All bacterial strains used in this study are listed in Supplementary Table S1.

We cultured M. abscessus ATCC 19977 (hereafter Mab) aerobically at 37°C in 7H9 Middlebrook (MB) (BD, Difco) broth supplemented with 0.2–0.5% glycerol, 0.05% Tween 80 and 10% ADC (albumin-dextrose-catalase) (BD, BD Biosciences) to an OD₆₀₀ of 1.3–1.6. This culture was used to prepare stocks in MB media with 20% glycerol and stored at −80°C. These stocks were used as an initial inoculum to culture Mab in Muller Hinton (MH) media for all our co-culture experiments as MH media supported the growth of all the bacteria used in this study. Glycerol stocks of P. aeruginosa ATCC 27853 (hereafter referred to as PaATCC) and other strains of P. aeruginosa (PA collectively used for all the P. aeruginosa isolates used in this study, unless specified as PaATCC for ATCC 27853), and S. aureus (SA) (hereafter SA, used collectively for MRSA and MSSA) stored at −80°C were streaked on blood agar plates and a single colony was used to inoculate cultures in MH media for overnight incubation to be used as an inoculum for all the experiments. We used two different approaches for bacterial co-culture studies; (1) Direct co-culture—when Mab and other bacteria were cultured together in MH media, (2) Indirect method—Mab grown in presence of filtrate (or supernatant) of PA and SA prepared in MH media. For infection experiments we harvested logarithmic phase bacterial cells by centrifuging bacterial inoculum at 8,000 rpm for 5 min and resuspended the pellet in experimental media (DMEM+5% FBS) for infection in RAW264.7 murine macrophages. Any clump of Mab was removed by passing the bacterial suspension through a 28G syringe needle (BD Sciences) at least 15 times (Kim et al., 2019). After infection, we determined bacterial colony forming units (CFU) by plating 10-fold dilutions of the bacterial suspension in duplicate on selective MH or LB agar plates after 4–5 days of incubation at 37°C.

Colony counts were determined by the drop plating method on selective media agar plates (Rodríguez-Sevilla et al., 2018b). Mab colonies were selected on Columbia Colistin Nalidixic Acid (CNA) agar plates when co-cultured with PA (Rodríguez-Sevilla et al., 2018b) and on vancomycin (2 mg/mL) when co-cultured and with SA. PA colonies were selected on MacConkey agar plate (Rodríguez-Sevilla et al., 2018b) that inhibits Mab and SA growth. SA when co-cultured with PaATCC were selected on CNA agar plates and on normal MH plates during co-culture with Mab as SA colonies appeared following day of plating while Mab colonies took 4–5 days to grow.

Bacterial supernatants after culturing PA isolates and SA in MH media were prepared with slight modifications from Beaudoin et al. (2017). A single colony of each PA strain and SA was inoculated separately in 5 mL MH media in 50 mL tubes and incubated overnight at 37°C in a shaking incubator. A 100 μL aliquot of this overnight culture was then added to a fresh 100 mL of MH media and incubated at 37°C in a shaking incubator until reaching an OD600_nm of 2.0–2.2. Following growth, bacterial cells were pelleted by centrifugation at 8,000 rpm for 15 min. We collected the supernatant and sterilized it by passing it through a 0.22 μm filter. Total protein concentration of the supernatant was determined using bicinchoninic acid (BCA) kit and the concentrations were maintained for all the experiments. We checked the sterility of supernatants by plating 1 mL of supernatant on agar plates (200 μL/plate). For all the indirect contact experiments, we diluted each of the supernatants from different strains to 1:1 with MH media. To account the use of supernatant (spent media) for our indirect contact experiments we used MH media diluted 1:1 with 1xPBS as a control. After filtration, supernatant was autoclaved in an autoclave for preparation of heat inactivated supernatant.

