Nucleoside decoys - metabolic interference in plant defence

Plant disease resistance (R) genes are widely deployed in plant breeding to help mitigate global crop losses to pests and pathogens which exceed 30%. Unfortunately, resistance is often overcome in the field as pathogens evolve ever sophisticated methods of deploying multi-functional "effectors" - key elements of the pathogens armoury the work collectively to avoid detection and suppress host immunity.
Despite having cloned R proteins more than 25 years ago, until 12 months ago we had little idea how these functioned. R proteins come in two flavours, TNLs and CNLs, both containing common key functional domains, the central nucleotide binding (N) domain and carboxyl terminal "leucine rich repeat" (L) region.
The recent major research breakthrough showed that the amino terminal TIR, (Toll Interleukin 1) domain of "T"NL disease resistance proteins dimerises to generate a complex capable of cleaving NADH or NADPH, key energy sources for cells. Critically, this "NADase" activity was essential to activate disease resistance. Notably, although animal and bacteria TIR domains have similar enzymatic activities, the products appear to differ. Plants and bacteria produce a compound called v-cADPR (variant cyclic ADP Ribose).
For a number of years we have studied the metabolic transition from defence to disease, specifically looking at how the bacterial plant pathogen Pseudomonas syringae overcomes host defences. The most exciting finding of comprehensive untargeted metabolite profiling of infected tissue was the identification of a totally novel molecule that accumulated rapidly in leaves that were infected with Pseudomonas that was going to cause disease but not a non-disease causing mutant. Remarkably, we have recently confirmed that this molecule, which we call "540" based on its molecular mass, is of identical molecular mass to v-cADPR formed by activated TNL disease resistance proteins. Critically, it has a different retention time, as also does the bacterial TIR produced v-cADPR suggesting some very subtle structural variations probably confer quite different specificities. To date, it is unclear whether either the TNL produced plant or bacterial v-cADPR have any biological activity.
Importantly, we had previously published two key pieces of evidence.
First, disease causing bacteria, but not disarmed bacteria induce a very specific locus of 6 truncated TNL genes (tTNs). This is at first counterintuitive. Why induce R genes?
Secondly, we know disease causing bacteria rapidly suppress a defense response - called a reactive oxygen burst - in the chloroplast. This is necessary for disease progression. A consequence of this is elevated NADP+ - an NADase substrate. These are really rapid events, occurring before TNL proteins are activated.
Putting this evidence together; the rapid accumulation of 540 co-incident with suppression of the oxidative burst and induction of the tTNs, we theorise that bacterial effectors both suppress the oxidative burst and simultaneously induce the tTNs to mop up accumulating NADP+, which would otherwise activate TNLs. We also cannot rule out the tTNs can also bind to, and interfere with, functional TNL TIR domains preventing activation.

Our previous work characterised 540 as a highly labile "cyclic phosphoriboside". In collaboration with an Australian group we elucidated the structure of a prokaryotic v-cADPR, which is subtly different from 540. We have recently developed a method to stabilize 540 and will determine the NMR structure and collaborate on the plant v-cADPR structure with US researchers.
Concomitantly, our work programme will fully characterise the tTNs using genetics, gene-editing, biochemistry and structural approaches. Finally we will determine the dynamics of NAD/P accumulation/loss during disease and defence development using state-of-the-art genetically encoded reporters. This multidisciplinary project benefits from collaborations in Australia, Hong Kong and the USA.

TNL resistance proteins use proximity induced dimerization to form functional NADases. NADase activity is essential to trigger the plant HR and leads to production of 'variant cyclic ADP ribose" (v-cADPR). Bacterial TIRs similarly produce v-cADPR, though with a different HPLC retention time. By untargeted metabolite profiling of Arabidopsis leaves infected with virulent Pseudomonas, we identified another v-cADPR, "540", with accurate mass identical to the plant and bacterial v-cADPR but with different retention time. 540 is also induced in leaves of tomato and Nicotiana benthamiana following virulent pathogen challenge. This implies subtle structural variations may confer different biological function.

Our data is indicative of a complex immune suppression mechanism, potentially operational against the whole class of TNLs. It involves 540 and a cluster of 6 truncated TNLs (tTNs) induced 4hpi by Pseudomonas, identified from our infection transcriptome dataset. The tTN promoters contain highly conserved motifs, one conferring effector specific induction. We hypothesise effectors induce tTNs to convert NADP+, which accumulates at Photosystem I following suppression of chloroplastic ROS ~4hpi, into 540. This sequestration likely precedes activation of functional TNLs.

We will determine the structure of 540 via NMR (as we did for bacterial v-cADPR), and address 540s biological significance - is it a "dead-end" product? In collaboration with leading US groups we will compare this to TNL produced v-cADPR.
In parallel we will comprehensively characterise the tTNs using genetics, biochemistry and structural approaches. We will collaborate with the Kobe lab in Brisbane to solve tTN TIR structures and use CRISPR to delete the entire tTN locus. We will also investigate whether tTNs additionally bind to, and interfere with, TNLs.
Finally we explore the metabolic changes during both disease and defense using genetically encoded reporter for both NADH and NADPH.