 you. Thanks very much for the invitation to present today. It's very nice to be here and very excited to show you what we've been doing part of the research in our lab. I just thought I'd start off by introducing our lab. We call ourselves the Sustainable Food Production Research Group at Griffith Union. We're a member of the Centre for Planetary Health and Food Security. There's a link there. You can go and find out more about the centre there if you'd like to. I'm a member of a group that also includes another presenter today, so Ido Bar. He'll be telling you about another one of our research programs, but we've got a few other research fellows and a number of PhD students that are currently enrolled with us. The sorts of programs we generally work on are today. I'll be talking to you about our biosensor for development for diagnostics of plant pathogens, but we also do work on looking at the evolution and adaptation of fungal species and again, you know, there's going to talk about one of those today. We lead the National Puppeye Breeding Programme, which has a lot of work in it to do with breeding for better quality and flavour, but also a bit of pathology in it, too, that we collaborate with the Queensland Department of Ag. Ido is also leading another program on flavour and sensory profile genomics work in Puppeye. So we've got a myriad of partners and investors that we work with there, but today I'm going to talk to you about the power of electrochemical biosensors and particularly focusing on this pathogen or disease, botrytis-grain mold and the pathogens that cause that in temperate legumes. So for those of you that don't know, botrytis species are necrotrophic fungal pathogens and they infect a lot of different host crops across the agricultural, horticultural and ornamental plant world. The main pathogens on temperate legumes are botrytis sinaria, the major pathogen, but this botrytis phage can co-occur with botrytis sinaria on a couple of different crops, including favour bean and lentil. Together, they cost the Australian temperate legume industry quite a lot of money each year and that was a survey that's done a few years ago now, so I expect it's probably potentially even worse now that we are experiencing a couple of more wet years and the cropping industries have started to expand. So why do we want better diagnostic tools for this particular pathogen for informing disease management? Well, it's really for us at the moment for application, for surveillance and monitoring and then the information that's generated would then go into better informing the decision-making and that includes the use and the timing of fungicides because currently there's a massive amount of fungicides that are poured onto the crops based on the best modelling that we currently have, but it's more around predictive rainfall and not necessarily on knowledge of what's actually in the paddocks. So to be a good diagnostic tool, it obviously needs to be accurate, so specific to the target organism, sensitive in other words able to identify or predict the presence of the pathogen prior to symptomology development for the best use of fungicides. Quantitative, so really good to know the loads of inoculum out there in the paddock and then marrying that information up with knowledge on the epidemic potential, which is growing. The diagnostic needs to be relatively fast to be useful in the paddock, affordable and portable as well as robust, so reliable within the cropping environment. So that comes to the validation work that I'll talk about at the end of my presentation. Really what happened for me and my group was I was sitting in a PhD confirmation seminar, actually no, it was the end of, it was their thesis candidate review milestone which is at the end of the PhD and I was listening to this particular student give a talk about how they produced these biosensors to detect analytes in human saliva and urine that were then useful for determining potential for cancer. And this fellow here, Mohamed Shadiki, and I then decided that we would look at translating that exact technology pretty much from what he was doing with cancer research to see whether we could make it work for plant pathogens and agriculture and water cultural applications. And that's when we found a suitable student, that's Marzia Bilkis there, and what I'm about to present is quite a lot of the work from her PhD thesis. So along the way we've published a couple of review articles, but I'm still working on that main scientific publication that you'll hear about the work that I'll present to them. So what is a nanobias sensor? Well it's a device that measures an event whether it be biochemical or biological using generally several different ways of detecting that event. They can be electronic, they can be an optical signal and it's usually collected using some sort of probe. So biosensors commonly comprise a biological recognition molecule or otherwise known as a bioreceptor that's immobilized onto the surface of a signal transducer that then recognizes the analyte which is being sought and subsequently transmits a signal that's detectable as I said. So the analyte and often in biomedical research is called a biomarker is from in our world usually the causal organism or the pathogen. It could be a secondary pathogen product or it could be a host product and most of what we work with relates to nucleic acid but it could also be the product from a protein expression or some other metabolite. They're fished out then using a catcher probe that attaches them to the bioreceptor surface