Alternative Proteins: Digestion, Absorption, Allergenicity | Masterclass

[00:00:18]Okay, good afternoon everyone. Uh, welcome to our master class in uh, protein digestion, absorption models and uh, allergenicity assessment. My name is Andre Bodkco. I'm based here in Chagas Mo Park in Ireland. And it is a great pleasure to introduce my uh, co-hosts.

[00:00:38]Um we start with uh my colleague uh Linda Giblin. Would you mind showing showing yourself there? My colleague Linda Giblin also here from Charismo Mop Park. Uh she will be a second speaker.

[00:00:53]The third speaker will be Claire Mills from University of Suri. Uh fourth speaker will be Kiara Nitrit from the University of Naples. And the last but not least speaker uh counter lanuzel from agroparite INRA in France. So without any further delay I will just introduce uh the project uh to you and that's uh the project is give me a second um as a giant leaps project a European project funded by the um European Commission. Uh so giant leaps stands for gap resolution and safety, nutritional allergenicity and environmental assessment to promote alternative protein utilization and the dietary shift. Uh we have a lot of partners, 35 partners altogether mainly in Europe and one uh one overseas. It's a four-year project. This is we are really at the end of the uh of the project now.

[00:01:57]started in september 22 and will finish later this this year. Um the uh topic of the project is really to accelerate the protein transition uh and make the food system as more sustainable and healthier but of course it is pretty complex uh is a complex uh task. So our solution is first of all fill some of the knowledge gaps in particular in uh uh alternative protein sources uh drive innovation uh improve sustainability safety and health of future uh diets and also the ambition would be to achieve a 50% of the total protein intake from plant and alternative uh sources. So the project has nine I'm sorry 10 work packages and today our speakers are from work package three uh led by cla safy debate by design and work package for digestibility and health and I just put in this slide some of you may be familiar with this uh others if you are in if you are interested in more uh more results please go to the website uh giant leaps so these are our uh protein source processes. That's the short list. So some of them um you know are low hanging fruit in a way fava bean lentils but also oat kinoa rape seed. We have a micro algae here a single cell protein um we have crickets insect and uh at the very end we are also dealing with cultured beef. So we have uh five speakers as I said all from excellent institutions and groups. So I will start off with in vitro food digestion models followed by Linda on absorption methods.

[00:03:44]Uh Clare Mills will give you a good introduction of uh invitro digestion and allergenicity risk assessment. Uh followed by Karan Treat on uh some new research on antutritional factors and co he will uh talk about their invivo assessment of protein uh digestibility.

[00:04:06]So as said my name is Andre Borco. I'm based here in Chagas Mo Park. Uh we are public research uh institute in the south of Ireland not far uh from Cork.

[00:04:17]Um and I'm interested in food digestion in in general. So why does some uh food behave like here on the left hand side you know highly protein highly soluble uh whey protein very fast digestion whereas other foods here commonly eaten uh breakfast cereal uh wheatix that's what it looks like in vitro and that's what it looks like in vivo uh and we are interested in uh finding out whether uh food structure matters and how does it influence the mechanism and the kinetics of food digestion. So um my my general interest in is in food digestion but one big topic is uh protein digestibility.

[00:05:05]So protein digestibility um is basically a number. Okay. And here are uh some of these different methods how to determine them. Some are very old, some are you know inaccurate, some are better than others. So that was nicely uh um um summarized by a recent review. Um but to determine the protein quality it generally requires invivo study right for regulatory and labeling purpose and generally that is done either in rats, pigs or humans. So when you look at rat models and pig models um you know you you would uh question the relevance especially with the rat rodent models. The pig models they are certainly better but there again you have the question um you know do you have to slaughter dead or alive or like here in the middle you have canulated pigs which are which is a very good uh model but again you have you know ethical questions you know is it relevant they're expensive and then uh uh blood and guts so uh as you know um animal trials are not as popular as they used to be so some of the companies that we dealing with they don't want animal tries anymore. Uh so we are lucky that we have a excellent group uh in agopitech in this project. So ka will cover the human intervention trials and that's really the the gold standard.

[00:06:35]When you look at protein quality or digestibility as I said there are several um several different methods out there. Um an older method would be pedicass.

[00:06:48]uh a more recent and more accurate and recommended method would be uh the DIS uh DAS method. So it's digestible indispensable amino acid score. And here we are looking at the uh digestibility um at the end of the elium. Okay. Okay.

[00:07:06]And in general that's done uh either when you when you slaughter pig or canulation or in terms of uh for for humans you use a tube and corner will will cover that. So human models the human model is certainly the gold standard. Uh there's no doubt about it.

[00:07:25]Um the animal studies um give me a second there. There we go. Um the annual studies they are good but they are certainly a second choice. They're also expensive uh um you know relevance could be could be questions and time um um yeah slow basically uh so there is a drive to replace or refine it spec uh especially to replace it. So there's a place for in vitro labbased models, right? Um so in vitro lab-based model in general they are uh relatively simple.

[00:08:04]They can be quite crude uh because they don't have this uh feedback loop that you would have in vivo. Okay. So when when you look at uh the different uh digestion models, first of all you you can there's a case for not doing any digestion. So if you eat one gram of protein, that's what it is. You write that on a packet, it's one gram of protein. If you uh drill a bit deeper, you have different methods and they uh all vary in ease of operation. Uh how um how crude they are or how refined they are. Uh and some of these are very accurate but also very complex and very expensive. So I just want to show you uh a number of uh these models. So there are some very sophisticated dynamic digestion models. So the most well-known is probably the the TIM model. That's the older uh TIM one model developed in TNO uh in the Netherlands. So here you see their stomach and their small intestine. Here they have a absorption cell which is a very good simulation.

[00:09:14]There are other models out there um about 10 to 12 models. There's another one here in France or the recently developed model from from China. If you're interested in these dynamic models uh a number of years ago 2022 uh I was hosting a um webinar where the 11 uh available dynamic models they all had a time slot of 20 minutes. So if you're interested in these go to our YouTube channel. So uh these dynamic models they're they're good models. They're really it's a good representation of the in vitro digestion. Uh disadvantage would be um access you know so you can only run one or two samples per day because that's that's what it takes.

[00:10:01]They are in general uh commercial or at least semicommercial. So there's a commercial interest in uh you know selling the instrument. uh they're not standardized. Okay, they're not standardized. So the conditions are very often proprietary and they are expensive to uh purchase and they're very expensive to run. Okay. So there is a case for simpler models and there are a lot of simple models and I want to um concentrate on one particular type of model and they are called static invitro digestion models and they are called so-called one pot model uh where the three uh digestion digestion phases the oral gastric and intestinal phase are simulated in one pot and the uh model that I want to talk about is the infoest model and that came out of a European uh EU funded cost action. So infoest stands for sharing information on food digestion and myself and my colleague Isida Reio from Madrid were leading one of the work packages uh working group uh to find international consensus and that's what it looks like uh really a one pot model. So you have basically one tube. Um you you um you simulate the oral phase with all its uh um ingredients. Then you change the condition in the same tube to assimilate the gastric phase. You add the pepsin uh if necessary the lipase. Uh you change the pH to pH3 over 2 hours. Uh and then after 2 hours you change the conditions again in the same tube to simulate the intestinal phase. Uh so you add your pancreatitin, you change the pH, you add bile and so forth and at the end of it you have to do your sample um sample preparation for your downstream analysis. So that was published uh initially in 2014 in food to function and then in 2019 in nature protocol where we really set out uh um a protocol a step-by-step protocol with all the uh individual steps outlined including uh troubleshooting. So that is a international consensus and it's based on available physiological data. Some of the data is just not available. So it's a best best guess and now for the last few years that's really the academic and industry standard for static in vitro digestion uh methods and we define enzyme assay in in the protocol. Uh u we give some recommendations of the downstream uh uh treatment is really case by case depends on what you want to look at. If you want to look at food structure or bioaccessibility we give some recommendation. We also provide some tools so spreadsheets how to um perform the invitro digestion. There's a YouTube channel um at food digestion one word. Uh we had some training schools and it has already been used and published um uh thousands of times already. So following on um we uh further developed um the method based on the infoest method to look at the uh protein digestibility. So that is currently happening uh um for a IDF ISO method ISO method for the uh uh assessment of protein digestibility. So that was led by Lotty Isid myself. Uh and only a few weeks ago uh Lotty ego and Rita Portman from Agroscope uh they were leading the this particular study and was published in nature protocol.

[00:14:01]It's called infoest quand that's basically based on the infoest method and then the downstream analysis to simulate uh an in vitro das method.

[00:14:14]So these static models um you know they they're pretty good they're standardized um they are simple but not easy um they're good for screening and they're good to estimate end points so the end of gastric phase or the end of the intestinal phase where they are not so good is for kinetics. Okay, so one pot models are not suitable to estimate the kinetics of um of uh food digestion and that's where the dynamic models come in.