We determined the growth pattern of Mab in co-culture with other CF pathogens such as PA and SA in MH broth media with modifications from Birmes et al. (2017) and Rodríguez-Sevilla et al. (2018b). For the direct co-culture method, we diluted overnight cultures of Mab to 5 × 10⁵ or 5 × 10⁷ CFU/ml separately and PA and SA to 10³ CFU/mL respectively, in 10 mL of MH media and incubated them at 37°C and 200 rpm in a shaker incubator for 120 h. The different inoculum sizes controlled for differences in growth rates. The monocultures of Mab, PA and SA served as controls. Every 24 h, aliquots were serially diluted and plated to enumerate viable CFU. For the indirect co-culture method, we grew Mab (5 × 10⁵ or 5 × 10⁷ CFU/ml) in supernatants of PA and SA diluted with MH media (1:1). As a control, Mab was also grown in MH media diluted with 1xPBS (1:1). All growth curve experiments were continued for 120 h.

We assessed the biofilm forming capability of Mab in the presence and/or absence of PA and SA as described by Rodríguez-Sevilla et al. (2018a, 2018b) with modifications. For direct contact, overnight cultures of PA and SA grown in MH media were diluted to a final concentration of 10³ CFU/mL. Mab overnight cultures were diluted to the final concentration of 5 × 10⁵ CFU/ml in the bacterial suspension. 200 μL of this bacterial suspension (mixture of Mab and other bacteria) was added in triplicate to a 96 well plate and covered with a permeable membrane and incubated at 37°C for 72 h without shaking. Single species inoculated in separate wells served as controls and wells with only MH media served as sterility controls.

For indirect contact, the diluted supernatants of PA and SA prepared in MH media were used to adjust Mab to a concentration of 5 × 10⁵ CFU/ml in 96 well plates same as above. In each experiment the initial count of Mab was determined by diluting and plating the initial cultures before incubation.

We obtained RAW264.7 murine macrophages from the American Type Culture Collection (ATCC) and cultivated them using in Dulbecco’s modified Eagle media (DMEM, Corning) media supplemented with 10% Fetal bovine serum (FBS), 1% penicillin/streptomycin in a humidified CO₂ incubator at 37°C, as recommended. Infection experiments to determine if the direct contact of bacterial organisms affect the internalization of Mab in macrophages were performed using infection assay with some modifications (Roux et al., 2016; Kim et al., 2017, 2019; Nandanwar et al., 2021). RAW264.7 cells were seeded in 24-well plates and were coinfected with Mab in presence/absence of PaATCC and SA respectively, at a multiplicity of infection (MOI) of 10 (for Mab) and 5 (for PaATCC and SA) for 2 h at 37°C with 5% CO₂ (Tian et al., 2019; Kim et al., 2022). Later, we washed the infected RAW264.7 cells with DMEM three times to eliminate extracellular bacteria. Fresh experimental media containing Amikacin 200 μg/mL, which is bactericidal for all tested organisms, was added, and the cells were incubated for an additional 2 h to kill residual attached and planktonic extracellular bacteria. After 2 h of antibiotic treatment we removed the medium containing Amikacin and washed the cells again three times with DMEM. RAW264.7 cells were lysed with 0.5% Triton-X-100 prepared in saline. The lysates were diluted in saline and plated onto relevant agar plates to count CFUs of internalized bacteria. To check internalization of Mab under indirect contact we cultured Mab in presence of supernatant from PaATCC and SA (diluted 1:1 with MH media) separately and performed the infection experiment as described above. The viability of infected macrophages compared to uninfected cells was evaluated by Trypan blue and was comparable before and after infection.

