All research was conducted in accordance with the University of Minnesota's Intuitional Biosafety Committee (IBC) approval protocol 0806H36901. All animal studies were conducted in compliance with the recommendations of the University of Minnesota's Institutional guidelines and approved animal care and use committee (IACUC) under approval protocol 1207A17288.

MAP strains K-10 (cattle isolate) and K-10 GFP (pWes4) were grown in Middlebrook medium MB7H9) supplemented with 10% glycerol, 1% oleic acid-albumin-dextrose (OADC), and mycobactin J (2 mg/L) (Allied Monitor, Fayette, MO) at 37°C with shaking at 200 rpm until the optical density at 600 nm reached 0.3. Cultures were tested for purity using IS900 PCR (Sorge et al., 2013) and IS1311 PCR-RFLP analyses (Amonsin et al., 2004). M. smegmatis mc² 155 parent and pSM417-MAP0403 and vector transformants were cultured in Luria-Bertani (LB) Lennox broth at 37°C with shaking at 200 rpm. Hygromycin (100 μg/mL) or kanamycin (50 μg/mL) was supplemented to LB and MB7H9, respectively, when appropriate.

Unless otherwise noted, all mammalian cell culture and experiments were conducted at 37°C in a humidified chamber containing 5% CO₂. MDMs from JD-free dairy cows were elutriated using a 58% percoll gradient and matured in teflon wells as described in Coussens et al. (2002) and Janagama et al. (2006). MDMs were maintained in RPMI 1640 supplemented with 20% autologous serum. Bovine mammary epithelial cells (MAC-T) were maintained in Dulbecco's modified Eagle medium (DMEM) supplemented with 10% fetal bovine serum (FBS).

The ORF of MAP0403 was PCR amplified from MAP K-10 genomic DNA using a high-fidelity taq polymerase and primers 0403-HF (5′ CCC AAG CTT GTG ACG CAC TCG AAT GA 3′), 0403-SR (5′ ACA TGC ATG CTC AAC TGA CGC AGG A 3′) engineered with Hind III and Sph restriction sites at the 5′ end, respectively. The amplified fragment was cloned into a restricted pSM417 vector (herein referred to as pSM417-MAP0403 using T4 DNA ligase and electroporated (single pulse generated at 2.5 kV, 1000Ω) into competent M. smegmatis mc² 155 cells. Competent M. smegmatis mc² 155 were created using the method described by Goude et al. (2015). Insert orientation and sequence fidelity were confirmed by classic Sanger sequencing at the University of Minnesota's Genomics Center. A vector control strain (pSM417 alone) in M. smegmatis mc² 155 was also created.

MDMs and MAC-T cells were used to study global gene expression profiles during initial MAP infection and/or MAP0403 expression. We used established methods described by our laboratory (Zhu et al., 2008; Lamont and Sreevatsan, 2010; Lamont et al., 2012). Briefly, 2 × 10⁶ MDMs/flask were seeded into 25 cm² flasks, incubated and allowed to adhere for 2 h. Successful cell attachment was confirmed by phase-contrast microscopy. Non-adherent cells were removed by washing with PBS prior to infection. MAP K-10 and M. smegmatis mc² 155 expressing pSM417 and pSM417-MAP0403 were pelleted at 500 × g for 15 min, separately re-suspended in RPMI 1640 containing 2% autologous serum, and passed multiple times through a syringe-driven 18 G needle. MDMs were infected separately with the above mycobacterial strains at a MOI of 10:1 for 2 h. Upon completion of incubation, cells were washed thrice with pre-warmed PBS and treated with amikacin for 2 h to remove extracellular bacteria. Fresh medium was added to 25 cm² flasks and incubated for 10, 30, and 120 min. After post-infection, MDMs were again washed in PBS and were either lysed in PBS containing 0.1% Triton X-100 or collected for RNA extraction. Triton X-100 lysates underwent differential centrifugation to separate host cellular debris from bacteria cells. Bacteria were resuspended in 1.0 mL of PBS, serially diluted on MB7H9 or LB-Lennox agar, and incubated at 37°C to determine CFUs. Hygromycin (100 μg/mL) was added to LB-Lennox agar when appropriate. MAC-T cells were seeded at a density of 2 × 10⁴ cells/well in a 24 well polysterene plate and allowed to reach 80% confluence. The same invasion method and post-infection processing as MDMs were applied to MAC-T cells with the exception that the initial infection time point was 3 h. Each time point and condition were conducted in biological triplicates. Each experiment was replicated three times.

