HspB4/αA-Crystallin Modulates Neuroinflammation in the Retina via the Stress-Specific Inflammatory Pathways

Abstract

Purpose: We have previously demonstrated that HspB4/αA-crystallin, a molecular chaperone, plays an important intrinsic neuroprotective role during diabetes, by its phosphorylation on residue 148. We also reported that HspB4/αA-crystallin is highly expressed by glial cells. There is a growing interest in the potential causative role of low-grade inflammation in diabetic retinopathy pathophysiology and retinal Müller glial cells' (MGCs') participation in the inflammatory response. MGCs indeed play a central role in retinal homeostasis via secreting various cytokines and other mediators. Hence, this study was carried out to delineate and understand the regulatory function of HspB4/αA-crystallin in the inflammatory response associated with metabolic stresses.

Methods: Primary MGCs were isolated from knockout HspB4/αA-crystallin mice. These primary cells were then transfected with plasmids encoding either wild-type (WT), phosphomimetic (T148D), or non-phosphorylatable mutants (T148A) of HspB4/αA-crystallin. The cells were exposed to multiple metabolic stresses including serum starvation (SS) or high glucose with TNF-alpha (HG + T) before being further evaluated for the expression of inflammatory markers by qPCR. The total protein expression along with subcellular localization of NF-kB and the NLRP3 component was assessed by Western blot.

Results: Elevated levels of IL-6, IL-1β, MCP-1, and IL-18 in SS were significantly diminished in MGCs overexpressing WT and further in T148D as compared to EV. The HG + T-induced increase in these inflammatory markers was also dampened by WT and even more significantly by T148D overexpression, whereas T148A was ineffective in either stress. Further analysis revealed that overexpression of WT or the T148D, also led to a significant reduction of Nlrp3, Asc, and caspase-1 transcript expression in serum-deprived MGCs and nearly abolished the NF-kB induction in HG + T diabetes-like stress. This mechanistic effect was further evaluated at the protein level and confirmed the stress-dependent regulation of NLRP3 and NF-kB by αA-crystallin.

Conclusions: The data gathered in this study demonstrate the central regulatory role of HspB4/αA-crystallin and its modulation by phosphorylation on T148 in retinal MGCs. For the first time, this study demonstrates that HspB4/αA-crystallin can dampen the stress-induced expression of pro-inflammatory cytokines through the modulation of multiple key inflammatory pathways, therefore, suggesting its potential as a therapeutic target for the modulation of chronic neuroinflammation.

Keywords: Müller glial cells; NF-kB; inflammatory markers; metabolic stress; αA-crystallin.

Figure 1

Metabolic-stress-induced expression of pro-inflammatory mediator is regulated by HspB4/αA-crystallin expression. rMC-1 cells were transfected with either empty vector (EV), wild-type HspB4/αA-crystallin (WT), phosphomimetic form of HspB4/αA-crystallin (T148D), or non-phosphorylatable form of HspB4/αA-crystallin (T148A). Then, 24 h post transfection, cells were either exposed to normal media (10% FBS), serum starvation (no FBS), high glucose (25 mM glucose), or to diabetic like stress (25 mM glucose + 100 ng/mL TNFα) for 4 h. The mRNA levels of (A) interleukin-1 beta (IL-1β), (B) interleukin-6 (IL-6), and (C) monocyte chemoattractant protein-1 (MCP-1) were normalized to the actin-encoding gene Actb. (D) Representative immunoblot depicting the expression of HspB4/αA-crystallin in transfected rMC-1 cells. ** p ≤ 0.01, *** p ≤ 0.001, significantly different from the respective EV-transfected cells. Each endpoint was measured on a minimum of six technical replicates. Statistical analysis was performed by one-way ANOVA followed by Student–Newman–Keuls test.

Figure 2

Primary Müller cells isolation from HspB4/αA-crystallin knockout mice retain MGCs’ characteristics. Primary mouse MGCs retain an elongated morphology during the first few passages but become very flat by passage 6 (A). The isolated Müller cells show expression of specific markers such as peroxiredoxin-6 (Prdx-6), glutamine synthetase (GLUL) and ATP-binding cassette 8a (Abc8a). Representative graphs of mRNA levels of (B) Prdx-6, (C) GLUL, and (D) Abc8a in primary Müller cells at passage 2 (P-2), passage 4 (P-4), and passage 6 (P-6) are shown normalized to the actin-encoding gene Actb. The Müller-cell-specific markers are highly expressed until they dramatically decrease after passage 4.

