Sex dependence of opioid-mediated responses to subanesthetic ketamine in rats

Abstract

Subanesthetic ketamine is increasingly used for the treatment of varied psychiatric conditions, both on- and off-label. While it is commonly classified as an N-methyl D-aspartate receptor (NMDAR) antagonist, our picture of ketamine's mechanistic underpinnings is incomplete. Recent clinical evidence has indicated, controversially, that a component of the efficacy of subanesthetic ketamine may be opioid dependent. Using pharmacological functional ultrasound imaging in rats, we found that blocking opioid receptors suppressed neurophysiologic changes evoked by ketamine, but not by a more selective NMDAR antagonist, in limbic regions implicated in the pathophysiology of depression and in reward processing. Importantly, this opioid-dependent response was strongly sex-dependent, as it was not evident in female subjects and was fully reversed by surgical removal of the male gonads. We observed similar sex-dependent effects of opioid blockade affecting ketamine-evoked postsynaptic density and behavioral sensitization, as well as in opioid blockade-induced changes in opioid receptor density. Together, these results underscore the potential for ketamine to induce its affective responses via opioid signaling, and indicate that this opioid dependence may be strongly influenced by subject sex. These factors should be more directly assessed in future clinical trials.

Fig. 1. Functional ultrasound imaging of intravenous ketamine administration.

a Schematic representation of the imaging setup. A surgical craniotomy enables ultrasound penetration, and an implanted chronic prosthesis allows imaging over repeated sessions. Drawing created in BioRender. b Coronal slices of the rat brain were imaged at bregma +2.5 mm and bregma −3.5 mm. The segmented regions of interest (ROIs) are highlighted on the Paxinos & Watson rat brain atlas. c Sequence of cerebral blood volume (CBV) coronal maps at bregma +2.5 mm and bregma −3.5 mm following administration of 10 mg/kg i.v. ketamine. The pixel intensity shows the CBV signals as a normalized difference with a pre-injection baseline (10 min). The time axis was zeroed at the time of ketamine injection. d The coronal maps were segmented and the CBV signals were averaged in the relevant ROIs. The plots show CBV time series in response to increasing doses of i.v. ketamine. Solid lines represent the mean values and shaded areas are SEM from n = 9 rats/group (10 and 5 mg/kg) or n = 8 rats/group (1 and 0 mg/kg). e Peak CBV in the segmented ROIs. Two-way mixed-effects ANOVA; within-subjects factor of region, F_3.08,92.55 = 17.82, P = 2.33E-09; between-subjects factor of dose, F_3,30 = 8.26, P = 3.74E-04; interaction, F_9.25,92.55 = 3.38, P = 0.001. Two-tailed unpaired t-test, *corrected P < 0.05; **P < 0.01; ***P < 0.001. n = 9 rats/group (10 and 5 mg/kg); n = 8 rats/group (1 and 0 mg/kg). Data are presented as mean +/− SEM. Source data are provided as a Source Data file. Details on the statistical analyses are provided in Supplementary Table 1. KET: ketamine. Cg1 cingulate area 1, PrL prelimbic cortex, IL infralimbic cortex, M2 secondary motor cortex, M1 primary motor cortex, S1 primary somatosensory cortex, Ins insular cortex, CPu caudate putamen, NAcC nucleus accumbens core, NAcSh nucleus accumbens shell, LPtA lateral parietal association cortex, MPtA medial parietal association cortex, RSG granular retrosplenial cortex, RSD dysgranular retrosplenial cortex, Hipp hippocampus, LHb lateral habenula, LPLR lateral posterior thalamic nucleus, VM ventromedial thalamic nucleus, VPM ventral posteromedial thalamic nucleus, VPL ventral posterolateral thalamic nucleus.

Fig. 2. Pharmaco-fUSI closely tracks ketamine-evoked gamma-band power in the prefrontal cortex.

a Schematic representation of the setup for recording intracranial electrocorticography over the Cg1 sub-region of the prefrontal cortex. Coronal slice drawing adapted from the Paxinos & Watson rat brain atlas. b Representative spectrogram for i.v. administration of 10 mg/kg ketamine (KET). c Time series of normalized electrocorticography (ECoG) power changes in each frequency band and cerebral blood volume (CBV) signal in the Cg1 region. Solid lines represent the mean values and shaded areas are SEM. d For each rat, the time series of normalized ECoG power changes and the Cg1 CBV signal were regressed using a four-parameter gamma-distribution function. The fitted β₁, β₂, and β₃ values were compared between each band and Cg1 CBV time series. A time-delay parameter β₄ was added to the model to improve the goodness of fit but was not included in the statistical analysis. No significant differences were observed in the least-squares minimization residuals (Supplementary Fig. 3f; P > 0.11). Two-tailed unpaired t-test comparing each band to CBV, *corrected P < 0.05; **P < 0.01; ***P < 0.001. n = 6 rats for ECoG; n = 9 rats for CBV. Data are presented as mean +/− SEM. Source data are provided as a Source Data file. Details on the statistical analyses are provided in Supplementary Table 1.

