Abstract
The leptin system plays a crucial role in the regulation of appetite and energy homeostasis in vertebrates. While the phenotype of morbid obesity due to leptin (Lep) or leptin receptor (LEPR) loss of function is well established in mammals, evidence in fish is controversial, questioning the role of leptin as the vertebrate adipostat. Here we report on three (Lepr) loss of function (LOF) and one leptin loss of function alleles in zebrafish. In order to demonstrate that the Lepr LOF alleles cannot transduce a leptin signal, we measured socs3a transcription after i.p. leptin which is abolished by Lepr LOF. None of the Lepr/Lepa LOF alleles leads to obesity/a body growth phenotype. We explore possible reasons leading to the difference in published results and find that even slight changes in background genetics such as inbreeding siblings and cousins can lead to significant variance in growth.
Introduction
Research into the genetic basis of obesity was initiated when breeding experiments at Jackson labs gave birth to a pair of plump mice in the 1960s. Tracing the genes responsible for the aptly named obese (ob/ob) and diabetes (db/db) alleles led to the milestone discovery of the leptin system. The hormone leptin (ob gene) is released by adipose tissue in proportion to fat stores. Leptin signals the status of peripheral adipose stores to the brain, where it is received by the melanocortin system which expresses the leptin receptor (db gene, reviewed by Coleman 2010). Leptin deficiency in humans as well as mice leads to morbid obesity, which can be cured by leptin supplementation (Farooqi & O’Rahilly 2014). The genes involved in energy homeostasis are conserved across all vertebrates (Volkoff 2016) including leptin (Lep), the hormone receptor (Lepr), and the leptin pathway which involves Janus kinase 2 (Jak2), signal transducer and activator of transcription 3 (Stat3) and suppressor of cytokine signaling 3 (Socs3) (Liongue et al. 2016). While fish Lep and Lepr share minimal amino acid similarity with their vertebrate counterparts, gene synteny and predicted protein structure are well conserved (Denver et al. 2011, Prokop et al. 2012). Recently, Jak2 and Stat3 were shown to be phosphorylated by trout leptin in trout cell culture, providing evidence for a functional leptin pathway (Gong & Bjornsson 2014). However, evidence for conservation of leptin function at a physiological level has been mixed and it is unclear whether fish leptin signals the status of peripheral adipose stores in the same manner as in mammals (Londraville et al. 2014, Deck et al. 2017).
To better understand the role of leptin in teleost physiology, several groups have turned to knockout models which have targeted mutations in either lep or lepr. There are currently six studies investigating a global loss of function (LOF) in fish: Lepr in medaka (Chisada et al. 2014), Lepr in zebrafish (Michel et al. 2016, Fei et al. 2017, Ahi et al. 2019), Lepa in zebrafish (Michel et al. 2016, Audira et al. 2018), and Lepb in zebrafish (He et al. 2021). The reported results, however, are not consistent amongst studies: two alleles for Lepr LOF in zebrafish have no significant obesity phenotype (Michel et al. 2016, Ahi et al. 2019) while a third allele was found to display several characteristics of obesity (Fei et al. 2017). Loss of Lepa was found to have an obesity phenotype in zebrafish (Audira et al. 2018) and LOF of Lepb led to obesity at 1.5 years of age (He et al. 2021). The obesity phenotype in medaka was only seen at an early age, after which WT and lof animals do not differ in weight or length (Chisada et al. 2014). It is worth noting at an equivalent developmental stage in mice, body weight doubles as a result of LEP or LEPR LOF and nearly all the additional weight is in adipose tissue (Ingalls et al. 1950, Hummel et al. 1966). Similarly, in children, leptin deficiency leads to morbid obesity early in childhood (Montague et al. 1997). While results differ among fish studies, the reported phenotype is comparatively mild.
We wondered whether there was a biological reason for the divergent results among the fish studies, as the differences were not only between species where life history could explain the divergence (medaka and zebrafish) but also among the different lof alleles in zebrafish. We hypothesized that the leprsa1508 mutation we studied could be a hypomorph. Importantly, none of the studies into fish leptin system dysfunction provide functional evidence indicating whether lepr or lepa gene function is ablated. Here, we show evidence that Lepr signaling in three different Lepr lof alleles is ablated by measuring socs3a transcription after leptin injection, including the previously studied leprsa1508 allele.
