Mycobacterial Lipids Orchestrate Infection Stage-Specific Remodeling of the Host Actin Cytoskeleton

Abstract

The actin cytoskeleton is a central target of host–pathogen interactions, yet how distinct microbial factors engage specific actin regulatory pathways remains poorly understood. Here, we show that mycobacterial envelope lipids directly modulate host actin dynamics through distinct molecular mechanisms. Using quantitative imaging and lipidomic analysis, we demonstrate that sulfoglycolipids (SL-1) and phenolic glycolipid-1 (PGL-1) promote actin depolymerization via cofilin-mediated severing, whereas phthiocerol dimycocerosate (PDIM) and trehalose dimycolate (TDM) induce actin polymerization through activation of the WASP–Arp2/3 pathway. These lipid-specific effects correspond to changes in lipid abundance across infection stages, linking cytoskeletal remodeling to temporal regulation of infection. Functionally, actin depolymerization enhances bacterial uptake or release depending on lipid context, while polymerization differentially impacts host cell entry and intracellular persistence. Together, our findings establish that mycobacterial lipids act as molecular determinants of cytoskeletal organization, engaging distinct actin regulatory pathways to shape infection dynamics.

Significance

Pathogens commonly manipulate the host actin cytoskeleton to promote infection, yet how distinct microbial factors selectively engage specific actin regulatory pathways remains unclear. Here, we identify mycobacterial envelope lipids as direct modulators of host actin dynamics, acting through distinct mechanisms to control cytoskeletal organization. We show that different lipid species activate either cofilin-mediated depolymerization or WASP–Arp2/3-dependent polymerization, resulting in stage-specific changes in the actin structure that influence bacterial entry and dissemination. These findings establish a lipid-driven framework for cytoskeletal regulation during infection and reveal a previously unrecognized level of control in host–pathogen interactions.

Introduction

Mycobacteria possess a lipid-rich cell wall that confers resistance to antibiotics and host immune defenses. , These lipids are grouped into major classes based on structure and biosynthesis, including fatty acyls (FAs), fatty esters (FEs), glycerolipids (GLs), glycerophospholipids (GPhLs), saccharolipids (ScLs), prenols (PRs), and polyketides (PLs). − They are distributed across the inner and outer membranes. The outer membrane contains mycolic acids (MA; FA), phthiocerol dimycocerosate (PDIM; FE), trehalose dimycolate (TDM; FA), sulfoglycolipids (SL; ScL), phosphatidylinositol mannosides (PIMs; GLs), lipoarabinomannan (LAM), phenolic glycolipid (PGL; glycosylated FE), and diacylglycerol (DAG; GL), whereas the inner membrane consists primarily of phospholipids and acylated phosphatidylinositol mannosides (Ac2PIM2, PIM6; GPhLs) (Figure ). ,,

1.

Chemical structures of representative lipids present in the mycobacterial cell envelope. COO-group marked as blue, acyl chains marked as green, headgroup modifications marked as red, glycerol backbone marked as brown, phosphate group marked as orange, glycosylation marked in red, amide bond marked as purple, and all factors causing kinks in the acyl chain are marked as pink.

Mycobacterial envelope lipids are key determinants of virulence. − Lipidomic studies show that envelope composition varies across infection stages, indicating dynamic regulation of host–pathogen interactions. Individual lipids exert distinct effects on host cells: PDIM increases membrane rigidity and promotes necrosis; LAM disrupts cholesterol-rich domains and vesicle fusion; TDM reduces membrane fluidity and inhibits vesicle trafficking; and SL-1 enhances membrane fluidity and blocks phagosome maturation. − Together, these lipid-specific effects support stage-dependent modulation of host responses during infection.

The host cytoskeleton is a primary target of pathogenic manipulation. It comprises a dynamic network of protein filaments and microtubules with actin as a central component. Actin dynamics is regulated by proteins controlling polymerization and depolymerization. Actin nucleation is mediated by the Arp2/3 complex, with branching driven by nucleation-promoting factors such as N-WASP and WASH. , Proteins including profilin and radixin contribute to filament assembly, − while cofilin and gelsolin regulate depolymerization. Cytoskeletal organization at the plasma membrane is further modulated by phosphatidylinositol 4,5-bisphosphate (PI­(4,5)­P2) and actin-binding proteins such as ezrin, vinculin, and talin, which anchor actin filaments to the membrane. −

Mycobacteria induce changes in host actin organization, often resulting in punctate or fragmented filaments indicative of depolymerization. However, the actin phenotype varies with Mycobacterium tuberculosis (Mtb) lipids likely produced at different infection stages or showing time-dependent differences, introducing heterogeneity in the intracellular responses ,,,, and indicative of stage-specific remodeling of host actin. In this study, we examine the effects of TDM and PGL-1 on the localization, abundance, and dynamics of interactions of key actin cytoskeletal components and associated lipids using microscopy and mass spectrometry. These effects are compared with previously reported activities of SL-1 and PDIM, to arrive at a prospective model of infection stage–dependent host cell regulation. SL-1 and TDM are trehalose-based lipids with distinct structural features. SL-1 harbors a sulfate group at the trehalose 2′ position and mycolic acid chains esterified at the 2-, 3-, 6-, and 6′-positions, whereas TDM contains mycolic acids only at the 6- and 6′-positions (Figure ). Similarly, PGL-1 is a glycosylated derivative of PDIM, with added sugar moieties that modify its functional properties (Figure ). These surface Mtb lipids are loosely associated with the mycobacterial envelope, are shed during infection with possible cell-type and time-dependent differences, and can interact with host cells independently of intact bacteria. , Their extracellular availability thus supports analysis in isolation. Accordingly, this study focuses on host responses to extracellular mycobacterial lipid exposure at physiologically relevant concentrations; i.e., within the range present in the outer bacterial cell membrane. , The framework of this work thus enables isolation of the individual lipid contribution within the integrated mycobacterial lipidome.

We show that SL-1 and PGL-1, despite structural differences, induce actin depolymerization primarily through cofilin-mediated severing. In contrast, PDIM and TDM promote actin polymerization via activation of the WASP–Arp2/3 pathway at the plasma membrane. Lipidomics analysis indicates that SL-1 and PDIM are enriched during early infection, whereas PGL-1 and TDM are more abundant at later stages. Functionally, PGL-1-induced actin fragmentation does not enhance bacterial uptake, unlike SL-1, but is associated with increased bacterial spread following host cell exit. Notably, actin fragmentation is ASAP1-dependent only in the presence of SL-1, indicating distinct underlying mechanisms. TDM reduces bacterial uptake, likely due to increased actin polymerization, while PDIM shows minimal effect. Overall, these findings support a model in which mycobacterial lipids drive infection stage-specific actin remodeling through distinct mechanisms, influencing bacterial uptake and dissemination.

Results and Discussion

Trehalose and Mycocerosate-Derived Mtb Glycolipids Differentially Regulate the Host Cytoskeletal Proteins

Distinct host actin phenotypes have been reported for Mtb lipids, either using purified glycolipids or mutant strains lacking specific components. ,, To compare their effects on host cytoskeletal regulation, we performed label-free proteomics mass analysis to assess changes in the abundance of key host cytoskeletal-associated proteins in macrophages exposed to selective Mtb lipids (Figure A). SL-1 treatment led to reduced abundance of proteins, including PIP4K2A, PICALM, ARHGAP, SACM1L, radixin (RDX), profilin (PFN1), Arp2/3 complex subunits, and cofilin (Figure A), whereas these proteins were generally upregulated in response to TDM. ARHGAP and PFN1 are associated with actin polymerization; , their reduced levels in the presence of SL-1 may contribute to actin puncta induction, possibly mediated by actin disassembly, , while increased levels with TDM are consistent with enhanced actin stability in accord with a previous study. Similarly, PIP4K2A regulates PI­(4,5)­P2 levels, and its upregulation with TDM may support increased actin density. , SACM1L, a phosphoinositide regulator, may further influence actin organization through modulation of membrane-associated signaling; however, this would need high-resolution localized analysis. Despite reduced cofilin abundance under SL-1 treatment, previous observations show increased actin punctate formation, suggesting that actin dynamics are influenced by cofilin activity and localization rather than total protein levels. Higher levels of Arp2/3 subunits (ARPC2–5) with TDM are consistent with enhanced actin polymerization and subsequently increased actin content per cell.

