ROBERT TANGUAY: Okay, thank you all for sticking
around. Getting ready for the good session when we
get to actually talk amongst ourselves. So, I am actually a molecular toxicologist
by trade, so what I do for a living is try to identify biologically active molecules
and then understand how they act biologically. I am not going to talk about that today, but
instead I am going to give an example of … many of us, we’ve heard how we stumbled upon
the need for nutrition for a number of other drivers. It really wasn’t the nutrition question
in most cases, it’s just “Oh, man, we need to consider this.” So, I think I am in that category, as well. Okay, so I am going … to talk a little bit
about my experience in developing and testing. And when I use the words “defined diet,”—Zoltan,
and I actually really mean it—a defined diet … And we’ve actually—and I’m
not going to show it today—we’ve also done complete chemical-defined diets, and
I will tell you, that was not very successful. We worked on it for about 3 1/2 years, and
that was for some SILAC studies we’re trying to do for stabilized stone labeling, and we
just didn’t get the recipe yet, so I won’t talk about that today. But it’s still something I think can be
done, it is just going to take a little bit of more work. So, I am going to talk about some of the challenges
with our defined diets and some of the impacts. And then, another one of the drivers, some
of you know this about our lab, is we’re really worried about biosecurity for a number
of reasons. We’re very concerned about the influence
of infection on research results. I know that is the topic of a different workshop
that this Institute handles, but I am going to talk about how the selection—the diet
really impacts your ability to handle the biosecurity, and that’s one of the drivers
that I am going to show you and some of the challenges and concerns, again, the impacts. And then finally I am going to put a toxicology
hat on just for the last slide or two to talk about the … really, how do we consider … Mark
talked about a little bit about toxicants, and as a toxicologist that was a nice primer,
but, I mean, it is a massively more complicated than what we suggested, so I’ll get into
that. So why did we want to define a … why did
we want to make a defined diet? We actually wanted to develop … evaluate
an individual component. We heard about that today, and one of my papers
was highlighted as one of the examples of quote “doing it right.” That wasn’t my doing; that was actually
a nutritionist, Maret Traber, who is an expert on vitamin E, who really helped us. So, we worked on that for—and we’re still
working that area—for over a decade. And so, what does it take to develop a defined
diet where you can actually isolate one component and ask the biological impact or the molecular
impact of that? And then some of the challenges operationally. We talked about this all day today is … What
is essential? What is normal? What is optimal? And I would say that nature doesn’t know
this either, right? You know, we want to benchmark it to—we
heard from our aquaculture experts in the room—we want to benchmark it to what they
normally eat. Well, they normally eat because what’s there
is what’s there. That hasn’t been selected in some top-down,
grand designer about what a fish needs to optimally survive. And when we think about what we’re trying
to do, all of us have different experimental objectives, and we now know that the diet
influences many of those. And—Steve mentioned this earlier—there
is a massive education effort that needs to happen. We’re not going to necessarily, at this
point, I view, dictate what people should be doing, but they need to understand that
diet influences almost everything they measure. And that is not widely accepted. We saw that in the survey results. Okay, so. And then we talked a little … Chris talked
about this a little bit, as well, the performance of the diets under the conditions that were
constrained with our design-efficient systems. See, these recirculating systems, they’ve
certainly massively expanded our capability and our ability to do more experiments, house
more animals and more types of animals, but they come with certain features that really
can fight against you when you’re trying to modify your diet formulations. So, we play around a lot with the formulations
and buoyancy in a very systematic way. We have our own diet kitchens, and there is
a lot of work that can be done in this area still. But, as you know, no one is really funded
to do any of this, so this is all kind of done as a hobbyist, per se. So, for the first diet that I am going to
describe—and we’re not going to go in in great details—but we did carefully go
through and try to identify, not only from the literature … from a lot of the aquaculture
literature, and some of the rainbow trout work that was done in the building that I
direct now, and really try to adapt it to zebrafish. But, importantly, identifying sources, and
then the purity of those sources. I can’t emphasize it enough from a toxicologist
how important it is to have a control over the quality, the stability, the lability,
the storage half-life, of everything. Because every … I mean, it’s in the diet
because it’s essential, but that means it has to be there. So if you have batch-to-batch variation, you
don’t have control of your diet. If you have a labile ingredient, we have examples
where we make defined diets and we’re asking for a component, and we’ll get a message
like, “Yeah, this one only has a 6-month half-life.” Okay, great. And we’ll order it in January, and we’ll
order it again in June, and we get the same batch with the same blot number. Okay, thank you. So that means you’ve shipped us at least
one expiration and maybe 100 expiration days. Is it labile or not? And if you do not measure—and nobody wants
