Sankalé

Essex: A recombinant is the same concept as a hybrid, if you think of an animal or plant, where it is a new species that has picked up genetic information from two distinctly unrelated parents. In the case of intersubtype recombinants, where each parent is a different subtype, the recombinant picks up part of its genetic information from one parent, part from the other. They are genetic hybrids.

Renjifo: It can be said that each of us is a “recombinant.” Each parent has its own genetic information, and the children are their recombinants. So recombination is a natural process. For us, the process doesn’t usually produce large genetic changes. For HIV, the recombination process can produce tremendous change in a single replication cycle.

HIV goes through the same process to be transmitted, to infect someone, to replicate. To survive successfully, the virus has to have some mechanism that allows it to survive in its host. This mechanism is its ability to change itself. HIV, in fact, changes, that is, mutates, quite fast compared with other viruses. Every time HIV replicates, 1 part in 10,000 may change. This may appear to be a small number, but a virus can experience drastic changes when a small mutation changes its phenotype, the way the virus behaves. Now with recombination, that process is increased a hundredfold, by changing a whole chunk of the virus at once. I think it is difficult for HIV to make a recombinant that is more infectious, but once it achieves this, it is a complete new entity. A tremendous change has occurred in one single step.

Marlink: How do we study these recombinants?

Irene Koulinska

Koulinska: For HIV, the first time people started talking about recombinants was when they began studying more than one part of the genome and realized that different genomic regions may cluster with different subtypes, meaning they originate from separate parental viruses. This recombination probably has been happening since the beginning of the epidemic. But it’s only recently, when people started classifying the viruses based on larger genomic regions, that we became aware of the presence of recombinants. And we now know even more about recombinants because of new techniques that allow us to amplify the whole genome.

Previously, we were just analyzing parts of the viral envelope gene or parts of the viral gag gene. Now we can study the whole viral genome. This has led to the reclassification of viruses that were previously known. Some of what we thought were “pure” subtypes, we now know are recombinants. We can study recombinants by sequencing their complete genomes and determining what subtypes contribute to specific genes. We can also look for common subtype patterns among recombinants of different origins in an attempt to identify genetic determinants important for transmission. In the future, we can make in vitro constructs in which the subtype of just one part of the genome is changed and then study how the modified virus behaves in culture. These experiments can show what the implications of naturally occurring recombination events are. This is important, because although so far we have described recombinants genetically, we don’t really know if they behave differently.

Marlink: Is there any evidence that recombinant HIV strains are taking over, especially in terms of new epidemics?

Koulinska: We have examples of the subtype E epidemic in Thailand and, more recently, the IBNG (AGIbNG) viruses in West and Central Africa that now account for about 50 percent of infections in the region. Both are recombinant viruses.

Boris Renjifo

Renjifo: For another example, we have the data from our work in Tanzania in our mother-to-infant transmission study. We have been gathering a large amount of data from these studies, and from these data, we may be able to link recombinants to transmission. The study of these recombinants is already giving us clues about the regions of the HIV genome that may be important to transmission.

In Tanzania we have a large number of distinct recombinants. One, a CD recombinant, was shown to be responsible for almost 10 percent of the HIV transmitted from mothers to infants. This recombinant appears to have been generated within this population and to have expanded.

Marlink: Essentially, then, recombination is an opportunity for the virus to improve its chances to survive?

Koulinska: Recombination is responsible for leaps in viral evolution.

Renjifo: I believe that one of the factors that leads to so much recombination is human behavior. Once different HIV subtypes are within human populations, a single person can be exposed to different subtypes, and intersubtype recombination can occur. And since people travel so much, they can be exposed to different viruses. So the virus drives recombination but intersubtype recombination is driven by human behavior.

The recombinants we see from Tanzania have such variety, suggesting that many people are being repeatedly exposed to different viruses.

Marlink: So all this may create viruses that are more adapted to spread, that can be more easily transmitted. Research into this transmissibility has important epidemiologic implications. But what other reasons are there for studying these recombinant viruses, other than simply understanding their contributions to its spread?

Tun-Hou Lee

Lee: I think what we’ve said so far has shown that some recombinants are doing better than others. Now there is a large number of recombinants that we can study to learn the full extent of what recombination can mean for the virus and for the epidemic.

Koulinska: Recombinants can present problems for vaccines and for treatments to halt mother-to-infant transmission.

For instance, if we consider timing of transmission, will recombinants get transmitted earlier, before delivery? Alternative therapies suggested for prevention of vertical transmission in developing countries are administered at birth. Those will most probably not be effective with viruses that transmit earlier. What are our alternatives?

Sankalé: We see a vaccine as essential to preventing transmission. Studying recombinants can tell you what region is found in viruses that are easily transmitted. Through these studies we can gain insight into what regions of the genome are responsible for transmission and design vaccines accordingly.

Photos by Paula Telch.