The Axis of Viral

The enemy of my enemy is my friend.

We might accept that viruses and bacteria at best instantiate the coincidental nature of such an alliance. The success of one bug might pave the way for another. But we’d be hard pressed to imagine that pathogens would whittle the syllogism to a sharper point and actively pursue our sorry asses in tandem or even in triplicate.

Karposi’s sarcoma-associated herpesvirus (KSHV or human herpesvirus 8 ) and the human immunodeficiency virus (HIV) appear to engage in just such a collaboration.

Kaposi’s sarcoma (KS) and AIDS, the diseases the two cause, have long been associated in the scientific literature. Indeed, KS lesions were important markers identifying AIDS as a novel syndrome in the first place. KS proved one of a multitude of ‘opportunistic infections’ that arise only when the immune response collapses, as it does during an HIV infection.

KS dynamics, however, may be more than opportunistic. A preponderance of circumstantial evidence tying KSHV and HIV together and several newly discovered mechanisms by which the two pathogens partake in reciprocal activation suggest the pathogens have a more functionally integrated relationship. That is, KSHV and HIV appear marked by a mutualism that on further analysis may explain several little understood aspects of the two viruses’ origins and pathogenesis.

The ecological and epidemiological circumstances KSHV and HIV share appear on their face coincidental. KSHV and HIV-1 share common xenospecific origins–chimpanzees. Both appear to have evolutionarily radiated out of the same general geographic region. Both are chronic infections of immune cells. HIV is capable of infecting B cells KSHV typically infects. Conversely, KSHV is capable of infecting the dendritic cells and macrophages HIV infects. KSHV and HIV appear to share overlapping modes of infection, including sexual transmission.

Indeed, coinfection may cause a convergence of modes of infection. Marcelin and colleagues (2004) show KSHV load increases in circulating blood cells when a patient is infected with HIV and expresses active KS, suggesting coinfection broadens KSHV’s cell range. Hmmmmm. In that vein, Henke-Gendo and Schultz (2004) reviewed accumulating evidence for a KSHV infection spread by reused needles.

But it’s in the molecular work where the relationship starts to get downright conspiratorial.

Huang and colleagues (2001) review evidence that KSHV and HIV regulate each other’s expression beyond the diffuse effects of immune suppression. HIV-1-induced cytokines can induce lysis, the virus-producing stage of KSHV’s life cycle. HIV-1’s Tat protein can activate epithelial growth factor KDR in enodthelial cells KSHV infects, helping bring on KS tumor growth.

Huang et al. offer additional evidence that KSHV and HIV undergo reciprocal transcriptional activation. KSHV’s ORF45 KIE2 protein can activate HIV’s regulatory Long Terminal Repeat (LTR). In turn, HIV’s Vpr and Tat proteins appear to activate KSHV intracellular expression, including of the major caspid protein.

Sun and colleagues (2003, 2005) meanwhile show KSHV K13 protein vFLIP, involved in blocking programmed cell death in KS lesions, also regulates HIV expression by way of nuclear transcription NF-KB. K13 and HIV-1 Tat synergistically activate HIV LTR. Guo and coworkers (2004) showed KSHV chemokine receptor vGPCR (ORF74) also synergistically activates NF-KB and NF-AT with HIV-1 Tat, generating KS tumors.

There appear other pathways for KSHV-HIV crosstalk:

• Several researchers (here and here) have reported that KSHV-encoded cytokine interleukin-6 (vIL-6), along with inducing vascular endothelial growth factor, increases HIV replication. In turn, HIV-infected cells produce elevated huIL-6, contributing to KSHV activation.
• Fc gamma receptors for immunoglobulin receptor IgG on the surfaces of effector immune cells mediate phagocytosis, antibody-dependent cell-mediated cytotoxicity, and activation of cytokine pathways. Lehrnbecher et al. (2000) showed polymorphic forms of Fc-gamma-R to be associated with differing KS outcomes in HIV-infected individuals. The FcgRIIIA receptor effects antibody-dependent cell-mediated cytotoxicity and the lysis of infected cells. Lerhrnbecher and colleagues suggest FF is protective because it reduces the kinds of inflammatory responses that induce KS pathogenesis.

We might be able to relate these activation mechanisms back to our seemingly coincidental transmission dynamics.

