Distinct Persistence Fate of Mycobacterium tuberculosis in Various Types of Cells

ABSTRACT Mycobacterium tuberculosis can invade different cells with distinct persistence fates because cells are equipped with different host restriction factors. However, the underlying mechanisms remain elusive. Here, we infected THP1 and Raw264.7 macrophages cell lines, A549 epithelial cell line, and hBMEC and bEnd.3 endothelial cell lines with M. tuberculosis and demonstrated that M. tuberculosis significantly inhibited lysosome acidification in THP1, hBMEC, A549, and Raw264.7 cells, while, in bEnd.3 cells, M. tuberculosis was mainly delivered into acidified phagolysosomes and auto-lysosomes. The systematic gene profile analysis of different cells and intracellular M. tuberculosis showed that the phagosome autophagy-pathway-related genes itgb3 and atg3 were highly expressed in bEnd.3 cells. Knockdown of these genes significantly increased the number of viable intracellular M. tuberculosis bacilli by altering phagosomal trafficking in bEnd.3 cells. Treatment with itgb3 agonist significantly decreased M. tuberculosis survival in vivo. These findings could facilitate the identification of anti-M. tuberculosis host genes and guide M. tuberculosis-resistant livestock breeding. IMPORTANCE As an intracellular pathogen, Mycobacterium tuberculosis could avoid host cell immune clearance using multiple strategies for its long-term survival. Understanding these processes could facilitate the development of new approaches to restrict intracellular M. tuberculosis survival. Here, we characterized the detailed molecular events occurring during intracellular trafficking of M. tuberculosis in macrophage, epithelial, and endothelial cell lines and found that ITGB3 facilitates M. tuberculosis clearance in endothelial cells through altering phagosomal trafficking. Meanwhile, the treatment with ITGB3 agonist could reduce bacterial load in vivo. Our results identified new anti-M. tuberculosis restriction factors and illuminated a new anti-M. tuberculosis defense mechanism.

epithelial and endothelial cells to find new avenues for spreading. When internalized into nonimmune cells, M. tuberculosis can replicate in epithelial cells but persist in brain microvascular endothelial cells (BMECs).
The ultimate killing of M. tuberculosis by host cells depends on multiple factors, such as the acidification of lysosomes, the process of autophagy, the delivery of antimicrobial peptides, and other restriction factors (2). Understanding these processes may greatly facilitate the development of new approaches to restrict intracellular M. tuberculosis survival. A few genome-wide studies have identified the host anti-M. tuberculosis restriction factors (3,4). Granulocyte-macrophage colony-stimulating factor (GM-CSF) produced by T cells, macrophages, or alveolar epithelial cells is vital for the host to inhibit intracellular M. tuberculosis growth (5,6). However, the restriction factors and their molecular defense mechanisms against M. tuberculosis in different cells are still not well understood.
The pathogenicity of microorganisms is often related to their capacity to survive and replicate within their hosts. While the phagosome-lysosome pathway functions effectively as the front line to kill the invading M. tuberculosis, the bacteria have evolved various elements of molecular machinery to escape from this clearance. M. tuberculosis avoids lysosomal fusion through alteration of endocytic trafficking, resulting in privileged multiplicative niches. In this scenario, it has been demonstrated that the phagosome markers Rab5 and Rab7, but not the lysosome markers LAMP and cathepsin, rapidly accumulate in the mycobacterial endosomal membrane in macrophages, epithelial cells, and endothelial cells (7)(8)(9). It has also been reported that the lack of acidification in the mycobacteria-containing phagosomes, due to the absence of lysosomal acid proteases on the phagosomal membranes, facilitates the survival of intracellular M. tuberculosis (10).
During the process of altering endocytic trafficking, ESAT6 secreted from M. tuberculosis can breach the phagosome membrane and lead to the translocation of M. tuberculosis into the cytosol for its survival (11,12). The permeabilization of the phagosomal membrane results in phagosomal and cytoplasmic mixing and allows extracellular mycobacteria DNA to access host cytosolic receptors, such as cGAS (13,14). The binding of mycobacteria DNA with cGAS induces type I interferon, which paradoxically facilitates M. tuberculosis replication and activates ubiquitin-mediated autophagy to restrict M. tuberculosis replication (15). Autophagy is involved in the innate defense against invading intracellular pathogens by sequestering them into autophagosomes and delivering them to the lysosomes (16,17). M. tuberculosis interferes with autophagyrelated pathways via various bacterial proteins. Dimycocerosate (DIM), a major virulence factor of M. tuberculosis, prevents phagosomal damage-independent autophagy and limits the acidification of LC3-positive M. tuberculosis compartments (18). Despite extensive documentation of the multiple functional intracellular pathogen-killing mechanisms in different cells, the underlying molecular mechanisms of M. tuberculosis clearance in different cells remains elusive.
