Corticotropin-releasing factor neurons in the bed nucleus of the stria terminalis exhibit sex-specific pain encoding in mice

Abstract

The bed nucleus of the stria terminalis (BNST) plays an emerging role in pain regulation. Pharmacological studies have found that inhibiting corticotropin-releasing factor (CRF) signaling in the BNST can selectively mitigate the sensory and affective-motivational components of pain. However, mechanistic insight on the source of CRF that drives BNST responses to these harmful experiences remains unknown. In the present study, we used a series of genetic approaches to show that CRF in the BNST is engaged in the processing and modulation of pain. We conducted cell-type specific in vivo calcium imaging in CRF-Cre mice and found robust and synchronized recruitment of BNST^CRF neurons during acute exposures to noxious heat. Distinct patterns of recruitment were observed by sex, as the magnitude and timing of heat responsive activity in BNST^CRF neurons differed for male and female mice. We then used a viral approach in Floxed-CRF mice to selectively reduce CRF expression in the BNST and found it decreased nociceptive sensitivity for both sexes and increased paw attending for females. Together, these findings reveal that CRF in the BNST influences multiple facets of the pain experience to impact the sex-specific expression of pain-related behaviors.

Figure 1

Measurement of Single-Cell Ca²⁺ Activity in BNST^CRF Neurons During Pain (A) Miniature microscope imaging in BNST^CRF neurons. (Left) Diagram of GCaMP6s infusion and endoscopic lens / baseplate implantation in the BNST of CRF-Cre mice. (Upper-Right) Imaging field-of-view (FOV) depicting GCaMP6s expression in BNST^CRF neurons. (Lower-Right) Experimental timeline with schematic of BNST^CRF imaging during behavioral assessment of thermal nociceptive sensitivity. Epochs relative to noxious heat exposure are color-coded: Before (blue), Heat Start (red), Heat End (orange), and After (green). To ensure that epochs represent equal amounts of time on a trial-by-trial basis, epoch lengths were quantitatively defined as half the duration of heat exposure for each specific subject and trial. (B–C) Peri-event heatmap of average BNST^CRF activity surrounding (B) heat onset (red line) and (C) paw withdrawal (orange line) for male (left column; n = 49 cells / 7–30 cells per mouse) and female (right column; n = 68 cells / 8–19 cells per mouse) subjects. Lighter shades of color correspond to higher z-scores, a metric indicative of greater changes in Ca²⁺ activity relative to baseline. (D–E) Total percentage of pain responsive BNST^CRF neurons for male (purple) and female (magenta) subjects across trials. Cells that respond to heat exposure with positive (darker), negative (lighter), and no (grey) changes in z-score were determined relative to the Before and After epochs, as demarcated by the onset of (D) heat exposure [“Relative to Before”] and (E) paw withdrawal [“Relative to After”] respectively (Wilcoxon rank-sum, P < 0.05). Data are shown as mean ± SEM. *P < 0.05; **P < 0.01; ***P < 0.001, ****P < 0.0001.

