Manno-oligosaccharides (MOS) were produced by the enzymatic hydrolysis of copra meal, a mannan-rich byproduct from coconut processing, as previously described by this research group (Di Pede et al., 2024). Two distinct MOS products were produced: low-molecular-weight MOS (LMW-MOS) in an offsite manufacturing plant and high molecular weight (HMW-MOS) in a 4 L batch process in-house. The HMW-MOS was produced using a shorter reaction time, resulting in less hydrolysis of mannan and a product with a greater proportion of oligomers in the higher-weight fractions. The reaction conditions for LMW-MOS resulted in a greater proportion of oligomers in the shorter DP fractions. The process for LMW-MOS included a spray-drying step, so LMW-MOS was dissolved in MilliQ water for all experiments and analyses. The concentrations of LMW-MOS and HMW-MOS were matched for all growth assays on a dry weight basis by assessing % solids (w/v) of LMW-MOS and HMW-MOS using an HE53 Moisture Analyzer (Mettler Toledo) and diluting liquid HMW-MOS to the equivalent solids % of LMW-MOS for each target concentration. A single batch of product was used for all experiments using LMW-MOS and HMW-MOS.

The oligomer and monomer composition of both LMW-MOS and HMW-MOS was determined using high-performance liquid chromatography (HPLC) and liquid chromatography–mass spectrometry (LCMS). The composition of mannose, glucose, galactose, xylose, arabinose, and cellobiose of both LMW-MOS and HMW-MOS was measured by HPLC using an Aminex HPX-87H column (Bio-Rad) with a column temperature of 80°C, a refractive index detector (55°C), a mobile phase of water with a flow rate of 0.6 mL/min, and an injection volume of 20 μL. Monomer and cellobiose standards were purchased from Sigma-Aldrich. To quantify the proportion of LMW-MOS and HMW-MOS in the oligomer fraction, liquid characterization was performed according to methods described elsewhere (Sluiter et al., 2008). Briefly, samples of LMW-MOS and HMW-MOS were hydrolyzed with 70% sulfuric acid (Sigma-Aldrich) and autoclaved at 121°C for 1 h to break down all carbohydrate polymers and oligomers into monomers. Hydrolyzed samples were analyzed using HPLC, and the total oligomers were quantified as the difference in monomers between the hydrolyzed and non-hydrolyzed samples, adjusted for differences in molecular weight due to the addition of water during hydrolysis.

The composition of mannose oligomers of DP 2-6 (mannobiose, mannotriose, mannotetraose, and mannohexaose) in both MOS types was determined using LCMS due to overlapping peaks and inconsistent separation of oligomers on our HPLC system. A UHPLC–MS system was used with an Aminex HPX-87H column (Bio-Rad) with a mobile phase of 0.1% formic acid and a heated electrospray ionization probe (HESI-II). Oligomer standards (mannobiose, mannotriose, mannotetraose, and mannohexaose) were purchased from Megazyme. Quantification of both monomers and oligomers was performed using a series of standards at various concentrations to create a standard curve. The % solids of both LMW-MOS and HMW-MOS were determined using an HE53 Moisture Analyzer (Mettler Toledo) to quantify monomers and oligomers on a % dry matter (DM) basis.

Bacterial strains Klebsiella pneumoniae strain 1101160, Listeria monocytogenes strain 1071/53, and Streptococcus mutans strain NCTC 10449 were supplied by ATCC (Cedarlane). Strain E. coli K12 MG 1655 was obtained from the Mahadevan lab (BioZone, University of Toronto). LB Lennox media (BioShop) was used for growing liquid cultures of E. coli, while K. pneumoniae was grown in TSB media (MilliporeSigma), and L. monocytogenes and S. mutans were cultured with BHI media (MP Biomedicals). All media was purchased as pre-formulated, dehydrated media prepared according to the manufacturer’s instructions. All media was heat sterilized with autoclaving before it was used in culturing experiments. For culturing on agar plates, liquid media was formulated with 2% agar, autoclaved, and then poured aseptically into plates. All bacterial strains were maintained on 30% glycerol stocks at −80°C. Cultures were revived by inoculating 1 μL of frozen bacterial stock into 3–5 mL of autoclaved media and incubating at 37°C overnight under shaking, aerobic conditions. To obtain isolates, liquid cultures were streaked on agar plates of their corresponding media and incubated at 24-48 h at 37° C. All plates were stored at 4° C until being used in growth studies.

