Identification of Tspan9 as a novel platelet tetraspanin and the collagen receptor GPVI as a component of tetraspanin microdomains

Abstract

Platelets are essential for wound healing and inflammatory processes, but can also play a deleterious role by causing heart attack and stroke. Normal platelet activation is dependent on tetraspanins, a superfamily of glycoproteins that function as 'organisers' of cell membranes by recruiting other receptors and signalling proteins into tetraspanin-enriched microdomains. However, our understanding of how tetraspanin microdomains regulate platelets is hindered by the fact that only four of the 33 mammalian tetraspanins have been identified in platelets. This is because of a lack of antibodies to most tetraspanins and difficulties in measuring mRNA, due to low levels in this anucleate cell. To identify potentially platelet-expressed tetraspanins, mRNA was measured in their nucleated progenitor cell, the megakaryocyte, using serial analysis of gene expression and DNA microarrays. Amongst 19 tetraspanins identified in megakaryocytes, Tspan9, a previously uncharacterized tetraspanin, was relatively specific to these cells. Through generating the first Tspan9 antibodies, Tspan9 expression was found to be tightly regulated in platelets. The relative levels of CD9, CD151, Tspan9 and CD63 were 100, 14, 6 and 2 respectively. Since CD9 was expressed at 49000 cell surface copies per platelet, this suggested a copy number of 2800 Tspan9 molecules. Finally, Tspan9 was shown to be a component of tetraspanin microdomains that included the collagen receptor GPVI (glycoprotein VI) and integrin alpha6beta1, but not the von Willebrand receptor GPIbalpha or the integrins alphaIIbbeta3 or alpha2beta1. These findings suggest a role for Tspan9 in regulating platelet function in concert with other platelet tetraspanins and their associated proteins.

Figure 1. Tetraspanin expression in primary mouse megakaryocytes

A LongSAGE library was generated from mouse megakaryocytes grown in vitro from bone marrow precursors [11]. The bar chart shows the number of SAGE tags, for those tetraspanins that were identified, out of 53046 total tags.

Figure 2. Tspan9 is relatively specific to human megakaryocytes when compared with other haematopoietic cell types

Whole genome expression data was obtained for eight different human blood cells using Illumina HumanWG-6 v2 Expression BeadChips. Mean normalized intensity values are shown, with error bars representing S.E.M from either 4 or 7 replicates. The dashed line represents the cut-off for present/absent.

Figure 3. Generation of a new antibody identifies the Tspan9 glycoprotein as a novel megakaryocyte/platelet tetraspanin

(A) HEK-293T cells were transiently transfected with empty vector (−) or FLAG-tagged mouse Tspan9 (+) and whole cell lysates were Western blotted with the peptide-purified rabbit anti-Tspan9 antibody. Essentially identical data were generated with the chicken anti-Tspan9 antibody (results not shown). (B) FLAG-tagged mouse Tspan9, transiently transfected into HEK-293T cells, and endogenous Tspan9 from human platelets, were immunoprecipitated with rabbit anti-Tspan9 antibody and subjected to control (−) or N-glycosidase (+) treatment with the enzyme PNGase F. Samples were analysed by Western blotting with chicken anti-Tspan9 antibody. (C) HEK-293T cells were transiently transfected with FLAG-tagged mouse Tspan9, fixed, permeabilized and analysed by fluorescence confocal microscopy with the rabbit anti-Tspan9 antibody. The left panel identifies the cells by DIC (differential interference contrast) microscopy and the right panel is a mid-plane confocal section showing Tspan9 expression in a subpopulation of the cells. (D) Whole cell lysates of the human cell lines DG75 and Raji (B-cells), HPB-ALL and Jurkat (T-cells), DAMI (megakaryocyte-like), HEL (erythroleukaemia) and MEG-01 (megakaryocyte-like), and human washed platelets were normalized for protein content and then Western blotted with the rabbit anti-Tspan9 antibody. (E) Whole cell lysates of various mouse tissues and washed mouse platelets were normalized for protein content and then Western blotted with the rabbit anti-Tspan9 antibody. In this panel and the previous panel, essentially identical data were generated with the chicken anti-Tspan9 antibody (results not shown). (F) Mouse megakaryocytes and (G) proplatelets were generated from cultured fetal liver cells, allowed to spread on fibrinogen, fixed, permeabilized and analysed by fluorescence confocal microscopy with the rabbit anti-Tspan9 antibody. The left panel identifies the cells by DIC and the right panel shows Tspan9 staining. Control staining with normal rabbit immunoglobulin did not yield a substantial signal (results not shown).

Figure 4. CD9 is expressed at 49000 surface copies per cell on human platelets

The number of surface copies of CD9 per platelet was determined using the mouse IgG1 anti-CD9 monoclonal antibody 1AA2 and the Platelet Calibrator Kit from Biocytex. (A) The calibrator beads, coated with the indicated numbers of mouse IgG1 antibody molecules, were stained with FITC-conjugated anti-mouse antibody and analysed by flow cytometry. (B) Washed human platelets from a representative donor were stained with an IgG1 control or the CD9 monoclonal antibody, followed by FITC-conjugated anti-mouse antibody and analysed by flow cytometry. (C) The geometric mean fluorescence intensities from (A) and (B) were compared, to quantify CD9 on human platelets. Data are shown for six donors of diverse ethnic backgrounds.

Figure 5. Quantification of Tspan9, CD63 and CD151, relative to CD9, in human platelets

(A) HEK-293T cells were transiently transfected with empty vector control, FLAG-tagged human CD9, CD63, CD151 or mouse Tspan9 (note that the Tspan9 antibody epitope is identical in protein sequence between human and mouse). Mouse anti-FLAG immunoprecipitates were Western blotted with a rabbit anti-FLAG antibody and each tetraspanin quantified, following ECL chemiluminescence, using the GeneGnome quantitative Western blotting system from Syngene. These FLAG–tetraspanin samples were then compared to whole cell lysates of washed human platelets by Western blotting for (B) CD9, (C) CD63, (D) CD151 and (E) Tspan9 and the relative intensities quantified as in (A). (F) The data in (A–E), for platelets from six human donors of diverse ethnic backgrounds, were used to quantify the relative expression levels of each tetraspanin in human platelets. The data are presented as mean relative expression levels±S.D.

Figure 6. Tspan9 is a component of platelet tetraspanin microdomains that include GPVI and α6β1, but not αIIbβ3, α2β1 or GPIbα

(A) Human washed platelets were surface biotinylated, lysed in 1% Brij 97 (left panel) or more stringent 1% Triton X-100 (right panels, including a longer exposure), immunoprecipitated with the indicated antibodies and the biotinylated proteins were detected by Western blotting with streptavidin using the Odyssey Infrared Imaging System from LI-COR. The position of certain proteins is indicated by arrows. These results are representative of four experiments using different donors. (B) Immunoprecipitations were performed as described in (A), with the exception that non-biotinylated platelets were used, that Western blotting was performed with Tspan9, GPIbα and αIIβ antibodies, and that blots were visualized using ECL and film. (C) CD151 and GPVI immunoprecipitations were performed from 1% Brij 97 (B) or 1% Triton X-100 (T) lysates of biotinylated platelets. Biotinylated proteins were detected by Western blotting with streptavidin and visualized using chemiluminescence and film. The dividing line indicates that images were grouped from two parts of the same gel. (D) GPVI, α2 and α6 immunoprecipitations were Western blotted with CD9 and CD151 antibodies as described in (B).