In this protocol, we demonstrate how to breed Astyanax mexicanus adults, raise the larvae, and perform whole-mount immunohistochemistry on post-larval fish to compare the phenotypes of surface and cave morphotypes.
River and cave-adapted populations of Astyanax mexicanus show differences in morphology, physiology, and behavior. Research focused on comparing adult forms has revealed the genetic basis of some of these differences. Less is known about how the populations differ at post-larval stages (at the onset of feeding). Such studies may provide insight into how cavefish survive through adulthood in their natural environment. Methods for comparing post-larval development in the laboratory require standardized aquaculture and feeding regimes. Here we describe how to raise fish on a diet of nutrient-rich rotifers in non-recirculating water for up to two-weeks post fertilization. We demonstrate how to collect post-larval fish from this nursery system and perform whole-mount immunostaining. Immunostaining is an attractive alternative to transgene expression analysis for investigating development and gene function in A. mexicanus. The nursery method can also be used as a standard protocol for establishing density-matched populations for growth into adults.
The Mexican tetra, Astyanax mexicanus, is a single species of fish that exists as river-dwelling populations (surface fish) and a number of cave-dwelling populations (cavefish) named for the caves they inhabit (i.e., Tinaja, Molino, Pachón). A growing number of researchers are using A. mexicanus to investigate the genetic and developmental bases of behavioral1,2,3,4, metabolic5,6,7,8, and morphological evolution9,10,11. Available resources for studies of A. mexicanus include a sequenced and annotated genome12; transcriptome13; developmental staging table14; and methods for breeding15,16,17, creating transgenics18, and editing genes19. Dissemination of additional tools and updated standard protocols will accelerate growth of the cavefish research community (see this methods collection20).
Our goal is to add to the existing repertoire of tools by providing a robust method for assessing gene activity in situ in post-larval A. mexicanus, in a manner comparable between laboratories. There are two challenges to achieving this goal. First, there is a need for standardized regimes for hatching and raising the fish between laboratories, as differences in parameters such as feeding and density affect growth and maturation, thereby impacting gene activity. Second, there is a need for a standardized yet adaptable method for examining patterns of gene activity in the post-larval fish. We address these issues here, establishing standard practices for raising fish to post-larval stages and introducing a robust whole-mount immunohistochemistry (IHC) protocol for assessing gene expression in A. mexicanus.
We first demonstrate how to breed the fish through natural spawning and identify fertilized eggs. Described next is how to hatch fertilized eggs (larvae) and transfer them to nursery containers, where they are maintained at a density of 20 fish per container for two weeks without recirculating or changing the water. At 5-days post fertilization, the fish have developed to post-larval stages (no longer having a yolk supply) and are provided algae-fed Brachionus plicatilis (rotifers) as a nutrient-rich food source that do not require daily replenishment. This method provides consistent growth parameters for larval and post-larval development.
To assess gene function, we demonstrate how to remove the fish from nursery containers and perform whole-mount IHC. The IHC method presented is adapted from protocols developed for use with Danio rerio21 and is effective for examining antigens in all A. mexicanus tissues tested, including the brain, intestine, and pancreas. IHC is a faster alternative to generating transgenic animals for examination of gene expression and protein localization. This protocol will be useful for studies aimed at growing A. mexicanus and comparing the phenotypes of surface fish and cavefish at post-larval stages.
The procedures described throughout this protocol have been approved by the Institutional Animal Care and Use Committee (IACUC) at Harvard Medical School.
1. Breeding
NOTE: There are several published methods for breeding15,16,17,20 that could also be used at this step. Prior to breeding, adult fish are maintained on a 10:14 light:dark cycle at 23 °C and fed a pellet diet (see Table of Materials) once daily. Fish can be bred in a recirculating system with mechanical filtration and UV sterilization. Breeding can also be accomplished in static (non-recirculating) tanks, but fish should not be left in a static tank for more than 3 days, as water quality quickly degrades.