Stock solutions of LMW-MOS and HMW-MOS for all bacteria assays were prepared at equal concentrations as described in section 2.2. For the growth inhibition and antibiotic potentiation assays, 100 mg/mL stocks of LMW-MOS and HMW-MOS were prepared, and 100 mg/mL stock of mannose was prepared by dissolving dry powder in MilliQ water. For the growth inhibition assays using individual MOS oligomers, 50 mg/mL solutions of mannobiose, mannotriose, and a 1:1 combination of mannotetraose and mannotriose were prepared by dissolving dry powders in MilliQ water. All substrates were filter-sterilized using a 0.22-μm syringe filter before being used in culturing experiments.

The broth microdilution method is a technique used to determine the minimum inhibitory concentration (MIC), the lowest concentration of a compound that inhibits the visible growth of bacteria under controlled conditions (Wiegand et al., 2008). While MICs have been reported for various types of oligosaccharides (Asadpoor et al., 2021b), complete inhibition is not commonly observed with sugar substrates. This is likely due to the ability of some bacteria to utilize them as a carbon source and the fact that their method of inhibiting bacterial growth tends to be bacteriostatic rather than bactericidal (Mizzi et al., 2020). Nonetheless, the broth microdilution method is an effective tool for screening the inhibitory ability of oligosaccharides at various concentrations.

Aseptic conditions were maintained for all culturing experiments by working next to a Bunsen burner. Working cultures of bacteria were obtained by inoculating one isolate colony from agar stocks using a flame-sterilized inoculating loop into 3–5 mL of prepared liquid media. The cultures were incubated at 37°C overnight under shaking, aerobic conditions. Each culture was diluted to OD 0.01 with fresh media.

A sample well plate design is included in the Supplementary material. Within a 96-well polystyrene well plate (Falcon), 90 μL of sterile media was added to each well, and then 90 μL of the 100 mg/mL stock solution of carbon substrate (LMW-MOS, HMW-MOS, or mannose) was added to the first column and serially diluted across each row using a multi-channel pipette, following the previously outlined protocol (Wiegand et al., 2008). To each well, 10 μL of bacterial culture was added to reach a final concentration of roughly 1.0 × 10⁵ CFU/mL. The plate was then incubated for 18–24 h under aerobic, shaking conditions and bacterial growth was measured continuously at OD600 using a microplate reader (Tecan Infinite 200 Pro). A blank reading of the plate was taken before adding bacterial cultures, and the blank reading was subtracted from each subsequent reading for the growth analysis. The media alone was used as the growth control, and the media alone without cultures added was used as a sterility control.

Antibiotics with a range of mechanisms of action against bacteria were selected for the antibiotic sensitization studies, using methods previously described (Craft et al., 2018b). Antibiotics tested included cell wall synthesis-inhibiting penicillin G (Cayman Chemical) and ceftazidime hydrate (Cayman Chemical), and protein synthesis-inhibiting clindamycin (Cayman Chemical) and gentamicin sulfate (AdipoGen). Stock solutions of antibiotics were prepared in a two-step dilution: penicillin and ceftazidime were prepared by dissolving powder in PBS and then diluting with PBS to the target concentration; clindamycin was prepared by dissolving powder in 90% ethanol and then diluting with PBS to the target concentration; and gentamicin was prepared by dissolving powder in water and diluting with water to the target concentration. All antibiotics were filter-sterilized before use in bacterial assays.

Each antibiotic was first screened against all four pathogen types to determine the minimum inhibitory concentration (MIC) using the broth microdilution method as described above, with some modifications. Working cultures of bacteria were obtained by inoculating one isolate colony from the agar plate aseptically into 3–5 mL of prepared liquid media. The cultures were incubated overnight at 37°C under shaking, aerobic conditions, and then diluted to OD₆₀₀ = 0.1 with fresh media the following morning. Antibiotics were prepared at twice the highest target concentration and then serially diluted within the well plate. Each well was inoculated with 10 μL of bacterial culture. The plates were then sealed with breathable plate seals and incubated aerobically at 37°C under shaking conditions (200 RPM). After 24 h, bacterial growth was measured as OD₆₀₀ using a microplate reader (Tecan Infinite 200 Pro). The MIC was defined as the lowest concentration at which no growth was observed, as measured by turbidity at OD₆₀₀.

