Phospho-CREB Regulation on NMDA Glutamate Receptor 2B and Mitochondrial Calcium Uniporter in the Ventrolateral Periaqueductal Gray Controls Chronic Morphine Withdrawal in Male Rats

Abstract

Opioid use disorder (OUD) is a chronic disease of the brain, and it currently continues at crisis proportions in the United States. Opioid physical withdrawal is a major driver of compulsive drug-taking behavior, triggering short-term relapse of opioid addiction. Early pharmacological evidence shows that midbrain periaqueductal gray (PAG) plays an important role in morphine withdrawal (MW). However, we still know few details about the underlying molecular mechanisms. Improving our understanding of such mechanisms will enable increasingly safe and effective treatments for patients with OUD. Here, MW was induced by the naloxone precipitation after chronic intraperitoneal administration of morphine for a period of 5 d in Sprague Dawley male rats. MW increased phosphorylation of cAMP response element binding protein (pCREB, a primary marker of CREB functional activation), NMDA glutamate receptor 2B subunit (NR2B), and mitochondrial calcium uniporter (MCU) within the ventrolateral PAG (vlPAG). Inhibition of pCREB, NR2B, or MCU within this brain region reduced the severity of MW. Chromatin immunoprecipitation (ChIP) assay and luciferase report assay demonstrated that pCREB mediated the transcription of the Grin2b (glutamate ionotropic receptor NMDA type subunit 2B, encoding NR2B) and Ccdc109a (encoding MCU) genes. These findings describe the role of pCREB in Grin2b and Ccdc109a gene transcription levels in the vlPAG during MW. The study may provide a novel therapeutic approach for OUD.

Keywords: MCU; NR2B; morphine withdrawal; pCREB; vlPAG.

Figure 1.

The role of pCREB in the vlPAG in MW rats. A, Scheme of the morphine treatment regimen used for MW behavioral tests. Animals were given escalating doses of morphine (10–50 mg/kg, i.p.) for a period of 5 d. On Day 5, animals received a morning injection of 50 mg/kg, and 1 h later, naloxone (4 mg/kg, i.p.) was administered to precipitate MW. One hour after naloxone, brains were harvested (B.H.) for neurochemical tests. M, morphine; nal, naloxone. B, Rats showed higher MW behavioral global scores in MW group compared with control group, p = 0.0012, Mann–Whitney test, n = 6–7. C, Western blots showed that MW did not significantly change total CREB expression, t₍₆₎ = 0.4768, p = 0.650, two-tailed t test, n = 4. D, MW boosted expression of pCREB in the vlPAG compared with control, t₍₈₎ = 5.622, p = 0.0005, two-tailed t test, n = 5. E, Brain block section at the level of the PAG in rats, the location of vlPAG sample collection and representative immunostaining image for pCREB in the vlPAG. Immunoreactivity of pCREB colocalized predominantly with NeuN, but not Iba1 nor GFAP; scale bar, 50 µm. F–H, To examine the effect of naloxone on CREB or pCREB in control group, rats received saline for 5 d; on Day 5 rats were injected either naloxone or vehicle (saline). Naloxone did not change MW-like behavior, p = 0.9048, Mann–Whitney test, n = 6 (F). Western blots showed that there was no significant difference between vehicle and naloxone in the expression of CREB, t₍₈₎ = 0.6407, p = 0.5396, two-tailed t test, n = 5 (G), or pCREB, t₍₈₎ = 0.1184, p = 0.9087, two-tailed t test, n = 5 (H). I, J, To further test the effect of naloxone on CREB or pCREB induced by CREB plasmid in in vitro study, cultured neuronal B35 cells were treated with vehicle or naloxone (see details in Materials and Methods). I, The effect of naloxone on expression of CREB in Western blots, F_(3,8) = 17.32, p = 0.0007, one-way ANOVA, n = 3. J, The effect of naloxone on expression of pCREB in Western blots, F_(3,8) = 7.453, p = 0.0105, one-way ANOVA, n = 3. In vehicle groups, cells treated with plasmid expressing CREB (pl-CREB/veh group) increased the level of CREB, p = 0.0012 (I), or pCREB, p = 0.0257 (J) compared with pl-veh/veh group, one-way ANOVA Holm–Sidak's multiple-comparisons test, n = 3. However, there is no significant difference in CREB, p = 0.1518 (I), or pCREB, p = 0.9862 (J) between pl-CREB/vehicle, and pl-CREB/naloxone, one-way ANOVA Holm–Sidak's multiple-comparisons test, n = 3.