to do that, nobody can do that, to measure the constituents and their integrity or their
purity over time. That is really not what anyone of us is really
trained or really want to do. And then so we … and then, so, making the
mineral mixes and defining those and then the vitamin mixes, defining those. So, this is what we worked on for a number
of years, and just so, again, I’ll go back real quick, the protein sources, we have … I
can’t go back in … its stuck. We’re stuck. Well, we’re dead. Do I have to escape out? Yeah. UNIDENTIFIED MALE: Maybe the mouse? Alright. Okay, good. So we had various diets where we had egg white
as our main protein source, and then we had casein as another source. And again, like we heard earlier, some of
those sources actually could have variations in them, as well, if you’re not … if you
don’t quantify them. Alright, so when we started developing these
diets, this was the first control experiment that we tried to do. We wanted to take … I want to point out,
we used this wild-type strain that I developed over a decade ago now, which its intent was
to have a diversity in the genetic stock in this, as this has a significantly higher diversity
of the … of some of the more inbred strains that are commonly used. And we’ve recently sequenced this genome
and demonstrated that diversity. So, our population is more diverse than most. And we used that for some screening purposes;
that’s why we did it. So, we’ve got large populations of these
fish, and then we split them up, and one thing you’ll notice is that we’re weaning at
1 month of age. We didn’t want to do this. So, we made a defined diet, and we wanted
to feed these guys right after they start feeding at 5 or 6 days, and this diet was
not sufficient to accomplish that. So right out of the gate, we could not isolate
the vitamin E because we couldn’t … the fish were not eating … they didn’t like
the taste, it didn’t look right. So, we figured out that about 30 days into
life, we can shift them over to this diet, and then they … then they would thrive on
it. So even in our … lot of effort, we still
haven’t solved that first part … So that’s also going to delay our studies. So, vitamin E, as you know, is a lipid-filled
component, so this … it takes time to get it in, and it also takes time to deplete it,
so we had to calculate those depletion kinetics. So then we’d feed them, and then about at
3 months of age, we start the spawning. And on the bottom, you see some initial experiments
to see, well, how is this diet doing? So, we looked at when the … after the embryos
are collected, we’d plate them—we do everything in an automated environment—so we’d plate
them in 96-well plate, we’d assess a number of automated morphological features. And then in this case, we collected RNA at
the 36-hour time point, and then we observe phenotypes up until the end. And so, we collect a massive amount of data
of every individual embryo that came from parents that were fed these diets. And just—it’s not important to look at
the top—but basically, if you look at the bottom, the affected individuals … you can
see the animals that are E-minus diet, even at 24 hours, you start seeing some effects
on the embryo and that increases a very high percentage—75, 80 percent are dead or arrested
development by 72 hours. And then our E-sufficient … so, our defined
diet that we supplemented with ethyl tocoferol base, they’re okay, but you can notice it’s
higher than background, so these guys were affected. These fish are not completely normal. And … but this is here just to demonstrate
the E deficiency. So, we were able to deplete … so we have
to … we deplete the parents, right? We can’t feed the embryos yet, because they’re
not eating. So, we feed the parents for … at 2 months
up to 3 months of age, and then the embryos that are produced are massively deficient
in vitamin E, and that deficiency is leading to these developmental consequences. So that was kind of the experimental system. We worked with this for a number of years. So, I won’t talk a lot about gene expression,
but this is just a first peek. So when the animals are indistinguishable
from each other—the E-sufficient, E-deficient, and our laboratory diet—we just sample embryos
and do full embryonic RNA sequencing at a very deep level … a very deep, deep sequence
number, and what we see, this is what … lots of ways to look at it, but we compared all
of the diets to the E-plus. So, you can see in the lab diet compared to
the defined diet that has enough vitamin E, there are still hundreds of gene expression
changes, and this is from a number of pooled replicates. And then if you actually look at the E-minus,
it is even greater. So, the E-deficiency in combination with the
defined diet has a massive impact on expression. So, again, it’s not that your individual
gene might be affected, but the point is that there are a lot of networks that are impacted. And the potential crosstalk of these disruptions
just by the diet could influence your interpretation of your experimental results if you are taking
a molecular approach. So, I just wanted … I brought this up recently. So, we heard earlier, so I actually run the
Sinnhuber Aquatic Research Laboratory, which has a long history in developing rainbow trout
diets, and it’s very different now. So, this entire 18,000-square foot facility
was all rainbow trout aquaculture and cancer studies. And so, I actually named the lab when I took
over the directorship about 15 years ago in honor of Russ Sinnhuber, who actually developed
the trout diets that was talked about earlier today. So, the reason I bring it up is it’s … not
only it’s a huge facility—we have four zebrafish rooms, we have a quarantine room—but
we also have a specific pathogen-free environment in about 70 percent of the working area, the
lab. So that means we want to control everything
