Thyroid hormone and vitamin D regulate VGF expression and promoter activity

Abstract

The Siberian hamster (Phodopus sungorus) survives winter by decreasing food intake and catabolizing abdominal fat reserves, resulting in a sustained, profound loss of body weight. Hypothalamic tanycytes are pivotal for this process. In these cells, short-winter photoperiods upregulate deiodinase 3, an enzyme that regulates thyroid hormone availability, and downregulate genes encoding components of retinoic acid (RA) uptake and signaling. The aim of the current studies was to identify mechanisms by which seasonal changes in thyroid hormone and RA signaling from tanycytes might ultimately regulate appetite and energy expenditure. proVGF is one of the most abundant peptides in the mammalian brain, and studies have suggested a role for VGF-derived peptides in the photoperiodic regulation of body weight in the Siberian hamster. In silico studies identified possible thyroid and vitamin D response elements in the VGF promoter. Using the human neuroblastoma SH-SY5Y cell line, we demonstrate that RA increases endogenous VGF expression (P<0.05) and VGF promoter activity (P<0.0001). Similarly, treatment with 1,25-dihydroxyvitamin D3 increased endogenous VGF mRNA expression (P<0.05) and VGF promoter activity (P<0.0001), whereas triiodothyronine (T3) decreased both (P<0.01 and P<0.0001). Finally, intra-hypothalamic administration of T3 blocked the short day-induced increase in VGF expression in the dorsomedial posterior arcuate nucleus of Siberian hamsters. Thus, we conclude that VGF expression is a likely target of photoperiod-induced changes in tanycyte-derived signals and is potentially a regulator of seasonal changes in appetite and energy expenditure.

Keywords: SH-SY5Y cells; Siberian hamster; VGF (non-acronymic); thyroid hormone; vitamin D.

Figure 1

Treatment of the SH-SY5Y cell line with 10 μM RA reduces cell proliferation and increases differentiation. (A) Images of SH-SY5Y cells at days 0 and 5 of differentiation in the absence or presence of 10 μM RA, showing differences in cell numbers and morphology (neurite lengths). (B) Treatment of the SH-SY5Y cell line with 10 μM RA significantly decreased cell number (P<0.0001) and (C) significantly increased neurite length (a marker of differentiation) (P<0.0001). (D) Neurite length was significantly greater 5 days post-treatment with differentiation media in wells coated with poly-l-lysine or type IV collagen (P<0.0001) than plastic. (E) Treatment of the SH-SY5Y cell line with differentiation media significantly increased Map2, Tau and Gap43 (neuronal markers of differentiation). Gene expression was quantified by QPCR, normalized to cyclophilin A mRNA, and then compared to the normalized expression in undifferentiated cells. All values are means±s.e.m. (n=6, *P<0.05, **P<0.01, ***P<0.001, and **** P<0.0001).

Figure 2

Treatment of undifferentiated or differentiated SH-SY5Y cells with NGF, RA or 1,25D₃ increases endogenous VGF mRNA expression, whereas T₃ decreases endogenous VGF mRNA. VGF mRNA was significantly increased by treatment with 50 ng/ml NGF or 10 μM RA for 24 and 48 h in both (A) undifferentiated SH-SY5Y cells and (B) differentiated SH-SY5Y cells. VGF mRNA was significantly increased by treatment with 10 nM 1,25D₃ but significantly reduced with 10 nM T₃ in both (C) undifferentiated SH-SY5Y cells and (D) differentiated SH-SY5Y cells. All values are means±s.e.m. (n=6, *P<0.05, **P<0.01 for comparisons between control and treatment).