Antibiotics with trends toward potentiation by MOS were used for further assays to measure the ability of MOS to increase protein concentration in the bacterial culture supernatant, adapted from methods described previously (Kruger, 2009; Moghimi et al., 2016). Starting cultures of all bacterial strains were grown within sterile media overnight to reach OD₆₀₀ = 1.0. Bacterial cells were washed by pelleting at 4000 g, removing the supernatant, adding sterile PBS (pH 7.2, Gibco), and then gently vortexing. The cells were washed three times with PBS to remove culturing media and then re-suspended in sterile PBS to reach a final concentration of OD₆₀₀ = 0.5. Stock solutions of MOS were prepared at 2.5x the target concentration, and antibiotics were prepared at 10x the target concentration. Within 96-well plates, 50 μL of bacterial culture was combined with 40 μL of MOS solution and 10 μL of antibiotics and then incubated under shaking conditions at 37°C for 1 h. The cultures were incubated under their respective MOS and antibiotic conditions for 1 h to assess protein concentration while limiting culture growth on the substrate. After 1 h, the plates were centrifuged in an Avanti J-E centrifuge (Beckman Coulter) at 4000 g for 5 min to pellet cells. A 10 μL aliquot of the supernatant was pipetted into a fresh well plate, and 200 μL of Bradford reagent (BioRad) was added along with standards of bovine serum albumin (BSA). After 15 min, the OD₅₉₅ was measured using a plate spectrophotometer, and protein concentration was quantified using a standard curve developed with BSA standards.

Data are reported as the mean values ± standard error (SEM) of at least three independent experiments routinely performed in triplicate. The experimental AUC of growth curves was calculated using the R package growthcurver (Sprouffske, 2020). Significant differences in the AUC or final OD₆₀₀ between groups were determined using Welch’s two-sample independent t-test, and p-values were adjusted using Bonferroni post-hoc adjustments for multiple comparisons. Statistical significance was calculated using the R package rstatix (Kassambara, 2023). Chemical structures were produced using ChemDraw (ChemDraw Professional, 2022). Figures were made using the R package ggplot2 (Wickham, 2016).

The carbohydrate composition of low- and high-molecular-weight manno-oligosaccharides (MOS) products is shown in Table 1. Low-molecular weight (LMW) MOS had a greater amount of mannobiose (14.6%) and mannose (32%) compared to HMW-MOS (3.6 and 24%, respectively) while HMW-MOS had a greater amount of oligomers in the DP 5–12 range (20%) compared to LMW-MOS (9%). Both LMW-MOS and HMW-MOS contain a small amount of proteins, lipids, and ash post-filtration.

Using the broth microdilution approach, we compared the growth of bacterial strains E. coli K12 MG 1655, K. pneumoniae 1101160, L. monocytogenes 1071/53, and S. mutans NCTC 10449 with and without LMW-MOS, HMW-MOS, and mannose, using turbidity as a measure of bacterial growth. We selected E. coli K12 as a model for type-1 fimbriated, mannose-mediated binding and K. pneumoniae, a fimbriated, highly virulent, and multidrug-resistant pathogen, in addition to two gram-positive pathogens L. monocytogenes and S. mutans.

In the first set of analyses, we compared the growth curves of E. coli K12 growing on various concentrations of LMW-MOS and HMW-MOS (Figure 2A). Notably, 50 mg/mL of LMW-MOS inhibited the growth of E. coli K12 compared to the control, while 25 mg/mL also inhibited growth but trended toward diauxic growth. Diauxic growth curves were also observed for E. coli K12 cultures growing on 50 mg/mL of HMW-MOS, while HMW-MOS at concentrations of 25 mg/mL and below had greater growth compared to the media alone. Lower concentrations of both LMW-MOS and HMW-MOS increased growth compared to the growth control, rather than inhibiting growth.