The aim of this study is, therefore, to investigate the persistent fate and the underlying mechanisms of M. tuberculosis in different types of cells. By combining cell biology and molecular biology techniques, we characterized the detailed molecular events occurring during intracellular trafficking of M. tuberculosis and identified restriction factors in different types of host cells.

RESULTS
Distinct survival and replication fate of intracellular M. tuberculosis in different types of host cells. After being inhaled by a human being, M. tuberculosis can locate to the pulmonary region and invade alveolar macrophages and type II epithelial cells. M. tuberculosis can break down the epithelial-endothelial barrier and cause extra-pulmonary tuberculosis. For example, it infects brain microvascular endothelial cells and disrupts the blood-brain barrier to cause central nervous system tuberculosis (Fig. 1A). To investigate the survival and replication fate of M. tuberculosis in different types of cells, an intracellular bacteria viability assay was carried out in endothelial cells (bEnd.3 and hBMEC), macrophages (Raw 264.7 and THP1), and epithelial cells (A549). First, the viability of the infected cells was examined on days 2, 4, and 6 postinfection. As shown in Fig. 1B to F, the persistence of M. tuberculosis strain H37Ra had a profound influence on the growth of THP1 cells, but less so on bEnd.3, hBMEC, Raw 264.7, and A549 cells. The numbers of bEnd.3, hBMEC, Raw 264.7, and A549 cells, but not THP1 cells, were increased over time (Fig. S1A to E in the supplemental material).
Next, we investigated the viability of intracellular M. tuberculosis. Our results showed that the total number of viable M. tuberculosis H37Ra bacilli decreased in bEnd.3, THP1, and Raw 264.7 cells over time, but increased in hBMECs and A549 cells ( Fig. S1F to J). Similarly, M. tuberculosis strain H37Rv could replicate in neither bEnd.3 cells nor in hBMEC cells ( Fig. S1K and L). Considering that the replication of M. tuberculosis could cause cell death and lead to the loss of M. tuberculosis during washing steps, the number of viable bacilli per cell was used to compare the survival and replication capability of M. tuberculosis in different cells. We found that at day 6, the number of viable M. tuberculosis per cell was significantly increased in comparison to day 2 in THP1 and A549 cells, but not in other examined cells ( Fig. 1G to K). These data showed that M. tuberculosis could replicate in THP1 and A549 cells and could persist in hBMECs, bEnd.3, and Raw 264.7 cells.
Distinct phagocytosis processes of M. tuberculosis in different types of cells. Next, we investigated the phagocytosis process of M. tuberculosis in various cell lines by co-staining with different phagosome markers, such as Rab5 and Rab7, or lysosome markers, such as LAMP2 and cathepsin L ( Fig. 2A). As shown in Fig. 2B to F, M. tuberculosis H37Ra-containing vacuoles (MCVs) colocalized with the phagosome and lysosome markers. The association of MCVs with Rab5 was significantly decreased in bEnd.3 cells over time and maintained at a low level in both hBMEC and A549 cells, but maintained at a high level in Raw 264.7 and THP1 cells over time (Fig. 2G). During the infection with M. tuberculosis H37Rv, the association of MCVs with Rab5 was maintained at a low level over time in both hBMEC and THP1 cells ( Fig. S2A and B). Further, we analyzed the colocalization of MCVs with the late phagosome marker (Rab7) and lysosome marker Confocal microscopic images as indicated. Cells were seeded onto coverslips and infected with dsRed-expressing M. tuberculosis for 6 h. Cells were fixed at the indicated time points and incubated with rabbit anti-Rab5, rabbit anti-Rab7, or rabbit anti-cathepsin L and rat anti-LAMP2 antibodies, respectively. Alexa Fluor 488 goat anti-rabbit IgG (green) was used to detect Rab5, Rab7, and cathepsin L. Alexa Fluor 647 goat anti-rat IgG (purple) was used to detect LAMP2. (G to K) The colocalization of M. tuberculosis with Rab5 (G), Rab7 (H), Rab7/LAMP2 (I), LAMP2 (J), and cathepsin L (K) were quantified from 100 bacteria. All specimens were imaged with the confocal microscope and experiments were performed in triplicate and repeated two times. Data are presented as mean 6 SEM and analyzed by unpaired t test with two-tailed P values, with asterisks indicating statistically significant differences: *, P , 0.05; **, P , 0.01; ***, P , 0.001; ****, P , 0.0001.
