2216 DOI 10.1002/pmic.200300779 Proteomics 2004, 4, 2216–2220 Short Communication Identification of swiprosin 1 in human lymphocytes Françoise Vuadens1*, Nathalie Rufer2*, Annegret Kress1, Patricia Corthésy2, Philippe Schneider1 and Jean-Daniel Tissot1 1 Service Régional Vaudois de Transfusion Sanguine, Lausanne, Switzerland Swiss Institute for Experimental Cancer Research (ISREC), NCCR Molecular Oncology, Epalinges, Switzerland 2 The aim of this work was to identify a new protein that discriminated CD8 from CD4 and CD19 lymphocytes. Proteins were separated by high-resolution two-dimensional electrophoresis. After silver staining, the gel images were captured with a laser densitometer, and studied with a dedicated software. This study confirmed the presence of two spots that appeared to be preferentially associated with CD8 lymphocytes, and mass spectrometry analyzes (liquid chromatography-tandem mass spectrometry, LC-MS/MS) identified six peptides for one spot and four for the other. The peptide sequences corresponded to an unknown protein that we named swiprosin 1 (Swiss-Prot Q96C19). Molecular analysis (reverse transcriptase-polymerase chain reaction, RT-PCR) and Northern blots confirmed that the gene expression was increased in purified populations of CD8 lymphocytes, when compared to CD19 and CD4 lymphocytes. Database mining revealed that swiprosin 1 contains two potential EF-hand domains, and therefore may have a role in calcium signaling. Its predominant presence in CD8 lymphocytes suggests that it may be involved in functions that are important for cytotoxic lymphocytes. Keywords: Human blood / Lymphocytes / Swiprosin 1 / Two-dimensional gel electrophoresis The lymphoid lineages, consisting of B, T, and natural killer cells, are generated from a common lymphoid progenitor [1]. B lymphocytes can differentiate into mature plasmocytes that produce antibodies, and are therefore implicated in the humoral immune response, whereas T lymphocytes are the major players in the cellular immune response [2, 3]. B cell differentiation is a highly regulated process with pathways and steps that have been well delineated [4]. B cells express uniquely CD19, a 95 kDa protein present throughout the B cell lineage until plasma cell differentiation [5]. T lymphocytes precursors originate in the bone marrow and mature in the thymus [6]. Mature T cells are separated in two main subsets, which can be differentiated according to the mutually exclusive expression of two polypeptides, CD4 and CD8. The cells being key components of cellular immunology, they have been Correspondence: Prof. Jean-Daniel Tissot, Service Régional Vaudois de Transfusion Sanguine, rue du Bugnon 27, CH-1005 Lausanne, Switzerland E-mail: jean-daniel.tissot@chuv.hospvd.ch Fax: 141-21-314-65-78 Abbreviation: FACS, fluorescence-activated cell sorter  2004 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim Received Revised Accepted 8/12/03 2/2/04 6/2/04 extensively studied, and several studies reporting on their proteome have been reported (reviewed in [7]). By comparative proteomic analysis of CD19, CD8, as well as of CD4 lymphocytes isolated from human blood, several spots discriminating the three populations have been recognized [8]. Here, we report on the identification of an unknown protein that appeared to be CD8 specific. Human CD19, CD8, and CD4 lymphocytes were purified from human blood as already described [8]. Proteomic studies were performed using identical protocols, except that nonlinear 4 to 7 immobilized pH gradients were used for IEF instead of nonlinear 3 to 10 pH gradients. For this study, three groups of gels were compared, that were CD19, CD8, and CD4 lymphocytes, and each group was constituted of four different 2-D gels, each corresponding to populations of cells purified from different pools of blood. 