Opioids Impair Intestinal Epithelial Repair in HIV-Infected Humanized Mice

Abstract

Intestinal barrier dysfunction and subsequent microbial translocation play crucial roles in persistent immune activation leading to HIV disease progression. Opioid use is associated with worse outcome in HIV-infected patients. The exacerbated disease progression by opioids is mainly driven by excessive intestinal inflammation and increased gut permeability. The objective of this study is to investigate how opioids potentiate HIV disease progression by compromising intestinal barrier function and impairing intestinal epithelial self-repair mechanism. In the present study, abnormal intestinal morphology and reduced epithelial proliferation were observed in bone marrow-liver-thymus humanized mice and in HIV-infected patients who were exposed to opioids. In bone marrow-liver-thymus mice, HIV, and morphine independently, and additively induced gut dysbiosis, especially depletion of Lachnospiraceae, Ruminococcaceae, and Muribaculaceae. We also observed that the abundance of Lachnospiraceae, Ruminococcaceae, and Muribaculaceae negatively correlated with apoptosis of epithelial cells, and intestinal IL-6 levels. Previous studies have shown that these bacterial families play crucial roles in maintaining intestinal homeostasis because they include most short-chain fatty acid-producing members. Short-chain fatty acids have been shown to maintain stem cell populations and suppress inflammation in the gut by inhibiting histone deacetylases (HDAC). In addition, we demonstrate that morphine exposure inhibited growth of intestinal organoids derived from HIV transgenic mice by suppressing Notch signaling in an HDAC-dependent manner. These studies implicate an important role for HDAC in intestinal homeostasis and supports HDAC modulation as a therapeutic intervention in improving care of HIV patients, especially in opioid-abusing population.

Keywords: BLT mice; HIV; gut microbiota; intestinal organoid; notch; opioids.

Figure 1

Morphine treatment potentiated intestinal epithelial damage in HIV-infected BLT (bone marrow, liver, thymus) mice. (A) Small intestinal sections harvested from the healthy controls, morphine-treated mice, the HIV-infected mice, and HIV-infected morphine-treated mice were stained with hematoxylin and eosin (H&E) (left anel; scale bar: 50 μm). Immunofluorescence was performed for claudin-1 (middle panel) and Ki67 (right panel; scale bar: 25 μm; white arrow: Ki67⁺ cells). (B) Intestinal tissue damages were scored based on a histological scoring system. Each dot represents one animal (n = 6). ANOVA followed by Bonferroni correction, F_{(3, 20)} = 4.001, ANOVA P = 0.0221. P-values for group comparison are shown in the figure if smaller than 0.05. (C) Claudin-1 (green) expression levels were quantified by measuring the intensity of green fluorescence within each villus. Quantification from two different fields for each animal is depicted (n = 12). ANOVA followed by Bonferroni correction, F_{(3, 44)} = 14.79, ANOVA P < 0.0001. P-values for group comparison are shown in the figure if smaller than 0.05. (D) The percentage of Ki-67⁺ (green) cells in each crypt was quantified. Quantification from three different crypts for each animal is depicted (n = 18). ANOVA followed by Bonferroni correction, F_(3,68) = 43, ANOVA P < 0.0001. P-values for group comparison are shown in the figure if smaller than 0.05.

Figure 2

Morphine treatment induced mucosal cell apoptosis and tissue inflammation in the intestine of HIV-infected BLT mice. (A) Terminal deoxynucleotidyl transferase–mediated deoxyuridine triphosphate nick end-labeling (TUNEL) assay was used to detect apoptotic cells in the intestinal sections. Apoptotic cells were stained red. Scale bar: 50 μm; white arrow: apoptotic cells. (B) Quantification of the TUNEL assay. The number of apoptotic cells in each villus was quantified. Quantification from three different villi for each animal is depicted (n = 18). ANOVA followed by Bonferroni correction, F_(3,68) = 52.32, ANOVA P < 0.0001. P-values for group comparison are shown in the figure if smaller than 0.05. (C) ELISA was used to determine the expression levels of IL-6 in the intestine. Each dot represents one animal (n = 6). ANOVA followed by Bonferroni correction, F_{(3, 20)} = 19.13, ANOVA P < 0.0001. P-values for group comparison are shown in the figure if smaller than 0.05.

Figure 3

Morphine treatment suppressed Notch signaling pathway in the intestine of HIV-infected BLT mice. (A) Immunohistochemistry (IHC) was performed to determine activated Notch1 in the intestinal section. The positive cells were stained brown. Scale bar: 10 μm. (B) Quantification of IHC images. The number of positive cells in each crypt was quantified. Quantification from two different crypts for each animal is depicted (n = 12). ANOVA followed by Bonferroni correction, F_{(3, 44)} = 33.86, ANOVA P < 0.0001. P-values for group comparison are shown in the figure if smaller than 0.05. (C) PCR was used to determine the expression level of hairy and enhancer of split-1 (HES1) in the intestine. Each dot represents one animal (n = 6). ANOVA followed by Bonferroni correction, F_{(3, 20)} = 5.589, ANOVA P = 0.0060. P-values for group comparison are shown in the figure if smaller than 0.05.