2.

Mtb lipid-mediated regulation of cytoskeletal remodeling and ASAP1 expression in THP-1 macrophages. (A) Protein abundance of indicated proteins in THP-1 macrophages following exposure to indicated Mycobacterium tuberculosis (Mtb) lipids (10 μg/mL) for 24 h, assessed by label-free mass spectrometry. Data represent mean ± SEM from three independent biological replicates. (B) Representative images showing effects of PGL-1 (10 μg/mL; 4h) on actin cytoskeleton organization in live THP-1 cells. Scale bar, 10 μm; images acquired using a 60× oil immersion objective. (C) Quantification of actin puncta per cell differentiated THP-1 cells in the presence and absence of Mtb PGL-1 (10 μg/mL; 4 h). Data is mean ± SD from three independent experiments (n = 60–70 cells each). (D) Transcriptional regulation of ASAP1 under lipid treatment. Relative mRNA expression of ASAP1 was quantified by qRT-PCR in differentiated THP-1 cells treated with TDM or PGL-1 (10 μg/mL; 24 h). Gene expression was normalized to GAPDH and expressed relative to untreated control cells. Data represent mean ± SEM from three independent biological experiments performed in technical triplicates. (E) Representative images depicting filopodia structures in control and TDM/PGL-1-treated macrophages. (F) Quantification of filopodia length. Data represent mean ± SD from three independent biological replicates (n = 60–70 cells each). (G) Isosurface rendering of F-actin architecture in control and TDM/PGL-1-treated macrophages. (H) Quantification of the number of filopodia/cell and (I) length of filopodia in differentiated THP-1 cells in the presence and absence of Mtb PGL-1 (10 μg/mL). Data is mean ± SD from three independent experiments (n = 60–70 cells each). Statistical significance was determined by one-way ANOVA and Tukey’s multiple comparison test. P-values <0.05 were considered statistically significant. Significance levels are indicated as (*) P < 0.1, (**) P < 0.01, (***) P < 0.001, and (****) P < 0.0001.

Between PDIM and PGL-1, the most prominent differences were observed for ARP2/3, ASAP1, and PIP4K2C (Figure A). PDIM exposure reduced ARP2/3 and ASAP1 protein levels, whereas both were increased in response to PGL-1. In contrast, PIP4K2C expression increased with PDIM but decreased with PGL-1. ASAP1, an ArfGAP implicated in actin remodeling and mycobacterial uptake, was reduced under PDIM, consistent with prior observations of limited actin puncta formation. In contrast, PGL-1 increased ASAP1 abundance. Actin imaging surprisingly revealed that despite ASAP-1 increase, PGL-1 exposure induced actin puncta formation and reduced the filament length (Figures B,C and S1). This indicates that actin fragmentation in this context is not dependent on ASAP1. Gene expression analysis further confirmed that ASAP1 is upregulated at both transcriptional and translational levels in response to PGL-1 (Figure C). TDM similarly increased ASAP1 transcript levels (Figure C), whereas PDIM had no effect and SL-1 reduced its expression. These findings indicate that PGL-1 and SL-1 induce actin fragmentation, possibly via distinct mechanisms. Of note, the % increase in puncta was not concomitant with a similar magnitude of decrease of other actin morphologies as seen in our previous work as well, implying that though smaller actin structures are generated in the presence of PGL-1, it is not accompanied by substantial loss of filamentous or cortical action. Despite increased actin fragmentation, PGL-1 treatment resulted in higher total actin content, suggesting redistribution into smaller fragments and a dynamic steady state (Figures E,F and S1). This could also suggest that PGL-1 induces actin polymerization into new short structures, rather than global depolymerization or a balance between these two processes. Next, we evaluated filopodia length upon exposure to Mtb PGL-1. Filopodia are hair-like structures or protrusions from the plasma membrane used for sensing the environment and modulated by pathogens to facilitate entry or alter host cell migration. PGL-1 induced extension of filopodia length (Figures G,H and S1); an increased actin volume (Figures E,F and S1) can lead to enhancement of the filopodia length to modulate host cell migration.

Mtb Lipid-Mediated Membrane Modulation Impacts Biomolecular In-Plane Diffusion

Previously, we showed that mycobacterial lipids upon insertion into the host plasma membrane differentially modulate the membrane properties: SL-1 and PGL-1 increase membrane fluidity, whereas PDIM and TDM induce membrane rigidification. , These changes are expected to alter lateral diffusion of lipids and proteins within the membrane, thereby influencing protein–protein and protein–lipid interactions. To assess the contribution of in-plane membrane diffusion of proteins/lipids in cytoskeletal regulation, we performed fluorescence recovery after photobleaching to measure diffusion coefficient and mobile fractions of actin, actin-binding proteins, and PI­(4,5)­P2 in live THP-1 cells upon exposure to TDM and PGL-1. THP-1 cells were transfected with various fluorescently labeled constructs (PLCδ1PH–PIP2-GFP; mCherry-Talin; mEmerald-ARPp34-C-14; pEGFP-C1 Lifeact-EGFP; mTagRFP-T-Vinculin-23; Cofilin-pmCherryC1 and mEmerald-WASP1-C-14) and the diffusion of these species was monitored. Both Mtb lipids increased PI­(4,5)­P2 and actin diffusion, although the effect was weaker for PGL-1; only PGL-1 increased the mobile fraction (Figures A,B, S2, and S3). Membrane–cytoskeleton linker proteins, such as talin and vinculin, exhibited biphasic recovery (Figure S3). The biphasic nature is attributed to a fast initial membrane recruitment of the cytosolic fraction followed by lateral in-plane membrane diffusion. TDM had minimal effect on talin diffusion but reduced its mobile fraction. PGL-1 reduced the fast diffusion component of talin but increased the mobile fraction, suggesting reduced membrane recruitment. For vinculin, TDM had no effect, while PGL-1 reduced the fast component and modestly increased the slower component, indicating decreased recruitment but likely an increased membrane-associated clustering. These results indicate that membrane rigidification (induced by TDM) restricts protein mobility, whereas increased fluidity (induced by PGL-1) enhances redistribution and plausibly membrane clustering of cytoskeletal regulators. ,

3.

(A) Diffusion coefficients and (B) Mobile fractions of indicated proteins and lipids in live differentiated host cells after treatment with indicated Mtb lipids (20 μg/mL; 24 h). Data is mean ± SD from three independent experiments (n = 60 cells each). Significance was determined by one-way ANOVA and Tukey’s multiple comparison test. P-values <0.05 were considered statistically significant. Legend: (*) P < 0.1, (**) P < 0.01, (***) P < 0.001, and (****) P < 0.0001.

We then extended the diffusional analysis toward other key actin regulators, including ARPC2 (polymerization), N-WASP (branching), and cofilin (depolymerization). Both TDM and PGL-1 increased the diffusion of ARPC2 and N-WASP, whereas only TDM enhanced cofilin diffusion (Figure A,B). TDM increased the mobile fraction of all three proteins, while PGL-1 reduced the mobile fraction of ARPC2 and N-WASP. Although TDM has been shown to preferentially reduce lipid diffusion within the ordered membrane domains, thus the observed increase in protein diffusion suggests redistribution or dampened clustering of ARPC2 and N-WASP proteins within the liquid-disordered (l_d) lipid domains in THP-1 plasma membranes; however, this needs further validation.