Another possible reason for the different results among the fish studies could be the experimental design. The studies which report an obesity phenotype raised WT or lof fish from genotype incrosses separated from each other (Chisada et al. 2014, Fei et al. 2017, Audira et al. 2018). In comparison, the studies that found no growth difference raised sibling WT and LOF animals together in a tank (Michel et al. 2016, Ahi et al. 2019). There are two factors in this difference – genetic distance and behavior in the same tank. On the genetic level, fish from heterozygous incrossing are siblings, and the genetic background varies only very little. In contrast, keeping WT and lof lines separated from each other increases the genetic distance and could segregate background mutations. This may not be a critical factor in a highly inbred species such as the laboratory mouse, but it could lead to significant differences in a highly outbred model such as the zebrafish (LaFave et al. 2014, Doran et al. 2016). On the behavioral level, raising WT and heterozygous fish together with homozygous LOF fish in the same tank could lead to food competition among genotypes, and homozygous Lepr LOF fish might not be able to develop an obese phenotype in this environment. Factors such as appetite, stress, and social dominance could be relevant. This is circumvented when genotypes are raised in separate tanks.
Here, we test body growth in three Lepr LOF lines and one Lepa LOF line. We test WT and lof siblings raised together in tanks and cousin WT and lof raised in separate tanks. A body growth phenotype indicative of obesity was not found in any of the Lepr LOF lines, nor in an independent Lepa LOF line when siblings were raised together in a tank. When cousins were raised separated from each other, significant growth differences between genotypes were observed, however, in some cases, the lof animals were smaller and in others bigger. A consistent obesity phenotype was not found, providing independent evidence that Lepr or Lepa LOF do not lead to obesity in zebrafish.
Materials and methods
Zebrafish strains and maintenance
Animals were raised in an AquaSchwarz system on a 14 h light:10 h darkness cycle according to Brand et al. (2002). Groups of 30 larvae were raised and were fed with 100 mL paramecia as well as a pinch of red breeze food supplement (Preis Aquaristik) each day. After 14 days until the end of the experiment, fish were fed with 2 × 3 mL of 2-days old dense Artemia per day plus either a pinch of red breeze food supplement or a pinch of spirulina powder on alternating days. We reduced density to 10 fish per 2.5 L tank at 30 dpf.
Zebrafish strains used
The Lepr mutant strain leprsa1508 was obtained from the Zebrafish International Resource Center (Oregon, USA). The other lines were established using CRISPR/Cas9 gene-editing according to Jao et al. (2013), with genotyping performed using a heteroduplex mobility assay according to Ota et al. (2013). Fish were outcrossed at least five generations before this study.
Specifically, the Lepr lines leprvu624 and leprvu625 were established using a gRNA against the target ATGATGAAGACAGACCTAGG in exon 3, from which two F1 lines were generated. Compared to GenBank BN000731.1 or RefSeq NM_001113376.1, the mutations are as follows: leprvu624: deletion of A34, insertion of GTC; leprvu625: deletion C29-G35, insertion of TGTCTTGATCATGCAGATGTCCTCTCTT. These mutations in exon 3 delete 1 bp and insert 3 bp in the leprvu624 allele and delete 7 bp and insert 28 bp in the leprvu625 allele resulting either in a frameshift and early truncation of the open reading frame or a disruption of the splice site depending on the predicted isoform analyzed. The Lepa mutant line lepavu622 was established using a gRNA to the target AATCTCTGGATAATGTCCTGG as described (Michel et al. 2016) which compared to RefSeq NM_001128576 has a deletion of C184-A188.
For genotyping, a heteroduplex mobility assay was employed using the following primers: leprvu624 and leprvu625 5’ACCCACACTAATGCGTCTCTG3’ × 5’AAGCTTAAAGATCGGACCATTCCA3’; lepavu622 5’CAAAGACCATCATCGTCAGA3’ × 5’ACCCAGAAGTGTGGATAGAT3’. For leprsa1508, we designed cleaved amplified polymorphic sequence primers according to Hodgens et al. (2017) and amplified with 5’ATGTGGAAGGATGTTCCAAATCCCAACAAGaG3’ × 5’ACAGGGGGTAAAATGACTTCAGAAAGATAATA3’; the product WT allele is cleaved by SacI, the loss of function is not.
All procedures were approved by the national animal care committee (LANUV Nordrhein-Westfalen) and the University of Cologne and/or the University of Michigan Animal Care and Use Committee.
Body growth
All animals from one experimental cohort were assessed on the same day by an experimenter blinded to the genotype. At indicated times, fish were anesthetized with tricaine, blotted dry, photographed on millimeter paper, weighed, fin-clipped, and placed in a recovery tank. Standard length was measured according to Parichy et al. (2009) from pictures with Fiji (ImageJ). Fulton’s condition factor (K) was calculated according to Stevenson & Woods (2006) and is a parameter analogous to BMI.
IP injection
Recombinant mouse leptin (SIGMA) was made up to a stock concentration of 1 mg/mL in DPBS (Gibco, Thermo Fisher) and injected into female zebrafish of 3–4 months of age at 2.5 mg/kg with a phenol red concentration of 1 µL/100 mg according to Kinkel et al. (2010) with the modification that we used a 10 µL Hamilton syringe with a 28 gauge needle.
