Reactivation of latent tuberculosis in cynomolgus macaques infected with SIV is associated with early peripheral T cell depletion and not virus load

Abstract

HIV-infected individuals with latent Mycobacterium tuberculosis (Mtb) infection are at significantly greater risk of reactivation tuberculosis (TB) than HIV-negative individuals with latent TB, even while CD4 T cell numbers are well preserved. Factors underlying high rates of reactivation are poorly understood and investigative tools are limited. We used cynomolgus macaques with latent TB co-infected with SIVmac251 to develop the first animal model of reactivated TB in HIV-infected humans to better explore these factors. All latent animals developed reactivated TB following SIV infection, with a variable time to reactivation (up to 11 months post-SIV). Reactivation was independent of virus load but correlated with depletion of peripheral T cells during acute SIV infection. Animals experiencing reactivation early after SIV infection (<17 weeks) had fewer CD4 T cells in the periphery and airways than animals reactivating in later phases of SIV infection. Co-infected animals had fewer T cells in involved lungs than SIV-negative animals with active TB despite similar T cell numbers in draining lymph nodes. Granulomas from these animals demonstrated histopathologic characteristics consistent with a chronically active disease process. These results suggest initial T cell depletion may strongly influence outcomes of HIV-Mtb co-infection.

Figure 1. SIV loads in plasma and PBMC of Mtb-SIV co-infected and SIV-only monkeys.

No significant differences in plasma, peripheral blood mononuclear cells (PBMC), bronchial alveolar lavage cells (BAL) or peripheral lymph node viral titer were observed between co-infected cynomolgus macaques (red line, n = 7) or the SIV-only group (blue line, n = 4). We were unable to obtain enough BAL cells from SIV-only monkeys to reliably measure virus load. The horizontal dashed line represents the detection limit for viral copy numbers. Lower error bars have been removed for clarity.

Figure 2. Acute CD4 T cell depletion correlates with reactivated TB.

A. Mean CD4 and CD8 T cell numbers from whole blood of co-infected macaques. Upper (CD8) and lower (CD4) error bars have been omitted for clarity. B. Mean CD4 T cell numbers over weeks 1–8 correlate with reactivation (red line: P = 0.011, R2 = 0.756) and necropsy time (blue line: P = 0.0007, R2 = 0.9172). C. Mean CD8 T cell numbers do not correlate with reactivation (red line: P = 0.112, R2 = 0.426) but do correlate to necropsy time (blue line: P = 0.020, R2 = 0.692). D. Animals that reactivate <17 weeks post-SIV infection have a trend toward lower frequencies of CD4 T cell in the BAL at 10 weeks post SIV infection than monkeys >26 weeks post SIV infection Man-Whitney (P = 0.057). 1207 (early) BAL cells were sampled 8 weeks post SIV infection.

Figure 3. Co-infected monkeys experience an increase in numbers of Mtb-specific T cells following early after SIV infection.

IFNγ ELISPOTs were performed on PBMCs stimulated with Mtb or SIV antigens. IFN-γ was measured in spot forming units (SFU) per 1×10⁶ cells.

Figure 4. Pathology associated with reactivation in co-infected monkeys.

A. Gross pathology at time of necropsy in co-infected monkeys. Weeks indicates time post-SIV infection for co-infected monkeys or post-Mtb infection for monkey 17706. Necropsy score represents the gross pathology score, CFU score indicates the bacterial number score. Lungs, thoracic lymph nodes, spleen, liver and kidney are represented. Granulomas are represented by blue circles, TB pneumonia by orange patches, enlarged thoracic lymph nodes by open ovals, and granulomatous thoracic lymph nodes by stars. Monkey 17706 is representative of pathology found in latent monkeys. B. Reactivation was quantified by gross pathology score, bacterial number score and percentage of tissue samples positive for Mtb in homogenized tissue. Co-infected monkeys are statistically different than latent monkeys (Mann-Whitney, gross pathology P = 0.0025, bacterial number score P = 0.004, % positive samples P = 0.005). Latent monkeys are also statistically different than active monkeys in gross pathology (Mann-Whitney, P = 0.001, bacterial number score P = 0.0004, and % positive P = 0.0004).

Figure 5. Histopathology associated with co-infected monkeys.

A. Actively disseminating tuberculous disease characterized by granulomatous inflammation infiltrating into the wall of a large bronchus and adjacent vessel. Hematoxylin and eosin (H&E) stain, 5× magnification. B. A “satellite” nodule (arrow) comprised of non-necrotizing granulomatous inflammation extends from the outer margin of a more chronic caseous granuloma, suggesting loss of immunological containment and reactivation. Arrowhead indicates mineralized area. H&E stain, 2× magnification C. A chronic granuloma with extensive central fibrous organization in conjunction with a dense circumscribing margin of peripheral fibrosis. H&E stain, 10× magnification. D. A solid fibrotic granuloma comprised of maturing fibroblasts depositing a collagenous matrix. H&E stain, 10× magnification.

Figure 6. Antigen specific T cells in involved and uninvolved tissue of co-infected monkeys.

Tissue homogenates were isolated from the granulomatous (involved) or tissues without grossly present granulomas (uninvolved) in lung. Monkeys not represented because too few cells were recovered from tissues at necropsy are indicated by asterisks.

Figure 7. CD4 and CD8 T cell numbers in lung and thoracic lymph nodes.

CD4 and CD8 T cell numbers in lung-draining thoracic lymph nodes (LN) are not significantly different between co-infected animals and animals with latent or active TB. T cell numbers in co-infected lung tissues were similar to animals with latent TB but significantly lower than animals with active TB (Mann-Whitney, CD4 P = 0.0023; CD8 P = 0.0078). Latent monkeys have fewer lung T cells than active monkeys (Mann-Whitney, CD4 P = 0.0014; CD8 P = 0.0018). Data points represent average T cell numbers from either thoracic LN or lung tissue from each monkey. Not all co-infected monkeys are not present due to limited tissue availability.