Body temperature regulates glucose metabolism and torpid behavior

Abstract

Glucose is a significant energy resource for maintaining physiological activities, including body temperature homeostasis, and glucose homeostasis is tightly regulated in mammals. Although ambient temperature tunes glucose metabolism to maintain euthermia, the significance of body temperature in metabolic regulation remains unclear owing to strict thermoregulation. Activation of Qrfp neurons in the preoptic area induced a harmless hypothermic state known as Q-neuron-induced hypothermia and hypometabolism (QIH), which is suitable for studying glucose metabolism under hypothermia. In this study, we observed that QIH mice had hyperinsulinemia and insulin resistance. This glucose hypometabolic state was abolished by increasing the body temperature to euthermia. Moreover, QIH-mediated inappetence and locomotor inactivity were recovered in euthermia QIH mice. These results indicate that body temperature is considerably more powerful than ambient temperature in regulating glucose metabolism and behavior, and the glucose hypometabolism in QIH is secondary to hypothermia rather than modulated by Qrfp neurons.

Fig. 1. QIH reorganizes glucose homeostasis with a fasting independent property.

a Diagram of viral injection for chemogenetics. b Representative thermography showing BAT temperature (T_BAT) of QIH mice at different time points after clozapine injection. c Traces of T_BAT after clozapine injection (n = 9). d Traces of T_core after clozapine injection (n = 9). e Representative micrographs showing immunofluorescent cFos staining 24 h after clozapine injection in the control (left) and QIH (right) groups. Scale bar: 100 μm. f Quantification of cFos-expressing Qrfp neurons in the preoptic area from the control (n = 4) and QIH (n = 4; two-tailed t test, p < 0.0001) groups 24 h after clozapine injection. g Twenty-four-hour food intake of control (n = 5) and QIH (n = 5; two-tailed t test, p < 0.0001) mice after clozapine injection. h Experimental design for evaluating glucose metabolism of QIH mice. i Body weight (BW) of control (n = 9) and QIH (n = 9; two-way ANOVA followed by Sidak multiple comparison test, p = 0.0088 at time = 24 h) mice during fasting. j Blood glucose levels of control (n = 9) and QIH (n = 9; two-way ANOVA followed by Sidak multiple comparison test, p = 0.0002 at time = 18 h, p = 0.0032 at time = 24 h) mice during fasting. k Plasma insulin levels in control (n = 8) and QIH (n = 8; two-way ANOVA followed by Sidak multiple comparison test, p = 0.0478 at time = 6 h, p = 0.0138 at time = 24 h) mice during fasting. l–o respiratory-gas analysis of control (n = 5) and QIH (n = 5) mice during fasting. (l, n) traces of VO₂ (l) and VCO₂ (n) of (n = 5) and QIH (n = 5) during fasting. (m, o) Average last 3 h fasting period (fasted 21 h to 24 h) VO₂ (two-tailed t test, p = 0.0002) (m) and VCO₂ (two-tailed t test, p = 0.0003) (o) of (n = 5) and QIH (n = 5) mice. Gray blocks in (l) and (n) indicate the dark period. All data are presented as means ± SEM; *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001.

Fig. 2. Glucose hypometabolic state in QIH mice.