2. Hatching Fertilized Eggs
3. Transfer of Hatched Larvae to Nursery Containers
4. Preparation of Rotifer-based Fish Food
5. Feeding of Post-larval Fish
6. Whole-mount Immunohistochemistry of Post-larval Fish
Table 1 shows success during one year of breeding surface fish and Tinaja, Molino, and Pachón cavefish in static breeding tanks. Surface and Pachón spawns with fertilized embryos always produced hatched larvae, while Molino and Tinaja were unsuccessful some of the time (2/6 and 2/18 spawning events did not produce hatched larvae, respectively). There is variation in clutch size that does not appear to be attributed to age of the parent fish. Table 2 shows the total number of hatched larvae resulting from some of the spawning events, and the age of the parent fish. In general, we found that surface fish produce the greatest number of larvae per spawn (average 1,550 ± 894, n = 5), followed by Pachón (average 879 ± 680, n = 6), Tinaja (average 570 ± 373, n = 11), and Molino (average 386 ± 276, n =3). The number of larvae produced is typically more than are needed per experiment or for growth into adults. We typically set up 6-18 nursery containers (120-360 larvae) and euthanize the remaining fish.
To measure the success of the nursery protocol we recorded the number of hatched larvae and surviving post-larval fish from successful spawning events. Table 3 shows data from 1 month of breeding in recirculating tanks and includes the number of larvae transferred to nursery containers that survived to 14 dpf. During this month, the survival rate ranged from 41-81 percent, resulting in 65-293 fish available per population for experiments or growth into adults.
To determine if whole-mount immunostaining is successful, we compared the fluorescence of samples incubated with primary antibody to those incubated with secondary antibody only. The fluorescent signal is only visible in the fish incubated with primary antibody. We have used this protocol to successfully label neurons10 (Figure 1) and pancreatic cells6 at stages up to 12.5 dpf in both surface and cave morphotypes.
Figure 1: Neuron labeling. Whole-mount immunostaining of A. mexicanus. Image of 12.5 dpf surface fish (a) and Pachón cavefish (b). Image of the mid-body region [hatched yellow outline shown in (a) and (b)] of surface fish (c) and Pachón cavefish (d) stained with pan-neuronal antibody (Hu). (e) Confocal image of a region of the surface fish intestine showing enteric neurons (Hu) and their projections (acetylated tubulin). For this image, the intestine was dissected out and mounted in medium containing DAPI to stain the nuclei. Please click here to view a larger version of this figure.
population | attempts | spawning events | clutches |
surface | 94 | 23 | 23 |
Molino | 110 | 6 | 4 |
Pachón | 167 | 13 | 13 |
Tinaja | 242 | 18 | 16 |
Table 1: Summary of data from one year of breeding A. mexicanus in static tanks. Number of breeding attempts, resulting spawning events, and number of spawning events that produced hatched larvae.
population | parent age | hatched larvae |
surface | 1 year | 989 |
surface | 1 year | 1050 |
surface | 3 years | 2214 |
surface | 3 years | 432 |
surface | 3 years | 1768 |
surface | 4 years | 2852 |
Pachón | 1 year | 1194 |
Pachón | 1 year | 1933 |
Pachón | 1.5 years | 371 |
Pachón | 3 years | 480 |
Pachón | 4 years | 1190 |
Pachón | 4 years | 110 |
Tinaja | 9 months | 259 |
Tinaja | 9 months | 253 |
Tinaja | 10 months | 1100 |
Tinaja | 11 months | 857 |
Tinaja | 1 year | 713 |
Tinaja | 1 year | 853 |
Tinaja | 1 year | 542 |
Tinaja | 1.5 years | 360 |
Tinaja | 1.5 years | 58 |
Tinaja | 1.5 years | 1100 |
Tinaja | 4 years | 185 |
Molino | 2.5 years | 460 |
Molino | 2.5 years | 619 |
Molino | 3 years | 81 |
Table 2: Approximate female age and number of hatched larvae from individual spawning events from the indicated populations of A. mexicanus.
population | attempts | spawning events | clutches | clutch size | larvae transferred to nursery cups | post-larval fish at 14dpf | survival (%) |
surface | 4 | 3 | 2 | 576 & 1728 | 360 | 174 | 48 |
Molino | 4 | 2 | 1 | 228 | 159 | 65 | 41 |
Tinaja | 4 | 1 | 1 | 1952 | 175 | 93 | 53 |
Pachón | 4 | 1 | 1 | 1696 | 360 | 293 | 81 |
Table 3: Summary of data from one month of breeding A. mexicanus in a recirculating system and raising the larvae. Number of breeding attempts, resulting spawning events, clutches that produced hatched larvae, average number of larvae per clutch, larvae transferred to the nursery containers, and post-larval fish present in the nursery containers after 14 days.
