Discussion

LSP1 was detected as a relatively abundant phosphoprotein in primary B cells and WEHI-231 cells as part of a larger effort to identify ligand-induced changes in protein phosphorylation. Because previous work indicated it was a target of multiple signaling pathways, it was important to identify the sites on LSP1 that are phosphorylated in intact cells. The availability of sensitive mass spectrometry methods made it possible to accomplish this with small amounts of material (e.g., approximately 1 pmol of protein per 2-D spot) resolved from whole cell lysates by 2-D gel electrophoresis. Both the long and short splice variants of LSP1 were present as multiply-charged variants that were separated during 2-D electrophoresis. The long splice variant was resolved into seven spots and the short variant into six spots. Each of the spots was labeled with ³²P, indicating that they were all phosphorylated forms of LSP1. Multiple phosphorylation is supported by previous studies in which immunoprecipitated T-cell LSP1 was treated with alkaline phosphatase and analyzed by 2-D electrophoresis(7). B-cell LSP1 was highly phosphorylated even under basal conditions. In quiescent B cells, cultured for 3 hours in serum-free medium, the predominant forms of LSP1 were those containing 4 and 5 moles of phosphate. LSP1 from WEHI-231 cells cultured in serum-free medium was predominantly present as forms containing 3 and 4 moles of phosphate. Previous studies have also shown a significant level of basal LSP1 phosphorylation in BAL17 B lymphoma cells(16), primary murine T cells (8), human B cells, and the human CESS B lymphoblastoid cell line(11). However, cells in these studies were cultured in the presence of 10% fetal bovine serum, which could have caused increased levels of LSP1 phosphorylation.

The phosphorylation of LSP1 in WEHI-231 and primary B cells was differentially affected by ligands. Stimulation of B cells with a concentration of IL-4 that elicited a robust response in the AfCS B-cell ligand screen had modest effects on LSP1 phosphorylation. In contrast, activation of the B-cell receptor with anti-IgM antibody or stimulation of PKC with PMA caused a significant increase in LSP1 phosphorylation. In WEHI-231 cells, stimulation resulted in a shift in the intensities of either fluorescent- or ³²P-labeled spots to the more acidic forms. Similar results were observed in primary B cells. In each case, the effects of PMA were stronger than those resulting from activation of the B-cell receptor. These data are consistent with previous results, indicating an important role for PKC in ligand-induced LSP1 phosphorylation. Phosphorylation of murine T-cell LSP1 is enhanced following stimulation with concanavalin A and PMA, but not the calcium ionophore A23187 (8). Phosphorylation of LSP1 was also enhanced by PMA treatment of peripheral blood B cells and the human CESS lymphoblastoid cell line(11). The phosphorylation of LSP1 induced by PMA is inhibited by the PKC-selective inhibitor staurosporine(11). However, other protein kinases are also likely to play a role in LSP1 phosphorylation. For example, human neutrophil LSP1 is phosphorylated in vitro by MAPKAP kinase 2, and introduction of a MAPKAP kinase 2 inhibitory peptide into intact neutrophils repressed LSP1 phosphorylation induced by PMA or fMet-Leu-Phe(13).

A total of seven in vivo phosphorylation sites were detected in LSP1 by mass spectrometry based methods (summarized in Fig. 8). This number is in concordance with the resolution of LSP1 into at least seven charged species by isoelectric focusing. These results suggest that most, or all, of the major sites of phosphorylation were identified. However, coverage of the LSP1 sequence by the peptides identified here was not complete, and additional sites could be present. Although the role of phosphorylation in regulating the function of LSP1 is unknown, the data provided here show that LSP1 is phosphorylated at multiple sites in vivo. Two of the phosphorylation sites we identified, Ser-205 and Ser-243, are located within the caldesmon-like domain of the protein, and two of the sites, Ser-180 and Ser-184, are immediately adjacent to this domain(18). It is possible that the phosphorylation of these sites may alter the interaction of LSP1 with actin or other proteins. The potential to be phosphorylated by multiple kinases raises the possibility that LSP1 serves as an important integration point for regulation of the actin cytoskeleton by multiple signaling pathways. Important goals for the future are to identify the protein kinases responsible for phosphorylating LSP1 in vivo and to determine how phosphorylation at each site affects the function of LSP1. On a broader scale, this work establishes a reproducible protocol for using 2-D gels and mass spectrometry for the determination of phosphorylation sites in future experiments by the AfCS.