TDM and PGL-1 Differentially Regulate Colocalization of Actin-Binding Proteins with PI­(4,5)­P2 and F-Actin

We previously reported that Mtb lipids alter the distribution of PI­(4,5)­P₂ lipid at the host plasma membrane, which may influence the localization of PI­(4,5)­P₂-interacting actin and associated proteins. To examine this hypothesis, we quantified colocalization of endogenous levels of PI­(4,5)­P₂ with actin regulators (ARPC2, N-WASP, Cofilin) using confocal microscopy in immune-stained THP-1 cells. In control cells, ARPC2 showed a strong association with PI­(4,5)­P₂, with a higher fraction of ARPC2 colocalizing with PI­(4,5)­P₂ than vice versa (Figure A). PGL-1 reduced this association, consistent with increased PI­(4,5)­P₂ mobility. TDM had no effect, aligning with minimal changes in PI­(4,5)­P₂ dynamics and clustering. Cofilin exhibited ∼60% colocalization with PI­(4,5)­P₂ in control cells (Figure B). This association decreased with PGL-1, likely due to increased PI­(4,5)­P₂ diffusion, while no change was observed with TDM. N-WASP showed high colocalization with PI­(4,5)­P₂ (M2 ≈ 0.72) (Figure C). This interaction was unchanged with PGL-1 and only slightly increased with TDM, consistent with enhanced actin polymerization under TDM conditions. Overall, PGL-1 reduces PI­(4,5)­P₂-associated interactions, whereas TDM maintains or modestly enhances them.

4.

(A) Colocalization of indicated host lipid and proteins in untreated or treated (TDM or PG-1; 20 μg/mL; 24 h) immunostained THP-1 cells. Mander’s correlation coefficient. (B) Quantification of colocalization for the indicated pairs (n = 3, N = 60). Scale bar: 10 mm, 63× oil objective. Significance was determined by one-way ANOVA and Tukey’s multiple comparison test. The white arrow shows the line over the cell along which the PCC was performed. Legend: (*) P < 0.1, (**) P < 0.01, (***) P < 0.001, and (****) P < 0.0001.

Pathogenic mycobacteria have been shown to exploit actin-based motility for intracellular spread by recruiting host cytoskeletal regulators such as ARPC2 and N-WASP. , To assess this aspect, we quantified colocalization of F-actin with actin regulator proteins in immune-stained THP-1 cells expressing endogenous levels of the proteins as investigated above (Figure ). In control cells, F-actin showed strong association with ARPC2 and N-WASP (M2 ≈ 0.65–0.68), particularly in cortical and podosomal regions (Figure A). TDM increased F-actin–ARPC2 association, while PGL-1 had no effect. Cofilin, a key actin-severing protein, showed no change in colocalization with F-actin in the presence of PGL-1, whereas TDM caused a modest increase (Figure B), consistent with reduced actin fragmentation. ,, In contrast, prior work shows SL-1 to decrease this association, promoting cofilin-mediated severing. This hints toward distinct mechanisms underlying similar actin phenotypes induced by SL-1 and TDM. TDM increased F-actin–N-WASP colocalization, whereas PGL-1 reduced it (Figure C). These results support enhanced actin polymerization/content with TDM and actin fragmentation with PGL-1 (Figure B). Notably, PGL-1–induced fragmentation appears to be independent of ASAP1 but dependent on N-WASP, further distinguishing its mechanism from SL-1.

5.

(A) Colocalization of indicated host lipid and proteins in untreated or treated (TDM or PG-1; 20 μg/mL; 24 h) immunostained THP-1 cells. Mander’s correlation coefficient. (B) Quantification of colocalization for the indicated pairs (n = 3, N = 60). Scale bar: 10 mm, 63× oil objective. Significance was determined by one-way ANOVA and Tukey’s multiple comparison test. The white arrow shows the line over the cell along which the PCC was performed. Legend: (*) P < 0.1, (**) P < 0.01, (***) P < 0.001, and (****) P < 0.0001.

Mycobacterial lipids are known to alter host membrane organization, including cholesterol-rich domains. , To assess the downstream effects of this perturbation, we examined colocalization of cholesterol with PI­(4,5)­P₂ and F-actin following exposure to TDM and PGL-1. TDM did not affect PI­(4,5)­P₂–cholesterol association but reduced F-actin colocalization with cholesterol (Figure S4A,B). PGL-1 increased PI­(4,5)­P₂ association with cholesterol while similarly reducing F-actin–cholesterol colocalization. These results indicate that both lipids preferentially disrupt F-actin association with cholesterol, with limited effects on PI­(4,5)­P₂ in the presence of TDM. Ezrin, a membrane–actin linker, showed increased colocalization with F-actin in the presence of PGL-1 and SL-1, but not TDM or PDIM (Figure S4C), suggesting enhanced membrane anchoring under fluidizing conditions. In contrast, colocalization of coronin with F-actin was unchanged under both TDM and PGL-1 (Figure S4D). This is consistent with the absence of large actin network formation, as coronin stabilizes actin filaments by binding and cross-linking them into larger networks and can also induce actin severing by recruitment of cofilin. , Finally, PGL-1 reduced F-actin–PI­(4,5)­P₂ colocalization (Figure S4E), likely reflecting decreased membrane association of fragmented actin structures.

Mtb Lipids Modulate Close-Range Interactions among Actin-Binding Proteins, Actin, and PI­(4,5)­P₂

Colocalization does not indicate the presence or absence of direct interactions necessitating further validation. This led us to monitor close-proximity interactions using fluorescence resonance energy transfer (FRET) analysis. FRET efficiencies for the same protein–lipid and protein–protein pairs as studied in colocalization assays were measured in live THP-1 cells in the presence of Mtb lipids (Figure ). Here, PI­(4,5)­P₂ or F-actin served as energy acceptors, while actin-binding proteins served as energy donors.

6.

Quantitative apparent FRET efficiency (in %) for the indicated acceptor–donor pairs in immunostained THP-1 cells. (A) PI­(4,5)­P₂ with N-WASP, (B) Actin with N-WASP, (C) PI­(4,5)­P₂ with ARPC, (D) Actin with ARPC, (E) PI­(4,5)­P₂ with Cofilin, and (F) Actin with Cofilin in untreated or treated (TDM or PG-1; 20 μg/mL, 4 h) cells. Significance was assessed using one-way ANOVA and Tukey’s multiple comparison test. Legend: (*) P < 0.1, (**) P < 0.01, (***) P < 0.001, and (****) P < 0.0001. PM: plasma membrane.

PI­(4,5)­P₂ activates N-WASP protein at the membrane, thereby initiating actin nucleation or recruiting ARP2/3 to initiate the actin filament assembly. Baseline FRET between PI­(4,5)­P₂ and N-WASP was higher at the plasma membrane than in the cytoplasm (Figure A). Both TDM and PGL-1 reduced this interaction, supported by increased lateral diffusion and reduced proximity. In contrast, TDM increased FRET between F-actin and N-WASP, whereas PGL-1 decreased it (Figure B,C), indicating enhanced association in the presence of TDM due to enhanced actin volume/cell. Reduced interaction with PGL-1 could be due to higher N-WASP diffusion. Both the lipids reduced FRET between PI­(4,5)­P₂ and ARPC2 (Figure D), and only PGL-1 decreased ARPC2–F-actin interaction. Similarly, PGL-1 reduced FRET between cofilin and both PI­(4,5)­P₂ and F-actin (Figure E,F), consistent with decreased association and increased actin fragmentation. This hints toward actin disassembly mediated by the perturbed interactome of Cofilin; actin; PI­(4,5)­P2 and N-WASP, underlying reduced filaments and enhanced punctate (Figure B). Overall, these results show that TDM and PGL-1 differentially regulate cytoskeletal organization by altering specific protein–protein and protein–lipid interactions.

TDM and PGL-1 Attenuate Intracellular Bacterial Burden

We previously showed that SL-1-induced actin remodeling is associated with enhanced bacterial uptake, whereas PDIM is not. Given that PGL-1 induced a similar actin phenotype, we evaluated the effects of PGL-1 and TDM on uptake and persistence of GFP-labeled Mycobacterium smegmatis, used as a fast-growing lab model for Mtb (Figure A).

7.

