Pyruvate carboxylase is critical for biofilm formation in Mycobacterium tuberculosis

Abstract

Mycobacterium tuberculosis (Mtb) forms biofilms. Biofilm formation is critical in Mtb’s virulence. The metabolic pathways governing mycobacterial biofilm formation remain largely unexplored. This study evaluated how different carbon sources influence Mtb biofilm formation. Fermentable substrates—glucose, glycerol, and pyruvate—significantly promoted thick, mature biofilm formation compared to non-fermentable alternatives. In this study, we analyzed the role of pyruvate carboxylase (Pca), an anaplerotic enzyme that converts pyruvate to oxaloacetate, in biofilm formation, as it regulates carbon metabolic flux into the tricarboxylic acid (TCA) cycle and gluconeogenesis. To this end, a transposon mutant of the pca gene was investigated for its ability to form biofilms. We observed that the pca transposon mutant is deficient in pellicle, submerged, and macrocolony biofilm formation. Confocal microscopy indicated that the pca mutant does not accumulate extracellular cellulose in the biofilms and has lower biomass. These defects could be rescued upon complementing the mutant with an episomal copy of the pca gene or adding glucose or pyruvate to the medium. These observations suggest that biofilm formation requires a regulated flow of carbon flux into gluconeogenesis and the TCA cycle to ensure the supply of precursors for EPS biosynthesis and sustained energy production to fuel EPS synthesis.

Introduction

Tuberculosis (TB) is caused by Mycobacterium tuberculosis (Mtb). It is one of the deadliest infectious diseases, leading to the deaths of more than a million people each year in the last decade¹. Importantly, TB is a chronic infectious disease, and chronic infections are often associated with biofilms^2,3. Recently, we have demonstrated that Mtb forms cellulose-rich biofilms⁴. We have used cellulose to detect the presence of Mtb biofilms in the lungs of experimentally infected mice, non-human primates, and in human autopsy samples⁵. We have further demonstrated that the biofilms protect resident Mtb cells from the immune system and antimycobacterial drugs⁵. These observations bring the mechanisms involved in mycobacterial biofilm formation, biofilm-mediated immune evasion & drug tolerance, the intricacies of dispersion, and the metabolic pathways governing biofilm formation and dispersion to the forefront of TB research.

Central metabolism comprises tightly interconnected core metabolic pathways, including glycolysis, the tricarboxylic acid (TCA), and the pentose phosphate pathway⁶. The metabolic pathways that govern bacterial growth during the planktonic mode of life are well studied, but are thought to be profoundly different from those that govern biofilm development, maintenance, and dispersion⁷. During planktonic growth, the cellular metabolism harvests reductive biochemical energy and invests it in the synthesis of macromolecules involved in bacterial multiplication. While in biofilm mode, reductive energy is harvested and invested in the biosynthesis of extracellular polymeric substances (EPS). This metabolic shift is crucial for biofilm development and the exhibition of drug tolerance. Only a few manuscripts have investigated the modulation of central metabolism during the shift from planktonic growth to the biofilm mode of life. In Bacillus subtilis, a concerted upregulation of the TCA, alterations in nucleotide biosynthesis, and amino acid synthesis were observed, which were viewed as a cellular effort aimed at boosting extracellular matrix (ECM) production⁸. On the other hand, ¹³C-labeling coupled with flux analysis suggested that Pseudomonas aeruginosa PAO1, the carbon flux in the central metabolic pathways and anaplerotic nodes is not much perturbed between the planktonic growth and biofilms⁹.

Although the metabolic pathways involved in mycobacterial biofilm formation have not been much explored, the central metabolism in Mtb growing in planktonic cultures is well studied. Extensive studies have suggested that Mtb possesses a highly flexible central metabolism¹⁰. This metabolic flexibility plays a key role in its ability to infect, survive, replicate, and cause productive infection. Inside the host, Mtb primarily utilizes host lipids, fatty acids, and cholesterol for fueling its central metabolism¹¹. Inside the host, Mtb is dependent on gluconeogenesis for the generation of glucose and other related carbon substrates in catabolic pathways^12–14. Gluconeogenesis is a metabolic pathway by which glucose is generated from non-carbohydrate carbon sources such as pyruvate, glycerol, acetate, lactate, and other glucogenic amino acids (Fig. 1). It is also known as the reverse of the glycolysis pathway, as the enzymes in glycolytic and gluconeogenic pathways are the same, catalyzing readily reversible reactions¹⁵. The only irreversible reaction in the gluconeogenesis pathway is catalyzed by fructose-1,6-bisphosphatase. Phosphoenolpyruvate is usually the entry point into the gluconeogenesis pathway, but it can vary depending on the carbon substrate¹⁶. This energy-consuming pathway generates carbohydrates for the bacteria¹⁷. Gluconeogenesis plays a critical role in mycobacterial survival inside the host^12–14. Gene encoding for pyruvate carboxykinase (pckA) is required for full virulence of Mtb¹². In addition to pckA, glpX, and gpm2, which encode fructose-1,6-bisphosphatase enzymes in the Mtb gluconeogenic pathway, are also required for virulence^13,18. Mutations of the above-mentioned genes disrupt gluconeogenesis, resulting in Mtb attenuation in in vivo models¹⁸.

Fig. 1. Schematic representation of the gluconeogenesis/glycolysis pathway.

The schematic shows the glycolysis pathway. The gluconeogenesis pathway is typically the reverse of glycolysis, with few key enzymatic steps involved in producing glucose from non-carbohydrate carbon sources, such as pyruvate or phosphoenolpyruvate. Additionally, the model represents the entry of glycerol molecules into the pathway via conversion to Dihydroxyacetone phosphate by glycerol kinase. This suggests that glycerol can also be utilized and fed into the gluconeogenesis pathway for producing glucose molecules. The enzymes in red are involved in the gluconeogenesis pathway, whereas those in blue are involved in the glycolysis pathway (Created in BioRender. Kumar, A. (2025) https://BioRender.com/c58q988).

Gluconeogenesis plays a crucial role in bacterial and yeast biofilm formation. Recent research on Saccharomyces cerevisiae, Staphylococcus aureus, Salmonella enterica serovar Typhimurium, Vibrio cholerae, and others have demonstrated that gluconeogenic intermediates act as critical substrates in the biosynthesis of EPS, thus playing a crucial role in biofilm development¹⁹. In S. cerevisiae, comparative RNA-seq analysis of biofilm versus planktonic cells revealed an upregulation of genes involved in gluconeogenesis during the initial attachment phase, followed by a metabolic switch to glycolysis as the biofilms mature. These observations elucidate a critical role of gluconeogenesis during the early phase of biofilm formation¹⁹. In S. aureus, itaconate inhibits glycolysis and redirects carbon flux into gluconeogenesis to support biofilm formation. This shift is marked by upregulation of pyruvate carboxylase (pca), which generates oxaloacetate for gluconeogenesis. Consequently, itaconate treatment enhances EPS production and enhances S. aureus biofilms²⁰. S. aureus, pca acts as a master switch that links the glycolysis and TCA cycle and regulates virulence and antimicrobial resistance²¹. In S. typhimurium, elevated c-di-GMP levels redirect carbon flux into gluconeogenesis, supplying substrates for polysaccharide synthesis during biofilm formation. High c-di-GMP concentrations induce upregulation of key gluconeogenic genes- fbp, ppsA, and pck. Thus, c-di-GMP not only suppresses motility to favor biofilm development but also reprograms carbon metabolism toward gluconeogenesis to enhance the biosynthesis of biofilm matrix²². In Candida albicans, mature biofilm formation involves dynamic switching between glycolysis and gluconeogenesis to generate hexose by-products supporting ECM production and bioenergetic requirements²³. Similarly, gluconeogenesis is essential for V. cholerae motility and biofilm development, supplying critical substrates required for in vivo growth²⁴. Overall, these studies delineate a crucial role of gluconeogenesis in different stages of biofilm formation. However, its role in mycobacterial biofilm formation remains largely unexplored.