Comparing gene activity between surface and cave A. mexicanus requires carefully controlled environmental parameters and methods that can be replicated across laboratories. Our protocol for raising A. mexicanus provides consistent nutritional content during post-larval development. Following this feeding regime, gene function can be confidently compared between populations using the robust immunohistochemistry protocol we present. Here we discuss the significance of this method as well as its limitations and future applications.
To achieve density-matched growth, we found that post-larval fish can be raised without recirculating water for two weeks in 1.5 L containers on a diet of rotifers. This protocol can also be used to raise fish on a recirculating system; however, rotifers must be added daily to compensate for those lost through the tank outflow. Newly hatched Artemia nauplii are commonly used as a food source in aquaculture, but we found that using rotifers has considerable advantages, including: reduced price, improved biosecurity, consistent nutrition, and better water quality (see below).
First, the weekly cost in consumables is 4 dollars for rotifers, compared to 14 dollars for Artemia. Regarding biosecurity, rotifers are raised in the laboratory under controlled conditions, while Artemia are collected from the wild and subject to natural variation in microbial or pathogen contents23. Additionally, the nutrient composition of Artemia nauplii are environmentally determined and therefore inconsistent. Nauplii thrive on their own energy stores after hatching; they quickly loose nutritional value as they develop and should optimally be fed to the fish within several hours. Artemia begin feeding at 12 house post-hatching, representing the first time that they could be nutritionally enriched; however, at this stage they have become too large for 5 dpf fish to consume. In comparison, rotifers continuously feed on marine microalgae resulting in high nutrient content regardless of when the rotifers are harvested. Rotifers are much smaller than Artmeia nauplii (160 vs. 400 microns), making them easier for the fish to capture and swallow. Post-larval surface fish and cavefish consume rotifers in similar quantities suggesting no difference in preference or ability to capture the rotifers10.
Finally, Artemia nauplii begin dying in fresh water several hours after being introduced. Uneaten nauplii will decay, rapidly decreasing the water quality if they are not manually removed. Removing dead Artemia is time-consuming and dangerous for post-larval fish that are not much bigger than Artemia and may be accidently removed or injured. Rotifers can live indefinitely in the nursery containers and provide food to the fish at all times without significantly impacting water quality.
While using rotifers as a food source has considerable benefits, to maintain the rotifer stock, algae must be added to the culture system daily. This can be achieved with an automatic feeder that dispenses liquid algae into the rotifer culture container (see Table of Materials). Rotifers must also be harvested from this set-up every 24-48 hours to maintain the health of the culture. Researchers that breed fish very infrequently (once a year, for example) and are not concerned with making comparisons between the populations at post-larval stages may prefer Artemia as a food source, since the encysted embryos can be hatched at any time.
We recommend tracking the number of hatched and surviving larvae to monitor the success of hatching and growth. If most of the embryos or larvae die, it may be due to bacterial or fungal contamination. It is recommended to monitor the water quality of the fish-ready water and sterilize any equipment with 70% ethanol. Nursery containers can be re-used after they are cleaned and sterilized. To minimize risk of disease, it is also critical to remove any dead fish from the nursery containers and not add rotifers before 5 dpf, when the fish begin to eat.
A. mexicanus are sexually mature at approximately one year old. This is a limitation for generating transgenic A. mexicanus compared to Danio rerio (zebrafish) that are able to breed at 10-12 weeks24. Immunohistochemistry (IHC) is an alternative method to investigate gene expression and protein localization. The protocol described here can be adapted at each step and used to detect antigens in any tissue of interest. It is important to note, however, that some tissues may be more difficult to visualize in surface fish due to pigmentation (a barrier that does not exist in unpigmented cavefish), which may influence the interpretation of comparative studies. To address this potential problem, surface fish pigment can be bleached after fixation using 3% hydrogen peroxide.
IHC requires successful tissue fixation, blocking, and antibody penetration. Methods for each vary depending on the tissue and protein of interest. The fixative and fixation time must preserve cell architecture while maintaining the antigen epitope. For this protocol, we use a cross-linking fixative (paraformaldehyde) and omit a denaturing fixative (such as methanol or acetone). We found that incubation in acetone diminished antibody signal for neuronal and pancreatic markers. The blocking step is essential to prevent antibodies from binding to non-target proteins in the tissue. This method uses a combination of normal serum (5%) and BSA (0.2%) in the blocking solution. The blocking solution contains antibodies and proteins that bind to reactive sites on proteins in the tissue, diminishing non-specific binding of the primary and secondary antibodies.