TDM and PGL-1 attenuate mycobacterial uptake or promote exit in THP-1 macrophages. (A) Representative images depicting internalization of GFP-expressing Msm in differentiated THP-1 cells pretreated with TDM or PGL-1 (10 μg/mL) and infected for 3 h. Nuclei were stained with DAPI (blue), and F-actin was visualized using TRITC–phalloidin (red) to outline cellular architecture. The cellular outline has been marked using white dashed lines. Yellow arrows indicate extracellular bacteria, while white arrows indicate intracellular bacteria. Images were acquired using a 100× oil immersion objective. Scale bar: 20 μm. Lipid-treated cells exhibit reduced intracellular bacterial signal compared to untreated controls. Data are representative of three independent biological experiments (N = 100–180 cells per condition). (B) Corrected total cell fluorescence (CTCF) was quantified from the imaging data shown in panel (A). Data is mean ± SEM from two independent experiments, demonstrating a reduction in bacterial uptake following TDM and PGL-1 treatment. (C) CFU-based quantification of intracellular bacterial burden in dTHP-1 macrophages following lipid treatment (10 μg/mL) and infection with Msm. CFU enumeration corroborates imaging-based observations, showing decreased intracellular bacterial counts in lipid-treated cells relative to controls, with TDM exhibiting a more pronounced effect than PGL-1. (D) Growth kinetics of Msm in the absence or presence of TDM or PGL-1 (10 μg/mL), assessed using a resazurin-based metabolic assay over time. (E) Growth dynamics of GFP-expressing Msm under identical conditions, confirming that TDM and PGL-1 do not significantly alter bacterial metabolic activity. (F) Extracellular bacterial association was quantified by measuring corrected total cell fluorescence (CTCF) per cell, representing the burden of surface-associated (noninternalized) bacteria. Fluorescence measurements were obtained from the GFP signal of GFP-Msm and normalized to cell number.

Quantitative imaging showed reduced intracellular bacterial signal in both TDM- and PGL-1-treated cells compared to control (Figure B), indicating decreased bacterial burden. CFU assays confirmed the reduced intracellular bacterial counts, with a stronger effect for TDM (Figure C). Increased actin stabilization, as observed previously with TDM, may impose a strong physical barrier for phagocytosis and likely limit the bacterial internalization. In contrast, PGL-1-induced actin fragmentation did not enhance uptake, unlike SL-1, suggesting distinct functional outcomes. Of note, actin undergoes polymerization during phagocytic uptake of bacteria; however, localized subdomains are also implicated to undergo tightly coupled assembly and disassembly. Here, we have performed global whole-cell averaging, which suggests loss of F-actin filaments in the presence of SL-1 and PGL-1, but not TDM or PDIM. These bulk measurements can miss rapid, transient bursts of polymerization that occur within seconds of contact. Furthermore, we propose actin depolymerization/fragmentation to underlie the enhanced actin puncta/reduced filaments, but the same can also result from actin polymerization into new but shorter structures. Currently, with the tools we have, we are unable to precisely delineate the contribution of both factors. Additionally, the use of Msm also limits the broader implication of these findings in the context of actual infection, and in future studies with a virulent mycobacterial strain closer to Mtb H37Rv would help address this.

Next, resazurin-based assays showed no effect of either lipid on bacterial growth or metabolic activity (Figure D,E), indicating that reduced intracellular burden likely reflects impaired uptake rather than growth inhibition. To delineate the functional consequence rendered by PGL-1-induced actin fragmentation, we assessed the abundance of extracellular bacteria and observed increased bacterial levels in PGL-1–treated cells (Figure F), consistent with enhanced bacterial release without host cell death: host cell viability was unaffected (Figure S5). Importantly, we cannot conclusively determine whether the observed actin fragmentation occurs specifically at the phagocytic cup/bacterial contact sites or reflects a more global reorganization of the actin cytoskeleton, precluding the arrival at a direct mechanistic link between actin disassembly and bacterial uptake and dissemination. Instead, the data describes these as associated phenotypes observed under the same experimental conditions. This could suggest (i) a localized actin remodeling at sites of bacterial engagement, which could directly facilitate uptake or exit, or (ii) a broader cytoskeletal disruption that indirectly influences the bacterial uptake/exit. In the future, high-resolution live-cell imaging of actin dynamics coupled with pharmacological perturbation of the actin network will facilitate deciphering the mechanistic relationship between actin reorganization and bacterial internalization and dissemination. Collectively, overall, TDM and PGL-1 differentially modulate macrophage responses by limiting bacterial uptake or promoting bacterial exit, highlighting lipid-specific association of cytoskeletal dynamics with distinct infection outcomes.

Proposed Model of Mtb Lipid-Driven, Infection Stage-Dependent Remodeling of the Host Actin Cytoskeleton

Several M. tuberculosis (Mtb) lipids interact with host cells and modulate key cellular processes, resulting in diverse infection outcomes. Although Mtb dynamically regulates its cell wall lipid composition during infection, − how temporal changes in the lipid landscape influence disease progression remains unclear. On the basis of experimental evidence demonstrating differential lipid expression across early and late stages of bacterial growth and infection, ,,, we propose a model in which infection outcomes are stage-dependent. Lipidomics mass analysis of intact bacteria from established models of early and late growth/infection stages indeed confirmed differential lipid abundance. While SL-1 and PDIM levels decreased at the late stage, PGL-1 (fatty ester class) levels increased (Figure S6). TDM could not be reliably quantified across stages, likely due to its metabolic competition with glucose monomycolate (GMM). , However, existing evidence indicates that TDM is present across growth states, with higher accumulation at late growth/infection stages.

FRET and colocalization analyses indicate that SL-1 and PGL-1 promote actin disassembly through selective disruption of cytoskeletal interactions. Both lipids reduce PI­(4,5)­P₂ association with N-WASP and ARPC2, attenuating actin polymerization. In addition, decreased cofilin association with PI­(4,5)­P₂, and with F-actin in the case of SL-1, suggests increased cytosolic cofilin activity, favoring actin severing. PGL-1 shows partial overlap with SL-1 effects but differs by maintaining ARPC2–F-actin association and showing weaker disruption of N-WASP–F-actin interactions. Moreover, PGL-1-induced actin fragmentation occurs independently of ASAP1, unlike SL-1, indicating distinct mechanisms. The observed actin phenotype with SL-1 and PGL-1 reflects fragmentation rather than complete depolymerization, generating smaller actin structures. This remodeling likely reduces the cortical barrier and facilitates bacterial uptake or host cell exit. Accordingly, actin dynamics shifts from organized filaments to fragmented short structures in a lipid- and infection stage-dependent manner. In contrast, PDIM and TDM do not induce actin depolymerization but modestly enhance actin polymerization. This is supported by increased PI­(4,5)­P₂–N-WASP and F-actin–N-WASP associations, along with enhanced membrane–cytoskeleton coupling, particularly via ezrin in the presence of TDM. These interactions are consistent with increased actin content per cell observed with both lipids.

Specifically, we speculate that regulated shedding of mycobacterial surface lipids, varying with time and host cell type, generates heterogeneity within the intracellular niche and directly influences the pathogenesis. This mechanism may provide Mtb with a sophisticated means of coordinately manipulating interconnected host pathways in a context-dependent manner. While speculative, this model (Figure ) offers a valuable conceptual framework to guide future investigations into the temporal roles of mycobacterial lipids during infection, leading to distinct outcomes. The contribution of individual lipid features, including sulfation, acyl chains, and glycosylation, remains unresolved and requires further investigation by using defined lipid analogues.

8.

Schematic representation of the effects of key mycobacterial glycolipids expressed at varied stages of infection on host cell actin dynamics. SL-1 and PDIM are expressed more during the early growth/infection stage, while TDM and PGL-1 are expressed more in the late growth/infection stage.