(LAMP2). MCVs were highly colocalized with Rab7 in bEnd.3 cells, but with a relatively low level in hBMEC, Raw 264.7, and THP1 cells. In A549 cells, the colocalization of MCVs with Rab7 significantly increased over time (Fig. 2H), whereas in both hBMEC and THP1 cells during infection with M. tuberculosis H37Rv, the association of MCVs with Rab7 was maintained at a low level over time ( Fig. S2A and B). In Raw 264.7 cells, the colocalization of MCVs with both Rab7 and LAMP2 significantly decreased over time but was maintained at a low level in other examined cells (Fig. 2I). The colocalization of MCVs with LAMP2 was maintained at a low level in all examined cells over time (Fig. 2J, Fig. S2A and B).
As intracellular pathogens can be degraded in acidified lysosomes, we then analyzed the colocalization of MCVs with the lysosomal endopeptidase enzyme cathepsin L in all examined cells. The colocalization of MCVs with cathepsin L was maintained at a high level in bEnd.3 cells, but at a low level in Raw 264.7 and A549 cells over time. In hBMEC and THP1 cells, the colocalization of MCVs with cathepsin L was significantly decreased during M. tuberculosis H37Ra infection (Fig. 2K), while it was significantly increased during M. tuberculosis H37Rv infection over time ( Fig. S2A and B). These data demonstrate that different M. tuberculosis strains underwent a distinct phagocytosis process in different cells. Overall, M. tuberculosis arrested more at the early phagosomal maturation stage in macrophages, while it arrested more at a late stage in epithelial cells. Of note, in murine endothelial cells, M. tuberculosis was differentially delivered into acidified phagolysosomes for degradation.
Distinct autocytosis processes of M. tuberculosis in different types of cells. As autophagosomes could target disrupted MCVs toward lysosomes for degradation ( Fig. 3A), we then analyzed the role of autophagy in M. tuberculosis survival in bEnd.3, hBMEC, Raw 264.7, THP1, and A549 cells. As shown in Fig. 3B, MCVs were associated with the autophagy marker LC3 in all the examined cells. In bEnd.3 and Raw 264.7 cells, the colocalization of MCVs with LC3 was maintained at a low level over time. The colocalization of MCVs with LC3 was significantly decreased in hBMECs and THP1 cells but significantly increased in A549 cells over time (Fig. 3C). During the infection with M. tuberculosis H37Rv, the association of MCVs with LC3 significantly increased over time in both hBMEC and THP1 cells ( Fig. S2A and B).
For the colocalization of MCVs with both LC3 and LAMP2, no significant difference was observed in Raw 264.7 and THP1 cells, whereas, in bEnd.3 cells, hBMECs, and A549 cells, the colocalization of MCVs with both LC3 and LAMP2 was significantly increased (Fig. 3D). During M. tuberculosis H37Rv infection, a similar colocalization of MCVs with both LC3 and LAMP2 was observed in THP1 and hBMEC cells ( Fig. S2A and B). In all examined cells, the colocalization of MCVs with LAMP2 was maintained at a low level over time (Fig. 3E, Fig. S2A and B). Further, a Western blot analysis was performed to examine autophagy by detection of Beclin-1 and LC3. As shown in Fig. S3, more Beclin-1 and matured LC3II were observed in M. tuberculosis-infected A549, hBMEC, Raw 264.7, and THP1 cells than in the the uninfected control cells. However, in M. tuberculosis-infected bEnd.3 cells, Beclin-1 and matured LC3II were decreased at day 3 postinfection. These data demonstrated that M. tuberculosis might block autophagosome fusion with lysosomes in bEnd.3, Raw 264.7 and A549 cells, but were delivered into auto-lysosome in THP1 and hBMEC.