Figure 3

Treatment of undifferentiated or differentiated SH-SY5Y cells with NGF, RA or 1,25D₃ increases VGF promoter activity, whereas T₃ reduces VGF promoter activity. (A) The VGF promoter (1.1 kb) was cloned into a mammalian expression vector, based on the backbone of pZsGreen1-1 (Clontech), in which the GFP reporter gene was substituted for an mRFP. A subsequent truncated promoter construct (0.5 kb), which lacked the potential TRE and VDRE (indicated as *), was generated via 5′ deletion. (B) Promoter activities (fluorescence) were similar in cells transfected with either the 1.1 or 0.5 kb VGF promoter constructs. Promoter activities are shown relative to the positive control (pZsGreen1-N1, fluorescence set at 100%). VGF promoter activity (pVGF1.1 construct only) was increased by 50 ng/ml NGF and 10 μM RA in both (C) undifferentiated and (D) differentiated SH-SY5Y cells. (C) SH-SY5Y cells were transfected with pVGF1.1 and treated with NGF or RA 72 h post-transfection. 50 ng/ml NGF rapidly induced pVGF1.1 promoter activity within 1 h (P<0.0001). Ten micromolar RA resulted in a slower, yet significant, increase in pVGF1.1 promoter activity, starting 6 h post-treatment (P<0.001). (D) Undifferentiated cells were transfected as described in (B), but 72 h post-transfection, cells were differentiated with 10 μM RA for 5 days. Differentiated transfected cells were then treated with 50 ng/ml NGF or 10 μM RA for 48 h. Treatment of differentiated SH-SY5Y cells with NGF (P<0.0001) and RA (P< 0.0001) had a similar effect to that observed in transfected undifferentiated cells. VGF promoter activity (pVGF1.1 construct only) was decreased by 10 nM T₃ in both (E) undifferentiated and (F) differentiated SH-SY5Y cells, but there were no effects on the pVGF0.5 promoter construct (which lacked the potential TRE). Cells were transfected and differentiated as described for (B) and (D) respectively, but cells were pre-treated with DMEM/F12 complete containing 50 ng/ml NGF for 1 h (to induce promoter activity) prior to addition of 10 nM T₃, which significantly reduced pVGF1.1 promoter activity (P<0.001) in both (E) undifferentiated and (F) differentiated cells. Similarly, treatment with 10 nM 1,25D₃ significantly increased pVGF1.1 promoter activities (P<0.0001) in both (G) undifferentiated and (H) differentiated cells, but there were no effects on the pVGF0.5 promoter construct (which lacked the potential VDRE). pRFP-basic was included as a negative control in all experiments to indicate background fluorescence, as this construct lacks a functional promoter. All values are means±s.e.m. (n=6).

Figure 4

Intra-hypothalamic T₃ administration reduces VGF mRNA expression in the SD Siberian hamster. (A) representative photomicrographs of coronal sections through the dmpARC counterstained with cresyl violet, VGF hybridization is revealed by dark silver grains in the overlying emulsion in Siberian hamsters exposed to LD or SD receiving intra-hypothalamic sham or T₃ implants for 8 weeks. Dotted line indicates approximate boundaries of the dmpARC, arrow indicates induces expression in a SD sham hamster, scale bar=100 μm. (B) analysis of VGF mRNA abundance, scores for individual animals are depicted; ^aP<0.001 vs LD sham group, ^bP<0.05 vs SD sham group. (C) overall change in body weight, values are mean±s.e.m., **P<0.01 vs LD-sham group. (D) individual paired testis weights at the end of the study, ***P<0.001 vs LD-sham group. Weekly mean body weight data and group mean testis weight data have been published previously (Barrett et al. 2007). Circled values (panels (B) and (D)) are data from the same individual.

Figure 5

Schematic summary of the proposed photoneuroendocrine control of VGF expression. Thyroxine (T₄) is taken up from the circulation into tanycytes via MCT8 transporters, and in LD is converted by DIO2 to T₃ (Ebling 2014). Furthermore, in LD, components of the retinoic acid-signaling pathway – CRBP-1, CRABP-2, RAR, and RXR – are all upregulated in the Siberian hamster (Ross et al. 2005, Barrett et al. 2006), while RALDH-1 is increased in the photoperiodic rat (Shearer et al. 2010). We demonstrate that VGF expression and promoter activity in SH-SY5Y cells is increased in response to treatment with RA and vitamin D and reduced in response to treatment with T₃. Furthermore, VGF mRNA expression is reduced in response to intra-hypothalamic T₃ administration in the SD-exposed Siberian hamster. In SD, expression of DIO3 is upregulated, and thus, inactive metabolites of T₄ such as rT₃ and T₂ are produced alongside reductions in components of the retinoic acid-signaling pathway, while vitamin D plasma levels are increased. This may account for the increase in VGF expression in the dmpARC whilst reducing VGF expression in the ARC. Adapted from Ebling (2014).