Lipid extraction from Mab after indirect co-culture was extracted using the chloroform-methanol method (Folch et al., 1957; Billman-Jacobe et al., 1999; Tatham et al., 2012) with some modifications, followed by thin-layer chromatography (TLC). Briefly, we cultured Mab in supernatant (diluted 1:1 with MH media) of PA and SA separately at 37°C in a shaking incubator for 120 h and used these cultures for GPL isolation. After incubation, the bacterial suspension was centrifuged at 4,700 × g for 10 min, the pellet was washed with 1xPBS and centrifuged again at the same speed. Equal amounts of the pellets (wet weight) for each sample were then dissolved in 2:1 chloroform/methanol (20 μL/mg of wet weight) and kept on a rocking shaker overnight at room temperature. The following day, we briefly sonicated the suspension four times each for 15 s and then centrifuged at 18,400 × g for 10 min to remove insoluble material. The chloroform/methanol supernatant was placed in a fresh tube and 0.2 volume of 0.9% NaCl was added and mixed vigorously by vortexing. Later, we centrifuged the mixture at 1,150 × g for 5 min. The organic phase containing GPL was collected and dried to evaporate any traces of alcohol. The dried lipid pellet was redissolved in chloroform/methanol (2:1), and 15 μL (300 μg) of each sample was loaded onto aluminum-backed silica gel TLC plate and irrigated with chloroform/methanol (100:7). The TLC plate was removed, dried, and sprayed with 0.1% resorcinol in 40% sulfuric acid, and glycolipids were detected by charring the plate at 140°C.

Since PA isolates are known to produce a secondary metabolite pyocyanin with antimicrobial properties (Gonçalves and Vasconcelos, 2021), we checked pyocyanin production in the isolates used in this study by extracting pyocyanin as described previously (Essar et al., 1990; Schmidt et al., 2023). Briefly, we cultured PA isolates in 5 mL MH media for at least 24 h and then centrifuged the bacterial suspension to pellet the bacteria. To the clear supernatant we added chloroform at 50% vol/vol. Samples were then vortexed vigorously and centrifuged at 6,000 × g for 5 min to separate the organic phase (lower layer containing chloroform) that was removed and transferred to a fresh tube and then 20% the volume of 0.1 N HCl was added and mixed vigorously. We then centrifuged the samples again at 8,000 rpm for 5 min. After centrifugation, the aqueous fraction (upper layer) was aliquoted in a 96-well plate and absorbance was measured at 520 nm. Pyocyanin concentration as μg/ml, was obtained by multiplying the OD520 nm by 17.072, as described previously (Essar et al., 1990; Schmidt et al., 2023).

P. aeruginosa produces different exoproducts that demonstrate antimicrobial and antibiofilm properties against other pathogens during co-infection, one such exoproduct is pyocyanin (Ozdal, 2019; Gonçalves and Vasconcelos, 2021). Hence, we checked the growth pattern and biofilm forming capabilities of Mab in presence of pyocyanin (pure pyocyanin procured from Sigma-Aldrich). We cultured Mab in MH media supplemented with different concentrations of pyocyanin and compared it with the growth of Mab in MH media supplemented with equivalent volume of Dimethyl sulfoxide (DMSO) as pyocyanin was dissolved in DMSO, serving as a control. We also determined biofilm formation by Mab in presence of pyocyanin followed by CV staining as described in above sections.

We performed statistical analyses using Prism Software (GraphPad version 9.0.0), and a p-value less than or equal to 0.05 was considered statistically significant. Normal distribution of the data was assessed using the Shapiro–Wilk test. Statistical significance was assessed by one way ANOVA-test for comparison of more than two conditions in an experiment. Independent Student’s t-test was used for comparison between two conditions.

Both PaATCC and SA significantly inhibited Mab growth in co-culture (direct contact) compared to the Mab monoculture (Figure 1A). PaATCC most strongly inhibited Mab growth, and MRSA was more inhibitory compared to MSSA in co-culture (Figure 1A). To account for different growth rates among these organisms we also tested higher inoculum of Mab during co-culture. Even at the higher Mab starting inoculum (5 × 10⁷ CFU/ml), PaATCC and SA significantly inhibited Mab growth compared to Mab monoculture (Figure 1B).