Next, we investigated the acidification of auto-lysosomes during M. tuberculosis infection in different cells. To this end, GML-bEnd.3 and GML-Raw 264.7 cell lines stably expressing an auto-lysosome formation and the chimeric fusion reporter chimeric gene GFPph-mCherry-LC3 were constructed by CRISPR/Cas9-mediated recombination and employed to compare the lysosomal acidification process of M. tuberculosis. As shown in Fig Considering the distinct trafficking processes of M. tuberculosis in examined cells, we compared the expression of phagosome-, lysosome-, and autophagy-related genes in different cells (Fig. 4D, Fig. S4B and C), and validated the expression of several genes Confocal microscopic images as indicated. Cells were seeded onto coverslips and infected with dsRed-expressing M. tuberculosis for 6 h. Cells were fixed at the indicated time points and incubated with rabbit anti-LC3 and rat anti-LAMP2 antibodies. Alexa Fluor 488 goat anti-rabbit IgG (green) was used to detect LC3. Alexa Fluor 647 goat anti-rat IgG (purple) was used to detect LAMP2. (C to E) The colocalization of M. tuberculosis with LC3 (C), LC3/LAMP2 (D), and LAMP2 (E) was quantified from 100 bacteria. (F and G) For the acidification assay, GML-bEnd.3 (F) and GML-Raw 264.7 (G) cells were seeded onto coverslips and infected with blue fluorescent protein (BFP)-expressing M. tuberculosis. The colocalization of M. tuberculosis with mCherry alone or with both GFP and mCherry was quantified from 100 bacteria. All specimens were imaged with the confocal microscope and experiments were performed in triplicate and repeated three times. Data are presented as mean 6 SEM and analyzed by unpaired t test with two-tailed P values, with asterisks indicating statistically significant differences: *, P , 0.05; **, P , 0.01; ***, P , 0.001; ****, P , 0.0001. by quantitative reverse transcriptase PCR (qRT-PCR) (Fig. S7). The protein-protein interactions (PPIs) network of DEGs (bEnd.3 versus Raw 264.7) was then constructed based on previous studies to further understand the function of DEGs (19) (Fig. 4E, Fig. S4D). By screening the potential genes involved in regulating M. tuberculosis trafficking in different host cells, we observed that the phagosome-related gene itgb3 and the autophagy-related gene atg3 were highly expressed in bEnd.3 cells (Fig. S4E, Fig. S7). As itgb3 is highly expressed in bEnd.3 cells, we then extracted the PPI partners of itgb3 from the DEG PPI network (Fig. 4F, Fig. S4F). Of note, the expression of most itgb3-interacting genes showed a significant difference between macrophages and endothelial cells (Fig. 4F, Fig. S4F). These data suggested that itgb3 might be involved in regulating M. tuberculosis phagosomal trafficking and contributing to the distinct persistence and replication fate of M. tuberculosis in different cells.
Upon invading host cells, M. tuberculosis needs to adapt to the intracellular hostile environment by reprogramming its gene expression profile. Thus, to explore the different gene expression profiles of intracellular M. tuberculosis in macrophages and endothelial cells, we isolated intracellular M. tuberculosis from Raw 264.7 and bEnd.3 cells, respectively, transcriptome sequencing (RNA-seq). As shown in Fig. S4G, significant divergence over the transcriptome level could be observed between the intracellular M. tuberculosis isolated from bEnd.3 and Raw 264.7 cells. According to the GSEA-KEGG analysis, we found the DEGs were enriched for fatty acid metabolism, biosynthesis of amino acids, oxidative phosphorylation, etc. (Fig. 4G). The DEGs involved in such pathways were significantly more highly expressed in M. tuberculosis isolated from Raw 264.7 cells than from bEnd.3 cells (Fig. S4H and I), suggesting these DEGs might be involved in the adaption to the hostile environment of Raw 264.7 cells.
As M. tuberculosis secreted proteins and cell wall proteins may interfere with host immune response to aid the bacterium's long-term survival, we constructed an M. tuberculosis-host PPI network between 17 M. tuberculosis DEGs and host proteins according to the previously established global M. tuberculosis-host PPI network (Fig. 4H). Among these genes, Rv1477, a secreted peptidoglycan hydrolase, was found to interact with ACTB, which is an interaction partner of ITGB3 (20). Meanwhile, we observed that Rv2280 interacted with CD14, which is also an interaction partner of ITGB3. It could be possible that these ITGB3-ACTB-Rv1477 and ITGB3-CD14-Rv2280 axes were involved in regulating the survival of M. tuberculosis in host cells, of which the underlying mechanisms needed to be further investigated.
Knockdown of itgb3 and atg3 enhanced M. tuberculosis survival in bEnd.3 cells. As the phagosome-related gene itgb3 and the autophagy-related gene atg3 showed the highest basal expression levels in bEnd.3 cells, we hypothesized that these genes might be involved in bacteria elimination and replication inhibition by interfering with trafficking pathways. To test this, we constructed itgb3 and atg3 knockdown bEnd.3 cell lines (Fig. 5A, Fig. S8A) and investigated the effect of knocked down itgb3 or atg3 on M. tuberculosis intracellular survival. Our data showed that upon atg3 or itgb3 knockdown, the intracellular survival of M. tuberculosis was significantly increased compared to the control (Fig. 5B). Furthermore, the colocalization of M. tuberculosis with Rab5 was increased upon itgb3 knockdown in bEnd.3 cells at day 3 postinfection, while the colocalization of M. tuberculosis with Rab5 was decreased upon atg3 knockdown at day 3 postinfection (Fig. 5C and D). Notably, the colocalization of M. tuberculosis with Rab7 was significantly increased upon itgb3 and atg3 knockdown in bEnd.3 cells (Fig. 5C and E). The colocalizations of M. tuberculosis with Rab7, LAMP2, cathepsin L, LAMP2, and LC3 were also investigated and showed no significant differences between  ITGB3 agonist treatment facilitated M. tuberculosis clearance in vivo. As ITGB3 may function as a potential host restriction factor for limiting M. tuberculosis replication, we further examined the role of ITGB3 in controlling M. tuberculosis survival in vivo through the treatment of mice with aspirin, an ITGB3 agonist (21). Upon treatment with aspirin, the bacterial load in the lung and spleen was significantly decreased compared to ethanol (ETOH)-treated control mice on day 7 postinfection (Fig. 5F and G). However, no significant difference was observed between ETOH-treated mice and aspirin-treated mice on day 21 postinfection (Fig. S9A and B).
