Role of feed management in nutrition in zebrafish Chris Lawrence

Role of feed management in nutrition in zebrafish Chris Lawrence

September 27, 2019 0 By Ewald Bahringer


CHRIS LAWRENCE: Hi, everybody. It’s nice to be here and thanks for the
invite. This a long time coming, so it’s really
nice to be able to have this kind of group all together in one room to talk about something
that’s certainly near and dear to many of our hearts, certainly mine. Today, or the next 15 minutes or so, I’m
going to talk to you a little bit about some practical aspects of feed management in a
zebrafish facility. Very basic logistical stuff. Oftentimes I am asked … Well, for those
of you that don’t know me, I run a very large zebrafish research program at Boston
Children’s, where we have one of the largest, most complex zebrafish research programs anywhere
in the world, and so I’ve got lots of experience in screwing things up, and I have lots of
experience in practical aspects of managing this kind of resource. And so oftentimes, when we talk about standardization,
definition, and improving some of the management aspects of this particular species, I’m
the guy that tells you the annoying things that you don’t really think about, because
those things are going to limit the application of anything that we come up with in this room. And so what I want to do is talk about two
or three things that are a thorn in our side now from a management perspective as it relates
to nutrition, and I think it’s really relevant to the conversation that we’re having here
today. So … let me see here … here we go. Sorry. Okay. So, we are in what I would say a transition
as it relates to zebrafish feeding from the traditional to a more progressive outlook. So, the traditional approach, as it relates
to nutrition in zebrafish and feeding zebrafish, is all about performance, and by performance
I mean primarily growth and particularly reproductive function. These are without … if you can’t do those
things in managing a fish facility of this nature, you’re not going to do any science,
right? And so, in particular there’s a really large
focus on egg production, embryo production, in zebrafish, and this is a historic fact. This has to do a lot with the fact that the
model as a biomedical model—the fish is a biomedical model—really is born in developmental
biology. So, everybody who’s making eggs and embryos
so that they can study them, right? And so, everything was about what diet made
lots of … made fish grow and produced lots of eggs for experiments. That was the focus and still is the focus
in many labs around the world, and it’s never going to go away as long as we’re
using this fish as a model. And so, not really concerned about what’s
in the diet, not really concerned so much about defining ingredients, defining amounts,
limiting any ingredients, any nutrients—just shotgun approach is the best way I can put
it. Just whatever works, right? And so, a lot of the early diets came out
of the fish hobbyist realm. So, a zebrafish is a tropical aquarium fish
and so the traditional feeding paradigm was built on a platform of Artemia, of brine shrimp,
and flake food—tropical flake food—and we’ll talk about this a couple of times
over the course of these next few minutes, but that was sort of the traditional approach
and still is the traditional approach. Many people still use that as the platform,
as a nutritional platform, for zebrafish—TetraMin flake and brine shrimp. We’re transitioning now more and more rapidly
to a more progressive approach about nutrition, which still we haven’t focused on performance. We have to have a focus on performance. But we’re also thinking about what diet
creates a healthy representative animal for use in a wide variety of different studies,
right? So, a fish that produces embryos is not necessarily
… at a copious rate is not necessarily a healthy representative animal. So, there’s focus on overall health and
then the definition of what I call the “nutritional landscape.” This is an experimental variable, a huge experimental
variable, that we have overlooked for as long as we have been using the zebrafish as a model
organism. You want to know what’s in the diet in terms
of the ingredients but also the quantity. This makes a huge difference … how much
you’re actually feeding the fish. Go back, if you don’t believe me in terms
of the lack of a definition on these things in the zebrafish field, go back … go into
the literature now and try to look at the materials and methods section and try to find
out how much and what fish were fed for a particular set of experiments; it’s not
going to be easy. In fact, it’s going to be very difficult. So, we are now … there’s a growing awareness
that this is important, as evidenced by the fact that we’re in this room together here
today, and so we’re moving in this direction. But even when we figure out what it is that
