Serine protease inhibition attenuates rIL-12-induced GZMA activity and proinflammatory events by modulating the Th2 profile from estrogen-treated mice

Abstract

Estrogen has potent immunomodulatory effects on proinflammatory responses, which can be mediated by serine proteases. We now demonstrate that estrogen increased the extracellular expression and IL-12-induced activity of a critical member of serine protease family Granzyme A, which has been shown to possess a novel inflammatory persona. The inhibition of serine protease activity with inhibitor 4-(2-aminoethyl) benzenesulfonyl fluoride hydrochloride significantly diminished enhanced production of proinflammatory interferon-γ, IL-1β, IL-1α, and Granzyme A activity even in the presence of a Th1-inducing cytokine, IL-12 from splenocytes from in vivo estrogen-treated mice. Inhibition of serine protease activity selectively promoted secretion of Th2-specific IL-4, nuclear phosphorylated STAT6A, signal transducer and activator of transcription (STAT)6A translocation, and STAT6A DNA binding in IL-12-stimulated splenocytes from estrogen-treated mice. Inhibition with 4-(2-aminoethyl) benzenesulfonyl fluoride hydrochloride reversed the down-regulation of Th2 transcription factors, GATA3 and c-Maf in splenocytes from estrogen-exposed mice. Although serine protease inactivation enhanced the expression of Th2-polarizing factors, it did not reverse estrogen-modulated decrease of phosphorylated STAT5, a key factor in Th2 development. Collectively, data suggest that serine protease inactivity augments the skew toward a Th2-like profile while down-regulating IL-12-induced proinflammatory Th1 biomolecules upon in vivo estrogen exposure, which implies serine proteases as potential regulators of inflammation. Thus, these studies may provide a potential mechanism underlying the immunomodulatory effect of estrogen and insight into new therapeutic strategies for proinflammatory and female-predominant autoimmune diseases.

Figure 1.

Inhibition of serine protease activity altered estrogen-induced extracellular GZMA activity but not extracellular and intracellular GZMA expression. Splenocytes from estrogen or placebo-treated mice were stimulated with rIL-12 (20 ng/mL) or left unstimulated in media with or without AEBSF (425 μM) for 3 hours. Extracellular GZMA (A) was determined in concentrated supernatants and analyzed with Western blot. Intracellular GZMA (B) and β-actin (protein loading control) expression in cytoplasmic extracts of cultured splenocytes were determined with Western blotting. Top portion of panels depict representative blots (mean ± SEM; panel A, n ≥ 8 mice per treatment; panel B, n ≥ 5 mice per treatment; *, P < .05). Dashed dividing lines delineate groups of specific bands from different sections of the same gel. Panel C demonstrates BLT esterase activity, a specific marker of GZMA activity, which was measured in supernatants of unstimulated (media) or rIL-12-stimulated lymphocytes treated with or without AEBSF from placebo or estrogen-treated mice at 3 hours. Values for BLT esterase activity are represented as percent expression of GZMA activity (±SEM) as paired to unstimulated cells (media), which was considered 100% (means of 9 or 7 experiments ± SEM; *, P < .05).

Figure 2.

Inhibition of serine protease activity down-regulates estrogen-modulated IFN-γ, IL-1β, and IL-1α production. IFN-γ levels in supernatants from unstimulated (media) or rIL-12-stimulated splenocytes from estrogen and placebo-treated mice cultured for 3 (A) and 24 hours (B) were measured with ELISA (means ± SEM; panel A, n ≥ 11 mice per treatment; panel C, n ≥ 5 mice per treatment). Supernatants from estrogen and placebo-treated rIL-12-induced splenocytes cultured with or without AEBSF were evaluated for IL-1 production via ELISA (A–D). Panels C and D show IL-1β production in picograms per mL (±SEM) at 3 and 24 hours, respectively. Level of IL-1α secretion at 3 (E) and 24 hours (F) culture was depicted as picograms per mL (±SEM). Panels C and D, n ≥ 6 mice per treatment; panels E and F, n ≥ 5 mice per treatment; *, **, and *** denote P < .05, P < .01, and P < .001, respectively).