Hematoxylin and eosin (H&E) staining showed that the alveolar wall was incomplete or thickened in the ETOH group. Of note, the aspirin treatment could recover the alveolar wall morphology on day 7 postinfection, but not on day 21 postinfection (Fig. 5H,  Fig. S9C), and there was no significant difference in the spleen between aspirin-treated groups and control groups (Fig. 5H to J, Fig. S9C and E). Furthermore, the histopathology quantification of inflammation showed that the aspirin treatment could reduce inflammation in the lung of mice on day 7 postinfection but not on day 21 postinfection (Fig. 5I, Fig. S9C and D). The possible reason might be that aspirin exerts its function mainly during the innate immunoresponse stage. Together, our data suggested that high expression of ITGB3 might facilitate bEnd.3 cells to eliminate invading M. tuberculosis by overwhelming phagolysosome maturation arrest and activation of ITGB3 could facilitate M. tuberculosis clearance in vivo (Fig. 5H).

DISCUSSION
As an aerosol transmission pathogen, M. tuberculosis is engulfed by macrophages, epithelial cells, and endothelial cells (7,8,22). Upon internalization into host cells, M. tuberculosis undergoes distinct trafficking processes in different cells. We observed M. tuberculosis strain H37Ra could arrest phagosome maturation at early stages in macrophages and late stages in endothelial and epithelial cells. M. tuberculosis H37Ra was degraded at a different level in acidified lysosomes (37.3% in THP1, 41% in Raw 264.7, 42% in hBMEC, and 19.6% in A549 cells). For strain H37Rv, the vacuoles maintained a low level of colocalization with phagosome markers over time in both THP1 and hBMEC cells. Notably, 86.3% of M. tuberculosis H37Ra was trafficked into the acidified lysosome for degradation in bEnd.3 cells, implying that the decreased numbers of viable M. tuberculosis in bEnd.3 cells might be caused by this acidified-lysosome-mediated degradation, although the M. tuberculosis could partially arrest late phagosome fusion with a lysosome in some unknown ways. These data suggest that different cells are equipped with different defense mechanisms to eliminate invading bacteria.
Autophagy is involved in innate defense against invading pathogens by sequestering them into autophagosomes and delivering them to lysosomes. We showed that M. tuberculosis could inhibit autophagosome formation in THP1 and hBMEC cells, but were delivered into auto-lysosomes for degradation in A549, Raw 264.7, and bEnd.3 cells. Our data demonstrate that M. tuberculosis can be degraded in acidified auto-lysosomes in bEnd.3 cells, but not in Raw 264.7 cells. By analyzing autophagy-related genes in different cells, we found that bEnd.3 cells highly expressed the atg3 gene. The E2 enzyme ATG3 conjugates the ubiquitin-like protein ATG8 to phosphatidylethanolamine to drive autophagosome formation (23). Upon atg3 knockdown, we observed that the viable intracellular M. tuberculosis was significantly increased and the phagosome maturation was inhibited at late stages. As bEnd.3 is a brain microvascular endothelial cell line often used as a blood-brain barrier model, it would be conceivable that this type of cell might be equipped with more robust immunologic machinery for brain pathogen defense.
Further, we analyzed the genes involved in the phagocytosis pathway and found that the basal expression of ITGB3 is significantly higher in bEnd.3 cells in comparison to other examined cells. Importantly, when ITGB3 was knocked down, M. tuberculosis survived better in bEnd.3 cells than in wild-type cells. Upon ITGB3 knockdown, the colocalization of MCVs with Rab7 was significantly enhanced, suggesting that less M. tuberculosis was delivered into lysosomes for degradation. Thus, we hypothesize that ITGB3 may facilitate M. tuberculosis clearance by delivering bacilli to phagolysosomes. Notably, in a cohort of TB patients, significant alteration of itgb3 mRNA was observed (24). Moreover, it has been reported that treatment with acetylsalicylic acid (aspirin), an ITGB3 agonist, in zebrafish could markedly reduce mycobacteria burden (21). Consistent with this, we found that aspirin treatment could attenuate M. tuberculosis survival in mice, indicating that ITGB3 might be a potential target for anti-M. tuberculosis drug development.
