Topological analysis of TMEM180, a newly identified membrane protein that is highly expressed in colorectal cancer cells

New target molecules for diagnosis of and drug development for colorectal cancer (CRC) are always in great demand. Previously, we identified a new colorectal cancer–specific protein, TMEM180, and successfully developed an anti-TMEM180 monoclonal antibody (mAb) for the diagnosis and treatment of CRC. Although TMEM180 is categorized as a member of the cation symporter family and multi-pass membrane protein, little is known about its function. In this study, we examined topology of this membrane protein and analyzed its function. Using a homology model of human TMEM180, we experimentally determined that the protein has 12 transmembrane domains, and that its N-terminal and C-termini are exposed extracellularly. Moreover, we found that the putative cation-binding site of TMEM180 is conserved among orthologs, and that its position is similar to that of melibiose transporter MelB. These results suggest that TMEM180 acts as a cation symporter. Our topological analysis based on the homology model provides insight into functional and structural roles of TMEM180 that may help to elucidate the pathology of CRC. Highlights A homology model of human TMEM180 was generated by secondary structure prediction. Putative cation-binding residues are conserved in TMEM180 orthologs. Both the N-terminus and C-terminus of TMEM180 are extracellularly exposed. TMEM180 is a 12-transmembrane protein. TMEM180 could act as a cation symporter.


Introduction
Colorectal cancer (CRC) is one of the most common cancers, and has high mortality and incidence rates around the world [1]. Consequently, there is a considerable incentive to identify new target molecules for diagnosis of and drug development for this disease. We previously identified a new CRC-specific protein, TMEM180, and developed the anti-TMEM180 monoclonal antibody (mAb) for use in diagnosing and treating CRC [2].
In order to elucidate the functional and structural role of TMEM180, we decided to analyze the topology of TMEM180. Here, we present our in silico analysis of TMEM180 and a homology model based on secondary structure prediction. We also present a reliable experimental topological model of TMEM180. We show that the both N-terminus and C-terminus of TMEM180 are extracellularly exposed. Immunofluorescence with or without cell membrane permeabilization was performed to confirm the topological model of TMEM180. We found that the putative cation-binding site in TMEM180 is highly conserved among TMEM180 orthologs. Our topological analysis based on the homology model provides insight into the functional and structural roles of TMEM180, which may help elucidate the pathology of CRC.

Cells and Cell culture
Human embryonic kidney (HEK) 293T cells were purchased from American Type Culture Collection. HEK293T cells were cultured in DMEM (Wako) supplemented with 10% FBS (Thermo Fisher Scientific) and 1% penicillin-streptomycin-amphotericin B suspension (Wako) at 37°C under a 5% CO 2 atmosphere.

Plasmid construction
The human TMEM180 gene was PCR amplified from plasmid pCMV-TMEM180 (OriGene) and fused with an EcoR1 restriction site in the 5' region and a BamH1 restriction site in the 3' region. The PCR products were cloned into pcDNA3. 3  were acquired on a BZ-X710 fluorescence microscope (Keyence) and analyzed using ImageJ version 1.52a (NIH).
All structures in this study were visualized using PyMol version 1.3 [23].
TM1, TM9, and TM12 were predicted by all programs, and TM3, TM5, and TM6 were predicted by all programs except one (Fig. 1A,B). Of the 12 TMs, 6 were predicted with high accuracy, whereas the other half were difficult to predict. There was no regularity of the positions of the TMs that were predicted with (Fig. 1B). Consequently, we decided to use secondary structure prediction and construct a homology model to investigate the topology of human TMEM180.

Homology modeling of human TMEM180
To address the function of TMEM180 based on structural information, we constructed a performed secondary structure prediction using DELTA-FORTE [18] and HHpred [19] independently. Both algorithms returned the melibiose transporter MelB from Salmonella typhimurium (stMelB) (PDBID 4M64) [24] (Fig. 2A, right) chain A as the top-scoring hit. We then used the on-line program MEDELLER [20], which is specialized for homology modeling of membrane proteins, and adopted stMelB chain A as the template structure. We obtained a homology model of human TMEM180 containing 12 TM ( Fig. 2A, left). Although human TMEM180 has only 19% sequence identity with stMelB, comparison of the crystal structure of stMelB with the homology model of human TMEM180 revealed that five putative cation-binding residues (Asn63, Asn66, Asp67, Thr139, and Asp142) in the homology model were located in positions similar to those of five cation-binding residues of stMelB (Asp55, Asn58, Asp59, Thr121, and Asp124) [24,25] (Fig. 2A,B). Multiple sequence alignment of TMEM180 from human, bovine, mouse, chicken, frog, and zebrafish revealed that these putative cation-binding residues are conserved across all species (Fig. 2B). In addition, we used the ConSurf Server [22] to identify evolutionarily conserved amino acid residues of TMEM180 from human, bovine, mouse, chicken, frog, and zebrafish. Mapping of the results onto the homology model of human TMEM180 revealed that the core helix regions are highly conserved, whereas the long loops are not (Fig. 2C). These results indicate that TMEM180 may have 12 TM regions and may act as a cation symporter. We then examined to verify this homology model experimentally.