We have recently demonstrated that polysaccharides are a significant component of the EPS of the mycobacterial biofilms⁴. Furthermore, we have shown that cellulose is a critical structural component of the EPS of mycobacterial biofilms⁴. It plays an essential role in maintaining biofilm architecture. In Mtb biofilms, cellulose forms fiber-like structures that act as a scaffold, connecting microcolonies and facilitating the attachment of bacterial cells to surfaces. Importantly, cellulase enzymes can effectively disrupt pre-formed biofilms and release bacterial cells⁴. In agreement with this, cellulase overexpression prevents spontaneous biofilm formation^5,25. Recent research has identified specific mycobacterial cellulases, including Rv0062 in Mtb and its ortholog MSMEG_6752 in M. smegmatis, which appear to regulate biofilm dispersal as part of the natural biofilm life cycle by degrading the cellulose matrix in response to metabolic signals such as decreased γ-aminobutyric acid (GABA) levels²⁵. The unique presence of cellulose in mycobacterial biofilms, absent in human tissues, has made it a valuable biomarker for detecting mycobacterial biofilm formation both in vitro and in clinical samples from infected patients⁵. Given that cellulose is a glucose biopolymer, we hypothesize that cellular glucose levels could play a vital role in Mtb biofilm formation. In this study, we examined whether the different carbon sources in the media modulate biofilm formation. We have also examined whether Pca (encoded by rv2967c in Mtb), a critical anaplerotic enzyme, plays a role in biofilm formation of Mtb.

Results

Impact of different carbon sources on biofilm formation of Mtb H37Rv

Different fermentable and non-fermentable carbon sources influence biofilm formation in microorganisms, including P. aeruginosa, M. avium, C. albicans, and C. glabrata, etc.^26–29. These carbon sources play an important role in the production of different polysaccharides present in the biofilm. Often, glucose is utilized by microorganisms as a carbon source to support different anabolic processes, such as protein and polysaccharide biosynthesis, thus facilitating the biofilm formation and maturation^30–32. To determine the appropriate concentration of different carbon sources that support mycobacterial growth, we analyzed Mtb growth in the presence of glycerol, glucose, pyruvate, acetate, succinate, malate, and oxaloacetate. In these experiments, we monitored the growth of Mtb in a standing culture, as it supports the pellicle biofilm formation. Pellicle biofilms are typically formed in Sauton’s media supplemented with glycerol³³. Hence, we used media with glycerol as the control for monitoring growth (Fig. S1). We evaluated the growth of Mtb on glucose ranging from 10 mM to 250 mM, pyruvate ranging from 2 mM to 150 mM, oxaloacetate ranging from 1 mM to 15 mM, succinic acid ranging from 0.5 mM to 25 mM, acetate ranging from 5 mM to 200 mM, and malic acid ranging from 0.5 mM to 10 mM (Fig. S1a–f). Glucose supported Mtb growth up to a concentration of 150 mM, after which it inhibited the mycobacterial growth (Fig. S1a). Pyruvate supported growth up to 20 mM (Fig. S1b), oxaloacetate supported growth of Mtb up to 10 mM (Fig. S1c), succinic acid supported mycobacterial growth up to 5 mM (Fig. S1d), acetic acid was tolerated up to 50 mM (Fig. S1e), and malic acid was well tolerated up to 6 mM (Fig. S1f).

After recording the growth of Mtb on different carbon sources, we next analyzed the capability of Mtb to form biofilms on different carbon sources. We first assessed pellicle biofilm formation across a broad concentration range of the different carbon sources (Fig. S2). Taking into account the concentrations that supported both growth and pellicle biofilm formation, we subsequently narrowed the range of each carbon source and re-evaluated the biofilm-forming capacity of Mtb. The selected concentration ranges were as follows: glucose (10–150 mM), pyruvate (2–70 mM), acetate (5–150 mM), succinic acid (0.5–5 mM), malic acid (0.5–8 mM), and oxaloacetate (1–8 mM) (Fig. 2a–m). These concentrations were derived from data described in Fig. S1 and Fig. S2 and are in good agreement with earlier published literature³⁴. We observed that Mtb forms biofilms with all the carbon sources used here, but mature and thick biofilms were observed when glycerol, glucose, pyruvate and acetate were used as the sole carbon sources (Fig. 2a–d). Even among these, the thickest biofilms, as determined by the crystal violet (CV) assay, were supported by 100 mM and 150 mM glucose and glycerol (Fig. 2b, e). However, the biofilms formed in the presence of succinate, malate, and oxaloacetate were very thin and fragile (Fig. 2h–m). We observed that, for most carbon sources, the concentrations that support planktonic growth also support biofilm formation, except for pyruvate and acetate, which support biofilm formation at higher concentrations. Specifically, pyruvate supports growth up to 20 mM and becomes toxic above this level, yet biofilms are supported by the pyruvate concentrations up to 70 mM. Likewise, acetate supports Mtb growth up to 50 mM, while biofilm formation was observed at concentrations up to 100 mM. These observations suggest that carbon source concentration required for biofilm formation are different than for growth and they depend on regulated metabolic flux, in which cellular homeostasis between polysaccharide biosynthesis and replenishment of TCA cycle intermediates is synchronized.

Fig. 2. Influence of carbon sources on biofilm formation in Mtb H37Rv.