Figure 2.

The effect of CREB knockdown in the rat vlPAG on MW. A, Representative photomicrograph of a coronal brainstem section showing a bilateral microinjection track (with guide cannula and injection cannula) for microinjection of antisense oligodeoxynucleotide (ODN) against CREB (AS-CREB, 2 µg, 0.5 µl, once a day) or mismatch oligodeoxynucleotide (mmO) to the vlPAG, A, aqueduct of Sylvius; the scheme of the injection regimen of oligodeoxynucleotide (ODN) in MW. B, The effect of AS-CREB on expression of CREB in Western blots, F_(2,11) = 4.619, p = 0.035, one-way ANOVA, n = 4–5; microinjection of antisense ODN against CREB (AS-CREB) into the vlPAG lowered expression of CREB in MW compared with mmO/MW, p = 0.023, one-way ANOVA Holm–Sidak's multiple-comparisons test, n = 4–5. C, The effect of AS-CREB on expression of pCREB in Western blots, F_(2,11) = 27.89, p < 0.0001, one-way ANOVA, n = 4–5; microinjection of AS-CREB reversed pCREB upregulation induced by MW, p < 0.0001, one-way ANOVA Holm–Sidak's multiple-comparisons test, n = 4–5. D, E, To further test the effect of AS-CREB on CREB or pCREB induced by CREB plasmid in in vitro study, cultured neuronal B35 cells were treated with mismatch ODN or AS-CREB (see details in Materials and Methods). D, The effect of AS-CREB on expression of CREB in B35 cells in Western blots, F_(2,6) = 15.73, p = 0.0041, one-way ANOVA, n = 3. E, The effect of AS-CREB on expression of pCREB in B35 cells in Western blots, F_(2,6) = 16.17, p = 0.0038, one-way ANOVA, n = 3. In mmODN groups, cells treated with plasmid expressing CREB (pl-CREB/mmO) increased the level of (D) CREB (p = 0.003) or (E) pCREB (p = 0.0036) compared with pl-veh/mmO group one-way ANOVA Holm–Sidak's multiple-comparisons test, n = 3. AS-CREB treatment reversed the upregulation of (D) CREB (p = 0.008) or (E) pCREB (p = 0.0045), one-way ANOVA Holm–Sidak's multiple-comparisons test, n = 3. F, The effect of AS-CREB or mmO microinjected into the vlPAG on MW behavioral tests, F_(3,15) = 36.28, p < 0.0001, two-way ANOVA, n = 6–7. AS-CREB/MW significantly reversed the increased the global scores compared with mmO/MW, two-way ANOVA multiple-comparisons test, n = 6–7.

Figure 3.

The role of NR2B in the vlPAG in rat MW. A, Western blots showed that MW increased NR2B expression (t₍₈₎ = 3.910, p = 0.0045, two-tailed t test, n = 5). B, Representative immunostaining image for NR2B in the vlPAG. Immunoreactivity of NR2B predominantly colocalized with NeuN, but not Iba1 or GFAP; scale bar, 50 µm. C, The scheme of the injection regimen of Ro25-6981 (Ro25) and MW. In control groups, veh or Ro25 (5 µg, 0.5 µl, once a day) was microinjected into the vlPAG 15 min before every morning morphine injection for 5 d. M, morphine; nal, naloxone. D, The effect of Ro25 on MW global scores, F_(3,13) = 63.83, p < 0.0001, two-way ANOVA, n = 5–7. There was an increase in MW global scores in rats with veh/MW compared with veh/cont, p < 0.0001; there was a significant decrease in MW scores in rats with Ro25/MW compared with that in veh/MW, p < 0.0001, two-way ANOVA multiple-comparisons test, n = 5–7.