Gandhi and colleagues (2004) showed CD4 cell count in KSHV-HIV coinfected individuals to be the strongest predictor for KSHV salivary shedding. Greater CD4 cell counts were associated with greater shedding. The results suggest KSHV loads should be higher early in the HIV-1 infection, during primary infection when CD4 cells counts are still relatively high.

In other words, the results imply KSHV piggybacks on HIV’s acute and epidemiologically-predominant first stage, within the first two months of infection. That stage, when few patients know they are infected, largely drives HIV’s epidemiological spread.

I previously described HIV infection as a dual life history. In an outbreak’s epidemic phase, when many new susceptible hosts are available, HIV acts much like a precocious semelparous organism, using its first burst of viral reproduction to rapidly infect the large pool of available susceptibles. In an endemic phase, when available susceptibles are comparatively rare, HIV uses its iteroparous nature, depending on multiple exposures over a long asymptomatic stage to wait out a new cohort of potential hosts.

It stands to reason that in a mutualism with HIV, KSHV’s life history may shift in turn, in such a way as to co-express an acute phase, with epidemiological dividends.

In perhaps a related phenomenon, HIV may better ‘prep’ KSHV for transmission by hobbling the immune response. Jacobson and coworkers (2000) showed males infected with KSHV after HIV infection were more likely to develop KS. Even if KS lesions themselves prove little involved in generating subsequent KSHV infection, although Krishnan and colleagues (2004) suggest that very possibility, a full exploration of its life cycle–from latent through lytic infection–may better permit KSHV infections greater pathogenic and epidemiological flexibility.

Such mutual amplification may explain in part the geographic distribution of Africa’s HIV strains.

Cohen (2000) describes several hypotheses for subtype C’s recent geographic expansion:

• Subtype C quasispecies rarely make the switch from coreceptors CCR5 and CXCR4. Those patients infected with subtype C virus will likely have more copies of the CCR5 virus that are typically involved in establishing subsequent infections.
• Individuals infected with multiple venereal pathogens produce more proinflammatory cytokine tumor necrosis factor (TNF)-alpha, boosting HIV replication, particularly for subtype C.
• A third possibility Cohen misses involves the socioeconomic conditions of many of the countries where subtype C is prevalent. Central Africa has been devastated by war and structural adjustment programs and undergone resulting changes in behavior regimens, including behaviors that spread HIV. Such ‘top-down’ increases in the number of susceptibles may select for more infectious variants that can burn through the supply of hosts at little cost to their epidemiological persistence.
• The KSHV-HIV mutualism offers a fourth, mutually inclusive, possibility. The HIV-1 LTR activation by KSHV protein K13 described above appears to be subtype-specific and dependent on the number of NF-KB sites. So KSHV prevalence–in some African populations as high as 87%–may help select for HIV subtypes. HIV subtype C, with three NF-KB sites, may use KSHV as an epidemiological amplifier in Central and East Africa.

A follow-up phylogeographic study could test whether the natural distributions of KSHV K13 and HIV LTR, including the clinical outcomes of both infections, are co-determined.

Such microscopic mutualisms may be more common than first thought. Indeed, HIV and the tuberculosis bacterium have long been shown to amplify each other’s transmission. Steve Lawn (2004) reviews other relationships HIV shares in Africa, including with malaria, schistosomiasis, and a number of STDs. Lawn catalogs some specifics. “Cytokine-mediated activation of the HIV-1 LTR,” for one,

is the main mechanism by which bacterial coinfections enhance proviral transcription. However, certain DNA virus coinfections such as human T lymphotropic virus type 1 (HTLV-1), herpes simplex virus type 2 (HSV-2) and cytomegalovirus (CMV) may also more directly enhance proviral transcription by encoding proteins that themselves transactivate the HIV-1 LTR via other specific receptors. By this mechanism, these chronic viral infections may exert a more direct and prolonged cofactor effect on HIV-1 replication.

The scope of possible combinations across circulating pathogens boggles the mind.

When a good friend of mine was a child he thought people who spoke other languages hadn’t as rich of an emotional life as those who spoke his native tongue. It isn’t just a kid’s miscue. We routinely confuse our failures in comprehension with the state of a system’s complexity. We think what we don’t see (or hear) can’t be there.

The bias appears to extend to our view of the lives of pathogens. Infectious agents must routinely learn new tricks with old tools. Their kit may include a shock of shocks: disparate viruses and bacteria, mixed and matched by historical happenstance humans have helped impose, may farm each other as assiduously as we do corn and cattle.