Representative images of Mtb H37Rv pellicle biofilms in 24-well plates with glycerol in the first well taken as a positive control. a A representative image of biofilm in media containing glycerol. Other wells contain (b) differing concentrations of glucose ranging from 10 mM to 150 mM, (c) differing concentrations of pyruvate ranging from 2 mM to 70 mM, and (d) differing concentrations of acetic acid ranging from 5 mM to 150 mM. Graph showing CV assay of pellicle biofilm formed in 24-well plates with (e) glycerol and glucose as carbon sources, (f) glycerol and pyruvate as carbon sources, and (g) glycerol and acetic acid carbon sources. Representative image of Mtb H37Rv pellicle biofilm formation in 24-well plates with (h) differing concentrations of succinic acid ranging from 0.5 mM to 5 mM, (i) differing malic acid concentrations ranging from 0.5 mM to 8 mM, and (j) differing concentrations of oxaloacetic acid ranging from 0.5 mM to 8 mM. Graph showing CV assay of pellicle biofilm formed in 24-well plates with (k) glycerol and succinic acid as carbon sources, (l) glycerol and malic acid as carbon sources, and (m) glycerol and oxaloacetic acid as carbon sources. The experiments shown in (b–d, h–j) are performed in duplicates. Data represent mean ± SEM from three independent biological experiments. GraphPad Prism 8 was used to plot the column bar graphs. Statistical significance was determined using Student’s t test, * indicates a p value of <0.05, ** indicates a p value of <0.01, *** indicates a p value of <0.001.

pca::Tn is defective for pellicle biofilm formation in Sauton’s medium with glycerol as the sole source of carbon

The above observations suggest that replenishing TCA cycle intermediates to generate reductive energy through respiration and synthesizing glucose (to be used as a precursor for biosynthesis of EPS) through gluconeogenesis are important for biofilm formation. The carbon flux into glycolytic carbon metabolism, TCA cycle, and gluconeogenesis is regulated by the phosphoenolpyruvate(-pyruvate-oxaloacetate anaplerotic node¹⁴. This node consists of four enzymes: pyruvate carboxylase (Pca), Phosphoenol Pyruvate carboxylase (PCK), malic enzyme, and pyruvate phosphate dikinase. Out of these, only two, Pca and PCK, are known to be critical for gluconeogenesis^12,14. Gluconeogenesis is assumed to be important for biofilm formation, as it can direct carbon flux from various carbon sources into glucose synthesis, thereby facilitating the biosynthesis of polysaccharides^19,20,22,24. It is important to note that Pca catalyzes the first step in gluconeogenesis and also regulates the flow of carbon into the TCA cycle. Also, Pca represents an important component of the anaplerotic node that could replenish TCA intermediates. For these reasons, we examined the role of Pca in mycobacterial biofilm formation. Towards this, we utilized Tn mutant of pca in Mtb CDC1551 (pca::Tn) and the Mtb CDC1551 wild-type strain (Mtb CDC1551). We observed that the pca::Tn is attenuated for pellicle biofilm formation in Sauton’s medium containing glycerol as the sole source of carbon, suggesting a critical role of Pca in the mycobacterial biofilm formation (Fig. 3a). Given that we used glycerol as the carbon source in these experiments, these results suggest that Pca-mediated conversion of pyruvate into oxaloacetate is critical to diverting carbon into generation of precursors for EPS biosynthesis as well as maintaining TCA cycle. These results indicate that the Mtb CDC1551 strain could utilize glycerol as the sole source of carbon to manage the metabolic flux for the synthesis of TCA intermediates and polysaccharides needed for EPS synthesis, but pca::Tn is attenuated for pellicle biofilm formation in media with glycerol as the sole source of carbon.

Fig. 3. pca::Tn is defective in the making of pellicle, submerged, and macrocolony biofilms.

a Differences in pellicle biofilm formation in 24-well plates between Mtb CDC1551 and pca::Tn strains, where pca::Tn is defective in pellicle biofilm formation in the presence of glycerol. b Differences in colony morphology between Mtb CDC1551 and pca::Tn on 7H11OADC CRCBB agar plates. pca::Tn colony is defective in color uptake of Congo red (CR) and Coomassie brilliant blue (CBB) dyes, depicting the absence of polysaccharide and structural protein content in the colony. c Differences in the submerged biofilm formation between the pca::Tn and Mtb CDC1551 strain. Cultures of both strains at an OD₆₀₀ of 1.0-1.5 were exposed to 6 mM DTT induction for 29 h and then imaged.

pca::Tn is defective for macrocolony biofilms and submerged biofilm formation

Next, we examined the role of Pca-mediated gluconeogenesis in macrocolony biofilm formation. For this, we spotted pca::Tn and Mtb CDC1551 on 7H11 Oleic acid-albumin-dextrose-catalase (OADC) media containing Congo Red (CR) and Coomassie Brilliant Blue (CBB) dyes. In this method, CR binds with polysaccharides such as cellulose, while CBB binds to the proteinaceous component of the biofilm matrix. This method provides details on biofilm formation, maturation, and the composition of EPS simultaneously³⁵. We observed that the Mtb CDC1551 strain could pick up the CR and CBB dyes, demonstrating the presence of carbohydrates and proteins in the macrocolony. The colony of pca::Tn was defective in the uptake of the CR; however, it could take up some amounts of CBB. These observations indicate that the pca mutant is unable to form polysaccharides, although it could produce structural proteins in the macrocolony. These data also suggest poor macrocolony formation and maturation (Fig. 3b). These data indicate the important role of Pca in macrocolony biofilm formation.

We also evaluated whether pca::Tn could form the submerged biofilm upon exposure to reductive stress mediated by dithiothreitol (DTT) and compared it with the Mtb CDC1551 strain. This method was specifically used because in this method, biofilm formation does not depend on growth; rather, it depends on a shift in metabolic flux from growth to EPS synthesis. Towards this, these strains were cultured to an OD₆₀₀ of 1.0-1.5 and then exposed to 6 mM DTT for 29 h. as described previously⁴. We observed that the pca::Tn was attenuated to form TRS-induced submerged biofilms as compared to the Mtb CDC1551 strain (Fig. 3c). Overall, these results suggest that the absence of the Pca plays a very important role in diverting flux into EPS synthesis and maintaining the TCA for the generation of reductive energy, and its absence leads to defects in biofilm formation.

Pyruvate could alter the metabolic flux and restore biofilm formation and maturation in pca::Tn

The above-described experiments suggest that the pca gene is important for mycobacterial biofilm formation in the presence of glycerol and also plays a critical role in the structural architecture of macrocolonies. To ensure that the pca gene is indeed responsible for the above phenotypes, we complemented the mutant strain by reintroducing the pca gene into the genome under the control of the constitutive hsp60 promoter. The vector control strain was also constructed, where only the vector containing the promoter without the presence of any gene was inserted into the mutant strain, and further referred to as pca::Tn VC strain. To investigate the complementation of the pca gene in the pca::Tn comp strain, we checked mRNA levels of pca in the Mtb CDC1551, pca::Tn, pca::Tn VC, and pca::Tn comp strains. It was observed that the pca::Tn comp showed elevated mRNA expression similar to that of the Mtb CDC1551 strain. Meanwhile, the pca::Tn and pca::Tn VC showed lower pca mRNA levels. The difference in expression levels of pca::Tn between Mtb CDC1551 was significant. In contrast, the fold difference in mRNA level between Mtb CDC1551 and the pca::Tn comp strain remains nonsignificant (Fig. 4a). We also analyzed whether pca::Tn is attenuated for growth in the presence of glycerol. Towards this, we analyzed the growth of pca::Tn, and the pca::comp strain, for growth in media containing glycerol as the sole source of carbon. Interestingly, all three strains could grow in the presence of glycerol as the sole source of carbon (Figure S3). We also evaluated the growth of pca::Tn and pca::Tn comp strains on different carbon sources and found that both the strains show similar growth patterns as Mtb CDC1551 when grown in glucose and pyruvate as sole carbon source (Figs. S1 and S4). These data suggest that the biofilm defects/regain phenotype observed were growth independent phenomenon.