MDMs and MAC-T cells were separately seeded at 2 × 10⁴ cells/well in a 24 well plate containing No. 1.5 glass cover slips. MAP invasion using MAP K-10(pWes4)-GFP was conducted using the same conditions described above with the exception of an 1 h pre-incubation step with 50 nM of bafilomcyin A1 (A.G. Scientific Inc., San Diego, CA). Phagosome acidification was determined by a LysoTracker staining method (Lamont et al., 2012). Upon the final 30 min of each post-infection time point, 25 nM of LysoTracker Blue DND-22 (Invitrogen, Carlsbad, CA) was added to cell medium. Cells were subsequently washed thrice in Dulbecco's PBS, incubated in pre-warmed Deep Red CellMask plasma membrane stain (2.5 μg/mL) (Invitrogen, Carlsbad, CA) for 5 min, and rewashed. Cells were fixed in absolute methanol for 5 min at −20°C and washed twice in ice-cold D-PBS. Slides were stored at 4°C until visualization by confocal microscopy. An Olympus FluoView 1000 upright confocal microscope (Olympus, South-end-on-sea, Essex, United Kingdom) was used to image infected and control cell slides slides with FITC, Cy5, and DAPI lasers. Z-series for each slide was taken in 1 μM steps and stacked to render a complete image. A minimum of three fields per slide were imaged. Blocking assay was repeated in triplicate for each condition.

All work surfaces and equipment were treated with RNase Away (Molecular Bioproducts, Inc., San Diego, CA). Total RNA was extracted from MAP infected MDMs treated with/out bafilomycin A1, acid treated and control bacteria using TRIzol reagent (Invitrogen, Carlsbad, CA) per manufacturer's instructions. Bacterial lysates were homogenized using a reported bead-beating method (Janagama et al., 2010). RNA samples were treated with the TURBO DNA-free Kit (Ambion, ThermoFisher Scientific, Rockford, IL) and subjected to PCR to confirm that the samples were devoid of genomic DNA. RNA quality and concentration was determined by measuring the 260/280 ratio on a NanoDrop ND 1000 spectrophotometer (Nanodrop Products, Wilmington, DE).

RNA extracted from MAP K-10 infected MDMs treated with/out bafilomycin A1at 30 min p.i. were processed and hybridized as described (Janagama et al., 2010). Briefly, total RNA treated with DNase was processed to remove host RNA and the 16S ribosomal RNA using MICROBEnrich and MicrobExpress kits (Ambion, Thermo Scientific, Rockford, IL) as specified by manufacturer. Microbial RNA was amplified using the MessageAmpII Bacteria kit for prokaryotic mRNA per manufacturer's instructions (Ambion, ThermoFisher Scientific, Rockford, IL). Labeled (Cy3 or Cy5) DNA was produced from microbial mRNA using the Superscript Plus Direct cDNA labeling system (Invitrogen, Carlsbad, CA) with aminoallyl-dUTP followed by a coupling of the aminoallyl groups to either Cyanine-3 or Cyanine-5 (Cy-3/Cy-5) fluorescent molecules. cDNA reactions were pooled by treatment group to obtain a sufficient concentration of labeled DNA. Effective labeling was achieved by incubating the aminoallyl-dUTP coupled cDNA with the dye for 2 h at room temperature (RT). Labeled cDNA was hybrized to 70-mer oligonucleotide microarray slides (National Animal Disease Center, Ames, IA) overnight at RT, washed in microarray buffer and scanned using the HP Scanarray 5000 (Perkin Elmer Inc., Waltham, MA). All images were stored. Microarray experiments were repeated three times. Raw microarray data files have been submitted to NCBI Gene Expression Omnibus (GEO) (Edgar et al., 2002) and are accessible through GEO series accession number GSE84708.

All microarray experiments were conducted using the minimal information about a microarray experiment (MIAME) guidelines. Microarray image analysis software, BlueFuse (BlueGnome Ltd., Cambridge), was used to extract numeric data from stored microarray images. Normalization by global LOWESS was performed and expression data was analyzed by GeneSpring GX 10 (Agilent Technologies, Foster City, CA). Normalized ratios were reported as fold change. Differentially expressed genes (DEGs) were cross-referenced to the MAP K-10 genome and the remaining mycobacterial genomes listed in National Center for Biotechnology Institute (NCBI) using Basic Local Alignment Search Tool (BLAST) alogorithm. MAP gene networks were analyzed by the STRING database (http://www.string-db.org).

