HIV-1 Tat and morphine decrease murine inter-male social interactions and associated oxytocin levels in the prefrontal cortex, amygdala, and hypothalamic paraventricular nucleus

Abstract

Many persons infected with HIV-1 (PWH) and opioid-dependent individuals experience deficits in sociability that interfere with daily living. Sociability is regulated by the prefrontal cortico-hippocampal-amygdalar circuit. Within this circuit HIV-1 trans-activator of transcription (HIV-1 Tat) and opioids can increase dendritic pathology and alter neuronal firing. Changes in sociability are also associated with dysregulation of hypothalamic neuropeptides such as oxytocin or corticotropin releasing factor (CRF) in the prefrontal cortico-hippocampal-amygdalar circuit. Accordingly, we hypothesized that the interaction of HIV-1 Tat and morphine would impair inter-male social interactions and disrupt oxytocin and CRF within the PFC and associated circuitry. Male mice were exposed to HIV-1 Tat for 8 weeks and administered saline or escalating doses of morphine twice daily (s.c.) during the last 2 weeks of HIV-1 Tat exposure. Tat attenuated aggressive interactions with an unknown intruder, whereas morphine decreased both non-aggressive and aggressive social interactions in the resident-intruder test. However, there was no effect of Tat or morphine on non-reciprocal interactions in the social interaction and novelty tests. Tat, but not morphine, decreased oxytocin levels in the PFC and amygdala, whereas both Tat and morphine decreased the percentage of oxytocin-immunoreactive neurons in the hypothalamic paraventricular nucleus (PVN). In Tat(+) or morphine-exposed mice, regional levels of CRF and oxytocin correlated with alterations in behavior in the social interaction and novelty tests. Overall, decreased expression of oxytocin in the prefrontal cortico-hippocampal-amygdalar circuit is associated with morphine- and HIV-Tat-induced deficits in social behavior.

Keywords: Aggression; Corticotropin releasing factor; Endogenous opioid system; HIV-associated neurocognitive disorders; Hippocampus; Opiate abuse; Paraventricular nucleus of the hypothalamus; Resident-intruder; Social anxiety; Social interaction.

Fig. 1.. Experimental design depicted on a timeline.

All Tat transgenic mice received DOX-containing chow for 8 weeks and were repeatedly injected with morphine (10 – 40 mg/kg, increasing by 10 mg/kg/2 day, s.c., b.i.d.) or saline for the last 2 weeks. Mice in cohort 1 were single housed starting at week 5 of DOX administration to establish territorial behavior. Then on day 14 of repeated morphine injections cohort 1 mice were tested in assays of aggressive sociability, the social dominance tube and resident-intruder tests. On day 14 of repeated morphine injections mice in cohort 2 were tested in the novel environment non-reciprocal social interaction and social novelty assays. The day following behavioral testing, mice in both cohorts were euthanized and tissues were randomly assigned to immunoblotting and ELISA or immunohistochemistry assays.

Fig. 2.. HIV-1 and morphine decreased aggressive, while morphine decreased non-aggressive interactions in the resident-intruder test.

Tat(+) and Tat(−) mice were fed DOX chow for 8 weeks and administered escalated morphine (10 – 40 mg/kg, s.c., b.i.d.) for 2 weeks before being tested in aggressive measures of reciprocal social interaction. Tat and morphine did not affect dominance in the social dominance tube test (A), but exposure to both decreased total time spent interacting with the unfamiliar mouse in the resident-intruder test (B). Tat and morphine did not alter the latency to interact non-aggressively with the unfamiliar mouse in the test mouse’s home cage (C), but morphine decreased the percentage of time spent interacting non-aggressively (D). Tat and morphine also did not alter the latency of the test mouse to interact aggressively with the novel mouse (E), but both Tat and morphine decreased the percentage of time the test mouse interacted aggressively (F). Data are presented as mean ± SEM; n = 9-10 mice per group. Main effect of Tat, *p < 0.05 vs Tat(−) mice. Main effect of morphine, ^#p < 0.05 vs saline treated mice.

Fig. 3.. HIV-1 Tat and morphine did not alter sociability in the non-reciprocal social interaction or social novelty tests.

Tat exposure for 8 weeks and escalated morphine (10 – 40 mg/kg, s.c., b.i.d.) administration for 2 weeks did not significantly affect the latency to interact (A) or percentage of time spent interacting (B) with the unfamiliar mouse, or distance traveled (C) in the social interaction test performed in a novel environment. Similarly, in the social novelty test Tat and morphine did not affect the latency to interact (D) or percentage of time spent interacting with the novel unfamiliar mouse (E), or distance traveled (F). Data are presented as mean ± SEM; n = 9-10 mice per group.