Fig. 4. Addition of pyruvate restores the formation of thick, voluminous pellicle biofilm structures in the pca::Tn strain and the colony morphology phenotype.

a Bar graph showing the relative mRNA levels between the four strains. The pca::Tn comp strain showed mRNA expression similar to the Mtb CDC1551 strain. Meanwhile, the pca::Tn and pca::Tn VC strains showed lower mRNA levels. b–d The pellicle biofilm was formed in 8-well-chambered glass slides with Mtb stained with Auramine O and IMT-CBD-mC to stain the presence of cellulose. The biofilms were then visualized with 488 nm and 561 nm lasers using CLSM. Representative 3D images and side-view images of pellicle biofilm of all four strains. b Under glycerol-fed conditions, (c) under 50 mM pyruvate-fed conditions, (d) under 70 mM pyruvate-fed conditions. e Biomass (in µm3/µm2), (f) thickness (in µm), and (g) relative intensity analysis of the biofilm structures. h Differences in colony morphology between Mtb CDC1551, pca::Tn, pca::Tn VC, and pca::Tn comp strains on 7H11 OADC CRCBB agar (i) containing glycerol as a carbon source (ii) containing 50 mM pyruvate as a carbon source (iii) containing 70 mM pyruvate as a carbon source. Images are representative of three independent experiments. Biomass and thickness analysis were performed through the COMSTAT plugin in ImageJ software, and relative intensity was calculated through IMARIS version 8.9 software. Data represent mean ± SEM from three independent experiments (e–g). GraphPad Prism 8 was used to plot the column bar graphs. Statistical significance was determined using Student’s t test, * indicates a p value of <0.05, ** indicates a p value of <0.01, *** indicates a p value of <0.001, **** indicates a p value of <0.0001 (e–g).

Data presented above suggested that the pca::Tn mutant is capable of growth in the presence of glycerol in standing medium but is unable to divert the flux towards gluconeogenesis and maintain the replenishment of TCA intermediate during pellicle biofilm formation. Glycerol is known to lead to the accumulation of triose phosphates that may result in the production of methylglyoxal, a dicarbonyl compound. Accumulation of methylglyoxal is known to be toxic to Mtb^36–38. We reasoned that the pca::Tn mutant is unable to form a biofilm in the presence of glycerol, since, in the absence of this critical anaplerotic reaction, triose phosphates may accumulate, leading to the production of methylglyoxal. The other possibility is that the lack of Pca leads to the accumulation of pyruvate, which prevents the simultaneous diversion of carbon flux into EPS synthesis and the TCA cycle.

To test this, we examined whether pyruvate could support biofilm formation in Mtb CDC1551 and pca::Tn mutant. We used different concentrations of pyruvate (the sole carbon source) in Sauton’s media. Sauton’s media containing glycerol served as a positive control. We imaged the pellicle biofilm formed by the above-mentioned strains via confocal laser scanning microscopy (CLSM) in the presence of glycerol and pyruvate. To characterize the biofilms in terms of cellulose content, biomass, and biofilm thickness, we have stained the mycobacterial cells using Auramine O dye and cellulose with an IMT-CBD-mC probe. In this probe, the cellulose-binding domain (CBD) of CenA from Cellulomonas fimi³⁹ is fused to the fluorescent protein mCherry through an octa-glycine linker. Since the CBD has a high affinity for cellulose, the IMT-CBD-mC probe binds cellulose with high affinity⁴⁰. Therefore, the intensity of mCherry directly correlates with the presence of cellulose^5,40. Here, we observed that the presence of glycerol leads to the formation of thicker and mature biofilms in the Mtb CDC1551 strain. In contrast, glycerol does not support biofilm formation in the pca::Tn and pca::Tn VC strains, validating our earlier observations. We observed that the biofilm biomass was lower than in Mtb CDC1551 and the pca::Tn comp strain when glycerol was used as a carbon source (Fig. 4b). However, in the presence of pyruvate, all strains formed thicker, fully mature biofilms. The biomass, biofilm thickness, and cellulose content significantly increased substantially upon addition of pyruvate as a carbon source in the pca::Tn and pca::Tn VC strains. We also observed that IMT-CBD-mC staining was diminished in the mutant strain in the case of the control panel, whereas increased staining was visualized in the case of 50 mM and 70 mM pyruvate panels for all four strains (Fig. 4c, d). Representative images of the side view of the biofilm panels are shown in Fig. 4c, d demonstrating the biofilm thickness of different strains. Biomass and thickness analysis were performed for the panels described above, and a statistical difference was observed between Mtb CDC1551 and pca::Tn, pca::Tn VC strains, and between the pca::Tn glycerol and pca::Tn pyruvate-containing media (Fig. 4e, f), highlighting the role of pyruvate in biofilm formation. Relative intensity was calculated between mCherry/Auramine O fluorescence to quantify cellulose content of the biofilms (Fig. 4g). 50 mM and 70 mM pyruvate were capable of supporting biofilm formation even in the pca::Tn and pca::Tn VC strains.

After observing the biofilm formation defect in the pca::Tn mutant strain, we sought to check whether the biofilm defect could be restored in the 7H11OADC CRCBB media plates. We took Mtb CDC1551, pca::Tn, pca::Tn VC, and pca::Tn comp strains for this experiment. We performed spot assay on 7H11OADC CRCBB containing 50 mM and 70 mM pyruvate as the sole carbon source and plate containing glycerol as a control (Fig. 4h). We observed that pca::Tn, which is otherwise defective in color uptake, takes up the CR stain indicating that pyruvate helps in polysaccharide biosynthesis even in the absence of Pca (Fig. 4h). These results suggested that availability of pyruvate helps in directing carbon flux towards polysaccharides biosynthesis and thus enhances biofilm formation even in the pca mutant Mtb strain. An excess of pyruvate can be utilized through flexible metabolic pathways to divert carbon flux into the synthesis of precursors for EPS biosynthesis and TCA cycle intermediates.

Pca mutant’s defect in pellicle biofilm formation is ameliorated by glucose

Given that we have earlier demonstrated that replacement of glycerol with pyruvate leads to the formation of thick and mature pellicle biofilms in Mtb H37Rv. It was reasonable to assume that Pca plays an important role in the modulation of carbon flux into gluconeogenesis, and the TCA cycle is critical for biofilm formation in the presence of glycerol. Gluconeogenesis provides a glucose precursor for EPS biosynthesis, while the TCA cycle could provide reductive energy for EPS biosynthesis. Next, we analyzed whether the pca::Tn could form biofilms in the presence of glucose. We observed that the pca::Tn strain formed mature biofilms in medium wherein glycerol was replaced by glucose as the sole carbon source (Fig. 5a–c). In these experiments, we have used 100 mM and 150 mM of glucose, as these concentrations facilitated biofilm formation in Mtb H37Rv. These biofilms of pca::Tn and pca::Tn VC have a cording pattern similar to the Mtb CDC1551 strain (Fig. 5a–c). These data suggest that the carbon flux towards gluconeogenesis and the TCA cycle plays a critical role in biofilm formation. If glucose is directly provided, the pca mutant has enough glucose to support polysaccharide biosynthesis and to maintain TCA cycle intermediates, thereby supporting biofilm formation.

