Mesothelin, Stereocilin, and Otoancorin are predicted to have superhelical structures with ARM-type repeats

Abstract

Background: Mesothelin is a 40 kDa protein present on the surface of normal mesothelial cells and overexpressed in many human tumours, including mesothelioma and ovarian and pancreatic adenocarcinoma. It forms a strong and specific complex with MUC16, which is also highly expressed on the surface of mesothelioma and ovarian cancer cells. This binding has been suggested to be the basis of ovarian cancer metastasis. Knowledge of the structure of this protein will be useful, for example, in building a structural model of the MUC16-mesothelin complex. Mesothelin is produced as a precursor, which is cleaved by furin to produce the N-terminal half, which is called the megakaryocyte potentiating factor (MPF), and the C-terminal half, which is mesothelin. Little is known about the function of mesothelin and there is no information on its possible three-dimensional structure. Mesothelin has been reported to be homologous to the deafness-related inner ear proteins otoancorin and stereocilin, for neither of which the three-dimensional structure is known.

Results: The BLAST and PSI-BLAST searches confirmed that mesothelin and mesothelin precursor proteins are remotely homologous to stereocilin and otoancorin and more closely homologous to the hypothetical protein MPFL (MPF-like). Secondary structure prediction servers predicted a predominantly helical structure for both mesothelin and mesothelin precursor proteins and also for stereocilin and otoancorin. Three-dimensional structure prediction servers INHUB and I-TASSER produced structural models for mesothelin, which consisted of superhelical structures with ARM-type repeats in conformity with the secondary structure predictions. Similar ARM-type superhelical repeat structures were predicted by 3D-PSSM server for mesothelin precursor and for stereocilin and otoancorin proteins.

Conclusion: The mesothelin superfamily of proteins, which includes mesothelin, mesothelin precursor, megakaryocyte potentiating factor, MPFL, stereocilin and otoancorin, are predicted to have superhelical structures with ARM-type repeats. We suggest that all of these function as superhelical lectins to bind the carbohydrate moieties of extracellular glycoproteins.

Figure 1

Amino acid sequence and predicted secondary structure for human mesothelin. The first line gives the residue serial numbers and the second the sequence. The mature mesothelin sequence starts from the residue number 1, after the furin cleavage at the ARG-ARG (rr) sequence of the precursor. The signalling sequence for the GPI attachment that was suggested in an earlier study [1] is shown in low-case italics; the signalling sequence predicted using a current prediction program [45] is colored green. The 9 lines that follow give the predicted secondary structural type for each residue. The names of the programs are indicated at the right-hand side. The blue color indicates residues that were predicted to be helical unanimously by all programs. The magenta indicates all other helical predictions. The last line gives the secondary structures of model#1 of the I-TASSER server, calculated using the DSSP program (H: Helix, program output states H, h, G; E: Beta, E; C: Coil, all other output states). The helical residues in this model are colored green. The vertical lines indicate the boundaries of the 8 repeats of model#1 of the I-TASSER server.

Figure 2

Multiple alignments of the 10 repeats of 1BK5A (1bk5-1 to 1bk5-10) and the 8 repeats of mesothelin model (msln-1 to msln-8). The three double-headed arrows at the top of the Figure indicate the boundaries of the three helical regions of the 7th repeat of 1BK5A. The first two form the outer helices and the third the inner helix. The helix boundaries of other repeats do not always agree with those of the 1BK5A 7^th repeat. The residues in the 7^th repeat of 1BK5A are shown in the next line in uppercase letters. The residues in all other repeats that follow are in upper or lower case letters depending on whether they align with residues of the 7^th repeat of 1BK5A or not, respectively. The columns labelled 10, 13, 16 and 17 at the bottom are the key positions in the outer helix that are hydrophobic in many ARM/HEAT repeats [40]. Similarly LxxxAxxxL at the bottom indicates the conserved LEU and ALA positions in the inner helix, x being any residue. The hydrophobic residues (A, V, I, L, M, F, Y, W) in these 7 columns are highlighted.

Figure 3

Stereo pair of the ribbon representation of the 3D structure of model #1 from I-TASSER server. The chain runs N-terminal to C-terminal from bottom to top in the figure. Repeats 1, 3, 5, and 7 are in orange; repeats 2, 4, 6, and 8 are in yellow; the residues highlighted in Figure 2 are in green. The drawing was made using Chimera [53].

Figure 4

A phylogenetic tree of mesothelin precursor and related proteins. The unrooted tree was made using the Neighbor-Joining method [54] implemented in the MEGA3 program [55]. The tree is based on a multiple alignment from the MUSCLE program of the conserved regions of selected mesothelin, MPFL, stereocilin, and otoancorin proteins (see additional file 1 for the alignment). The number 0.2 on the scale bar indicates the number of substitutions per site.