[00:14:48]But as I mentioned earlier, dynamic models uh the access um the availability or accessibility is not that great. So a number of years ago we came up with a standardized semi dynamic uh digestion method. semi-dnamic because the gastric phase is dynamic and the intestinal phase is static and we used uh commercially available equipment titrator um um syringe pumps uh the vessels uh and here we basically assimilate um the um the change in the gastric phase here in this case is the gastric volume but the same for pH you start with a high pH and bit by bit the pH changes and more and more enzyme is added and the gastric volume also changes because we simulate the gastric emptying points. So as I said we use commercially available uh equipment including the stirer uh that can be 3D printed. So that way uh whether we perform the digestion here in Ireland or in the UK or in Australia we can actually uh compare the results and you see something something like this. This is an example from Annabel Mole Cabaro's work where we can see the digestion of in this case it is 100% whey protein as a function of time. So we can actually see the um uh digestion over time. Look at the kinetics uh and you can see fine differences and these differences would be very hard to see in simpler static digestion models. Um so I think that's my last slide. Uh that's where we are at the moment. So we have a static digestion model uh for adults. That's the international consensus. There's also a model for older adults, elderly and a wellestablished infant method. In terms of semi-dnamic, we have established international consensus on the adult uh and there's currently in progress an infant um uh model and a model for older adults. So I would like to thank you and we move on to our next speaker uh my colleague Linda Giblin here from Charismo Park and she will talk about in vitro absorption models.

[00:17:12]Over to you Linda.

[00:17:14]>> Hi everybody. So I'll just share my screen.

[00:17:30]Yeah, pointer. That looks good. Okay, so um as Andre said, my name is Linda Giblin and I'm from Chus as well and I'm going to talk to you about in vitro methods to assess specifically protein absorption.

[00:17:46]So um the first thing is what is absorption? So this is when nutrients or food compounds are transported from the gut lumin across the gut barrier. So for digestive proteins that means small peptides die try and tetropeptides and free amino acids. So they must cross from the gut lumin um they must cross the gut barrier into the bloodstream. So once nutrients cross the gut barrier they're bioavailable and true bioavailability is of course when the nutrient reaches the target cells, organs or tissues.

[00:18:20]So where is um proteins and amino acids absorbed? So they are absorbed in the small intestine and the small intestine is made up of the geodinum, the ginum and the illium. The geodinum is closest to the stomach and the illium closest to the colon. So the protein is digested from your food in the in the small intestine into these peptides and free amino acids. And the small intestine then is six major in length. So it's quite long but it has a much more increased surface area because of these wormlike projections or villi. So its surface area is 250 m squared. So it's huge absorptive capacity.

[00:19:03]So this is what the villi look like. So here you'll have the gut lumen and the nutrients have to cross a single layer of cells so they can be taken up by the lymph system or blood system and this is just a hisytologology slide of pigs here in chus and you can see that this is the geodenum. This is what they look like these villi in the geodenum.

[00:19:24]So let's talk about this gut barrier. So this is the gut barrier here. It's a single layer of cells and it's made up of four different types of cell types.

[00:19:35]And you can see that there's a green layer over it. It's green here in this diagram. It's not actually green, but this is the mucus layer. And this protects those cells from the gut lumen contents. So the four different types of cells are the entrites, which are the main primary cells in this gut barrier.

[00:19:55]They are the absorptive cells. They they're most abundant and they occupy at least 80% of the gut barrier. Then there are goblet cells which are about 4 to 12% of the population and they produce that mucous layer which protects the barrier. Then there are entraendentric cells which is about 2% of the population and they produce satiety hormones. How full do you feel in response to the food that you're eating.

[00:20:19]And then there are panthead cells which are about 2% of the population and really they produce digestive enzymes, antimicrobial peptides, cytoines and they're down here in the crypt because they maintain the health of the villi.

[00:20:34]So how is nutrients absorbed across this gut barrier? Well, the first thing I would say is that the individual cells in the gut barrier hold hands with their neighbors by tight junctions. So there are two mechanisms in which nutrients can pass from the gut lumen to um across this barrier. So parasellular is where they go through these tight junctions.

[00:20:58]Transcellular is where they grow go through the cells and that can be passive or it can be active where they're taken up by a transporter or a carrier and brought across to the other side. So if we specifically look at peptides and amino acids from uh the digesttor, they will primarily use an active transcellular transport like for example pept one is the protein trans the peptide transporter and there are also various amino acid transporters on both the apical side which is the gut lumen side and on the basilateral side.

[00:21:35]So what are the methods to create gut barriers so we can do absorption experiments? Well, um they're primarily cell based obviously. So we can look at intestinal cell lines or we can grow or culture organoids in a two-dimensional or three-dimensional shape or now you'll see these guts on a chip. But there are also artificial ways we can do it. We can do it by using artificial membranes or we can use Xvivo which is uh biopsy samples from animal tissue where we can set out a set of using chambers or indeed now we can do incilico where we can predict knowing the permeability characteristics of the nutrient and knowing how the barrier operates we can predict what the permeability would be like but I'm just going to concentrate on cellbased in this in this presentation.

[00:22:28]So in the cell-based methods you have um either intestinal cell lines and they can be grown as monocultures into a barrier or as co-cultures and there's a wide range of those cell lines but the most popular are kaka 2 and then HD29 MTX. There are also organoids as I said and there also the gut on the chip which introduces a flow of nutrients.

[00:22:52]So let's get back to the first one intestinal cell lines. As I said, the most popular are kakatu. These are derived from human colurectal carcinoma from a 72 year old caucasian male back in 1977. And what they found was that when they grow these cells in these transwell plates for 21 days, they'll actually form a polarized barrier which has tight junctions. It has secretary vesicles on the apical side. It has microvilli. It has a brush border. So are these kaca 2 barriers suitable for protein absorption? So this table lists all of the peptide transporters that we know the apical amino acid transporters and the basilateral amino acid transporters. And here is the expression levels in the small intestine both mRNA transcripts and protein by western blotss. And you can see that they're um they're different levels of expression but they're all present. And here are kacatu cells. And you can see that they have all the peptide transporters. On the apical side, they have all but one of the amino acid transporters. And on the basilateral side, they've all but two. So they do have the full practically the full complement of peptide and amino acid transporters although expression levels may differ within vivo.

[00:24:14]So what is the method to create a polaroid stable kaka 2 monollayer? The first thing that you must do is obviously take it out of uh liquid nitrogen storage and resuscitate it. And we recommend that you do three passages of this in a flask first. Um we usually work between passage numbers of 20 to 30. And the me the media is usually DMAM plus 10% FPS fetal bovine serum plus antibiotics. And then after three passages you're re ready to seed into transwell plates. We recommend a low density seeding so that you ensure that you will produce a monollayer. And so we do about 6x 10 to the four cells per well in a 12 plate. Then we allow these to culture for 21 days. But you have to replace the spent media every 2 to 3 days. And the volumes you're talking about is about 0.5 ms in the apical and 1.5 mls in the basilateral. And the last step then is as they are growing for the 21 days you have to monitor that a barrier is being formed and you can do that by electrical impedance which is transepithelial electrical resistance and you measure that every 7 days and you want that to increase over time but you never do this measurement on a day in which you're changing the media.

[00:25:32]Now when day 21 comes you then have to make sure that the barrier is formed. So for this particular kaku clone HTB37, you'd expect your TUR values at day 21 to be above 800 ohms by cm squared. You want to make sure that the barrier is stable, that there's no further increase in ter values. Often at day 21, people then also run a barrier permeability test. They make sure that the parisellular transport is working and the transcellular transport is working.

[00:26:02]And they do that by using fluorescent compounds like lucifer yellow or with far pharmaceutical compounds. You can also check that the barrier is mature and you can do that by the expression of brush border enzymes for example alkaline phosphatase.

[00:26:17]So what does the absorption experiment look like then on day 21? Well, here you have your food and as Andre talked about is you can then do an in vitro um digestion with the infogest method so that your protein in the food is now digested into uh peptides and free amino acids. You incubate, you add that food to the apical side of your polarized monoler that has been grown for 21 days and you do the incubation for about 2 to four hours in a simple buffer like HPSS with an energy source or simple media.

[00:26:53]And then what you want to do is you want to see that the peptides and amino acids are crossing from the apical chamber into the basilateral chamber. And what I would a word of warning is that you need to um apply sufficient food digesttor material here so that you can detect it in the basilateral. Um you can detect the individual amino acids and the peptides in the basilateral. Now um that will all that will really depend on the type of instrument you're using for detection and its detection limits.

[00:27:29]So before you start however there are biompatibility issues between food digesttor and cell monolers and the main challenge is that the food digesttor contains digestive enzymes for example trison and trison will lift your cell monolers. Bile salts are toxic and there could also be osmolality issues. So the solutions are to remove or inactivate the digestive enzymes and to dilute the salts before you put the food digesttor down on your monollayers.