2-D gel image studies using Melanie 3.0 (Genebio, Geneva, Switzerland) confirmed that the two spots that appeared to be CD8 specific using pH 3–10 pH gradient for IEF were also preferentially expressed in CD8 lymphocytes when studied using 4 to 7 pH gradient (Fig. 1). LC* These authors contributed equally to this work. www.proteomics-journal.de Proteomics 2004, 4, 2216–2220 Swiprosin 1 in lymphocytes 2217 Figure 1. (A) Ammoniacal silver stained high-resolution 2-D polyacrylamide gel of CD8 lymphocytes isolated from healthy blood donors (first dimension: immobilized 4 to 7 nonlinear pH 4–7 gradient; second dimension: 9–16% gradient polyacrylamide gel electrophoresis). (B) Detail of the area corresponding to swiprosin 1 (pI 4.6–5.3; molecular weights 30 000– 26 000). Two control spots were expressed (%Volume) without significant differences between the three groups (N = 4 in each group, mean 6 SD, P = 0.257, panel C, and P = 0.16, panel D; one-way analysis of variance), contrasting with the two spots corresponding to swiprosin 1, which appeared significantly more expressed in CD8 lymphocytes (all pairwise multiple comparison, method of Holm-Sidak: P , 0.001, CD8 vs. CD19; P , 0.001, CD8 vs. CD4: panel E; P , 0.001, CD8 vs. CD19; P = 0.001, CD8 vs. CD4: panel F). No significant differences were observed between CD4 and CD19 lymphocytes (P = 0.11: panel E; P = 0.254: panel F). MS/MS analyzes were performed as described [8]. Six peptides were characterized for one spot and four for the other. These two spots corresponded to an unidentified protein, that we named swiprosin 1 (Swiss-Prot: Q96C19). As already mentioned, swiprosin 1 appeared as two spots, showing different pI and molecular masses. The biochemical origin of this heterogeneity is unknown, but could be due to either glycosylation (potential N-glycosylation site at position 202), phosphorylation or proteolytic degradation. The gene coding to swiprosin 1 was recently sequenced, and is located on chromosome 1 [9]. The deduced sequence of swiprosin 1, as well as its internal peptide sequences determined by LC-MS/MS are shown in Fig. 2.  2004 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.proteomics-journal.de 2218 F. Vuadens et al. Proteomics 2004, 4, 2216–2220 Figure 2. Sequence of swiprosin 1. Amino acids in italics represent the potential EF-hand domains, whereas the amino acids underlined represent those identified by LC-MS/MS analysis. To further understand at which level the differential expression of swiprosin 1 in various lymphocyte population is achieved, studies of swiprosin 1 mRNA with reverse transcription-polymerase chain reaction (RT-PCR) were performed. For this purpose, we have used a modified RTPCR protocol that relies on the detection of specific cDNAs after global amplification of expressed mRNAs [10, 11]. Because the method yields sufficient cDNA from as few as five cells, it allows the analysis of gene expression in small highly purified subpopulations of lymphocytes. CD19, CD4, and CD8 lymphocytes were prepurified on magnetic beads (Dynal Biotech, Oslo, Norway). Gene expression analysis was performed on five cells sorted from each purified lymphocyte subpopulation by fluorescence-activated cell sorter (FACS; FACS-Start SE; Becton Dickinson, San Diego, CA, USA) (see R1 gating region, Fig. 3A). The procedures for cDNA preparation, cDNA amplification as well as the RT-PCR were recently described in details [12]. We used the following primers for the positive controls: (i) glyceraldehyde phosphate dehydrogenase (GAPDH) 5’-GGACCTGACCTGCCGTC TAG-3’, Rev-5’-CCACCACCCTGTTGCTGTAG-3’; and (ii) CD3: 5’-CGTTCAGTTCCCTCCTTTTCTT-3’, Rev-5’-GATTAGGGGGTTGGTAGGGAGTG-3’. For swiprosin 1, we prepared the following primers; Swi1.3: 5’-CTCGAGGCCAC AGTTATGCAA-3’, Rev-5’ATCCGCTAAGGCAAAC GCA-3’; Swi1.4: 5’-CAGTCCAAACAAGTGAGTGGCC3’, Rev-5’-TAAAACCTCTGCAAGTCCGCG-3’. The analyzes were either performed on lymphocytes pooled from different donors (experiments #1 and #2) or from two individual donors (experiments #3 and #4). These results first showed that T lymphocytes (CD31CD41 and CD31CD81), but none of the aliquots of the B lymphocytes (CD32CD191), yielded a detectable CD3-specific product (Fig. 3B). Second, we showed that swiprosin 1 mRNA was more expressed in CD8 lymphocytes compared to either CD19 or CD4 lymphocytes (P , 0.001, chi-square = 22.873 with one degree of freedom) (Table 1). A representative example Figure 3. (A) Example of FACS analysis of the lymphocytes purified for RT-PCR. CD19-PE, CD4-PE, and CD8-PE represent phycoerythrin coupled to either anti-CD19, CD4, or CD8 antibodies, whereas CD3-FITC represents anti-CD3 antibodies coupled to fluorescein isothiocyanate. CD19, CD4, and CD8 populations were purified by using Dynal beads (purity was . 