RNA extraction and qPCR
Animals were sacrificed 50 min after injection and whole livers were isolated, immediately placed on dry ice, and stored at −80°C until RNA extraction was performed. Total RNA was extracted using Trizol (Invitrogen) and purified using a PureLinkTM RNA mini kit (Thermo Fischer) following the manufacturer’s protocols. Residual genomic DNA was removed using the PureLink DNase mixture included in the kit. Final RNA concentration was determined by optical density reading at 260 nm using NanoDrop 2000c and 1 µg of RNA with a 260/280 ratio > 1.8 was used for RT using a high-capacity cDNA kit (Applied Biosystems). Primers were designed using Primer-BLAST (NCBI) with similar melting temperatures. Additionally, we attempted to span an exon–exon junction or an intron where possible. The sequences are listed in Table 1. Oligonucleotide primers were ordered from IDT.
Oligonucleotides used for qPCR experiments.
| Gene name | Refseq No. | Primer sequence (5′–3′) | Amplicon length (bp) |
|---|---|---|---|
| eef1a1l1 | NM_200009 | Forward AACATGCTTGAGGCCAGTCC Reverse ACGGTTCCGATGCCTCCA |
188 |
| socs3a | NM_199950 | Forward AAGCAGGGAAGACAAGAGCC Reverse AGAGCTGGTCAAAAGAGCCTAT |
95 |
| lepr | NM_001309403 | Forward GGGGAACCGGCTCGATACAC Reverse GCTCTCCATGGTGTCAATCTGCC |
210 |
| lepa | NM_001128576 | Forward GGAACACATTGACGGGCAAA Reverse ATGGGTTTGTCAGCGGGAAT |
87 |
Q-RT-PCR was performed with Sybr Select Master Mix (Life Technologies, Thermo Fisher Scientific) on an ABI-Prism 7500 Fast Detect system under the following conditions: initial PCR activation 50°C for 20 s, 95°C for 2 min; PCR cycles, 95°C for 15 s denaturation, 60°C for 1 min annealing and 72°C for 30 s extension, repeated for 40 cycles. Each assay was performed in triplicate including a no-RT and no-template control. Relative expression levels were calculated following a modified ΔΔCt method to include primer efficiency (Pfaffl 2001) with eef1a1l1 as the reference gene. Efficiencies and R2 were assessed using a five-point cDNA serial dilution. All the primer pairs had PCR efficiencies between 96% and 102%, and the amplicon for each primer pair was sequenced. Data are presented as fold change relative to WT, calculated based on primer efficiency using a pooled sample of the relative dataset’s cDNA, and then standardized to eef1a1l1 expression. The reference gene has previously been shown to be a stable housekeeping gene (McCurley & Callard 2008) and was tested here to verify that it did not vary among genotypes.
μCT imaging and analysis
Zebrafish were anesthetized, fixed in 4% PFA, and imaged using a micro-CT (μCT) device (U-CT OI, MILabs B.V., Utrecht, the Netherlands). Zebrafish were placed on an acrylic glass plate and an ultra-focus scan over the whole body was performed in a full-rotation in step-and-shoot mode. One thousand four hundred and forty projections (1944 × 1536 pixels) were acquired per sub scan with an x-ray tube voltage of 60 kV, power 0.16 mA and exposure time of 75 ms and 4 averages. All CT images were reconstructed at an isotropic voxel size of 20 µm using a Feldkamp type algorithm (filtered back-projection). The fat-containing tissue regions, which appear hypo-intense in the μCT data, were segmented, and the volumetric fat percentage was computed as the ratio of (s.c. and visceral) fat volume to the entire body volume using Imalytics Preclinical (Gremse-IT GmbH, Aachen, Germany (Gremse et al. 2016).
Statistical analysis
Statistics were carried out in GraphPad Prism version 9.1.1 for Windows, GraphPad Software. Two-tailed, unpaired t-tests were carried out to compare lof and WT data. One-way ANOVA was followed by Dunnett’s multiple comparisons test. P-values less than 0.05 were considered significant. Violin plots were chosen to represent the data because in addition to the frequency distribution of the data, these also highlight the median and quartiles.
Results
All lepr alleles abolish Lepr signal transduction
To determine whether we had functional Lepr ablation, we tested whether the mRNA undergoes non-sense-mediated decay (NMD) in any of the Lepr LOF alleles. Interestingly, none of the alleles underwent NMD (Fig. 1). No significant difference was found in lepr mRNA in WT compared to lof siblings in the leprsa1508 allele (t(16) = 0.09094, P = 0.9287, Fig. 1A), the leprvu624 allele (t(16) = 0.6938, P = 0.4978, Fig. 1B) or the leprvu625 allele (t(16) = 0.03766, P = 0.9704, Fig. 1C). Consequently, the lepr mRNA is stable in each of the Lepr LOF alleles compared to their respective WT siblings.