a Glucose tolerance test (GTT) of control (n = 8) and QIH (n = 9; two-way ANOVA followed by Sidak multiple comparison test, p = 0.0416 at time = 30, p = 0.001 at time = 60, p = 0.0303 at time = 120; two-tailed t test in area under the curve (AUC) during GTT, p = 0.0004) mice 18-h after clozapine injection and fasting. b Plasma insulin levels in control (n = 6) and QIH (n = 6; two-way ANOVA followed by Sidak multiple comparison test, p = 0.0448 at time= 0, p = 0.0319 at time = 30) mice after during first 30 mins of GTT. c Insulin tolerance test (ITT) in control (n = 8) and QIH (n = 8; two-way ANOVA followed by Sidak multiple comparison test, p = 0.0272 at time = 30, p = 0.0088 at time = 60) mice 6-h after clozapine injection and fasting. d Pyruvate tolerance test in control (n = 8) and QIH (n = 8; two-way ANOVA followed by Sidak multiple comparison test, p = 0.0819 at time= 0, p = 0.0413 at time = 30, p = 0.0108 at time = 60, p = 0.0467 at time = 120) mice 18-h after clozapine injection and fasting. e–g clearance of ¹³C₆ glucose in control and QIH mice after fasting. (e) blood glucose and comparative amount of ¹³C₆ glucose after tracer injection in control (n = 5) and QIH mice (n = 5; two-way ANOVA followed by Sidak multiple comparison test, p < 0.0001 at time = 15). f Rate of disappear of blood glucose during 3 to 15 min ¹³C₆ glucose postinjection (control, n = 5; QIH, n = 5; two-tailed t test, p < 0.0001). g Rate of appearance of blood glucose during 3 to 15 min ¹³C₆ glucose postinjection (control, n = 5; QIH, n = 5; two-tailed t test, p = 0.0001). h–q 2DG uptake with or without insulin stimulation in the soleus (h, i), BAT (j, k), heart (l, m), epididymal white adipose tissue (eWAT) (n, o), and brain cortex (p, q). (h, j, l, n, p) Absolute 2DG uptake with saline (control, n = 4; QIH, n = 4; two-tailed t test, p = 0.003 at (h) p = 0.0096 at (j) p = 0.0008 at (l), p = 0.0015 at (p)) or insulin (control, n = 4; QIH, n = 4; two-tailed t test, p = 0.0001 at (h) p < 0.0001 at (j) p < 0.0001 at (l) p = 0.001 at (n) p = 0.0015 at (p) injection in the soleus (h) BAT (j) heart (l) eWAT (n) and brain cortex (p). (i, k, m, o, q) Enhancement of 2DG uptake by insulin stimulation in the soleus (i) BAT (k) heart (m) eWAT (o) and brain cortex (q) of control (n = 4) and QIH (n = 4; two-tailed t test, p = 0.0007 at i, p = 0.0047 at (k), p = 0.0002 at (m), p < 0.0001 at (o) mice. All data are presented as means ± SEM; ns = not significant, *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001.

Fig. 3. QIH-induced diabetes-like metabolic states are regulated by body temperature.