90%). Five cells of each purified subpopulations were isolated by flow cytometry (see gating region R1; CD19 cells were CD191CD32; CD4 cells were CD41CD31; and CD8 cells were CD81CD31). (B) Results of RT-PCR. Each lane corresponds to a five-cell sample, and results are presented according to the primers used. GAPDH was the positive control to ensure that cDNA material was present. CD3 was the positive control for the T lymphocytes. Swi1.3 and Swi1.4 are two different sets of primers of swiprosin 1, which yield the same results. Data from ten independent five-cell aliquots are shown. (2) negative; (1) positive controls. The amplification products were visualized on a 1.2% agarose gel after ethidium bromide staining.  2004 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.proteomics-journal.de Proteomics 2004, 4, 2216–2220 Swiprosin 1 in lymphocytes 2219 Table 1. Summary of RT-PCR results Experiment Ratio of positive five-cell samples for Swi1.3 and Swi1.4 primers/negative five-cell samples (% of positive) CD19 CD4 CD8 #1 (pool of donors) #2 (pool of donors) #3 (donor #A) #4 (donor #B) 2/5 (28.6%) 4/10 (28.6%) 1/16 (5.9%) 3/13 (18.8%) 4/3 (57.1%) 5/9 (38.5%) 2/12 (14.3%) 10/6 (62.5%) 5/1 (83.3%) 10/4 (71.4%) 5/12 (29.4%) 15/1 (93.8%) Total 10/44 (18.5%) 21/30 (41.2%) 35/18 (66%) of the gene expression pattern found in one individual donor #B (experiment #4) is depicted in Fig. 3B. The expression of mRNA was also tested by Northern blot analysis. Eight mg of total RNA from sorted CD19, CD4 and CD8 cells was prepared with the QIAGEN RNeasy kit (Qiagen, Basel, Switzerland), separated by formaldehyde 1.2% agarose gel electrophoresis and transferred to positively charged nylon membranes (Hybond N1; Amersham Pharmacia Biotech Europe, Dübendorf, Switzerland) by electroblotting. RNA ladders (New England Biolabs, Beverly, MA, USA) were loaded, as markers, next to the samples. Probes for either human swiprosin 1 or human GAPDH were prepared from the RT-PCR product obtained with primers Swi1.3 and Swi1.3 rev and primers GAPDH and GAPDH rev, respectively. Briefly, gel purified PCR products were labeled by random priming reactions, while the probe for the RNA ladder was obtained by reverse transcription of a 0.1 mg RNA ladder aliquot using 1 mM random-labeled hexamer primers. Hybridizations were carried out in 0.5 M phosphate buffer, pH 7.4, 1 mM EDTA, 7% SDS, 1% bovine serum albumin. After washing with 150 mM NaCl, 15 mM sodium citrate, pH 7.0, the quantification of hybridization signals was carried by phosphorimager. A typical result is depicted in Fig. 4. When normalized to the amount of GAPDH mRNA, swiprosin 1 mRNA appeared sixfold higher in CD8 when compared to CD19 lymphocytes, and threefold higher when compared to CD4 lymphocytes. Taken together, our results provide evidences that swiprosin 1 is a protein that is preferentially expressed by CD8 lymphocytes when compared to either CD4 or CD19 lymphocytes. Of note, our studies revealed a variable expression of swiprosin 1 mRNA in different five-cell samples (Fig. 3B). At this stage, we cannot decide whether this observation reflects the purity of sorted five-cell sample (around 98%), a stochastic element of the PCR-based amplification from very small transcript numbers or cell biological heterogeneity among each subsets. In addition, it must be kept in mind that the relation between mRNA  2004 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim Figure 4. Northern blot analysis of swiprosin 1 mRNA. Upper panel, positive control with GAPDH. Lower panel, expression of swiprosin 1 mRNA. After normalization of the band corresponding to GAPDH mRNA, the swiprosin 1 mRNA band was higher in CD8 lymphocytes when compared to either CD19 lymphocytes or CD4 lymphocytes. and protein expression can be quite variable [13]. Nevertheless, the most likely explanation is related to the fact that CD8, CD4 as well as CD19 five-cell samples were probably heterogeneous. The