No evidence of lepr mRNA decay in any of the Lepr LOF alleles (A) the leprsa1508 allele; (B) the leprvu624 allele; (C) the leprvu625 allele. (ns, no significant difference; t-test). Expression data are shown as fold change relative to controls. Data represented as violin plots (n = 9 fish/group).
Citation: Journal of Endocrinology 251, 1; 10.1530/JOE-21-0037
No evidence of lepr mRNA decay in any of the Lepr LOF alleles (A) the leprsa1508 allele; (B) the leprvu624 allele; (C) the leprvu625 allele. (ns, no significant difference; t-test). Expression data are shown as fold change relative to controls. Data represented as violin plots (n = 9 fish/group).
Citation: Journal of Endocrinology 251, 1; 10.1530/JOE-21-0037
No evidence of lepr mRNA decay in any of the Lepr LOF alleles (A) the leprsa1508 allele; (B) the leprvu624 allele; (C) the leprvu625 allele. (ns, no significant difference; t-test). Expression data are shown as fold change relative to controls. Data represented as violin plots (n = 9 fish/group).
Citation: Journal of Endocrinology 251, 1; 10.1530/JOE-21-0037
To determine whether leptin can induce socs3a expression in WT control animals, heterologous mouse leptin was injected intra-peritoneally into adult fish. We found that 2.5 mg/kg was the minimal effective dose to induce hepatic socs3a expression 50 min after injection compared to vehicle-injected animals (Fig. 2A). When leptin was injected into control and leprsa1508 LOF animals, we saw a significant increase in socs3a expression in WT but not lof siblings (ANOVA F(3,40) = 145.6, P < 0.0001, Fig. 2B). The loss of leptin-induced socs3a expression replicated in the leprvu624 line (ANOVA F(3,32) = 107.5, P < 0.0001, Fig. 2C) as well as the leprvu625 line (ANOVA F(3,32) = 8.644, P < 0.0002, Fig. 2D). These data provide evidence that all three Lepr LOF alleles abolish leptin-induced socs3a transcription.
Intraperitoneal leptin injection can induce hepatic socs3a expression and Lepr LOF can block this effect. (A) titration of IP leptin on hepatic socs3a expression; (B) The effect of IP leptin in the leprsa1508 allele; (C) the leprvu624 allele and (D) the leprvu625 allele. One-way ANOVA was followed by Dunnett’s multiple comparisons test. Asterisks indicate significant effects against vehicle-injected control (***P < 0.001, **P < 0.01, and *P < 0.05; ANOVA). Expression data are shown as fold change relative to controls. Data represented as violin plots (n = 9–11 fish/group).
Citation: Journal of Endocrinology 251, 1; 10.1530/JOE-21-0037
Intraperitoneal leptin injection can induce hepatic socs3a expression and Lepr LOF can block this effect. (A) titration of IP leptin on hepatic socs3a expression; (B) The effect of IP leptin in the leprsa1508 allele; (C) the leprvu624 allele and (D) the leprvu625 allele. One-way ANOVA was followed by Dunnett’s multiple comparisons test. Asterisks indicate significant effects against vehicle-injected control (***P < 0.001, **P < 0.01, and *P < 0.05; ANOVA). Expression data are shown as fold change relative to controls. Data represented as violin plots (n = 9–11 fish/group).
Citation: Journal of Endocrinology 251, 1; 10.1530/JOE-21-0037
Intraperitoneal leptin injection can induce hepatic socs3a expression and Lepr LOF can block this effect. (A) titration of IP leptin on hepatic socs3a expression; (B) The effect of IP leptin in the leprsa1508 allele; (C) the leprvu624 allele and (D) the leprvu625 allele. One-way ANOVA was followed by Dunnett’s multiple comparisons test. Asterisks indicate significant effects against vehicle-injected control (***P < 0.001, **P < 0.01, and *P < 0.05; ANOVA). Expression data are shown as fold change relative to controls. Data represented as violin plots (n = 9–11 fish/group).
Citation: Journal of Endocrinology 251, 1; 10.1530/JOE-21-0037
Growth parameters of leprvu624, leprvu625 and lepavu622 – evidence from a heterozygous incross
We tested whether the leprvu624 or leprvu625 lines develop an obesity phenotype. The study of Fei et al. suggests that there is a feeding-induced component of the obesity phenotype (Fei et al. 2017), and therefore, animals raised at low density were fed excess food according to Leibold & Hammerschmidt (2015). Offspring of a heterozygous lepr+/− x lepr+/− cross were raised, and these experiments were performed in animals derived from the same cross to control for cryptic background alleles in the genome. The leprsa1508 allele was previously tested under these conditions with no observed differences in weight (Michel et al. 2016).
