BALB/c mice were purchased from Orient. Co. Ltd. (Gyeonggi-do, Korea) and maintained in an animal environmental control chamber (Myung Jin Inst. Co., Seoul, Korea). Animals were fed Purina Laboratory Rodent Chow 5001 ad libitum. Eight to twelve weeks age of mice were used in this study. Animal care was in accordance of with the institutional guidelines set forth by Kangwon National University.

The murine B lymphoma cell line, CH12F3-2A was provided by Dr. T. Honjo (Osaka University, Japan) (16). These cells were cultured at 37℃ in a humidified atmosphere containing 5% CO₂ in RPMI 1,640 medium (Sigma) supplemented with 10% FBS, 50µM 2-ME, 5 mM HEPES, penicillin (100 U/ml)/streptomycin (100µg/ml). Mouse spleen B cell suspensions were prepared as described before (4) and whole mesenteric lymph node cells were prepared as teasing the tissues with forceps.

ELISAs were performed as described previously (17). The reaction products were measured at 405 nm with an ELISA reader (Molecular Devices VERSAMAX reader, Molecular Devices, Sunnyvale, CA, USA).

Isolated mouse spleen B cells were labeled with a CFSE kit (Invitrogen Life Technologies, Carlsbad, CA, USA), according to the manufacture's instruction, and added with LPS, activin A (10 ng/ml), and TGF-β1 (0.2 ng/ml). Dilution of CFSE was measured by counting 30,000 viable cells with a FACSCalibur (BD Biosciences).

We have recently shown that activin A selectively stimulates IgA expression (15). However, the relationship between actions of activin A and TGF-β1 in the context of IgA CSR is unclear at the moment. In this regard, we have demonstrated in the earlier study that anti-proliferative activity of TGF-β facilitates IgA CSR leading to an increase of IgA production (18). Therein, we first examined the effect of activin A on the growth of murine spleen B cells. LPS is generally adopted for the non-specific stimulation of mouse B cells. As shown in Fig. 1, LPS increased cell viability up to 4 day of culture. Effect of Activin A on the cell viability was marginal while TGF-β1 substantially inhibited it. Subsequently, we tested the effect of activin A on B cell proliferation. In parallel to cell viability, Activin A had no influence on cell proliferation though anti-proliferative activity of TGF-β1 was evident (Fig. 2). These results suggest that activin A increases IgA expression through the mechanism different from that of TGF-β1 and that activin A has such activity without inhibiting cell proliferation.

Then, how does activin A stimulate B cells to express IgA isotype expression? To assess its underlying mechanism, we examined the possible involvement of TGF-β in the activin A-induced IgA secretion. In the culture of CH12F3-2A B cell lymphoma, the activin A-induced IgA production was abrogated by anti-activin A Ab in a dose dependent manner (Fig. 3). In contrast, anti-TGFβ1 Ab failed to eliminate the IgA response enhanced by activin A. Moreover, in normal spleen B cell cultures, a similar pattern of IgA secretion was observed when anti-activin A Ab and anti-TGFβ1 Ab were treated as shown in Fig. 4. On the other hand, anti-activin A Ab or anti-TGFβ1 Ab little affected secretion patterns of IgM, IgG2b, and IgG3. These results indicate that activin A selectively induces IgA expression and that this activity of activin A is independent of that of TGF-β1. Furthermore, these results exclude the possibility that activin A has such activity through induction of TGF-β1 production.

Peyer's patch (PP) is the most important gut-associated lymphoid tissue, where most of IgA B cells are committed (19). Therein, it is difficult to isolate surface IgA negative B cells from PP. On the other hand, mesenteric lymph node (MLN) B cells are also known to be the site for IgA B cell commitment (20,21). Here, we were interested if activin A affects IgA production by MLN B cells. As shown in Fig. 5, activin A, as similar to TGF-β1, increased IgA secretion by MLN cells but neither IgG1 nor IgG2b.

In summary, we demonstrate in this study that activin A, as similar to TGF-β1, selectively increases IgA expression. Nevertheless, Activin A had no influence on cell proliferation, which is in contrast to the anti-proliferative property of TGF-β1. We have recently shown that activin A, like TGF-β1 adopts Smad3 and Smad4 as the main signaling mediators (15). Therefore, it still remains to determine the mechanism by which activin A causes IgA isotype switching, distinguishing from that of TGF-β1. Finally we observed that activin A can stimulate mesenteric lymph node B cells to produce IgA. These results raise the possibility that activin A may have important effects on the expression of IgA in the gut.