Fig. 5. Addition of glucose or pca gene complementation restores the formation of thick, voluminous pellicle biofilm structures in the pca::Tn strain.

a Representative image of pellicle biofilm formation in 24-well plates of Mtb CDC1551, pca::Tn, pca::Tn VC, and pca::Tn comp strains with glycerol and glucose as carbon sources. b Representative image of CV assay performed on pellicle biofilm setup. c Graph showing CV assay of pellicle biofilm of Mtb CDC1551, pca::Tn, pca::Tn VC, and pca::Tn comp strains with glycerol and glucose as carbon sources. The pellicle biofilm was formed in 8-well-chambered glass slides with Mtb stained with Auramine O and IMT-CBD-mC to stain the presence of cellulose. The biofilms were then visualized with 488 nm and 561 nm lasers using CLSM. Representative 3D and side-view images of pellicle biofilm of all four strains i.e., Mtb CDC1551, pca::Tn, pca::Tn VC, and pca::Tn comp (d) under glycerol-fed conditions. e under 150 mM glucose-fed conditions. f Biomass (in µm³/µm²), (g) thickness (in µm), and (h) relative intensity analysis of the biofilm structures. Biomass and thickness analysis were performed through the COMSTAT plugin in ImageJ software, and relative intensity was calculated through IMARIS version 8.9 software. Data represent mean ± SEM from three independent experiments (c, f–h). GraphPad Prism 8 was used to plot the column bar graphs. Statistical significance was determined using Student’s t test, * indicates a p value of <0.05, ** indicates a p value of <0.005, *** indicates a p value of <0.0005, **** indicates a p value of <0.0001 (f–h). Statistical significance was determined using Two-way ANOVA, * indicates a p value of <0.05, ** indicates a p value of <0.01, *** indicates a p value of <0.001, **** indicates a p value of <0.0001 (c).

In addition to quantification via CV assay, we also imaged the pellicle biofilm formed by the above-mentioned strains via CLSM. Pellicle biofilms were formed in 8-well chambered glass slides. Mtb cells were visualized in the biofilms using Auramine O dye, and the cellulose was imaged using IMT-CBD-mC stain, which specifically stains the cellulose⁴⁰. In agreement with the above observations, we found that the Mtb CDC1551 strain formed thicker, more voluminous biofilms in the presence of glycerol. In contrast, the pca::Tn mutant and pca::Tn VC strain biofilms possessed less biomass than the Mtb CDC1551 and pca::Tn comp strains (Fig. 5d, e), when glycerol is available as the sole carbon source. Three-dimensional images of the biofilms suggested that the pca::Tn mutant and pca::Tn VC biofilms were more scattered, while the Mtb CDC1551 and the pca::Tn comp strains formed more voluminous biofilms with higher biomass content. Pellicle biofilms have been previously reported to have high polysaccharide content in their biofilms^4,5,40. Biomass and thickness analysis were performed for the panels described above, and a statistical difference was observed between Mtb CDC1551 and pca::Tn, pca::Tn VC strain, wherein glycerol acted as a carbon source. The biomass content and thickness of the pca::Tn mutant and pca::Tn VC were also significantly less than those of the Mtb CDC1551 and the pca::Tn comp strain (Fig. 5f, g). This defect of the pca::Tn mutant and pca::Tn VC strains in the biofilm architecture, biomass, and thickness recuperates upon using glucose as the sole carbon source (Fig. 5f, g), and the pca::Tn and pca::Tn VC could form mature and differentiated biofilms (Fig. 5f, g). We also analyzed the presence of cellulose as a measure of polysaccharide biosynthesis, a key polysaccharide, through calculation of mCherry/Auramine O fluorescence in media containing glycerol or glucose as a carbon source (Fig. 5h). We observed that providing glucose reinstates polysaccharide biosynthesis during maturation of biofilms in the pca mutant strain.

Glucose leads to the restoration of DTT-induced submerged biofilm formation in the pca::Tn strain

Our group has recently demonstrated that Mtb, M. smegmatis, and other mycobacterial species can form biofilm in response to thiol reductive stress (TRS)^4,40. This model differs from pellicle biofilm formation, which depends on bacterial growth. In this method, the mycobacterial cultures that have achieved an OD₆₀₀ of 0.8–1.2 are exposed to TRS mediated by DTT. TRS leads to the switching of metabolism from cellular growth to the production of polysaccharides and structural proteins⁴. Given that this process involves metabolic switching and Pca could be a critical enzyme in this change of carbon flux, we examined whether the pca::Tn strain is attenuated for forming biofilms in response to TRS induced by DTT. Towards this, logarithmic cultures of Mtb CDC1551, pca::Tn, pca::Tn VC, and pca::Tn comp were grown to an OD₆₀₀ of 1.2 and then exposed to 6 mM DTT for 29 h. We observed that pca::Tn and pca::Tn VC lacked a rigid biofilm structure. At the same time, the pca:: Tn comp formed a thick biofilm similar to the Mtb CDC1551 (Fig. 6a, b). We have earlier observed that the biofilm-forming capability of the pca::Tn could be restored upon the addition of glucose in the 7H9 + 5% OADC medium. We observed that the addition of glucose restores biofilm formation in the pca::Tn strain to comparable levels as that of Mtb CDC1551, as well as the pca::Tn comp strain. This phenomenon was quantified using the CV assay. We observed a significant difference between the Mtb CDC1551 and pca::Tn and pca::Tn VC strain when glycerol was used as a carbon source (Fig. 6b). These differences became nonsignificant upon replacing glycerol with glucose. These results clearly suggest that during biofilm formation, a large amount of glucose (or glucose-derived monosaccharides) is required for polysaccharide biosynthesis and for replenishing the TCA cycle intermediates. The enzyme Pca plays a critical role in this process and, hence, is unable to form biofilms when glycerol is used as a carbon source.

Fig. 6. The addition of glucose reinstates the formation of thick, voluminous cellulose-rich TRS-induced submerged biofilm structures in the pca::Tn strain.

a Representative image of TRS-induced submerged biofilm formation in 24-well plates of Mtb CDC1551, pca::Tn, pca::Tn VC, and pca::Tn comp strains; 6 mM DTT for 29 h was given to induce biofilm formation. Glycerol and glucose were utilized as carbon sources. b Graph showing CV assay of submerged biofilm of Mtb CDC1551, pca::Tn, pca::Tn VC, and pca::Tn comp strains with glycerol and glucose as carbon sources. Data represent mean ± SEM from three biological replicates, each with two technical repeats. The submerged biofilm was formed in 8-well-chambered glass slides with Mtb stained with Auramine O and IMT-CBD-mC to detect cellulose. The biofilms were then visualized with 488 nm and 561 nm lasers using CLSM. Representative 3D and side-view images of submerged biofilm of all four strains, i.e., Mtb CDC1551, pca::Tn, pca::Tn VC, and pca::Tn comp, (c) under glycerol-fed conditions. d Under 150 mM glucose-fed conditions. e Biomass (in µm³/µm²), (f) thickness (in µm), and (g) relative intensity analyses of the biofilm structures. Biomass and thickness analyses were performed through the COMSTAT plugin in ImageJ software, and relative intensity was calculated through IMARIS version 8.9 software. Data represent mean ± SEM from three independent experiments (b, e–g). GraphPad Prism 8 was used to plot the column bar graphs. Statistical significance was determined using Two-way ANOVA, * indicates a p value of <0.05, ** indicates a p value of <0.01, *** indicates a p value of <0.001, **** indicates a p value of <0.0001 (b). Statistical significance was determined using Student’s t test, * indicates a p value of <0.05, ** indicates a p value of <0.01, *** indicates a p value of <0.001 (e–g).