these diets should actually look like, there are still going to be things that we need
to … problems that we’re going to need to solve in terms of, again, leveraging—so,
being able to apply what we’ve learned. So, we know exactly what the fish need; it’s
still a whole other matter of how we get it into the fish in a practical way, and so that’s
really what I’m going to talk about. So, a few years ago at Children’s, we started
this transition in terms of trying to get away from the black box of nutrition and starting
to define what it is we were doing and trying to have a data-driven biological basis for
our feeding practices. And so, what we did was we published a study
which was … looked at a number of different things, but one was feed frequency: How many
times a day do you need to feed a zebrafish in order to drive certain basic parameters? In this case, it was growth and, importantly,
reproduction. We also wanted to look at simplifying the
diet and moving away from this shotgun approach in feeding one diet—one fish at the adult
stage—that we could define or define to a much greater degree than what we had done
before—control it. Also, the quantity, right? So, we wanted to move away from this shotgun
approach where we were using a mixture of zooplankton and various off-the … you know,
sort of off-the-shelf fish feeds, and we moved towards an aquaculture diet, what I would
say is a commercial feed that was designed to grow cold-water marine fish, but it’s
a fairly refined diet. It’s a Skretting diet, which many of you
may know as general microbes. So, we used this diet, we used feed frequency
experiments, and we found lots of things—important things—out. Number one, we found that you could actually
feed the fish as few times as once per day and still get the same kind of performance
as you could get with three or five times a day at that same amount. We also found that you could move away … you
could remove something like Artemia from the diet, so I often call Artemia sort of a third
rail of the zebrafish feeding set of beliefs, if you will. Most people think if you take Artemia out
of the diet for zebrafish, everything’s going to end, right? And so, we showed that you didn’t need Artemia;
you could go to one commercial feed as little as once a day and still get good performance. This sounds elementary to some of you that
aren’t zebrafish people, but no one had ever worked this out; we just followed what
was done before. So, the interesting thing is we did this study—this
was published in 2011—and in the discussion section we noted some of the problems that
may arise as a result of what we were going to do. So, we were going to transition to this kind
of feeding approach, but we acknowledged that there may be challenges. And what I’m going to explain to you or
illustrate to you over the next couple of slides are three of the primary challenges
that we identified early on and are now grappling with now, and they’re all very straightforward,
particularly the first two. So, challenge number one: feed delivery. Anyone who has worked in zebrafish facility
in any sort, particularly as they scale up, realizes what a challenge feeding is. So, a
zebrafish facility is a very … research facility is a very challenging and, in some
ways, I think, a unique challenge for feed delivery, and that’s different than commercial
aquaculture, that’s different from most other lab animals—mammalian lab animals. We have, in some cases, hundreds or even thousands
of tanks, and so this particular … this photo on the background is a room that has
8,000 tanks in them—8,000 inputs, right? That in and of itself … so it’s one of
the benefits of the model, so it’s very … we can fit 200,000 fish in a room of this
size, but there are consequences to that, and one of them is that there are many thousands
of inputs for feeding … feed delivery. These entities are not all uniform. If they were, it would be a lot easier, but
they’re not, right? So, you’ve got different life stages, you
have different densities, so we average a way of target density of 10 fish per liter,
but we don’t … it’s not absolute, so we have some tanks that have one or two fish
per liter or less, and then we have some that have 15 or 20. And it depends on the experiment, it depends
on life stage, and then the experimental state is also very different. So, you may have fish that are being used
just to profligate … they’re just being grown to profligate a subsequent generation,
so you may be making a transgenic or a knock-out, and all you’re doing is growing up that
generation to breed them once and identify carriers, right? Or you may be growing fish that are long-term
tumor models or stem cell transplant models or anything. As we become more and complex with what we’re
doing with fish, the number of different experimental states that can exist within one facility
changes, and all of those fish, all of those experiments, actually there are different