Figure 3

Primary MGCs (HspB4−/−) express pro-inflammatory molecules upon stress exposure, which is modulated by HspB4/αA-crystallin. Primary Müller cells (HspB4−/−) were transfected with either empty vector (EV), wild-type HspB4/αA-crystallin (WT), phosphomimetic form of HspB4/αA-crystallin (T148D), or non-phosphorylatable form of HspB4/αA-crystallin (T148A). Then, 24 h post transfection, cells were either exposed to normal media (10% FBS), serum starvation (no FBS), high glucose (25 mM glucose), or to diabetic-like stress (25 mM glucose + 100 ng/mL TNFα) for 4 h. The mRNA levels of (A) interleukin-1 beta, (B) interleukin-6, (C) interleukin-18, and (D) monocyte chemoattractant protein-1 were normalized to the actin-encoding gene Actb. (E) Representative immunoblot depicting the expression of HspB4/αA-crystallin in transfected MGCs (HspB4−/−). ** p ≤ 0.01, *** p ≤ 0.001, significantly different from the respective EV-transfected cells. The primary Müller cells (HspB4−/−) were used up to passage 4 for the experiments. Each endpoint was measured on a minimum of six technical replicates in three independent experiments. Statistical analysis was performed by one-way ANOVA followed by Student–Newman–Keuls test.

Figure 4

HspB4/αA-crystallin distinct interaction with inflammatory pathways might play key role in regulating neuro-inflammation in MGCs. Primary Müller cells (HspB4−/−) were transfected with either empty vector (EV), wild-type HspB4/αA-crystallin (WT), phosphomimetic form of HspB4/αA-crystallin (T148D), or non-phosphorylatable form of HspB4/αA-crystallin (T148A). Then, 24 h post transfection, cells were either exposed to normal media (10% FBS), serum starvation (no FBS), high glucose (25 mM glucose), or to diabetic like stress (25 mM glucose + 100 ng/mL TNFα) for 4 h. The mRNA levels of (A) NOD-, LRR-, and pyrin-domain-containing protein 3 (Nlrp3); (B) caspase-1 (Casp-1); (C) apoptosis-associated speck like protein containing a caspase recruitment domain (Asc); and (D) nuclear factor-kappa B-1 (NF-kB1) were normalized to the actin-encoding gene Actb. ** p ≤ 0.01, *** p ≤ 0.001, significantly different from the respective EV-transfected cells. The primary Müller cells (HspB4−/−) were used up to passage 4 for the experiments. Each endpoint was measured on a minimum of six technical replicates. Statistical analysis was performed by one-way ANOVA followed by Student–Newman–Keuls test.

Figure 5

HspB4/αA-crystallin regulates neuro-inflammation in MGCs by inhibiting NLRP3 induction and NF-kB activation. Primary Müller cells (HspB4−/−) were transfected with either empty vector (EV), wild-type HspB4/αA-crystallin (WT), phosphomimetic form of HspB4/αA-crystallin (T148D), or non-phosphorylatable form of HspB4/αA-crystallin (T148A). Then, 24 h post transfection, cells were exposed to either serum starvation (no FBS) or diabetic-like stress (25 mM glucose + 100 ng/mL TNFα) for 4 h. For the inflammasome, NLRP3 subcellular localization was assessed in cytosolic and nuclear fractions and normalized to the empty vector (A). Representative images of immunoblot signal for NLRP3, HspB4/αA-crystallin as well as GAPDH and histone-H3, which served as the cytosolic (CF) and nuclear fraction (NF) controls, respectively (B). For NF-kB activation, total lysate levels were analyzed for total and phosphorylated NF-kB (Ser536), HspB4/αA-crystallin, and actin, the latter being used as loading control (representative images in (C). Fold change of phosphorylated NF-kB (Ser536) (D), total NF-kB (E), and ratio of phosphorylated NF-kB (Ser536) to total NF-kB (F) are shown normalized to the empty vector. * p ≤ 0.05, ** p ≤ 0.01, significantly different from the respective EV-transfected cells. The primary Müller cells (HspB4−/−) were used up to passage 4 for the experiments. Each endpoint was measured on a minimum of three technical replicates. Statistical analysis was performed by one-way ANOVA followed by Student–Newman–Keuls test.