N-terminal and C-terminal topology of human TMEM180
First, we decided to determine the N-and C-terminal topology of human TMEM180. For this purpose, FLAG-tags were introduced at the N-or C-terminus of TMEM180 (Fig. 3B,D), and HEK293T cells were transiently transfected with the FLAG-tagged proteins. HEK293T cells transfected with N-terminally 3xFLAG-tagged TMEM180 were stained positively with both anti-TMEM180 mAb and anti-FLAG mAb after permeabilization (Fig. 3A,B upper panel).
Non-permeabilized cells were also stained positively by both antibodies (Fig. 3A,B lower   panel). Similarly, cells transfected with the C-terminally 3xFLAG-tagged construct were stained by both antibodies whether the cells were permeabilized (Fig. 3C,D upper panel) or not permeabilized (Fig. 3C,D lower panel). These results indicate that both the N-and C-terminus of TMEM180 are extracellularly exposed. Moreover, TMEM180 must have an even number of transmembrane segments in the plasma membrane.

The entire topology of human TMEM180
To elucidate the topology of TMEM180 in more detail, we constructed TMEM180 genes with an N-terminal HA-tag in the FLAG-tag inserted in loops predicted based on the homology model (Fig. 4A). We chose to use the HA-Tag to confirm expression rather than the anti-TMEM180 mAb that we developed previously [2] because the location of the epitope for this mAb remains unknown. HEK293T cells transfected with a TMEM180 gene with FLAG inserted into five predicted loops (79FLAG, 153FLAG, 236FLAG, 329FLAG, and 398FLAG) were stained by both anti-HA pAb and anti-FLAG mAb regardless of permeabilization (Fig. 4B). On the other hand, HEK293T cells transfected with a TMEM180 gene with FLAG inserted into the other six predicted loops (46FLAG, 123FLAG, 188FLAG, 298FLAG, 357FLAG, and 445FLAG) were stained with both anti-HA pAb and anti-FLAG mAb only in permeabilized HEK293T cells (Fig. 4B). These results indicate that TMEM180 has 12 TMs, and that the topology of our homology model is accurate.

Discussion
The normal counterpart to cancer commonly contains not only normal colonocytes but also fibroblasts, white blood cells, and smooth muscle cells even when laser micro-dissection is used to isolate the normal colonocytes. Therefore, it is difficult to obtain truly cancer-specific molecules under such conditions, even when comprehensive expression analysis is performed. Previously, in our laboratory, we subjected 100% pure normal colonocytes obtained from lavage solution after colonoscopy to comprehensive expression analysis along with colorectal cancer cell lines [26]. Following this analysis, we performed RT-PCR and in-situ hybridization; as a result, we identified TMEM180, a CRC-specific multi-pass membrane protein of unknown function [2]. In this study, we analyzed protein structure, including topology, to elucidate the molecular function of TMEM180.
In human TMEM180, cation binding residues (Asn63, Asn66, Asp67, and Thr139) and residues shared by the cation and sugar bindings sites (Asp142 and Lys412, but not Leu138) are located in similar positions (Fig. S1, red and purple) and are highly conserved in TMEM180 orthologs [3]. On the other hand, the galactoside-binding residues (Fig. S1, blue) are not conserved at all in TMEM180. Based on our topological analysis, TMEM180 may act as a cation symporter, but does not transport galactosides. Further functional analysis is necessary to determine the substrate of TMEM180.
Previously, we reported that TMEM180 is upregulated under low-oxygen conditions, and that TMEM180 may play an important role in the uptake or metabolism of glutamine and arginine in cancer cell proliferation [2]. We also showed that expression of TMEM180 is not essential for mouse development, as Tmem180-knockout mice do not exhibit embryonic, neonatal, or postnatal lethality [2]. During the preparation of this manuscript, a group from Huazhong University of Science and Technology reported that SNP rs2001389 is significantly associated with pancreatic cancer risk, weakens the binding activity of the transcriptional repressor CCCTC-binding factor (CTCF), and is associated with the lower expression of MFSD13A/TMEM180. In addition, they showed that siRNA-mediated knockdown of MFSD13A/TMEM180 promotes proliferation of pancreatic cancer cell lines [27]. On the other hand, we found that shRNA-mediated knockdown of TMEM180 suppressed proliferation of CRC cell lines (data not shown). Thus, the relationship between expression of TMEM180 and cancer proliferation remains controversial.
This is the first report of the topological structure of TMEM180. We performed in silico analysis (Fig. 1A,B) and constructed a homology model of human TMEM180 based on secondary structure prediction ( Fig. 2A). We then experimentally confirmed that TMEM180 has 12 TM domains (Fig. 4B) and showed that its N-and C-termini are extracellularly exposed (Fig. 3A,C). Moreover, we found that the putative cation-binding site of TMEM180 is conserved among orthologs, and that the positions of the important residues are similar to those in the melibiose transporter MelB ( Fig. 2A,B). These results suggest that TMEM180 acts as a cation symporter. We expect that our model will be useful for analysis of biological function of TMEM180, as well as in determination of its high-resolution x-ray crystal structure or cryo-electron microscopic structure.

Disclosure statement
Y.M. is co-founder, shareholder, and Board Member of RIN Institute, the company that owns the anti-TMEM180 mAb. T.A. declares no relevant conflicts of interest.