The TRiC/CCT chaperone is implicated in Alzheimer's disease based on patient GWAS and an RNAi screen in Aβ-expressing Caenorhabditis elegans

Abstract

The human Aβ peptide causes progressive paralysis when expressed in the muscles of the nematode worm, C. elegans. We have exploited this model of Aβ toxicity by carrying out an RNAi screen to identify genes whose reduced expression modifies the severity of this locomotor phenotype. Our initial finding was that none of the human orthologues of these worm genes is identical with the genome-wide significant GWAS genes reported to date (the "white zone"); moreover there was no identity between worm screen hits and the longer list of GWAS genes which included those with borderline levels of significance (the "grey zone"). This indicates that Aβ toxicity should not be considered as equivalent to sporadic AD. To increase the sensitivity of our analysis, we then considered the physical interactors (+1 interactome) of the products of the genes in both the worm and the white+grey zone lists. When we consider these worm and GWAS gene lists we find that 4 of the 60 worm genes have a +1 interactome overlap that is larger than expected by chance. Two of these genes form a chaperonin complex, the third is closely associated with this complex and the fourth gene codes for actin, the major substrate of the same chaperonin.

Figure 1. GWAS genes +1 interactome.

The +1 interaction network for the 52 GWAS genes with intermediate and high significance (p<10⁵). For the 52 genes in the GWAS white+grey zones, there were 703 interactions in the +1 interactome.

Figure 2. Worm genes +1 interactome.

The +1 interaction network for the human orthologues of the 60 worm genes that were highlighted in our worm RNAi screen. For the 60 worm-screen genes with human orthologues, there were 3191 interactions in the +1 interactome.

Figure 3. Distribution of rankings of worm genes.

The x-axis represents the log of the gene ranking bin boundaries, arranged in decreasing gene-product connectedness from left to right. The y-axis represents the log of the fraction of genes in each bin (where 100% is 60 genes). The dashed line shows the linear regression for the worm screen results. The results of the screens are shown as black diamonds and results of random simulations are shown as empty triangles.

Figure 4. Distribution of rankings of worm models of neurodegenerative diseases.

The x-axis represents the log of the gene ranking bin boundaries, arranged in decreasing gene-product connectedness from left to right. The y-axis represents the log of the fraction of genes in each bin (where 100% is 135, 23, 22 & 52 genes for panels A, B, C & D respectively). The dashed line shows the linear regression for the worm screen results. The results of the screens are shown as black diamonds and results of random simulations are shown as empty triangles. (A) polyglutamine screen; (B) α-synuclein screen; (C) tau screen; (D) GWAS candidate white+grey zone genes for AD.

Figure 5. The interaction network of 4 significant worm gene products.

Four human orthologues of worm screen gene products (MOB4, TCP1, CCT8 and ACTB) have +1 interactomes that overlap more than expected with the +1 interactome of the GWAS white+grey gene products. Two of these are components of the abundant cytoplasmic TRiC/CCT chaperone, the third (MOB4) interacts closely and fourth (ACTB), along with tubulin, is an important substrate. TRiC/CCT interacts with STK24 and PP2a to form a complex that regulates the MAPK pathway. This network of interactions may have a bearing on tau phosphorylation and neuronal regeneration.