Figure 4.

The role of MCU in the vlPAG in MW in rats. A, Western blots showed that MW upregulated MCU expression (t₍₈₎= 3.531, p = 0.0077, two-tailed t test, n = 5). B, Double immunostaining showed that MCU immunoreactivity was colocalized with NeuN, but not GFAP or Iba1; scale bar, 50 µm. C, The scheme of the injection regimen of Ru360 and MW. Ru360 (5 µM, once a day) was microinjected into the vlPAG 15 min before every morning morphine injection for 5 d. M, morphine; nal, naloxone. D, The effect of Ru360 on MW global scores, F_(3,13) = 53.8, p < 0.0001, two-way ANOVA, n = 5–6. There was an increase in MW global scores in rats with veh/MW compared with veh/sal, p < 0.0001; there was a decrease in the global scores in rats with Ru360/MW compared with that in veh/MW group, p < 0.0001, two-way ANOVA multiple-comparisons test, n = 5–6.

Figure 5.

Transfection of CREB by HSV vectors into B35 cells increased expression of CREB, pCREB, NR2B, or MCU. We incubated the cultured B35 neuronal cells (Yi et al., 2018) with HSV-GFP or HSV-CREB (MOI = 0.1) for 24 h, and then cells were collected for neurochemical analysis (A). We found that inoculation with HSV-CREB significantly upregulated the expression of (B) CREB (t₍₁₀₎ = 3.959, p = 0.0027, two-tailed t test, n = 6), (C) pCREB (t₍₁₀₎ = 7.786, p < 0.0001, two-tailed t test, n = 6), (D) NR2B (t₍₁₀₎ = 4.931, p = 0.0006, two-tailed t test, n = 6), and (E) MCU (t₍₁₀₎ = 4.782, p = 0.0007, two-tailed t test, n = 6).

Figure 6.

HSV-mediated CREB overexpression in the vlPAG of naive rats induced MW-like responses. A, To determine if overexpression of CREB induces MW-like behavior in rats, we microinjected HSV vectors expressing GFP or CREB (0.5 µl, 3 × 10⁷ pfu/ml) into the naive rat vlPAG. The scheme of the injection regimen is shown. HSV-GFP or HSV-CREB vector was microinjected into the vlPAG. Two days after HSV, naloxone (nal, 4 mg/kg, IP) was injected. MW-like behavior was observed for 30 min. B, HSV-CREB induced the global scores, p = 0.0498, Mann–Whitney test, n = 5–6. One hour after naloxone, the brain was harvested (B.H.) for Western blotting. Western blots demonstrated that HSV-CREB significantly increased protein expression of (C) CREB (t₍₈₎ = 3.433, p = 0.0089, two-tailed t test, n = 5), (D) pCREB (t₍₈₎ = 6.550, p = 0.0002, two-tailed t test, n = 5), (E) NR2B (t₍₆₎ = 5.776, p = 0.0012, two-tailed t test, n = 4), and (F) MCU (t₍₈₎ = 2.877, p = 0.0206, two-tailed t test, n = 5).

Figure 7.

