Circadian factors BMAL1 and RORα control HIF-1α transcriptional activity in nucleus pulposus cells: implications in maintenance of intervertebral disc health

Abstract

BMAL1 and RORα are major regulators of the circadian molecular oscillator. Since previous work in other cell types has shown cross talk between circadian rhythm genes and hypoxic signaling, we investigated the role of BMAL1 and RORα in controlling HIF-1-dependent transcriptional responses in NP cells that exist in the physiologically hypoxic intervertebral disc. HIF-1-dependent HRE reporter activity was further promoted by co-transfection with either BMAL1 or RORα. In addition, stable silencing of BMAL1 or inhibition of RORα activity resulted in decreased HRE activation. Inhibition of RORα also modulated HIF1α-TAD activity. Interestingly, immunoprecipitation studies showed no evidence of BMAL1, CLOCK or RORα binding to HIF-1α in NP cells. Noteworthy, stable silencing of BMAL1 as well as inhibition of RORα decreased expression of select HIF-1 target genes including VEGF, PFKFB3 and Eno1. To delineate if BMAL1 plays a role in maintenance of disc health, we studied the spinal phenotype of BMAL1-null mice. The lumbar discs of null mice evidenced decreased height, and several parameters associated with vertebral trabecular bone quality were also affected in nulls. In addition, null animals showed a higher ratio of cells to matrix in NP tissue and hyperplasia of the annulus fibrosus. Taken together, our results indicate that BMAL1 and RORα form a regulatory loop in the NP and control HIF-1 activity without direct interaction. Importantly, activities of these circadian rhythm molecules may play a role in the adaptation of NP cells to their unique niche.

Keywords: BMAL1; HIF-1; Pathology Section; RORα; intervertebral disc; nucleus pulposus.

Figure 1. Expression analysis of BMAL1 and other related factors in NP cells

A. Immunohistochemical localization of BMAL1 in rat intervertebral disc. Sagittal sections of the mature rat intervertebral disc, immunostained with BMAL1 antibody, showed prominent nuclear expression in NP tissue. B. Western blot analysis of BMAL1 and RORα expression in NP tissues isolated from three rats showed positive expression for both the proteins. C. qRT-PCR analysis of BMAL-1 and RORα mRNA expression from NP and AF tissues from rat discs (n=3 animals/group) D. qRT-PCR analysis of BMAL1, RORα, ARNT, ARNT2, ARNTL2 and CLOCK expression in rat NP cells cultured under hypoxia (1% O₂). None of the genes showed significant increase in hypoxia. E. Western blot analysis of BMAL1 and RORα in NP cells cultured under hypoxia. F., G. Densitometric analysis of multiple blots shown in (E) above. No significant differences were seen between normoxic and hypoxic levels of BMAL1 and RORα. Data is represented as mean ± SE, n=3, p<0.05.

Figure 2. BMAL1 synergizes HIF-1 dependent HRE activity in NP cells

A., B. Evaluation of HRE activity in NP cells transfected with ARNT, ARNT2, and BMAL1 (150 ng/each) with or without 25 ng of HIF-1α following 24 h culture in hypoxia or normoxia. Co-transfection of BMAL1 and ARNT but not ARNT2 with HIF-1α significantly increased the activity of HRE compared to HIF-1α alone. C.-F. Evaluation of HRE activity in NP cells co-transfected with increasing dose of BMAL1(C, D) or ARNT (E, F) with 25 ng of HIF-1α, after 24 h hypoxia and normoxia treatment. The HRE activity showed a significant increase at highest dose of BMAL1 and ARNT compared to HIF-1 alone. Data is represented as mean ± SE, n=3, * p<0.05.

Figure 3. BMAL1 controls HIF-1 activity without affecting HIF-α-TAD function

A. Evaluation of BMAL1 and RORα expression by Western blot in NP cells stably transduced with lentivirus expressing BMAL1 shRNA. B., C. Densitometric analysis of multiple blots shown in (A) demonstrated decreased BMAL1 (B) and RORα (C) expression in BMAL1-silenced cells under both normoxia and hypoxia. D. HRE activity of BMAL1-silenced NP cells is significantly lower than cells transduced with control shRNA. E.-G. Evaluation of BMAL1 control of HIF-α-TAD function. BMAL1 overexpression had no effects on HIF-1α-C-TAD (E), HIF-1α-N-TAD (F), as well as HIF-2α-TAD (G) regardless of oxygen tension. Data is represented as mean ± SE, n= 3, * p<0.05.