Figure 3.

Inhibition of serine protease activity via AEBSF transiently induced IL-4 production at 3 hours from rIL-12-cultured cells from estrogen-treated mice. Supernatants after 3 (panel A, n = 8) or 24 hours (panel B, n = 3) of cell cultures with rIL-12 or media with or without AEBSF were analyzed for IL-4 production with ELISA (means pg/mL ± SEM; *, P < .05).

Figure 4.

Selective enhancement of nuclear pSTAT6A, total STAT6A, and STAT6A DNA binding activity in splenocytes from estrogen-treated mice was dependent on serine protease inactivation. Nuclear extracts (25μg) from splenocytes cultured in media or rIL-12 with or without AEBSF for 3 hours were subjected to Western blot and depicted as mean relative densitometry for pSTAT6A (A), pSTAT6B (B), STAT6A (C), and STAT6B (D). Top portions of panels A and B show representative blots for pSTAT6A/B and STAT6A/B, respectively. Panel E shows representative blot for cytoplasmic STAT6/B expression as detected using 25 μg cytoplasmic extracts of splenocytes from estrogen or placebo-treated mice. β-Actin was used as a protein loading control (n = 6; mean ± SEM; *, P < .05). Dashed dividing lines delineate groups of specific bands from different sections of the same gel. Panel F, EMSAs were performed using 5 μg of nuclear extracts from splenocytes from estrogen- or placebo-exposed mice cultured with media or rIL-12 in the presence or absence of AEBSF for 3 hours. For the controls, samples without nuclear extract with only STAT6-labeled probe, HeLa control nuclear extract with STAT6-labeled probe, with no probe and unlabeled cold STAT6 probe (cold competition) instead of labeled STAT6 probe were used. Dashed dividing lines indicate the grouping of sections of gel run at the same time under the same conditions. Data shown represent at least 4 independent experiments.

Figure 5.

Inhibition of serine proteases reverses estrogen-mediated nuclear accumulation of GATA3 and c-Maf. Nuclear extracts (25 μg/sample) from media or rIL-12-treated splenocytes cultured with or without AEBSF for 3 hours were analyzed for GATA3 (A) and c-Maf (B) with Western blot analysis. Representative blots and mean relative densitometry data of GATA3 (A) and c-Maf (B) are presented. Results in panel A (GATA3) are a representative of least 7 independent experiments, and data from panel B (c-Maf) are representative of at least 5 independent experiments (mean± SEM; * and ** denote P < .05 and P < .01). Dashed dividing lines delineate groups of bands from different sections of the same gel.

Figure 6.

AEBSF treatment does not alter estrogen-mediated hindrance of nuclear pSTAT5. Nuclear expression phosphorylated STAT5A (pSTAT5A) and total STAT5 from media or rIL-12-activated splenocytes with or without AEBSF for 3 hours were detected with Western blot analysis. β-Actin served as protein loading control. Dashed dividing lines indicate the grouping of sections of the different gel run at the same time under the same conditions from cell cultures. Representative blot pictures and mean relative densitometry data (± SEM) from at least 3 experiments are shown (*, P < .05).

Figure 7.

A proposed schematic representation showing how inhibition of serine protease activity in the presence of estrogen and/or IL-12 can still shift a Th1 signature toward Th2. Estrogen-enhanced Th1 profile (ie, IFN-γ, extracellular GZMA activity, IL-1β, and IL-1α) was diminished following AEBSF treatment. Serine protease inactivation selectively enhanced pSTAT6A and STAT6A DNA binding whereas nuclear pSTAT6B was down-regulated. Up-regulation of pSTAT6A activation could lead to increased GATA3 and c-Maf. Interestingly, estrogen diminished nuclear pSTAT5A expression, which was not reversed by AEBSF. Although enhanced pSTAT6, c-Maf, and GATA3 levels could increase IL-4, which could in turn further activate STAT6, the absence of pSTAT5A could fail to further up-regulate a dominant and stable Th2 profile.