Selectively Frequent Expression of CXCR5 Enhances Resistance to Apoptosis in CD8 Þ CD34 Þ T Cells from Patients with T-Cell-Lineage Acute Lymphocytic Leukemia

Selectively frequent expression of CXCR5 enhances resistance to apoptosis in CD8 þ CD34 þ T cells from patients with T-cell-lineage acute lymphocytic leukemia

Zhang Qiuping1,7, Xiong Jie1,2,7, Jin Youxin2, Wu Qun1, Ju Wei1, Liu Chun1, Wang Jin1, Liu Yan1, Hu Chunsong3, Yang Mingzhen4, Gao Qingping5, Li Qun3, Zhang Kejian5, Sun Zhimin6, Liu Junyan1 and Tan Jinquan*,1,3

1Laboratory of Allergy and Clinical Immunology, Department of Immunology, Institute of Allergy and Immune-related Diseases, Centre for Medical Research, Wuhan University School of Medicine, Wuhan University, Dong Hu Road 115, Wuchang, Wuhan 430071, China; 2The State Key Laboratory of Molecular Biology, Institute of Biochemistry and Cell Biology, Shanghai Institutes for Biological Sciences, Chinese Academy of Science, Shanghai, China; 3Department of Immunology, College of Basic Medical Sciences, Anhui Medical University, Hefei 230032, China; 4Department of Hematology, The Afﬁliated University Hospital, Anhui Medical University, Hefei 230031, China; 5Department of Hematology, The First and Second Afﬁliated University Hospital, Wuhan University, Wuhan 430071, China; 6Department of Hematology, The Provincial Hospital of Anhui, Hefei 230020, China

We investigated CD4 þ CD34 þ , CD8 þ CD34 þ , advantages of CXCR5/CXCL13 for infiltration, resis- CD4 þ CD34À, and CD8 þ CD34À T cells from cord blood tance to apoptosis, and inappropriate proliferation. and from typical patients with T-cell-lineage acute Oncogene (2005) 24, 573–584. doi:10.1038/sj.onc.1208184 lymphocytic leukemia and T-cell-lineage chronic lympho- Published online 6 December 2004 cytic leukemia in terms of expression and functions of CXCR5/CXCL13. We found that CXCR5 was selectively Keywords: leukemia; T cells; chemokine receptor; apop- frequently expressed on T-cell-lineage acute (chronic) tosis; chemotaxis lymphocytic leukemia (T-ALL) CD8 þ CD34 þ T cells, but not on T-ALL CD4 þ CD34 þ , CD4 þ CD34À, and CD8 þ CD34À T cells. CXCR5 was rarely expressed on þ À all types of CD34 and CD34 CB or T-CLL T cells. Introduction CXCL13/B cells attractingchemokine 1 induced significant resistance to TNF-a-mediated apoptosis in T-ALL þ þ CXCR5 can be detected on mature B cells, small subsets CD8 CD34 T cells, instead of induction of chemotactic of normal CD4 þ and CD8 þ T cells, skin-derived DCs, and adhesive responsiveness. A proliferation-inducing þ þ follicular DCs, and also other stromal cells (Fo¨ rster ligand expression in T-ALL CD8 CD34 T cells was et al., 1994; Gunn et al., 1998; Ansel et al., 2000; Saeki upregulated by CXCL13/BCA-1 (B-cell attracting che- et al., 2000; Wu and Hwang, 2002; Yu et al., 2002). mokine 1). The CXCR5/CXCL13 pair by means of CXCR5 is essentially responsible for guiding B cells into activation of APRIL (A proliferation-inducingligand) þ þ the B-cell zones of secondary lymphoid organs (Fo¨ rster induced resistance to apoptosis in T-ALL CD8 CD34 T et al., 1996), as well as for T-cell migration (Fo¨ rster et al., cells in livin-dependent manner. In this process, cell–cell 1996; Okada et al., 2002). CXCR5, together with CCR7, contact in culture was necessary. Based on our findings, controls lymphocyte and DC homing to secondary we suggested that there were differential functions of lymphoid organs and functions lymphoid organ orga- CXCR5/CXCL13 in distinct types of cells. Normal nogenesis and organization (Reif et al., 2002; Ohl et al., lymphocytes, especially naı¨ ve B and T cells, utilized 2003). The CXCL13/B-cell attracting chemokine 1 CXCR5/CXCL13 for migration, homing, maturation, and (BCA-1), the only ligand for CXCR5 (Muller et al., cell homeostasis, as well as secondary lymphoid tissue 2003), is responsible for compartmental homing of organogenesis. Meanwhile, certain malignant cells took CXCR5-bearing B-lymphocytes and directing T-helper cells into the lymphoid follicle (Luther et al., 2000; Finke et al., 2002). CXCR5 is also highly expressed on *Correspondence: T Jinquan, Laboratory of Allergy and Clinical Immunology, Department of Immunology, Institute of Allergy and vascular endothelium in the primary central nervous Immune-related Diseases, Centre for Medical Research, Wuhan system lymphoma (Smith et al., 2003), and in the mantle University School of Medicine, Wuhan University, Dong Hu Road zone of all secondary lymphoid follicles in Hp-induced 115, Wuchang, Wuhan 430071, China; gastric mucosa-associated lymphoid tissue (Mazzucchel- E-mail: [email protected] li et al., 1999). However, the functional importance of 7These authors contributed equally to this work Received 16 February 2004; revised 3 September 2004; accepted CXCR5-CXCL13 in malignant T-cell trafficking, hom- 8 September 2004; published online 6 December 2004 ing, and survival is not fully understood. Apoptotic resistance via CXCR5/livin/APRIL pathway Z Qiuping et al 574 A common manifestation of T-cell-lineage acute and Results chronic lymphocytic leukemia (T-ALL and T-CLL) is infiltration of various organs by leukemic cells (Till et al., CXCR5 is selectively frequently expressed on T-ALL 2002). CCR9 is selectively and functionally overex- CD8 þ CD34 þ T cells pressed on CD4 þ T cells from patients with T-ALL and The percentages of CD4 þ CD34 þ , CD8 þ CD34 þ T cells T-CLL (Qiuping et al., 2003). Despite these findings, from patients with T-ALL were higher than that from little is known about the exact mechanisms and molecules regulate the homing, migration, inappropriate proliferation, and resistance to apoptosis of T-ALL and T-CLL T cells. The TNF family contains 16 members including TNF-a, lymphotoxin a, lymphotoxin b, CD40L, CD30L, CD27L, OX40L, 4-1BBL, FasL, TRAIL/APO-2 L, TL-1, TRANCE/RANKL, LIGHT, TWEAK, BLyS/BAFF, and a proliferation-inducing ligand (APRIL). APRIL is a rather new member of is family, named for its capacity to stimulate the proliferation of tumor cells in vitro (Hahne et al., 1998). APRIL transcript levels are low in normal tissues. In contrast, higher mRNA levels have been detected in tumor cell lines and in a variety of primary tumor tissues (Kelly et al., 2000). APRIL promotes tumor cell proliferation and resistance to apoptosis (Hahne et al., 1998; Kern et al., 2004). A glioblastoma cell line is resistant to apoptosis by transfection with APRIL (Roth et al., 2001). However, there are still contradictory findings with regard to its overall biological effects (Rennert et al., 2000). So far, eight human inhibitors of apoptosis proteins (IAPs) have been identified: c-IAP1, c-IAP2, NAIP, survivin, XIAP, Bruce, ILP-2, and livin. IAPs can block apoptosis mainly through their ability to bind and inhibit specific caspases such as caspase-3 and -7 (Chai et al., 2001; Riedl et al., 2001). A novel IAP member, designated livin/ML-IAP/KIAP (Lin et al., 2000; Kasof and Gomes, 2001), is suggested to mediate suppression of cell death (Sanna et al., 2002). A key function of livin suggests that this protein may be an important target for immune-mediated tumor destruction (Schmollinger et al., 2003). Except these points, little is known about the antiapoptotic effect of livin, and its regulatory mechanism.