pCREB mediated Grin2b transcription by binding to a CRE site in its promoter region. A, The sequence alignment of rat Grin2b gene promoter regions and a putative CRE site (Alexandre et al., 1991; Manna et al., 2009) based on homology with the mouse sequence (Klein et al., 1998). The CRE site was found at −413 to −420 bp from the putative TSS (+1, Chr 4, 169999938, chromosome 4 NCBI Reference Sequence: NC_005103.4 Rnor_6.0 Primary Assembly). ChIP-qPCR primer areas on the Grin2b gene are shown as blue. B, Grin2b gene promoter luciferase reporter plasmid, in conjunction with a CREB overexpression plasmid (pl-CREB), was cotransfected into B35 cells for 30 h. The luciferase assay demonstrated that the overexpression of CREB significantly increased NR2B luciferase activity compared with control (pl-vehicle) group, t₍₆₎= 12.6, p < 0.0001, two-tailed t test, n = 4. C, Map of luciferase reporter plasmids of the Grin2b promoter CRE region (WT) and CRE region mutation (Mut) (TGACGTGA to AGTCTCGA). D, DNA Sequencing confirmed the CREB-binding site in Grin2b promoter plasmid as well as the Grin2b promoter mutant refers to CREB-binding sites. E, The plasmids were transfected into the B35 cells using Lipofectamine 3000 Transfection Reagent (Miyatake et al., 2002). Luciferase reporter assay showed that the mutation of CRE sites (Mut) significantly decreased NR2B luciferase activity compared with control, t₍₆₎ = 3.122, p = 0.0205, two-tailed t test, n = 4. F, The effect of AS-CREB on enrichment of pCREB at the Grin2b gene promoter region, F_(3,12) = 6.614, p = 0.0069, one-way ANOVA, n = 4. ChIP-qPCR assay showed that enrichment of pCREB in the mismatch oligodeoxynucleotide (mmO)+morphine withdrawal (mmO/MW) group was increased compared with the mmO/cont group at the Grin2b gene promoter CRE region, p = 0.0055; enrichment of pCREB in the AS-CREB/MW group was lower than that in the mmO/MW group on Grin2b gene promoter at region, p = 0.0077, one-way AVONA with Holm–Sidak's test, n = 4. G, The effect of AS-CREB on NR2B mRNA expression, F_(3,19) = 5.309, p = 0.0079, one-way ANOVA, n = 5–6. RT-PCR revealed that Grin2b mRNA in the mmO/MW group was higher than in the mmO/cont group, p = 0.0088; Grin2b mRNA in the AS-CREB/MW group was lower than that in the mmO/MW group, p = 0.0064, one-way ANOVA with Holm–Sidak's test, n = 5–6. H, The effect of AS-CREB on NR2B protein expression, F_(2,11) = 11.57, p = 0.002, one-way ANOVA, n = 4–5. NR2B expression in the mmO/MW group was higher than that in the mmO/cont group, p = 0.002; NR2B in the AS-CREB/MW group was lower than that in the mmO/MW group, p = 0.002, one-way ANOVA with Holm–Sidak's test, n = 4–5. I, Double label immunostaining showed that pCREB immunoreactivity was colocalized with NR2B immunoreactivity; scale bar, 50 μm.

Figure 8.

pCREB mediated Ccdc109a transcription by binding to several CRE sites in its promoter region. A, We analyzed the sequence of rat Ccdc109a gene promoter region. The base C (29199224 at chromosome 20, NCBI Reference Sequence: Rnor_6.0 Primary Assembly, NC_005119.4) of the first intron of Ccdc109a gene was referred to as number +1. We found four putative CRE sites (Ikuyama et al., 1992; Taupenot et al., 1998; Berhane and Boggaram, 2001), Region 1 (TGACGTAA, −3,304 to −3,311), Region 2 (TGAGGTCA, −2,878 to −2,885), Region 3 (TGAGGTCA, −1,117 to −1,124), and Region 4 (TGAGGTCA, −908 to −915). The TATA box of the Ccdc109a gene is shown. B, Motif of the CRE sites based on JASPAR data sets. C, The Mcu gene promoter luciferase reporter plasmid, along with a CREB overexpression plasmid (pl-CREB), was cotransfected into B35 cells for 30 h. The luciferase assay demonstrated that the overexpression of CREB resulted in a significant increase in MCU luciferase activity compared with the pl-vehicle group, t₍₆₎ = 11.76, p < 0.0001, two-tailed t test, n = 4. D, The map of luciferase reporter plasmids was shown in which the four CRE regions of Ccdc109a gene promoter were inserted. E, Schematic of luciferase reporter constructs of the Ccdc109a promoter CRE sites in luciferase reporter constructs. The four putative CRE sites in the Ccdc109a gene promoter were mutated. F, DNA sequencing confirms the sequences of CREB-binding regions (R1–R4) in either normal (WT) or mutated (Mut) MCU promoter plasmid referred to CREB-binding sites. G, The effect of mutation of CRE binding sites at the Ccdc109a promoter region on expression of MCU luciferase, F_(4,14) = 113.5, p < 0.0001, one-way ANOVA, n = 3–4. B35 cells were transfected with WT or four different mutated plasmids of Ccdc109a-driven luciferase constructs along with CREB overexpression plasmid for 30 h. Then luciferase activity was detected by Luciferase Reporter Assay System. Ccdc109a-driven luciferase activity compared with the WT plasmid was lowered with R1-Mut (p < 0.0001), R2-Mut (p < 0.0001), R3-Mut (p < 0.0001), and R4-Mut (p < 0.0001), one-way AVONA with Holm–Sidak's test, n = 3–4.