that gets in, and then we want to make sure that we’re not introducing disease into
our system. And so that actually creates a lot of challenges. So, what is it? So, its initially just specific pathogen-free
for Pseudoloma. Again, we instituted over a decade ago a strict
biosecurity program, so we need it massive for the high-throughput screening part of
my lab. We need 10,000 to 80,000 embryos every single
day, and that’s Monday through Friday and then some on Saturday and Sunday, as well. We have a lot of experience in doing feeding
studies, and I am always trying to drive—and many of you are doing the same thing—drive
labor cost down, so that was the main driver for me: It was to get the labor cost down. So, we removed the small live feed. So, this is in 2010, we stopped feeding paramecium
and rotifers for the little … the little fish. And this was for a number of reasons. There are biosecurity problems that are well
understood by many, and then also the labor involved, and so we wanted to get rid of that. So, we have successfully done that for 8 years
now. But then … but when we tried to remove Artemia
with our traditional commercial diets, it was not successful. So that combination of pulling the Artemia
and then going to that flake diet was not successful. So we said, let’s try again. So, again, the Artemia … why are we worried
about Artemia? Many of you know this. The quality varies, the cost varies, the shortages
are a problem, massive problem for biosecurity breaches, it’s not sterile. But worse than that is in terms of a toxicant
load. We measured a number of batches, mercury is
one example because these are collected in open environments. So, the mercury load, although at a low concentration,
bioaccumulates, so that’s a concern for us. And then also just having so many components
entering our system. We wanted to clean that up. So, the commercial diets … there has been
issues with inconsistency, I know others have noticed that feathers showing up in your diets
and clumps of other food, plastics, so it just seems to be a … less control over quality
assurance in this sector. It’s gotten better. So, then it would arrive, again, weeks after
the expiration date. That’s really helpful, particularly when
you have your animal care inspectors look at your food and say, “Oh, this is brand
new, but it’s expired.” So then we get slapped. So, again, relying on mix of feeds are often
the case, and we’ve heard examples of that today. So, so the first study goal was the second
one: Could we completely remove Artemia from our program if we changed our defined diet? And then, again, that would minimize our biosecurity
challenges, minimize labor costs, and hopefully maintain or improve our embryo production. So right now, my driver—and I know it’s
different for many of you—is to achieve all of these. Better, faster, cheaper. Lots of eggs, and get them fast and high quality. So that’s what we wanted to see, if we could
achieve that. Okay, there we go. There you go. Okay. So, in our first initial trial, we tried our
control diet, which was the Zeigler AP1000 Golden Pearl with Artemia, and then we … which
we’ve heard a lot about, the GEMMA Micro, and then we had the Zeigler but with Artemia,
and then the Zeigler alone. So we tried these combinations, just an initial,
fairly large trial to see how did these, how did they perform in terms of viability
and egg production. What is going on? So again, the idea was these guys we can feed
right away, so right when they start eating, we can track survival beginning at 30 days,
and then at 90 days. And then we would spawn the fish at intervals
at 10 or 14 days and do a cumulative count. So we’re getting a count and a quality score—I’m
not going to show you the quality score here—and see how they did. And so in this first trial, you can see that
of the—I labeled them—the GEMMA outperformed our traditional lab diet, and the AP100 Plus
and with Artemia and minus were the least performing. So, we already had a hint that the GEMMA in
our … for our drivers was actually going to perform better for our needs. So, what we did is we repeated these studies
with those two diets, our control diet and the GEMMA plus or minus Artemia, and just
did the rest of the studies. You can see that in terms of the viability
at 83, approximately 80 percent in the controls, and where it starts dropping in these other
diets. And then it just kind of shows you … this
is three replicates of very large groups, so at 30- and 90-day survival in the control
and the GEMMA diet and you can see that the GEMMA is … I mean they’re close, but the
GEMMA is outperforming it certainly in the 90-day timepoint, so these are pretty good
results for us that would meet our criteria. And then we want to … large fecundity studies,
since we do large production, we don’t do a lot of little spawning tanks. We have large 100-gallon tanks, we have 35-gallon
tanks, so we can collect lots of eggs very quickly. So, in this example, we have 375-liter tanks
with about 900 fish, 1:1 ratio, and this one we have 150 fish … 150 liters and 400 fish. Then we … these are what I am going to describe,
two different groups. And then what we’re going to do is we are
going to count the quality and the number of eggs. And so we group spawn them, and this is kind
of cumulative. When you get this many eggs, you don’t really
count anymore, although we have an automated counter now, but when we did this, we didn’t. So, it’s actually done by mass and you divide
it. You can see that in the lab diet, and we had
pretty good production, with 207,000 eggs cumulatively. But even the GEMMA did better with 250,000,
and then the fish group three was, again, the smaller set of fish, fewer eggs, but still
the GEMMA outperformed it. So, for our production needs, this one did
really well. And then if you start tracking again, we need