Conclusions

The actin cytoskeleton is a dynamically changing network regulated by coordinated polymerization and depolymerization processes mediated by proteins such as cofilin/ADF, the ARP2/3 complex, profilin, and nucleation-promoting factors, including WASP. Pathogens exploit this machinery to modulate the actin structure and generate environments that support entry, intracellular survival, and dissemination. The molecular mechanisms of each of these processes can be distinct, involving distinct host and pathogenic components.

This study demonstrates that mycobacterial lipids directly regulate host actin remodeling. The data proposes a lipid-dependent, infection stage-specific model of cytoskeletal reorganization (Figure ). During early infection, higher SL-1 abundance promotes actin fragmentation or the generation of short actin structures, enhancing bacterial uptake by facilitating membrane remodeling. In later stages, increased PGL-1 induces actin restructuring that seems to favor bacterial release from the host cell. The latter has indeed been shown for other pathogens like Chlamydiae or Clostridium difficile, and could be a tactic employed by mycobacterial species as well. In contrast, PDIM and TDM promote actin polymerization, potentially supporting intracellular survival or motility, although TDM is associated with reduced bacterial uptake. Previously, actin polymerization has been observed with Listeria monocytogenes or SFG Rickettsia, which leads to remodeling of actin structure on the pathogen surface in the form of actin comets or tails. These facilitate mobility of the pathogen within host cells to infect bystander cells via mechanisms of protrusion. , Specifically, Group Rickettsia stimulates actin polymerization via the WASP-ARPC-2 axis to form an actin tail that leads to early slow movement of bacteria through the cytoplasm shortly after uptake into a eukaryotic cell. The apparent miscorrelation between intact actin filaments and reduced intracellular bacterial count in the presence of TDM is surprising and mandates further experiments to decipher the underlying mechanisms. These also indicate that always reduced actin assembly may not always be associated with higher intracellular bacterial loads.

These findings indicate that distinct mycobacterial lipids differentially regulate actin dynamics to control key stages of infection, including entry, persistence, and dissemination. More work is mandated to fully study the functional consequences of these lipid-specific effects on host–pathogen interactions and yield a cohesive picture of the host-cytoskeleton–pathogen interface in tuberculosis pathogenesis.

Methods and Materials

All experiments were performed using the human THP-1 macrophage cell line obtained from Prof. Sarika Mehra (IIT Bombay). Cells were cultured in RPMI-1640 (HiMedia, AL028A) with fetal bovine serum (HiMedia RM112), 20× antibiotic-antimycotic solution (HiMedia, A002A), penicillin G (Sigma, 13752), and streptomycin (HiMedia, MB287). Cell differentiation was induced using PMA (phorbol 12-myristate-13-acetate; Sigma, P8139), and fixation was carried out with paraformaldehyde (HiMedia, TC703). Mycoplasma contamination was routinely monitored using the MycoAlert Mycoplasma Detection Kit (Lonza, LT07–701). Primary antibodies included Anti-Cofilin (ab42824), Anti-N WASP (EPR6959), Anti-ARPC2 (EPR8533), Anti-Ezrin (3C12), and Anti-PIP2 (2C11) (all from Abcam), with corresponding secondary antibodies conjugated to anti-rabbit Alexa Fluor 488 (Cat #A–11008), anti-mouse Alexa Fluor 561 (Cat #A–11004), or anti-rabbit Alexa Fluor 633 (Cat #A 21070) (Thermo Fisher). The ASAP1 primers were purchased from Sigma. Recombinant DNA constructs used for live-cell imaging and functional assays included PLCδ1PH-GFP (Laboratory of Tamas Balla, NIH, USA), mCherry-Talin, mTagRFP-T-Vinculin-2, mEmerald-ARPp34-C-143, pEGFP-C1 Lifeact-EGFP, Cofilin-pmCherryC1, mEmerald-WASP1-C-14, and PH-Akt-GFP (all from Addgene). The GFP-expressing Msm (GFP-Msm) is Msm mc²155, which is a kind gift from Prof. Amit Singh, IISc and Prof. Siddharth Chopra, CDRI, respectively. Data analysis and visualization were performed using ChemDraw 21.0.0, Inkscape 1.1, GraphPad Prism v10, ImageJ, and Fiji.

Methods

Mammalian Culture

THP-1 cells were cultured in RPMI-1640 medium with 10% heat-inactivated FBS and penicillin (100 U/mL), streptomycin (100 mg/mL), and amphotericin B (20 μg/mL) at 37 °C in 5% CO₂. THP-1 monocytes were cultured (at 2.5 × 10⁵/mL) with media replacement every 3 days. For the experiments, THP-1 monocytes were differentiated using 20 nM PMA for 48 h; cells were then incubated in complete media (no PMA) for 24 h. HEK293 cells overexpressing TLR2-YFP were cultured in DMEM media with 10% FBS, penicillin (100 U/mL), and streptomycin (100 mg/mL) at 37 °C in 5% CO₂. All experiments were performed with low cell passages [# 18–20]. Cells were tested for Mycoplasma and found to be noncontaminated. For transfection, differentiated cells seeded on glass coverslips were transfected with Lipofectamine (Life Technologies) and the indicated plasmids for 8–10 h following media replacement.

FRAP Analysis

Transfected cells were incubated with the Mtb lipids (24 h; 20 μg/mL) followed by washing with 1x PBS and replacement with phenol-free complete media before data acquisition. Three regions of interest (ROI) were selected. (i) bleached, (ii) reference (scan region that is not bleached), and (iii) background regions using 63× objective. Continuous confocal scanning frame mode with 12-bit intensity resolution (256 × 256 pixels; digital zoom 4×), and a pixel dwell time of 1.58 μs was employed. Bidirectional scanning was done with the 1 Airy unit pinhole. A region of 8-pixel radius was chosen for photobleaching, and the total experiment consisted of 500 cycles; each cycle had a run time of 242.04 ms (Time Series tab). Bleaching was done using 30 iterations (100% laser power transmission), including 10 scans of the prebleach area before starting the photobleaching. For reduced electronic noise and better data quality, 2–4 scans were considered. The detailed procedure for data normalization, correction, and processing can be found in ref . In brief, double normalization correction was applied to the raw data, followed by full scale normalization and fitting the data to a one-phase association to obtain the half-lives, % mobile fraction, and diffusion coefficients.

Colocalization and FRET Analysis

Differentiated cells were treated with indicated Mtb lipids for 24 h and fixed in 4% paraformaldehyde/PBS for 10 min at RT, followed by 1× PBS wash and permeabilization using 0.1% Triton-X-100/PBS for 5 min at RT. Thereafter, blocking was performed using 1% BSA in 1× PBST for 1 h followed by washing with 1× PBS/PBST for 5 times. TRITC–Phalloidin (1:1000) was used for staining in 1× PBS for 30 min in the dark, washed with 1× PBS, and coverslips were dried for 2–3 min. Samples were incubated with primary and secondary antibodies against Cofilin, ARPC-2, N-WASP, Filipin, and PIP₂ (1:1000 and 1:500 dilutions, respectively) overnight and 45 min in the dark, respectively. Stained coverslips were washed with 1× PBS, followed by mounting. Images were acquired on a laser scanning confocal microscope (Carl Zeiss 780) with a 63×/1.4NA objective. All postacquisition analyses were done using Imaris, Zen 3.0, and ImageJ Software (Colac and Colocalization analyzer). The fluorescence intensity in both channels was quantified by drawing a line over the cells using the line profiler module in the Zen 3.0 software. The FRET and colocalization experiments were done on the indicated pairs. (1) F-actin with various actin-binding proteins (ABPs; N-WASP, ARPC-2, Cofilin, Filipin, Ezrin, and PI­(4,5)­P₂) in which F-actin acted as an acceptor and ABPs as donors. (2) PI­(4,5)­P₂ as an acceptor, and various ABPs (N-WASP, ARPC-2, Cofilin, Filipin) served as donors.

For FRET experiments, the bleaching was performed in the acceptor channel using the acceptor photobleaching FRET module, while the emission intensity was monitored in the donor emission channel (postbleach). Using AccPbFret and Fret calculator software in ImageJ, cross-talk calibration, photobleaching correction, and partial photobleaching correction were done. FRET efficiency was then obtained after applying other correction factors explained in detail in refs , .