Fig. 3. Pharmaco-fUSI reveals a sex-dependence of opioid-mediated effects of ketamine administration.

a Rats received an s.c. injection of either vehicle (VEH; saline) or naltrexone (NTX; 10 mg/kg) followed by ketamine (KET; 10 mg/kg, i.v.) or vehicle after 10 min. Each animal was imaged three times under the treatment conditions of VEH + KET, NTX + KET, and NTX + VEH. b, c Functional maps in male (b) and female (c) rats. The t scores were calculated by contrasting the pixel-wise peak cerebral blood volume (CBV) in the NTX + KET versus VEH + KET groups. Statistically significant clusters are displayed overlaid on a power Doppler template (one cohort of n = 9 females and n = 9 males imaged at bregma +2.5 mm; one cohort of n = 9 females and n = 9 males imaged at bregma −3.5 mm; two-tailed paired t-test, corrected P < 0.05). In male rats, functional maps show that naltrexone reduced peak activity in M1/2, Cg1, NAc, and CPu, and increased activity in RSG, LHb, and LPRL. There were only minor clusters in female rats. d, e CBV time series in Cg1, Cpu, NAcC, RSG, and LHb in male (d) and female (e) rats. Solid lines represent the mean values and shaded areas are SEM. f A different cohort of male rats received an i.v. dose of 0.1 mg/kg MK-801 with naltrexone or vehicle pretreatment. The bar plots display the peak CBV in the Cg1 and NAcC regions. Two-tailed paired t-tests (NTX + KET vs VEH + KET and NTX + MK-801 vs VEH + MK-801), corrected P: Cg1 = 0.016 (KET), 0.572 (MK-801); NAcC = 0.024 (KET), 0.922 (MK-801). Hedge’s g effect sizes: Cg1 = −1.2 (KET), −0.3 (MK-801); NAcC = −0.97 (KET), 0.04 (MK-801). n = 9 male rats (KET groups); n = 6 male rats (MK-801 groups). Data are presented as mean +/− SEM. g CBV time series in Cg1 and NAcC in male rats receiving MK-801. n = 6 male rats. Solid lines represent the mean values and shaded areas are SEM. Source data are provided as a Source Data file. Details on the statistical analyses are provided in Supplementary Table 1.

Fig. 4. Sex dependence of opioid-mediated effects of subanesthetic ketamine are region-specific.

a Peak cerebral blood volume (CBV) changes in individual male and female rats receiving vehicle or naltrexone pretreatment. Two-tailed paired t-test (NTX + KET vs VEH + KET), corrected P: Cg1 = 0.024 (M), 0.23 (F); M1 = 0.024 (M), 0.954 (F); NAcC = 0.048 (M), 0.954 (F); CPu = 0.096 (M), 0.386 (F); LHb = 0.084 (M), 0.77 (F). Group-level differences were significant before multiple comparisons correction in CPu (uncorrected P = 0.032) and LHb (P = 0.021) in male rats. Hedge’s g effect sizes (NTX + KET vs VEH + KET): Cg1 = -1.20 (M), 0.53 (F); M1 = -1.28 (M), 0.03 (F); NAcC = −0.97 (M), 0.08 (F); CPu = −0.78 (M), 0.40 (F); LHb = 0.87 (M), 0.17 (F). n = 9 male rats; n = 9 female rats. b Peak CBV differences between the NTX + KET and VEH + KET treatments in individual rats were compared between males and females. One-way ANOVA for sex factor, F_1,178 = 7.52, P = 0.007. Two-tailed unpaired t-test, corrected P: Cg1 = 0.0117; CPu = 0.0473; LPLR = 0.0473. Group-level differences did not pass multiple comparisons correction in NAcC (uncorrected P = 0.0245). Hedge’s g effect sizes: Cg1 = 1.77; CPu = 1.24; LPLR = −1.32. c A different cohort of male rats received a surgical orchiectomy. CBV time series show that the removal of gonadal hormones completely blocked the opioid-mediated effect of subanesthetic ketamine. n = 7 rats. Solid lines represent the mean values and shaded areas are SEM. d Peak CBV in the segmented ROIs. Two-way ANOVA; within-subjects factor of region, F_9,54 = 3.81, P = 8.9E-04; within-subjects factor of treatment, F_1,6 = 0.31, P = 0.6; interaction: F_9,54 = 1.33, P = 0.25. n = 7 rats. Data are presented as mean +/− SEM. Source data are provided as a Source Data file. Details on the statistical analyses are provided in Supplementary Table 1.

