Characterization of the mimivirus cyclophilin

An interesting article showed up today on the Articles in Press page at the Journal of Molecular Biology. Vu Thai, the most talented young crystallographer I have never wanted to punch in the face, and Elan Eisenmesser, a man of sharp wits and sharper dress, along with several collaborators, have performed a really extensive characterization of the cyclophilin encoded in the mimivirus genome. This is an intriguing study because the role of cyclophilins in viral life cycles is still very poorly understood, even though they are known to aid the infectivity of HIV and SARS.

Typically, these viruses do not encode their own cyclophilins -- this would be a pretty costly thing for such a small and limited organism -- but rather hijack them from the local cellular machinery. What still hasn't been cleared up is exactly why that happens. The catalytic activity of cyclophilins is to convert proline residues from a cis conformation, in which the C alpha atoms of sequential residues lie on the same side of the C'-N bond, to a trans conformation in which they are on the opposite side. It's been demonstrated that human cyclophilin A (hCypA) catalyzes this conversion on the HIV capsid. However, it's not known whether this is the key role and reason for its incorporation. hCypA is also known to mediate some cell-binding events for HIV, and it's hypothesized that this is the only reason for inclusion, and the cis-trans isomerization is incidental.

The mimivirus cyclophilin (mimicyp) has the potential to answer some of these questions. One of the keys here is that it has been carried along in the viral genome. Because viruses are such efficient parasites operating under strong selective pressure, it stands to reason that anything the virus considers important enough to carry along with it is critical to its survival. Granted, mimivirus is a bit of an odd duck, one of the largest viruses known. Additionally, its primary targets are amoebae, though mimivirus has also been known to cause pneumonia by direct attack of human cells. And indeed, Thai et al. find that mimicyp most closely resembles an amoeboid cyclophilin.

The results start to get progressively weirder from there. Firstly, mimicyp has a vanishingly low affinity for cyclosporin, the molecule that gave cyclophilins their name. Moreover, it doesn't appear to have catalytic activity towards a standard proline isomerase substrate at all. The sequence of the protein is missing key residues that typically interact with isomerization substrates and contribute to catalysis, and its putative active site is uncharacteristic of active cyclophilins. Mimicyp crystallizes with an unusual trimeric arrangement, and Elan has demonstrated that it also forms multimers in solution that can be dissociated by adding arginine. Nonetheless, mimicyp localizes to the outer capsid of the mimivirus and is thought to be critical to infectivity, though Thai et al. were unable to demonstrate this for infection of Acanthamoebae polyphaga.

So what we have here is a cyclophilin that is catalytically dead, forms strange multimer arrays, and can't yet be shown to be important for infectivity, that nonetheless is always carried along by mature virions, and is so critical to the success of the virus that it is carried along in the genome and expressed despite significant selective pressure in favor of exclusion, mutation, or silenced expression. Granted some of these pressures are alleviated by the sheer size of the genome and virus, but others (particularly expression) are not. Thai et al. suggest a number of reasons why mimicyp might be important, including charge neutralization of the capsid, mediation of entry and viral disassembly, or capsid-masking similar to that performed by hCypA for HIV. And, of course, the possibility exists that mimicyp has some other, as-yet-unidentified catalytic activity that is essential for mimivirus infectivity, or is only critical for infecting certain amoebae or cells.

One might also surmise that the ability of mimivirus to infect vastly different hosts to some degree relies on its carrying along this particular cyclophilin, which is probably quite dissimilar to any protein abundantly found in human cells. This only brings us back to the question of why mimicyp is so important in the first place.

Ultimately this paper does not reveal the precise reason why cyclophilins are used by viruses, and indeed it will probably do more to spur debate than end it. Nonetheless, Thai et al.'s research can be taken as adding weight to the proposition that the catalytic activity of cyclophilins on virions is incidental, or at least that its importance is an idiosyncratic feature of particular viruses. Mimicyp may also be our first introduction to a new class of cyclophilins, catalytically inactive, cyclosporin-free (cyclophobins? cyclomehs?), and multimeric. What it is that these cyclophilins do and why they are important will be an intriguing inquiry to follow.