Table 1.
Primers used for real time qPCR.
Figure 1.
Cohort Diagram illustrating the assignment of intestinal biopsy and plasma samples to each analytical method.
Table 2.
Demographics and clinical parameters of study cohorts.
Figure 2.
Relative abundance of epithelial cells in intestinal biopsies is not decreased in HIV+ individuals.
Fixed paraffin-embedded intestinal biopsy sections were stained with DAPI and imaged. (A) Representative images of biopsy sections from a healthy control (left panel) and an HIV+ individual (right panel) are shown with nuclei staining in blue. Abundance of epithelial (EPI) cells, represented by either the area occupied by epithelial cell nuclei (arrow) or the length (Length) of epithelial-luminal border (white line, inset), was compared to the abundance of lamina propria (LP) cells indicated by area occupied by LP cell nuclei (arrowhead). (B) Relative abundance of epithelial cells expressed as the ratio between EPI/LP nuclei area or Length/LP nuclei area, and (C) epithelial cell packing density denoted by number of epithelial nuclei per 100 µm of epithelial-luminal border, were determined in the ascending, transverse, and descending colon and compared between healthy controls and HIV+ individuals. Box-and-whisker plots were constructed using Tukey's method, where black dots identify the outliers (n = 5 for both cohorts).
Figure 3.
Subcellular localization of tight junctional proteins, occludin and ZO-1, is unaltered in the HIV colon.
Fixed, frozen biopsy sections of the ascending, transverse, and descending colon in HIV+ individuals and healthy controls were stained for occludin or ZO-1 (green), and counterstained with DAPI (blue). Representative images at the luminal surface (A) and in the crypt (B) of the ascending, transverse, and descending colon are shown.
Table 3.
Average fluorescence intensity for occludin and ZO-1 staining of colonic surface epithelium sections.
Table 4.
Average fluorescence intensity for occludin and ZO-1 staining of colonic crypt epithelium sections.
Figure 4.
Tight junctional transcripts are decreased in the colon, not the terminal ileum, of HIV+ individuals.
Calibrated normalized relative quantities (CNRQ) of ZO-1, occludin, claudin-2, and claudin-4 transcripts were determined in total RNA isolated from (A) colonic (n = 9 for HIV+, n = 13 for healthy controls) and (B) terminal ileal (n = 10 for HIV+, n = 8 for healthy controls) biopsies of HIV+ individuals and healthy controls. Levels are normalized to β-actin and eef1α1 expression. Box-and-whisker plots were constructed using Tukey's method, where black dots identify the outliers. Statistical analysis was performed on all data points, including the outliers (* p<0.05, ** p<0.01 between HIV+ and healthy controls).
Figure 5.
Tight junctional transcript levels in HIV+ individuals decrease progressively from proximal-to-distal colon.
Total RNA was isolated from biopsies obtained from the terminal ileum, as well as the ascending, transverse, and descending colon, of HIV+ individuals (red square) and healthy controls (blue circle). CNRQ of (A) ZO-1, (B) occludin, (C) claudin-2, and (D) claudin-4 transcripts were compared at each location. The Kruskal-Wallis test, a non-parametric version of the one-way ANOVA, was performed, with p-values for the HIV+ (HIV) and control (HC) cohorts shown in the top left of each panel, after which r2 and p-value were determined when appropriate using a post-test for linear trend for each transcript across location in the two cohorts separately. Linear regression is shown as a representation (For HIV+ n = 9 at all colonic locations and n = 10 at the terminal ileum; for healthy controls n = 8 for terminal ileum, n = 7 for ascending colon, and n = 9 at transverse and descending colon; # p<0.08, * p≤0.05, and ** p<0.01 between HIV+ and healthy controls).
Figure 6.
Tight junctional protein levels are decreased in the descending colon of virally-suppressed HIV+ individuals.
Total protein lysate extracted from descending colonic biopsies of HIV+ individuals and healthy controls were immunoblotted for occludin, claudin-2, claudin-4, cytokeratin-18, GAPDH, and β-actin. (A) A representative blot is shown, with H denoting an HIV+ sample, C denoting a healthy control sample, and N denoting the inter-gel normalizing control sample. (B) Specific bands within the linear density range for occludin, claudin-2, and claudin-4 were quantitated by densitometric analysis, and compared between virally-suppressed HIV+ individuals and healthy controls. Target protein levels were normalized against cytokeratin-18 protein levels. Box-and-whisker plots were constructed using Tukey's method, where black dots identify the outliers. Statistical analysis was performed on all data points, including the outliers (For HIV+ n = 10 for all proteins; For healthy controls n = 13 for occludin and claudin-4, n = 12 for claudin-2; # p<0.07, * p<0.05).
Figure 7.
Human β defensin-3 and E-cadherin expression varies differentially from proximal-to-distal HIV+ intestine.
Total RNA was isolated from intestinal biopsies, and CNRQ of transcripts for (A) human β defensin-3 (n = 11 for HIV+, n = 12 for control) and (B) E-cadherin (n = 9 for HIV+, n = 13 for control) were measured in the colon of HIV+ individuals and healthy controls (upper panel). Box-and-whisker plots were constructed using Tukey's method, where black dots identify the outliers. Statistical analysis was performed on all data points, including the outliers. Transcript levels were also compared between HIV+ individuals (red square) and healthy controls (blue circle) at each location: terminal ileum, ascending, transverse, and descending colon (lower panel). The Kruskal-Wallis test, a non-parametric version of the one-way ANOVA, was performed for each transcript in the HIV+ (HIV) and control (HC) cohorts separately, with p-values shown in the top left of each panel. Linear regression is shown as a representation (hBD3: For HIV+ n = 8 for transverse colon, and n = 9 for ascending and descending colon; for healthy controls n = 6 for ascending colon, and n = 7 for transverse and descending colon; E-cadherin: For HIV+ n = 9 for the terminal ileum and ascending colon, n = 7 for transverse colon, and n = 8 for descending colon; for healthy controls n = 9 for the terminal ileum and descending colon, and n = 7 for ascending and transverse colon; # p<0.06, * p<0.05, and ** p<0.01 between HIV+ and healthy controls).
Figure 8.
Microbial translocation and systemic immune activation marker levels inversely correlate with colonic TJ transcript levels.
(A) Plasma samples from HIV+ individuals and healthy controls were thawed and analyzed by ELISA to measure levels of soluble CD14 (sCD14). Medians are denoted, with bars representing the interquartile range (n = 21 for both cohorts; # p<0.07). Plasma levels of (B) the immune activation marker sCD14 and (C) the microbial product LPS inversely correlate with transcript levels of claudin-4 and claudin-2, respectively, in the descending colon of subjects from both the HIV+ and healthy control cohorts. Spearman's rank test was used to determine correlations. The black lines represent the linear regression estimate.