Portal de Periódicos da CAPES

The leukaemia‐associated antigen, SSX2IP, is expressed during mitosis on the surface of myeloid leukaemia cells

2007; Wiley; Volume: 138; Issue: 5 Linguagem: Inglês

10.1111/j.1365-2141.2007.06706.x

1365-2141

Frances Denniss, Angela Breslin, Wendy Ingram, Nicola Hardwick, Ghulam J. Mufti, Barbara‐ann Guinn,

RNA Research and Splicing

The cancer-testis antigen SSX2 (Tureci et al, 1996) is a member of a highly conserved family of SSX genes which are only expressed in the testis and at very low levels in the thyroid of normal individuals. SSX2 is also expressed in various cancer types including approximately 50% of melanomas (Tureci et al, 1996). SSX2 interacting protein (SSX2IP) was identified through its interaction with SSX2 (de Bruijn et al, 2002); however, it is ubiquitously expressed in a range of normal tissues. Human SSX2IP has been shown to co-localize to the nucleus of HeLa cells transfected with SSX2, while SSX2IP fragments localize to the nucleus or cytoplasm depending on whether they included a putative nuclear localization signal (de Bruijn et al, 2002). In mice, SSX2IP (also known as afadin DIL domain-interacting protein) has been shown to strictly colocalize with afadin and nectin at cell–cell adherens junctions and in the perinuclear regions such as the Golgi complex (Asada et al, 2003), while in normal rat and Madin Darby canine kidney cells, SSX2IP localizes to the Golgi complex and nucleus (Asada et al, 2004). We have previously shown sera reactivity against SSX2IP through the immunoscreening of a testis cDNA library with pooled presentation acute myeloid leukaemia (AML) sera (Guinn et al, 2005) and subsequently showed that SSX2IP was preferentially recognized by AML patient, rather than normal donor sera. Reverse transcription polymerase chain reaction was used to show that SSX2IP was expressed in 33% of AML patient samples at presentation but not in normal donor peripheral blood and bone marrow samples. To examine the potential targeting of SSX2IP for immunotherapy, the present study examined SSX2IP protein expression in AML cells by immunocytochemistry (ICC). Four of seven myeloid leukaemia cell lines (K562, P39, VLB and KG1 but not NB4, HL60 and U937) (Fig 1A) expressed SSX2IP protein compared with only one of nine presentation primary AML samples. However, when 106 AML cells/ml were incubated in 10 ng/ml stem cell factor and 20 ng/ml interleukin 3 (IL-3) in X-VIVO 15TM for ≥ 3 days, to induce cell cycling, an increased frequency of SSX2IP protein expression was found in three of nine AML patient samples (Fig 1B). SSX2IP expression was not found in 18 normal donor haematopoietic cell samples (peripheral blood mononuclear cells or total white blood cells) in the presence or absence of cytokine treatment, or in three normal donor CD34+ samples purified from bone marrow or mobilized peripheral blood. Of note, SSX2IP expression was most apparent on the surface of myeloid leukaemia cells with condensed chromosomes (in mitosis) as seen by confocal microscopy (Fig 1C). We costained K562 cells, transduced with the HR'SINctwSV lentiviral vector containing CD80 cDNA (Chan et al, 2005), and showed that SSX2IP protein expression predominated on the cell surface, mimicking the cell surface localization pattern of CD80 (Fig 1D). SSX2IP was not found to colocalize with SSX2 or AF6 on K562 cells (data not shown). SSX2IP protein expression (SSX2IP) in cell lines and acute myeloid leukaemia (AML) patient samples. Immunocytochemistry on samples were air dried onto slides; SSX2IP expression (seen in red) was detected on the surface of (A) myeloid cells lines and (B) primary AML but not normal donor haematopoietic samples. SSX2IP antibody (Abcam, Cambridge, UK) was used at a dilution of 1/200 and all data are representative of ≥ 3 independent experiments. Confocal microscopy showed that (C) SSX2IP expression (stained green by fluorescein isothiocyanate) was detectable on the surface of K562 cells whose chromosomes had condensed (nuclei stained blue by 4′,6-diamidino-2-phenylindole [DAPI]), two frames shown; (D) Transduction of K562 cells with the CD80 lentiviral vector confirmed that, like CD80, SSX2IP was expressed on the cell surface. Overlays of SSX2IP and CD80 (stained red by PE) demonstrated that these two proteins did not colocalize; (E) A subpopulation of K562 cells (accounting for ≤ 16% of the total cell population) had an elevated level of SSX2IP as shown by flow cytometry. Isotype control shown in purple block, staining for SSX2IP surface expression (x-axis) with a green line, cell counts on the y-axis; (F) 0·3 mmol/l hydroxycarbamide for 72 h was used to block the cells at the G1/S interface of the cell cycle; x-axis, propidium iodide staining indicated DNA content, y-axis, cell count; (G) K562 cells were analysed hourly for 24 h following release back into the cell cycle; peak SSX2IP expression (seen by virtue of a red colouration) was detectable at the t = 14–17 h time points. To determine whether the low frequency of SSX2IP-positive cells observed by ICC and ≤ 16% expression detected by surface staining in flow cytometry (Fig 1E) was due to a cell cycle-related expression, we synchronized K562 cells using either serum starvation (105 cells/ml in 0·1% foetal calf serum for 5 days to cause a block at G0) or hydroxyurea (0·3 mmol/l for 3 days to block at the G1/S interface) (Fig 1F). On release back into the cell cycle, there was a 4-h period, post-synchronization, at 24–26 h for serum starvation and 14–17 h for hydroxycarbamide treatment (Fig 1G), during which time SSX2IP expression peaked. Although SSX2IP protein expression was observed in a limited number of AML patient samples at presentation and on slightly more samples following stimulation with cytokines, this probably reflects the low frequency of cycling blasts, and the even lower frequency of malignant cells in mitosis, in AML patient samples when cultured ex vivo (Chan et al, 2005). Here, we describe for the first time the increased surface expression of SSX2IP on AML cells during mitosis. The role SSX2IP plays on the surface of AML cells, and the reason for its upregulation during mitosis, remains unknown, although SSX2IP plays a role in cell–cell adhesion in rodents (Asada et al, 2003). However, in contrast to rodent cells, we found no colocalization between SSX2IP and afadin on K562 cells (data not shown). Therapeutically, the removal of SSX2IP-positive cells from AML patient stem cell harvests has the potential to reduce the tumour burden for patients whose graft fails or for patients who lack a suitable stem cell donor, although these possibilities require more thorough investigation. We would like to thank Drs Julie Richards, Steve Devereux, Shaun Thomas and Danny MacKay and Mr Steve Orr for technical advice. This study was supported by the Leukaemia Research Fund.