Figure 1 CXCR5 distribution on CD4 þ CD34 þ , CD8 þ CD34 þ CD4 þ CD34À, or CD8 þ CD34À T cells. Triple-color ﬂow cytometric analysis of the distribution of CXCR5 on CD4 þ CD34 þ , CD8 þ CD34 þ CD4 þ CD34À, or CD8 þ CD34À T cells from T-ALL (a) and T-CLL (b) patients. The CD4 þ and CD8 þ T cells were freshly isolated and stained in triple colors of CD4 or CD8 (PE), CD34 (FITC), and CXCR5 (PerCP) as described in Materials and methods. The indicated numbers in the graphs were percentages of CD4 þ CD34 þ , CD8 þ CD34 þ CD4 þ CD34À, or CD8 þ CD34À T cells. The indicated percentages in the graphs were numbers of CXCR5 þ T cells. The data were from a single experiment, which was representative of 20 (T-ALL) and 15 (T-CLL) similar experiments performed. Isotype Ab controls were expressed as dished curves. CXCR5 expression on CD8 þ CD34 þ T cells was examined by real-time quantitative RT–PCR (c). The CD8 þ CD34 þ T cells were freshly isolated from CB of uncomplicated births, T-ALL, or T-CLL patients. In (c), a linear relationship between CT and log starting quantity of standard DNA template or target cDNA (CXCR5) was detected (data not shown). Bars show mean values7s.d. of eight similar experiments conducted

Oncogene Apoptotic resistance via CXCR5/livin/APRIL pathway Z Qiuping et al 575 patients with T-CLL (Figure 1). Interestingly, CXCR5 CXCL13 induces neither chemotaxis nor adhesion in was selectively frequently expressed on T-ALL T-ALL CD4 þ CD34 þ and CD8 þ CD34 þ T cells CD8 þ CD34 þ T cells (89%) (Figure 1a), whereas We studied the functional activity of CXCR5 on T-ALL significantly lower on T-ALL CD4 þ CD34 þ T cells CD4 þ CD34 þ and CD8 þ CD34 þ T cells. CXCL13 failed (14%). Both T-CLL CD4 þ CD34 þ and CD8 þ CD34 þ T to induce chemotaxis and adhesion in freshly isolated T- cells expressed low CXCR5 (16 and 19%, respectively) ALL CD4 þ CD34 þ and CD8 þ CD34 þ T cells, whereas (Figure 1b). T-ALL CD8 þ CD34À T cells expressed CXCL12, a ligand for CXCR4, had a significant marginally but significantly higher frequent CXCR5 chemotactic and adhesive effect on T-ALL (28%) than that on T-ALL CD4 þ CD34À T cells, and T- CD4 þ CD34 þ T cell and T-ALL CD8 þ CD34 þ T cells CLL CD4 þ CD34À and CD8 þ CD34À T cells (18, 17, and 10%, respectively) (Figure 1). CXCR5 was at a (Figure 2). CXCL13 also failed to induce chemotaxis and adhesion in T-ALL CD4 þ CD34À and CD8 þ CD34À rather low level on CD4 þ CD34 þ , CD8 þ CD34 þ , T cells (data not shown). Thus, the CXCR5-CXCL13 CD4 þ CD34À, or CD8 þ CD34À T cells from CB (Table 1). CXCR3, CXCR4, and CXCR6 expression pair had no ‘common’ chemotactic and adhesive functions in T-ALL CD8 þ CD34 þ T cells. were either in agreement with previous reports or there were no differences among the three types of cell sources (Table 1). CXCR5 mRNA was detected CXCL13 selectively rescues T-ALL CD8 þ CD34 þ T cells at very a low level in freshly isolated normal CB from TNF-a-mediated apoptosis CD4 þ CD34 þ and CD8 þ CD34 þ T cells. CXCR5 By knowing that activation of CCR9 led to phosphor- mRNA in T-ALL CD8 þ CD34 þ T cells was significantly ylation of GSK-3b and FKHR and provided a cell upregulated, whereas in T-ALL CD4 þ CD34 þ T cells survival signal (Youn et al., 2001), and that selective they had no significant difference from that in frequently CXCR5 expressed on T-ALL CD8 þ CD34 þ CD4 þ CD34 þ and CD8 þ CD34 þ T cells from normal T cells, which had no ‘common’ chemotactic and CB and T-CLL patients (Figure 1c). The same adhesive functions in the cells, we examined the patterns of CXCR5 mRNA and protein expression protective effects of CXCL13, CCL25, and CXCL12 in distinct cells were confirmed by Northern and on different types of cells on TNF-a-mediated apoptosis. Western blot (data not shown), as well as by immuno- The number of apoptotic and necrotic cells were fluorescence digital confocal microscopy (data not significantly decreased on culture of T-ALL shown). CD8 þ CD34 þ T cells in the presence of CXCL13

Table 1 CXC chemokine receptor expression on CD4+CD34+, CD8+CD34+, CD4+CD34À, or CD8+CD34À T cellsa Cell typeb CXCR3 CXCR4 CXCR5 CXCR6

CB CD4+CD34+ o1c 62711d 57232713 CD4+CD34À o155714 4732578 CD8+CD34+ o165718 774NDe CD8+CD34À o152710 673ND

T-ALL CD4+CD34+ 227569712 11731772 CD4+CD34À 187870717 8731574 CD8+CD34+ 157469715 81714 1276 CD8+CD34À 117758719 12771075

T-CLL CD4+CD34+ 21710 71716 1178ND CD4+CD34À 197866718 873ND CD8+CD34+ ND 65713 24710 ND + CD8 CD34À ND 68711 1277NDFigure 2 Chemotaxis and adhesion of CD4 þ CD34 þ and CD8 þ CD34 þ T cells. Chemotaxis (a) and adhesion (b) of freshly aCD4+ and CD8+ T cells were freshly isolated and stained in triple isolated CD8 þ CD34 þ T cells from T-ALL patients toward colors of CD4 or CD8 (PE), CD34 (FITC), and indicated chemokine CXCL13 (solid bars) or CXCL12 (gray bars) (100 ng/ml) or PBS receptor antibodies (PerCP) as described in Materials and methods. control (open bars) are shown. All results were determined as bCD4+ or CD8+ T cells were isolated from CB of uncomplicated described in Materials and methods and expressed as C.I. or births, T-ALL, or T-CLL patients. co1 was under detectable level. percentage of adhesive cells with standard deviation (7s.d.), and dListed numbers in the table were percentages of CXC chemokine based on triplicate determination of chemotaxis and adhesion on receptor-positive T cells. The data were mean values7s.d. of at least each concentration of chemokine indicated as ng/ml (x-axis). The eight similar experiments conducted. For detection of CXCR5, shown results were sums of eight experiments conducted. Statistical data were mean values7s.d. of 20 T-ALL and 15 T-CLL patients. signiﬁcant differences as compared to controls are indicated eND, no determination *Po0.001. Values P>0.05 are considered nonsigniﬁcant

