Research

I synthesise nanoparticles using plant extracts and bacteria rather than harsh chemical routes, then characterise them and test what they actually do to the pathogen. Chitosan formulated with Moringa oleifera leaf extract by ionic gelation gives semi-spherical particles around 46 nm; the same chemistry run with mangrove-derived Bacillus velezensis gives 96 nm particles with a different activity profile. Copper, zinc oxide and selenium have followed the same pipeline against different pathosystems.

The interesting question is not whether a nanoparticle kills a fungus, which most do at some concentration, but by what route and at what cost. Transcriptomics on treated Magnaporthe oryzae pointed to disruption of cell wall biosynthesis and membrane transport, and deletion mutants for the implicated chitin synthase and major facilitator superfamily genes were markedly more sensitive to treatment, which turned a correlation into a mechanism. Metabolomic profiling showed where fungal metabolism collapses. Electron microscopy showed hyphal shrinkage and intracellular leakage consistent with both.

Alongside the nanomaterials, I work on the organisms themselves and the compounds they produce. Mangrove rhizosphere soil has been a productive source: Pseudomonas aeruginosa R64 from the Mai Po Nature Reserve produces a fermentation extract that suppresses mycelial growth, conidiation, germination and appressorium formation in Pyricularia oryzae, hitting four stages of the infection cycle rather than one. Bacillus velezensis J69 lipopeptides do comparable work against strawberry anthracnose.

Genetics is how these move from promising isolate to usable agent. Genome mining across the Paenibacillus polymyxa complex identified the biosynthetic clusters behind biocontrol and growth promotion in nodule-inhabiting strains. Screening 997 transposon mutants of Burkholderia seminalis traced production of the antifungal cyclic dipeptide cyclo(Phe-Pro) to the Dct C4-dicarboxylate transport system, which tells you what to feed the organism if you want more of the compound.

A treatment that clears the pathogen and sterilises everything around it has not solved the problem, it has moved it. So every intervention I test is assessed against the resident community as well as the target. Metagenomic sequencing of treated rice showed Magnaporthe abundance falling by 93.5% while overall microbial diversity rose in both compartments, with Thiobacillus, Nitrospira, Nocardioides and Sphingomicrobium enriched in the rhizosphere and Azonexus, Agarivorans and Bradyrhizobium in the phyllosphere.

The same lens applied to systems without any treatment at all is revealing. Comparing ancient and cultivated bayberry showed that the older trees carry tissue-specific microbial and metabolite signatures associated with their higher stress resistance, which suggests the community is doing work that breeding has partly discarded.

Losses after harvest are as costly as losses in the field, and the constraints are tighter, since anything applied to fruit is going to be eaten. Work on strawberry anthracnose therefore measured pathogen suppression, fruit quality, water loss and acute ecotoxicity in the same experiment rather than treating safety as a separate follow-up study. Treated fruit lost 1.8% of its weight against 8.5% in the pathogen control, and Artemia salina survival was 66.6% against 28% for azoxystrobin at the same concentration.

This line began early. My first publications were first reports of postharvest pathogens in Pakistan, Geotrichum candidum causing sour rot in loquat and Alternaria mali causing apple core rot, identified by morphology and ITS sequencing. More recently the same interest in what microbes do to stored food has extended to fermentation, where selected lactic acid bacteria starters lowered biogenic amines in a traditional rice-fish product while improving its flavour chemistry.