nutritional requirements for them. So, you shouldn’t be feeding breeding fish
the same way you feed fish with a high tumor burden—adult tumor burden. But it’s easier said than done. So, this is a big challenge. And there is … I’m just going to do it
like this … so it wasn’t just you and your technological … although … Alright. So delivery is a big challenge, and over the
last 5 or 6 years, in particular, as we have transitioned away from the traditional feeding
approach or even if you’re still doing it, there are lots of solutions for delivering
known quantities of feed to thousands and hundreds of thousands of tanks. They include the venerable squirt bottle or
squeeze bottle or wash bottle. They have … we have … sorry …
UNIDENTIFIED FEMALE: More technical issues. We’re just very lucky. CHRIS LAWRENCE: So we have insect pesticide
sprayers that have been used to deliver Artemia. We have modified and stainless steel versions
of effectively what were spice dispensers to use to deliver known quantities of a dried
feed to a tank, and then all the way at the highest pinnacle of technological innovation,
we have robots that will deliver feed to hundreds or thousands of tanks, and these are all sort
of representations of the fact that we have a challenge in terms of delivery. So, any diet that we design has to be deliverable,
and preferably it’s one that is easier to deliver. So, if we come up with a diet that needs to
be fed 15 times a day, this is going to be a practical challenge that we’re going to
have to deal with. UNIDENTIFIED FEMALE: The video feed has stopped
for some strange reason. CHRIS LAWRENCE: Are we good? UNIDENTIFIED FEMALE: Yes. CHRIS LAWRENCE: So, delivery is one of the
big challenges. So, when we transitioned away from a diet
that included zooplankton and a mixture of other dry feeds and we went to a single dry
feed, we had to come up with a way of actually delivering it to thousands of tanks in a reproducible
way—not simple. And we’ve spent enormous amounts of time
and money, actually, trying to figure out what’s the best way to deliver diet. The second challenge, if I can advance the
slide … UNIDENTIFIED MALE: Did you push the arrow
on the computer? CHRIS LAWRENCE: Yes I am. As long as you guys don’t mind, I don’t
mind. The second challenge is waste production. So, what I have here is a figure that shows
you a generalized version of the classic commercially available zebrafish housing tank. In most of the major vendors they differ in
some respect or another, but they mostly utilize the same approach, which is driven by hydrating
flow into the tank where solid wastes that are produced in the tank are then sucked out
or siphoned out the bottom of the tank and up behind a baffle and out into the overflow
drain. This is the standard approach, whether you
do it with a siphon, whether you do it with just a baffle like this. These systems are all designed to be self-cleaning
in the sense that anything solid that’s going in the tank that’s free, that’s
up in the water column, will get sucked out pretty quickly. This is the way these systems are designed,
and these systems are not changing anytime soon. This is a realization that I’ve come to
over the last couple of years when I’ve actually asked people to start … housing
vendors to start thinking about this issue. It’s not easy because we’ve got tens or
hundreds of thousands of units in production right now, and this isn’t changing, so this
is the way these systems are made, and it’s a challenge. It’s less of a challenge when you’re feeding
a traditional diet, which is low levels, low biomass, of Artemia going in, or rotifers,
and floating or quasi-floating TetraMin. That’s really what these things were designed
to deal with. Soft cleaning in that low biomass, low solids-producing
… that dietary regime produces relatively little solid waste. If you move to an aquaculture diet with commercial
feed—a pelleted diet or a crumbled diet—and you feed it 5 percent of the body weight—3,
4, 5 percent of body weight a day for an adult—this is a sludge-producer. Sludge producer. And this is something that we knew going in
when we were transitioning to this approach. We knew this was going to be a challenge,
but we did not appreciate how much of a challenge it was going to be. So, this is sort of the back side of one of
the housing units that’s out there on the market today. We’re looking at the back of a tank, and
that green is the baffle and those plaques, if you will, those are solids that are starting
to accumulate behind that baffle. And the second that that starts to happen
… and they’re starting to accumulate because what happens first is you get a biofilm that
forms, and then the biofilm is sticky and then it starts to … and it’s a very narrow
gap of a millimeter or two or whatever it is, I don’t know exactly what the gap is,
but it’s very narrow. It’s very easy to start sticking solids,