Colocalization Analysis

All image analyses were done using Zen 3.0 software, and the fluorescence intensity was quantified using a line drawn across the cell with the line profiler module. In transfected cells, pixel intensity histograms were generated with the intensity not saturated at the highest level. Greater than 1 line was drawn on each cell to account for the reproducibility of intensities. Approx. 60 cells were scored/condition/three independent experiments using blue edition Zen software (Zeiss, Jena, Germany). The data were then exported and plotted using Microsoft Excel 2007 (Redmond, WA) and GraphPad Prism 5.0. Mander’s colocalization coefficient and Pearson correlation coefficient were calculated using COLAC and IMARIS colocalization software, respectively. Coefficient values >0.8 indicate acceptable colocalization, and values <0.5 indicate less colocalization.

Analysis of Actin and Filopodia

Differentiated THP-1 cells were incubated with PGL-1 for 4 h at a concentration of 10 μg/mL. Later on, cells were fixed with 4% paraformaldehyde for 10 min at RT and rinsed with 1× PBS three times. Then, cells were permeabilized using 0.1% Triton-X-100 in 1× PBS, followed by blocking in 1% BSA for 1.5 h and rinsed with PBS/PBST. TRITC–Phalloidin was used to stain F-actin (2 μg/mL) for 30 min in the dark; it was rinsed thoroughly with 1× PBS, and slides were mounted. Both Fiji and Olympus CellSens software were used in this investigation to quantify the actin area. Phalloidin-stained pictures were processed using Fiji, a 3D object counter for analysis, which allowed the distribution of actin within cells to be quantified. Area measurements were used to ascertain the degree of actin coverage after thresholding techniques were used to separate actin filaments from background noise. In order to evaluate actin distribution in a volumetric setting, three-dimensional picture stacks were analyzed using the Olympus CellSens software’s Voxel Tool Window for voxel imaging. Actin morphology and distribution in two- and three-dimensional regions were thoroughly investigated using this dual method.

Further, data were analyzed and plotted using Microsoft Excel 2019 and GraphPad Prism 5.0. Postacquisition image analysis for F-actin structures was performed using the Phyllo Quant plugin in Fiji software. Relative quantification was determined by calculating the number of puncta (ITCN plugin, ImageJ) per cell and dividing by the total cell number. The filopodia number and length were quantified using the FiloQuant plugin in Fiji. A minimum of 60–70 cells was considered per condition per experiment.

Lipidomic Analysis

Mycobacterial lipids were extracted and quantified as previously reported using an Agilent 1260 HPLC (Agilent Technologies; Palo Alto, CA) and an Agilent AdvanceBio 6545XT quadrupole time-of-flight (Q-TOF) mass spectrometer equipped with an Agilent ESI/atmospheric-pressure chemical ionization (APCI) multimode source. Mass spectrometric analyses were performed in positive electrospray ionization (ESI+) mode. The instrument was operated with a nebulizer pressure of 45 psig, a capillary voltage of 4.0 kV, a drying gas flow rate of 8 L min^–1, drying gas and vaporizer temperatures of 300 °C, fragmentor, charging, and skimmer voltages of 175, 1000, and 65 V, respectively. The octopole radio frequency voltage was kept at 750 V. Spectral data were acquired in 4 GHz high-resolution mode at a scan frequency of 1 spectrum/s, corresponding to 5882 transients per spectrum. Profile-mode spectra were collected over an m/z window of 100–3200. Instrument calibration was achieved using Agilent Tune Mix standards. Accuracy was maintained through continuous infusion of a reference ion at m/z 922.009798. Data acquisition was carried out using Agilent MassHunter Workstation software (vB.10.00). Product-ion analyses for compound verification were conducted on an Agilent 6545 quadrupole time-of-flight (Q-TOF) mass spectrometer. Identical source parameters were employed, while tandem MS experiments utilized collision-induced dissociation with a collision-energy slope of 6.5 V per 100 Da and an offset value of 2.0 V. LC-MS data sets were processed using MassHunter Qualitative Analysis software (vB.10.00; Agilent Technologies, Santa Clara, CA, USA). Molecular features were generated using the Molecular Feature Extraction (MFE) workflow, which identifies and consolidates covarying ions, including isotopic species and different charge-state forms, into a single molecular feature based on accurate-mass chromatographic data. Small-molecule extraction settings were applied with count-based peak filtering. The analysis considered protonated ions (+H), a peak-spacing tolerance of 0.0005 m/z combined with 5 ppm mass accuracy, a common-organic-molecule isotope model, and charge states of +1 and +2. No restrictions based on compound class, mass range, or mass defect were implemented during feature extraction. The resulting molecular features were subsequently screened against formula libraries compiled from the Mtb LipidDB and MycoMass databases using the MassHunter compound-screening workflow. Candidate assignments were generated based on exact-mass matching with a tolerance threshold of 5 mDa. Both positive and negative charge carriers (+H and −H) were included during the search procedure, and charge states ranging from 1 to 2 were considered for compound annotation. From the list thus obtained, the quantitative abundance of each compound was derived by using the following formula.

$$\mathrm{c}\mathrm{oncentration}\mathrm{of}\mathrm{lipids}(\mathrm{ppm})=\frac{{\mathrm{a}\mathrm{r}\mathrm{e}\mathrm{a}}_{\mathrm{l}\mathrm{i}\mathrm{p}\mathrm{i}\mathrm{d}\mathrm{s}}\times \mathrm{c}\mathrm{o}\mathrm{n}\mathrm{c}\mathrm{e}\mathrm{n}\mathrm{t}\mathrm{r}\mathrm{a}\mathrm{t}\mathrm{i}\mathrm{o}\mathrm{n}\mathrm{o}\mathrm{f}\mathrm{r}\mathrm{e}\mathrm{s}\mathrm{e}\mathrm{r}\mathrm{p}\mathrm{i}\mathrm{n}\mathrm{e}(\mathrm{p}\mathrm{p}\mathrm{m})}{{\mathrm{a}\mathrm{r}\mathrm{e}\mathrm{a}}_{\mathrm{r}\mathrm{e}\mathrm{s}\mathrm{e}\mathrm{r}\mathrm{p}\mathrm{i}\mathrm{n}\mathrm{e}}}$$ 1

$$\mathrm{m}\mathrm{ol}\%\mathrm{of}\mathrm{\hspace{0.17em}\hspace{0.17em}}\mathrm{lipids}=\frac{{\mathrm{c}\mathrm{o}\mathrm{n}\mathrm{c}\mathrm{.}}_{\mathrm{l}\mathrm{i}\mathrm{p}\mathrm{i}\mathrm{d}\mathrm{s}(\mathrm{p}\mathrm{p}\mathrm{m})}\times 0.001}{\mathrm{M}{\mathrm{W}}_{\mathrm{l}\mathrm{i}\mathrm{p}\mathrm{i}\mathrm{d}\mathrm{s}}}\times 100\%/\mathrm{t}\mathrm{o}\mathrm{t}\mathrm{a}\mathrm{l}\mathrm{n}\mathrm{o}\mathrm{.}\mathrm{o}\mathrm{f}\mathrm{m}\mathrm{o}\mathrm{l}\mathrm{e}\mathrm{s}$$ 2

Liposome Preparation and Lipid Treatment

Trehalose dimycolate (TDM) and phenolic glycolipid-1 (PGL-1) liposomes were prepared using a thin-film hydration method. Lipids were initially dissolved in chloroform:methanol (2:1, v/v) at concentrations between 250 and 500 μg/mL to 1 mg/mL and stored at −80 °C until further use. Organic solvents with a working concentration of 10 μg/mL were evaporated under a gentle stream of nitrogen gas, followed by lyophilization to ensure complete removal of residual solvent and formation of a dry lipid film. The lipid films were hydrated in serum-free growth media and subjected to five cycles of vortexing and sonication to facilitate liposome formation.