M. tuberculosis has a characteristic lipid-rich cell wall to counter host intracellular hostile events and allow its proliferation in a niche. The gene expression profiles of intracellular M. tuberculosis isolated from endothelial cells and macrophages showed that DEGs highly expressed in intracellular M. tuberculosis isolated from macrophages were enriched in fatty acid metabolism pathways. These data suggest that M. tuberculosis may upregulate its lipid synthesis-related gene expression to synthesize more lipids to counteract the hostile environments in host cells. Moreover, the expression of ATP synthase and NADH-quinone oxidoreductase genes were significantly lower in intracellular M. tuberculosis of bEnd.3 cells, which may reflect a reduced metabolic level and restricted life cycle of M. tuberculosis in these cells. Notably, we found that 17 DEGs interacted with host genes, of which Rv1477 (RipA) showed significantly lower expression in endothelial cells than in macrophages. Rv1477 is essential for M. tuberculosis cell wall synthesis and division and the depletion of Rv1477 could promote the clearance of M. tuberculosis in infected mice (25). As Rv1477 formed a complex with ITGB3 through ACTB, it would be important to further investigate whether Rv1477 may participate in ITGB3-mediated phagosomal maturation arrest.
Together, our data demonstrated that different cells are equipped with different levels of pathogen-host restriction factors, such as ITGB3 and ATG3, to eliminate pathogen invasion. M. tuberculosis could significantly inhibit lysosome acidification to promote replication or long-term persistence in THP1, hBMEC, A549, and Raw 264.7 cells, whereas M. tuberculosis is unable to replicate in bEnd.3 cells because the host cell can deliver them into acidified phagolysosomes and auto-lysosomes. Our data identified that ITGB3 and ATG3 may function as host restriction factors to eliminate M. tuberculosis in bEnd.3 cells. These findings may shed a light on the understanding of the innate immunity against M. tuberculosis invasion and facilitate the identification of anti-M. tuberculosis factors for drug development and transgenetic breeding of M. tuberculosisresistant animals.

Distinct Persistence Fate of M. tuberculosis
July/August 2021 Volume 6 Issue 4 e00783-21 msystems.asm.org 11 Construction of gene knockdown cell lines. Short hairpin RNAs (shRNAs) targeting atg3 or itgb3 were designed using BLOCK-iT RNAi Designer (Invitrogen) (shown in Table 1) and inserted into the pHKO vector. To obtain lentiviruses, HEK293T cells were seeded into 6-well plates and cotransfected with 1 mg pMD.2G, 2 mg pSPAX, and 2 mg pHKO-atg3-kd-shRNAs or pHKO-itgb3-kd-shRNAs. pHKO vector alone was used as the negative control. At 48 h posttransfection, the supernatants were collected and filtered with a 0.45-mm filter. Viruses were pelleted by centrifuging 30 min at 5,000 Â g at 4°C after overnight incubation with virus concentration solution. Viral pellets were resuspended, mixed with polybrene, and added to bEnd.3 cells. After 48 h, the medium was replaced with fresh medium containing 10% FBS and 4 mg/ml puromycin (s7417, Selleck) to purify the cell lines.
Quantitative reverse transcriptase PCR (qRT-PCR) was performed to detect the knockdown efficiency of target genes. RNAs (1 mg) were reverse transcribed into cDNA using ReverTra Ace qPCR RT Master Mix (Toyobo). Gene expression was analyzed by qRT-PCR using SYBR green Master Mix (Invitrogen) on the ABI 7300 system (Applied Biosystems). Data were analyzed using the threshold cycle (2 -DDCt ) method relative to GAPDH levels for each sample. The primers used for qRT-PCR are shown in Table 1.
Intracellular bacterial viability assay. Cells (5.0 Â 10 4 ) were seeded into a 24-well plate for 24 h and were incubated with M. tuberculosis (MOI = 10) for 6 h at 37°C in a 5% CO 2 incubator. The cells were washed three times with prewarmed phosphate-buffered saline PBS to remove extracellular M. tuberculosis and supplied with a fresh medium with 5% FBS containing amikacin (50 mg/ml) (referred to as day 0). The medium was changed every 2 days to avoid serum starvation-induced autophagy. The infected cells were lysed at the indicated time points using sterile 0.1% Tween 80 in water, and viable M. tuberculosis were enumerated by serial dilution of lysates and plating as described above.
