Histone deacetylase 6 gates the synaptic action of acute stress in prefrontal cortex

Abstract

The prefrontal cortex (PFC), a region responsible for high-order cognitive functions, such as decision-making, attention and working memory, is highly influenced by stress and corticosteroid stress hormones. Recently it has been shown that acute stress affects PFC functions by potentiating glutamatergic transmission via a mechanism dependent on glucocorticoid receptor (GR) and its downstream target, serum and glucocorticoid-inducible kinase (SGK). To identify the key regulators of stress responses, we examined the role of histone deacetylase 6 (HDAC6), a unique member of the HDAC family that could regulate the GR chaperone protein heat shock protein 90 (HSP90), in the synaptic action of acute stress in PFC. We found that HDAC6 inhibition or knockdown blocked the enhancement of glutamatergic transmission and glutamate receptor trafficking by acute stress in vivo or corticosterone treatment in vitro. In addition, HDAC6 inhibition blocked the up-regulation of SGK in animals exposed to acute stress. HSP90 inhibition or knockdown produced a similar blockade of the acute stress-induced enhancement of glutamatergic signalling. These findings have identified HDAC6 as a key molecule gating the effects of acute stress on synaptic functions in the PFC.

Figure 1. HDAC6 inhibition blocks the enhancement of AMPA receptor-mediated current by acute stress or corticosterone treatment

A, input–output curves of AMPAR-EPSC evoked by a series of stimulus intensities in PFC pyramidal neurons from control vs. animals exposed to a forced-swim stress with or without i.p. injection of trichostatin A (TSA, a pan-HDAC inhibitor, injected at 1 h prior to stress). *P < 0.01, ANOVA. B, dot plots showing the amplitude of AMPAR-EPSC evoked by the same stimulus in PFC pyramidal neurons from control vs. stressed animals with or without i.p. injection of TSA (0.5 mg kg⁻¹) or trapoxin A (TpxA, 0.3 μg kg⁻¹, a non-HDAC6 inhibitor). *P < 0.01, ANOVA. Inset: representative synaptic current traces. Scale bar: 50 pA, 20 ms. C, bar graphs (mean ± SEM) showing the mEPSC amplitude and frequency in cultured cortical neurons treated with DMSO or corticosterone (20 min; 100 nm) in the absence or presence of tubacin (20 μm, a specific HDAC6 inhibitor), TSA (1 μm) or TpxA (10 nm). *P < 0.01, ANOVA. D, representative mEPSC traces. Scale bar: 50 pA, 2 s.

Figure 2. HDAC6 knockdown blocks the enhancement of glutmatergic signalling by corticosterone treatment or acute stress

A, immunoblots of HDAC6 expression in cortical cultures infected with HDAC6 shRNA lentivirus or a scrambled control shRNA lentivirus. B, bar graphs (mean ± SEM) showing the mEPSC amplitude and frequency in GFP- or HDAC6 shRNA-transfected cortical cultures treated with DMSO or corticosterone (20 min, 100 nm). *P < 0.01, ANOVA. C and D, cumulative plots of mEPSC amplitude distribution and mEPSC traces in representative neurons under different conditions. Scale bar: 50 pA, 2 s. E, confocal images of neurons from animals with stereotaxic injection of GFP or HDAC6 shRNA lentivirus to the PFC region. F, dot plots showing the amplitude of AMPAR-EPSC in PFC pyramidal neurons from control vs. stressed animals with stereotaxic injection of GFP or HDAC6 shRNA lentivirus (injected at 1 week prior to stress). *P < 0.01, ANOVA. Inset: representative synaptic current traces. Scale bar: 50 pA, 20 ms.

Figure 3. Inhibition of HDAC6 blocks the up-regulation of surface and synaptic glutamate receptors by acute stress or corticosterone treatment

A, immunoblots and quantification analysis of surface and total GluR1 and GluR2 in lysates of PFC slices from control vs. stressed animals with or without TSA injection (0.5 mg kg⁻¹, i.p.). *P < 0.01, ANOVA. B, immunocytochemical images and quantitative analysis of synaptic GluR1 (PSD-95 colocalized, yellow puncta), total GluR1 clusters (green puncta) and PSD-95 clusters (red puncta) along dendrites of PFC cortical cultures treatment with DMSO or corticosterone (20 min, 100 nm) in the absence or presence of tubacin (20 μm, added 1 h beforehand). *P < 0.01, ANOVA.

Figure 4. HDAC6 inhibition blocks the acute stress-induced upregulation of SGKs

A and B, immunoblots (A) and quantitative analysis (B) of SGK1 and SGK3 expression in PFC slices from control vs. stressed animals with or without i.p. injection of TSA (0.5 mg kg⁻¹) or TpxA (0.3 μg kg⁻¹). *P < 0.01, ANOVA.

Figure 5. HSP90 inhibition or knockdown blocks the acute stress-induced regulation of AMPARs and SGK

A, dot plots showing the amplitude of AMPAR-EPSC in PFC pyramidal neurons from control or stressed animals with or without i.p. injection of 17-DMAG (30 mg kg⁻¹, a brain-permeant HSP90 inhibitor) or geldanamycin (2 mg kg⁻¹, a non-permeant HSP90 inhibitor). *P < 0.01, ANOVA. Inset: representative synaptic current traces. Scale bar: 50 pA, 20 ms. B, representative Western blots in HEK293 cells transfected with Flag-tagged rat HSP90 in the absence or presence of a HSP90 shRNA. HSP90 was detected with anti-Flag. C, bar graphs (mean ± SEM) showing the mEPSC amplitude and frequency in GFP- or HSP90 shRNA-transfected cortical cultures treated with DMSO or corticosterone (20 min, 100 nm). *P < 0.01, ANOVA. D, immunoblots and quantification analysis of surface and total GluR1 and GluR2 in PFC slices from control vs. stressed animals with or without 17-DMAG injection (30 mg kg⁻¹, i.p.). *P < 0.01; #P < 0.05, ANOVA. E, immunoblots and quantitative analysis of SGK1 and SGK3 expression in PFC slices from control vs. stressed animals with or without i.p. injection of 17-DMAG (30 mg kg⁻¹). #P < 0.05, ANOVA.