Honours Project:

Computational structure and function prediction of dementia risk protein TMEM106B and its interacting protein ATP6AP1

Advisor: Daniel Rigden

Abstract

Frontotemporal Lobar Degeneration with TDP-43 (FTLD-TDP) is a lethal neurodegenerative disease and TMEM106B has been identified as an important disease modifier. The roles of TMEM106B in lysosomal dysfunction has been revealed, and it has been found to physically interact with the vacuolar-ATPase accessory protein 1 (ATP6AP1) in lysosome. Precise molecular mechanisms of TMEM106B remain unclear due to the absence of protein structures. This project aims to use structural bioinformatics methods to predict the structure and molecular mechanisms of TMEM106B and its interacting protein ATP6AP1. Homology modeling servers and ab initio servers were used to build models for TMEM106B and ATP6AP1. The following pocket analyses identified four pockets on the TMEM106B lumenal domain and the largest pocket may serve as a molecular binding sits. Conservation analyses identified highly conserved residues surrounding the largest pocket, suggesting the functional roles of the pocket. Electrostatics showed the largest pocket of TMEM106B is negatively charged. It was hypothesized that the largest pocket of TMEM106B lumenal domain is a molecular binding sites that preferentially interact with the positive charged molecules. Four possible TMEM106B-ATP6AP1 lumenal docking states were predicted but not determined. The top β-sheet of the ATP6AP1 C-terminal lumenal domain was hypothesized to interacts with c’’ subunit of V-ATPase V0. Fold recognition of ab initio models of the TMEM106B C-terminal cytosolic domain imply the cytosolic domain may harbor a Zinc ion binding site. Finally, the intact model of TMEM106B was assembled.

Introduction

Frontotemporal lobar degeneration (FTLD) is a lethal neurodegenerative disease, characterized by progressive degeneration of the frontal and anterior temporal lobes (Mohandas and Rajmohan, 2009). The predominant subtype of FTLD is FTLD-TDP, characterized by TAR DNAbinding protein 43 (TDP-43) positive and tau negative inclusions, but the pathologic mechanism has not been elucidated (Zhou et al., 2017). One of the most frequent genetic cause of FTLD-TDP is the heterozygous loss-of-function mutations in Progranulin gene (GRN), which result in haploinsufficiency of progranulin (PGRN) (Zhou et al., 2017). A genome-wide association study identified that an uncharacterized gene TMEM106B significantly associated with FTLD-TDP, especially in patients with GRN mutations (Van Deerlin et al., 2010). It was then regarded as a risk factor causing FTLD-TDP by finding that three SNPs (rs1990622, rs6966915, and rs1020004) encompassing TMEM106B are correlated to the FTLD-TDP (Suzuki and Matsuoka, 2016). Later studies have demonstrated TMEM106B is a disease modifier by modifying the effects of PGRN haploinsufficiency. Intricate interplays between TMEM106B and PGRN has been revealed by in vivo and in vitro studies, and the current researches have revealed their opposite roles lysosomal dysfunctions which is a converging mechanism of various neurodegenerative diseases (Brady et al., 2013, Klein et al., 2017).

TMEM106B is a glycosylated (Figure 1B) type Ⅱ transmembrane protein locating in late endosomes and lysosomes (Brady et al., 2014). It has a length of 274 amino acids and consists of three domains: a lumenal domain of 157 residues, a transmembrane domain of 21 residues and mechanisms of TMEM106B. Brady et al. (2014) and Suzuki and Matsuoka (2016) reported that TMEM106B undergoes serial regulated intramembrane proteolysis events (RIP) in which the lumenal domain is firstly shed leaving a N-terminal fragment (NTF) (Figure 1C) anchored at the lysosomal membrane, and the cytosolic domain is then liberated generating intracellular domain (ICD) (Figure 1D). Although the functions of TMEM106B RIP remain unknown, the overexpression of full length TMEM106B, NTF or ICD have been shown to induce cell death, possibly by the caspase-dependent mitochondrial pathways and other possible pathways such as lysosomal cell death pathway (Suzuki and Matsuoka, 2016). Regarding to the lumenal domain, an updated research reveals that the lumenal domain of TMEM106B physically binds to vacuolar-ATPase accessory protein 1 (ATP6AP1) which is required for assembling the vacuolar (V)-ATPase, a protein pump responsible for lumenal acidification (Klein et al., 2017). The deficiency of TMEM106B results in a decline of ATP6AP1 and V0 subunits of V-ATPase and consequently impairs lysosomal acidification (Klein et al., 2017). Thus, Klein et al. (2017) hypothesize that TMEM106B regulates lysosomal acidification by interacting with V-ATPase.

Albeit a variety of evidence has shown TMEM106B is an essential disease modifier and a potential therapeutic target, its molecular mechanisms have not been elucidated. Many biochemical techniques and gene knockout mice have been used to investigate the gene expression changes. However, no structural analyses have ever been done to TMEM106B. As protein structures determine protein functions, structural biology methods may shed lights on TMEM106B functions. Unfortunately, 3D structures of either TMEM106B or its interacting protein ATP6AP1 are currently unavailable, possibly due to the difficulty of determining membrane proteins by current experimental methods. Nevertheless, computational methods, such as homology modeling and ab initio modeling, can provide efficient and trustworthy structure prediction. In addition, homology detection in the process of homology modeling can provide extra hints for protein function in the light of evolution. Hence, this project aims to predict and analyze 3D structures of protein TMEM106B and its interacting protein ATPAP1 by using bioinformatics approaches to form new hypotheses regarding to the molecular mechanisms.

The results of this project has not been published, please contact me if you are interested.

Selected figures of results.

Figure 1. In silico generated intact model of TMEM106B. The homology model of the lumenal domain and the transmembrane domain and the ab initio model 5 of the cytosolic domain were placed together.

Related publications

Brady, O.A., Zhou, X. and Hu, F., 2014. Regulated intramembrane proteolysis of the frontotemporal lobar degeneration risk factor, TMEM106B, by signal peptide peptidase-like 2a (SPPL2a). Journal of Biological Chemistry, 289(28), pp.19670–19680.

Brady, O.A., Zheng, Y., Murphy, K., Huang, M. and Hu, F., 2013. The frontotemporal lobar degeneration risk factor, TMEM106B, regulates lysosomal morphology and function. Human Molecular Genetics, 22(4), pp.685–695.

Zhou, X., Sun, L., Brady, O.A., Murphy, K.A. and Hu, F., 2017. Elevated TMEM106B levels exaggerate lipofuscin accumulation and lysosomal dysfunction in aged mice with progranulin deficiency. Acta Neuropathologica Communications, [online] 5(1), p.9. Available at: <https://doi.org/10.1186/s40478-017-0412-1>.

Suzuki, H. and Matsuoka, M., 2016. The lysosomal trafficking transmembrane protein 106B is linked to cell death. Journal of Biological Chemistry, 291(41), pp.21448–21460.