Projects

KLA is the single most important cause of monomicrobial liver abscess and its subsequent complications in many parts of Asia, including Singapore. We examine bacterial factors and the ensuing host immune responses in the encounter between hypervirulent and its mammalian host. It is believed that many infections initiate from bacteria colonized in the intestines as they translocate across the intestinal barrier and travel via the hepatic portal vein to the liver. We are interested in examining both bacterial and host factors at play at each stage of the infection, including during the colonization stage. We routinely use a hypervirulent strain isolated from a liver abscess patient designated as a reference strain, SGH10, for our investigations. Using murine colonization and infection models, as well as primary human intestinal organoids, we hope to understand when and how the bacteria transition from a gut colonizer to a potentially deadly pathogen.

Klebsiella induced liver abscess (KLA)

Fig 1. Scanning Electron Microscope (SEM) images of SGH10 WT strain at 40,000X magnification (left) and SGH10 ΔwcaJ capsule mutant strain at 60,000X magnification (right).

Multidrug resistance plasmid transmission in Klebsiella pneumoniae

We are interested in understanding the factors driving the transmission of carbapenem-resistance and multi-drug resistance plasmids in hospitals. We track plasmid and resistance genes dominance over time in a multi-centre collaboration, then apply molecular genetics to determine plasmid and bacterial host factors favouring spread and stability of carriage. In particular, we focus on the emergence of hypervirulent and carbapenem-resistant K. pneumoniae strains and aim to define factors and conditions leading to this convergence.

Fig 2. Conjugation frequency of pKPC2 and pNDM1 among various Enterobacterales donor-recipient pairs. (A) Conjugation frequency of pKPC2 and pNDM1 from K. pneumoniae SGH10 plasmid donor strain to a panel of Enterobacterales recipient strains; and B) from the panel of Enterobacterales as plasmid donor strains to K. pneumoniae SGH10 recipient strain. Each symbol represents 1 experimental replicate with a total of 12 replicates. Data points that are not seen on the graphs indicate no detectable transconjugant. 🌐

Novel strategies to target transmission of carbapenem-resistance in Enterobacterales

We have extensive collaborations with our clinical colleagues, computational biologists and chemists to devise strategies to combat carbapenem resistant Enterobacterales. Our efforts involve evaluating the efficacy and clinical potential of innovative synthetic antibacterial compounds and materials. Furthermore, we explore combinatorial therapy guided by AI platform to identify novel drug combinations capable of effectively eradicating CRE.

Fig 3. OIM1-6's DNA binding properties and resulting DNA damage. (A) Flow cytometry analysis demonstrates picogreen displacement by OIM1-6 inside bacterial cytoplasm, indicating the DNA-binding properties of OIM1-6. Picogreen is a DNA binding dye with fluorescence level measured under Alexa Fluor 488 channel. (B) Confocal images showing the treatment of OIM1-6 leads to more foci formation in E. coli Gam-GFP strain as compared to the untreated cells. The hydrogen peroxide cells served as positive control with elevated level of foci formation. The presence of foci is indicative of double-stranded DNA breaks caused by the drug treatment. 🌐

Fig 4. Workflow of the IDentif.AI-guided drug combinatorial study against CRE. Two clinical isolates from the most representative CRE species (hvKp ENT646 and E. coli C31) are investigated by IDentif.AI. The workflow begins by selecting 12 FDA-approved drugs, and they are individually assessed via dose-response experiment in vitro. Relevant concentration levels are selected. Subsequently, 155 OACD-designed combinations are experimentally validated, and the respective data are analyzed by IDentif.AI. Top combinations selected by the platform are comprehensively analyzed. 🌐

Interaction of Burkholderia pseudomallei with host cells

B. pseudomallei is unique among bacterial pathogens in its ability to fuse host cells into multinucleated giant cells (MNGCs). Fusion is dependent on bacterial intracellular motility through the polymerization of actin into “actin comet tails” and the Type 6 Secretion System (T6SS5). We discovered that the process of cell fusion results in genomic instability and formation of micronuclei. Host cells sense danger brought on by unnatural cell fusion and this then triggers the cGAS-STING innate immune signalling pathway that ultimately leads to autophagic cell death. This could be the body’s defense against cellular transformation.

One of the routes of infection is through breaks in skin, such as those experienced by farmers with exposure to soil. We also examine the interaction of the bacteria with primary keratinocytes and the response of the host at this encounter.

Fig 5. MNGC formation in HepG2 liver epithelial cell. Bacteria (Burkholderia spp.) is shown in red, the cell periphery in green and the cell nuclei in blue.

Fig 6. B. thailandensis infected multinucleated giant cell containing highly condensed mitotic-like chromosomes. Cell rounds up, initiates cytokinesis but eventually abort the process. HepG2 cells were infected with mApple fluorescent B. thailandensis (in Red) and stained with CellMask plasma membrane dye (in Magenta) and Hoechst 33342 DNA dye (in Blue).