After determining that the highest concentration tested (50 mg/mL) was the most inhibitory, we assessed the area under the curve (AUC) over the growth cycle for all strains via separate experiments with each of LMW-MOS, HMW-MOS, and mannose, at 50 mg/mL (Figure 2B). Compared to the growth control, 50 mg/mL LMW-MOS resulted in significantly lower AUCs for E. coli and K. pneumoniae and trended toward lower AUCs for L. monocytogenes. In contrast, 50 mg/mL HMW-MOS showed more variability and did not significantly decrease the AUCs for E. coli K12, L. monocytogenes, and S. mutans, but resulted in greater AUCs for K. pneumoniae. Notably, mannose at a concentration of 50 mg/mL inhibited the growth of K. pneumoniae, while not significantly impacting the growth of E. coli K12 or L. monocytogenes. The full growth curves of all pathogens for all concentrations of LMW-MOS, HMW-MOS, and mannose are included in Supplementary files.

Next, we aimed to determine how growth inhibition of type-1 fimbriated bacteria differed for distinct oligomer fractions. High-purity, HPLC-grade MOS oligomer products of DP 1 (mannose), DP 2 (mannobiose), and a 50:50 mixture of DP 4 and 6 (mannotetraose and mannohexaose) were used as substrates for cultures of E. coli and K. pneumoniae at concentrations of 25 mg/mL. Mannobiose and the mixture of mannotetraose and mannohexaose inhibited the growth of E. coli (Figure 3A), while mannobiose promoted the growth of K. pneumoniae (Figure 3B). Substrates LMW-MOS, mannose, and mannotetraose/hexaose trended toward lower AUCs for K. pneumoniae cultures than the growth control, and all substrates trended toward inhibiting the growth of E. coli.

The broth microdilution method was used to determine the minimum inhibitory concentrations (MICs) of antibiotics against all pathogens before potentiation assays with MOS. The MICs are included in Table 2.

Next, we were interested in studying whether MOS can potentiate the action of antibiotics against the studied bacterial strains. A sub-MIC concentration was selected to determine what concentration of MOS, if any, would result in lower bacterial growth than a sub-MIC concentration of the antibiotic alone (Figure 4). A summary of the synergistic effects of 50 mg/mL MOS with tested antibiotics against pathogens is included in Table 2.

Against E. coli, only penicillin demonstrated trends of potentiation by MOS, while MOS promoted the growth of cultures with ceftazidime and gentamicin in a dose-dependent manner. Both 50 mg/mL and 25 mg/mL MOS potentiated clindamycin and ceftazidime against K. pneumoniae, resulting in very little final bacterial growth with clindamycin and half as much growth by 50 mg/mL MOS with ceftazidime (final OD₆₀₀ of 0.19 [SE: 0.03, 0.35] compared to 0.37 [0.34, 0.41], p.adj < 0.0001). All antibiotics against L. monocytogenes trended toward being potentiated by MOS in a dose-dependent manner, with 50 mg/mL MOS significantly potentiating ceftazidime (final OD₆₀₀ of 0.01 [−0.01, 0.01] compared to 0.33 [0.16, 0.49], p.adj < 0.0001) and clindamycin (final OD₆₀₀ of 0.04 [−0.01, 0.19] compared to 0.41 [0.01, 0.79], p.adj < 0.001). The addition of 50 mg/mL MOS potentiated the action of clindamycin against S. mutans and, to a lesser degree, ceftazidime. In some cases, however, antibiotics were rendered less effective by the addition of MOS to the media; for example, both 50 mg/mL and 25 mg/mL MOS reduced the effectiveness of gentamicin against K. pneumoniae.

The total protein concentration in the supernatant of cultures was measured after 1 h of incubation with LMW-MOS or HMW-MOS in combination with antibiotics, to explore possible mechanisms of inhibition and antibiotic potentiation (Figure 5). Significant differences in the relative increase in protein concentration were observed across cultures for both MOS types. While LMW-MOS had a greater protein content at baseline, as evidenced by the comparison of each growth control (CTRL) to the blank condition (BLANK), cultures grown with LMW-MOS had a greater amount of protein released into the supernatant than those grown with HMW-MOS, especially for gram-positive pathogens L. monocytogenes and S. mutans.

The protein concentration was also assessed in cultures incubated with both types of MOS and antibiotics (Figure 5). Cultures of E. coli incubated with penicillin and LMW-MOS had a statistically significant increase in protein concentration compared to LMW-MOS alone. For all other pathogen-antibiotic conditions, incubating bacterial cultures with MOS and antibiotics resulted in lower protein concentrations in the supernatant than incubating cultures with MOS alone.