Oncogene Apoptotic resistance via CXCR5/livin/APRIL pathway Z Qiuping et al 576 (Figure 3Al), in comparison with that in T-ALL CD95L, and TRAIL were not changed (Figure 4a). CD4 þ CD34 þ T cells (Figure 3Ak), and with that in APRIL mRNA was expressed at a very low level in normal CB CD4 þ CD34 þ and CD8 þ CD34 þ T cells freshly isolated T-ALL CD8 þ CD34 þ T cells (Figure 4b). (Figure 3Ai and j). CCL25 selectively protected T-ALL It was significantly altered within 8 h after stimulation CD4 þ CD34 þ T cells from TNF-a-mediated apoptotic with CXCL13. The peak of expression appeared at 24 h. response (Figure 3Ao), but protected neither T-ALL The same patterns of APRIL mRNA expression were CD8 þ CD34 þ T cells nor normal CB CD4 þ CD34 þ and confirmed in T-ALL CD8 þ CD34 þ T cells by Northern CD8 þ CD34 þ T cells (Figure 3Am, n, and p). CXCL12 blot (Figure 4c). Elevated APRIL protein expressions could rescue malignant cells (Figure 3Ag and h), as well were also observed in T-ALL CD8 þ CD34 þ T cells by as normal CB CD4 þ CD34 þ and CD8 þ CD34 þ T cells Western blot (Figure 4d), confirming that CXCL13 from TNF-a-mediated apoptosis (Figure 3Ae and f), in selectively increased one member of the TNF family comparison with that in the untreated cells (Figure 3Aa, b, c, and d). Ab against CXCR5 could completely block the protective effect of CXCL13, indicating that rescuing effect was indeed induced by means of interaction of CXCL13 and CXCR5 (data shown in Figure 5). The observation was also confirmed in T-CLL CD4 þ CD34 þ and CD8 þ CD34 þ T cells (data not shown). As shown in Figure 3B, the total fractions of dead cells (including apoptotic and necrotic) in different types of the cells tested were the same patterns as the results in Figure 3A. These results could confirm the count effects of CXCR5 on death signaling of TNF-a in T-ALL CD8 þ CD34 þ T cells. We also measured CXCR4 ligand CXCL12 and CXCR5 ligand CXCL13 expression (Table 2). CXCL12 and CXCL13 were expressed in CD8 þ CD34 þ and CD8 þ CD34À T-ALL cells. There was no difference in CXCL12 and CXCL13 expression between CB and T- ALL T cells.

APRIL expression in T-ALL CD8 þ CD34 þ T cells is selectively increased by CXCL13 We examined the expression levels of several members of TNF family on T-ALL CD8 þ CD34 þ T cells during the stimulation with CXCL13. Freshly isolated T-ALL CD8 þ CD34 þ T cells expressed low levels of APRIL (11%) (Figure 4a). After a 24 h culture with CXCL13, the levels of APRIL expression could be signiﬁcantly altered by 78%. Interestingly, level of the APRIL expression on normal CB CD8 þ CD34 þ T cells was unchanged after CXCL13 stimulation. Other examined members of the TNF family, for example, CD95,

Figure 3 Analysis of apoptotic and total dead (necrotic and apoptotic) cells in CD4 þ CD34 þ and CD8 þ CD34 þ T cells. Flow cytometric analysis of apoptotic (A) and total dead cells (B). The CD8 þ CD34 þ T cells were freshly isolated from CB of uncomplicated births or T-ALL patients as described in Materials and methods, and were pretreated in the absence or presence of chemokine as indicated (all at 100 ng/ml) for 24 h at 371C, following stimulation with TNF-a (100 ng/ml) for 24 h at 371C. The cells were analysed by flow cytometry for PI- (y-axis) and FITC-conjugated annexin V (x-axis) as described in Materials and methods. The gating in the forward scatter and side scatter histograms was adhered to the lymphocyte region. The percentages of PIÀ annexin V þ cells and PI þ annexin V þ cells were indicated in the figure. The data (A) were from a single experiment, which was representative of six experiments performed. The data for total dead cells (PIÀ annexin V þ þ PI þ annexin V þ )(B) were mean values7s.d. of six experiments performed. Statistically significant differences as compared with untreated cells were indicated (*Po0.001)

Oncogene Apoptotic resistance via CXCR5/livin/APRIL pathway Z Qiuping et al 577 Table 2 CXCL12 and CXCL13 expression on CD8+CD34+ and APRIL expression on T-ALL CD8 þ CD34 þ T cells. CD8+CD34À T cells TWE-PRIL, TWEAK–APRIL fusion protein on the cell Cellsa CB T-ALL surface (Pradet-Balade et al., 2002), was expressed at a very low level in freshly isolated cells (Figure 4e). TWE- CD8+CD34+ CD8+CD34À CD8+CD34+ CD8+CD34À PRIL mRNA expression was significantly increased CXCL12 8.370.5b 7.470.6 7.670.8 7.970.4 within 16 h after stimulation with CXCL13. The peak of CXCL13 4.470.6 5.570.6 4.370.5 4.870.3 expression appeared at 24 h. The same patterns of TWE- PRIL mRNA expression were seen in T-ALL aCD8+CD34+ and CD8+CD34À T cells from CB or T-ALL were CD8 þ CD34 þ T cells by Northern blot (Figure 4f). prepared as described in Materials and methods. bThe data shown were mean values7s.d. of six experiments conducted in each group. The procedure for real-time quantitative RT–PCR amplification was CXCL13 via APRIL upregulation induces resistance to described in Materials and methods ( Â 103 mRNA copies/50 ng apoptosis in T-ALL CD8 þ CD34 þ T cells cDNA) An a-CXCR5Ab significantly inhibited apoptotic protection effect in T-ALL CD8 þ CD34 þ T cells by

Figure 4 CD95, CD95L, TRAIL, APRIL, and TWE-PRIL expression on CD8 þ CD34 þ T cells. Membrane TNF family member expression on CD8 þ CD34 þ T cells was examined by flow cytometry (a), real-time quantitative RT–PCR (b, e), Northern blot (c, f), and Western blot (d). The CD8 þ CD34 þ T cells were purified from T-ALL patients or CB of uncomplicated births (E) as described in Materials and methods, and were either freshly purified or treated in the presence of CXCL13 (at 100 ng/ml) for 24 h (or time intervals as indicated) at 371C. In (a), the cells were then stained with PE-labeled CD8mAb and appropriate Ab indicated as described in Materials and methods. The indicated numbers in the graphs were percentages of double-positive cells of CD8 and TNF family member protein as indicated. The data are from a single experiment, which is representative of six similar experiments performed. Isotype Ab controls were for a-APRIL, while other isotype Ab controls were not shown. In (b and e), the procedure for quantitative RT–PCR amplification was described in Materials and methods. A linear relationship between CT and log starting quantity of standard DNA template or target cDNA (APRIL and TWE-PRIL) was detected (data not shown). Bars show mean values7s.d. of six experiments conducted. In (c) and (f), the detection of mRNAs of APRIL and TWE-PRIL by Northern blot for CD8 þ CD34 þ T cells from T-ALL patients either freshly isolated or treated in the presence of CXCL13 (at 100 ng/ml) for time intervals as indicated. Total RNA from different cells were isolated, electrophoresed, and blotted as described in Materials and methods. The hybridization signals for APRIL or TWE-PRIL mRNA from different cells are shown in the upper panel. The 28S rRNAs in the lower panel confirm that comparable amounts of total RNA were used. In (d), the APRIL protein was examined using Western blot analyses. The cells, from T- ALL patients as described above, were lysed and total protein content electrophoresed and blotted as described in Materials and methods. Actins in the lower panel indicated the quantity of total cellular protein from the tested samples loaded in each lane. Arrows indicated markers used to verify equivalent molecular weights of appropriate proteins in each lane

