Fig 1.
FDG PET CT analysis of Mtb infection in lymph nodes of cynomolgus and rhesus macaques.
A. PET/CT scans of 3 different macaques (monkey numbers 16213, 9811, 16113 showing different trajectories of thoracic lymph nodes at different time points post infection. B. Representative serial PET CT FDG SUVR plots showing several lymph nodes visible by PET at 2 weeks post infection in four different animals. Trajectories of individual lymph nodes in an animal is shown to be independent of each other. Each line is a lymph node. Dotted line represents the cut-off for calling FDG+ LNs “hot” (SUVR≥5). C. Most lymph nodes visible (SUVR≥2.3) on scan by PET 1–2 days before necropsy harbor live Mtb (top left panel), while only a small proportion of those that are not seen by PET have live Mtb (top right panel). Most”hot” lymph nodes (SUVR≥5) were CFU+ compared to only half of “warm” lymph nodes (SUVR 2.3–4.99) (bottom panels).
Fig 2.
Mtb burden and killing in lymph nodes of cynomolgus and rhesus macaques.
A, B. Live Mtb burden (CFU) in thoracic LNs from cynomolgus (A) and rhesus (B) macaques at various time points post-infection (at necropsy). Lymph node CFU of cynomolgus macaques decreases over the course of infection while rhesus macaques do not. C, D. Total (live+dead, CEQ) Mtb burden in cynomolgus (C) and rhesus (D) macaque LNs. There was no difference in the level of CEQ in cynomolgus macaque LNs over the course of infection, while a decline in CEQ was found in rhesus macaque LNs at later time points post-infection. E, F. Mtb killing in thoracic lymph nodes, as calculated as the ratio of live (CFU) compared to total (CEQ) bacteria. Cynomolgus macaque LNs (E) exhibit poor Mtb killing at 4 weeks post infection but improve over the course of infection. Highest Mtb killing capacity was observed in monkeys with latent infection (34–54 weeks post infection). Little killing was observed in rhesus macaque LNs (F). The CFU was transformed by adding 1 to reflect sterile LNs with CEQ and/or granulomas either by gross or microscopic examination. For C-F, only LNs in which CEQ were detected were included. Each macaque is shown in a different color. Each data point is one lymph node. Open symbols are sterile lymph nodes (CFU-). The number of macaques per time point post-infection is as follows: a.) 4–6 weeks (Cynos n = 8, Rhesus n = 4); b.) 11–14 weeks (Cynos n = 9, Rhesus n = 7); c.) 16–29 weeks (Cynos n = 9, Rhesus n = 8); d.) 34–54 weeks (Cynos n = 6). The number of lymph nodes analyzed ranged from 4 to 13 per macaque. Dotted line represent the limit of detection of our qPCR assay. Statistics are Kruskal-Wallis with post-hoc Dunn’s multiple test comparisons; p values are shown on figure.
Fig 3.
Proportion of thoracic lymph nodes infected with Mtb in cynomolgus and rhesus macaques at necropsy.
A. Percent of thoracic LNs that were CFU+ (red) or CFU- (blue) at necropsy. Rhesus macaques had higher proportion of Mtb-infected lymph nodes than cynomolgus macaques. B. Proportion of lymph nodes that were uninfected (CFU-/CEQ-, light blue), infected but were sterile (CEQ+/CFU-, purple), had culturable Mtb but no detected genome (CFU+/CEQ-, yellow) and with culturable Mtb and Mtb genome (CFU+/CEQ+, pink). Threshold for detection of CEQ is 1000 genomes/LN. CFU limit of detection is 20/LN. Number of macaques and lymph nodes at each time point is in Table 1.
Fig 4.
Most peripheral lymph nodes that had detectable Mtb genome were sterile.