At 3 months of age (89 dpf), animals of the leprvu624 allele displayed no difference in standard length (SL) in either gender (Fig. 3A; SL female: t(41) = 1.583, P = 0.1212; male: t(20) = 0.7716, P = 0.4494). We scored the mass of the different genotypes and again saw no significant difference in either gender (Fig. 3B; weight, female: t(41) = 1.574, P = 0.1233; male: t(20) = 1.337, P = 0.1963). The condition factor of these fish was also calculated and again no significant difference was found between genotypes in either gender of fish (Fig. 3C; CF female: t(41) = 0.1413, P = 0.8883; male: t(20) = 0.9979, P = 0.3303).
Offspring of a heterozygous × heterozygous cross raised in tanks of mixed genotypes do not show a growth phenotype. Animals of either line were raised in tanks of ten sibling animals under the same feeding conditions until (A, B and C) leprvu624 – 89dpf; n = 21/22/11/11 (as seen on the graph female +/+ & −/−; male +/+ & −/−); (D, E and F) leprvu625 – 92dpf; female n = 17/32/9/16; (G, H and I) lepavu622 – 110dpf; female n = 13/16/12/12; Indicated is the standard length (A, D and G), weight (B, E and H) and condition factor (C, F and I). Unpaired t-tests were used to compare between genotypes. Asterisks indicate significant effects against the control (*P < 0.05; t-test). Data represented as violin plots (+/+ wildtype; −/− loss of function).
Citation: Journal of Endocrinology 251, 1; 10.1530/JOE-21-0037
Offspring of a heterozygous × heterozygous cross raised in tanks of mixed genotypes do not show a growth phenotype. Animals of either line were raised in tanks of ten sibling animals under the same feeding conditions until (A, B and C) leprvu624 – 89dpf; n = 21/22/11/11 (as seen on the graph female +/+ & −/−; male +/+ & −/−); (D, E and F) leprvu625 – 92dpf; female n = 17/32/9/16; (G, H and I) lepavu622 – 110dpf; female n = 13/16/12/12; Indicated is the standard length (A, D and G), weight (B, E and H) and condition factor (C, F and I). Unpaired t-tests were used to compare between genotypes. Asterisks indicate significant effects against the control (*P < 0.05; t-test). Data represented as violin plots (+/+ wildtype; −/− loss of function).
Citation: Journal of Endocrinology 251, 1; 10.1530/JOE-21-0037
Offspring of a heterozygous × heterozygous cross raised in tanks of mixed genotypes do not show a growth phenotype. Animals of either line were raised in tanks of ten sibling animals under the same feeding conditions until (A, B and C) leprvu624 – 89dpf; n = 21/22/11/11 (as seen on the graph female +/+ & −/−; male +/+ & −/−); (D, E and F) leprvu625 – 92dpf; female n = 17/32/9/16; (G, H and I) lepavu622 – 110dpf; female n = 13/16/12/12; Indicated is the standard length (A, D and G), weight (B, E and H) and condition factor (C, F and I). Unpaired t-tests were used to compare between genotypes. Asterisks indicate significant effects against the control (*P < 0.05; t-test). Data represented as violin plots (+/+ wildtype; −/− loss of function).
Citation: Journal of Endocrinology 251, 1; 10.1530/JOE-21-0037
We also tested growth in the leprvu625 allele. Comparing to the other two alleles under these raising conditions, no obesity phenotype at 3 months (92 dpf) of age was noted in standard length (Fig. 3D; SL female: t(47) = 1.464, P = 0.15; male: t(23) = 0.9099, P = 0.3723), animal weight (Fig. 3E; weight, female: t(47) = 0.1956, P = 0.8458; male: t(23) = 0.9468, P = 0.3536) or in animal condition factor (Fig. 3F; CF female: t(47) = 1.326, P = 0.1912; male: t(23) = 0.6883, P = 0.4981).
In the mouse, mutations in the leptin receptor and its ligand, leptin, result in nearly identical phenotypes. We previously reported that a trans-heterozygote for lepavu621/vu622 alters β-cell number in a similar way to the leprsa1508 allele in overfeeding conditions (Michel et al. 2016), but the effect of mutations in leptin on growth was not examined. lepavu622 fish raised in the same conditions as described above did not show or size phenotype: fish from a heterozygous × heterozygous cross of the lepavu622 allele were unaltered in standard length (Fig. 3G; SL, female t(27) = 0.5626, P = 0.5784; male: t(22) = 1.657, P = 0.1116), body weight (Fig. 3H; female: t(27) = 0.2222, P = 0.8258; male: t(22) = 2.045, P = 0.0530) and condition factor (Fig. 3I; female: t(27) = 0.8852, P = 0.3839; male: t(22) = 0.3593, P = 0.7228).
These results confirm that loss of leptin signaling function, either through loss of the receptor or the ligand, does not result in a change in mass, size, or condition factor in zebrafish.
Does competition among genotypes influence weight gain in lof animals?