Oncogene Apoptotic resistance via CXCR5/livin/APRIL pathway Z Qiuping et al 578 CXCL13 (Figure 5Ab). An Ab against APRIL was used Figure 5B, the total fractions of dead cells (including to block APRIL after the cell stimulating with CXCL13, apoptotic and necrotic) in different experimental condi- which could significantly increase APRIL expression. tions were the same patterns as the results in Figure 5A. The Ab against APRIL could obviously inhibit apopto- These results could confirm the importance of APRIL tic protection effect in T-ALL CD8 þ CD34 þ T cells by expression for induction of resistance to apoptosis in T- CXCL13 (Figure 5Ac), compared with that in cell ALL CD8 þ CD34 þ T cells by CXCL13. Recombinant culture in the absence of Ab (Figure 5Aa), suggesting human APRIL was added to TNF-a- and CXCL13- that APRIL expression was crucially important for treated T-ALL CD8 þ CD34 þ in the absence or presence induction of resistance to apoptosis in T-ALL of mAb against CXCR5 (Table 3). Higher concentra- CD8 þ CD34 þ T cells by CXCL13. As shown in tions of APRIL (0.1–1 mg/ml) could in vitro indeed rescue T-ALL CD8 þ CD34 þ T cells from TNF-a- induced apoptosis in a dose-dependent manner. Caspase-3 and -8 expression was essential for this type of cell death (Thornberry and Lazebnik, 1998). Expres- sion levels of activated caspase-3 (Figure 6a) and caspase-8 (Figure 6b) were measured by intracellular staining of T-ALL CD8 þ CD34 þ T cells within the distinct cell cultures. In freshly isolated T-ALL CD8 þ CD34 þ T cells, both cleaved caspase-3 and -8 were at very low levels (11 and 15%, respectively). After stimulation with TNF-a, cleaved caspase-3 and -8 were significantly upregulated (44 and 54%, respectively). In the presence of CXCL13 in cell culture, the levels of cleaved caspase-3 and -8 in T-ALL CD8 þ CD34 þ T cells were stabilized. Ab against CXCR5 as well as Ab against APRIL could separately block the stabilizing effects of CXCL13 on caspase-3 and -8 during TNF-a- mediated apoptosis. The results of Western blotting also confirmed the observation (Figure 6c). Thus, the observation suggested that CXCL13, by means of activation of frequently expressed CXCR5-induced membrane APRIL expression, subsequently stabilized caspase-3 and -8 in T-ALL CD8 þ CD34 þ T cells, and rescued the cells from TNF-a-mediated apoptosis. The inhibitor of the apoptosis family of proteins was found to protect against the broadest spectrum of apoptotic signals (Duckett et al., 1996; Miller, 1999). Western blot showed that the protein levels of antiapoptotic members Bcl-2, Bcl-X, and c-FLIPL in the cells under the distinct experimental conditions were identical to that in freshly isolated T-ALL CD8 þ CD34 þ T cells (Figure 6d). Interestingly, after apoptotic induction with TNF-a and stimulation with CXCL13, the expression level of one antiapoptotic member Figure 5 Analysis of apoptotic and total dead (necrotic and protein in IAP family livin was significantly increased apoptotic) cells in T-CLL CD8 þ CD34 þ T cells. Flow cytometric analysis of apoptotic (A) and total dead cells (B), the CD8 þ CD34 þ T cells were isolated from T-ALL patients as described in Materials and methods, and were pretreated in the absence or presence of Table 3 Effects of recombinant human APRIL on rescuing T-ALL Ab(s) (or isotypes) as indicated (all at 5 mg/ml) for 6 h at room CD8+CD34+ T cells from TNF-a-induced apoptosis temperature, following subsequent stimulations in the presence of CXCL13 (at 100 ng/ml) for 24 h, and stimulation with TNF-a Treatmenta rh APRIL (100 ng/ml) for 24 h at 371C. APRIL Ab blocking was after CXCL13 stimulation. AbÀ, without antibody block. Abs, blocking 0 0.01 mg/ml 0.1 mg/ml 1 mg/ml with both CXCR5 and APRIL antibodies. The cells were analysed by flow cytometry for PI- (y-axis) and FITC-conjugated annexin V a-CXCR5 (À)1673b 15751372874 (x-axis) as described in Materials and methods. The gating in the a-CXCR5 (+) 63718 51714 28710 1175 forward scatter and side scatter histograms was adhered to the lymphocyte region. The percentages of PIÀ annexin V þ cells and aT-ALL CD8+CD34+ T cells were pretreated CXCL13 (100 ng/ml) PI þ annexin V þ cells were indicated in the figure. The data (A) in the absence (À) or presence (+) of mAb against CXCR5 (5 mg/ml) were from a single experiment, which was representative of five as described in Materials and methods, followingstimulation with experiments performed. The data for total dead cells (PIÀ annexin TNF-a (100 ng/ml) for 24 h at 371C before apoptotic assay. bData for V þ þ PI þ annexin V þ )(B) were mean values7s.d. of five apoptotic cells were mean values7s.d. of six experiments performed in experiments performed. Statistically significant differences as each group. Recombinant human APRIL was obtained as described compared with absence of Ab (AbÀ) were indicated (*Po0.001) previously (Yu et al., 2000)

Oncogene Apoptotic resistance via CXCR5/livin/APRIL pathway Z Qiuping et al 579

þ þ Figure 6 Activation of caspases and occurrence of Bcl-2, Bcl-X, c-FLIPL, and IAP expression in T-CLL CD8 CD34 T cells. Flow cytometric analysis of active caspase-3 (a) and caspase-8 (b), and Western blot assay of caspase-3 and -8 (c), Bcl-2, Bcl-X, and c-FLIPL, XIAP, c-IAP1, c-IAP2, survivin and livin (b) in T-ALL CD8 þ CD34 þ T cells. The CD8 þ CD34 þ T cells were isolated from T-ALL patients as described in Materials and methods, and were pretreated in the absence or presence of Ab ( þ ) (or isotypes) as indicated (all at 5 mg/ml) for 6 h at room temperature, following subsequent stimulations in the presence of CXCL13 (at 100 ng/ml) for 24 h, and stimulation with TNF-a (100 ng/ml) for 24 h at 371C. APRIL Ab blocking was after CXCL13 stimulation. They were then permeabilized and fixed as indicated in Materials and methods, and subsequently stained for intracellular activated (cleaved) caspase-3 or -8. Activated caspase-3- or -8-specific fluorescence intensity was measured by flow cytometry. The indicating percentages of cells with activated caspase-3 or -8 were quantitated from relative-frequency histograms. Isotype Ab controls were expressed as dished curves. In (c and d), The caspase-3 (upper panel) or caspase-8 (lower panel), Bcl-2, Bcl-X, and c-FLIPL, XIAP, c-IAP1, c-IAP2, survivin and livin proteins were examined using Western blot analyses. The cells, from different subjects as described above, were lysed and total protein content electrophoresed and blotted as described in Materials and methods. Actins in lower pictures indicated the quantity of total cellular protein from the tested samples loaded in each lane. Arrows indicated markers used to verify equivalent molecular weights of appropriate proteins in each lane

in T-ALL CD8 þ CD34 þ T cells (Figure 6d), in CXCR5-CXCL13 increases resistance to apoptosis in comparison with that in freshly isolated cells. Ab T-ALL CD8 þ CD34 þ T cells in a livin-dependent manner against CXCR5 as well as Ab against APRIL could separately and signiﬁcantly inhibit the elevation of The 293T cells were transfected with plasmids encoding expression level of livin in these cells (Figure 6d). Other CXCR5 and different combinations among wt livin, wt antiapoptotic members in the IAP family XIAP, c-IAP1, APRIL, mutant livin, and mutant APRIL as indicating c-IAP2, and survivin expressed identical levels in the in Figure 7. Only the cells transfecting both wild-type different conditions in T-ALL CD8 þ CD34 þ T cells livin and wild-type APRIL could be rescued by (Figure 6d). Thus livin was involved in the events of CXCL13 from TNF-a-induced apoptosis (Figure 7a). CXCR5-CXCL13-induced resistance to TNF-a- Expression levels of activated caspase-3 and -8 were mediated apoptosis in T-ALL CD8 þ CD34 þ T cells. measured by intracellular staining of both wild-type