Conversely, PaATCC growth was not affected significantly by either Mab or SA (Figure 1C). PaATCC also inhibited the growth of SA, although more strongly and significantly for MSSA than MRSA (Figures 1D,E). Lastly, Mab did not significantly affect the growth of and SA (Figures 1D,E). Collectively, the presence of PaATCC and SA inhibited Mab growth, with a predominant inhibition observed in the presence of PaATCC. Also, during indirect co-culture, PaATCC supernatant significantly inhibited Mab growth more predominantly than SA supernatant compared to the control (Figures 2A,B). In summary, PaATCC and SA inhibited Mab directly and indirectly. But SA inhibited Mab to a lesser extent than PaATCC in direct co-culture and indirect interaction. Neither SA nor Mab affected PaATCC growth significantly during direct co-culture.

Since PaATCC significantly affected Mab growth, even during indirect contact, we also checked if the same is observed for strains of other sub. species of Mab. PaATCC also inhibited the growth of Mms and Mbl1518 strains, indicating that PaATCC has an inhibitory effect on growth of other M. abscessus sub. species (Figures 3A–C). The heat-inactivation of supernatant of PaATCC, however, did not inhibit the growth of Mab strains as strongly and significantly as the normal supernatant (supernatant without heat inactivation), although it did affect the growth of Mab strains to certain extent (Figures 3A–C). Hence, certain heat labile metabolites are produced by PaATCC that inhibit Mab strains growth compared to MH media control (Figures 3A–C) and heat inactivation reduces the inhibitory effect of PaATCC supernatant by affecting heat liable metabolites present in the supernatant.

We found that the presence of PaATCC and SA in co-culture significantly inhibited biofilm forming capabilities of Mab as there was significant reduction in the Mab biofilm biomass after 72 h of incubation in co-culture compared to the Mab monoculture biofilm biomass (Figure 4A). On the other hand, PaATCC biofilm forming capabilities did not change during co-culture with Mab or SA when compared to PaATCC monoculture biofilm after 72 h (Figure 4B). Apart from Mab, PaATCC also significantly inhibited MRSA and MSSA biofilm in co-culture compared to monoculture, while being unaffected by the presence of Mab in co-culture after 72 h of incubation (Figures 4C,D).

As observed for the direct contact, the biofilm biomass of Mab was significantly reduced after 72 h of incubation in presence of PaATCC supernatant compared to the MH media control (Figure 5A) indicating that even the indirect interactions of PaATCC can inhibit Mab biofilm. Mab biofilm biomass was also reduced in SA supernatant but was not significant compared to the MH media control (Figure 5A). However, CV staining after 72 h of incubation demonstrated significant inhibition of Mab biofilm in supernatant of PaATCC and SA compared to the MH media control (Figure 5B).

Besides using diluted MH media (diluted 1:1 with 1xPBS) as our control, we also determined biofilm forming capabilities of Mab in supernatant prepared by culturing Mab in MH media as a control to match the conditions of using spent media (supernatant) for our indirect contact approach. We observed comparable biofilm formation by Mab in MH media control and Mab supernatant (Figures 5A,B; Supplementary Figure S1). This indicates that the reduction of Mab biofilm is due to the antibiofilm properties of extracellular metabolites produced by PaATCC and SA that inhibit biofilm forming capabilities of Mab and not due to the depleted nutrients in the supernatant.

In summary, PaATCC and SA directly and indirectly reduce the biofilm formation capabilities of Mab.

We observed that growth pattern of Mab was not affected significantly when grown in MH media supplemented with pyocyanin (Figure 5C). However, the biofilm formation capabilities of Mab were significantly reduced in the presence of pyocyanin compared to the media control (Figure 5D). We used pyocyanin because it is one of the few commercially available metabolites produced by PA isolates with known antimicrobial properties and even inhibits biofilm formation by other pathogenic microbes during coinfection (Hauser et al., 2011; Korgaonkar et al., 2013). We observed differential expression of pyocyanin by PA isolates used in this study (Supplementary Figure S2A). We also found that PaATCC produced a higher amount of pyocyanin when co-cultured with Mab compared to PaATCC monoculture (Supplementary Figure S2B).