Figure 2

Sex-Dependent BNST^CRF Responses to Noxious Heat Exposure by Epoch. (A–B) Relative fluorescence change (ΔF/F) of BNST^CRF activity during noxious heat exposure. Z-scores of (A) male (n = 49 cells / 7–30 cells per mouse) and (B) female (n = 68 cells / 8–19 cells per mouse) subjects are represented over time by aligning traces that were averaged across trials to the onset of the heat stimulus. Individual cell activity is indicated in lighter colors, with total average cell activity in darker colors, and maximum heat duration (20 s) in the transparent blue block. Scale bars, x = time (10 s), y = z-score based on relative ΔF/F (1 z-score). Imaging timelines for individual subjects are located above to represent epochs via the color-coding system described in Fig. 1A: Before (blue), Heat Start (red), Heat End (orange), and After (green). Similar color designations apply for all panels of Fig. 2. (C–D) Heatmap exhibiting the average z-score of individual BNST^CRF neurons by epoch for (C) male and (D) female subjects. Darker shades of color correspond to greater changes in Ca²⁺ activity relative to baseline. (E–F) Cumulative distribution function plotting the z-score frequency by epoch in (E) male and (F) female subjects (KS test: By Sex-Before [D(117) = 0.3466, P value = 0.0014], Heat Start [D(117) = 0.3850, P value = 0.0002], Heat End [D(117) = 0.1815, P value = 0.2759], After [D(117) = 0.2611, P value = 0.0334]; Comparing each epoch within males results in significance with Heat Start and Heat End over Before and After [Before vs. Heat Start, D(98) = 0.4694, P value = 2.183e−05; Before vs. Heat End, D(98) = 0.4286, P value = 0.0001; Heat Start vs. After, D(98) = 0.4082, P value = 0.0003; Heat End vs. After, D(98) = 0.3265, P value = 0.0079], whereas comparing each epoch within females results in significance with Heat End over Before, Heat Start, and After [Before vs. Heat Start, D(136) = 0.2647, P value = 0.0135; Before vs. Heat End, D(136) = 0.4558, P value = 7.372e−07; Heat Start vs. Heat End, D(136) = 0.2352, P value = 0.0386; Heat Start vs. After, D(136) = 0.2941, P value = 0.0041; Heat End vs. After, D(136) = 0.3970, P value = 2.634e−05]). (G) Comparing average z-scores of BNST^CRF neurons by epoch for male and female subjects (Two-way mixed-model ANOVA with Tukey’s post hoc: Sex × Epoch interaction [F(3, 345) = 5.996, P = 0.0005], main effect of Sex [F(1, 115) = 13.20, P = 0.0004] and Epoch [F(2.153, 247.7) = 26.31, P < 0.0001]). denotes comparison of epochs for each sex, ^ denotes comparison of the same epoch between sexes. (H) Neuronal coactivity as measured by mean proportion of active BNST^CRF neurons. At each frame (0.2 s), the fraction of cells exhibiting positive z-scores was calculated and binned by epoch to reveal the proportion of concurrent BNST^CRF activity at phases relative to heat exposure (Two-way mixed-model ANOVA with Sidak’s post hoc: no Sex × Epoch interaction [F(3, 18) = 0.5414, P = 0.6601] or main effect of Sex [F(1, 6) = 0.9203, P = 0.3744]; main effect of Epoch [F(1.840, 11.04) = 4.163, P = 0.0475]). Data are shown as mean ± SEM. *P < 0.05; **P < 0.01; ***P < 0.001, ****P < 0.0001.

Figure 3

Progression of BNST^CRF Activity Across Noxious Heat Exposure Trials. (A) Average BNST^CRF traces of male (purple; n = 49 cells / 7–30 cells per mouse / 3 mice total) and female (magenta; n = 68 cells / 8–19 cells per mouse / 5 mice total) subjects by trial. Within each representative time window, the maximum heat duration (20 s) is indicated in the transparent blue block. (B-C) Percentage of pain responsive BNST^CRF neurons for male (purple) and female (magenta) subjects were determined relative to epochs surrounding the onset of (B) heat exposure [“Relative to Before”] and (C) paw withdrawal [“Relative to After”] in trials 1–4. Positive, negative, and non-responsive cells are indicated by darker purple/magenta, lighter purple/magenta, and grey respectively (Wilcoxon rank-sum, P < 0.05). Comparisons by trial and sex relative to the epochs surrounding heat onset (Two-way mixed-model ANOVA with Sidak’s post hoc: no Trial × Sex interaction [F(3, 18) = 0.0466, P = 0.9862] or main effect of Trial [F(1.092, 6.553) = 0.4473, P = 0.5435] and Sex [F(1, 6) = 1.510, P = 0.2651]) and paw withdrawal (Two-way mixed-model ANOVA with Sidak’s post hoc: no Trial × Sex interaction [F(3, 18) = 0.3670, P = 0.7777] or main effect of Trial [F(2.740, 16.44) = 0.3368, P = 0.7818] and Sex [F(1, 6) = 0.0008, P = 0.9776]) were not statistically significant. (D) Heatmap displaying the average z-score of individual BNST^CRF neurons by epoch for male and female subjects in trials 1–4. Epochs are color-coded: Before (blue), Heat Start (red), Heat End (orange), and After (green), with darker shades of color corresponding to greater changes in Ca²⁺ activity. Data are shown as mean ± SEM. *P < 0.05; **P < 0.01; ***P < 0.001, ****P < 0.0001.