A. The majority of peripheral lymph nodes assayed were CEQ- and CFU- (sterile). B, C. Live Mtb burden (CFU)(B) and total (live+dead, CEQ) Mtb burden (C) are significantly lower in peripheral lymph nodes than in thoracic lymph nodes. D. CFU/CEQ for thoracic and peripheral LNs. Killing capacity of peripheral lymph nodes is significantly higher (lower CFU/CEQ) compared to thoracic lymph nodes. These are data from 10 monkeys that had CEQ in both thoracic and peripheral lymph nodes. Each data point is one lymph node. Open symbols are sterile (CFU-) lymph nodes. The CFU was transformed by adding 1 to reflect sterile but CEQ+ LNs. Statistics are Mann-Whitney.
Fig 5.
Mtb infection results in granuloma formation and in some instances lymph node effacement.
A. Examples of gross pathology of thoracic lymph nodes from cynomolgus macaques that are minimally involved (left), with focal granuloma (middle) and severely effaced (right). The yellow arrow is pointing to a granuloma. B. Examples of microscopic histopathology of cynomolgus macaque lymph nodes that are not involved (left), with focal granuloma (middle) and severely effaced (right). The arrows are pointing to granulomas.
Fig 6.
Histologic and immunohistochemical characterization of Mtb-infected macaque lymph nodes with varying levels of disease.
FFPE tissue sections from Mtb-infected macaques were stained with hematoxylin and eosin (H&E) to show the tissue morphology and immunohistochemistry was performed on serial sections to identify the lymph node’s cellular, vascular, and structural elements. The box in the full-lymph node image indicates the region for the magnified panels (below) A. Lymph node showing no histologic evidence of disease and normal lymph node architecture. B. Lymph node demonstrating histologically-moderate disease where focal granulomas are present in T cell regions but not yet distorting the overall nodal architecture. Arrow indicates a granuloma. C. Severe lymph node disease showing large-scale disruption of the normal nodal structure in the vicinity of large coalescing granulomas. Arrow indicates a granuloma. Black scale bar (lower left) for the full-scale lymph nodes represents 2 mm. White scale bar (lower left, second column) for magnified image fields, represents 100 μm.
Fig 7.
Lymph node effacement promotes Mtb growth.
A. H&E and Auramine-rhodamine (A-R) staining of a severely-effaced thoracic lymph node (previously depicted in Fig 6C). The location of a large granuloma, indicated by white and grey dashed lines in the H&E and A-R panels, respectively, corresponds with substantial numbers of A-R-stained Mtb. Inset regions (black and white boxes in the H&E and A-R panels, respectively) show the interface between granulomatous and non-granulomatous lymph node regions. B, C. CFU (B) and CEQ (C) of thoracic lymph nodes with ≤50% or >50% effacement (effacement determined by H&E section). D. Mtb killing capacity (CFU/CEQ) of lymph nodes are not affected by the degree of effacement. Each data point is one lymph node. The CFU was transformed by adding 1 to reflect sterile but CEQ+ lymph nodes. Statistical test is Mann-Whitney.
Fig 8.
Immune response in thoracic LNs of cynomolgus macaques with granulomas.
Cytokine production in thoracic LN with granuloma in response to ESAT6+CFP10 (A-B) or PDBu and ionomycin (C-E) between LNs with bacterial burden (CFU+, red) and those that cleared (CFU-, blue). A. Frequency of CD11b+ cells producing IL-10 (n = 10 macaques; 27 LNs); B. Frequency of CD4+ T cells producing TNF (n = 24 macaques; 48 LNs). C. Frequency of CD3+ cells producing IFNγ(n = 12 macaques; 35 LNs); D. Frequency of CD8+ cells producing IFNγ(n = 12 macaques, 34 LNs) IL-2 (n = 10 macaques; 28 LNs) and TNF (n = 12 macaques; 34 LNs); E. Frequency of CD20+ cells producing IL-2 (n = 11 macaques; 20 LNs). Each symbol represents a LN. Statistical test is Mann-Whitney.
Table 1.
Animals used in the CFU, CEQ, and histological studies.