We observed that the presence of PaATCC during co-infection resulted in higher internalization of Mab in RAW264.7 macrophages compared to Mab infection alone (Figure 6A). However, no significant difference was observed during co-infection with SA compared to Mab-only infection in macrophages (Figure 6A). Mab on contrary, did not affect the internalization capabilities of PaATCC and SA during co-infection compared to experiments without Mab (Figure 6B). As observed for the direct contact during co-infection, Mab grown in the presence of PaATCC supernatant (indirect contact) also showed higher internalization in RAW264.7 macrophages compared to the controls (Mab grown in diluted MH media and in Mab supernatant). In contrast, Mab grown in the presence of SA supernatant did not affect the internalization of Mab in macrophages (Figure 6C).

In summary the PaATCC directly and indirectly enhanced internalization of Mab in RAW264.7 macrophages, while SA did not affect the internalization of Mab in RAW264.7 macrophages either directly or indirectly.

Since the presence of PaATCC significantly enhanced the internalization capabilities of Mab in RAW264.7 macrophages, we also performed GPL isolation to check if PaATCC alters GPL expression in Mab during indirect contact using supernatant, as the lower expression of GPL by Mab is associated with higher internalization of Mab in macrophages (Nandanwar et al., 2021). Indeed, thin layer chromatography (TLC) analysis demonstrated lower GPL expression in Mab grown in PaATCC supernatant compared to the controls (Mab grown in MH media and Mab supernatant) as shown in Figure 6D. However, GPL expression in Mab did not alter when grown in SA supernatant (Figure 6D). Since PaATCC supernatant altered GPL expression, we also checked Mab GPL expression after exposure to heat-inactivated PaATCC supernatant and observed that the GPL production was normalized in Mab grown in heat-inactivated PaATCC supernatant (Figure 6E). Therefore, we can hypothesize that heat labile extracellular compounds are produced by PaATCC strain that alter GPL expression in Mab. We were unable to isolate GPL from Mab grown in co-culture with PaATCC and SA; hence GPL determination was possible only in supernatant experiments.

Since we found Mab growth, biofilm formation, and GPL expression to be significantly altered by PaATCC during direct and indirect interaction, we also assessed if other P. aeruginosa isolates (P. aeruginosa reference strain MPAO1 and two P. aeruginosa clinical isolates, CPA53 and CPA87 respectively) also affected Mab characteristics. The direct co-culture of Mab with other isolates of P. aeruginosa also significantly inhibited Mab growth (Figure 7A). However, during indirect contact, Mab growth was inhibited more in the supernatant of MPAO1 compared to CPA53 and CPA87. Although it did not affect the growth as strongly as observed for PaATCC (used as a control here) as shown in Figure 7B. During both direct and indirect contact of Mab with MPAO1, CPA53, and CPA87, Mab biofilm forming capability was significantly inhibited (Figures 7C,D) as determined by CFU counts of biofilm biomass and CV staining. Interestingly, Mab GPL expression did not change during growth in supernatants of these isolates compared to the control (Mab supernatant) as shown in Figure 7E.

Since the direct and indirect contact of PaATCC with Mab significantly affected Mab growth pattern, biofilm formation we also evaluated if pre-exposure of Mab to PaATCC (using supernatant) durably changes Mab phenotypic characteristics. We grew Mab in presence of PaATCC supernatant, diluted with MH media to different concentrations (50, 25, and 10% respectively). After incubation, we centrifuged and washed Mab with 1xPBS before setting up experiments to check the growth pattern, biofilm formation and GPL expression. We observed that Mab pre-exposed to PaATCC supernatant and then grown in normal MH media, regains its normal growth pattern (Figure 8A) with no significant difference compared to the media control. Moreover, Mab also retains its biofilm formation capabilities and GPL expression once it is not under the stress of PaATCC supernatant (Figures 8B,C). These observations indicate that PaATCC alters Mab phenotypic characteristics in a reversible manner.

Similar observations were also observed after pre-exposure of Mab to supernatant of MPAO1, CPA53, and CPA87, wherein, Mab regained its normal growth pattern, biofilm formation and GPL expression once the stress of supernatant is replaced by normal MH media (Supplementary Figures S3A–C).