One-step qRT-PCR was performed using QuantiFast SYBR Green One-Step QRT-PCR mix (Qiagen, Valencia, CA) in a LightCycler 480II (Roche, Madison, WI) with corresponding software. The following program was used: 50°C for 10 min, 95°C for 5 min (activation), 95°C for 10 s (denature), and 95°C for 15 s, 60°C for 1 min repeated for 40 cycles (PCR amplification). Primers used for qRT-PCR were designed using Primer 3 software (http://frodo.wi.mit.edu/primer3/) and are listed in Table S1. Fold changes were calculated using the 2^−ΔΔCt method (Livak and Schmittgen, 2001). The value of the housekeeping gene, secA, was normalized to untreated bacteria cells. All samples were conducted in technical triplicates.

MAP K-10 (OD₆₀₀ = 0.3) in MB7H9 was vortexed for 5 min and passed 10 times through a sterile 18 G syringe-driven needle to disperse bacterial clumps. The culture was allowed to stand for 5 min to facilitate the sedimentation of bacterial clumps. The upper two-thirds of the culture (containing a single cell suspension) were used in all subsequent experiments. Acidity of test cultures was adjusted to a pH of 5 by 2N HCl, while control cultures were maintained at a pH of 6.6 ± 0.2. All cultures were mixed and allowed to rotate at 37°C for 10, 30, and 120 min. Each condition was conducted in biological triplicates. All experiments were repeated three times. CFUs were determined by plating serial dilutions of the suspension in duplicate on MB7H9 agar.

Log-phase M. smegmatis mc² 155 parent and pSM417-MAP0403 transformants were labeled with 5(6)-carboxyfluorescein N-hydroxysuccinimide ester (5(6)-CFDA) (Sigma, St. Louis, MO) as described in Gaggìa et al. (2010). Briefly, 5.0 mL of bacterial cells were harvested by centrifugation at 13,000 rpm for 5 min, resuspended in 980 μl of filter (0.45 μM pore size) sterilized citric acid–phosphate-buffer (pH 7.0) supplemented with 10 μL of 1 M glucose and 10 μL of 5(6)-CFDA, and incubated for 1 h at 37°C. Upon completion of incubation, cells were sedimented at 13,000 rpm for 5 min, washed thrice in phosphate buffered saline (PBS), and resuspended in LB broth (pH = 7). A standard curve relating fluoresce with decreasing pH was established by irreversible membrane permeablization of M. smegmatis mc² 155 parent using ethanol (63% v/v) incubation at 37°C for 30 min. Bacteria were pelleted as described above, resuspended in 5 mL of LB broth adjusted to pH range of 4–8, and incubated at 37°C for 30 min. to equilibrate pH_IB to medium pH. Fluorescence at each pH was measured using a SpectraMax 2000 spectrophotometer (492 nm excitation, 517 nm emission) (Molecular Probes, San Diego, CA) and the mean fluorescence units were plotted against pH. Acid treatment for parent and pSM417-MAP0403 strains was repeated with the exception of irreversible membrane permeabilization. Fluorescence units were compared against the standard curve to determine pH_IB. The experiment was thrice repeated.

The results pertaining to relative fold changes and CFUs were analyzed by two-way analysis of variance (ANOVA) with Bonferroni correction in Graphpad Prism software (GraphPad Software, La Jolla, CA). P < 0.05 were considered to be statistically significant. Fitness of M. smegmatis transformants in MDM infection is shown as a box-whisker plot to demonstrate the actual distribution of observed CFUs.

In order to identify the acid stress transcriptome during initial MAP infection, we first sought to characterize the phagosome maturation process in bovine MDMs. MAP K-10 was allowed to invade MDMS for 10–120 min post-infection (p.i.) and phagosomes were subsequently assessed for acidification using LysoTracker Blue, a fluorescent, acidotropic probe, that emits upon protonation of its basic amine. We observed phagosome acidification as early as 10 min in MDMs (Figure 1A). Phagosome acidification continued for 1 h (Figure 1A); however, the acidification process completed by 2 h p.i. (data not shown). Phagosome acidification was validated by pre-treatment of MDMs with bafilomcyin A1, a potent vacuolar ATPase inhibitor. Bafilomcyin A1 treatment of MDMs abrogated phagosome acidification at all p.i. time points (Figure 1A).