and the bioreceptor surface will often then comprise nanomaterials that enable the separation of that initial captured analyte from the rest of the soup of the background that it's been fished out of then allows concentration of that analyte and then subsequent diagnosis and quantification and often this uses a secondary catcher probe to do this. So I'm sort of very simplistically going through the steps here because it is complex it's our civil it is quite complex and it's taking me a while to get my head around this as well and the complex binding then creates a chemical signal that's captured by the transducer and in the work that we've done that signal is then converted into an electrical signal that's then detected in other work a lot of other previous work and ongoing work the catcher probe may use an enzyme that's convert converts the signal to a cop color change and you know a good example of that would be a pregnancy test kit and some of the diagnostic test kits we're seeing evolving currently without without particular pandemic that's occurring. So just to go through this again the target analyte for us is a leaf or a sample for a folio sample because we're working with necrotrophic folio pathogens the analyte that we actually extract then is a nucleic acid that's then captured using a functionalised biosensor molecule that then stimulates an electrochemical change that's then detected. So why have we gone for the electrochemical route as opposed to the you know assay or down the sort of color change route? There's reasons why you would go down a much faster color change route you get a very instantaneous signal but you can't quantify that necessarily so and particularly not using the methodologies that we're trying to apply here so really it's a way of using very small easy to use of low cost devices. The binding leads to a change in current or voltage and the change in that charge distribution after binding is captured by the transducer as I said and in a nano biosensor the transducer comprises a nano material with the advantages that yes they're very small nano particles but they have very high total surface area so you're likely to capture quite a bit more there. The ability to bind probes for secondary capture is also there and there are other chemical properties that different nano particles can be loaded with to enable further functionality such as magnetic separation and in the case of the particles we're using an oxidative reaction that occurs that leads to the stimulation of that electrochemical signal. So really what I want you to do is just imagine in real time if we could implement nano biosensors for informed disease management we're really supporting real-time decision-making by providing the growers with rapid and accurate diagnosis and quantification but also feeding into more accurate modeling around epidemics and so therefore these are very useful for that continuous surveillance and monitoring that would feed into this holistic decision-making process. Now our target pathogens as I told you before were the species that caused botrytis disease, botrytis-grain mold so our the aims of this particular part of our research was to develop first of all molecular probes that are sorry nucleic acid probes that are attached to the nanoparticles for electrochemical biosensors and then we also wanted to quantify those target organisms but also discriminate them so they're discrete from each other knowing that they co-occur in the field and therefore understanding the ratio of that co-occurrence into the future would probably better reform that disease management. So we needed to determine the limits of detection and we did that using titrated pure fungal DNA standards to begin with to create some standard curves. We then wanted to diagnose discriminate and quantify the species within the plant host backgrounds because potentially there's going to be interference from other molecules with tab lights that occur in that particular environmental background. We then wanted to use them to quantify them in the field again under the environmental conditions that the plants actually grow in and that the pathogens actually occur and then we wanted to determine if the licenses were useful to detect potential latent infection knowing that latent infection of botrytis species is a really big problem in many industries. So the first thing we had to do was to find our best possible target analytes, nucleic acids, probes, all names for the same thing. So we needed to find sequences that were conserved for the target species but discriminatory among them and they're quite closely related taxonomically. So the first effort was to source a published primer pair, a couple of primer pairs from literature from Farnett L back in 2015 and we just went ahead and used those primers as pros and just initially did some traditional PCR on a set of isolates that we collected over the years and from the most recent cropping seasons mostly from Faber bean, lentil and we had one from Chickpea and grape as well. So we had a few isolates from Cineria, of Cineria, a few of Faden and what we found was that the, we also included the type isolates from the ATCC collection for Cineria Faden and what we found was that the published Cineria primers were not able to amplify from the typed isolate despite us trying multiple times to get this to work and also it didn't pick up one of our isolates from the Kingsford region down in South Australia. So it was not consistent. The good thing was that it didn't amplify anything from Betryda's baby. So