To achieve antibody penetration, the tissue must be permeabilized. This can be achieved with detergents or denaturing solvents but must be optimized to preserve the antigen epitope. Our protocol uses a combination of Triton and dimethyl sulfoxide (DMSO). Triton and DMSO are included during the blocking and antibody incubation steps at concentrations of 0.5% and 1%, respectively. Using this concentration, we have observed staining in the brain, pancreas, intestine, and muscle, suggesting that it is likely effective for penetration of all tissues. Fish size may also influence penetration. This protocol has not been tested on fish that are greater than 14 days old (approximately 7 mm in length). To troubleshoot staining, it is recommended to alter the fixation, blocking, and antibody penetration steps. It is also important to examine the sequence conservation of the immunogen with the A. mexicanus protein of interest using the available genome25.
A. mexicanus is an excellent model to investigate evolution as populations of the same species that have evolved in dramatically different environments can be directly compared in the laboratory. Standard husbandry protocols, both within and among laboratories, are essential to understanding the biological differences between surface fish and cavefish. Our article provides a method to examine development and gene activity in post-larval fish exposed to consistent growth parameters.
The authors have nothing to disclose.
This work was supported by grants from the National Institutes of Health [HD089934, DK108495].
methylene blue | Kordon | B016CBHZUS | antifungal |
heater | Finnex | 4711457836017 | 100W Digital Control Heater |
airstone | Lee's Aquarium & Pet Products | 10838125202 | disposable air stone |
salt | Instant Ocean | 51378014021 | Sea Salt |
nursery container | IPC | 21545-002 | 40 oz or 1.5 L clear containers |
transfer pipette | VWR | 414004-002 | plastic bulb pipettes |
compact culture system(CCS) starter kit with Brachionus plicatilis (L-type) rotifers | Reed Mariculture | na | fish food |
RGcomplete | APBreed | 817656016572 | 32 oz bottle of rotifer food |
Programmable Auto Dosing Pump DP-4 | Jebao | DP-4 | automatic feeder for rotifers |
Tricane-S | Western Chemical | MS 222 | fish anesthetic |
sodium bicarbonate | Sigma-Aldrich | S5761-500G | for tricane solution |
nylon mesh strainer | HIC (Harold Import Co.) | 735343476235 | 3-inch diameter |
Formalin solution, neutral buffered, 10% | Sigma-Aldrich | HT501128-4L | fixative |
10X PBS | Invitrogen | AM9625 | buffer, dilute to 1X using distilled water |
Triton-x 100 | Sigma-Aldrich | T8787-250ML | detergent |
sodium azide | Sigma-Aldrich | S2002-25G | anti-bacterial |
bovine serum albumin | Sigma-Aldrich | A9647-100G | blocking reagent |
glass vial with screw-top cap 4mL | Wheaton | 224742 | staining vial |
plastic mesh screen for breeding tank | Pentair | N1670 | Cut into a rectangle 6mm larger on all edges than the dimensions of the bottom of the breeding tank. Cut a 6mm square from each corner of the rectangle. Bend the edges of the screen down along all four edges.Place a pair of 6mm vinyl-coated disk magnets on either side (top and bottom) of the mesh on each corner. The screen should be as snug as possible to the sides of the tank. The screen can be removed from the tank with a metal fish net. |
vinyl-coated disk magnets | Kjmagnets | D84PC-AST | |
New Life Spectrum Thera-A pellet fish food | New Life International | na | Adult fish food. A list of retailers for this product is available on the company website |
Antibodies | |||
insulin antibody from guinea pig | Dako | A0564 | 1:200 |
glucagon antibody from sheep | Abcam | ab36215 | 1:200 |
acetylated tubulin antibody from mouse | Sigma | T6793 | 1:500 |
HuD/HuC antibody from mouse | Life Technologies | A-21271 | 1:500 |
nitric oxide synthase (nNOS) antibody from rabbit | Abcam | ab106417 | 5μg/mL |
choline acetyltransferase (ChAT) from rabbit | Abcam | ab178850 | 1:2000 |
seratonin (5HT) from rabbit | Immunostar | 20080 | 1:500 |