We selected a 30 min p.i. time point to perform a microarray analysis of genes expressed during phagosome acidification in MDMs. We compared expression profiles from MDMs infected with MAP K-10 with and without bafilomycin pre-treatment. Microarray analysis identified a MAP acid stress related network composed of genes MAP0403, MAP_RS20120 (MAP3922), MAP_RS12430 (MAP2439c), MAP_RS02045 (MAP0401), MAP_RS02035 (MAP0399c), rlmN, mutY, and nth (Figure 1B). All genes within the acid stress network were validated by qT-RT-PCR (Figure 1B).

MAP0403 was the center node of the acid stress network. A Basic Local Alignment Search Tool (BLAST; https://blast.ncbi.nlm.nih.gov/Blast.cgi) protein analysis predicted that MAP0403 encodes a serine protease that is 86% identical and 92% similar to MarP found in M. tuberculosis. MAP0403 is composed of 397 amino acids and is projected by transmembrane helix prediction software (TMHMM) to have 4 transmembrane helices at the N-terminus, similar to MarP (Figure S1A). The C-terminal portion is predicted to contain a trypsin-like protease domain and is likely localized to the periplasm. In MarP, oxidative stress triggers autocleavage of the protease domain and stabilization of the active site through reduction of disulfide bonds and subsequent activation of the domain's protease activity (Cameron et al., 1994; Vandal et al., 2008, 2009; Biswas et al., 2010). Cysteine disulfide bond prediction of MAP0403 using DiANNA ver. 1.1 web software showed a disulfide bond located at amino acid positions 214-397 (SLAPSCQKVLE–VGTGSCVS), which is also present in MarP. Further motif analysis showed that like MarP and other serine proteases, including eukaryotic chymotrypsin proteases, MAP0403 contains the conserved catalytic triad that functions at the active site of transferase enzymes (Figure S1B). Together these data suggest that MAP0403 is likely the equivalent of MarP in MAP. Given the role of MAP0403 as the central node in an acid stress network and its conservation to MarP, we focused on elucidating the association of MAP0403 with acid stress and pH_IB.

To further understand MAP0403 and its role in bacterial survival during early events in host infection, we analyzed MAP0403 transcription in MAP K-10 exposed to MDMs treated with and without bafilomycin A1. As previously performed, MAP K-10 was allowed to invade MDMs for 10, 30, and 120 min p.i. and MAP0403 transcription was analyzed by qT-RT-PCR. MAP0403 was up-regulated by a 30-fold difference compared to control MAP K-10 within 10 min (Figure 2A). MAP0403 upregulation was sustained throughout all p.i. time points; however, expression was reduced by 14-fold at 120 min (Figure 2A). Pre-treatment of MDMs with bafilomycin A1 decreased MAP0403 expression at all p.i. times (Figure 2A). We next sought to analyze MAP0403 transcription during infection within another cell type to determine if expression occurs only in a specific host cell. We used MAC-T cells, a mammary epithelial cell line, as a surrogate (Patel et al., 2006) for the intestinal epithelium—the first tissue encountered by MAP during natural infection. In a previous study, we showed that phagosome acidification of MAC-T cells infected with MAP K-10 occurred within 10 min and was sustained until 60 min p.i. (Lamont et al., 2012). Unlike MDMs, MAP0403 differential expression was not detected until 30 min p.i. (Figure 2B). The 30-fold upregulation of MAP0403 compared to the bacteria control corresponded to peak acidification in MAC-T cells (Figure 2B) (Lamont et al., 2012). MAP0403 expression decreased by 120 min p.i. and showed a 16-fold difference compared to control (Figure 2B). As expected, pre-treatment of MAC-T cells with bafilomycin A1 abolished MAP0403 expression at all p.i. time points (Figure 2B).

