SalmoSim: Building a salmon gut from scratch
What does it take to build a healthy, synthetic salmon gut? When it comes to fish nutrition, a lifetime of health and performance can be greatly influenced by the early stages of the gut microbes. Dr. Martin Llewellyn and Raminta Kazlauskaite of the University of Glasgow are creating new ways to improve sustainable fish feed and drug stability for salmon aquaculture farming with SalmoSim.
The following is an edited transcript of Tom Martin’s interview with Dr. Martin Llewellyn and Raminta Kazlauskaite. Click below to hear the full audio.
Tom: We're entering a new frontier in fish nutrition and among those conducting cutting-edge research in the field are Dr. Martin Llewellyn, a research scientist at the University of Glasgow in Scotland, and PhD candidate Raminta Kazlauskaite. Llewellyn, author of over 50 peer-reviewed research articles, has expertise in salmonid parasitology and nutrition. Kazlauskaite's focuses are in the fields of bioengineering and molecular biology, and together, they have been working on creating an in vitro system replicating the Atlantic salmon gut. They call it SalmoSim and they're here to talk to us about it. Thank you for joining us.
Martin: It's nice to be here.
Raminta: Thank you.
Martin: Thanks for having us on.
Tom: What is the problem that you set out to address resulting in the development of SalmoSim?
Martin: Atlantic salmon are a funny old fish. They're a carnivore and we don't farm many carnivores, so there's a big problem around sustainability of salmon feed, as well as its price because what you have to do to feed a salmon is you have to go out -- when we first started farming salmon, we'd have to go out and catch an awful lot of wild fish, grind it up, and feed that to the salmon.
These days, things are changing quite a lot. We're feeding new feeds to salmon all the time with a focus on insect-based, plant-based feeds including things like soy, gluten, protein. What this means is we're basically challenging the salmon gut every time with things it's not entirely used to. There are so many different feed additives in different protein sources out there on the market that there's an awful lot of in vivo testing. That means testing of salmon to see what they're happiest on, what they're healthiest on, what they grow best on.
Essentially, we've built this system to act as a pre-screening tool because these trials are really expensive and there aren't many places that do them, so it's a massive bottleneck. You're looking at estimates for a single trial of around £150,000, maybe $200,000 just to try a couple of different ingredients in the sea cage to see whether your fish are going to perform well on your feed. What we've developed the system for, as well as to do pure science because we're from the University of Glasgow, we're not strictly a commercial entity, but as well as doing pure science, we'd like this system to be a useful way for people to triage or pre-screen, so you come with ten different alternatives or ten different ingredients and be able to narrow that down to two that then take on to a trial in real salmon and reduce the cost of the whole process of getting from these new feed ingredients, new additives that people dream up in the lab all the way through to having the right feeds to bring up healthy, happy, productive fish.
Tom: What are you seeing in terms of cost savings?
Martin: At this stage, we're right at the beginning. We know that we could do it substantially under the cost of a current feed trial but where ours doesn't completely replace a feed trial in vivo. It's one of these things that's kind of like for like cost saving. It's hard to estimate that. I don't quite know what that is at this stage, but it's likely to be very substantial just because the cost of what we need to do in the lab is just really a tiny fraction of the cost of what people have to pay for these in vivo trials.
Tom: This technology is created in vitro. Do you want to tell us more? Raminta, do you want to talk about that?
Raminta: Yeah. What we do, we literally just transfer all the salmon gut inside the lab. We have three bioreactors, but they represent three different salmon gut pieces, stomach, pyloric cecum, and midgut, so it's similar to us humans like stomach, small intestine, and large intestine. What we do, we literally just take bacteria from these three different gut compartments and transfer them into three bioreactors or jars and then we set the right conditions and we get it running.
Tom: I was going to ask you what it looks like in the lab, but I guess you've just described that.
Raminta: Yeah. Literally we have a feed bottle. We have three bioreactors and then we have waste and it's continuously going. In other words, we just use SalmoSim to produce really expensive salmon poo.
Tom: Okay. If you could expand on that and give us an idea in a practical sense how this tool is used.
Martin: First, I'll tell you what a bioreactor is. A bioreactor is like a very fancy fermentation vessel and a lot of the people that we work with at the University of Glasgow work on anaerobic digestion of food waste. All the stomach is essentially is an anaerobic digestion tool, so we've taken that same technology of these closed units where you can change the pH and monitor the pH, change the temperature, monitor the temperature.
We're putting in enzymes directly extracted from salmon. Also, we're putting in the microbial communities the bacteria, as Raminta says. What we would do when we get a new feed, what we tend to do is stabilize the system on a feed, our control feed, and then change the feed and then essentially begin to introduce that into the system, which is essentially grinding it up, feeding it through the pipes, and beginning to take it into this three-compartment model, so it'll go into the stomach. It'll be exposed to these different pHs, enzymes, and then it'll get transferred into the pyloric cecum. Again, there'll be different pH, different enzymes it's going to get exposed to, the microbes as well, and then moving on to the midgut. It'll get exposed there.