[00:28:00]The warnings however are that if you use enzyme inhibitors they can often be toxic to your cells. They can inhibit brush border enzymes and the alternative would be to heat it the food digestive sample but then you can um include structural changes in your food um in your food sample. You can remove the enzymes by using for example molecular weight cut offs but then you could also remove food components and you can dilute but then you may be back to being unable to detect your nutrient in the basilateral chamber. Um but we do recommend that you do a biompatible check pre and post a detoxification. And what that means is you take your digesttor um sample at different concentrations and you incubate um overnight cells with this and you run just simply a cell viability assay like an MTS assay.

[00:28:55]So we have recently published a paper back in 2024 on a series of recommendations which give you a step-by-step guide depending on your research question type of digesttor and food component. So you'll be able to make decisions on how you're going to detoxify your food digesttor before you put it down on your cell monomers. And it is something like this. This is a graphical of a yes no yes no.

[00:29:21]So what are the advantages with cac with cacao 2 cells? Well the I suppose the first advantage is it does have good permeability correlation with human gut.

[00:29:31]It has an R squared of about 0.84 with 21 drug components. It is tried and tested. So there's lots of data in the literature. It is commercially available. It's quite a robust uh cell line. It's easy to culture. If you do 12 well plates, you can do three plates per day per person. So it's not high throughput, but you can get through a number of samples. It is suitable for food digesttor provided you clean it up.

[00:29:57]Um there is an option to add additional cells directly to the monollayer like HD29 or to the basilateral site like for example if you want to add M cells and it does express the transporters particularly for peptides and amino acids it's seems to have almost the full complement of those transporters.

[00:30:16]The limitations are that it does have narrow tight junctions compared to human gut cells. So you'll see that kaka 2 cells are about 0.4 four nanometers between each individual cell where it's about 1 to 1.3 nanometers for human cells. It also has the electrical resistance over it is greater than 800 ohms per centime squared. Whereas with humans it's much less. It's 50 to 100 ohms by centime squared. Absorption is influenced by solubility by hydrophobicity polar surface of the compounds. They are cancerous cells. So they do not use butyrate as an energy source and butyrate would be the main energy source of intestinal cells. They lack the metabolism enzyme cytochrome P450 and then they don't have a mucous layer and we know that mucus is present in the intestine and there's substantial interlab var variability due to different protocols and different kaka 2 clones. So the last two there I'll show you what we have done to remedy that. So many labs now use they add goblet cells without disturbing the kacatu monolers.

[00:31:24]So this goblet cell this this cell line HD29 MTX or HD29 was derived from a human colon adino carcinoma from a 44 year old female and they found that if they treat it with methotex rate the subclones mature into actual goblet cells producing mucus. So what people do now is they often seed the cacao 2 cells with these HT 29 MTX at a ratio of 90 to 10 and then they allow this the 21 days to form the polarized barrier and it will have a mucous layer.

[00:32:01]To overcome the variability issue within giant leaps, we set up an international ring trial where we agreed protocols on the Kakatu HD29 MTX culturing on seeding the trans wells on measuring the tur on measuring the parisellular and transcellular permeability. So this was across eight European countries across 17 laboratories and the outcomes of that are that we want to reduce the variability via these consensus protocols much like what Andre was talking with the infest standardized protocols. We also want to be able to identify a stop no go where you have to have a minimum terror at day eight or that barrier will never form by day 21.

[00:32:44]So it gives people a chance to stop their experiments.

[00:32:48]Um we want to be we also can now define a healthy versus an inflamed barrier and we can understand try to understand and unravel the relationship between electrical impedance and also how that tur value relates to luciferous lucifer yellow permeability.

[00:33:06]So just then to mention the other types that are out there I'm not going to mention methods for these they are the organoids and the gut on the on a chip.

[00:33:15]the organoids uh the 2D organoids are intestinal and epithelial barrier is formed from starting from stem cells and they will have the full complement of cell types they'll have a mucous layer entrt goblet cells panate cells stem cells entend cells and tough cells and it's a complex protocol however to start with stem cells and differentiate it into intest intestinal epithelial cells but you do have multiple cell types there you have a cytochrome P450 and you do because you're growing them as a as a monoler in a transwell you do have an accessible apical side if you go to 3D organoids that's talked about again we talk about taking stem cells um usually from crypts actually and they will self-organize into intricate tissue-like structures which are three-dimensional and therefore they mimic the architecture of the intestine However, there is an accessibility issue to the apical side. So now they've done an apical uh in anteroid and an apical out version.

[00:34:27]What additional um uh characteristics will got on a chip? These are now commercially available and what they actually will be is they will have intestinal epithelial cells but they'll also have a micrfluidic device that mimics fluid flow and therefore nutrient movement and that simulates peristalsis.

[00:34:48]So that's that additional characteristic to add to the system.

[00:34:53]So just to recap this is your absorption experiment. You start with food. You digest the food in the background. You grow your cell lines or you differentiate your cell lines in into a a polarized monollayer. Then you add your digesttor to one side of the chamber and you measure what crosses. So that is nutrient absorption.

[00:35:19]So thank you very much.

[00:35:24]>> Excellent. Thank you. Thank you Linda.

[00:35:27]Uh so our next speaker will be Cla Cla Mills. Um I just forgot to say if you have any questions you know type them in the question and answer box. Would be actually good if you have any question if you type them now because we we can see them dur during the talk. Okay very good over to you Claire.

[00:35:46]>> Yeah and just to say I'm doing Kiara is also contributing um to what I'm talking about today. So let's get going.

[00:36:03]That all okay Andre?

[00:36:07]>> Yep. Perfect. Perfect.

[00:36:08]>> Okay. So, in vitro digestion and allergenicity risk assessment which is in work package three of the giant leaps project and I'm going to be um talking to you about this today together with Kiara Nitrida who also contributes to this work in giant leaps. So um hopefully the slides will move on. Um first of all my declaration of interest just to say that I do work for the UK food standards agency and I'm on the advisory committee on novel foods and processes and I do work for the European Food Safety Authority. And I leave this disclaimer in because some of the work that we talk about has also been done outside the framework of giant leaps but under a FSA procurement. Um so they require this disclaimer um to be added.

[00:37:04]So um just going back to the whole concept and why digestion is important in allergenicity is that hydrarolysis is acknowledged as a way of removing the allergenicity of foods. So um and I don't quite know how my um mouse was moving without me doing anything but anyway there we go. Um, so small peptides lack the multiple sites that are required to elicit an allergic reaction. In order to cause an allergic reaction, you make a molecule called um, which is normally produced to help protect you from parasitic infections.

[00:37:49]And in some people, for some reasons, they they begin to make this towards their food. But those antibodies have to be directed towards multiple epitopes.

[00:37:59]Um, I'm not sure where the are there two pointers on the screen because there's one that seems to be moving around randomly. I'm not sure where that's coming from. I'm not used to using zoom, I'm afraid.

[00:38:13]I I don't know whether you can see mine.

[00:38:15]Um, but there is there are two laser pointers on my screen.

[00:38:21]That's not mine.

[00:38:24]Don't know who's moving up. Anyway, we will keep going. Um, so if you start breaking your protein down into small peptides, you lose the epitopes. I have to say I'm finding this very distracting. Um, completely. I take that away. I'm not going to bother. Um, if you have enzyatic hydrarolysis, it's used actually to prepare hypoallergenic ingredients. And what you can see here in the bottom left of my slide is a study done by Katrina Bow in actual uh animal models looking at the capacity of hydrayed betalacttoglobbulin from uh cow's milk whey um and a hydrayed whey protein to cause the development of antibodies in an animal model compared to the intact protein. And basically when you hydrarolyze it, the capacity to both cause an allergy and then elicit an allergy is very much reduced. And that work is also backed up with a study that was done by Nestle with Sophie Newton where they showed that if they hydrayze the betalacttolobulin they have almost undetectable protein um and it abolishes antibbody binding. So that's the uh hypothesis that we work to that if you hydrayze your protein you're going to reduce its allergenicity and that underpinned the original work done back in 1996 by Jim Award and the guys in Monsanto which actually took us all on this journey and the stability of food allergens to digestion in vitro and in fact the paper of Jim Awards published in nature biotech hypothesized that food allergens must present sufficient gastric stability to reach the intestinal mucosa where absorption and sensitization to the development of allergic disease can occur. And this gave birth to something called the pepsin resistance test. Um, which is now really considered a physicochemical marker of allergenic potential. And the test was developed using simulated gastric fluid derived from US Pharmacopia and recipes used essentially for drug dissolution testing. And the original paper purports to show that resistance was greater for allergens than putitive non-allergens. And back in 1996, we didn't know very much about food allergies. And this has basically become enshrined in legislation as part of the safety assessment of GMOs. And so we have had this integrative uh weight of evidence approach for looking at predicting new allergens, denovo sensitization and prediction of crossreactive allergies which has included this pepsin resistance test and in vitro digestibility tests for quite some time now. So we know a lot about it. However, what has come clear over the time is that the relationship between resistance to digestion and allergenicity is not clear-cut. So, this is work that I and my team did in Manchester back in 2013 looking at different purified proteins digested using the pepsin resistance test or varants of it. And really what we found was that resistance to digestion was not predictive of allergenicity alone but it did give you a very good diagnostic readout of the susceptibility of different types of protein structure different scaffolds. So for example um the 2S alb album albamins the lipid transfer proteins uh were highly resistant to digestion whereas the bet v1 homalologues which are considered to be less problematic allergens were way much more digested um than the than the trou albamins and ltps.