Cell proliferation analysis. Cells (5.0 Â 10 4 ) were seeded onto coverslips in a 24-well plate for 24 h and were incubated with dsRed expressing M. tuberculosis (MOI = 10) for 6 h at 37°C with 5% CO 2 . The

TABLE 1 Oligonucleotides used in this study
CCAGGTGCCCGTGGTCGGT gyrB-qPCR-F AAGGACCGCAAGCTACTGAA gyrB-qPCR-R CAGACCTTCTGCACAAACGA cells were washed three times with prewarmed PBS to remove extracellular M. tuberculosis and supplied with fresh medium with 5% FBS containing amikacin (50 mg/ml). At the indicated time points, cells were detached with 0.25% (wt/vol) trypsin-0.03% (wt/vol) EDTA solution (Gibco) for 5 min at 37°C. Live cells were counted after staining with 0.4% trypan blue solution.
For trafficking assays, the infected cells were fixed at the indicated time points with 4% paraformaldehyde. After blocking with 5% normal goat serum in 0.3% Triton X-100 PBS for 1 h at room temperature, specimens were incubated with a 1:300 dilution of anti-Rab5 (Abcam), anti-Rab7 (Santa Cruz Biotechnology), anti-LAMP2 (Abcam), anti-cathepsin L (Abcam) or anti-LC3 (Sigma) antibody for 2 h at room temperature. Alexa Fluor 488 goat anti-rabbit IgG (Invitrogen) or Alexa Fluor 405 goat anti-rabbit IgG was used with a 1:300 dilution to detect Rab5, Rab7, cathepsin L, and LC3. Alexa Fluor 647 goat antirat IgG (Invitrogen) was used with a 1:300 dilution to detect LAMP2. The specimens were incubated with DAPI in PBS for 5 min. After washing 3 times with PBS, the specimens were mounted onto microscope slides using Prolong antifade reagent.
For auto-lysosomal acidification analysis, 5.0 Â 10 4 GML-bEnd.3 and GML-Raw 264.7 cells were seeded onto coverslips in a 24-well plate for 24 h and infected with BFP-expressing M. tuberculosis (MOI = 10) for 6 h at 37°C in 5% CO 2 . The infected cells were fixed with 4% paraformaldehyde overnight at the indicated time points and then mounted onto microscope slides using Prolong antifade reagent.
Images were obtained with an Olympus confocal laser microscope system equipped with FV10 ASW Imaging Software (Version 4.2; Olympus. Japan). One hundred bacteria in total were analyzed for each assay. All experiments were performed in triplicate.
Western blotting. Cells (1.0 Â 10 6 ) were seeded into a 6-well plate and then infected with M. tuberculosis (MOI = 10) for 6 h at 37°C in 5% CO 2 . The cells were lysed with radioimmunoprecipitation assay (RIPA) lysis buffer at the indicated time points. The cell lysates mixed with loading buffer were boiled for 10 min and cooled on ice for 10 min. The proteins were separated by 12.5% SDS-PAGE gels and then transferred to polyvinylidene difluoride (PVDF) membranes. The membrane was blocked with 5% (m/v) skim milk in Tris-buffered saline with Tween 20 (TBST) buffer (50 mM Tris, pH 7.5, 150 mM NaCl, 0.05% Tween 20) at room temperature for 1 h and incubated with primary antibodies against LC3, Beclin, ITGB3, or ATG3 (Abclonal) overnight at 4°C. The membrane was washed three times with TBS-0.05% Tween 20 and then incubated with horseradish peroxidase (HRP)-conjugated goat anti-rabbit or anti-mouse secondary antibody (1:5,000, ABclonal) for 1 h. The specific proteins were visualized using an electro-chemiluminescence ECL kit under Image Lab software 4.0.1 (Bio-Rad).
Host cells RNA-seq library preparation and sequencing. Cells (1.0 Â 10 6 ) were seeded into a 6well plate and then infected with M. tuberculosis (MOI = 10) for 6 h at 37°C in 5% CO 2 . The cells were lysed with 1 ml of TRIzol reagent at the indicated time points. Lysates were mixed with 0.2 ml of chloroform centrifuge for 10 min at 12,000 Â g at 4°C. Then, 0.4 ml of the aqueous phase was mixed with 0.4 ml of isopropanol in a new nuclease-free Eppendorf tube. The RNA was precipitated at room temperature for 10 min and pelleted by centrifugation for 10 min at 12,000 Â g. Eluted RNAs were washed twice with 75% ethanol (ETOH). The quality and quantity of RNAs were examined by Nanodrop 2000c spectrophotometer (Thermo Fisher Scientific). DNAs were removed using Ambion TURBO DNA-free DNase treatment and removal reagents (Invitrogen) following the user guide.
