Aggregated Mycobacterium tuberculosis Enhances the Inflammatory Response

Abstract

Mycobacterium tuberculosis (Mtb) bacilli readily aggregate. We previously reported that Mtb aggregates lead to phagocyte death and subsequent efficient replication in the dead infected cells. Here, we examined the transcriptional response of human monocyte derived macrophages to phagocytosis of aggregated Mtb relative to phagocytosis of non-aggregated single or multiple bacilli. Infection with aggregated Mtb led to an early upregulation of pro-inflammatory associated genes and enhanced TNFα signaling via the NFκB pathway. These pathways were significantly more upregulated relative to infection with single or multiple non-aggregated bacilli per cell. Phagocytosis of aggregates led to a decreased phagosome acidification on a per bacillus basis and increased phagocyte cell death, which was not observed when Mtb aggregates were heat killed prior to phagocytosis. Mtb aggregates, observed in a granuloma from a patient, were found surrounding a lesion cavity. These observations suggest that TB aggregation may be a mechanism for pathogenesis. They raise the possibility that aggregated Mtb, if spread from individual to individual, could facilitate increased inflammation, Mtb growth, and macrophage cell death, potentially leading to active disease, cell necrosis, and additional cycles of transmission.

Keywords: Mycobacterium tuberculosis; TB pathogenesis; TNF-alpha; aggregation; inflammation; phagocytosis.

Figure 1

Mycobacterium tuberculosis (Mtb) aggregation changes the macrophage transcriptional response. (A) Monocyte-derived macrophages (MDMs) were infected with mCherry labeled, aggregated Mtb (bottom panel), single Mtb bacilli (middle panel), or uninfected (top panel) and sorted for RNA-Seq 3 h post-infection. Populations consisted of cells infected with aggregated Mtb (Gate P3—bottom panel), cells infected with single or few bacilli (Gate P2—middle panel), cells infected with multiple single bacilli (Gate P3—middle panel), or uninfected (Gate P1—top panel). Dead cells were excluded. X-axis on histograms is signal from mCherry fluorescent Mtb in MDMs, y-axis is count. (B) Principal component analysis (PCA) of the top 0.1% most variably expressed genes following rlog normalization and batch correction in R. Small circles are individual experiments (3 repeats from each of 5 blood donors). (C) Percentage contribution of individual genes used in the PCA. Color bar represents the percent contribution of individual genes to the first two principal components.

Figure 2

Individual genes and gene sets were differentially regulated in aggregate Mycobacterium tuberculosis (Mtb) infection. (A) Venn diagram showing the number of differentially regulated genes in single (blue), multiple (green) and aggregate Mtb (salmon) infected macrophages relative to uninfected macrophages as identified by DESeq2 differential expression analysis. (B) Heatmap showing read counts of genes identified as most variably expressed between infection conditions and selected by DESeq2 as significant (21 genes). (C) Normalized enrichment score (NES), expressed as percentage of maximum enrichment for the gene sets defined as “TNFα signaling via NF-κB” and “Inflammatory response.” Enrichment scores were calculated for all treatment comparisons and were significantly different at nominal p < 0.001 and FDR < 0.005, with the exception of the aggregate to multiple comparison for the “inflammatory response” gene set (where p < 0.05 and FDR = 0.24).

Figure 3

Aggregation state alters macrophage transcriptional response at the single gene level. Box plots of median and interquartile range of log normalized read counts from 15 independent infections of monocyte-derived macrophage (MDM) from 5 blood donors. Shown are expression levels as log transformed read counts in uninfected, single infected, multiple infected and aggregate infected macrophages. p-values are ns = not significant; * < 0.01; ** < 0.001; *** < 0.0001; **** < 0.00001; as determined by Kruskal–Wallace non-parametric test with Hochberg multiple comparison correction, with comparisons performed to nearest neighbor in graph (3 comparisons total).

Figure 4

Macrophage death is dependent on infection with live Mycobacterium tuberculosis (Mtb) aggregates. (A) Timelapse microscopy showing mCherry labeled Mtb (red) induced monocyte-derived macrophage (MDM) death as detected by DRAQ7 (green). (B) The number of dead cells in MDM infected with live Mtb (red bar), heat-killed Mtb (orange bar), or uninfected (blue bar) after 24 h. Shown are mean ±SD of DRAQ7 positive cells after 24 h. p-values are ns = not significant and * < 0.01.

Figure 5

Mycobacterium tuberculosis (Mtb) acidification per Mtb bacillus decreases with increasing Mtb aggregate size. (A) Image of lysotracker (green) colocalization with phagocytosed mCherry expressing Mtb (red). Scale bar is 20 μm (B) lysotracker fluorescence as a function of total aggregate area. The linear regression line is shown in black (R² = 0.63, p < 0.0001). (C) Ratio of lysotracker fluorescence to Mtb fluorescence as a function of Mtb area. Black line shows a model based on the surface area to volume ratio (R² = 0.25, p < 0.0001, black line).

Figure 6

Mycobacterium tuberculosis (Mtb) and Mtb aggregates are found near the periphery of a TB cavitary lesion. (A) Ziehl-Neelsen stain of a lung section. Aggregated bacilli are highlighted with a red circle and single bacilli with a blue circle. Scale bar is 3 mm. A black dashed circle is overlaid if the Mtb are in close proximity to a cell nucleus (blue stain). Sub-areas 1–9 are magnified in separate panels. Scale bars are 20 μm in the sub-areas. (B) Stacked histogram of the number of Mtb objects with varying numbers of Mtb bacilli found to be cell free or cell associated. (C) Stacked histogram of the total number of Mtb observed to be single bacilli or aggregates, and were found to be in close association with host cell nuclei.