[00:42:36]So what we now have is pepsin resistance test that is really now much more considered to be a marker of biochemical stability of a protein. But gastrodiodenal digestion is still valuable because it provides understanding of the context in which a protein is presented to the immune system in a a physiologically relevant context. So now we have a new approach and this is from ESSA um published um I think late last year um which is now moving towards a stepwise I love it they slightly amend it stepwise weight of evidence approach but what you can see here is that our um in vitro gastrointestinal digestion has very much moved into this protein stability looking at pH eat and gastrointestinal digestion. And what is important is that it is now expanded not just to look at allergens but also look at t toxins and anti-nutritional factors.

[00:43:44]So how do we do that in vitro digestion?

[00:43:47]And the processes for allergenicity have really come from that pepsin resistance test which was grossly unphysiological to see how can we develop conditions that are more relevant to the physiological situation and that also reflect conditions found in vibo in perhaps what we would consider susceptible groups. So infants where we know the gut is immature may not produce as much pepsin may not reduce the pH down as low as you find in in uh adults.

[00:44:20]um looking at fasted versus fed states and also just think 25 30% of the European population will take ant acids at some point and the high very high uh pH that you get in the stomach as a consequence of their use which almost stops pepsin digestion completely. And then you have intestinal conditions again which can reflect those perhaps found in vivo in adults and infants where there are fewer lower levels of of enzymes produced. So in giant leaps, one of the things we've done in work package three is really map out all of these conditions and it's been applied in a 96 well plate format with physiologically uh relevant biosactants because the original pepsin resistance test and the intestinal digestion test there don't take account of um surfactants.

[00:45:17]So the other thing that is important and Andre touched on in his great introduction to this whole um master class is uh about sampling and for allergenicity risk assessment. We want to know not the end point but actually how is the protein digested and how will it get the exposure of the immune system. So the important part of the immune system in the gut wall that's involved in sampling of the pest patches these are found in the small intestine.

[00:45:50]Most of them are found towards the terminal illium um but they are scattered throughout the small intestine and the first pair patches are present in the dueum. So understanding early gastric event emptying events is really important. The time courses and this is just a wonderful study from 2019 looking at the impact of calories and also the difference between particullet and soluble material and how that influences gastric emptying and what has been arrived at. But again, like Andre said with the infoest protocol, you come up with a a a calculation based on what you're hoping is going to be right. Um or as an approximation is generally where taking a gastric digestion that at 10 minutes is the important one because it's when you get the first gastring emptying event into the intestinal um into the intestinal phase. So that's one of the challenges for the allergenicity risk assessment is that we need to sample and we need multiple time points to provide data for kinetic analysis. So I'll now pass over to Kiara who's going to pick up the baton and I will change the slide as you tell me.

[00:47:10]>> Thank you Cla. So now we move on to show you showing you some applicative examples of how we can um actually analyze and collect readouts from uh these uh samples that we have collecting at different time points of gastrointestinal digestion to inform about the fate of these allergenic proteins. So the first step can be electropharesis analysis can that can be coupled also with imunolotting.

[00:47:39]So in this electropharesis analysis what we can do is to collect information about proteins and polyeptides above a certain molecular weight that would be around 10 kyodalon and it is very important and there are some consideration to make and this is related with the type of enzymes that we actually select for simulating the gastric endoden phase and this actually will become even more clear later on but we tend to use purified enzymes and this is because this is crucial as we want to follow the breakdown of the proteins electrophoretically visually basically the use of this purified enzyme would actually help maximizing the clarity of the information that we are following and another advantage of electropharesis is that this technique can be semiquantitative through densitometry analysis so means that we can actually follow persisting peptides And we can actually calculate the halflife of specific buns and quantify the fate uh throughout the digestion um throughout digestion the time of digestion. Uh next slide please.

[00:48:57]Oh, so it's a >> to support electropharesis which can tell us together with the monoling whether a band is a given protein and how actually is breaking down. What mass spectrometry can provide us is basically the name and the surname of this protein can provide us with sequences of the polyeptides.

[00:49:21]Um, as mass spectrometry can work out to a lower range of molecular weight, meaning that we can target much shorter peptide as short at four, five amino acid length, which may not be that relevant for allergenicity risk assessment. Yet is the shorter length of a peptide that can be inferred to a protein with a high confidence. As when we go down to a lower molecular weight with two and three peptides, it is more difficult to uniquely infer and associate that sequence to a protein.

[00:49:58]And then we can monitor as I has can be 3,00 m over zed, which means length that can go up to 30 35 um amino acid. So quite large polyeptides.

[00:50:12]Next slide please.

[00:50:16]So in order to increase the confidence of peptide identification and this is true also for proteins identification through mass spectrometry uh we need very strict identification criteria. uh we need a very high repeatability and reproducibility because if you think about it, we are talking about a very complex system where we simulate the digestion of proteins and we also then um actually implement downstream process to prepare the samples for mass spectrometry analysis. So repeatability and reproducibility are really really important to make very strong the outcome of the analysis.

[00:50:58]With mass spectrometry for identification we do rely on protein databases.

[00:51:05]This is also quite of um a critical point, a critical aspect as um when we think about novel foods, alternative foods, um the presence of well annotated and curated databases is not a given. Uh sometimes we have to work with omology and this can actually in some ways influence the type of readouts that we have. Nevertheless, there are biioninformatic tools which are really interesting and really help out uh with what we call denovo sequencing, meaning that we can from our masspec data um sequence peptides and then with this peptide sequences, we can perform what is called identification biomology and still have an idea of the persistent peptides that um can be identified in our food ingredient and uh can be in uh can be associated to the ingredient itself.

[00:52:10]Once we have collected all this information then we can move into trying to understand um whether in this long peptides there is a epitop that can be persistent epitope that align with the uh protein sequence. And uh when we look at the molecular level our proteins and we try and we manage to identify portion of the proteins which is more resistant to digestive enzymes. In that case we can start understanding more in depth for example the influence the sequence of the protein up to the post-transational modification that can naturally occur. how this can then influence the way the protein is broken down and also um the way actually is folded with dissoli bonds and in this context we can even start looking into uh induced post-transational modification that can be related with processing.

[00:53:08]Next slide please.

[00:53:13]So just a few example what you see here these are electropharesis that have been performed on purified proteins and you can see these are time points throughout um the digestion which was performed at different phes. So also challenging actually the ability of the enzyme to degrade and you can clearly see that it's easy to follow the breakdown of the protein throughout the time points. So we can actually build up a real kinetics in a simplistic model.

[00:53:47]Um next slide please.

[00:53:50]And thanks to this readout and thanks densometry that we can um we can collect information and we can actually collect um objective information that would inform about the kinetics uh behind the breakdown of the proteins. we can clearly see which are the transient digestion product and which are the most persistent product uh that would lead to a longer exposure in the gut.

[00:54:20]Next slide please.

[00:54:23]So we have been talking about proteins, we have been talking about purified proteins but this can it is applicable also to more complex ingredients and even foods. And this is an example where we have been following the faith of allergenic ingredients in the context of a baked muffin. This was meant for oral imunotherapy um to deliver eggic allergenic ingredients. And what you can clearly see here is that we can follow and monitor several different type of ingredients allergenic ingredients from the egg, peanut and even milk. And we can see that in this context even at a further stage of the uh digestion we can clearly see that there is still resistance immuno activity. So there is still polyeptide that are more resistant and this is actually more evident when there is a lack of the pepsin digestion.

[00:55:23]Um so when actually uh at the very early stage of gastric camping.

[00:55:30]Next slide please.

[00:55:34]Again complex food matrix. Here is an application of mass spectrometry to the study of the digestion of a pizza which is just a becket product is a is a model of a becket product where we follow the digestion of gluten proteins as well as soybean proteins. So one of the first outcome from mass spectrometry would be the length of the peptide at the abundance the number of peptides with that given length at that given time point of digestion. And what you we can see is that um during the digestion gluten peptides gluten proteins would lead to much longer peptides as compared for example to soybean protein. And this is intrinsic to the primary structure of this protein as gluten of course is much more rich in proline and amino acid that would reduce the liability of this peptide to be digested in the eveninal phase.