Figure 4. RORα controls HIF transcriptional activity and TAD function in NP cells

A., B. Evaluation of HRE activity in NP cells transfected with 50-150 ng of RORα with or without 25 ng of HIF-1α, after 24 h hypoxia and normoxia treatment. Under normoxia, compared to HIF-1α alone, cells co-transfected with RORα and HIF-1α showed a significant increase in HRE activity at all doses, while under hypoxia this increase was seen at 150 ng of RORα. C., D. Treatment of NP cells with increasing dose of highly specific, small molecular RORα inhibitor, ML176 (2-10 μM), resulted in decrease in activity of endogenous HIF protein under both normoxia (C) and hypoxia (D). E.-G. Regulation of HIF-α-TAD function by RORα in NP cells. Treatment with ML176 (10 μM) under both normoxia and hypoxia resulted in decrease in HIF-1α-C-TAD (E), HIF-1α-N-TAD (F) as well as HIF-2α-TAD (G) activities. Data is represented as mean ± SE, n=3, *p<0.05.

Figure 5. BMAL1 and RORα do not bind to HIF-1α in NP cells

A. Immunoprecipitation (IP) of BMAL1 and CLOCK from NP cells cultured under normoxia or hypoxia for 24 h followed by Western blot analysis using anti-HIF-1α, anti-BMAL1 and anti-CLOCK antibodies. BMAL1 bound to CLOCK, but neither BMAL1 nor CLOCK bound to HIF-1α irrespective of oxygen tension. Preimmune rabbit IgG was used as a negative control for IP assays. B. Pulldown of HIF-1α did not show co-precipitation of BMAL1, CLOCK, or RORα. HC: heavy chain of IgG, nsp: non-specific (C) Pulldown of HIF-1α showed co- precipitation of HIF-1β/ARNT, association was higher under hypoxia. D., E. Treatment of NP cells with RORα inhibitor, ML-176 (10 μM), showed no effect on nuclear levels of HIF-1α. Densitometric analysis shown in (E) was performed on blots from 3 independent experiments, Data is represented as mean ± SE, * p < 0.05.

Figure 6. BMAL1 and RORα inhibition results in decreased expression of select HIF-1 target genes

A.-B. qRT-PCR analysis of mRNA expression of select HIF-1 target genes following stable silencing of BMAL1 (A) or pharmacological inhibition of RORα activity by ML176 (10 μM) treatment. Knockdown of BMAL1 resulted in significant decrease in VEGF-A, enolase 1 (Eno1), PFKFB3 as well as RORα under both normoxia and hypoxia. B. NP cells treated with RORα inhibitor showed decreased mRNA expression of BMAL1, VEGF, PFKFB3 and Eno1. Expression of PHD2, another known HIF-1 target gene, remained unaffected in BMAL1 silenced and ML176 treated NP cells. Data is represented as mean ± SE. n=5, *p<0.1.

Figure 7. BMAL1 knockout mice (KO) evidence compromised disc height and vertebral bone health

A. Representative μCT scans of lumbar spine (L1-L5) of a wild type and a KO littermate (10 week). Left panels show a coronal cut-plane through 3D reconstructions of the full lumbar segments of a WT and BMAL1 KO animal, showing changes in DHI. The right panels show a representative 3D reconstruction of ROI chosen for the bone trabecular morphometric analysis of each vertebral body showing changes in trabecular bone. The ROI contour only the outer boundary of the trabecular bone, excluding the cortical bone. B., C. The average of three disc height measurements (ventral, center and dorsal) (B) and disc height index (DHI) (C) were significantly lower in BMAL1-KO mice. D.-H. BMAL1 KO animals showed significant decreases in trabecular bone volume (BV) and total volume (TV) (D), bone volume fraction (BV/TV) (E), trabecular number (Tb.N) (F), and trabecular thickness (Tb. Th) (G), and a significant increase in trabecular separation (Tb. Sp) (H). I. Connectivity density (Conn. dens) was unaffected in KO animals. Data is represented as mean ± SE. n = 4 animals/genotype, 5 lumber vertebrae were measured/animal, *p<0.05.

Figure 8. BMAL1 null mice shows altered disc morphology

A.-D. Representative images of H&E and alcian blue staining of sagittal sections of lumbar disc from WT and KO mice. Compared to that of WT animals (A, B) NP tissue in null animals show decreased accumulation of matrix (lower matrix-to-cell ratio) (C, D), Scale bars: A, C: 50 μm, B, D: 10 μm. Disc height on images is outlined with dotted line. E.-H. Representative image of Picosirius red staining of WT and KO under bright field (E, G) and polarized light (F, H). Annulus fibrosus (AF) tissue in KO animals showed hyperplasia and increased thickening (white arrow). Scale bars: 100 μm.