In addition to quantification of the biofilm formation via CV assay, we also performed CLSM analysis of these biofilms. DTT-induced submerged biofilms were formed in 8-well chambered glass slides. Mycobacterial cells were stained using Auramine O dye, and cellulose was visualized using IMT-CBD-mC. Similar to the observations described above, we observed that in the control panel with glycerol as the carbon source, the Mtb CDC1551 strain formed thicker and fully matured biofilms. In contrast, the pca::Tn mutant and pca::Tn VC strain formed thin biofilms (with a smaller Z-stack) having large spaces between mycobacterial cells (Fig. 6c, d). The addition of glucose to the medium could restore this biofilm-forming capability of the pca::Tn mutant and pca::Tn VC strain. We also observed that IMT-CBD-mC staining was absent from the mutant strain when glycerol was used as a carbon source. In contrast, increased staining was visualized in all the strains, including the pca::Tn mutant and pca::Tn VC strain, upon addition of glucose (Fig. 6c, d). Representative images of the side view of the biofilm panels are shown in (Fig. 6c, d), which depict the biofilm thickness of different strains. Biomass and thickness analysis were performed for the panels described above, and the statistical difference was observed between Mtb CDC1551 and pca::Tn, pca::Tn VC strains, and between the pca::Tn glycerol and pca::Tn glucose panels, suggesting that the presence of glucose could compensate for the absence of pca (Fig. 6e, f). These observations point towards the importance of diverting carbon flux towards gluconeogenesis and the TCA cycle in biofilm formation. To ascertain the cellulose content of biofilms formed by different strains in the presence or absence of glycerol and glucose, the relative intensity was calculated as the ratio of mCherry/Auramine O fluorescence. This analysis indicated that the presence of glucose compensates for the absence of the pca gene (Fig. 6g). Similar results were obtained when DTT-induced submerged biofilms of Mtb CDC1551 and pca::Tn, pca::Tn VC and pca::Tn comp were provided with pyruvate as a carbon source (Fig. S5). These observations suggest that the diversion of carbon source from DNA synthesis to the synthesis of polysaccharides and structural proteins of EPS, and the critical role played by Pca in this diversion.

Addition of glucose leads to restoration of the defect in macrocolony phenotype in the pca::Tn strain

After observing the defect in biofilm formation in the pca::Tn mutant strain, we sought to analyze whether the addition of glucose to the solid media plates could restore the macrocolony biofilm formation capability of the pca::Tn mutant strain. For this, we analyzed the macrocolony biofilm development on 7H11OADC CR-CBB plates. We observed that pca::Tn, which is otherwise defective in color uptake, takes up the CR stain upon using glucose as the sole carbon source (Fig. 7a, b). These results suggested that the availability of glucose molecules plays a pivotal role in the biofilm formation and the architecture of the macrocolony of mycobacterial strains.

Fig. 7. The addition of glucose or pca gene complementation restores the colony morphology phenotype in the mutant strain.

Differences in colony morphology between Mtb CDC1551, pca::Tn, pca::Tn VC, and pca::Tn comp strains on 7H11OADC CRCBB plates, (a) containing glycerol as a carbon source, (b) containing 150 mM glucose as a carbon source. Images are representative of three independent experiments. c Gluconeogenesis plays a critical role in the formation of mycobacterial TB biofilm. Mtb H37Rv or Mtb CDC1551 strains, under normal conditions, can feed the carbon flux through the gluconeogenesis pathway, thereby generating glucose molecules. In wild type, glycerol leads to the production of pyruvate, which is pushed to gluconeogenesis and the TCA cycle, and it helps in biofilm formation. The generation of glucose molecules is required to produce cellulose and other polysaccharides whose primary subunit is glucose. These polysaccharides are required to integrate into the pellicle, submerged, and macrocolony biofilm models. The integration of cellulose contributes to the structural backbone of these biofilm models. A possible hypothesis for this defect can be that in the case of pca:: Tn mutant strain, the flux is stopped at pyruvate, which leads to the accumulation of triose phosphates, which are converted into methylglyoxal (toxic to bacteria). As methylglyoxal is known to be toxic to Mtb, it can in turn inhibit the growth and biofilm formation in the case of the pca:: Tn mutant strain. This obstruction at the entry point of the gluconeogenesis pathway can result in low/no glucose generation. This results in low polysaccharide production, thereby defective pellicle, submerged, and macrocolony biofilm models. This hypothesis ensures the essentiality of the gluconeogenesis pathway in mycobacterial biofilm formation. (Created in BioRender. Kumar, A. (2025) https://BioRender.com/d39e222).

Discussion

The shift down in central metabolic pathways during the mycobacterial biofilm formation has not been previously explored. This study analyzed the impact of different carbon sources on mycobacterial biofilm formation. We observed that fermentable carbon sources glucose, glycerol, and pyruvate support thick and mature biofilm formation, while the TCA intermediates do not support the formation of mature and thick biofilms. These results also point towards the capability of Mtb to synchronously direct the metabolic flux towards gluconeogenesis to fulfill the requirement of glucose and related monosaccharides for the synthesis of EPS polysaccharides and towards the TCA cycle to enable the generation of reductive energy in the form of NADH to support the biosynthesis of EPS. A pca::Tn mutant strain was utilized to examine its role in biofilm formation. This enzyme was selected because the conversion of pyruvate to oxaloacetate is a key step in gluconeogenesis and also helps replenish TCA cycle intermediates. The pca::Tn mutant was defective in pellicle biofilm formation when glycerol was used as the sole carbon source, but upon glucose supplementation or genetic complementation with the pca gene, biofilm formation was successfully restored, demonstrating its critical role in metabolic switching and rerouting the carbon flux towards gluconeogenesis and the TCA cycle. The study expanded its scope beyond traditional pellicle biofilms to investigate the impact of Pca on TRS-induced submerged biofilms and in the macrocolony biofilms. It was observed that the pca::Tn mutant is attenuated for submerged biofilm formation and possesses structural defects. This attenuation in biofilm formation was remedied by supplementing the medium with glucose and/or pyruvate or by complementation with an episomal copy of the pca gene. Using CLSM imaging and cellulose detection with the IMT-CBD-mC probe, we confirmed the presence of cellulose in WT Mtb CDC1551 and complemented strains. The pca::Tn strain could not accumulate significant levels of cellulose; however, this defect was corrected upon addition of external pyruvate or glucose. These observations suggest that the pca gene plays a crucial role in biofilm formation by modulating polysaccharide content in both biofilm models. These findings align with previous reports in other microbes like P. aeruginosa, where pca is important for pathogenicity⁴¹ and biofilm formation, and S. aureus, where the absence of the related pca gene decreased biofilm formation but was restored through genetic complementation and shown to be essential for pathogenesis and intracellular survival⁴².