a Experimental design for evaluating glucose metabolism of QIH mice in 34 °C. b Representative thermography showing BAT temperature of QIH and control mice 24 h after clozapine injection and fasting. c BAT temperature of QIH (n = 8) and control (n = 8) mice housed after 24-h fasting in 34 °C. d Tracer of core temperature after clozapine injection in 34 °C (n = 9). e Representative micrographs showing immunofluorescent cFos staining in the Qrfp^POA in control and QIH mice. Scale bar: 100 μm f Quantification of cFos expressing Qrfp^POA in control (n = 3) and QIH (n = 3; two-tailed t test, p = 0.0362) mice. g Representative thermography showing BAT temperature of control or QIH mice 0 h, 1 h, 2 h after being moved from 34 °C to 23 °C. h, i Traces of BAT temperature (h) and core temperature (i) in control (n = 8) and QIH (n = 8) mice during cooling. j Body weight changes in control (n = 7) and QIH (n = 6) mice after clozapine injection and food deprivation in 23 °C or 34 °C. k, l Comparisons of body weight change in control (n = 7) and QIH (n = 6) mice after 6-h (k) and 24-h (l) fasting in 23 °C or 34 °C(One-way ANOVA followed by Sidak multiple comparison test. For 6-h fasting: p = 0.0302, 34 °C Ctr vs. 23 °C QIH, p = 0.0037, 34 °C QIH vs. 23 °C QIH. For 24-h fasting: p = 0.0022, 34 °C Ctr vs. 23 °C Ctr, p < 0.0001, 34 °C QIH vs. 23 °C QIH, p < 0.0001, 23 °C Ctr vs. 23 °C QIH). m Blood glucose levels of control and QIH mice housed in 23 °C (control, n = 7; QIH, n = 6) and 34 °C (control, n = 7; QIH, n = 6) after clozapine injection and food deprivation. n–q Comparisons of blood glucose levels in control (n = 7) and QIH (n = 6) mice 0 h (n) 6 h (o) and 24 h (p) after clozapine injection and food deprivation in 23 °C or 34 °C. (One-way ANOVA followed by Sidak multiple comparison test. For 0-h fasting: p = 0.0061, 34 °C Ctr vs. 23 °C Ctr, p = 0.047, 34 °C QIH vs. 23 °C QIH. For 6-h fasting: p = 0.01, 34 °C Ctr vs. 23 °C Ctr, p = 0.0654, 23 °C Ctr vs. 23 °C QIH. For 24-h fasting: p = 0.0002, 34 °C QIH vs. 23 °C QIH, p = 0.0003, 23 °C Ctr vs. 23 °C QIH.) (q) Plasma insulin levels in control (n = 8) and QIH (n = 8; one-way ANOVA followed by Sidak multiple comparison test, p < 0.0001, 34 °C QIH vs. 23 °C QIH, p < 0.0001, 23 °C QIH vs. 23 °C Ctr) mice housed in 23 °C and 34 °C after 24-h fasting. r–u respiratory-gas analysis of control and QIH mice during fasting in 23 °C (control, n = 5; QIH, n = 5) or 34 °C (control, n = 6; QIH, n = 5). (r, t) traces of VO₂ (r) and VCO₂ (t) of control and QIH mice during fasting in 23 °C (control, n = 5; QIH, n = 5) or 34 °C (control, n = 6; QIH, n = 5). (s, u) Average VO₂ (s) and VCO₂ (u) of and QIH mice during last 3 h fasting period (fasted 21 h to 24 h) in 23 °C (control, n = 5; QIH, n = 5) or 34 °C (control, n = 6; QIH, n = 5)(One-way ANOVA followed by Sidak multiple comparison test, for VO2: p = 0.0001, 34 °C Ctr vs. 23 °C Ctr, p = 0.0191, 34 °C QIH vs. 23 °C QIH, p < 0.0001, 23 °C Ctr vs. 23 °C QIH). Gray blocks in (r) and (t) indicate the dark period. All data are presented as means ± SEM; ns = not significant, *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001.

Fig. 4. Increased body temperature abolishes QIH-mediated glucose hypometabolism.

a GTT in control and QIH mice housed in 23 °C (control, n = 8; QIH, n = 8) and 34 °C (control, n = 8; QIH, n = 8). b Area under the curve (AUC) of (a) (One-way ANOVA followed by Sidak multiple comparison test, p < 0.0001, 23 °C Ctr vs. 34 °C QIH, p < 0.0001, 23 °C Ctr vs. 34 °C QIH). c Plasma insulin levels of control (n = 6) and QIH (n = 6) mice during GTT in 34 °C. d, ITT in control and QIH mice housed in 23 °C (control, n = 8; QIH, n = 8) and 34 °C (control, n = 8; QIH, n = 8). e AUC of d (One-way ANOVA followed by Sidak multiple comparison test, p = 0.0099, 23 °C Ctr vs. 23 °C QIH, p = 0.0056, 23 °C QIH vs. 34 °C QIH). f–j 2DG uptake of the soleus (f), BAT (g), eWAT (h), heart (i), and brain cortex (j) of control and QIH mice housed in 23 °C (control, n = 6; QIH, n = 6) and 34 °C (control, n = 6; QIH, n = 5) after insulin injection (One-way ANOVA followed by Sidak multiple comparison test, Soleus: p < 0.0001, 23 °C Ctr vs. 23 °C QIH, p = 0.0002, 34 °C QIH vs. 23 °C QIH, BAT: p < 0.0001, 23 °C Ctr vs. 23 °C QIH, p < 0.0001, 34 °C Ctr vs. 23 °C Ctr, eWAT: p = 0.0012, 23 °C Ctr vs. 23 °C QIH, p = 0.0188, 34 °C QIH vs. 23 °C QIH, heart: p < 0.0001, 23 °C Ctr vs. 23 °C QIH, p = 0.0179, 34 °C Ctr vs. 23 °C Ctr, p = 0.0409, 34 °C QIH vs. 23 °C QIH, p = 0.0173, 34 °C Ctr vs. 34 °C QIH, brain: p = 0.0155, 23 °C Ctr vs. 23 °C QIH, p = 0.0464, 34 °C Ctr vs. 23 °C Ctr, p = 0.0001, 34 °C QIH vs. 23 °C QIH). All data are presented as means ± SEM; ns = not significant, *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001.

