Phosphoproteomes of Strongylocentrotus purpuratus shell and tooth matrix: identification of a major acidic sea urchin tooth phosphoprotein, phosphodontin

Abstract

Background: Sea urchin is a major model organism for developmental biology and biomineralization research. However, identification of proteins involved in larval skeleton formation and mineralization processes in the embryo and adult, and the molecular characterization of such proteins, has just gained momentum with the sequencing of the Strongylocentrotus purpuratus genome and the introduction of high-throughput proteomics into the field.

Results: The present report contains the determination of test (shell) and tooth organic matrix phosphoproteomes. Altogether 34 phosphoproteins were identified in the biomineral organic matrices. Most phosphoproteins were specific for one compartment, only two were identified in both matrices. The sea urchin phosphoproteomes contained several obvious orthologs of mammalian proteins, such as a Src family tyrosine kinase, protein kinase C-delta 1, Dickkopf-1 and other signal transduction components, or nucleobindin. In most cases phosphorylation sites were conserved between sea urchin and mammalian proteins. However, the majority of phosphoproteins had no mammalian counterpart. The most interesting of the sea urchin-specific phosphoproteins, from the perspective of biomineralization research, was an abundant highly phosphorylated and very acidic tooth matrix protein composed of 35 very similar short sequence repeats, a predicted N-terminal secretion signal sequence, and an Asp-rich C-terminal motif, contained in [Glean3:18919].

Conclusions: The 64 phosphorylation sites determined represent the most comprehensive list of experimentally identified sea urchin protein phosphorylation sites at present and are an important addition to the recently analyzed Strongylocentrotus purpuratus shell and tooth proteomes. The identified phosphoproteins included a major, highly phosphorylated protein, [Glean3:18919], for which we suggest the name phosphodontin. Although not sequence-related to such highly phosphorylated acidic mammalian dental phosphoproteins as phosphoryn or dentin matrix protein-1, phosphodontin may perform similar functions in the sea urchin tooth. More than half of the detected proteins were not previously identified at the protein level, thus confirming the existence of proteins only known as genomic sequences previously.

Figure 1

The phosphorylation site of protein SM30-E. This peptide was also identified in a non-phosphorylated version in the present and previous studies [10,11], indicating that this site is only partially modified. The spectrum shows an uninterrupted series of y ions (y2-y7). This sequence tag, supplemented by some b ions, and the accurate mass of the complete peptide measured in the orbitrap, allowed the identification of this peptide by database searches. The most intense ions, y7 and y4, are due to preferential cleavage N-terminal of proline residues in position 7 and10 of the peptide sequence. This is a well known feature of Pro-containing peptides frequently used for manual validation of peptide assignments. Loss of H₃PO_4, indicated by —P, is first observed in b9 and y5, indicating phosphorylation of Thr in position 9 of the peptide sequence. Loss of NH₃, indicated by -17, frequently occurs upon fragmentation of Asn-containing peptides. Cyclization of N-terminal Gln to pyroglutamate is common in peptides with N-terminal Gln.

Figure 2

The phosphorylation site of protein P19. This peptide was also identified in a non-phosphorylated version in the present and a previous study [11] indicating that the site is only partially modified. P19 was implicated in biomineralization events previously [45] and was identified as a phosphoprotein in L. variegatus tooth tissue by phosphor-specific staining [46]. However, previous results also indicated that this protein was at best a very minor component of the intracrystalline matrix [11]. A high match of observed to theoretically expected fragments, including a sequence tag of y3-y7, together with the accurate measurement of the peptide mass, allowed the identification of this peptide. Loss of H₃PO₄, indicated by —P, and first observed with y10 indicated the presence of a phospho group at the only serine in the peptide sequence. Loss of NH₃ and water is indicated by -17 and -18, respectively. These neutral losses are frequently observed in peptides containing Glu and Gln.

Figure 3

Identification of the phosphorylation site of peptide NVAEAAGLSSNEVTQVK. Similar to the spectrum in Fig. 2, this spectrum shows the complexity to be expected from fragmentation of a relatively long peptide by multistage activation. However, the presence of extended series of y and b ions and the accurate mass of the intact peptide measured in the orbitrap enabled the unequivocal identification of the sequence. The phosphorylation site is identified by the increase of 80Da in the y-ion series starting with y8 and the absence of such an increase in b-ions up to b9. Loss of H₃PO₄, H₂O, and NH₃ is indicated by -P, -18, and -17, respectively. The neutral loss of water and NH₃ is frequently observed upon fragmentation of peptides containing Asn, Gln, Glu, Ser and Thr, also contained in this peptide.

Figure 4

Analysis of the [Glean3:20139] sequence. The Thr- and Pro-rich sequence (25% Thr, 26% Pro) is shaded light grey and the Gln-rich sequence (63% Gln) is shaded dark grey. Identified phosphopeptides are underlined, phosphorylated Ser are in bold italics. A phosphorylated region comprising three phospho sites is sandwiched between these two domains, while one phospho site is in the N-terminal region of the Gln-rich domain.

Figure 5

Alignment of [Glean3:18649] (Similar to pleckstrin homology domain-containing protein) phosphopeptide to its human counterpart. Phosphoserine 239 [39,47] of the human pleckstrin homology domain-containing family F member 2 protein is also phosphorylated in the homologous sea urchin protein. The phosphorylated serine 248 of human PKHF2 [47] is not conserved in the sea urchin protein.

Figure 6

The sequence of phosphodontin ([Glean3:18919]), the major phosphoprotein of tooth matrix. The predicted signal sequence and the Asp-rich (72% Asp) C-terminus are doubly underlined. Alternating sequence repeats are shaded light and dark grey. Possible phosphorylation sites derived from experimental data are printed in bold and underlined. Because several repeats occur in multiple identical copies and also non-phosphorylated copies were detected for several of them, it was not possible to determine the extent of phosphorylation of this protein. Furthermore, obviously only one Ser of each repeat can be phosphorylated at a time. Experimental data favor the second Ser of each repeat as the phosphorylation site, but do not completely exclude modification of the first one. Peptides sequenced by MS/MS are printed in italics.