One of the major findings of this study is that the carbon sources influence the biofilm formation in Mtb. The role of carbon sources on the biofilm formation and architecture is well known but has not been analyzed in Mtb. This study revealed that carbon sources in the media influence the thickness and maturation of Mtb biofilms, with fermentable sources such as glucose, glycerol, and pyruvate supporting robust biofilm formation, whereas TCA cycle intermediates, such as succinate, malate, and oxaloacetate, result in thin, fragile biofilms. These observations highlight the importance of fermentable carbon substrates in promoting polysaccharide biosynthesis, which is crucial for biofilm architecture. They are consistent with the requirement of robust anabolic processes in biofilm formation, including the biosynthesis of extracellular polysaccharide matrices and the energy requirement for their synthesis. These findings also point towards synchronously directing the carbon flux into the gluconeogenic pathway and the TCA cycle. Gluconeogenic enzymes support the requirement of monosaccharide subunits for the biosynthesis of polysaccharides, a major component of the EPS of Mtb biofilms. While an active TCA cycle provides NADH required for the EPS synthesis. TCA cycle intermediates could also channel carbon flux into amino acid biosynthesis and general cell maintenance. Importantly, these observations suggest that successful biofilm formation in Mtb is tightly linked to carbon flux homeostasis between ECM biosynthesis and the replenishment of TCA intermediates. Gluconeogenesis plays a critical role in this process, as the pca mutant was defective in forming mature biofilms. Interestingly, enzymes involved in gluconeogenesis play a critical role in Mtb’s survival in mice^12,18 Recently, biofilms were shown to be important for establishing a productive infection⁵. These observations suggest that gluconeogenesis plays a dual role in mycobacterial survival within a host by enabling biofilm formation and regulating metabolic flux for intracellular macrophage survival.

One of the key findings of this study was that when glycerol is used as the sole source of carbon, the Pca mutant could grow similarly to the WT strain in the standing culture model. However, it is defective in biofilm formation under the same conditions. These findings are important, given that the same carbon source supports planktonic growth but requires Pca to maintain flux for biofilm formation. It is plausible that when glycerol is used as the sole carbon source to support biofilm formation, it may lead to excessive production of triose phosphates, which may result in the production of methylglyoxal^36–38. Methylglyoxal, a dicarbonyl aldehyde compound, is inhibitory to mycobacterial growth and is also produced by macrophages to control the growth of intracellular Mtb³⁸. It is tempting to speculate that the presence of Pca may ameliorate this effect of glycerol metabolism. Interestingly, when pyruvate is directly provided as a carbon source, it supports biofilm formation even in the absence of Pca. However, low concentrations of pyruvate are not effective in supporting biofilm formation, suggesting that excessive pyruvate perhaps leads to replenishment of oxaloacetate levels by entering the TCA cycle via pyruvate dehydrogenase, and then could utilize the reverse methyl citrate cycle and the glyoxylate shunt for production of oxaloacetate and it utilization in gluconeogenesis⁴³. In fact, pyruvate is preferred by Mtb compared to glucose and fatty acids⁴³. Observations from this study suggest that the formation of thick and mature biofilm in the presence of pyruvate further strengthen this notion.

In addition to the classical pellicle biofilm formation, we have also checked the role of pca in TRS-induced submerged biofilm formation. This biofilm model was discovered by Trivedi and coworkers. The EPS of these biofilms is rich in polysaccharides such as cellulose and other components, i.e., proteins, extracellular DNA, etc⁴. pca::Tn mutant also showed a defective submerged biofilm formation with gaps and spaces present in between the biofilm structure. This phenotype was complemented upon the superficial addition of different glucose concentrations in the medium or the pca::Tn comp (pca gene complemented) strain. We also visualized the pellicle and submerged biofilm via CLSM imaging and checked for the presence of cellulose in the biofilms using an IMT-CBD-mC probe as described earlier^5,40. The presence of cellulose in mycobacterial biofilms was also observed ubiquitously across all biofilm models of Mtb⁴⁴, non-infectious M. smegmatis⁴⁴, and non-tuberculous mycobacteria⁵, suggesting its integral role in biofilm formation. The presence of cellulose was observed via IMT-CBD-mC staining in Mtb CDC1551 and pca::Tn comp strains and was enhanced upon increasing the concentration of glucose and pyruvate. These results suggest that pca is important in biofilm formation and modulates the polysaccharide content in pellicle and submerged biofilm models. Previous reports in microbes such as P. aeruginosa and S. aureus have shown the importance of the pca gene in biofilm formation. In P. aeruginosa, in addition to being involved in several metabolic pathways, this gene is shown to be important for pathogenicity and biofilm formation during in vitro growth⁴¹. In S. aureus, the absence of the pca gene led to decreased biofilm formation, which was restored upon genetic complementation. It has also been shown to be involved in pathogenesis and required for intracellular survival⁴².

Through these results, it can be concluded that pca is important for biofilm formation. We hypothesize that the pca feeds the carbon flux simultaneously into the gluconeogenesis pathway and TCA cycle. Gluconeogenesis plays an important role in production of glucose molecules. These glucose molecules are then utilized to generate energy or as a substrate for EPS production under biofilm-forming conditions. Recent research on S. cerevisiae, S. aureus, S. enterica serovar Typhimurium, C. albicans, V. cholerae, and others has demonstrated that gluconeogenesis intermediates are necessary for EPS synthesis and biofilm development^{19,20,22–24}. Since the biofilm models represented in this study show the presence of cellulose, confirmed via the IMT-CBD-mC probe staining, it can be suggested that the generation of cellulose is achieved through gluconeogenesis under biofilm development conditions. These observations suggest a critical role of diverting carbon flux into gluconeogenesis and TCA in mycobacterial biofilm development, as the non-functionality of this pathway leads to defective biofilm formation. The model depicts the role of gluconeogenesis in mycobacterial biofilm development, illustrated in Fig. 7c.

Methods

Bacterial medium and culturing conditions

Mtb H37Rv (ATCC 27294), Mtb CDC1551 (Bei Resources, NR-13649), Mtb CDC1551 Tn Mutant 2242 (MT3045, rv2967c) (Bei Resources, NR-15722) were grown in complete 7H9 broth [7H9 (Becton Dickinson Difco, 271310) supplemented with 0.2% glycerol (Rankem, G0040), 0.1% Tween 80 (MP, 103170) and 10% (OADC)]. The vector control strain pstHIT vector in Mtb CDC1551 Tn Mutant 2242, i.e., pca::Tn VC, and complementing strain containing pstHIT+pca gene i.e., pca::Tn comp were grown in Middlebrook 7H9 broth or solid Middlebrook 7H11 OADC agar along with additional hygromycin (50 μg/ml). The cultures were incubated at 37°C, 90 rpm in 30 mL or 60 mL square polyethylene terephthalate glycol media bottles until the desired OD₆₀₀ was reached. 7H11 agar (BD Difco, 212203), along with 0.5% glycerol (Rankem, G0040) and 10% OADC, was used as solid media.

