RPMI8226 and Jurcat cells (American Tissue Culture Collection, Rockville, MD, USA) were cultured in RPMI-1640 media supplemented with 10% (v/v) fetal bovine serum, penicillin G and streptomycin. Anti-CD99 monoclonal Abs (YG32 and DN16-PE) were purchased from Dynona Inc. (Seoul, Republic of Korea). For the engagement of CD99, cells were stimulated with 10 µg/ml YG32 Ab and 17 µg/ml anti-mouse immunoglobulin G (IgG) monovalent F(ab) fragments (Jackson ImmunoResearch, Westgrove, PA, USA) for crosslinking. All anti-MAP kinase antibodies were purchased from Cell Signaling Technology (Beverly, MA, USA).

Total RNA was extracted from cells using the NucleoSpin kit (Macherey-Nagel, Düren, Germany) and reverse-transcribed using the iScript cDNA synthesis kit (Bio-Rad Laboratories, Hercules, CA, USA). The cDNA was amplified using specific primers for CD99 type-I and -II, as described previously (513). Levels of glyceraldehyde-3-phosphate dehydrogenase (GAPDH) mRNA were used to normalize expression. Real-time PCR was performed using an iCycler iQ system with the iQ SYBR Green Supermix (Bio-Rad Laboratories). Fold induction was calculated with the comparative Ct method (23), using the expression of ribosomal protein S18 as the reference. The primers used in real-time PCR analyses are listed in Table S1.

The human BATF gene was PCR-amplified from cDNA of RPMI8226 cells using the following primers: 5'-GGC GCTAGCGCCACCATGCCTCACAGCTCCGAC-3' and 5'-GCCCTCGAGTCAGGGCTGGAAGCGC-3'. The amplified products were digested using the NheI and XhoI restriction enzymes, and were ligated into the pCDNA3.1- hygro expression vector (Invitrogen, Carlsbad, CA, USA). The resulting expression construct was sequenced to verify the cDNA sequence vectors were used in the reporter assays.

The luciferase assay was performed using the luciferase assay system and the β-galactosidase assay system (Promega, Madison, WI). RPMI8226 and Jurkat cells were transfected using Lipofectamine LTX (Invitrogen) in accordance with the manufacturer's protocol. A standard transfection reaction used 1 µg of DNA, including AP1-Luc (Clontech, Mountain View, CA), pM1-β-Gal (Roche Diagnostics, Indianapolis, IN) and expression plasmid (pCDNA3.1 or pCDNA3.1-BATF). Luciferase activities were normalized with respect to β-galactosidase activities.

BATF-targeting and control siRNAs were purchased from Genolution Pharmaceuticals Inc. (Seoul, Korea) (Table S2). RPMI8226 cells (2.0×10⁶) were electroporated using 1 µM BATF-targeting siRNA or control siRNA using a microporator (Invitrogen, Carlsbad, CA). BATF expression was determined using real-time PCR at 48 hours after electroporation.

1.0×10⁷ cells were labeled with 0.2 µM CFSE and incubated at 37℃ for 10 minutes. Cold fetal calf serum was added to stop the staining reaction. After 3 days of culture, the fluorescence intensity of CFSE was measured using a FACSCalibur flow cytometer and analyzed using FlowJo software (Ashland, OR, USA).

Apoptosis was analyzed by the apoptosis detection kit (BD Biosciences, San Jose, CA, USA) and tetramethylrhodamine ethyl ester perchlorate (TMRE; Molecular Probes, Eugene, OR, USA) staining. For annexin V staining, cells 1×10⁶ cells were suspended in binding buffer and incubated with Annexin V-FITC and propidium iodide for 15 min at room temperature in the dark. For TMRE staining, 1×10⁶ cells were incubated with 50 nM TMRE for 20 minutes at 37℃. The samples were then measured using a FACSCalibur flow cytometer and analyzed using CellQuest-Pro software (BD Biosciences).

To select a B-cell line for studies of CD99 function, CD99 surface expression was measured in the B-cell lines RPMI8226, ARH77, NCI-H929, and IM-9 by flow cytometry (Fig. 1A). Among the cell lines tested, RPMI8226 cells had the highest surface expression of CD99. In particular, the expression level of CD99 on RPMI8226 cells was higher than on IM-9 cells, in which CD99 engagement triggers intracellular signaling (5). There are two major isoforms of CD99, termed type-I and -II, which have distinct roles in various cell types (513). RT-PCR analysis with primers designed to specifically amplify each isoform showed that only the type-I isoform was detected in RPMI8226 cells, whereas both the type-I and -II isoforms were detected in MDA-MB-435 breast cancer cells, as reported previously (13).