Figure 4

BNST^CRF Activity Changes with Repeated Noxious Heat Exposure. (A–B) Average z-score of BNST^CRF neurons by epoch for trials 1–4 in (A) male and (B) female subjects. Epochs are color-coded: Before (blue), Heat Start (red), Heat End (orange), and After (green) (Males = Two-way mixed-model ANOVA with Sidak’s post hoc: Epoch × Trial interaction [F(4.352, 208.9) = 8.148, P < 0.0001], main effect of Epoch [F(2.375, 114.0) = 16.66, P < 0.0001] and Trial [F(2.822, 135.4) = 10.63, P < 0.0001]; Females = Two-way mixed-model ANOVA with Sidak’s post hoc: Epoch × Trial interaction [F(3.150, 211.0) = 21.05, P < 0.0001], main effect of Epoch [F(1.679, 112.5) = 8.880, P = 0.0006] and Trial [F(2.491, 166.9) = 6.030, P = 0.0014]). (C–D) Average neuronal coactivity of BNST^CRF neurons by epoch for trials 1–4 in (C) male and (D) female subjects. Epochs are color-coded with the same system as (A–B): Before (blue), Heat Start (red), Heat End (orange), and After (green) (Males = Two-way mixed-model ANOVA with Sidak’s post hoc: no Epoch × Trial interaction [F(1.526, 3.052) = 1.046, P = 0.4252] or main effect of Epoch [F(1.419, 2.838) = 3.149, P = 0.1859] and Trial [F(1.178, 2.355) = 0.5741, P = 0.5447]; Females = Two-way mixed-model ANOVA with Sidak’s post hoc: no Epoch × Trial interaction [F(2.082, 8.329) = 0.9274, P = 0.4366] or main effect of Epoch [F(1.745, 6.981) = 2.102, P = 0.1935] and Trial [F(1.581, 6.326) = 1.041, P = 0.3871]). (E) (Left) Maximum z-score response to heat exposure by trial in male (purple) and female (magenta) subjects (Two-way mixed-model ANOVA with Sidak’s post hoc: Trial × Sex interaction [F(3, 345) = 14.96, P < 0.0001], main effect of Trial [F(2.609, 300) = 20.28, P < 0.0001] and Sex [F(1, 115) = 13.69, P = 0.0003]). (Right) Average maximum z-score response to heat in each trial by sex (unpaired t-test: t(115) = 3.70, P = 0.0003). (F) (Left) Latency of maximum z-score response to heat exposure by trial in male (purple) and female (magenta) subjects (Two-way mixed-model ANOVA with Sidak’s post hoc: Trial × Sex interaction [F(3, 345) = 16.90, P < 0.0001], main effect of Trial [F(2.717, 312.5) = 7.890, P < 0.0001], no main effect of Sex [F(1, 115) = 3.20, P = 0.0763]). (Right) Average latency of maximum z-score response to heat stimulus in each trial by sex (unpaired t-test: t(115) = 1.789, P = 0.0763). Data are shown as mean ± SEM. *P < 0.05; **P < 0.01; ***P < 0.001, ****P < 0.0001.