big batches, so you start tracking over time every time you do these spawns. Our lab diet, which we use for a long period
of time, produced less embryos, took longer to achieve numbers over 5,000 eggs in a spawn,
and then fewer spawns greater than 10,000 embryos. So, everything with the GEMMA is shifted up
and looking at better production. And then we start looking at the adults. We have some histopathology on a lot of these
that Mike Kent does for us, but you can even look in the weight gain for the females and
the males, when you compare the control diet versus the GEMMA. So the GEMMAs are getting bigger, like we
heard in many of these talks today, and they get longer, as well. And then, if you actually look at … if you
start looking at the K-factor, again, the GEMMA diet produces females with a higher
body condition factor, and then the males are just marginally better, so again, the
GEMMA seems like these fish are growing better and they produce better, more eggs, and high-quality
eggs. So, the study outcomes … we wanted to know
could we remove Artemia without additional live foods, and yes we can, and under, again,
under our conditions. We want to minimize potential biosecurity,
so we reduced our incoming stuff by five to one. And then limited hundreds of staff hours in
making live food, in maintaining it, and, again, cost. And then finally, maintain or approve existing
survival. So for our metrics, we have more eggs, higher-quality
eggs consistently with the GEMMA diets. So last thing, I told you I was going to talk
about, the very last slide, is the toxicant load. How do you define your toxicants? The problem is there are some really bad actors
that many of us have identified in the literature—you know, high loads of copper are very bad for
aquatic organisms, for example: mercury we’ve heard about, arsenic, lead, etc. So, it’s actually easy to measure those. You can do that on a single run on ICPMS at
very quantitated numbers, and there are formulas and models to calculate what the bioaccumulation
would be—that’s actually pretty easy. So, that’s not a problem. So, you should be able to, if you’re designing
diets … commercially, you should be able source components where you keep those levels
down; that’s actually very easy. The tougher one is the rest of the chemical
universe. That’s where I work in. Many … so, we’ve screened hundreds of
thousands of chemicals specifically to identify chemicals that modulate the developmental
progression, or developmental events, so we have a better understanding of how many can
do that, and so you start thinking of pesticides. Pesticides are, quote, “residues in many
of these compounds.” Some of these are actually bioaccumulate. Plastic components, we’ve heard about a
little bit of that. Bisphenol A, phthalates … And you start
thinking about industrial chemicals—polyfluoroalkyl substances, flame retardants, PCBs, PAHs,
dioxins … These things are particularly important because they bioaccumulate, and
they’re active at really low concentrations. So, they may not notice it, even for a couple
generations, but these things are accumulating in your system. What I mean by that is, if you have these
hydrophobic compounds in a diet, these can bioaccumulate, even if they are at really
low levels, particularly if you have these long-term studies that we’ve heard about. Again, the chemical properties, which you
can predict fairly well, influence a degree of uptake or accumulation. Again, there are models to predict that. So, if you use the logKow ratios … if you
have a high logKow, those are compounds that tend to get into the lipids. That means that if they are in the diet, they’ll
accumulate, and then they’ll accumulate into the yolk. The last point that I really want to make
is for compounds that are delivered either, like Mark was saying earlier, you … a lot
of these studies are done by adding the chemicals to the water. That’s actually okay, because some of these
compounds partition from the water to the yolk really fast; I’m talking within minutes. So, it’s almost as if you got it from a
paternal load into the yolk. It’s not the same if there is metabolism,
but the point is that the parents, what they eat—and we’ve measured a lot of these,
they really are loading the cargo, the yolk that is going to be the building blocks to
build all of those developing systems that we’re all trying to study in zebrafish—are
going to be influenced by those accumulations in the cargo. So, I think we’re at a better stage to start
increasing what chemicals we should be looking at to see whether or not they are entering
into our diets and therefore our systems. And just one example from literature … an
example from that group that … I don’t know that group, but they showed that an accumulation
of mercury just by the number of days at these low levels of mercury in the diet, and they
accumulate at fairly high levels, and when you removed it from the diet, it still takes
a fair amount of time for it to be removed. And then a paper from Mike Carvan’s lab
recently demonstrated that very low levels of mercury epigenetically modified the embryo
and affects behavior in, like, four generations. These are things we talked a little bit about
in here about epigenetics, but identifying chemicals that have the propensity to modulate
the machinery that affects epigenetics should be, in my view, a priority, and those are
not well understood in any system. Alright, with that, I’ll shut up. Thank you. {applause}
MODERATOR: We have time for maybe one question. UNIDENTIFIED MALE: {off microphone} {indiscernible} ROBERT TANGUAY: That’s a great question. The advice we got from the nutritionist. The second we … were able to infer that
they were fairly clean. So, there were other protein sources that
we were considering. If at … also, the way they behaved in the
water column, those combinations.