CD99 engagement with the functional anti-CD99 antibody (Ab) is the typical way to trigger CD99-mediated intracellular signaling that culminates in the aggregation, apoptosis, and proliferation of immune cells (456789101124). To elucidate the function of CD99-mediated signaling in RPMI8226 cells, we employed an agonistic anti-CD99 Ab, YG32, to trigger intracellular signaling (1225). We measured changes in AP-1 activity upon CD99 engagement by YG32 in RPMI8226 cells because CD99 has been reported to stimulate AP-1 signaling in other cells (1113). Surprisingly, CD99 engagement reduced AP-1 activity in RPMI8226 cells, whereas it enhanced AP-1 activity in Jurkat cells, as reported previously (Fig. 2A) (11). We next examined MAP kinase activity because MAP kinases are upstream of AP-1, and CD99-mediated signaling activates MAP kinases in Jurkat cells (1112). However, phosphorylation of both extracellular signal-regulated kinase (ERK) and c-Jun N-terminal kinase (JNK) was increased by CD99 engagement in RPMI8226 cells (Fig. 2B). We postulated that AP-1 activity could be regulated by a negative regulator of AP-1 in RPMI8226 cells, rather than by altered MAP kinase signaling mediated by CD99. We measured changes in the expression of AP-1 transcription factors, including BATF, using real-time PCR (Fig. 2C). BATF was identified as being potentially responsible for the reduction in AP-1 levels because it is a negative endogenous regulator of AP-1-mediated transcription. BATF expression was 12-fold higher in CD99-engaged cells than in control cells, whereas the expression levels of Jun, JunB, Fos, and FosB were all less than 2-fold higher in CD99-engaged cells than in control cells (Fig. 2C). Time-course experiments showed that BATF expression was rapidly induced within 8 hours after YG32 Ab treatment, and was sustained at a high level up to 24 hours after CD99 engagement (Fig. 2D). This result is completely different to the effects observed in Jurkat cells, where BATF expression did not increase upon YG32 Ab treatment (Fig. 2E). These results suggest that the induction of BATF expression following CD99 engagement suppresses AP-1 activity in RPMI8226 cells.

To investigate the function of BATF in RPMI8226 cells, we measured AP-1 activity when BATF was ectopically overexpressed. As previously reported in other cells (171819), overexpression of BATF reduced AP-1 activity in RPMI8226 cells, as well as in 293T cells (Fig. 3A). To investigate the extent of BATF involvement in the CD99-induced reduction of AP-1 activity, BATF expression was knocked down by siRNA transfection. Transfection of BATF-targeting siRNA reduced BATF expression by about 50% in comparison to that in cells transfected with non-silencing control siRNA (Fig. 3B). In cells transfected with control siRNA, CD99 engagement reduced AP-1 activity. However, in cells transfected with BATF-targeting siRNA, the CD99-mediated reduction in AP-1 activity was abolished (Fig. 3C). These results demonstrate that induction of BATF expression by CD99 engagement suppresses AP-1 activity in RPMI8226 cells.

Given that CD99 engagement suppressed AP-1 activity in RPMI8226 cells, we investigated the effect of CD99 signaling on the proliferation of these cells. Engagement of CD99 by its agonistic Ab YG32 significantly reduced the number of cells recovered after 3 days of culture (Fig. 4A). To investigate whether growth suppression triggered by CD99 engagement was owing to a reduction in proliferation or an increase in apoptosis, apoptosis was evaluated by the TMRE assay and annexin V staining, and proliferation was evaluated by CFSE labeling. After 3 days of culture, the decrease in CFSE fluorescence intensity observed in control cells was evidently retarded in YG32-treated cells (Fig. 4B). In the TMRE assay, there was no marked difference between YG32-treated and control cells (Fig. 4C). Over 24 h of culture, the ratio of annexin V-positive cells to the total number of cells did not significantly differ between YG32-treated and control samples, whereas this ratio significantly increased in bortezomib-treated samples (Fig. 4D). These results suggest that the reduction in the number of cells recovered after CD99 engagement was mostly caused by the suppression of cell division, and not by a marked increase in the rate of apoptosis. We measured the expression of cyclin D1, D2, and D3 using real-time PCR after CD99 engagement because cyclin D proteins are major regulators of proliferation and cyclin D1 is a target gene of active AP-1 (26). The expression levels of cyclin D1 and D3 were significantly lower in CD99-engaged RPMI8226 cells than in untreated cells (Fig. 4E). These results suggest that the decrease in the recovery of RPMI8226 cells following treatment with an anti-CD99 Ab is mainly owing to a reduction in cell proliferation, possibly through reduced AP-1 activity via BATF. Nonetheless, the possible contribution of pro-apoptotic mechanisms cannot be entirely excluded.