[00:56:38]Next slide please. So going back to the drawbacks of pancreatine. So we tend to use quankadine as a complex pool of enzymes for dudinal digestion. This is what is used in infogest. However, as I said at the beginning, a pancreatin is quite of a problem when we want to use electropharasis as one of the technique to collect readouts. And this is because it's clearly seen on the right hand side of the slide where you really see in the pancreatic phase in the intestinal phase we can really see that the food proteins are masked by the complexity of the pancreatic enzymes as compared to on the left hand side a digestion that has been performed with the um purified um enzymes. So how to overcome this? One of the ways to use simplified to use purified enzymes or what can be done is to use imonolotting where we can use actually antibodies to mark the proteins and follow the bands related to the given allergenic protein. Um nevertheless um also in this case we have a limitation from one side antibodies they have to be very well characterized um uh to um have a confident readouts and on the other side is also true that they may fail in recognizing very large um fragments and so we may fail at monitoring properly these digestive enzymes.

[00:58:19]So just to finish up um we've given you a bit of a a big overview and you can see how in vitro digestion has evolved over the years since 1996 in Jim Aswood's paper and I think we have much better um tools and approaches now that can help us with assessing um potential allergenicity but there are still always aren't there always gaps and needs. I mean one of the things really important is for reporting that people do report in their studies protease to protein ratios providing enzyme activity units um where you actually do your studies that you use different pH conditions and those enzyme ratios that might represent physiological conditions found in either infants, adults or those with an impaired digestive capacity. Um something that in addition to ant acids, many older people these days are put on to statins and they have a profound effect on steroid metabolism including bile salts. And you also have people with um issues over bile uh production which may also affect digestibility quite considerably.

[00:59:38]One of the things in addition to the 96 well plate format, we've been trying to refine the conditions for allergenicity risk assessment and so that we can actually connect a bit more to the um approaches taken in infoest for looking at the nutritional quality of proteins and we definitely coming up with harbonized and perhaps more standardized methodology. But what we're still needing going forward beyond giant leaps I think is a need for interlaboratory assessment. It provides more credibility if you show the transferability of test um approaches and using appropriate methodology for monitoring um the outcomes of digestion and readouts and and we've done a lot of work to actually move towards readouts that can actually uh provide better interpretation of test results. And lastly, and this is something that I have a bit of a bug bear about, but maybe Andre also would like to see more standardization in infoest, is this need to include comparator and standardized proteins.

[01:00:43]And very very few laboratories include those and I think that is something that is going to be really important for the future um use of this methodology uh more widely. So just to say thank you to everybody um who's contributed from work package 3 but particularly uh to Javanni Daria who works with Kiara and two PhD students Sha Kunin and Lorie Stevenson who work with me and their PhD projects are associated with the giant leaps project and thank you for listening.

[01:01:22]>> Thank you Cla. Wonderful. Um, so Kiara, you're the next speaker.

[01:01:30]>> Yes.

[01:01:31]>> Excellent.

[01:01:33]>> So I'm going to share my screen.

[01:01:49]>> Can you see the screen?

[01:01:50]>> Yes, perfect.

[01:01:51]>> Okay, perfect. So, good afternoon again.

[01:01:54]I'm Karen Nitra from the University of Naples and I'm going to um I'm going to take you through exploring antiutritional factors in the context of um the um digestibility and the safety of um alternative protein ingredients.

[01:02:15]So when we think about anti-nutritional factors, we are actually bridging the gap in between nutrition and safety. And the identification, characterization and quantification is part of a um is part of the premarket evaluation of alternative pertinent ingredients.

[01:02:38]whether it is part of um the nutritional assessment because they are considered nutritional limitation or if they are considered naturally occurring toxins.

[01:02:49]Um their assessment and the evaluation is actually has the final goal of um providing um ingredients with high quality uh that are nutrition and safe.

[01:03:04]So what are antiutritional factors?

[01:03:06]These are molecules that are able to um actually interfere with natural processes of digestion, absorption of nutrients. So basically are molecules that would interfere with metabolism of nutrients. And if we think about plants and traditional ingredients like pulses and cereals, this anti-inritional factor are actually present um in these ingredients naturally present in these ingredients.

[01:03:40]And from a plant perspective, they are chemical molecules are chemical complexes which are produced because plants do not have legs. So they cannot run away. Therefore they have uh to they have developed a mechanism that would protect them from predators from pest but also from environmental stresses like can be the extreme heat. So in this context and in the context of climate change and global warming the assessment of antutritional factor becomes even more important in traditional food as well as in um more alternative and noble food products. So when it comes to the classification we have categorized anti-nutritional factors in nonproinatious anti-inritional factor primarily small molecules and protein anti nutritional factors um within the non proteinious anti-nutritional factors actually we have a lot of molecules some that are also associated with um um taste and of flavors like bitterness ness a stringency in the mouth and just to name a few we can we have the phytic acid which is known to bind the minerals and impair their absorption but in the context of protein digestions also polyphenols and dietary fiber can be considered anti-nutritional factors given their ability of binding to proteins and this can be both food proteins as well as the gastrodigestive enzymes s um impairing their ability actually to digest macroolelecules.

[01:05:27]When we look instead at protein anti nutritional factors, one of those that we may be more familiar with are lectins. Uh lectins are proteins that are capable of binding sugar moyotes and these can bind to the interosite in intestine altering the natural permeability and they are considered interrotoxins.

[01:05:50]And then we have enzyme inhibitors both proteins as well as amaz inhibitors which are again proteins that have the ability of blocking gastrointestinal digestive enzymes from breaking down macroolelecules which we can only absorb after digestion to small small amino acids.

[01:06:16]So when we needed to approach the assessment of antiutritional factors within the giant lips project and we actually screened ingredients that at very beginning Andre showed you uh which can go from plant proteins so plant ingredients all the way to single cell proteins or insects. We decided to use a stepwise approach. Uh we started from an incilico bional uh b um bioninformatic analysis where we started gathering information from literature which may be available for more um traditional alternative ingredients. But then when you move into more new ingredients, literature review can be much more limited. And then we took a step further and we approach through filogenetic analysis. We try to see all these ingredients whether they could be related with more described and more characterized food ingredients. And we started digging and analyzing whether available genomic data, proteomic data and metabolomic data. There are biiniratic tools that can help you out when you want for example screen the potential presence of lectins based on genomic data that are available on organism.

[01:07:42]So uh we move a step further and we kind of after this um actually review uh that we have done which was an exercise of uh that involved multiple work packages including work package three. We moved a step further and we started working in the lab to try to collect information about our ingredients and this was done through proteomic analysis and metabolomic analysis. So through eSer resolution mass spectrometry.

[01:08:15]So what is interesting and why this is needed? This is needed because genomic data are static data. So they tell us about the potential that something can happen within an ingredient within an organism. But actually proteomic and metabolomic are dynamic. They work and protein expression metabolites are expression of environmental events. that can occur around our ingredients around our plant around our microorganism.

[01:08:48]So collecting at the molecular level protein level evidence of expression or the presence of a metabolites can actually tell us how that food is that ingredient that organism is reacting to the environment. Now this exercise can be challenging and can be challenging because we need very well curated protein databases for proteomic analysis as well as we really need well annotated spectral library for metabolomic analysis and this is not a given and it's something we may need to work towards particularly for more innovative ingredients. So this step is rather crucial and important because would tell us how to move forward. Meaning what are we going to do next? What are we going to analyze in the lab? What are we going to look and measure the activity for?

[01:09:46]And this is done through a combination of invitra assessment that would quantify the molecules and also tell us the activity of the molecule. And then a step further would be assess the activity the effect of that molecule under invitrophysiological condition. So basically an inhibitor how much would inhibit the gastrointestinal digestion in vitro. So in our info using our infoest model.

[01:10:17]So just to give you some practical example, this is one of the outcome of this incilico screening we have done combining literature data together with um also uh metabolic pathway analysis, genomic um and proteomic fishing. uh we have analyzed several different type of plant proteins, single cell protein and insects and we have a different level of information like for lectins we can see that they may possibly be quite ubiquitarious molecule. They are they are meta they have metabolic reasons uh for being um actually expressed uh both in plants but also in single cell proteins and the presence of lectins could be collected mostly at genomic level meaning that we can say that potentially there could be this protein expressed.

[01:11:16]Um we have some type of antiutritional factors that are food dependent meaning that for example kitin as a non-digestible carbohydrate may be impairing the digestion of protein in insects to a lower extent in a slightly different form. We can find them also in fungi but predominant this would be and non-digestive diffs would be um anti-nutritional factor in insects and when we look at the lower molecular weight antiutritional factors we can see for example phase a phytic acid to be quite ubiquitarious in different ingredients. So this helped us basically move forward um when we had to start our invitro analysis and this is where we started facing challenges and one of the first thing we need to understand is that we do eat ingredients. Well, actually we do eat foods containing ingredients, but when we want to analyze this anti-nutritional factor, we work with extracts. And this is rather important because when we approach the screening, this this proteomic and metabolomic screening we have done to collect as many information as possible using mass spectrometry. We could use quite harsh extracting buffer.