In zebrafish work, WT x WT (lepr+/+ × lepr+/+) and lof × lof (lepr−/− × lepr−/−) crosses are sometimes used instead of het × het (lepr+/− × lepr+/−) as this can simplify the experimental setup. Fish are subsequently raised in separate tanks. For body growth, that means that the genetic background is further removed than in siblings of a het × het crossing (i.e. cousins or further removed) and that genotypes do not compete in the same environment (tank) for resources. Importantly, the studies in zebrafish that do not report a growth phenotype raise animals as a mixed genotype (Michel et al. 2016, Ahi et al. 2019), while the studies that do report a growth phenotype do not (Fei et al. 2017, Audira et al. 2018). To investigate whether this methodological difference could underlie the different outcomes among the studies, we raised offspring of a WT × WT cross and offspring of a Lepr LOF × Lepr LOF cross in separate tanks. Interestingly, significant growth differences were observed at sporadically 146 dpf.
In the leprvu624 allele, no difference in standard length between genotypes was found (Fig. 4A; SL leprvu624; female: t(40) = 1.877, P = 0.0678; male: t(24) = 0.3463, P = 0.7322), however standard length in female animals trended toward shorter LOF animals. Moreover, a significant reduction in leprvu624 weight was observed in female but not male animals, while there was a trend for reduced weight in male animals as well (Fig. 4B, weight female: t(40) = 2.079, P = 0.0441; male: t(24) = 1.769, P = 0.0896). In the conditioning factor, a trend for a reduced K in male animals was seen (Fig. 4C; female: t(40) = 0.9529, P = 0.3464; male: t(24) = 2.055, P = 0.0509).
Offspring of a WT × WT and lof × lof crosses raised separately in tanks of ten animals with the same access to food until 146 dpf. (A, B and C) the leprvu624 allele, n = 18/24/10/16 (as seen on the graph female +/+ & −/−; male +/+ & −/−); (D, E and F) the leprvu625 allele, n = 20/12/10/27; (G, H and I) the leprsa1508 allele n = 12/16/19/13 and (J, K and L) the lepavu622 allele n = 17/4/12/8. Indicated are the standard length (A, D, G, and J), weight (B, E, H and K) and condition factor (C, F, I and L). Unpaired t-tests were used to compare between genotypes. Asterisks indicate significant effects against the control (*P < 0.05; t-test). Data represented as violin plots (+/+ wildtype; −/− loss of function).
Citation: Journal of Endocrinology 251, 1; 10.1530/JOE-21-0037
Offspring of a WT × WT and lof × lof crosses raised separately in tanks of ten animals with the same access to food until 146 dpf. (A, B and C) the leprvu624 allele, n = 18/24/10/16 (as seen on the graph female +/+ & −/−; male +/+ & −/−); (D, E and F) the leprvu625 allele, n = 20/12/10/27; (G, H and I) the leprsa1508 allele n = 12/16/19/13 and (J, K and L) the lepavu622 allele n = 17/4/12/8. Indicated are the standard length (A, D, G, and J), weight (B, E, H and K) and condition factor (C, F, I and L). Unpaired t-tests were used to compare between genotypes. Asterisks indicate significant effects against the control (*P < 0.05; t-test). Data represented as violin plots (+/+ wildtype; −/− loss of function).
Citation: Journal of Endocrinology 251, 1; 10.1530/JOE-21-0037
Offspring of a WT × WT and lof × lof crosses raised separately in tanks of ten animals with the same access to food until 146 dpf. (A, B and C) the leprvu624 allele, n = 18/24/10/16 (as seen on the graph female +/+ & −/−; male +/+ & −/−); (D, E and F) the leprvu625 allele, n = 20/12/10/27; (G, H and I) the leprsa1508 allele n = 12/16/19/13 and (J, K and L) the lepavu622 allele n = 17/4/12/8. Indicated are the standard length (A, D, G, and J), weight (B, E, H and K) and condition factor (C, F, I and L). Unpaired t-tests were used to compare between genotypes. Asterisks indicate significant effects against the control (*P < 0.05; t-test). Data represented as violin plots (+/+ wildtype; −/− loss of function).
Citation: Journal of Endocrinology 251, 1; 10.1530/JOE-21-0037
In the leprvu625 allele, we found that standard length was significantly higher in male but not female LOF animals (Fig. 4D; female: t(30) = 0.5241, P = 0.6040; male: t(35) = 2.2, P = 0.0345). No differences in animal weight were observed (Fig. 4E; female: t(29) = 0.6418, P = 0.5261; male: t(36) = 1.086, P = 0.2846), but there was a significant reduction in condition factor in male LOF animals (Fig. 4F; female: t(29) = 0.04389, P = 0.9653; male: t(36) = 2.449, P = 0.0193). In this case, the significant increase in length without an equal increase in weight resulted in a slightly reduced condition factor in Lepr LOF animals.