Fig. 5. Acute cooling down of body temperature recovers QIH-mediated glucose hypometabolism.

a, h Diagram of experimental design for 6-h (a) and 18-h (g) fasting. b, c, Traces of BAT temperature (b) and core temperature (c) of 6-h–fasted control (n = 8) and QIH (n = 8) mice after being moved to 23 °C. d Blood glucose levels of 6-h–fasted control (n = 10) and QIH (n = 9; one-way ANOVA followed by Sidak multiple comparison test, p < 0.0001, 34 °C Ctr vs. cooled Ctr, p < 0.0001, cooled Ctr vs. cooled QIH) mice with or without 2-h cooling. e ITT in 34 °C-housed control (n = 10) and QIH (n = 9) mice with or without cooling. f AUC of (e) (One-way ANOVA followed by Sidak multiple comparison test, p = 0.0378, cooled Ctr vs. cooled QIH, p = 0.003, cooled QIH vs. 34 °C QIH). g Minimal glucose level during ITT (One-way ANOVA followed by Sidak multiple comparison test, p = 0.0316, cooled Ctr vs. cooled QIH, p = 0.0158, cooled QIH vs. 34 °C QIH). i, j Traces of BAT temperature (i) and core temperature (j) of 18-h–fasted control (n = 8) and QIH (n = 8) mice after being moved to 23 °C. k Blood glucose levels of 18-h–fasted control (n = 10) and QIH (n = 9; one-way ANOVA followed by Sidak multiple comparison test, p = 0.0047, cooled Ctr vs. Cooled QIH, p = 0.0004, 34 °C QIH vs. cooled QIH) with or without 2-h cooling. l Plasma insulin levels of 18-h–fasted control (n = 8) and QIH (n = 8; one-way ANOVA followed by Sidak multiple comparison test, p < 0.0001, cooled Ctr vs. cooled QIH, p < 0.0001, 34 °C QIH vs. Cooled QIH) with or without 2-h cooling. m GTT in 34 °C-housed control (n = 10) and QIH mice (n = 9) with or without cooling. n AUC of (m) (One-way ANOVA followed by Sidak multiple comparison test, p < 0.0001, cooled Ctr vs. cooled QIH, p < 0.0001, cooled QIH vs. 34 °C QIH). o–r, respiratory-gas analysis of control and QIH mice of 18-h–fasted control (n = 6) and QIH (n = 5) with or without 2-h cooling. o, q traces of VO₂ (o) and VCO₂ (q) of 18-h–fasted control (n = 6) and QIH (n = 5) mice with or without 2-h cooling. p, r Comparisons of VO₂ (p) and VCO₂ (r) of control (n = 6) and QIH (n = 5; one-way ANOVA followed by Sidak multiple comparison test, for VO₂: p = 0.0199, cooled Ctr vs. cooled QIH, p = 0.0037, 34 °C QIH vs. Cooled QIH, for VCO₂: p = 0.0103, cooled Ctr vs. cooled QIH, p = 0.0008, 34 °C QIH vs. Cooled QIH, p = 0.0156, 34 °C Ctr vs. cooled Ctr) mice before and after 2-h cooling. All data are presented as means ± SEM; ns = not significant, *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001.

