Ge Zhang , Meiyi Xiang , Wei Tian , Dajun Deng
Peking University Cancer Hospital and Institute
Background: Autophagy is an indispensable, conserved process to degrade intracellular components, such as soluble or aggregated proteins and dysfunctional organelles. There are three types of autophagy, including macroautophagy, microautophagy, and chaperone-mediated autophagy (CMA), in which mechanisms of substrate delivery to lysosomes are different. CMA is a selective process of autophagy through interactions of target proteins with chaperone protein heat shock protein 8 (HSPA8/HSC70). The substrate-HSPA8 complexes in the lysosome surface are further interacted with lysosomal-associated membrane protein 2A (LAMP2A) and transported into lysosomes for degradation. CMA deficiency has been linked with the pathogenesis of certain diseases. TTC28 (Tetratricopeptide Repeat Domain 28) is a conservative gene. According to the public databases in The Cancer Genome Atlas (TCGA), TTC28 was a gene with recurrent somatic mutations and copy-number variation in several kinds of human cancers. Furthermore, the expression level of the TTC28 gene is comprehensively downregulated in many human cancers, according to the GEPIA analysis based on TCGA transcriptome databases. The low level of TTC28 expression was significantly associated with the poor overall survival (OS) of patients with lung or ovarian cancer. These phenomena imply that TTC28 might function as a tumor suppressor. Both zebrafish Ttc28 and macroautophagy have been reported to be a regulator of chromosomal instability (CIN). However, the biological functions and downregulation mechanisms of human TTC28 in cancer cells remain to be studied.
Results: According to the results of the GEPIA analysis, the level of TTC28 mRNA was significantly decreased in many types of human cancers and OS of patients with lung cancer with low TTC28 expression was significantly shorter than that with high TTC28 expression. Thus, we selected lung cancer as the target cancer for further study. Currently only rabbit polyclonal antibody against the C-terminus of TTC28 is available for us. Thus, we compared the polyclonal antibody staining patterns of endogenous and Flag-labeled TTC28 protein in H1299 cells with anti-Flag antibody. We found that both endogenous TTC28 and Flag-labeled TTC28 were mainly localized in the cytoplasm of H1299 cells under the confocal microscopy. In the nucleo-plasmic separation analysis, the polyclonal antibody could bind to cytoplasmic Flag-TTC28 either.Thus, the TTC28 polyclonal antibody is suitable for further studies. We immunoprecipitated endogenous TTC28-binding complexes in H1299 cells using the TTC28 antibody, separated these protein complexes in SDS-PAGE gel, excised and pooled the differential bands into two subgroups for LC–MS/MS identification. The MS/MS analysis results revealed that 222 candidate proteins might interact with TTC28. We observed that endogenous TTC28 in deed interacted with endogenous HSPA8 protein in the CoIP analysis.Endogenous TTC28 and endogenous HSPA8 proteins were colocalized in the cytoplasm of H1299 and H460 cells in the confocal analysis. In contrast to TTC28, the level of HSPA8 expression was upregulated in many cancers in the TCGA project. TTC28 is a highly conserved protein in vertebrates, which contains 28 tetratricopeptide (TPR) motifs and bind to different ligands. There is 96% similarity and 91% identity of TTC28’s TPR domains between zebrafish (Danio Rerio) and human. Therefore, we used a set of established zebrafish ttc28 HA-Ttc28-N, HA-Ttc28-N-1 and HA-Ttc28-ΔN-1 truncation mutants to determine their binding to HSPA8 in H1299 cells. The results of CoIP experiments demonstrated that HSPA8 not only bound to the full-length zebrafish Ttc28, but also bound to all three kinds of TPR-containing Ttc28 mutants. The C-terminal PTIEEVD motif of human HSPA8 is the amino acid sequence to bind many TPR proteins. To confirm whether the PTIEEVD motif is involved in the TTC28-HSPA8 interaction, we also constructed the HSPA8 mutant with PTIEEVD truncation. While the full-length HSPA8-wt strongly bound TTC28, the PTIEEVD truncated mutant very weakly bound TTC28 in H1299 cells. Inhibition of CMA by lysosome inhibitor chloroquine (CQ) or siRNA-knockdown of the expression of LAMP2A, an essential gene for CMA, significantly increased the amount of TTC28 in lung cancer cells. However, CQ treatment did not affect the amount of TTC28 in cells with HSPA8 knockdown, indicating the degradation of TTC28 via CMA. In addition, there are 16 “KFERQ”-like motifs in the amino sequence of TTC28, which may involve in the degradation of CMA substrates. Besides, LAMP2A could be co-immunoprecipitated by TTC28 in H1299 cells. Confocal analysis showed the frequency of micronuclei in TTC28 knockout (TTC28-KO) cells was more than 3-fold that in TTC28 wildtype (TTC28-WT) cells (7.75% vs. 2.33%, p=4.862591e-009), while the rescue of Ttc28 expression decreased the frequency of micronuclei in TTC28-KO cells from 11.91% to 4.80% (p=2.833800e-011), indicating the role of TTC28 in maintaining chromosomal stability.
Conclusion: We demonstrated for the first time that human TTC28 was a HSPA8-binding protein. Tetratricopeptide repeat domains of TTC28 bound to C-terminal motif (PTIEEVD) of HSPA8, and subsequently led to the degradation of TTC28 by CMA pathway. TTC28 could maintain the chromosomal stability, which may contribute to cancer development.