Oncogene Apoptotic resistance via CXCR5/livin/APRIL pathway Z Qiuping et al 580 Table 4 CXCL13 requires livin and APRIL for protection against apoptosis in cell–cell contact mannera Cell treatmentd Mixingb Transwellc

APRIL+livin+ and APRIL+- livin+ livin+ (1 : 1) livin+

CXCL13À 52718e 62715 60714 CXCL13+ 11710f,*1277* 59719

aMCF7 cells were cotransfected with vectors encoding CXCR5 (400 ng) and either wt APRIL and wt livin or wt livin alone (400 ng each). bAPRIL+livin+ and livin+ cells were mixed in ratio 1 : 1 in culture during the procedure. cTranswell experiments were carried out in 24-well plates as described previously (Jonuleit et al., 2000). APRIL+livin+ and livin+ cells were separately placed in transwell chambers (Millicell, 0.4 mm; Millipore) in culture during the procedure. dCells were pretreated in the absence (À) or presence (+) of CXCL13 (100 ng/ml) described in Materials and methods, following stimulation with TNF-a (100 ng/ml) for 24 h at 371C before the other assay as indicated. The analysis of total dead (necrotic and apoptotic) cells. eCells were analysed by ﬂow cytometry as described in the legend for Figure 3. The data were mean values from six experiments performed. fStatistically signiﬁcant differences were compared between the absence and presence of CXCL13 (*Po0.001)

livin and APRIL, caspases could be stabilized (upper panel, Figure 7b), but not in cells transfected with wild- type APRIL and mutant livin (lower panel, Figure 7b), and vice versa (data not shown). We observed similar results in other cells transfected with mutant livin or mutant APRIL alone (data not shown). The results mentioned above strongly suggested the necessity of both livin and APRIL in CXCL13 rescuing cells from TNF-a-mediated apoptosis. MCF7 cells were cotransfected with vectors encoding CXCR5 and either wt APRIL and wt livin, or wt livin alone. APRIL þ livin þ and livin þ cells were mixed or separately placed in transwell chambers in culture Figure 7 CXCL13 requires livin and APRIL for protection during the procedure. As show in Table 4, the cells in against apoptosis. HEK 293T cells were cotransfected with vectors the mixed culture (APRIL þ livin þ : livin þ ¼ 1 : 1) and encoding CXCR5 (400 ng) and different combinations of wt livin, cotransfected cell culture (APRIL þ livin þ cells alone) wt APRIL, mutant livin, and mutant APRIL (400 ng each) as were equally rescued in the presence of CXCL13, indicated. The amount of transfected cDNA was kept constant in þ each sample by adding control pcDNA3 vector. Cells were whereas livin cells alone could not be rescued in the pretreated in the absence or presence of CXCL13 (100 ng/ml) presence of CXCL13. This observation strongly sug- described in Materials and methods; some of the cells were gested that upregulated APRIL could link to binding stimulated with TNF-a (100 ng/ml) for 24 h at 371C before the sites (presumably be APRIL receptors) (Yu et al., 2000) other assay as indicated. (a) Analysis of total dead (necrotic and apoptotic) cells. The cells were analysed by flow cytometry as on the neighboring cells in the culture in cell–cell contact described in the legend for Figure 4. The data were mean values manner to selectively activate livin, and consequently to from six experiments performed. Statistically significant difference rescue the cells. In this process, APRIL receptors might as compared with vector transfected only was indicated also directly activate livin. (*Po0.001). (b) Flow cytometric analysis of active caspase-3 and -8 in transfected cells as indicated. Cells were pretreated as mentioned above. They were then permeabilized and fixed as indicated in Materials and methods, and subsequently stained for intracellular activated (cleaved) caspase-3 or -8. Activated caspase- Discussion 3- or -8-specific fluorescence intensity was measured by flow cytometry. The percentages of cells with activated caspase-3 or -8 CXCR5 and CXCL13 play a crucial role in lymphocyte were quantitated from relative-frequency histograms. Isotype Ab controls were expressed as dished curves traffic within secondary lymphoid tissues, and are involved in the formation of the B-cell compartment. Mice lacking either CXCR5 or BLC (the murine APRIL- and wild-type livin-transfected cells that as well homolog of human BCA-1) have gross aberrations in as wild-type APRIL- and mutant livin-transfected were the follicular architecture of the spleen and reduced measured. Only in cells transfected with both wild-type numbers of Peyer’s patches and lymph nodes. CXCR5