[01:12:36]We do not need to keep the native structure of the proteins or to keep the molecule intact. We just want to get as much as we can and this harsh extracting are very useful in this respect and so they can provide us quite of a good understanding of the ingredients.

[01:12:55]Nevertheless, when we want to measure the activity, so the inhibition for example that that molecule can um have in a physiological environment like a gastrointestinal digestive system. Well, we cannot use this ar extractant but we heavily rely on pre preventing so preserving the native confirmational state of the molecules. So we need to use physiological extractants and this is where we have to balance our decision as actually extractants the the method need to be aligned to the matrix complexity.

[01:13:34]For example we need to have different approaches whether we are working with pulses or whether we are working with single cell proteins has microbials and fungal sources they actually have cell walls. So this is where physiological extraction needs to be associated with um bit milling for example or may require homogenization phases at the beginning.

[01:14:01]So I'll take you through a few example starting with phytic acid as an example of a small molecule. So what you see on the left hand side of the screen, this is the structure of the phytic acid where we can clearly see that there are six phosphate groups that are negatively charged and these work like they are magnets and they actually bind to um to essential civil cations they block them. They work like they were a cage and they actually create insoluble complexes and they reduce they by accessibility. Basically they are not bioavailable anymore and these cartoons cannot be absorbed. So the presence of high level of phytic acid in the diet can actually in the long term cause mineral deficiencies.

[01:15:03]So when we started looking at the different method that available for measuring fatty acid in foods, we found several methods and actually um the the the pioneer method one that is very widely used as is very simple is the weight reagent base method. Uh nevertheless this is a method that cause quite high variability in the results.

[01:15:30]There are more recent developed methods.

[01:15:33]We have for example the annual exchange chromatography based quantitative method. In this case this is also a standard method is a validated method that would monitor the uh more active form of phytic acid. And there is a more recent method that should be more specific as it is actually using an enzyme, a specific enzyme looking at the phytic acid in its most active form meaning that it has the six phosphate groups. um specifically overcoming some of the limitation of the Iote exchange chromatography where there is coalition of different forms of phytic acid more active with low to no activity forms of phytic acid.

[01:16:24]So uh one issue when we have different methods and this is actually true for everything is even true for quantifying proteins in a given ingredient is very important that we always trace the amount of molecule we are quantifying to the method that has been used because the method is what would make that quantity comparable across different type of ingredients and um um so it's always important important to not only provide a quantity but also provide the method that was used for that given quantity and um and of course a step forward would be the harmonization of all these different type of methods um trying to provide um a a common um quantity.

[01:17:18]So we also have looked into as part of the giant tips project methods to reduce uh the level of phytic acid in the ingredients even in processed foods through fermentation there are thermal processes that can actually lead to the reduction of phytic acid in this active form. Um nevertheless when we start breaking down the phytic acid to forms that have low to no activity um these methods they need to be supported by a certain level of validation in vitro where the and the by accessibility of the valention um of the processed uh ingredients of the processed food should always be validated in comparative studies where untreated control are used. So we actually do not measure how much left active fatty acid we have in the ingredient but actually we try to see how much of the divent ions are still bioaccessible at the end of the simulated gastrointestinal digestion as compared to the untreated sample.

[01:18:32]And this is just very rapidly an example of phytic acid screening in different uh pulses and cereals ingredients. Um these are not only flowers, we do also have protein isolate and protein concentrates from pulses which are becoming quite mainstream ingredients in the production of these alternative foods. And what is quite evident is that fatty acid is enriched in isolated proteins ingredients and this is due to the processing. Uh these ingredient goes through for the fractionation. As we know that proteins are solubilized in an alkaline environment and then are precipitated through um acidic precipitation and actually phytic acid solubility mirrors that of proteins and so is enriched in the alkaline environment and crash out with protein through acidic precipitation.

[01:19:32]Let's move to enzyme inhibitors. So hurt nutritional factors. So how do they work? We have here on the left hand side trin and kamotripsin. These are the most abundant proteases in the pancreatic fluids in the blanc juice and in the presence of inhibitors which are themsel proteins and this is a serin serin inhibitor what they do basically is to block the enzyme in their active site. So once this enzyme is blocked in the active site, it is not capable anymore of catalyzing the hydraulysis. So they impair protein digestion. We cannot digest any more proteins. And this is true also for amid inhibition.

[01:20:20]So how do we analyze and how do we test whether there is active enzyme inhibitors in our ingredient? We technically incubate our trin with the ingredients protein extract and we test the enzyatic activity of the trin. And uh what is important again is to have a protein extract prepared under physiological conditions because we really need to preserve the native structure of the inhibitors in order it to be to work in our system.

[01:20:59]Another very important thing is that as I said at the beginning is not only protein inhibitors that can impair digestion. We also have small molecules that that could impair digestion. So in order to link univocally the inhibition to the proteas inhibitor so to our protein inhibitor, we um have uh taken a step where we have been uh thermal treating our protein extract in order to denature the inhibitors. Uh so that if the inhibition was still um uh was um still active was still present in our invitro essay this would suggest us that there are heat stable molecules that are causing that inhibition other that the inhibitor themselves.

[01:21:51]And last but not least, to ensure biological relevance, the inhibitory activity should also be tested under simulated physiological conditions as to test the uh enzyme inhibition. We work with protein extracts while actually under simulated physiological condition, we do not remove the ingredient matrix effect. We take into account biochemical dynamics um of the gastrointestinal environment meaning that we account for pH fluctuation but also the dynamism of the enzyatic digestion that could have an effect.

[01:22:32]And what you see here is just an example again of trine inhibition tested by protein extracts. And we see that there are some of these ingredients that have inhibition of thin particular these are protein concentrates. And this is because we have an enrichment of these protein inhibitors that are not modified by any mean because the protein concentrate is obtained through our classification. So basically is a physical enrichment and this is instead the testing of the pancreatic trips inhibition under physiological digestive conditions where you can see these are the same ingredients. You can see that the inhibition becomes negligible in the duodenum and uh even after autoclabbing.

[01:23:23]So means even after the thermal treatment and the autocribing is the most effective thermal treatment in denaturing inhibitors um uh protease inhibitors.

[01:23:36]So last but not least lectins we will only touch base on this. So we said that lectins can potentially be present in several ingredients in many of the organism and potentially because we have been addressing this from a genomic perspective. Um we have also collected later on also some protein level evidence of expression. So lectins are very complex. These are proteins that can bind carbohydrates. Why are they important? Because we have carbohydrates on the interosytes and so these proteins can actually bind to intestinal brush border membrane enzymes and can compromise nutrients absorptions.

[01:24:18]Lectins uh can be really different families and they can be uh dramatically different because of the specific car carbohydrates moatis they would recognize.

[01:24:32]So how do we test whether there is active lectins in our matrix? Again we need to start off with uh physiological extracts of proteins. So we need to solubilize our proteins in physiological buffers and we use something that is called the retrautination essays where we incubate our lectins with red blood cells and whether the lectins are present and they are active they would start cross-linking specific carbohydrate on red blood cells. And basically what happens is that the negative control we don't have lectins we would see actually a spot a red spot which you see here at the bottom of our well actually if the lectins are active and they start cross linking carbohydrates we would start seeing actually a network and is what you would see here which is um a cover that on the on the well that would um suggest the presence of the lectins and this method is also semi-quantitative.

[01:25:37]So just to summarize uh for advanced identification the coordination of all this information is basic integrating filogenetic analysis and omix would help moving beyond the basic analysis uh which would make the information way more um informative and useful and also the development and validation of multi-analytical approach for active anti-nutritional factors quantification is actually crucial.

[01:26:08]Because of this issue and the protein solubility um and bio accessibility, we really need the strong methods to be able to test um these anti-nutritional factors in the alternative ingredients like for example um uh single cell proteins or even more new ingredients may come up from the bio um from biotechnology. ology and um we have seen that actually even processes like fractionation or processing can actually influence anti-nutritional factor profiles. And so the use of invitro to this for um uh simulating gastrointestinal digestion and understanding the effect of fractionation and processing on this anti-nutritional factor would help validating protocol for effective antiutritional reduction.

[01:27:09]So thank you so much for the attention.

[01:27:16]>> Excellent. Thank you Kiara.

[01:27:18]Uh so I hand over to K uh Kiara and Claire. There are a number of questions uh in the box. You can actually start typing some of the answers because most of them are about the risk assessment and we can go through uh through them uh after last talk after Kan's talk. Okay.

[01:27:40]Uh Kan that looks good. Very good.

[01:27:42]>> Okay. Thank you Andre. Good afternoon everyone. I will try to add the last layer of of this session. And uh in my presentation I would like to to take you through how do we measure actually protein bioavailability in vivo in humans by the means of stable isotope. And for this uh for this task I will use the the example of fabin um which is a paper that has been published two years ago. if you want to to have a deeper look. Um, so I'm actually a post-doctoral researcher working at In Agroparite Tech in the research unit of uh nutrition physiology and eating behavior and um and I will I will go through the the presentation. So uh before we go through the method uh of measuring protein bio availability, I would like just to refresh you a bit with the concept of uh of protein bioavailability.