However, as these results are not completely congruent with the results from the other Lepr LOF allele (Fig. 4A, B and C), we also tested the previously published leprsa1508 allele which has no altered growth phenotype in a heterozygous incross breeding setup (Michel et al. 2016). When raised in separate tanks as offspring of a WT × WT and lof × lof crossing scheme, no difference was found between leprsa1508 lof and WT animals in standard length (Fig. 4G; female: t(26) = 0.5140, P = 0.6116; male: t(30) = 0.2472, P = 0.8064) or weight (Fig. 4H; female: t(26) = 0.2504, P = 0.8043; male: t(30) = 0.9736, P = 0.3381). Surprisingly, the condition factor of leprsa1508 LOF male fish but not female fish was significantly higher (Fig. 4I; female t(26) = 0.1948, P = 0.8471; male: t(30) = 2.858, P = 0.0077). This result is in contrast to the overall trend seen with the other two alleles.
Lastly, we also tested the Lepa LOF allele lepavu622. Here, a significant reduction was seen in male standard length but not female standard length (Fig. 4J; female: t(19) = 0.04163, P = 0.9672; male: t (18) = 4.908, P = 0.0001), as well as in male but not female weight (Fig. 4K; female: t(19) = 0.4742, P = 0.6408; male: t(18) = 4.349, P = 0.0004). As the significant growth reduction is represented in both LOF length and weight, this did not reflect as a difference in condition factor in either male or female fish between the genotypes (Fig. 4L; female: t(19) = 0.8227, P = 0.4209; male: t(18) = 0.7406, P = 0.4685).
Since we saw differences in size that were inconsistent among alleles, we re-calculated the average change in LOF over WT animals and expressed the data in a table to gain a better overview (Table 2). For comparison, we included examples from published vertebrate studies.
Summary of growth experiments performed for the different alleles and crossing conditions.
| % Change lof compared to WT | ||||||
|---|---|---|---|---|---|---|
| SL, female | SL, male | Weight, female | Weight, male | CF, female | CF, male | |
| Het inx | ||||||
| leprvu624 | −3.1 | +2.0 | −9.8 | +9.2 | +0.4 | +4.1 |
| leprvu625 | +3.8 | −3.0 | +1.5 | −7.1 | −7.1 | +2.8 |
| lepavu622 | −1.4 | −3.6 | +1.9 | −13.2 | +5.5 | −2.5 |
| WT and LOF inx | ||||||
| leprsa1508 | +0.7 | −0.4 | +1.6 | +5.8 | −1.0 | +8.2 |
| leprvu624 | −2.8 | −0.6 | −12.4 | −8.2 | −3.9 | −6.6 |
| leprvu625 | −1.0 | +4.7 | −4.3 | +6.2 | +0.2 | −10.0 |
| lepavu622 | −0.1 | −9.4 | +5.2 | −27.7 | +5.1 | +2.4 |
| Mouse | Weight |
|||||
| ob/ob | +210.3 | Male/female mixed | ||||
| db/db | +58.8 | Male/female mixed | ||||
| Human | ||||||
| ob/ob | +100 | 9-year-old boy | ||||
Data are expressed as % increase or % decrease from control animals. Significant differences for increases and decreases are shown as bold. The mouse and human data are adapted from: ob/ob (Ingalls et al. 1950), db/db (Hummel et al. 1966) and a human patient from Montague et al. (1997).
Is there a lipodystrophy phenotype underlying the growth parameters?
One possibility is that LOF affects body fat accumulation and/or distribution, even without leading to a change in overall weight or length. Indeed, a recent paper reported that Lepb LOF zebrafish show an increased body fat percentage at 1.5 years of age (He et al. 2021). μCT analysis was carried out in 1-year-old leprsa1508 WT and LOF fish of comparable weight and standard length, but no differences were found in total body fat (Fig. 5A, t(8) = 0.1315, P = 0.8986), s.c. adipose tissue (SAT, t(8) = 0.5358, P = 0.6067) or visceral adipose tissue (VAT, t(8) = 1.179, P = 0.2722). In order to ascertain that this is not a very late developing phenotype, we followed animals with the leprvu625 and lepavu622 LOF alleles to 1.5 years of age. Again, no differences were observed among genotypes for either leprvu625 (Fig. 5B, total fat t(7) = 0.8490, P = 0.4240; SAT t(7) = 0.4168, P = 0.6893 or VAT t(8) = 0.1573, P = 0.8789) or lepavu622 (Fig. 5C, total fat t(6) = 0.5592, P = 0.5962; SAT t(5) = 1.055, P = 0.3396 or VAT t(7) = 0.4059, P = 0.6944).