Figure 9.

pCREB mediated Ccdc109a gene transcription and expression. A, The effect of AS-CREB on enrichment of pCREB at the Ccdc109a gene promoter region-1, F_(3,12) = 6.045, p = 0.0095, one-way ANOVA, n = 4. B, The effect of AS-CREB on enrichment of pCREB at the Ccdc109a gene promoter region-2, F_(3,12) = 11.06, p = 0.0009, one-way ANOVA, n = 4. C, The effect of AS-CREB on enrichment of pCREB at the Ccdc109a gene promoter region-3, F_(3,12) = 4.848, p = 0.0196, one-way ANOVA, n = 4. D, The effect of AS-CREB on enrichment of pCREB at the Ccdc109a gene promoter region-4, F_(3,12) = 8.173, p = 0.0031, one-way ANOVA, n = 4. A–D, ChIP-qPCR assay showed that enrichment of pCREB in the mmO/MW group was increased compared with the mmO/cont group at the Ccdc109a gene promoter region 1 (p = 0.0122), region 2 (p = 0.0009), region 3 (p = 0.0192), and region 4 (p = 0.0028); enrichment of pCREB in the AS-CREB/MW group was lower than that in the mmO/MW group on Ccdc109a gene promoter at region 1 (p = 0.0082), region 2 (p = 0.0085), region 3 (p = 0.019), and region 4 (p = 0.0181), one-way AVONA with Holm–Sidak's test, n = 4. E, The effect of AS-CREB on expression of MCU mRNA, F_(3,19) = 7.872, p = 0.0013, one-way ANOVA, n = 5–6. RT-PCR revealed that Ccdc109a mRNA in the mmO/MW group was higher than in the mmO/cont group (p = 0.002); Ccdc109a mRNA in the AS-CREB/MW group was lower than in the mmO/MW group, p = 0.0019, one-way ANOVA with Holm–Sidak's test, n = 5–6. F, The effect of AS-CREB on expression of MCU protein, F_(2,11) = 9.087, p = 0.0047, one-way ANOVA, n = 4–5. MCU expression in the mmO/MW group was higher than in the mmO/cont group, p = 0.0038; MCU in the AS-CREB/MW group was lower than that in the mmO/MW group, p = 0.0066, one-way ANOVA with Holm–Sidak's test, n = 4–5. Double label immunostaining showed that either pCREB (G) or NR2B (H) immunoreactivity was colocalized with MCU immunoreactivity; scale bar, 50 μm.

Figure 10.

Signaling pathways regulating pCREB-dependent transcription in neurons of the PAG in MW. MW induces pCREB in vlPAG neurons. As a transcriptional factor, pCREB binds to CREs present within the promoter regions of the Grin2b or Ccdc109a genes, mediating the transcription of Grin2b or Ccdc109a genes, and production of NR2B and MCU, contributing to MW behavior. Inhibition of pCREB, NR2B, or MCU within this brain vlPAG region, reduces the severity of MW.