The most important finding of this study was the identification of PATMK as a member of the TMK family that participates in erythrophagocytosis and is uniquely required for intestinal but not hepatic infection. In addition this manuscript introduced a novel technique of gene knockdown using shRNA in E. histolytica that could find general usefulness in a parasite for which there is not the ability to replace genes.
The participation of PATMK in ingestion of the dead human red cell was demonstrated not only by shRNA but by identification in the early phagosome proteome, co-localization with erythrocytes, and inhibition with anti-peptide antibodies and by expression of a carboxy-truncated mutant. Enzymatic activity of the kinase domain was not demonstrated when PATMK was expressed and purified from E. coli or immunoprecipitated from E. histolytica, Further studies will therefore be required to understand if PATMK is acting as a receptor for apoptotic red cells or as a regulator of ingestion via its kinase domain.
A comprehensive analysis of the phagosome proteome was not the goal of this work, as several groups have accomplished this for E. histolytica
. However it is interesting that each group has identified new molecules of interest. The first analysis of this type in E. histolytica
, published by Okada et al.
, revealed many of the GTP binding proteins that are important for maturation of phagosomes [24
]. However, this screen did not reveal any of the endoplasmic reticulum resident proteins identified in metazoan systems [17
]. More recent work by Marion et al
. identified ER resident proteins in E. histolytica
phagosomes such as calreticulin as well as significant involvement by myosin IB, actin and actin accessory proteins [25
]. Later this group identified a small number of surface proteins which were suggested as possible receptors [27
]. The screen reported here was focused on early time points following ingestion with the goal of identifying new surface molecules with roles in this process. In addition proteins were identified such as amoebapore A and B, which were known to be involved in phagocytosis but had remained absent in other screens. The most obvious conclusion from this collection of proteomics is that not one effort has taken its screen to saturation. Incompleteness of these screens may also explain the differences that have been published between clinical isolates of E. histolytica
Manipulation of PATMK by binding the ectodomain with antibody (75% reduction of ingestion of calcium treated erythrocytes in M199S with 55mM D-galactose), reducing the protein levels with interfering RNA (65% reduction in M199S with 55mM D-galactose), or by expressing a truncated protein (81%), all produced a similar reduction of erythrophagocytosis in vitro. The mechanism of this interference is not entirely clear. Given that anti-PATMK serum co-localizes with erythrocytes in contact with ameba, we assume that antibody against this protein blocks uptake of erythrocytes by interfering with receptor function, but it is very possible that this could be blocking interaction with a different molecule by steric hinderance or by interfering with signaling (as opposed to receptor) functions of PATMK. Clearly all of the experimental approaches supported a role for this protein in host cell ingestion by E. histolytica.
Although no kinase activity has been demonstrated, we hypothesize that PATMK is in fact a receptor kinase. The Mer family of tyrosine kinases provide a paradigm for receptor kinase involvement in phagocytosis [30
]. Mer tyrosine kinase interacts with the bridging molecule GAS6 to recognize PS on apoptotic cells. The identification of the TMK family was one of, if not the most exciting finding from the genome project, because of its lack of precedent in a unicellular eukaryote. Implication of PATMK in erythrophagocytosis is an important step towards understanding the role of this family of kinases in parasite biology.
Virulence of E. histolytica
has been long associated with the parasite's ability to ingest host cells. This work suggests that this may be more important in intestinal disease than in liver abscesses. Ameba expressing PATMKΔ932
had reduced ability to infect the intestine but were not impaired in causing liver abscess when directly injected into the liver. A varying requirement for virulence factors in different environments for the amebae has been seen before.
For example cysteine protease 2 over-expression was found to reduce in vitro
monolayer destruction but had no effect on liver abscess formation [33
], and amoebapore A silencing led to inability to cause liver abscesses although these parasites still caused tissue damage in a colonic xenograft model of amebiasis [34
]. The data presented here concerning a retention of virulence in the liver abscess model as well as the analysis showing that lectin expression on the surface has not been altered illustrate that the ameba are otherwise competent to attach to the host and cause disease. This may indicate that expression of PATMKΔ932
does not simply produce an impaired ameba, but an ameba which fails to colonize the intestine possibly because of failure to clear dead and dying host cells. It also suggests that the virulence program required for the parasite's success in the gut differ from that of the liver.
The rationale behind phagocytosis of host cells by this parasite is still a mystery, but erythrophagocytosis is a hallmark of E. histolytica infection. One could envision that this behavior provides an advantage during infection by clearing dying and dead cells and thereby reducing the infiltration of inflammatory cells and release of toxic cellular content. PATMK may have a pivotal role in allowing this parasite to persist longer in the host. Interfering with this pathway may produce a more robust immune response to E. histolytica and clearance of infection.