and when that happens, solids begin to … the tanks don’t work the way they should, and
so what you end up getting is actually more solids going directly into the system. You also get a lot of this waste directly—this
Artemia—going into a tray filter. This is Artemia fed at high amounts, but we’re
trying to actually limit what’s going into sump. It’s not easy. You just can’t turn a system off and feed
it. Like, when I started in zebrafish 20 years
ago, we actually had these homemade systems that you had timers on, and you put the timers
on for 20 minutes to be off, feed Artemia, and we would reduce this, but now the systems
aren’t very compatible. Our relay is not compatible with this at all,
so you can’t do it, so you end up flushing that down the drain. That’s not cheap. It’s very expensive, and it also causes
all sorts of problems. And if you’ve gone through the pain and
suffering of culturing Artemia, encapsulating Artemia, and purchasing Artemia, when you
see this, it’s painful. So, like I said, a sludge producer. So, these are just pictures of sumps in a
particular … on the left, that’s a picture of a sump. That is sludge that’s just accumulating. On the right is a tank that doesn’t have
any fish in it. So that’s all the suspended solids in the
water and the fish keep up that suspended solids by design, it gets up into the water
column so it’s gets pulled out. But if the fish isn’t there anywhere where
you have sort of settling, this is what you’ll get in a fish system that’s fed intensively
with a commercial feed—any commercial feed. It doesn’t matter which one it is. And that sludge is not a good thing. One, it promotes the growth of lots of things. Do you see … I don’t have a … does this
have a laser … doesn’t really. At any rate, if you see the little lines on
either … in the sump and on the tank, those are Bryozoa. That is a major problem, a following organism
that acts as a scaffold for lots of other microorganisms; that could be problematic. It’s also nasty. It’s a major problem in the shipping industry,
not this particular species, but Bryozoa is a major nuisance species, and it’s a nuisance
species in zebrafish, particularly when there’s lots of suspended solids in the system. That solid waste is a problem … I don’t
know if this video is going to work. Probably because I’m a Mac person it won’t
… oh it is. Look at that. So, these are fish. They are between feedings, and they are grazing
all along those solids. They’re grazing on this stuff, so, microorganisms
that are in the sludge. So, if you have sludge in your system and
you want a fun little observational experiment, pull some of that sludge out and look at it
under the microscope; it’s going to be crawling with microorganisms. What the impact of these microorganisms are
is not necessarily known, but certainly we know that certain microorganisms are vectors
for different pathogens. This is something you generally want to avoid,
but we are producing systems that are chock-full of it. The last challenge that I’ll talk about
is a little bit more complex, and it has to do with nutritional imbalance. So, we transition from a zooplankton-based
diet to one that is completely reliant on a commercial feed that was designed, again,
to grow cold-water fish. It’s designed to grow fish quickly, and
it works really well for zebrafish. It’s not nutritionally formulated for zebrafish. Obviously, we don’t know what the nutritional
requirements are; we obviously know that. But this diet and other diets work well to
grow fish, and they grow zebrafish very, very well. And one of the things that you see is that
when the fish are in rapid growth phase … so zebrafish displays sort of an indeterminate
growth, but the growth slows down. So if you leave a fish alone for a long period
of time, it will be gigantic in 3 years, but that growth slows when they hit a growth plateau. So these diets and this diet that we use is
very good to get fish to that plateau, very good. It’s good at growing and driving somatic
growth, but once they hit that plateau, where do all those excess nutrients go? So this … the feed that we use is 62 percent
crude protein, which is a lot more than is required for an adult zebrafish or even a
growing zebrafish. This, in part, is because it’s designed
to mimic Artemia, which has that similar protein content, about the same crude protein content
in Artemia. But the idea is, once they hit that plateau,
what happens? Where does all that excess nutrient go, right? We don’t change what we feed. We don’t change the formulation of the diet. We don’t necessarily even change how much
we feed, not deliberately. So, we feed 1 percent a day, 5 percent of
body weight, 4 to 5 percent of body weight, but we’re not generally adjusting that for
every single tank in the representation of thousands of different units. So, what happens is you end up feeding them