Figure 4

Morphine treatment and HIV infection induced gut microbial dysbiosis. (A) Alpha diversity from BLT mice infected with vehicle or HIV and treated with placebo or morphine was assessed by species richness [presented as observed Operation Taxonomic Units (OTUs)]. (B) Quantification of observed OTUs. ANOVA followed by Bonferroni correction, F_{(3, 28)} = 8.086, ANOVA P = 0.0005. P-values for group comparison are shown in the figure if smaller than 0.05. (C) Principal coordinate analysis (PCoA) of samples using the Bray_Curtis metrics at the OTU level shows distinct clustering of all four groups. (D) HIV + morphine group showed significant distance from all other groups. The test of significance was PERMANOVA with 999 permutations, generating false discovery rate-adjusted P-value (Q-value) (E) Relative abundance of Muribaculaceae. ANOVA followed by Bonferroni correction, F_{(3, 28)} = 107, ANOVA P < 0.0001. P-values for group comparison are shown in the figure. (F) Relative abundance of Lachnospiraceae. ANOVA followed by Bonferroni correction, F_{(3, 28)} = 16.48, ANOVA P < 0.0001. P-values for group comparison are shown in the figure if smaller than 0.05. (G) Relative abundance of Ruminococcaceae. ANOVA followed by Bonferroni correction, F_{(3, 28)} = 18.47, ANOVA P < 0.0001. P-values for group comparison are shown in the figure if smaller than 0.05.

Figure 5

Morphine inhibited the growth of small intestinal organoids of Tg26 mice. (A) Saline- or morphine- treated WT or Tg26 intestinal organoids were isolated and cultured in the IncuCyte for live-imaging analysis for 5 days. The representative image of organoid culture is depicted. (B) The surface areas of each organoid on day 0 and 5 were quantified using ImageJ. Each dot represents one organoid (n = 15). (C) The increase of surface area was calculated by subtracting the surface area of day 0 from that of day 5 (n = 15). ANOVA followed by Bonferroni correction, F_{(3, 56)} = 8.711, ANOVA P < 0.0001. P-values for group comparison are shown in the figure if smaller than 0.05. (D) The percentage of organoids that can form villi (n = 3). ANOVA followed by Bonferroni correction, F_{(3, 8)} = 7.898, ANOVA P = 0.0089. P-values for group comparison are shown in the figure if smaller than 0.05. (E) Representative images of organoids stained with anti-Ki-67 antibody. (F) Quantification of Ki-67-stained organoids. DAPI (40,6-diamidino-2-phenylindole) was used as a nuclear stain. Green fluorescence intensity was normalized by DAPI (n = 10). ANOVA followed by Bonferroni correction, F_{(3, 36)} = 4.197, ANOVA P = 0.0120. P-values for group comparison are shown in the figure if smaller than 0.05.

Figure 6

Morphine modulated Notch signaling in the intestinal crypts of Tg26 mice. (A) Representative Western blot analysis for cleaved Notch1 and HDAC1 in organoids. (B) Densitometric analysis of cleaved Notch1 (n = 3). ANOVA followed by Bonferroni correction, F_{(3, 8)} = 4.343, ANOVA P = 0.0429. P-values for group comparison are shown in the figure if smaller than 0.05. (C) Densitometric analysis of HDAC1 (n = 3). ANOVA followed by Bonferroni correction, F_{(3, 8)} = 5.562, ANOVA P = 0.0233. P-values for group comparison are shown in the figure if smaller than 0.05. (D) Representative Western blot analysis for HES1 in organoids. (E) Densitometric analysis of HES1 (n = 3). ANOVA followed by Bonferroni correction, F_{(3, 8)} = 4.171, ANOVA P = 0.0472. P-values for group comparison are shown in the figure if smaller than 0.05. (F) Representative Western blot analysis for HDAC2 in organoids. (G) Densitometric analysis of HDAC2 (n = 3). ANOVA followed by Bonferroni correction, F_{(3, 8)} = 0.0812, P = 0.9684. (H) Representative Western blot analysis for cleavd-Notch1 following HDAC inhibitor treatment. (I) Densitometric analysis of cleaved-Notch1 (n = 3). ANOVA followed by Bonferroni correction, F_{(3, 8)} = 1.348, P = 0.3258. Values were normalized against β-actin and presented as a ratio, compared with the control group. Data represent the means of three independent blots.