If we're looking at drug stability, let's say we were taking one of these sea lice drugs, and quite often, some of them are in-feed, some of them aren't in-feed, but with an in-feed one, you absolutely want to make sure that it's not getting released in the stomach, so there's very little absorption that happens in the stomach in Atlantic salmon. Most of the absorption happens in the pyloric cecum. If your capsules, if you like, if your microcapsules you've absorbed your drug onto are releasing all that drug and the drug is potentially getting degraded inside your stomach compartment then really you're losing a lot of efficacy with the delivery of your drug dose. So what you want to do is make sure that the highest pharmacologically relevant concentrations of the drug are in the pyloric cecum. Again, we could take various different combinations of the drug, maybe try them alongside different kinds of feeds, for example, and see which feed combination or encapsulation combination delivers the best dose of drug to where we want it to get it to. That's I guess an example of how --
Tom: So that's the ultimate holy grail that you're going for?
Raminta: No. We look at [0:07:03] [Indiscernible] stability hopefully, as well as probiotic survival in different gut compartments and how do we affect bacterial communities, VFA analysis.
Martin: Certainly, anything that you want to do to a real fish, we can give you a sort of an early warning system to rank the different possibilities in order of their likely efficacy in an in vivo model.
Tom: How could a greater understanding of these processes reveal pathways to improve growth efficiency of fish fed on plant-based diets?
Martin: Lots of different ways. The first and most obvious way is I think probably digestibility, so just how readily large complex organic molecules like protein is broken down into smaller organic molecules like amino acids and then presumably absorbed by the fish.
The slightly further on the line things are the impacts on the microbiome. So at the moment, we could detect whether there was a significant perturbation in the microbial communities by bringing in sort of a plant-based feed. The science is less developed there, but if you're shifting your microbial communities away from a stable, potentially complex microbial community to one microbial community dominated by a small number of microorganisms, that's normally a bit of a red flag. Those types of communities are normally more easily invaded like pathogens, for example. There are also other things beyond digestibility related to the microbial composition of the gut that I think we can predict with our system in relation to what happens if you feed them on a soy meal-based feed certainly for plant protein. That's the main thing.
Raminta: For feed, yeah.
Tom: Alltech as a leading agricultural biotechnology company has been making important inroads into the aquafeed sector. What is the company's connection with this research that you're conducting?
Martin: We were dreaming this idea up about three years ago. There's a guy called John Sweetman and someone else called Philip Lyons and these guys have both been working in aquaculture for a long time. I was interested in this initially as a tool to understand what are the ecological processes that underpin microbial community assembly, so why are microbiomes like what they are from a pure academic perspective, but talking to these guys at actually this kind of conference where industry and academia are brought together and you get these lovely link-ups and explosive, potentially explosive, disruptive combinations then we got to talking about this. They were enthused. They saw potentially some of the early promise for some of the applications I've just described to you, so Alltech very kindly funded a PhD student, and that's Ruminta.
Yeah, Alltech have been really a major catalyst for actually getting this idea off the door and brought it into reality. Ruminta actually made that reality happen. I never believed we'd get as far as we have today. It's all Ruminta's hard work.
Tom: I'm really interested in something you just said too about this conference and how it brings together industry and academia. Have you seen sparks fly this time?
Martin: Yes. Certainly, I've had some interesting conversations around the place and you can see interesting conversations are being had I think across all different sectors. I'm really pleased to see there's increasing people talking about sustainability in livestock production that's all the way from agriculture to chickens and beef, et cetera. I'm particularly impressed by how far and how much industry is recognizing that as important with an aquaculture session, so I sit within my aquaculture sessions, having sat with a few other sessions and I really think we in aqua -- I don't want to blow our trumpet too much, but we're really thinking about is what we're doing sustainable.
We've got these brilliant feed conversion ratios compared to huge amounts of other -- we're down at FCRs of 0.8 whereas poultry, I don't know, but they're around 3. We're talking about ingredients, where are they coming from. We can't rely on marine protein, so there's a real environmental responsibility, I think, that is there in the aqua sessions and I really hope it's building in the other sessions too. I think there's been lots of really good open discussion around that here, which is being really encouraging, I think.
Tom: Yeah. I think it's safe to say that sustainability has quickly become something of a watchword or a guide star for [0:12:04] [Indiscernible] our clients.
Martin: Yeah, that's right.
Tom: What's beyond this research? Are there uses for this technology in other species, do you think?
Raminta: Definitely. It's just time to validate and to see what is happening inside of a SalmoSim or other species that are representative of what happens in real fish, so just time and money, yeah, but it's definitely, definitely possible.
Martin: Yeah. We've taken about two years to get this far with Atlantic salmon, but we've learned an awful lot along the way, so I do think it is transferable, but like Ruminta says --
Raminta: It just takes time.
Martin: It takes time, yeah.
Raminta: And some resources, yeah.
Martin: Yeah. You could apply it to tilapia. You could apply it to trout, just a bit of time, a bit of resource, and demand really.
Tom: Dr. Martin Llewellyn, a research scientist at the University of Glasgow in Scotland and PhD candidate Raminta Kazlauskaite.