Latency Confusion

I've noticed a lot of confused/conflicting posts over the last week or two regarding the old doozy of HIV latency. The concept isn't helped by the fact that the word has two meanings, and BOTH apply to HIV and AIDS. One article talking about latency may have nothing at all in common with another. The difference is explained during undergraduate virology classes, but since so many of the dissidents are either untrained or deliberately choose to mislead their readers, this makes excellent fodder for anti-HIV "logic".

Clinical Latency
HIV infection has a well known asymptomatic period, in between the seroconversion illness (which doesn't always happen - 50% of the time) and end-stage disease (AIDS, which nearly always does happen - 95% of the time). This period is known as clinical latency, because clinically (meaning, in terms of wellness, signs and symptoms) there isn't much going on. Generally speaking, only lab tests will give a hint that something is amiss, such as depressed CD4 T cells, and of course virus-specific tests like antibodies and viral load.

However, it is now well established that during this time "the virus" is replicating at a significant rate. This rate is reflected in the amount of virus in the peripheral blood, the viral load. Although the virus replicates in lymph nodes, of course some gets out into the blood, and this can be detected using standard tests like RT-PCR (the RT bit refers to the fact that the virus is made of RNA, and so as to be converted to DNA for the PCR to work). If the virus were dormant then there would be no viral load. The only reason why the infection doesn't become run-away is that the immune system is able to partially control it. Contrary to popular dissident belief, it is not due to antibodies. As with most viruses, it is the CD8 T cells than control the infection. In an ideal situation the CD8 response would be able to clear the infection, but HIV has the ability to persist. That ability is tied into the following:

Virological Latency
Now, on the level of the individual infected cells, some will NOT be producing viruses. And by producing viruses, we're talking thousands at a time. Viruses do not replicate by multiplying, like bacteria, or through offspring like sexual animals or plants. Rather they literally turn an infected cell into a virus factory. One virion can result in thousands or millions of other virions, depending on what virus we're talking about. This is required because we know that not all viruses are active, in the case of HIV it might only be 1 in 60,000 peripheral viruses - the others are either genetically defective or physically damaged.

Because HIV has the requirement for cellular transcription factors (NF-kB specifically) if that protein isn't around, it won't do anything. The provirus will be latent. This is virological latency, and is found also in herpes viruses. However, since in a population of many millions of T cells some will be producing NF-kB, say in response to a pathogen they recognise, not all the proviruses will be latent.

The analogy is perhaps a snowstorm on a TV screen. Some of the dots are white, some are black. Overall the image is a grey fuzzy mess! It doesn't mean that each virus is "half active", in fact there is good evidence that for HIV it really is an all-or-nothing response. Either the provirus is shut down, of it's fully active. Black, or White.

So when you hear people talking about "reactivating latent virus" they're referring to turning all the black dots white. When they're talking about "HIV replicating throughout infection" they're referring to the fact that it's a snowstorm rather than a black-out. As a virologist, this makes perfect sense to me, because comparing clinical and virological latency is like comparing chalk and cheese. Sometimes they might coincide (e.g. herpes breakouts) but there is no logical reason why they MUST coincide. The two concepts are not contradictory or confusing, although it's easy to see how they might be to some people!