Oncogene Apoptotic resistance via CXCR5/livin/APRIL pathway Z Qiuping et al 581 and CXCL13 are not strictly obligatory for cellular and understanding how apoptosis is altered in leukemia humoral immune responses (Voigt et al., 2000). CXCR5 cells and how it can be modulated to overcome is expressed uniformly on B cells, except plasma cells resistance and improve clinical outcomes. In the present (Hargreaves et al., 2001). CXCR5 is expressed on a study, with sufficient and direct data we have managed subset of peripheral blood and tonsillar memory to show that CXCL13 selectively increases APRIL CD4 þ T cells, but is absent from naı¨ ve CD4 þ T cells, expression in T-ALL CD8 þ CD34 þ T cells to induce and Th1 and Th2 cells (Breitfeld et al., 2000; Kim et al., resistance to apoptosis in a livin-dependent manner. 2001). In the present study, we have investigated four Livin is a potent antiapoptotic protein and not different types of cells, for example, CD4 þ CD34 þ , detectable in most normal adult tissues. Elevated CD8 þ CD34 þ ,CD4þ CD34À, and CD8 þ CD34À T cells expression of livin renders melanoma cells resistant to from normal CB, T-ALL, and T-CLL patients in terms apoptotic stimuli and potentially contributes to the of expression and functions of CXCR5. We have found pathogenesis of this malignancy (Vucic et al., 2000). Our that CXCR5 is selectively frequently expressed on T- findings, together with others, indicate that APRIL and ALL CD8 þ CD34 þ T cells. CXCL13 selectively induces livin may be some critical cellular factors whose resistance to apoptosis in T-ALL CD8 þ CD34 þ T cells. increased expression confers resistance to apoptotic Chemokine ligand-receptor signaling is able to provide stimuli, thereby contributing to the pathogenesis and antiapoptotic activity to hematopoietic cells (Han et al., progression of malignant cells such as melanoma cells, 1997; Scheuerer et al., 2000; Youn et al., 2001, 2002). and T-ALL CD8 þ CD34 þ T cells. However, there are some controversial even contra- The TWE-PRIL is a functional membrane-bound dictory reports, such as CXCR4 induces programmed TWEAK–APRIL fusion protein and shares the same cell death of human peripheral CD4 þ T cells, malignant receptor-binding domain with APRIL (Rennert et al., T cells, and CD4/CXCR4 transfectants (Berndt et al., 2000; Pradet-Balade et al., 2002). In the present study, 1998). The interaction between HIV R5 Env and CCR5 we have observed that only higher concentrations of activates the Fas pathway and caspase-8 as well as APRIL can in vitro rescue cells from apoptosis (Table 3). triggering FasL production, ultimately causing CD4 þ T- CXCL13 rescues APRIL- and livin-cotransfected cells in cell death (Algeciras-Schimnich et al., 2002). CCR3 cell–cell contact manner (Table 4). Taking into account expression induced by IL-2 and IL-4 functions as a the same expression patterns of TWE-PRIL as APRIL death receptor for B cells (Jinquan et al., 2003). The after stimulation with CXCL13 (Figure 4f), we presume results in this study, together with other observations, that CXCL13 in real life needs functional synergy of suggest that normal lymphocytes utilize CXCR5/ APRIL and TWE-PRIL to rescue cells from apoptosis. CXCL13 for migration, homing, development, matura- The exact mechanism should be subjected to further tion, selection, and cell homeostasis, as well as investigation. secondary lymphoid tissue organogenesis. Meanwhile, some malignant cells, particularly T-ALL CD8 þ CD34 þ T cells, take advantages of CXCR5/CXCL13 for Materials and methods infiltration, resistance to apoptosis, and inappropriate proliferation. To our knowledge, this study is the first Patients and cell purification report on differential functions of CXCR5/CXCL13 in All patients with T-ALL or T-CLL were diagnosed according distinct types of cells in terms of induction of apoptotic to The French–American–British (FAB) Cooperative Group resistance, and is the direct evidence of the pathophy- criteria (Bennett et al., 1981) or to the guidelines of the siological activity of T-ALL CD8 þ CD34 þ T cells National Cancer Institute Working Group on B-CLL (Bennett induced by CXCL13. et al., 1989; Cheson et al., 1996). All patients gave informed APRIL-transfected cells and addition of APRIL to consent according to institutional guidelines. CD4 þ or cells increase their proliferation and in vitro growth rate, CD8 þ CD34 þ cells were purified PBMCs as described else- and enhance tumor growth rate in nude mice (Hahne where (Jinquan et al., 2000; Qiuping et al., 2003). After et al., 1998). B-cell maturation antigen (BCMA), an enrichment of PBMCs by LymphPrept gradient, CD4 þ ,or þ þ APRIL receptor, blocks the growth of APRIL-expres- CD8 CD34 , double-positive cells were sorted using a FACSstarPlus and FITC- or PE-conjugated antibodies (BD sing cell lines in nude mice (Rennert et al., 2000). T cells Pharmingen) against CD4, CD8, and CD34 from cord blood from APRIL transgenic (Tg) mice show increased (CB) of uncomplicated births or patients with T-ALL or T- proliferation and elevated IL-2 production (Stein et al., CLL. The malignancy of purified T-ALL or T-CLL CD4 þ or 2002), suggesting APRIL as an in vitro T-cell stimulator CD8 þ CD34 þ cells was checked by expression of CD25, (Yu et al., 2000; Siegel and Lenardo, 2001). A highly CD45RO, and HLA-DR (Qiuping et al., 2003). The cell lines significant increase in vitro survival of Tg APRIL- were human embryonic kidney 293T cells and MCF7-Fas cells expressing T cells is also observed (Medema et al., 2003). (from the American Type Culture Collection, Manassas, VA, A genetic change that leads to blocking apoptosis USA). The a-CXCR5mAb and chemokines (CXCL13, may allow a cell to acquire further mutations, survive CCL25, and CXCL12) were purchased from R&D Systems, inappropriately, and eventually become malignant. Abingdon, UK. Additionally, a defect in the inherent ability of a cell to undergo apoptosis may account for much of the Flow cytometry resistance to different therapies observed in leu- For detection of the chemokine receptor, the cells were triple- kemic cells. Therefore, much effort has gone into stained PE-labeled CD4 or CD8, FITC-labeled CD34