For growth curve experiments, exponentially growing cultures of Mtb were washed and inoculated in 7H9 medium containing all the supplements to achieve an initial OD₆₀₀ of 0.1. Their growth rate was recorded for 22 days until they obtained the stationary phase of growth. The other carbon sources utilized in this study were replaced with glycerol, with glucose conc. ranging (10 mM- 250 mM). pyruvate conc. ranging (2 mM–150 mM), acetate conc. ranging (5 mM–200 mM), succinate conc. ranging (0.5 mM–25 mM), malate conc. Ranging (0.5 mM–10 mM), oxaloacetate conc. ranging (1 mM–10 mM). All the experiments utilizing Mtb were carried out inside the institute’s BSL-3 facility in accordance with the standard operating procedures.

Construction of pca complemented strain

Cloning was carried out using KOD polymerase (Toyobo, KOD-201) and restriction digestion with Fast Digest range of enzymes from Thermo Scientific (EcoRV; FD0304 and HindIII; FD0504). The gene rv2967c was amplified using primers listed in Table 1. and was cloned into a mycobacterial integration vector, pstHIT, with a hygromycin gene cassette used as a selection marker. The N-terminal His-tag was already present in the vector map. Finally, it was transformed in Mtb CDC1551 Tn Mutant 2242 (MT3045, rv2967c) to generate the complementing strain, and the vector alone without the gene sequence was also transformed to create a vector control strain. For the qRT-PCR experiment, RNA was isolated using a Direct-Zol RNA isolation kit (ZYMO). The primers used for RT-qPCR to detect mRNA levels are listed in Table 2. cDNA synthesis and PCR reactions were carried out using total RNA extracted from each bacterial culture and Superscript III platinum-SYBR green one-step qRT-PCR kit (Invitrogen) with appropriate primer pairs (2 µM) using an ABI real-time PCR detection system. Endogenously expressed Mtb rpoB was used as an internal control. The ∆∆CT method calculated the fold difference in gene expression⁴⁵.

Table 1.

List of primers used for creating a complementing strain of the rv2967c (pca) gene

S. no.	Primer name	Sequence	Restriction site
1	rv2967c FP	GCATCTAGAATGGTGTTTTCCAAGGTGCTCGTCGCCAATC	EcoRV
2	rv2967c RP	CGCGCTTAATTAATCAGCTCACCACCACCAACAG	HindIII

Table 2.

List of oligonucleotide primer sequences used in RT-qPCR to detect mRNA levels in the mutant and complemented strains

S. No.	Primer name	Sequence
1	rv2967c forward primer-	CTACCCGAGTACGCACCAG
2	rv2967c reverse primer-	CAACACTCCACGGACAACAG

Pellicle biofilm assays and colony morphology

For pellicle biofilm, Sauton’s media is typically used, which contains glycerol as the primary carbon source. For other carbon sources, glycerol was replaced with different concentrations of carbon sources as depicted in the figure. For pellicle biofilm formation, saturated cultures of Mtb CDC WT, pca::Tn, pca:: Tn VC, and pca::Tn comp strains were inoculated (10% inoculum) in 2 ml of Sauton’s media in 24-well plates. Pellicle biofilms were formed in 8-well-chambered glass slides for CLSM. A 10% inoculum was added in 500 µl of Sauton’s media in 8-well chambered glass slides and incubated for approximately 3–4 weeks at 37°C in standing conditions. To form submerged biofilms, logarithmic-phase cultures of Mtb CDC WT, pca::Tn, pca::Tn VC, and pca::Tn comp strains (grown in 7H9 media + 5% OADC) were induced with 6 mM DTT for 29 h at 37°C in 24-well plates. Submerged biofilms were also formed in 8-well-chambered glass slides for CLSM. For inoculation in 8-well chambered glass slides, cultures of OD₆₀₀ 0.8–1.5 were taken, 500 µL was transferred to each well, and was induced with 6 mM DTT.

Crystal violet (CV) assay of biofilms

The CV assay of Mtb biofilms was performed as described earlier⁴. Briefly, the CV assay was performed in 24-well plates. After forming the Mtb pellicle or submerged biofilm (in 24-well plates), the media was removed and washed with 1X phosphate buffer saline (PBS), and 2 ml of 1% CV (Sigma) was gently added to the biofilm. It was incubated for 20 min at 37°C. The stain was removed, and the biofilm was gently washed twice with 1XPBS. The bound CV was then extracted by a 10-minute incubation at 37°C with 2 ml of 95% ethanol. The absorbance of extracted CV was measured at 595 nm on a spectrophotometer. The CV assay was performed in three independent replicates for each dataset.

Preparation of plates containing CR and CBB dye, containing glucose and pyruvate

For the construction of 7H11OADC, media plates containing glycerol were prepared as described previously⁵. Glycerol was eliminated from the 7H11 OADC CRCBB media plates and was instead replenished with glucose and pyruvate. For glucose 100 mM and 150 mM conc. of dextrose (Merck) was added as per the required media volume. Similarly, for pyruvate, 50 mM and 70 mM conc. of sodium pyruvate (Sigma) was added as per the required media volume. The media was also supplemented with CR (40 μg/ml) and CBB (20 μg/ml) along with the respective carbon sources.

For spot assay, 2 μl of each logarithmic-phase culture with similar OD₆₀₀ were spotted onto solid Middlebrook 7H11 OADC CRCBB agar containing either glycerol, glucose, or pyruvate as a carbon source. The plates were incubated at 37°C for approximately four weeks.

Confocal laser scanning microscopy of pellicle and submerged biofilms

Mature pellicle and submerged Mtb biofilms were produced on chambered glass slides for CLSM imaging. After the maturation of biofilms, the sample was fixed with 4% paraformaldehyde for 1 hour and washed thrice with 1XPBS. After washing, it was stained with fluorescent probes such as Auramine O (for Mtb staining) and IMT-CBD-mC (for cellulose) (10 µM in TRIS buffer (50 mM, pH 7.5) with 1 M NaCl for 1–1.5 h). After staining, samples were washed thrice with PBS. Stained biofilms were viewed using a Nikon A1R confocal microscope.

Statistical analysis

Results are expressed as mean ± SEM from at least three biological replicates. Statistical significance was determined using GraphPad Prism 8 employing the Student’s t test (unpaired, non-parametric Gaussian distribution, Mann-Whitney test) and Two-way ANOVA. * depicts p value < 0.05, ** depicts p value < 0.01, *** depicts p value < 0.001, and **** p value < 0.0001.

Author contributions

S. Singh and S. Sharma performed the experiments. S. Singh performed the analysis of the experiments. S Singh wrote the original draft, and A.K. and S. Sharma reviewed and edited the manuscript. A.K. conceived and coordinated the study.

Data availability

All the data pertaining to the results reported in this study are part of the main text.

Competing interests

The authors declare no competing interests.

Supplementary information

The online version contains supplementary material available at https://doi.org/10.1038/s41522-026-01044-1.