[01:28:45]So so um first you have the protein intake.

[01:28:50]So here we take the the example of of fabins. So the protein from fab beans as was said today um earlier today they will enter the stomach. In the stomach you have the combination of acid and pepsin that will contribute to protein hydraysis and from that you will have the formation of peptide and polyeptide from the initial protein and some free amino acid will be released. Then the digestion will continue in the small intestine. In the small intestine, you have the secretion of pancreatic enzymes and you also have brush powder enzymes that will continue to chop the peptides into amino acid but also as Linda said in some d and tripeptide and um and this smaller molecules such as amino acid they will be absorbed and will enter the bloodstream.

[01:29:47]However, some of the some of the some of the amino acid will um wait, hold on.

[01:29:56]Some of the amino acid will escape digestion by the end of the small intestine and will end up in the colon.

[01:30:03]In the colon, you have a microbiota that is capable of fermenting uh this uh protein breakdown product. But um the the product generated by the microbiota will not be directly used by the host for its own needs. So we consider this fraction of protein to be to be lost and we call it the digestive losses.

[01:30:27]Um of the amino amino acid that are absorbed most of them will be used for um for the for the body requirements such as protein synthesis.

[01:30:39]But of this absorbed amino acid some will be catabolized and transformed into ura and will be eventually excreted in the urine. So these amino acid that are catabolized are lost and we call it the metabolic losses.

[01:30:57]So this leads me to the definition of protein bioavailability as we quantify it in our invivo studies. This is actually the protein intake minus uh the digestive losses so the fraction that is not digested minus also the fraction that is uh catabolized and transform into ura after absorption.

[01:31:22]So in the title of my presentation I also mentioned that to quantify protein bioavailability in our case we use stable isotope. So what are stable isotope? they are uh non-raactive atoms and here is the example of hydrogen.

[01:31:42]So in uh in nature the most common isotope of of nitrogen is proteium. Um that's the the hydrogen we we we find mostly and it's it's uh its atomic weight is one because it's composed of one proton and zero neutron. And then there is the heavy uh isotope of hydrogen which is called darium. And uh this darium has an atomic weight of two because it contain one proton and one neutron.

[01:32:17]And in nutritional studies, but also in in in other biological studies, those um stable isotope such as 15N or 13C are used to track molecule of interest in the biological system and in our case it's in the in the human body.

[01:32:40]Um today we are focusing on protein bioavailability. So the molecule we are actually trying to track is amino acid and you can label this amino acid with stable isotope at different position.

[01:32:57]You can label the carbon. You can also label the hydrogen or the nitrogen. And in the case of giant lips we labeled the nitrogen atom with 15N.

[01:33:12]So how do we label actually the the amino acid and the the protein of the the plant of interest? So in this case fabin. So here is a is a schematic representation of of fabin that is growing and you also see the the picture that was taken at chas. So what they did is that they sprayed the fabin with a fertilizer that contain the stable isotope 15N. And by spraying this uh this isotope the plant will integrate um the this isotope in into its own into its protein. Sorry.

[01:33:54]And how does that look like? Well, here what you see is the amount of atoms that are actually 15 N atoms out of all the nitrogen atoms for the label fabin. So here we reach uh 15N atom percent of 1.2. And you may wonder whether that's that's a lot or or not. Well, if you compare it to the natural abundance, so how much 15 N atoms would you find in nature without enriching the system? you see that it's around 04.

[01:34:29]So it means that we actually multiply uh the enrichment by three by by three times and uh this is sufficient for us to to track the the protein of interest during digestion and after absorption.

[01:34:45]So once we have labeled the food uh that we want to test in our invivo study um we we have to uh place a nazo illeal tube in our subject. So our the nazoil tube is a tube that goes from the nose until the end of the small intestine.

[01:35:06]And as I said in the in the first slide, we target the end of the small intestine because that's where we consider the uh enzyatic digestion by the host to be completed.

[01:35:18]And if you look at it in a with an X-ray, this is this is how it looks like. So you see here the tube uh in a in a subject and it start making loops because it's actually um following the shape of the small intestine.

[01:35:36]Then if you zoom at the end of the tube, this is how it looks like. So we have uh at the end of the tube a ballast. The ballast is actually a way to give some weight to the tube and this will help the migration of the tube throughout the gastrointestinal tract until we reach the target position.

[01:36:00]Before um this ballast we have a collection zone. So the correction zone is simply a tube with some holes and with those holes we are going to pump out the digesttor uh from the from the from the intestine before uh this collection zone. So a few cm before what we do is that with another tube we are going to infuse uh an indigestible marker in our case polyent decol and we are going to uh catch back this indigestible marker with our collection zone and by measuring the dilution of our polyatylenol in the digesttor we will be able to quantify uh the intestinal flow at the collection zone. So we have an idea of how much digesttor is passing at the collection zone and we can determine the amount of nitrogen that is escaping digestion.

[01:37:06]Um so this is our sample collection scheme. So the day before uh we start the experiment the volunteer they they come to the hospital and they are intubated by a trained gastroenterenterologist.

[01:37:21]We check the position of the tube by radiography at around 8:00 p.m.

[01:37:29]the volunteers they receive their last meal and then they go for an overnight fasting to empty the proximal gastrointestinal tract. And in the morning around 10:00 a.m. uh we feed them the test meal that is labeled and contain the protein that uh that we want to we want to to assess and then for 8 hours we are going to sample uh different um different types of of samples. We are going to to collect the digesttor and with this digesttor we are going to quantify the digestive losses.

[01:38:07]Um with these digestive losses we can uh determine the digestibility of nitrogen and nitrogen here is used as a proxy for protein digestibility and we also measure the the digestibility of single amino acids with the urine and blood sample. We are going to uh look at the metabolic losses and those metabolic losses as I said they are representing the fraction of absorbed nitrogen that is catabolized and transformed into ura and with that uh with the digestive losses and the metabolic losses we can calculate the bioavailability of of a protein of interest.

[01:38:56]So uh this is a video of of a real live intestinal sampling. So you see the volunteer that is incubated with this nazoial tube and we are just pumping out the digesttor from the body using a siren.

[01:39:15]So once we collected our sample so as I said we collect intestinal content, blood and urine. We are going to analyze them in the lab. And in our case, most of our anal an analysis they rely on isotope ratio mass spectrometry.

[01:39:32]Um and this uh this kind of equipment unable to uh determine the ratio between 15N and 14N atoms in our samples and from that ratio we are capable of telling how much uh dietary and endogenous protein is in the sample.

[01:39:53]Um if we have a closer look at how the method works. So here you take a sample in this case intestinal content. It can also be blood or urine. And um this uh these samples contain a mixture of dietary protein and endogenous protein.

[01:40:14]Um so the sample will will first go uh through an oven where the where combustion will occur and this will lead to the formation of uh nitrogen oxides and this uh NOx will be reduced um and will form uh N2. So uh this N2 this uh we are we are forming a gas. This gas will be um transferred towards the ionization source. And in this ion ionization source we are going to um to impact uh the the N2 with electrons in order to form ions. And during this ion formation we will obtain 14 and 15 N2 ions. And uh these ions will be accelerated and then they will be directed towards the detectors. And uh on the trajectory we have a magnet that helps to separate the two uh ions and it will also focus the the ions to the associated detector. And by measuring uh the quantity in each detector, we are able again to obtain our ratio between uh 15 and 14N. And with that uh with that ratio as I said earlier we are able to to quantify the quantity of dietary protein in our sample.

[01:41:46]So with that said we are going to look at the digestive losses.

[01:41:50]Uh here u what you see is the the digestive losses for dehold and cooked fabin. So the graph that you see here is the dietary nitrogen flow rate at the end of the tube. So at the collection zone and uh so when we start feeding the volunteers so at t0 there is no dietary nitrogen because the food is just ingested and slowly you will see that the dietary nitrogen will appear at the tube and um and at some point there will be a decrease because digestion is completed.

[01:42:29]So this dietary nitrogen fluorite is representing the amount of dietary nitrogen that is escaping digestion.

[01:42:38]So if you sum all these points on the curve um you can calculate what we call the digestive losses and in the case of the held and cooked fabin they represent 16% of the ingestive nitrogen.

[01:42:54]And if you subtract these digestive losses from the ingested amount of nitrogen, you can calculate uh the true ilial nitrogen digestibility which we also called the true protein digestibility. And for fabin 100us 16 give you a digestibility of 84%.

[01:43:20]So I just mentioned that we we do this uh quantification of digestive losses for the nitrogen to get to the digestibility and you can apply the same principle for every amino acid. So you can measure the digestive losses of each amino acid and in this way you can calculate the true uh ilial amino acid digestibility.