Fig. 6. Body temperature regulates food intake and locomotion.

a Twenty-four-hour food intake of control and QIH mice housed in 23 °C (control, n = 7; QIH, n = 7) and 34 °C (control, n = 8; QIH, n = 8)(One-way ANOVA followed by Sidak multiple comparison test, p < 0.0001, 23 °C Ctr vs. 23 °C QIH, p < 0.0001, 34 °C Ctr vs. 23 °C Ctr, p = 0.0022, 34 °C QIH vs. 23 °C QIH, p = 0.0077, 34 °C Ctr vs. 34 °C QIH). b Accumulative refed food intake of control and QIH mice housed in 23 °C (control, n = 7; QIH, n = 7) and 34 °C (control, n = 6; QIH, n = 6) after 24-h fasting. c Food intake of control and QIH mice housed in 23 °C (control, n = 7; QIH, n = 7) and 34 °C (control, n = 6; QIH, n = 6; one-way ANOVA followed by Sidak multiple comparison test, p < 0.0001, 23 °C Ctr vs. 23 °C QIH, p = 0.0075, 34 °C Ctr vs. 23 °C Ctr, p < 0.0001, 34 °C QIH vs. 23 °C QIH, p < 0.0001, 34 °C Ctr vs. 34 °C QIH) during 4-h refed. d Representative micrographs showing immunofluorescent cFos staining in the arcuate nucleus (ARC) of control and QIH mice housed in 23 °C and 34 °C. Scale bar: 100 μm. e Quantification of cFos expression in the ARC of control and QIH mice after 24-h fasting in 23 °C (control, n = 3; QIH, n = 3) and 34 °C (control, n = 4; QIH, n = 5; one-way ANOVA followed by Sidak multiple comparison test, p = 0.0003, 23 °C Ctr vs. 23 °C QIH, p = 0.0044, 34 °C Ctr vs. 23 °C Ctr, p = 0.0149, 34 °C QIH vs. 23 °C QIH). f Experimental design to measure the locomotion activity of control and QIH mice. Blue bar indicates the recording period. g–j Traces of locomotion activity of control and QIH mice housed in 23 °C (control, n = 5; QIH, n = 5) (g, h) and 34 °C (control, n = 5; QIH, n = 5) (i, j). k, l Locomotion activity during light (k) and dark (l) periods before clozapine injection and food deprivation (One-way ANOVA followed by Sidak multiple comparison test, for dark periods: p = 0.0337, 34 °C Ctr vs. 23 °C Ctr, p = 0.0018, 34 °C QIH vs. 23 °C QIH). m, n Locomotion activity during light (m) and dark (n) periods after clozapine injection and food deprivation (One-way ANOVA followed by Sidak multiple comparison test, for light periods: p < 0.0001, 23 °C Ctr vs. 23 °C QIH, p = 0.0217, 34 °C Ctr vs. 23 °C Ctr, p < 0.0001, 34 °C QIH vs. 23 °C QIH, p = 0.0083, 34 °C Ctr vs. 34 °C QIH, for dark periods: p < 0.0001, 23 °C Ctr vs. 23 °C QIH, p = 0.047, 34 °C QIH vs. 23 °C QIH, p < 0.0001, 34 °C Ctr vs. 23 °C Ctr). Gray blocks in (g) to j indicate the dark period. All data are presented as means ± SEM; ns = not significant, *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001.

Fig. 7. Hypothermia regulates glucose metabolism.

Qrfp^POA activation induces hypothermia by decreasing BAT thermogenesis and increasing vasodilation without directly affecting glucose metabolism. QIH-mediated hypothermia results in glucose hypometabolism with increased plasma insulin and decreased insulin sensitivity. Hypothermic mice show inappetance and low locomotor activity. The glucose hypometabolism and torpid behaviors during QIH are recovered by increasing body temperature to euthermia.