five cell-samples probably contained different subsets of cell populations, each characterized by different functions and cell surface markers [14–21]. Thus, the variations of swiprosin 1 mRNA expression observed in five-cell samples probably express the diversity of cells that were purified by cell sorting using two cell surface markers (CD32CD191 or CD31CD41 or CD31CD81). Exploring the function of swiprosin 1 should provide insights into the functional diversity of various lymphocyte populations. The identification of the location of swiprosin 1 in lymphocytes, notably the search of its presence in lipid rafts [22] or other cellular compartments using specific antibodies, will be an important step to attain this goal. In addition, cell sorting of various CD8 populations using CD7 as surface marker will allow the study of the expression of swiprosin 1 in the three populations of CD8 cells that are CD7 high (subset containing naïve and memory cells), CD7 low and CD7 negative (subsets containing effector cells) [18]. Database analyses revealed that proteins similar to swiprosin 1 are expressed in other species. The protein appears to be absent in primitive species like yeast or worms but it is present in Drosophila, that does not poswww.proteomics-journal.de 2220 F. Vuadens et al. sess an adaptive immune system and T cells. This suggests that swiprosin 1 expression is not limited to lymphocytes. Moreover, it was shown, by microarray assays, that swiprosin 1 mRNA is present in different tissues, but more importantly in bone marrow, thymus and spleen (http:// genome-www.stanford.edu/cgi-bin/genecards/carddisp? MGC4342). Until now, nothing is known about its function. Proteomic data and database mining (http://harvester. embl.de/harvester/Q96C/Q96C19.htm) showed that the protein is characterized by the presence of two potential EF-hand motifs that may be involved in calcium binding. The fact that swiprosin 1 is predominantly expressed in CD8 lymphocytes suggests that it may play a role in the cellular pathways involved in cytotoxicity. References [1] Kondo, M., Weissman, I. L., Akashi, K., Cell 1997, 91, 661– 672. [2] Delves, P. J., Roitt, I. M., N. Engl. J. Med. 2000, 343, 37–49. [3] Delves, P. J., Roitt, I. M., N. Engl. J. Med. 2000, 343, 108–117. [4] Hardy, R. R., Hayakawa, K., Annu. Rev. Immunol. 2001, 19, 595–621. [5] Tedder, T. F., Haas, K. M., Poe, J. C., Biochem. Soc. Trans. 2002, 30, 807–811. [6] Busslinger, M., Nutt, S. L., Rolink, A. G., Curr. Opin. Immunol. 2000, 12, 151–158.  2004 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim Proteomics 2004, 4, 2216–2220 [7] Tissot, J. D., Schneider, P., in: Hondermarck, H. (Ed.), Proteomics: Biomedical and Pharmaceutical Applications, Kluwer Academic Publisher, Amsterdam 2004, in press. [8] Vuadens, F., Gasparini, D., Deon, C., Sanchez, J. C. et al., Proteomics 2002, 2, 105–111. [9] Strausberg, R. L., Feingold, E. A., Grouse, L. H., Derge, J. G. et al., Proc. Natl. Acad. Sci. USA 2002, 99, 16899–16903. [10] Brady, G., Iscove, N. N., Methods Enzymol. 1993, 225, 611– 623. [11] Sauvageau, G., Lansdorp, P. M., Eaves, C. J., Hogge, D. E. et al., Proc. Natl. Acad. Sci. USA 1994, 91, 12223–12227. [12] Bigouret, V., Hoffmann, T., Arlettaz, L., Villard, J. et al., Blood 2003, 101, 3198–3204. [13] Anderson, N. L., Anderson, N. G., Electrophoresis 1998, 19, 1853–1861. [14] Diehl, S., Rincon, M., Mol. Immunol. 2002, 39, 531–536. [15] Ellefsen, K., Harari, A., Champagne, P., Bart, P. A. et al., Eur. J. Immunol. 2002, 32, 3756–3764. [16] Harari, A., Ellefsen, K., Champagne, P., Nobile, M. et al., Adv. Exp. Med. Biol. 2002, 512, 155–164. [17] Murphy, K. M., Reiner, S. L., Nat. Rev. Immunol. 2002, 2, 933–944. [18] Aandahl, E. M., Sandberg, J. K., Beckerman, K. P., Tasken, K. et al., J. Immunol. 2003, 170, 2349–2355. [19] Harari, A., Petitpierre, S., Vallelian, F., Pantaleo, G., Blood 2004, 103, 966–972. [20] Otero, D. C., Anzelon, A. N., Rickert, R. C., J. Immunol. 2003, 170, 73–83. [21] Rufer, N., Zippelius, A., Batard, P., Pittet, M. J. et al., Blood 2003, 102, 1779–1787. [22] Bini, L., Pacini, S., Liberatore, S., Valensin, S. et al., Biochem. J. 2003, 369, 301–309. www.proteomics-journal.de