We further interrogated the role of MAP0403 during cell infection by asking whether or not expression was critical to bacteria survival. Attempts by our laboratory to create a MAP0403 deletion in MAP have been unsuccessful. Therefore to answer this question, we utilized the non-pathogenic mycobacteria, M. smegmatis mc² 155, as a surrogate host to express MAP0403 due to its inability to survive in macrophages past 48 h (Lagier et al., 1998; Kuehnel et al., 2001; Anes et al., 2003, 2006). It is important to note that M. smegmatis mc² 155 contains a gene, MSMEG_6183, that is predicted to encode a serine protease that shares 66 and 68% amino acid identities with MarP and MAP0403, respectively. However, we found that MSMEG_6183 is not up-regulated during MDM infection or under in vitro acid stress (pH = 5) compared to control bacteria at 10-120 min p.i. (data not shown). M. smegmatis mc² 155 was electroporated with a plasmid expressing MAP0403 from an Hsp60 promoter (M. smegmatis pSM417-MAP0403) or an empty vector [M. smegmatis pSM417(empty)]. M. smegmatis mc² 155 parent and transformants were exposed to MDMs for 0, 10, 30, and 120 min p.i. and assessed for bacterial viability. M. smegmatis expressing MAP0403 showed a greater than 1.5 log₁₀ increase in CFU recovery compared to M. smegmatis controls within 10 min p.i. (Figure 2C). Increased recovery of M. smegmatis pSM417-MAP0403 vs. controls was maintained throughout MDM infection (Figure 2C). MAP0403 transcription was evaluated in M. smegmatis mc2 155-MAP0403 in response to MDM infection (Figure S2). However, MAP0403 expression was not differentially expressed to the control time point at 0 min (Figure S2A). This is likely due to its constitutive expression under the pSM417 Hsp60 promoter. In summary, these data provide initial support that MAP0403 is expressed during host infection in different cell types undergoing phagosome acidification and likely aids in bacterial cell survival.

Although phagosome acidification was observed in both MDMs and MAC-T cells, acid stress represents one of several host stresses against MAP and other factors including Reactive Oxygen Species (ROS) and Reactive Nitrogen Intermediates (RNI), etc. were likely present within the phagolysosome. Therefore, the sole contribution of acid to MAP0403 transcription could not be determined. In order to understand the direct influence of acid stress to MAP0403 expression and bacteria survival, we took a reductionist approach by employing acid stress in vitro to MAP K-10 and M. smegmatis mc² 155 pSM417-MAP0403. MAP K-10 was separately incubated with acidified (pH = 5) and untreated broth (pH = 6.8) for 10–120 min and interrogated for MAP0403 differential expression. Acid treatment resulted in a four-fold upregulation of MAP0403 within 10 min (Figure 3A) compared to control. MAP0403 transcription increased to ~8-fold by 30 and 60 min, but decreased by 120 min (Figure 3A).

MAP K-10, M. smegmatis mc² 155 pSM417-MAP0403, and M. smegmatis mc² 155 pSM417(empty) were incubated in the same conditions described above and analyzed for recovery using direct plating. Within 10 min of acid treatment, CFUs for all strains were reduced by 0.5-1 log₁₀ compared to inoculum (Figure 3B). However, by 120 min of acid exposure MAP K-10 and M. smegmatis mc² 155 pSM417-MAP0403 recovered and stabilized in contrast to M. smegmatis mc² 155 pSM417(empty), which showed a 1-fold reduction in CFUs (Figure 3B). We analyzed MAP0403 transcription in M. smegmatis mc²155 pSM417-MAP0403 in response to acid in vitro (Figure S2B). Similar to pSM417-MAP0403 transcription during MDM infection, data showed that MAP0403 was not differentially expressed (Figure S2B).

Lastly, we asked if M. smegmatis mc² 155 pSM417-MAP0403 viability during in vitro acid stress corresponded to maintenance of its pH_IB. A 5(6)-CFDA fluorescence ratio microplate assay was used to determine pH_IB (Siegumfeldt et al., 1999; Gaggìa et al., 2010). A calibration curve was created to establish the link between fluorescence and pH (range 4–8; Figure S3). Increasing fluorescence was positively correlated to pH (Figure S3). M. smegmatis mc² 155 pSM417-MAP0403 and M. smegmatis mc² 155 stained with 5(6)-CFDA were exposed to a pH range of 4–7 for 30 min, measured for fluorescence, and compared against the calibration curve. Both strains maintained a neutral pH_IB when exposed to extracellular alkaline pHs as evidenced by an increased fluorescence (Figure 3C). However, only M. smegmatis mc² 155 pSM417-MAP0403 was able to sustain a neutral pH_IB when incubated with medium acidified to pH = 5 (Figure 3C). All strains at pH = 4 failed to show a neutral pH_IB(Figure 3C). Overall, data showed that extracellular acid exposure is one signal for MAP0403 expression and similar to intracellular survival, MAP0403 supports bacteria viability when met with an acid stressor.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.