and on the flip side the Betryda's baby published pros, they didn't amplify from Cineria but they also amplified all the Faden isolates that we tried it again. So we sort of thought we could do better, we needed to find more consistent probe for Betryda's Cineria for the Australian situation. So we then sought other literature that specifically published around Betryda's Cineria, specific sequences and tried a few of those. I won't go through all of the trials and errors, you can imagine a lot of PCRs and repeat amplifications trying to get these to work. We ended up actually redesigning some primers and we actually did use the net 1G, the necroprosis ethylene-inducing protein sequence and we re-designed these and we're able to come up with very consistent Betryda's Cineria specific amplification and Faden specific amplification. So this was pretty much where we landed with determining how analytes as targets to go forward with. We then wanted to see how sensitive these were on pure fungal DNA. So we developed a 10-fold titration series from 10 photographs up to 10 nanograms and you can see clearly here a very nice standard curve was produced for both of those with very high mass square amounts. So pretty happy with those and we then went on to generate our genome calculation from the amplification that we achieved and I won't go through that but we've had it tried and tested. We're pretty sure we're onto a good thing and you can have a look at this if you want to go back and look at the slides later on. So effectively what we came up with was from a traditional QPCR perspective and using that titration where we had a threshold of detection around 100 photographs per microliter which equated to about two genomes per QPCR reaction which is pretty good and probably sensitive enough to be honest but then we wanted to convert this to something else to another sort of sensitive device as I've explained. So the next thing we needed to do is just check the sensitivity in a plant background. So we created a bio-SA and we sprayed some plants with different concentrations of the fungus separately and our controls were obviously just water. We also used positive controls from our ATCC collection isolates and what we found was that just using PCR, traditional PCR, we had detection threshold in a plant background was when the plants were sprayed with about 10 to the 5 spores per mil which equated to about 100 spores per microliter. So not particularly sensitive. So then moving on to QPCR we again inoculated the plants we actually used the same plant material. We collected the tissues at 24, 36 and 48 hours after inoculation and we then took 100 milligrams of leaf tissue and extracted the total DNA from that and used that as our template. So interesting observations and thoughts here. First of all it's unlikely that the inoculum was evenly distributed across the whole of the plant material so you've got to think about what does 100 milligram sample actually represent in the real world. It was obvious that the fungal amounts increased over time so so there's obviously growth as the fungal material occurring on the plant material and interestingly despite the fact that we assume the same amount of inoculum is applied in all cases. Consistently more botrytocineria was detected than botrytous baby amounts of fungus. So this sort of led us to an initial thought around maybe the scenario by receptor was potentially more sensitive than the botrytous baby one. So the next step really was for us to convert this to an electro-nano biosensor. There's a mouthful and the first thing was to show you here is how we captured the targets, the analytes. So again material the plants were inoculated and the DNA was extracted and again a reminder this is plant and fungal material. The DNA is overheated to produce single strands and then incubate it with target capture probe or bireceptor and that's that probe is either scenario or baby specific. It's also has a biotinylated sequence bound to it that's useful for then fishing out using dinobits which have streptavidin coated on them. Once that's fished out we can magnetically separate that from the soup from the background solution and we can then release that once it's separated out and concentrated to go forward with. So that's then put prepared it onto a screen printed carbon electrode SPC in you'll see that come up later which once it's on the electrode it binds with the nanoparticles which are gold plated the gold bloated ferric oxide nanoparticles which also create creates in a redox reaction through a rune hexide electro catalytic cycle which exchanges electrons. I'm not a chemist that's how I can explain it I'm sure they're made chemists onboard that that know more about this reaction but it's a very well published phenomenon but that electro catalytic exchange then creates an electrochemical signal and that the intensity of that signal then translates back to the amount of fungal target analyte that was captured in the very first place. So it's a pretty elegant method which doesn't need amplification that's the big point here you don't need a PCR reaction the PCR reaction was just to test the specificity and sensitivity of those of those virus set more sequences that we use. So we then wanted to test the specificity of our virus set to probes if you like similarly to the way we've done using PCR. So we used our our ATCC isolates and we tested those that that method I've just described in various ways with various controls where we just use a bare electrode we then used the electrode with the RU hex on