Cell Viability Using Resazurin Assay

Differentiated cells (2 × 10⁴ cells per well) were treated with TDM or PGL-1 liposomes (10 μg/mL; 24 h). Untreated cells served as negative controls, while wells containing medium alone were used as blanks. Positive controls were generated using 50% DMSO to induce complete cell death. All conditions were performed in triplicate. Following treatment, culture medium was replaced with 200 μL of freshly prepared resazurin working solution (44 μM in RPMI), and plates were incubated for 4 h at 37 °C in the dark. Absorbances were measured at 570 and 600 nm using a plate reader. The following equations were used to monitor cell viability and then relative cell viability.

$$\mathrm{r}\mathrm{eduction}\mathrm{percentage}=\frac{E_{\mathrm{O}\mathrm{X}\mathrm{\_}600\mathrm{n}\mathrm{m}}\times A_{570\mathrm{n}\mathrm{m}\mathrm{\_}tx}-E_{\mathrm{O}\mathrm{X}\mathrm{\_}570\mathrm{n}\mathrm{m}}\times A_{600\mathrm{n}\mathrm{m}\mathrm{\_}tx}}{E_{\mathrm{R}\mathrm{E}\mathrm{D}\mathrm{\_}570\mathrm{n}\mathrm{m}}\times A_{600\mathrm{n}\mathrm{m}\mathrm{\_}t0}-E_{\mathrm{R}\mathrm{E}\mathrm{D}\mathrm{\_}600\mathrm{n}\mathrm{m}}\times A_{570\mathrm{n}\mathrm{m}\mathrm{\_}t0}}$$ 3

$$\mathrm{r}\mathrm{elative}\mathrm{c}\mathrm{ell}\mathrm{v}\mathrm{iability}=\frac{{\mathrm{c}\mathrm{e}\mathrm{l}\mathrm{l}\mathrm{v}\mathrm{i}\mathrm{a}\mathrm{b}\mathrm{i}\mathrm{l}\mathrm{i}\mathrm{t}\mathrm{y}}_{\mathrm{s}\mathrm{a}\mathrm{m}\mathrm{p}\mathrm{l}\mathrm{e}}-{\mathrm{c}\mathrm{e}\mathrm{l}\mathrm{l}\mathrm{v}\mathrm{i}\mathrm{a}\mathrm{b}\mathrm{i}\mathrm{l}\mathrm{i}\mathrm{t}\mathrm{y}}_{\mathrm{P}\mathrm{C}}}{{\mathrm{c}\mathrm{e}\mathrm{l}\mathrm{l}\mathrm{v}\mathrm{i}\mathrm{a}\mathrm{b}\mathrm{i}\mathrm{l}\mathrm{i}\mathrm{t}\mathrm{y}}_{\mathrm{N}\mathrm{C}}-{\mathrm{c}\mathrm{e}\mathrm{l}\mathrm{l}\mathrm{v}\mathrm{i}\mathrm{a}\mathrm{b}\mathrm{i}\mathrm{l}\mathrm{i}\mathrm{t}\mathrm{y}}_{\mathrm{P}\mathrm{C}}}$$ 4

Bacterial Uptake Assay Using Confocal Microscopy

Differentiated THP-1 cells (5 × 10⁴–6 × 10⁴ cells per well on 12 mm coverslips) were treated with TDM or PGL-1 liposomes (10 μg/mL; 24 h). Cells were infected for 3 h with GFP-expressing Msm with a multiplicity of infection (MOI) of 10:1. Extracellular bacteria were removed by washing twice with sterile phosphate-buffered saline (PBS), followed by a recovery period of 1–2 h in complete RPMI medium without antibiotics. Cells were fixed with 4% paraformaldehyde for 10 min at RT and permeabilized using 0.1% Triton-X-100 for 5–10 min. Nonspecific binding was blocked with 1% bovine serum albumin (BSA) for 1 h. F-actin was stained using TRITC–phalloidin (1 μg/mL), and nuclei were stained with DAPI (1:1000 dilution; 300 nM stock). Coverslips were mounted and imaged with an Olympus Fluoview FV3000 confocal microscope using a 100× oil objective. Z-stack images were acquired and analyzed using Fiji (ImageJ). The Corrected Total Cell Fluorescence was calculated as follows

$$\mathrm{CTCF}=\mathrm{i}\mathrm{ntegrated}\mathrm{d}\mathrm{ensity}-(\mathrm{a}\mathrm{rea}\mathrm{of}\mathrm{selected}\mathrm{cell}\times \mathrm{m}\mathrm{ean}\mathrm{fluorescence}\mathrm{of}\mathrm{background}\mathrm{readings})$$ 5

Intracellular Infection and CFU Enumeration

THP-1 cells were seeded in 96-well plates at a density of 2 × 10⁴ cells per well and differentiated with 100 nM PMA. Cells were treated with TDM or PGL-1 liposomes (10 μg/mL) for 24 h prior to infection. Cells were infected with M. smegmatis at an MOI of 1:10 for 3 h in serum-free, antibiotic-free RPMI medium. Post infection, any extracellular bacteria were removed by washing with PBS and kept for 1–2 h in complete medium but without antibiotics. Cells were lysed using sterile water, and lysates were serially diluted and plated onto Luria agar plates and incubated at 37 °C for 48–72 h, after which colony-forming units (CFUs) were calculated as per eq .

$$\frac{\mathrm{C}\mathrm{F}\mathrm{U}}{\mathrm{m}\mathrm{l}}=\frac{(\mathrm{n}\mathrm{u}\mathrm{m}\mathrm{b}\mathrm{e}\mathrm{r}\mathrm{o}\mathrm{f}\mathrm{c}\mathrm{o}\mathrm{l}\mathrm{o}\mathrm{n}\mathrm{i}\mathrm{e}\mathrm{s}\times \mathrm{d}\mathrm{i}\mathrm{l}\mathrm{u}\mathrm{t}\mathrm{i}\mathrm{o}\mathrm{n}\mathrm{f}\mathrm{a}\mathrm{c}\mathrm{t}\mathrm{o}\mathrm{r})}{\mathrm{v}\mathrm{o}\mathrm{l}\mathrm{u}\mathrm{m}\mathrm{e}\mathrm{o}\mathrm{f}\mathrm{c}\mathrm{u}\mathrm{l}\mathrm{t}\mathrm{u}\mathrm{r}\mathrm{e}\mathrm{p}\mathrm{l}\mathrm{a}\mathrm{t}\mathrm{e}\mathrm{d}(\mathrm{m}\mathrm{l})}$$ 6

Growth Kinetics of Msm in the Presence of Lipids

Msm cultures (including GFP-expressing strain) were grown to OD₆₀₀ ≈ 0.8 and diluted to OD₆₀₀ = 0.05. Aliquots (100 μL) were treated with liposomes (10 μg/mL), followed by the addition of 8 μL of 0.02% solution of resazurin. Samples were incubated in 96-well plates at 37 °C with shaking at 100 rpm. Absorbances at 570 and 600 nm were measured every 30 min for 24 h using a reader. Untreated bacteria and sterile medium served as controls. All experiments were conducted in triplicate. The following equation was used to monitor the reduction percentage, which was then normalized to cell viability percentage.

$$\mathrm{r}\mathrm{eduction}\mathrm{percentage}=\frac{E_{\mathrm{O}\mathrm{X}\mathrm{\_}600\mathrm{n}\mathrm{m}}\times A_{570\mathrm{n}\mathrm{m}\mathrm{\_}tx}-E_{\mathrm{O}\mathrm{X}\mathrm{\_}570\mathrm{n}\mathrm{m}}\times A_{600\mathrm{n}\mathrm{m}\mathrm{\_}tx}}{E_{\mathrm{RE}\mathrm{D}\mathrm{\_}570\mathrm{n}\mathrm{m}}\times A_{600\mathrm{n}\mathrm{m}\mathrm{\_}t0}-E_{\mathrm{R}\mathrm{E}\mathrm{D}\mathrm{\_}600\mathrm{n}\mathrm{m}}\times A_{570\mathrm{n}\mathrm{m}\mathrm{\_}t0}}$$ 7

Gene Expression Analysis by Quantitative Real-Time PCR

THP-1 cells (2 × 10⁶ per well) were seeded in 6-well plates and differentiated using 100 nM PMA for 48 h, followed by sequential recovery periods in antibiotic-containing and antibiotic-free media. Cells were treated with liposomes (10 μg/mL; 24 h) or infected with Msm for 3 h, followed by a 24 h incubation. Total RNA was extracted using the RNeasy Mini Kit, and 0.5 μg RNA was reverse transcribed using QuantiTect reagents. qRT-PCR was done using SYBR Green on a QuantStudio 5 system. Reactions included 500 ng cDNA and 10 μM gene-specific primers. Gene expression was normalized to GAPDH and calculated using the 2^–ΔΔCt method.