that same high-protein, high-nutrient-load diet that they don’t necessarily need, and
so there are lots of repercussions to this that many people in this room can talk about
with greater specificity than I can. But certainly one of the things that happens
from a practical and logical standpoint, particularly in females, is that once they’re done growing,
they are allocating those excess resources to reproduction, right? So, a female is making eggs, and so one of
the things that we see is they become egg machines. But the question is, what happens then? It becomes very difficult to balance things
like liability and embryo production, so one thing that we’ve done over the years and
we’ve seen happen is our viability rates go down. So, the number of eggs that are good has decreased. A viable egg is an embryo that on day one
is normal and healthy. And so, one of the things that we look at
what’s the relationship between that ration size and the breeding interval in an adult
fish. So, female is taking all that excess nutrient
and using it to build and to produce eggs. You’ve got to take those eggs from her to
sort of compensate. Do you understand what I’m saying? And so, how many eggs do you have to take? What’s that interspawning … what does
that breeding interval look like? Is it every day? Is it every other day? Is it once a week? Is it every 2 weeks? And again, this is once the bulk of somatic
growth is completed, so what you get are these great big females that produce lots and lots
of eggs. So, we did a very simple experiment. Actually, it wasn’t very simple, it was
actually enormously burdensome to do, but we looked at … we tried to look at what
the balance was in females, and I’m just telling you this to illustrate the difficulty
of this one aspect without having a feed that’s designed for the species in question and the
life stage. So, we fed fish either 1, 3, or 5 percent
of body weight a day once a day, and then we bred them either twice a week, once a week,
once every other week, once a month, or once every 3 months, and we wanted to look and
see what the relationship between embryo number produced and viability was … to think about
it is that somewhere there’s a sweet spot where you get the right amount of feed in
the female such that you have the highest amount of embryo production possible with
high viability, so there’s no waste. Because what happens is they produce and produce
and produce. They’re ovulating all of the time, and if
you’re not taking those embryos away, they expire, essentially. They … within the body cavity they expire,
and when they are expressed during spawning, they don’t fertilize, so that’s the relationship
that we’re trying to look at. So, this is just graphs that show the relationship
between fecundity, which is the number of embryos produced, and the viability. And you can see that the relationship is not
necessarily … we couldn’t figure out where that sweet spot was. So, regardless of whether or not we bred them
once … so A is twice a week, B is once a week, C is once every other week, and D is
once a month, E is once every quarter, essentially. So, we weren’t seeing much of a difference
with the interspawning interval, but we were seeing a difference with the number of embryos
spawned overall. And, conversely, we saw that we had higher
viability when we fed them less. But we’re still seeing viability rates which
… so there’s a balance, so you either don’t feed them … if you feed them 1 percent
of body weight a day, you get pretty low fecundity. If you want higher fecundity, you get much
higher-end viability rates, and somewhere in there is a sweet spot that we can’t necessarily
hit. The sweet spot would be that you would have
higher fecundity, higher viability, so that you’re not wasting time and energy, because
if you have high … so, you need lots of eggs, but they need to be good to be able
to do these experiments. Just an illustration of how difficult it is,
actually, to get it right. In summary, what I would list as requirements
for the next generation of zebrafish diets is that obviously they have to be comprised
of defined ingredients that can be controlled. They should be able to produce and maintain
representatively healthy fish by default; that should be the default diet, whether we
have maintenance diets, whether we have breeding diets; these are the things that I think probably
should be discussed. Delivery of whatever the feed is has to be
practical. The feed itself has to be compatible with
the current housing, such that the administration does not overtax them. Any pelletized diet is going to cause problems,
so we’re going to have to think about different ways to formulate a diet physically so that
you don’t have that wastage going into the system. But at the same time we’re going to have
to be able to deliver it in a practical way, and it has to be life stage- and experimental
context-specific. So, there you have it. I’ll take questions if you have any. (applause)
