Mycobacterium ulcerans challenge strain selection for a Buruli ulcer controlled human infection model

Abstract

Critical scientific questions remain regarding infection with Mycobacterium ulcerans, the organism responsible for the neglected tropical disease, Buruli ulcer (BU). A controlled human infection model has the potential to accelerate our knowledge of the immunological correlates of disease, to test prophylactic interventions and novel therapeutics. Here we present microbiological evidence supporting M. ulcerans JKD8049 as a suitable human challenge strain. This non-genetically modified Australian isolate is susceptible to clinically relevant antibiotics, can be cultured in animal-free and surfactant-free media, can be enumerated for precise dosing, and has stable viability following cryopreservation. Infectious challenge of humans with JKD8049 is anticipated to imitate natural infection, as M. ulcerans JKD8049 is genetically stable following in vitro passage and produces the key virulence factor, mycolactone. Also reported are considerations for the manufacture, storage, and administration of M. ulcerans JKD8049 for controlled human infection.

Fig 1. Experimental approaches for evaluating candidate M. ulcerans isolates for use in a CHIM.

MIC: minimum inhibitory concentration. Created with BioRender.com.

Fig 2. Procedure for diluting cryopreserved cell banks of M. ulcerans in phosphate-buffered saline.

200 μL of red dye (bottom syringe, representing a thawed cell bank) and 800 μL of phosphate-buffered saline (top syringe) are connected using a luer lock connector and mixed by repetitive plunging. Four mixing steps are shown. This is performed a total of 15 times to mix samples thoroughly, creating a homogeneous dilution. Arrows represent the direction of plunging to mix the sample seen in the image.

Fig 3. Phylogenomic tree of the M. ulcerans/M. marinum complex illustrating the relationships among the isolates examined in this study.

The tree, constructed via maximum likelihood based on whole genome SNPs, is rooted with M. marinum as an outgroup. The three major lineages within the mycolactone producing mycobacteria complex are highlighted. The horizontal bar provides a scale to represent the number of core SNPs defining branch lengths.

Fig 4. Growth characteristics of M. ulcerans in SMVT.

(A) JKD8049 cultured for 12 weeks in SMVT, photographed after brief agitation; the growth is in large, dense clumps that rapidly settle with gravity, leaving a background with minimal turbidity, and visible biofilm; (B) 100 μL spread plate of M. ulcerans NM20/02 direct from stationary culture in SMVT after 10 weeks of incubation at 30°C; (C) filtrate of NM20/02 shows evenly distributed colonies of approximately similar size and shape in the filtrate; nevertheless, loss of CFU is demonstrated. (D) Enumeration of isolates from stationary cultures in SMVT media and filtered to remove clumps. The upper border of each isolate represents the mean CFU/mL obtained in the filtrate. Enumeration was poor from all isolates except NM20/02 after 2 weeks. Recovery of cells from the filtrate reduced over time, despite visible, clumpy growth in the media.

Fig 5. M. ulcerans JKD8049 yield in SMVT shaking incubation over time.

JKD8049 was cultured in SMVT shaking cultures with 20 to 25 glass beads (3 mm diameter), and samples for CFU enumeration were drawn from the centre of the shaking culture, after pausing shaking for 15–20 minutes. (A) represents the average of six biological replicates, except for time 0, which was established from 3 replicates after 48 hours of shaking incubation (S2 Table). Mean and standard deviation are shown. Comparison of means was performed by Student’s t test, * represents p < 0.05, when comparing 2 weeks’ incubation with subsequent timepoints. Image (B) illustrates visible turbidity, in addition to surface pellicle growth and biofilm formation, in these conditions. The initial inoculated biomass is seen on the far left (note hydrophobic floating material), followed by the disrupted material 48 hours later, with subsequent turbid growth.

Fig 6. CFU/mL of various M. ulcerans isolates after shaking incubation and filtration.

The upper border of each isolate represents the mean CFU/mL obtained; error bars represent standard deviation. Pre-filtration (left columns) and post-filtration (right columns) for each strain demonstrate a loss of CFUs from 35 to 85%. On the far right, the closely-spaced replicates (round spots) illustrate the improved accuracy of CFUs obtained after filtration (JKD8049 shown in purple).

Fig 7. CFU/mL of various M. ulcerans isolates after filtration and cryopreservation in glycerol at -80°C (accounting for dilution factor) compared to CFU/ml expected from original filtrate (F) sample.

CFU/mL were counted immediately after thawing (0 hours), and at two additional timepoints. The upper border represents the mean CFU/mL obtained; error bars represent standard deviation. Viability loss (%), is shown for JKD8049, others are available in S4 Table.

Fig 8. (A) JKD8049 cryopreserved with and without supplement.

( Filtered JKD8049 was stored in either glycerol with ‘Veggietones’ supplement (left columns) or glycerol without supplement (right columns). Samples were tested immediately after the sample was thawed on ice (‘0 hours’) or at subsequent timepoints (‘2 hours’ and ‘4 hours’). The upper border of each isolate represents the mean CFU/mL obtained; error bars represent standard deviation. (B) CFU/mL obtained before and after dilution using needle and syringe system. There was no significant (ns) difference in CFU before (left) or after (right) dilution in PBS and injection through 30G needle and LDS syringe (mean and standard deviation are shown, comparison of means was performed using Student’s t test).

Fig 9. CFU counts obtained from five vials of cryopreserved M. ulcerans JKD8049 from a cell bank of 46 cryovials.

There was no significant difference in CFU count across all five vials tested (presented here as CFU count in each technical replicate); mean and standard deviation are shown (one-way ANOVA, p = 0.13).

Fig 10. Comparisons of JKD8049 in various culture conditions.

Image (A) shows 5 μL spot plates of JKD8049 on 7H10/OADC agar compared to (B) SMVT agar. Image (C) shows 100 μL spread plates of M. ulcerans JKD8049 spread onto SMVT agar, compared to (D) Sauton’s minimal media agar. Although triplicates were produced of each agar, only one colony was visible from all three Sauton’s media agar plates, whereas colonies on all three SMVT agar replicates demonstrated spreading growth. Images A-D were captured after 12 weeks of incubation at 30°C. (E) M. ulcerans JKD8049 cultured at 37°C (top) and 30°C (bottom) confirmed thermal restriction of this isolate. (F) Shown are mean and standard deviation of biological triplicate experiments of M. ulcerans JKD8049 cultured in both aerobic and microaerophilic conditions.

Fig 11. ELISA measurement of IL-8, IL-1b, IL-6 and TNFα secreted by human macrophage in response to 0.4 mL of M.

ulcerans JKD8049 culture sample, before and after filtration. SMVT media only was analysed in addition to unstimulated controls.