[01:43:45]And with this digestibility you can calculate uh the score that Andre mentioned in his introduction the DAS which stands for the digestible indispensable amino acid score which is the score recommended by the FAO to assess the nutritional quality of of protein.

[01:44:06]Um so now we've we've looked at the digestive losses. Now we are going to look at the the metabolic losses.

[01:44:15]which are uh which represent the absorbed nitrogen lost through ura and in this case we are looking at two pool the blood which is actually a proxy for the total uh body water and the the urine so I'll come back to that in the in the next slide so here what you see is the amount of dietary nitrogen recovered in body ura over time and you see that at start of the digestion there is no nitrogen from the food catabolized into ura in the body and this will increase throughout time and then it start stagnating or or even decreasing and as I said here we measure this uh this body ura in the blood and we extrapolate to the to the rest of the body to quantify the total amount of uh of uh ura from the food that is still in that is present in the body.

[01:45:18]And you see that for fabinh after 8 hours we still have 12% of the ingested nitrogen that is present in the body in the form of ura and um we also have another pool which is the dietary nitrogen in urinary ura.

[01:45:39]So actually this pool here from the body is being emptied in in the urines and you see here the cumulative uh recovery of dietary nitrogen uh in urinary ura uh for the 8 hours after feeding. And you see that at 8 hours we recovered 6% of the ingested nitrogen in the urine in the form of ura.

[01:46:09]So if you want to quantify the metabolic losses, what you have to do is um is quantify the amount of dietary nitrogen in the body ura after 8 hours that is still present and you also add to that the cumulative amount of nitrogen that you actually recovered uh in urinary ura over the eight hours. So for our fibin example, if you add 12 and six, you get to uh to 18%. So the metabolic losses of for fabin for fabin here represent 18% of the ingested nitrogen.

[01:46:55]Um so this lead me to to the the conclusion or the the calculation of the protein bioavailability using 15N stable isotope. Um what we saw is that uh we measured the digestive losses um and for fabin it was 16%. And we also quantified uh the metabolic losses in the body and in the urine. and for fabin it represented uh 18%.

[01:47:28]So if you plug those numbers in the protein bioavailability formula that I showed uh in my uh intra second slide, you can calculate this number and if you subtract 100, so the protein intake minus the digestive losses minus the metabolic losses, you get to a protein bioavailability of 66% for the famine.

[01:47:55]So with that said, uh I thank you for your attention.

[01:48:04]>> All right. So I would like to thank all speakers. We have a bit of time. We have about 10 minutes. Um and we have a number of questions. If you wouldn't mind showing yourself there, we might start with the easier one. Linda, is that okay? You typed a response in there. I just read out the question. Um um as bioavailability is a major issue for food peptides to harness their full biological potential at the target site inside the body. What real world strategies can be employed to enhance their bio availability and accessibility and stability and Linda already typed typed an answer there. Is there anything else you want to add Linda or go for that matter? I suppose all I wanted to say is that we are looking at how um protein and food is digested, how it's absorbed and how it's bioavailable. We are not looking to uh extend its halflife in the circulatory system uh to ex to extend and make sure that it is uh bioavailable and reach the target organs would be something that the pharmaceutical industry has been working a long time at. And they would have things like osmpic which is a GLP-1 peptide. And their mechanisms for extending the halflife the circulatory halflife of peptides includes amino acid modifications adding fatty acid chains to bind stybon maybe covering it in PEG.

[01:49:35]Um these are strategies that are used by the pharma industry but we're looking at how proteins in food are digested.

[01:49:43]>> Yeah that's okay. Thank thank you Linda.

[01:49:45]uh if you have any further questions you know even after the uh webinar either send us an email or you can also type it in the YouTube channel if you if you wish um I just go to the next one Kiara uh it is about the risk assessment of anti nutritional uh factors in vitro so you uh typed your answer an extensive answer do you want to add to that >> Kiara yeah so just yes I can add that actually um when we approached this and we approached this issue, we started looking at mainstream ingredients that were already used like soybean as being a comparator. But what we realized later on is that when we look into anti-nutritional factors, we actually needed to look at those foods that of that anti-nutritional factor would have high level and I use those ones as benchmark and this is what I just studied as would be lectin since beans and common beans or um parise inhibitors in potatoes uh which are quite rich of that. Yeah, very good. Uh there's another question for you and Claire um from Jessica Schwarz. Um for evaluating allergenicity risks, what do you think about using size exclusion chromatography for characterizing digestive peptide size?

[01:51:16]Claire, >> yes. I I wrote and a response to Jessica in the chat, but basically I think GPC is a useful method if you've because you can retain dulfide bonds and if you do it under physiologically relevant conditions if you've got non-covalent interactions that hold peptides together in larger aggregates that is very useful and you might want to couple it with other types of analysis whether that's SDS PAGE or mass spectrometry but I I think it is a useful method and it's been used to characterize hydraulicates in allergenicity um before >> yeah very good there's another question by Sophie Kendler uh how does EPSA include the invitro digestibility and risk assessment in their recommendation are those recognized methods so I basically typed in the response in terms of uh the digestibility so the invitro das and digestibility that is happening at the moment so the paper has published the ISO method has been submitted has already gone through two rounds of um uh revisions. So I'd say it will be published within this calendar year. And Claire, you mentioned that food allergy risk assessment that there are several guide guidance documents. Uh and you can also look at opinions. Is that correct?

[01:52:37]>> Yeah. So they look I put the guidance links in the chat thing. Um >> Oh yeah, you did. Oh yeah, I see that now. Yeah. Yeah. And because there are several there and and I I I think I I think you need to go and ask Essa.

[01:52:54]>> Yeah. Yeah. Yeah.

[01:52:56]>> But there is a lot out there and it's what you always have to remember is it's a scientific process and you need good science and if it's something that you can get you need scientific high high quality scientific approaches and if you stick with that you'll be all right.

[01:53:14]It's when you're people try to cut corners and the science isn't sound that there are challenges and difficulties.

[01:53:22]>> Yeah. Very good. So just to mention uh at the peak of the webinar we were about 80 people u present live. So that's very good. Thanks for everyone you know sticking to the very end. Now uh so we have uh three more minutes. Uh there's one question by Antonio uh Fernandez Dumont. Interpretation of out of outcome of GI models in the context of risks assessment. So protein toxicity and allergen allergenicity prediction might be challenging. Oh yes they are. Expert judgment is crucial but what do you think about developing criteria for the interpretation of GI models for risk assessment purposes would be possible?

[01:54:07]Thank you for this event. very informative. Claire, you started typing there already. Do you want to add anything? Well, it it really I mean um it it is a it is a technically challenging thing for the toxins and it depends where you're coming from because the way in which the digestibility is used in for example newly expressed proteins in genetically modified organisms is slightly different because there you have things like Bt toxin that are actually introduced into the plant and if it's there in the edible tissue to you know understanding how that gets broken down and I actually think that there are ways you can actually use the methodology that we've got and to calculate half lives and look at persistence and put numbers on there to actually say well how much gets exposed to the mucosa you can actually do those calculations and I think that would be a very interesting experiment to do a thought experiment um based on information that might be there some of it is not in the public domain know um for some of those they they would be proprietary to the a biotech industry.

[01:55:18]So I think you can do that for that it it that's quite simple I think for allergenicity risk assessment and actually Antonio has written specifically allergenicity prediction. I think that this is much more complicated and we already know that allergenicity is a feature not simply of the material that you're giving to someone but a feature of that person's physiological predisposition. And I think that is is is much more complex um to actually interpret. Um but it might be that if you do have foods that are very very readily or molecules are very readily digested um they will pose a lower risk in the same way that you know otherwise you would not I mean hypoallergenic extensively hydrayed formula you know this is used for kids who are allergic and you can make a hydraicate that still has significant peptides left that actually they will not react to so you know it I think there is a way there through of at least saying this is very low risk this is the one you go to the next step I don't know whether that helps makes any sense to you >> just about clear >> it is complicated >> yeah I just have a one last question for Kiara one minute yeah one minute that's a question from Andrea Ria de Rio Have you also looked into the role of polyphenols as anti-nutritional factors?

[01:56:58]>> Um so yes we have been quantifying and profiling polyphenols in these ingredients. Uh nevertheless the ingredients that were analyzed within the project um did not contain very high level of polyphenols. Mostly these are fractionated. So within the context of giant lips I cannot say that we have done that in separated uh projects which are side projects where we have working primarily with lifts for example we have seen quite dramatic reduction in digestion.

[01:57:34]>> Very good. Okay. I think we are done. I would like to thank all the the presenters. Uh really nice nice talks and uh everyone still present. There are about 60 people present now between uh Zoom and and um and YouTube. Thanks for your patience and sticking around and thank you all for your attention. Okay, bye bye. Bye. Thank you all. Bye. Bye.

[01:57:57]Bye.

[01:57:57]>> Bye.