Oncogene Apoptotic resistance via CXCR5/livin/APRIL pathway Z Qiuping et al 582 (DAKO, Denmark), and PerCP-labeled chemokine receptor Pharmacia Biotech, Little Chalfont, UK). All Abs (APRIL, antibody (R&D Systems, Abingdon, UK) at 5 mg/ml in PBS Bcl-2, Bcl-X, c-FLIPL, c-IAP1, c-IAP2, XIAP, and survivin) for 20 min, followed washing twice in staining buffer (Qiuping were from Santa Cruz Biotech Inc., Santa Cruz, CA, USA, et al., 2003). For detection of apoptosis, cells were stained except for a-livin mAb from Imgenex Corp., Sorrento Valley, in staining medium with 1 mg/ml propidium iodide (PI) San Diego, USA, and a-b-actin mAb from Sigma Chemical for 30 min at 41C, then stained with FITC-conjugated annexin Co. V with binding buffer (BD Pharmingen) (Youn et al., 2001; Jinquan et al., 2003). For detection of intracellular active Chemotaxis assay caspases, cytofix/cytoperm buffer (BD Pharmingen) was used to permeabilize cells, and cells were subsequently stained The chemotaxis assay was performed in a 48-well micro- with anti-active-capsase-3 or anti-active-caspase-8mAb chamber (Neuro Probe, Bethesda, MD, USA) technique (BD Pharmingen). The measurements were performed with (Qiuping et al., 2003). Briefly, chemokines in RPMI 1640 with a flow cytometer (COULTERs XL, Coulter Corpo- 0.5% BSA were placed in the lower wells (25 ml). A measure of 6 ration, Miami, FL, USA). Data were analysed by the 25 ml cell suspension (2 Â 10 cells/ml) was added to the upper WinList program (The Scripps Research Institute, La Jolla, well of the chamber, which was separated from the lower well CA, USA). by a 5 mm pore size, polycarbonate, polyvinylpyrolidone-free membrane (Nucleopore, Pleasanton, CA, USA). The chamber was incubated for 60 min at 371C and 5% CO2. The membrane Real-time quantitative reverse transcription (RT)–PCR assay was then carefully removed, fixed in 70% methanol, and All real-time quantitative RT–PCR reactions were performed stained for 5 min in 1% Coomassie brilliant blue. The as described elsewhere (Kruse et al., 1997). Briefly, total RNA migrating cells were counted using light microscopy. Approxi- from purified cells (1 Â 105, purity >99%) was prepared by mately 6% of the cells will migrate spontaneously (known as using Quick Preps total RNA extraction kit (Pharmacia MCNC). The results were expressed as chemotactic index Biotech, USA). RNA was reverse transcribed by using (C.I.) with standard deviation (s.d.). oligo(dT)12–18 and Superscript II reverse transcriptase (Life Technologies, Grand Island, USA). The real-time quantitative Adhesion assays s PCR was performed with ABI PRISM 7700 Sequence As described previously (Qiuping et al., 2003), 96-well plates Detector Systems (Applied Biosystems, Foster City, CA, s were coated with laminin (20 mg/ml, Sigma Chemical Co.) USA). By using SYBR Green PCR Core Reagents Kit, in PBS for 1 h at 371C. Plates were washed with PBS fluorescence signals were generated during each PCR cycle via 0 0 and incubated with medium containing 0.2% BSA for 1 h to the 5 to 3 endonuclease activity of AmpliTaq Gold to provide block nonspecific adhesion. The single-cell suspensions (4 Â 105 real-time quantitative PCR information. The sequences of the cells/ml) with 0.2% BSA were added to the appropriate specific primers are as follows: CXCR5 sense, 50-GGTC 0 0 chemokine. The cell suspension was added 100 ml/well in TTCATCTTGCCCTTTG-3 ; CXCR5 antisense, 5 -ATGC 1 0 0 triplicate to the plates, and incubated for 60 min at 37 C. The GTTTCTGCTTGGTTCT-3 ; APRIL sense, 5 -CCAGCCT- wells were then washed with 0.2% BSA in PBS, followed by CATCTCCTTTCTTGC-30; APRIL antisense, 50-TCACAG 0 0 careful aspiration. Subsequently, the adherent cells were fixed TTTCACAAACCCCAGG-3 ; TWE-PRIL sense, 5 -GCCTT with 1% formaldehyde and stained with 1% crystal violet. TCCTGAACCGACTAGTTC-30; TWE-PRIL antisense, 50-T 0 0 Crystal violet was then extracted by the addition of a 1 : 1 CAGTGGGCCCGACCCGAGATGT-3 ; CXCL12 sense, 5 - M 0 mixture of sodium citrate (0.1 ) and ethanol (pH 4.2). The ACACTCCAAACTGTGCCCTTCA-3 ; CXCL12 antisense, absorbency was then read at 540 nm. Background cell 50-CCACGTCTTTGCCCTTTCATC-30; CXCL13 sense, 50-T 0 adhesion to 2% BSA-coated wells was subtracted from all CTCTGCTTCTCATGCTGCTG-3 ; and CXCL13 antisense, readings. 50-TTCGATCAATGAAGCGTCTAGG-30. All unknown cDNAs were diluted to contain equal amounts Plasmids and cell transfection of b-actin cDNA. PCR retaining conditions were 2 min at 501C, 10 min at 951C, 40 cycles with 15 s at 951C, and 60 s at Plasmids encoding CXCR5, livin, and APRIL and the 601C for amplifications. catalytically inactive mutants livin and APRIL (RKRR motif deletion) used in this study have been described previously (Kanbe et al., 1999; Rennert et al., 2000; Vucic et al., 2000; Northern and Western blot assays Lopez-Fraga et al., 2001). The cells were transiently trans- For mRNA detection (Northern blot), as previously described fected with vectors encoding target genes as described (Sica et al., 1997), 5 mg of total RNA was electrophoresed elsewhere. Briefly, the cells were cultured with DMEM under denaturing conditions, followed by blotting onto Nytran containing 10% FCS, penicillin, and streptomycin. Cells were membranes, and crosslinked by UV irradiation. CXCR5 grown to approximately 70% confluence in 60-mm dishes for cDNA probes were obtained by PCR amplification of the 24 h before transfection. The DNA construct of expression sequence mentioned above from total RNA from PBMC from vectors (0.4 mg) or vector only was mixed with 12 mlof normal adults. The membranes were hybridized overnight with LipofectAMINE (Gibco BRL) in 2 ml of opti-DMEM 1 Â 106 c.p.m./ml of 32P-labeled probe, followed by intensive serum-free medium and added to cells, and incubated for washing with 0.2 Â SSC and 0.1% SDS before being auto- 6 h. The cells were further cultured in 2.5 ml of DMEM radiographed. For protein detection (Western blot), the cells containing 10% FCS in 5% CO2. were lysed in lysis buffer (Massari et al., 2000). Cell lysis was performed for 30 min at 41C with lysis buffer. Lysates were centrifuged at 10 000 r.p.m. for 5 min at 41C. Protein (around 40 mg) was loaded onto 16% SDS–PAGE, transferred onto Abbreviations PVDF membranes after electrophoresis, and incubated with APRIL, A proliferation-inducing ligand; BCA-1, B-cell the appropriate Abs at 0.5 mg/ml. Analyses were conducted attracting chemokine 1; BCMA, B-cell maturation antigen; using enhanced chemiluminescence detection (Amersham C.I., chemotactic index; CXCL (CC), CXC (CC) chemokine