it to stimulate catalytic exchange with and without the the DNA that the virus setters and with and without the pure fungal DNA being applied to it. So you can see this on this table there are instances where we're testing the Fabie virus set to the with Cineria DNA and we're testing the Cineria receptor with Fabie DNA. So they're like the cross reactions that we're testing for and then these are the positive response reactions that we're hoping to see work really well with the Fabie or the Cineria receptor being tested with the Fabie or the Cineria analyte or DNA and here's here's some results from that. So you can see the bare electrode next to know nothing there nothing significant. Then you're stepping through using the electrode with without the RU hex you still get a little bit of a signal but it's still considered the background signal. Again when you add the the DNA there's a there's a non-significant difference I would consider that between there there's the no template again non-significant and then you start to see what happens when you put this one is Cineria DNA on a Fabie probe next to no signal the same with the Cineria Fabie on the Cineria probe and then you see the response the positive response of the Fabie DNA on the Fabie probe and the Cineria DNA on the Cineria probe and the note here is around the strongest signal again that we're seeing with the positive reaction from the Cineria probe which is really interesting and needs to be considered later on. So looking at the sensitivity of these probes again using the same titration series that we used for the QPCR reactions we were able to show a very nice standard curve here again and that our kind of our threshold of detection was around 10 phantograms which is relates down to less than a single fungal score so you know it's as sensitive as it could possibly be in other words and we had a very similar reaction for the Fabie biosensor probe using again the pure fungal DNA and the titration series some really nice ask where it's at and by no means that this happened overnight this was a huge amount of optimization and replication that's sitting behind all of these very easy to present graphs that Marcia did a huge another work to produce again it's really interesting to see if you have a look at the the axis there that the signal detected using the Fabie probes if I quickly flip back was about a third of that detected using the sorry the Fabie probe was about a third of that detected using the the Cineria probe so again showing probably a drop in sensitivity for whatever reason we then looked at the electro biochemical sensitivity of detection in in planter so we treated the plant similarly to what I've shown you before with different spore loadings again just spray inoculated so again you consider the homogeneity of the spores on the plant surface and again extracted 100 milligrams of leaf tissue collected at different time points after inoculation and this is just showing you that at 24 hours after inoculation we're able to detect down to 10 to the 10 to the 2 spores per mil so you know shift of is it two or three fold from the detection using traditional PCR and again you can see a very nice standard curve here that instead of titration of DNA pure fungal DNA it relates to the spore loading onto the plants here again just pointing out that the detection the signal detected with the pure fungal DNA with the Fabie was about half of the signal detected with with the Cineria DNA and probes so we're looking this is just the results in a time series here and just to show you that the sensitivity was of detection was robust across the time points assessed and importantly this is just the three time points important it's this is long before characteristics of the disease are visible so long before you'll see any symptomology or even that's kind of fluffy white mycelial growth occurring in the field and we're really talking you know within within two days here so the next really important step for us to was to then take this to the field and and make it as compact as possible to validate for various reasons and we undertook this using two different trials two different experiments sets of experiments one in the Shade House at the Wake campus in Adelaide and working alongside partners at Saudi there Jenny Davidson and and colleagues and then the other field site testing was also in partnership with with those colleagues but going out to real field sites our favor being across South Australia and is I mean it's not a single handheld device at this point in time it is still a set of of small instruments that would be able to be plugged into a car and you know sort of a quick kit that's used to do a rough extraction and a quick spin and to get the the eluded sort of nucleic acid containing solution here through onto onto our handheld modulizer and then into the the sensing device so it in honor in all honesty this fits in a suitcase that nice here took on a plane and went down to Adelaide to use so we still portable at this stage so at the Wake campus we looked at three lentil cultivars and this was part of their annual trials that were going on so we were lucky enough to coincide with what was going on for them and we kind of worked in with that and those trials have been sprayed with same method I've just talked about just handheld sprayed inoculum with equal parts of sporinoculum from the trial scenario and the trials baby in a mixed spore state two by ten to the five spores per mill of each of those to spores and then we extracted the leaves at 24 and 48 hours after