IL-1β is overexpressed and aberrantly regulated in corticosteroid nonresponders with autoimmune inner ear disease

Abstract

Autoimmune inner ear disease is an enigmatic disorder characterized by recurring episodes of sudden or progressive sensorineural hearing loss. Hearing loss can be improved by timely corticosteroid administration, but only half of those treated respond, and for many responders, that response is lost over time. The mechanisms that control corticosteroid responsiveness in this disorder are largely uncharacterized. We have previously identified that the induction by dexamethasone of IL-1R type II (IL-1R2) expression in PBMC predicts corticosteroid responsiveness in this disorder. In this study, we asked whether IL-1β was overexpressed, and whether clinical corticosteroid responders differentially regulated IL-1β expression or release in response to dexamethasone, as compared with nonresponders. IL-1β has been reported to induce matrix metalloproteinase-9 (MMP-9) expression. Given that metalloproteinases can cleave IL-1R2, we also asked whether MMP-9 expression was altered in this disorder. In this study, we demonstrate that corticosteroid nonresponders have elevated plasma levels of IL-1β and MMP-9 as compared with clinically responsive patients (p = 0.0008 and p = 0.037, respectively). Increasing MMP-9 expression correlated with increasing IL-1β concentration, suggesting that IL-1β expression regulates MMP-9 expression. As expected, monocytes were the predominant producers of IL-1β. In vitro exposure of PBMC to dexamethasone from clinical corticosteroid responders suppressed IL-1β release. PBMC of corticosteroid nonresponders have substantially higher release of IL-1β into the conditioned media, and when exposed to dexamethasone, failed to repress IL-1β release (p = 0.05). Treatment of PBMC from clinical corticosteroid nonresponders with anakinra resulted in repression of IL-1β release, suggesting that IL-1β blockade may be a viable therapy for these patients.

Figures

FIGURE 1

IL-1β and MMP-9 are over-expressed in plasma of patients with corticosteroid-resistant hearing loss. Plasma values for 47 patients who received corticosteroid therapy for an acute decline in hearing. The majority of the patients had autoimmune inner ear disease, whereas a few had a SSNHL of suspected immunologic origin. Plasma values for IL-1β, MMP-9, sIL-1R2, and IL-17 were measured for clinical corticosteroid responders and nonresponders. Corticosteroid nonresponders demonstrated significantly higher levels of IL-1β than nonresponders. *p = 0.0008 (A). Similarly, a significantly higher level of MMP-9 was observed in corticosteroid nonresponders when compared with responders. **p = 0.037 (B). Neither sIL-1R2 (C) nor IL-17 (D) plasma values correlated with clinical corticosteroid responsiveness.

FIGURE 2

IL-1β induces MMP-9 expression; however, dexamethasone is required for MMP-9 to reciprocally induce IL-1β PBMC from control patients were obtained to determine whether IL-1β induces expression of MMP-9. RNA expression was measured by Q-RT-PCR. PBMC from three control subjects were treated with either dexamethasone or increasing amounts of IL-1β (A). Fold change is shown relative to the unstimulated condition. IL-1β clearly induces MMP-9 mRNA expression, p = 0.004, by a repeated measures ANOVA. Conversely, control PBMC were treated with either the catalytic domain of MMP-9 alone, or in combination with dexamethasone (D+M). The catalytic domain was used at either 100 or 500 ng/ml (B). Although MMP-9 alone had no effect on IL-1β transcription (data not shown), increasing the MMP-9 concentration in combination with dexamethasone resulted in a strong induction of IL-1β transcription. Similarly, MMP-9 alone had no effect on sIL-1R2 transcription or release; however, in combination with dexamethasone, increased MMP-9 concentration augmented both sIL-1R2 transcription and release (C).

FIGURE 3

Dexamethasone fails to reduce IL-1β transcription and release in patients unresponsive to corticosteroid therapy. PBMC from seven corticosteroid responders and seven nonresponders were evaluated prior to clinical treatment. In this figure, IL-1β and MMP-9 RNA expression (measured by Q-RT-PCR) and IL-1β and MMP-9 release into the culture supernatant (measured by ELISA) were evaluated in response to anakinra, minocycline, and dexamethasone as compared with unstimulated PBMC. Results for RNA expression are shown as fold change relative to unstimulated for IL-1β (A) and MMP-9 (C). Release of IL-1β (B) and MMP-9 (D) was measured for each of these culture conditions. Strikingly, in clinical corticosteroid nonresponders, coculture with dexamethasone resulted in increased IL-1β mRNA production, *p = 0.05, Mann–Whitney U test (A). And it failed to prevent IL-1β release, **p = 0.05, Mann–Whitney U test (B). In these nonresponders, anakinra effectively reduced IL-1β release, suggesting a role for anakinra therapy in corticosteroid nonresponders. No significant difference was observed for MMP-9 mRNA expression (C) or release (D), n.s. (not significant).

FIGURE 4

IL-1β mRNA expression is not reduced by dexamethasone in monocytes. Monocyte and nonmonocyte fractions were isolated from a group of AIED patients (n = 3) and a group of controls (n = 2). mRNA expression was determined by Q-RT-PCR, and expression was measured as fold change relative to the unstimulated condition. mRNA expression for IL-1β (A) and MMP-9 (B) in monocytes was performed. The nonmonocyte fraction was also evaluated for IL-1β (C) and MMP-9 (D) expression. Unstimulated monocytes from patients demonstrated substantially greater IL-1β mRNA than in either the nonmonocyte fraction of patients, or from either fraction of controls (detection threshold by Q-RT-PCR 19.4 as compared with 24.0 for nonmonocytes and 25.9 and 26.2 for controls). Dexamethasone had no substantial ability to repress IL-1β mRNA expression in monocytes from AIED patients (A). Given that transcriptional repression of IL-1β was observed in the PBMC of corticosteroid responders, it is possible that monocyte–T cell interactions may be necessary for the observed reduction in IL-1β transcription. MMP-9 transcription was repressed by dexamethasone in both fractions.

FIGURE 5

IL-1β is predominantly released from monocytes. Release of IL-1β and MMP-9 into conditioned media was determined by ELISA from the monocyte and nonmonocyte fractions in AIED patients and controls from Fig. 4. Monocytes clearly produce the majority of IL-1β (A, mean 122.7 pg/ml in patients versus 15.1 pg/ml in controls) as compared with the nonmonocyte fraction (B, mean 60.0 pg/ml in patients versus 1.2 pg/ml in controls), although both monocyte and nonmonocyte fractions were statistically significantly different in AIED patients than controls (monocyte fractions AIED versus controls, *p = 0.05; nonmonocyte fractions, AIED versus control, **p = 0.003). Anakinra was comparable to dexamethasone in its ability to reduce IL-1β release from monocytes (A). MMP-9 release from monocytes (C) and nonmonocyte fractions (D) was similar. Presence of minocycline, anakinra, and dexamethasone in the culture has no significant effect on MMP-9 release from either fraction.