Oncogene Apoptotic resistance via CXCR5/livin/APRIL pathway Z Qiuping et al 583 ligand; CXCR (CC), CXC (CC) chemokine receptor; T-ALL Province, China (no. 98436630), and Education and Research (T-CLL), T-cell-lineage acute (chronic) lymphocytic leukemia. Foundation of Anhui Province, China (no. 98JL063), and Research Foundation from Health Department of Hubei Acknowledgements Provincial Gvernment, China (no. 301140344). This work was supported by the National Science Foundation of China (no. 39870674), Science Foundation of Anhui

References

Algeciras-Schimnich A, Vlahakis SR, Villasis-Keever A, Kern C, Cornuel JF, Billard C, Tang R, Rouillard D, Stenou Gomez T, Heppelmann CJ, Bou G and Paya CV. (2002). V, Defrance T, Ajchenbaum-Cymbalista F, Simonin PY, AIDS, 16, 1467–1478. Feldblum S and Kolb JP. (2004). Blood, 103, 679–688. Ansel KM, Ngo VN, Hyman PL, Luther SA, Forster R, Kim CH, Rott LS, Clark-Lewis I, Campbell DJ, Wu L and Sedgwick JD, Browning JL, Lipp M and Cyster JG. (2000). Butcher EC. (2001). J. Exp. Med., 193, 1373–1382. Nature, 406, 309–314. Kruse N, Pette M, Toyka K and Rieckmann P. (1997). Bennett JM, Catovsky D, Daniel MT, Flandrin G, Galton J. Immunol. Methods, 210, 195–203. DA, Gralnick HR and Sultan C. (1981). Br. J. Haematol., Lin JH, Deng G, Huang Q and Morser J. (2000). Biochem. 47, 553–558. Biophys. Res. Commun., 279, 820–831. Bennett JM, Catovsky D, Daniel M-T, Flandrin G, Galton Lopez-Fraga M, Fernandez R, Albar JP and Hahne M. (2001). DA, Gralnick HR and Sultan C. (1989). J. Clin. Pathol., 42, EMBO Rep., 2, 945–951. 567–584. Luther SA, Lopez T, Bai W, Hanahan D and Cyster JG. Berndt C, Mopps B, Angermuller S, Gierschik P and (2000). Immunity, 12, 471–481. Krammer PH. (1998). Proc. Natl. Acad. Sci. USA, 95, Massari P, Ho Y and Wetzler LM. (2000). Proc. Natl. Acad. 12556–12561. Sci. USA, 97, 9070–9075. Breitfeld D, Ohl L, Kremmer E, Ellwart J, Sallusto F, Lipp M Mazzucchelli L, Blaser A, Kappeler A, Scharli P, Laissue JA, and Forster R. (2000). J. Exp. Med., 192, 1545–1552. Baggiolini M and Uguccioni M. (1999). J. Clin. Invest., 104, Chai J, Shiozaki E, Srinivasula SM, Wu Q, Datta P, Alnemri R49–R54. ES, Shi Y and Dataa P. (2001). Cell, 104, 769–780. Medema JP, Planelles-Carazo L, Hardenberg G and Hahne Cheson BD, Bennett JM, Grever M, Kay N, Keating MJ, M. (2003). Cell Death Diff., 10, 1121–1125. O’Brien S and Rai KR. (1996). Blood, 87, 4990–4997. Miller LK. (1999). Trends Cell. Biol., 9, 323–328. Duckett CS, Nava VE, Gedrich RW, Clem RJ, Van Dongen Muller G, Hopken UE and Lipp M. (2003). Immunol. Rev., JL, Gilﬁllan MC, Shiels H, Hardwick JM and Thompson 195, 117–135. CB. (1996). EMBO J., 15, 2685–2694. Ohl L, Henning G, Krautwald S, Lipp M, Hardtke S, Finke D, Acha-Orbea H, Mattis A, Lipp M and Kraehenbuhl Bernhardt G, Pabst O and Forster R. (2003). J. Exp. J. (2002). Immunity, 17, 363–373. Med., 197, 1199–1204. Fo¨ rster R, Emrich T, Kremmer E and Lipp M. (1994). Blood, Okada T, Ngo VN, Ekland EH, Forster R, Lipp M, 84, 830–840. Littman DR and Cyster JG. (2002). J. Exp. Med., 196, Fo¨ rster R, Mattis AE, Kremmer E, Wolf E, Brem G and Lipp 65–75. M. (1996). Cell, 87, 1037–1047. Pradet-Balade B, Medema JP, Lopez-Fraga M, Lozano JC, Gunn MD, Ngo VN, Ansel KM, Ekland EH, Cyster JG and Kolfschoten GM, Picard A, Martinez-A C, Garcia-Sanz JA Williams LT. (1998). Nature, 391, 799–803. and Hahne M. (2002). EMBO J., 21, 5711–5720. Hahne M, Kataoka T, Schroter M, Hofmann K, Irmler M, Qiuping Z, Qun L, Chunsong H, Xiaolian Z, Baojun H, Bodmer JL, Schneider P, Bornand T, Holler N, French LE, Mingzhen Y, Chengming L, Jinshen H, Qingping G, Kejian Sordat B, Rimoldi D and Tschopp J. (1998). J. Exp. Med., Z, Zhimin S, Xuejun Z, Junyan L and Jinquan T. (2003). 188, 1185–1190. Cancer Res., 63, 6469–6477. Han ZC, Lu M, Li J, Defard M, Boval B, Schlegel N and Caen Reif K, Ekland EH, Ohl L, Nakano H, Lipp M, Forster R and JP. (1997). Blood, 89, 2328–2335. Cyster JG. (2002). Nature, 416, 94–99. Hargreaves DC, Hyman PL, Lu TT, Ngo VN, Bidgol A, Rennert P, Schneider P, Cachero TG, Thompson J, Trabach Suzuki G, Zou YR, Littman DR and Cyster JG. (2001). L, Hertig S, Holler N, Qian F, Mullen C, Strauch K, J. Exp. Med., 194, 45–56. Browning JL, Ambrose C and Tschopp J. (2000). J. Exp. Jinquan T, Jacobi HH, Jing C, Millner A, Sten E, Hviid L, Med., 192, 1677–1684. Anting L, Ryder LP, Glue C, Skov PS, Jarman E, Lamberth Riedl SJ, Renatus M, Schwarzenbacher R, Zhou Q, Sun C, K, Malling HJ and Poulsen LK. (2003). J. Immunol., 171, Fesik SW, Liddington RC and Salvesen GS. (2001). Cell, 1722–1731. 104, 791–800. Jinquan T, Quan S, Jacobi HH, Jing C, Millner A, Jensen B, Roth W, Wagenknecht B, Klumpp A, Naumann U, Hahne M, Madsen HO, Ryder LP, Svejgaard A, Malling HJ, Skov PS Tschopp J and Weller M. (2001). Cell Death Differ., 8, and Poulsen LK. (2000). Blood, 96, 1230–1238. 403–410. Jonuleit H, Schmitt E, Schuler G, Knop J and Enk AH. Saeki H, Wu MT, Olasz E and Hwang ST. (2000). Eur. (2000). J. Exp. Med., 192, 1213–1222. J. Immunol., 30, 2808–2814. Kanbe K, Shimizu N, Soda Y, Takagishi K and Hoshino H. Sanna MG, da Silva Correia J, Ducrey O, Lee J, Nomoto K, (1999). Virology, 265, 264–273. Schrantz N, Deveraux QL and Ulevitch RJ. (2002). Mol. Kasof GM and Gomes BC. (2001). J. Biol. Chem., 276, Cell. Biol., 22, 1754–1766. 3238–3246. Scheuerer B, Ernst M, Durrbaum-Landmann I, Fleischer J, Kelly K, Manos E, Jensen G, Nadauld L and Jones DA. Grage-Griebenow E, Brandt E, Flad HD and Petersen F. (2000). Cancer Res., 60, 1021–1027. (2000). Blood, 95, 1158–1166.

Oncogene Apoptotic resistance via CXCR5/livin/APRIL pathway Z Qiuping et al 584 Schmollinger JC, Vonderheide RH, Hoar KM, Maecker B, Till KJ, Lin K, Zuzel M and Cawley JC. (2002). Blood, 99, Schultze JL, Hodi FS, Soiffer RJ, Jung K, Kuroda MJ, 2977–2984. Letvin NL, Greenﬁeld EA, Mihm M, Kutok JL and Voigt I, Camacho SA, de Boer BA, Lipp M, Forster R and Dranoff G. (2003). Proc. Natl. Acad. Sci. USA, 100, Berek C. (2000). Eur. J. Immunol., 30, 560–567. 3398–3403. Vucic D, Stennicke HR, Pisabarro MT, Salvesen GS and Dixit Sica A, Saccani A, Borsatti A, Power CA, Wells TN, Luini W, VM. (2000). Curr. Biol., 10, 1359–1366. Polentarutti N, Sozzani S and Mantovani A. (1997). J. Exp. Wu MT and Hwang ST. (2002). J. Immunol., 168, 5096–5102. Med., 185, 969–974. Youn BS, Kim YJ, Mantel C, Yu KY and Broxmeyer HE. Siegel RM and Lenardo MJ. (2001). Nat. Immunol., 2, (2001). Blood, 98, 925–933. 577–578. Youn BS, Yu KY, Oh J, Lee J, Lee TH and Broxmeyer HE. Smith JR, Braziel RM, Paoletti S, Lipp M, Uguccioni M and (2002). Apoptosis, 7, 271–276. Rosenbaum JT. (2003). Blood, 101, 815–821. Yu G, Boone T, Delaney J, Hawkins N, Kelley M, Stein JV, Lopez-Fraga M, Elustondo FA, Carvalho-Pinto CE, Ramakrishnan M, McCabe S, Qiu WR, Kornuc M, Xia Rodriguez D, Gomez-Caro R, De Jong J, Martinez-A C, XZ, Guo J, Stolina M, Boyle WJ, Sarosi I, Hsu H, Senaldi G Medema JP and Hahne M. (2002). J. Clin. Invest., 109, and Theill LE. (2000). Nat. Immunol., 1, 252–256. 1587–1598. Yu P, Wang Y, Chin RK, Martinez-Pomares L, Gordon S, Thornberry NA and Lazebnik Y. (1998). Science, 281, Kosco-Vibois M H, Cyster J and Fu YX. (2002). 1312–1316. J. Immunol., 168, 5117–5123.

Oncogene Oncogene (2011) 30,2798 & 2011 Macmillan Publishers Limited All rights reserved 0950-9232/11 www.nature.com/onc RETRACTIONS Androgen activates PEG10 to promote carcinogenesis in hepatic cancer cells

X Jie, C Lang, Q Jian, L Chaoqun, Y Dehua, S Yi, J Yanping, X Luokun, Z Qiuping, W Hui, G Feili, J Boquan, J Youxin and T Jinquan

Oncogene (2011) 30, 2798; doi:10.1038/onc.2011.66

Retraction to: Oncogene (2007) 26, 5741–5751; doi:10.1038/ sj.onc.1210362; published online 19 March 2007

This paper has been retracted.

Selectively frequent expression of CXCR5 enhances resistance to apoptosis in CD8 þ CD34 þ T cells from patients with T-cell-lineage acute lymphocytic leukemia

Z Qiuping, X Jie, J Youxin, W Qun, J Wei, L Chun, W Jin, L Yan, H Chunsong, Y Mingzhen, G Qingping, L Qun, Z Kejian, S Zhimin, L Junyan and T Jinquan

Oncogene (2011) 30, 2798; doi:10.1038/onc.2011.67

Retraction to: Oncogene (2005) 24, 573–584; doi:10.1038/ sj.onc.1208184; published online 6 December 2004

This paper has been retracted.