inoculation of more observations and thoughts you can see here that um generally again more fungus was detected over time 24 at 48 hours compared to 24 hours potentially there's a cultivar influence here and I'm yet to check but is cultivar bold for instance more susceptible to the trial scenario and you know if so potentially there's you know there's something on the other cultivars that's uh retarding the growth of the fungus over that time period that initial sort of establishment period and again more betrata scenario was detected um then betrata's baby despite the same quantity of inoculum applied at the very beginning um and then equated to about a 1.5 times more growth of betrata's scenario than betrata's baby in a single mix for inoculum on a single plant um so in other words and that had been adjusted for the probe sensitivities of one to three that we kind of determined from our previous work so the big question is is this is this an example of real-time competition going on and um we're yet to to confirm that with some further work um by doing some nice histopathology the other field trial that we did was in three field sites across um South Australia so real field trials and um they were sampled in 2020 from both both this is important to note both symptomatic and the asymptomatic um tissues and this is what the tissues look like so you can see the symptoms there versus the clean tissues that were collected um and the inoculum was natural infection and so the expectation it was that we would um because this was favor being that we would be able to detect both scenario and favor we collected five symptomatic and five asymptomatic clients per site and the results showed us that we had similar levels of fungus detected across all sites slightly more at Mandela but also surprisingly similar levels were detected in asymptomatic tissues as they were collected in symptomatic tissues so wow that's that's interesting for us because it's sort of then starts us thinking about that whole kind of increase in fungal biomass and and whether uh the fungus is is increasing in biomass in in that time frame without creating symptoms um there were some outliers with significantly more fungus detectants um and we're not sure how to explain that except that we know that epidemics and what always occur in evenly across our paddock so that then sort of brings me to the last point I want to make um is around the importance of diagnostics of latent infection because if we're detecting similar amounts of of fungal essentially amount of fungus on asymptomatic tissues as we are on symptomatic tissues without symptoms being apparent it does indicate that there is some latent growth of the fungus occurring which is a huge problem we know in other cropping industries and if we can understand more about this it would definitely inform our predictive epidemic modeling more much as the presence of the pathogen but presence prior to symptomology and of course then also relates to post-harvest sanitation and and you know the use of concedes going forward for the industry so um in summary um we've developed biosensors to detect both of our type pathogens they're highly specific we're able to discriminate among those two pathogens and I'll also add that there was quite a bit of background bioinformatics work done to make sure that nothing else was going to be picked up um that was genetically closely related to those targets and would occur in a chickpea paddock or a favorbean paddock or a lentry paddock um the highest sensitive so we can detect down to single genomes or spores it's pretty fast using our suitcase kit takes about 45 minutes to get a quantified positive result um it's portable um it's durable so we've shown that those um those those the methods that we use they don't degrade the the um they're they're there and they're present for us to use over and over again in fact you can use the same nanoparticles if you want to wash them off and use them again they're ready to use um it's affordable um and um just depending on numbers that obviously you want to chug through the system it's definitely going to be less than one block it's only proof of salary um and it's reasonably easy easy to use although we're still working on that um so you know I think the results of this research really will help to inform better disease management um on being able to guide biologically and geographically on the use of chemicals uh reducing the excess or unnecessary application of chemicals although that scenario is pretty endogenous and spread throughout systems um hopefully that will then lead to better product returns and reduced environmental impacts and I want to finish up very much by acknowledging the whole team because it's it's not just the people that I'll mention quickly but it you know it's all of the breeders and economists and the growers that that are helping us with the research and making field sites available etc um I want to really thank Shadiki because he's the brains behind the electro-biochemical um biosensor methodology Marcia's the person's done one of the work he knows helped with the bioinformatics side of this Sam is our go-to person for sort of confirming things biologically through his his the biology etc a Jenny in the industry Jeremy is very much helped with supervision and and um the ideas um and there's some references to those two review articles that we've got out so far so and then these sponsors with that exact piece of work so thank you very much for listening and I'll walk up any questions and quickly I can answer them