So as I said previously, there are (thus far) two genes that undergo DNA recombination in lampreys, VLRA and VLR-B. Both VLR's have a very similar structure, much more so than immunoglobulin-domain based BCR's and TCR's. VLR cassettes essentially consist of an N-terminal leucine rich repeat (LRR) and a C-terminal LRR that probably serve as anchor points for gene conversion (a flavour of DNA recombination), and a stalk that anchors the VLR to the cell surface via a molecule called glycophosphatidylinositol (GPI). Pseudogene VLR's upstream and downstream of the VLR cassette copy themselves into the cassette via gene conversion (probably) to generate receptor diversity, and the end products of both VLR-A and VLR-B look fairly similar at least on a protein domain level.
Of course, VLRA and VLRB are not equivalent or else we'd have nothing to talk about. The paper I mentioned in my last post finally came out in a print edition of Nature.[1] Max Cooper's group were able to generate antibodies specific to invariant stalk of either VLR-A or VLR-B, thus allowing them sort out blood cells by flow cytometry. They did an initial sort of the blood cells by "gating" for lymphocytes based on their characteristic size and light scattering; in that respect, lamprey lymphocytes are interestingly similar to mouse lymphocytes.[2] Anyway, lamprey lymphocytes can sort into distinct populations of either VLRA+VLRB- or VLRA-VLRB+ (or double negative) cells. The fact that VLRA and VLRB are on different lymphocyte populations suggests functional differences between the two genes. Both lymphocyte populations could also be sorted fairly purely, which allowed for further characterisations. Using primers that flank the coding sequences of VLRA and VLRB, they were able to demonstrate by genomic PCR that only the VLRA gene and not VLRB was recombined in VLRA lymphocytes. Similarly, only the VLRB gene was recombined in VLRB lymphocytes. Moreover, when they quantified the band intensities of germline and mature PCR products of VLRA in VLRA lymphocytes, they were equal. This indicates that somatic recombination of VLRA only happens on one allele of VLRA, from either Mom or Dad.[3] This monoallelism was previously reported with VLRB according to the authors. I think this is a really neat result because it fits in so well with the clonal selection hypothesis, which requires immune cells that generate variable receptors like TCR and BCR to only carry one receptor specificity. You can make sense of this by a simple thought experiment with a hypothetical immune cell with both a specificity to foreign antigen and to self-antigen. If this immune cell were to become activated by its cognate foreign antigen, it will also react against self. In order to avoid autoimmune disease, immune cells must only have one receptor specificity-- this monoallelic recombination is seen with the immunoglobulin heavy chain and TCRβ genes (although not the Ig light nor TCRα genes).
They also perform some quantitative PCR on a bunch of immune genes and demonstrate that the gene expression profiles of VLRA and VLRB lymphocytes differ from each other. In fact, blocks of expressed immune genes normally associated with T and B lymphocytes respectively were also associated with VLRA and VLRB lymphocytes. This result in and out of itself is interesting, but it's hard to say how it fits in with what's known about lamprey adaptive immunity. The same can be said with their experiment with PHA, which preferentially causes T cells to divide-- the gist is that VLRA lymphocytes preferentially proliferate after administration of PHA. I think the main thing to take away is that, indeed, VLRA and VLRB cells are not functionally the same. Quibbling aside, there are some especially interesting data like the expression of Notch1 on VLRA cells.[4]
The real fun stuff is when they immune challenge lamprey larvae with killed anthrax spores. Both VLRA and VLRB populations proliferate after immune challenge, but curiously only VLRB cells bind to spores and only VLRB is secreted in much the same manner as antibodies. This is analogous to how only B cells bind native unprocessed antigen, and how only B cells secrete their receptor. The obvious speculative idea is that perhaps VLRA cells are like T cells, in that they don't bind native antigen, don't secrete their receptor, and have a similar gene expression profile and response to cell-specific mitogens like PHA. The authors bolster this line of speculation by pointing out the need to control how an anticipatory variable receptor repertoire is deployed, re: helper T cells. On a related note, I think that functional diversification of VLRA and VLRB cells also reflects functional diversification of lymphocytes in jawed vertebrates. For instance, other classes of T cells like CTL and TH17 cells mediate different arms of the immune response to different pathogens in topologically-distinct compartments (i.e. intracellular/extracellular).
In summary, VLRA and VLRB lymphocytes are clearly functionally different. This may reflect certain common challenges to the evolution of adaptive immunity, such as the spectre of autoimmunity and forming responses to the huge variety of possible infectious routes. Lamprey immunity is still getting its legs, but this paper goes a long way to making a reality of a robust comparison between two separate but convergent adaptive immune systems. If even some of the speculative ideas I mentioned above can be confirmed, it would be a truly elegant validation of how we understand the evolution of adaptive immunity.
1. Guo P, Hirano M, Herrin BR, Li J, Yu C, Sadlonova A, Cooper MD. 2009. Dual nature of the adaptive immune system in lampreys. Nature. 459: 796-801.
2. Mayer WE, Uinuk-Ool T, Tichy H, Gartland LA, Klein J, Cooper MD. 2002. Isolation and characterization of lymphocyte-like cells from a lamprey. PNAS. 99(22):14350-14355.
3. My initial reaction and that of my supervisor to the reported intensity values was that, wow, they're really close to 50:50. They did genomic PCR on populations of lymphocytes and not on single cells, which means their conclusion of monoallelism is probably correct but not beyond all doubt since it's conceivable that some lymphocytes could have recombined both alleles and some didn't recombine diddly. After some meditation, however, I think their conclusion is more than just probably correct since they used a pure VLRA lymphocyte population, which would preclude those lymphocytes with two germline VLRA alleles. In addition, the essentialness of only one receptor per a lymphocyte would undoubtedly force values very close to 50:50.
4. I looked at their phylogenetic tree of Notch1 and lamprey Notch1 does seem to cluster with Notch1 of jawed vertebrates. I presume that lampreys have more than just Notch1, but either they haven't found the other Notchs or they just didn't want to include lamprey Notch2 into their Notch tree. A quick search in Entrez didn't give me anything on lamprey Notch. Anyway, Notch1 is an absolutely necessary and sufficient signalling molecule in T cell development, which is interesting because lampreys don't have a thymus to express Notch1 ligand (Delta-like 4?). Then again, my supe reminds me that Notch used many times during development, and that VLRA cells expressing Notch doesn't mean anything without more info. So like I said, interesting.
1 comment:
poop
Post a Comment