Sample Preparation for Proteomic Analysis of Lipid-Treated Macrophages

For lipid treatment conditions, differentiated THP-1 macrophages (dTHP-1) were exposed to liposomal formulations of mycobacterial lipids, TDM, and PGL-1, at a final concentration of 10 μg/mL for 24 h. Liposomes were prepared as described previously, and all treatments were performed under identical culture conditions to ensure experimental consistency.

Following treatment, cells were rapidly harvested on ice to preserve protein integrity and post-translational modifications. Monolayers were washed with ice-cold PBS, and cells were then pelleted by centrifugation at 5000 rpm for 15 min at 4 °C. The cell pellets were either used immediately or flash-frozen and stored (−80 °C) until further use.

For proteomic analysis, cell pellets were resuspended in 100–200 μL of denaturing lysis buffer (6 M urea, 2 M thiourea, 50 mM Tris–HCl (pH 8.0), 75 mM NaCl, and 1 mM MgCl₂). The buffer was supplemented with 1× PhosSTOP phosphatase inhibitor (Roche) and a broad-spectrum protease inhibitor cocktail (Thermo Scientific Pierce Protease Inhibitor Cocktail). Cells were lysed by intermittent vortexing on ice for 30–45 min to facilitate solubilization of proteins under denaturing conditions. This was followed by probe sonication (2 s ON/2 s OFF cycles at 40% amplitude) for a total duration of 5 min to ensure efficient mechanical disruption and nucleic acid shearing. Lysates were incubated on ice for an additional 10 min, and insoluble debris was removed by centrifugation (10,000g for 15 min at 4 °C). The total protein concentration in clarified supernatant was measured using the Pierce 660 nm protein assay kit.

Next, 50 μg of protein for each sample was adjusted to a final volume of 20 μL using double-distilled water. Proteins were reduced using 20 mM tris­(2-carboxyethyl)­phosphine (TCEP) at 37 °C for 60 min or 5 mM dithiothreitol (DTT) at RT for 55 min. Following reduction, free thiol groups were alkylated using 10 mM iodoacetamide (IAA) under dark conditions for 30 min (TCEP-treated samples) or 45 min (DTT-treated samples). Next, samples were diluted 4-fold with 50 mM Tris–HCl (pH 8.0) to reduce the urea concentration to <1 M. Proteolytic digestion was carried out using sequencing-grade trypsin at an enzyme-to-substrate ratio of 1:50, followed by overnight incubation (∼16 h) at 37 °C in a shaking dry bath.

Following digestion, peptide samples were acidified with formic acid to terminate enzymatic activity and promote peptide stability. Peptides were subsequently desalted using in-house prepared C18 StageTips. Briefly, columns were activated with methanol (2 × 50 μL), equilibrated with acetonitrile (ACN) with 0.1% formic acid, and conditioned with 0.1% formic acid. Peptide samples, reconstituted in 0.1% formic acid, were loaded onto the columns in aliquots and centrifuged at 1000g. After washing with 0.1% formic acid to remove salts and contaminants, bound peptides were eluted sequentially using 40%, 50%, and 60% ACN in 0.1% formic acid. Eluted peptides were dried using a SpeedVac vacuum concentrator and stored at −20 °C for short-term storage (up to 6 months) or at −80 °C for long-term storage until LC–MS/MS analysis.

Liquid Chromatography–Tandem Mass Spectrometry (LC–MS/MS) and Protein Identification

Peptide samples following desalting were analyzed using an EASY-nLC liquid chromatography system (Thermo Scientific) interfaced with a Q Exactive Plus Orbitrap mass spectrometer (Thermo Fisher Scientific). For each analysis, approximately 500 ng of peptides was introduced onto the analytical column, with column loading performed at pressures not exceeding 750 bar. Separation of peptide mixtures was achieved using a binary solvent system composed of 0.1% formic acid in water (mobile phase A) and 0.1% formic acid in 80% acetonitrile (mobile phase B). Elution was carried out at a flow rate of 300 nL min^–1 across a 120 min run. The gradient profile began with 2% mobile phase B held for 2 min, followed by a gradual increase to 15% over 78 min. The proportion of solvent B was then raised to 45% during the subsequent 20 min and finally increased to 95% over 15 min. This high-organic phase was maintained for 5 min to ensure complete column cleaning. To reduce sample carryover, the autosampler was rinsed between injections using a 100 μL wash step. Mass spectrometric measurements were acquired in positive-ion mode using a data-dependent acquisition workflow. Survey scans (MS1) were collected over an m/z interval of 350–2000 at a resolving power of 70,000 (m/z 200 reference), with an AGC target value of 1 × 10⁶ and a maximum ion accumulation time of 60 ms. For each duty cycle, the 15 most abundant precursor ions were selected for fragmentation by higher-energy collisional dissociation (HCD). Fragmentation was performed using a normalized collision energy of 27 and an isolation width of 1.2 m/z. Product-ion spectra (MS2) were recorded at a resolution of 17,500 with an AGC target of 2 × 10⁵ and a maximum injection time of 120 ms. To enhance proteome coverage and avoid repeated analysis of identical precursor ions, dynamic exclusion was applied with a duration of 35 s. In addition, peptide matching and isotope exclusion settings were enabled during precursor selection. Internal mass recalibration was achieved using a lock mass at m/z 445.12003 to maintain high mass measurement accuracy throughout the analysis. Full-scan spectra were acquired in profile format, whereas fragmentation spectra were recorded in centroid mode, using a single microscan for each scan event. The acquired LC-MS/MS files were processed in Proteome Discoverer software (version 2.2; Thermo Fisher Scientific) for peptide and protein identification, label-free quantification, chromatographic alignment, and validation of spectral assignments. Subsequently, the processed data sets were exported for statistical evaluation and downstream bioinformatic analyses using Microsoft Excel and MetaboAnalyst.

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acsomega.6c07536.

• Additional confocal images of actin morphology, thermal FRAP images and recovery curves, additional colocalization data sets, cell viability tests, and lipidomic data (PDF)

∥.

Department of Infectious and Geographic Medicine, Stanford School of Medicine, California 94305, United States

§.

M.L., S.L.R., and A.D. contributed equally to this work. Microscopy-based experiments were performed by M.M., S.L.R., A.D., and M.M. A.D. performed proteomics, and A.P.M. did lipid analysis. S.K., S.L.R., and A.D. made the figures and wrote the manuscript. S.K. supervised the project.

This work was supported by the DBT/Welcome Trust India Alliance Intermediate Fellowship (IA/I/21/1/505624) awarded to S.K. Central CSIF facilities at IIT Bombay are gratefully acknowledged. Mycobacterium leprae Phenolic Glycolipid-I (PGL-I), NR-19342, and trehalose dimycolate, NR-14844, were obtained from BEI Resources, NIAID, NIH. MP Biomedical is acknowledged through the MP Biomedical and WRCB joint venture. We thank Purushothaman R for helping A.D. in experimentation and data analysis as an IRCC, IIT Bombay Intern.

The authors declare no competing financial interest.

Published as part of ACS Omega special issue “Celebrating Women in the Chemical Sciences in India”.