Edith Dalgaard posted an update 3 months, 1 week ago
Thus, a small portion of UTCsoluble proteins is excluded from the analysis. Remarkably, color selection provided by the 2D-DIGE procedure in PSIA makes this approach especially convenient for the identification of proteins derived from aggregates which differ in compared samples, such as prion proteins or proteins of amyloids whose appearance is related to pathology. Indeed, besides proteins detected to validate this approach, six huntingtin-associated proteins were also identified in this way, for two of which, Pub1 and Def1, the ability to form amyloid fibrils was shown earlier. It should be noted that the use of sarkosyl in PSIA is not well suited for the identification of constitutive amyloids that may be present in both test and control samples. This is due to the fact that sarkosyl treatment does not solubilize some protein complexes of non-amyloid nature. Most of the presumed nonamyloid proteins represent proteasomal components, which indicates resistance of 20S catalytic core of yeast proteasome to 3% sarkosyl rather than their amyloid origin. Earlier, the proteasome was shown to be resistant to some non-ionic detergents, such as Triton X-100. However, other proteins, especially those which were derived from the most stable aggregates resistant to such a strong detergent as SDS, e.g. Gas1, Ape1 and Ape4, may represent proteins of constitutive amyloids. Of course, in all cases, a set of additional experimental assays is PB 203580 152121-47-6 necessary to verify whether a new protein candidate identified by PSIA is indeed amyloidogenic or behaves like a prion. The developed approach also allowed us to extend the work on characterization of the ability of amyloids to induce polymerization of endogenous yeast proteins. Earlier it was demonstrated that polymers of proteins with extended polyQ domains, including mutant human huntingtin, caused appearance of SDSinsoluble aggregates of some chromosomally-encoded Q/N-rich proteins,,. Here, we show that amyloids of mutant human huntingtin induce appearance of SDS-insoluble aggregates of at least six host proteins, Def1, Pub1, Rpn10, Bmh2, Sgt2, and Sup35. Notably, with one exception, all these proteins contain Q- or Q/N-rich tracts of different lengths supporting our suggestion that such proteins can interdependently form amyloids. It may however seem surprising that among the multitude of yeast Q/N-enriched proteins only these formed detectable aggregates in response to appearance of amyloids of mutant huntingtin. It is clear that the capability of a protein to aggregate should depend on its expression level, however with the exception of Sup35, identified proteins are only modestly expressed. Nevertheless, proteins enriched with Q and N may differ from each other by their intrinsic propensity to polymerize and the detected proteins may be among those which are most prone to polymerization. Except for Sup35, whose polymerization substantially contributes to toxicity of mutant huntingtin in yeast, other identified proteins are non-essential and therefore their polymerization-mediated inactivation modulates rather than causes a cytotoxic effect. In support of this, deletion of the DEF1 gene alleviated huntingtin toxicity in yeast, while Def1 was shown to colocalize with huntingtin aggregates. Sgt2 was also detected in huntingtin inclusions and has been proposed to be an amyloid sensor. Rpn10 and Bmh2 have not been previously shown to interact with huntingtin polymers, however Bmh1, which is highly similar to Bmh2, has been detected in huntingtin aggregates and shown to play a role in huntingtin toxicity via aggresome formation. The finding that many functionally unrelated proteins form polymers in response to the appearance of huntingtin amyloids may explain the diversity of lesions typical for Huntington pathology, especially keeping in mind that such polymers can sequester other proteins, which interact with these polymers, as was earlier shown in a yeast model for the Sup45 protein which binds to Sup35 amyloid polymers,,. Another recently published approach to the identification of amyloid proteins, called TAPI is based on inability of SDS-resistant aggregates to migrate into polyacrylamide gel, if boiling of the sample is omitted. However, it seems that the selectivity of TAPI is not sufficiently high and, therefore, proteins which, in all probability, do not form detergent-resistant complexes are also trapped at the top of the gel. In favor of this, non-overlapping sets of non-prion proteins were identified in and cells. Remarkably, TAPI and PSIA resulted in identification of different huntingtin-associated yeast proteins, among which only one protein, Pub1, was identified in both screens. However, most of the proteins revealed by PSIA in cells expressing mutant huntingtin contain Q/N-rich domains and were included in the list of potential prion proteins, while only two, Pub1 and Ynl208w, of those identified by TAPI are of this class. At last, contrary to PSIA, TAPI did not identify Sup35 among proteins whose polymerization is crossseeded by mutant huntingtin, despite it being a major source of its toxicity in yeast.