STEPHEN WATTS: Any questions for Chris? UNIDENTIFIED MALE: So, you decided to use
the flake diet? CHRIS LAWRENCE: No, we don’t. UNIDENTIFIED MALE: Do you use the extruded
diet? CHRIS LAWRENCE: Yes, it’s an extruded pelletized
diet. UNIDENTIFIED MALE: So, why standardize an
aquaculture pellet diet rather than on flakes, like TetraMin? Did you understand that on something that
you could use a small (indiscernible, 03-32:21) and it’s not optimally suited to zebrafish,
initially? CHRIS LAWRENCE: Yeah, that’s a good question,
and there are lots of different reasons as to why we moved away from flake diets. Number one, a flake diet, to feed it properly,
is not easy, so a pellet … the difference between the physical characteristics of a
pellet make it easy to deliver at least known quantities; a flake is a much harder to deliver
known quantity. Oftentimes, historically, people would take
flake and they would mix it with Artemia and fluidize that. That, among other problems, would lead to
leaching of all of the water-soluble nutrients from the flake directly into the water without
getting to the target fish. The other thing about flake diets is you have
almost no control over what’s in them, right? And so, there are people in this room who
can talk at length about the problems of flake feeds for a research fish. Now, does it work in terms of supporting growth
and reproduction? On its own, I would doubt it, but in conjunction
with other things, yes. But we moved away from that for those reasons
and others. Yes? UNIDENTIFIED MALE: I just had a comment that
the switch to GEMMA that’s happened, to GEMMA Micro food, it really came down to the
fact that the food would stay viable and palatable for a couple of hours at the bottom of the
tank, and until that food came out, most of the time you’d never see fish eating from
the bottom. But that pellet material … they just will
eat it all over many hours, where the flake and the other stuff quickly just rotted and
the fish could care less. So it was that feature to how Skettering is
cornering the market, if you will, and Chris’ paper showing that you could just do this
once a day. But it has a lot of problems, as you pointed
out. STEPHEN WATTS: Okay, last question. UNIDENTIFIED MALE: So, me or him? STEPHEN WATTS: Ron, go ahead. You won. (laughs)
UNIDENTIFIED MALE (RON): I won. Thanks. This is kind of a dumb question. I’m not a zebrafish guy, but do you know
the point at which these fish begin to switch energy allocation from somatic growth to ovarian
and reproduction? I know it’s not an abrupt change, but there’s
a … I’m looking at, say, gene expression, (indiscernible). CHRIS LAWRENCE: I don’t know that specifically
what it is, what that landmark is, and obviously there are lots of things that would influence
this, but it’s typically when they are … well, there’s a lot of spread in this, but it’s
at 3 to 6 months of age when the fish is about … males probably about 0.3 or 0.4 grams,
females 0.5 to 0.8 grams. Once they hit that size they’re definitely
allocating anything towards reproduction and very little of it to growth. This switch probably happens earlier than
that. Steve probably has the best information on
that, but it’s very variable, and we don’t have, really, any published metric for that
that’s standardized. UNIDENTIFIED MALE: There is—I can’t remember
the details—but there is a very good paper on this in PLoS One by Matthias Hammerschmidt’s
group that looks at basically allocation to energy being deposited in lipids, somatic
growth, and egg production. And like I said, I can’t remember the specifics,
but we should definitely consider that paper in any further kind of discussion on this. CHRIS LAWRENCE: Yeah, did that come out, like,
in the last year or two? UNIDENTIFIED MALE: In 2016, I think, yeah. And I think it’s a lot earlier. I think it’s between 1 month and 2 months
of age that they start. STEPHEN WATTS: One more comment, and we can
talk about this later. UNIDENTIFIED MALE: This has to do with … I
like the work that you do with the frequency. Have you ever thought about the time of day
of feeding as being something important, or is it just logistically prohibitive? CHRIS LAWRENCE: That’s a good question. It’s logistically prohibitive, generally,
but it’s a good question that should probably be answered. I think fish are generally fed during the
9-5 work day. That’s probably the best way to put it. UNIDENTIFIED MALE: Again, there is a good
paper on time—and I can’t remember the specifics—but on when fish are fed with
regard to how much fat they store in adipose tissue. So, I can dig that out as well. UNIDENTIFIED MALE: ApoB and tests for ApoB
is on the clock. UNIDENTIFIED MALE: Yeah. I think everything …
UNIDENTIFIED MALE: Which makes all the atherogenic lipoproteins is cycles. STEPHEN WATTS: It’s time, maybe save it
for the break, I think, or for the panel discussion.