The libraries were generated with VAHTS Stranded mRNA-seq library prep kit (Vazyme Biotech Co., Ltd, China). Briefly, mRNAs were isolated using mRNA capture beads from 1 mg of total RNA and incubated for fragmentation at 94°C for 8 min to get mRNAs with a length of around 150 bp to 200 bp. The mRNAs were reverse transcribed to cDNA and converted into double-stranded cDNA molecules. Following end repair and dA tailing, the paired-end sequencing adaptors were ligated to the ends of the cDNA fragments and then subjected to library amplification and purification. The purified libraries were validated and quantified using the Agilent 2100 Bioanalyzer (Agilent Technologies) and sequenced with the Hi-Seq 10Â instrument (Illumina).
Intracellular M. tuberculosis RNA-seq library preparation and sequencing. Raw 264.7 and bEnd.3 cells were seeded into T225 flasks and infected with M. tuberculosis (MOI = 10) for 6 h at 37°C in 5% CO 2 . Infected cells were lysed with 30 ml of guanidine thiocyanate (GTC) buffer at day 3 postinfection and centrifuged for 20 min at 5,000 Â g to pellet bacteria. The bacteria pellets were suspended in 1 ml of GTC and centrifuged for 2 min at 15,000 Â g. The bacteria pellets were disturbed in 1 ml of TRIzol reagent by vortex for 10 min. Total RNAs were extracted as described above. Ribosomal RNA (rRNA) was removed under the guidelines of Ribo-Zero rRNA removal kit (Illumina). Library construction and sequencing were performed as described above.
For the differential analysis, all of the gene expression count tables were merged into a combined matrix showing the expression count of each gene from every individual sample. DEseq2 was used to export the foldchange list and normalized count matrix. The gene ID in the matrix was annotated with DAVID (https://david.ncifcrf.gov/).
The expression levels of autophagy, phagosome, and lysosome pathway genes were normalized corresponding to each sample. GSEA-KEGG analysis and KEGG enrichment were performed by R package clusterProfiler (28) and visualized by R package enrichplot (https://github.com/GuangchuangYu/ enrichplot). The polarized heatmap was generated by TBtools with the cells filling indicating the expression level. The interaction network was constructed by Cytoscape (29) with the host protein interaction based on string-db (20) and M. tuberculosis-host PPI based on our previous study (19). The bar plotted heat map was generated by R package pheatmap (https://CRAN.R-project.org/package=pheatmap). Differential expression genes (DEG) between cell lines were selected with the following principle: Abs (log 2 foldchange [gene expression in A cell comparing to uninfected A cell] 2log 2 change [gene expression in B cell comparing to uninfected B cell]) . 1 or gene expression cannot be detected in A or B cells, but abs (log 2 foldchange [gene expression in B or A cell comparing to uninfected B or A cell]) . 1.
Animal experiment. The animal experiment was carried out under the administrations of animal care and usage for research of the Ministry of Health in China. Animals were raised in specific-pathogen-free condition rooms. All the experiments were approved by the committee on the ethics of animal experiments of the College of Veterinary Medicine, Huazhong Agricultural University (HZAUMO-2020-0004).
C57BL/6 mice (6-week-old, female, 20 6 2 g) were purchased from the Laboratory Animal Center of Huazhong Agricultural University. All animals were randomly distributed into 3 groups, 5 to 7 of each. Mice were infected intravenously with M. tuberculosis at a dose of 5 Â 10 6 CFU. Then, mice received oral aspirin (ASA) at a daily dose of 10 mg/kg (30). The ETOH intake group was used as an untreated control and PBS was used as an uninfected control. These mice were administered with drugs for 6 days or 20 days and euthanized at day 7 or 21 postinfection, respectively. Lungs and spleens were used for histopathological observation and CFU analysis. A part of the left lung and spleen were fixed with 4% paraformaldehyde for 48 h, and tissues were paraffin embedded to make a section for hematoxylin and eosin (H&E) staining. Homogenates of lungs were plated onto Middlebrook 7H11 agar containing OADC enrichment and BBL MGIT PENTA antibiotics (Becton, Dickinson) for bacterial burden enumeration. CFU were counted after incubation at 37°C for 3 to 4 weeks. Experiments were repeated two times.
Statistical analysis. Data were analyzed and plotted using GraphPad Prism 6.0 (La Jolla, CA). Data are expressed as mean 6 standard error of the mean (SEM). Evaluation of the significance of differences between groups was performed by unpaired t test as indicated.
Data availability. The data sets and computer code produced in this study are available in the following databases: Host RNA-Seq data: Gene Expression Omnibus GSE125907 and M. tuberculosis RNA-Seq data: Gene Expression Omnibus GSE162200.

SUPPLEMENTAL MATERIAL
Supplemental material is available online only.