Fig. 5. Ketamine infusions increase postsynaptic density in a sex-dependent and opioid-mediated fashion.

a Rats received an s.c. injection of naltrexone (NTX; 10 mg/kg) or vehicle (VEH) followed by ketamine (KET; 10 mg/kg, i.p.) or vehicle after 10 min and were perfused 24 h post ketamine infusion. Brain slices were analyzed by immunohistochemistry (IHC). b Representative PSD-95 stains in the rat prefrontal cortex at 20× magnification. Scale bar: 100 μm. Coronal slice drawing adapted from the Paxinos & Watson rat brain atlas. c Quantification of prefrontal cortex PSD-95 and DAPI intensity in male rats. Two-tailed unpaired t-test (PSD-95), corrected P: VEH + VEH vs VEH + KET = 1.42E-05; VEH + KET vs NTX + VEH = 1.86E-05; VEH + KET vs NTX + KET = 1.86E-05. Hedge’s g effect sizes: VEH + VEH vs VEH + KET = 4.36; VEH + KET vs NTX + VEH = −3.75; VEH + KET vs NTX + KET = −3.48. n = 4 rats/group. d Quantification of prefrontal cortex PSD-95 and DAPI intensity in female rats. Two-tailed unpaired t-test (PSD-95), **corrected P: VEH + VEH vs VEH + KET = 0.00709. Hedge’s g effect sizes: VEH + VEH vs VEH + KET = 2.88. n = 4 rats/group. Data are presented as mean +/− SEM. Source data are provided as a Source Data file. Details on the statistical analyses are provided in Supplementary Table 1.

Fig. 6. Sex dependence of opioid-mediated behavioral effects of subanesthetic ketamine.

a Locomotor activity was measured for 20 min in an open-field arena. Male and female rats received an s.c. injection of either vehicle (VEH; saline) or naltrexone (NTX; 10 mg/kg) followed by ketamine (KET; 10 mg/kg, i.p.) or vehicle 10 min later on each of four days. All animals were habituated for 2 days (Hab1-2) with vehicle only, followed by 4 daily sessions with ketamine with or without naltrexone pretreatment (D1-4). b Total distance traveled by male rats normalized to the mean of Hab1-2. Two-way mixed-effects ANOVA; between-subjects factor of treatment, F_2,17 = 5.24, P = 0.017; within-subjects factor of session, F_2.74,46.6 = 5.25, P = 0.004; interaction, F_5.48,46.6 = 3.40, P = 0.009. VEH + KET vs VEH + VEH: two-tailed unpaired t-test, ^#corrected P < 0.05; Hedge’s g effect sizes: D3 = 2.79; D4 = 1.70. NTX + KET vs VEH + KET: two-tailed unpaired t-test, *corrected P < 0.05; Hedge’s g effect sizes: D3 = −1.25; D4 = −1.49. n = 8 male rats/group for VEH + KET and NTX + KET. n = 4 male rats for VEH + VEH. c Total distance traveled by female rats. Two-way mixed-effects ANOVA; between-subjects factor of treatment, F_2,17 = 9.89, P = 0.001; within-subjects factor of session, F_5,85 = 15.1, P = 1.35E-10; interaction, F_10,85 = 4.16, P = 0.001. VEH + KET vs VEH + VEH: two-tailed unpaired t-test, ^#corrected P < 0.05; ^###P < 0.001; Hedge’s g effect sizes: D2 = 2.69; D3 = 4.65; D4 = 1.83. NTX + KET vs VEH + KET: two-tailed unpaired t-test, **corrected P < 0.01; Hedge’s g effect sizes: D2 = −0.03; D3 = −1.97; D4 = −0.35. n = 8 female rats/group for VEH + KET and NTX + KET. n = 4 female rats for VEH + VEH. d Distance traveled by male and female rats as a function of the time on Day 1 and Day 4. VEH + KET vs VEH + VEH: two-tailed unpaired t-test, ^#corrected P < 0.05; ^##P < 0.01. VEH + KET vs NTX + KET: two^-tailed unpaired t-test, *corrected P < 0.05. n = 8 male and female rats/group for VEH + KET and NTX + KET. n = 4 male and female rats for VEH + VEH. e Representative traces of body position during the open-field session at D4 in male and female rats. Data are presented as mean +/− SEM. Source data are provided as a Source Data file. Details on the statistical analyses are provided in Supplementary Table 1.

Fig. 7. Chronic naltrexone administration increases the density of mu opioid receptors in female rats.

a For four consecutive days, male and female rats were pretreated with naltrexone (10 mg/kg, s.c.) or vehicle 10 min prior to ketamine (10 mg/kg, i.p.) or vehicle and were euthanized 24 h after the last treatment. Autoradiography was performed to quantify [³H]DAMGO binding in the mPFC, CPu, and NAc. The images show representative autoradiography slices. b Quantified [³H]DAMGO binding potential in Cg1/PrL, CPu, and NAc. In NAc: two-way ANOVA; sex, F_1,40 = 5.12, P = 0.029; treatment, F_3,40 = 2.59, P = 0.066; interaction F_3,40 = 2.03, P = 0.125. Two-tailed unpaired t-test, *corrected P < 0.05. Hedge’s g effect sizes: CPu, NTX + KET, male vs female = 1.38; NAc, NTX + VEH, male vs female = 1.26; NAc, female, VEH + KET vs NTX + VEH = 1.79; NAc, female, VEH + KET vs NTX + KET = 1.33. n = 6 male and female rats/group. Data are presented as mean +/− SEM. Source data are provided as a Source Data file. Details on the statistical analyses are provided in Supplementary Table 1.