Fig. 4.. HIV-1 Tat, but not morphine decreased PFC and amygdalar, but not hippocampal oxytocin levels.

After 8 weeks of Tat, 2 weeks of morphine exposure, and behavioral testing in assays of sociability Tat(+) mice, regardless of morphine exposure, exhibited lower expression of oxytocin in the PFC (A) and amygdala (B), but not the hippocampus (C) as measured by western blotting. Representative blots show decreased oxytocin in Tat(+) compared to Tat(−) mice in the PFC (A, top) and amygdala (B, top). All oxytocin western blots are represented as relative intensity to GAPDH normalized to Tat(−)/Saline. Data are presented as mean ± SEM; n = 9-10 mice per group. Main effect of Tat, *p < 0.05 vs Tat(−) mice.

Fig. 5.. HIV-1 Tat and morphine did not alter serum corticosterone or PFC, hippocampus, or amygdala corticotropin releasing factor (CRF) expression.

Mice that were behaviorally tested in assays of sociability after 8 weeks of Tat and 2 weeks of morphine exposure did not exhibit altered levels of CRF in the PFC (A), amygdala (B), or hippocampus (C) of behaviorally-tested mice as measured by western blotting. Tat and morphine also did not alter serum corticosterone levels (D). All CRF western blots are represented as relative intensity to GAPDH normalized to Tat(−)/Saline. Data are presented as mean ± SEM; n = 9-10 mice per group.

Fig. 6.. HIV-1 Tat and morphine decrease hypothalamic paraventricular nucleus (PVN) oxytocin expression.

Representative images of cells immunoreactive for both oxytocin (green) and the neuronal marker NeuN (blue) in the PVN (A-D) imaged with Keyence VHX-7000 digital microscope at 40× magnification and stitched together. Mice exposed to Tat for 8 weeks, administered 2 weeks of morphine, and tested in social interaction assays did not demonstrate changes in hypothalamic PVN volume (E). Tat and morphine decreased oxytocin- (F), but not corticotropin releasing factor (CRF)- (G) or oxytocin- and CRF-immunoreactive colocalized (H) neurons in the PVN of the hypothalamus. These data suggest that oxytocin-expressing neurons in the PVN are selectively vulnerable to morphine and Tat, but do not show if this vulnerability is due to changes in the metabolism or release of oxytocin, or the loss of the oxytocin-expressing neuron subpopulation. Data are presented as mean ± SEM; n = 4-6 mice per group. Main effect of Tat, *p < 0.05 vs Tat(−) mice. Main effect of morphine, ^#p < 0.05 vs saline treated mice. Scale bar = 200 μm.

Fig. 7.. Cellular localization of oxytocin, corticotropin releasing factor (CRF), and NeuN immunoreactivity in the hypothalamic paraventricular nucleus (PVN).

Representative images of oxytocin (green), CRF (red), and Hoechst (blue) positive cells were taken using a Zeiss LSM 700 microscope at 63× magnification (Zeiss, Oberkochen, Germany). Tat(−)/morphine (B), Tat(+)/saline (C), and Tat(+)/morphine (D) PVN tissue sections had less oxytocin immunoreactive-cells compared to Tat(−)/Saline (A) control sections. Scale bar = 20 μm.

Fig. 8.. In HIV-Tat- or morphine-exposed mice, oxytocin or corticotropin releasing factor (CRF) expression correlates with social behaviors.

Despite the lack of morphine-induced changes in the social interaction test, a higher percentage of oxytocin-immunoreactive neurons within the hypothalamic paraventricular nucleus of morphine-exposed mice correlated with decreased time spent interacting with the unknown mouse in the social interaction test (A). Although Tat also did not change the overall oxytocin or CRF levels or social behaviors, in Tat(+) mice, hippocampal oxytocin levels positively correlated with time spent interacting with the novel unknown mouse in the social novelty test (B). In the social interaction test, higher amygdalar CRF levels in Tat(+) mice correlated with an increased latency to interact (C) and decreased time spent interacting (D) with the unknown mouse. The latency to interact and time spent interacting with the unknown mouse also positively correlated with PFC (E) and hippocampal (F) CRF levels, respectively. n = 8 mice per group.