Figure 5

CRF Deletion in BNST Reduces Thermal Nociceptive Sensitivity. (A) Diagram of Cre-dependent approach for CRF deletion in BNST. (B) Representative histology of CRF mRNA expression in BNST of Floxed-CRF mice following virus infusion. (C) Quantification of CRF mRNA expression in BNST of male (n = 25–26) and female (n = 25–26) subjects by cells/mm² (Two-way ANOVA with Tukey’s post hoc: no Sex × Virus interaction [F(1, 98) = 2.548, P = 0.1136] or main effect of Sex [F(1, 98) = 0.2667, P = 0.6067]; main effect of Virus [F(1, 98) = 66.74, P < 0.0001]). (D) (Left) Schematic of the Hargreaves test. Each paw was assessed using three replicate exposures, alternating between the right (R) and left (L) limb, resulting in six trials total. (Right) Thermal nociceptive sensitivity of male (n = 25–26) and female (n = 25–26) subjects (Two-way ANOVA with Tukey’s post hoc: no Sex × Virus interaction [F(1, 98) = 0.1253, P = 0.7242] or main effect of Sex [F(1, 98) = 2.471, P = 0.1192]; main effect of Virus [F(1, 98) = 5.388, P = 0.0223]). Data are shown as mean ± SEM. *P < 0.05; **P < 0.01; ***P < 0.001, ****P < 0.0001.

Figure 6

CRF Deletion in BNST Selectively Alters Sensory-Discriminative and Affective-Motivational Behaviors. (A) Schematic of hot plate test. (B) Thermal nociceptive sensitivity of male (n = 16–17) and female (n = 15–17) Floxed-CRF mice, as defined by latency to first paw withdrawal for each mouse (Two-way ANOVA with Tukey’s post hoc: no Sex × Virus interaction [F(1, 61) = 1.853, P = 0.1785] or main effect of Sex [F(1, 61) = 0.1124, P = 0.7386] and Virus [F(1, 61) = 0.1524, P = 0.6976]). (C) Quantification of sensory-discriminative and affective-motivational behaviors throughout the 45-s exposure to hot plate. Averaged area under the curve (AUC) of paw withdrawal (Two-way ANOVA with Tukey’s post hoc: no Sex × Virus interaction [F(1, 61) = 1.072, P = 0.3047] or main effect of Sex [F(1, 61) = 1.885, P = 0.1747] and Virus [F(1, 61) = 0.6208, P = 0.4338]), paw attending (Two-way ANOVA with Tukey’s post hoc: no Sex × Virus interaction [F(1, 61) = 2.605, P = 0.1117] or main effect of Sex [F(1, 61) = 0.8298, P = 0.3659]; main effect of Virus [F(1, 61) = 4.695, P = 0.0342]), and paw guarding (Two-way ANOVA with Tukey’s post hoc: no Sex × Virus interaction [F(1, 61) = 0.09333, P = 0.7610] or main effect of Sex [F(1, 61) = 1.026, P = 0.3152] and Virus [F(1, 61) = 0.04448, P = 0.8337]) are shown, with corresponding cumulative distribution functions displayed in (D–E). (D–E) Sensory-discriminative and affective-motivational behaviors are exhibited across time with cumulative distribution functions for (D) male and (E) female subjects (paw withdrawal: CON (M): 391.9 [341.0–442.9], CRE (M): 311.7 [263.2–360.2], CON (F): 285.9 [249.1–322.7], CRE (F): 296.8 [261.1–332.5]; paw attending: CON (M): 11.00 [4.672–17.33], CRE (M): 14.03 [6.087–21.97], CON (F): 7.147 [2.442–11.85], CRE (F): 27.87 [17.95–37.78]; paw guarding: CON (M): 189.3 [171.7–207.0], CRE (M): 179.3 [156.7–201.9], CON (F): 163.8 [147.6–180.0], CRE (F): 165.6 [147.8–183.4]). Data are shown as mean ± SEM. *P < 0.05; **P < 0.01; ***P < 0.001, ****P < 0.0001.