Size-matched adult fish were compared for whole-body adiposity using μCT imaging. (A) representative transverse section through leprsa1508 WT and lof zebrafish including the abdominal cavity. Fat-containing tissue regions appear hypo-intense and are colored. (B) Representative three-dimensional volume rendering of segmented bones and fat (left panel) as well as (C) fat alone (right panel). (D, E and F) quantification of fat subdivided into total body fat, s.c. adipose tissue (SAT) and visceral adipose tissue (VAT) for the (D) leprsa1508 allele, the (E) leprvu625 allele, and the (F) lepavu622 allele. Unpaired t-tests were used to compare between genotypes. Asterisks indicate significant effects against the control (*P < 0.05; t-test). Data represented as violin plots (+/+ wildtype; −/− loss of function).
Citation: Journal of Endocrinology 251, 1; 10.1530/JOE-21-0037
Size-matched adult fish were compared for whole-body adiposity using μCT imaging. (A) representative transverse section through leprsa1508 WT and lof zebrafish including the abdominal cavity. Fat-containing tissue regions appear hypo-intense and are colored. (B) Representative three-dimensional volume rendering of segmented bones and fat (left panel) as well as (C) fat alone (right panel). (D, E and F) quantification of fat subdivided into total body fat, s.c. adipose tissue (SAT) and visceral adipose tissue (VAT) for the (D) leprsa1508 allele, the (E) leprvu625 allele, and the (F) lepavu622 allele. Unpaired t-tests were used to compare between genotypes. Asterisks indicate significant effects against the control (*P < 0.05; t-test). Data represented as violin plots (+/+ wildtype; −/− loss of function).
Citation: Journal of Endocrinology 251, 1; 10.1530/JOE-21-0037
Size-matched adult fish were compared for whole-body adiposity using μCT imaging. (A) representative transverse section through leprsa1508 WT and lof zebrafish including the abdominal cavity. Fat-containing tissue regions appear hypo-intense and are colored. (B) Representative three-dimensional volume rendering of segmented bones and fat (left panel) as well as (C) fat alone (right panel). (D, E and F) quantification of fat subdivided into total body fat, s.c. adipose tissue (SAT) and visceral adipose tissue (VAT) for the (D) leprsa1508 allele, the (E) leprvu625 allele, and the (F) lepavu622 allele. Unpaired t-tests were used to compare between genotypes. Asterisks indicate significant effects against the control (*P < 0.05; t-test). Data represented as violin plots (+/+ wildtype; −/− loss of function).
Citation: Journal of Endocrinology 251, 1; 10.1530/JOE-21-0037
Discussion
Can the study of different alleles explain the difference in growth phenotypes?
Several groups working with engineered mutations in the zebrafish leptin signaling system have reported different phenotypes, either finding a role for leptin in body weight regulation (Chisada et al. 2014, Fei et al. 2017, Audira et al. 2018, He et al. 2021) or not (Michel et al. 2016, Ahi et al. 2019). Several scenarios could explain these divergent results. It could be that some alleles are hypomorphs and do not, in fact, ablate function. The leprsa1508 mutation, for example, leads to a stop codon in the 17th exon. This truncation allele is predicted to abolish JAK/STAT signaling, however, the mutation is far downstream, raising the possibility that the intact N terminal domain may retain some cryptic signaling ability. Indeed, none of the above-published articles presents functional evidence regarding the ablation of leptin signaling. Therefore, we tested whether the canonical downstream Lepr pathway was intact in our alleles by measuring socs3a expression induced by leptin injection.
The canonical signaling pathway of the Lepr is through Jak/Stat activation (Gong & Bjornsson 2014, Peelman et al. 2014) and downstream socs3a transcription (Kwon et al. 2016). Recombinant mouse leptin was chosen because the predicted structure of leptin is well conserved between vertebrates (Denver et al. 2011, Prokop et al. 2012). Indeed, a previous study in tilapia showed comparable effects of recombinant mammalian leptin with recombinant tilapia leptin (Baltzegar et al. 2014). The major concern against the use of heterologous leptin is that the heterologous hormone will have a different dissociation constant and higher off-target activity than the species-specific source and will possibly activate a different receptor system altogether. Previous studies have reported different kinetics between species-specific and heterologous leptin in fish (Baltzegar et al. 2014, Shpilman et al. 2014). That is possible in this system as well; however, we picked a timepoint where socs3a is significantly elevated after heterologous leptin injection and a minimal effective dose to achieve this. Therefore, the induction of the canonical Lepr signaling pathway through socs3a expression is specifically blocked by the presented mutations in lepr. It remains possible that aspects of the Lepr signaling system which do not act through socs3a remain intact especially since the mRNA of lepr does not decay in any of these lines.
As sometimes an allele can have unintended effects such as being an unexpected hypomorph rather than the predicted amorph, we generated two new alleles of lepr using CRISPR-mediated mutagenesis. These two alleles were tested with an injection of recombinant mouse leptin, and again leptin-induced socs3 expression was abolished in these Lepr LOF fish. The presented evidence suggests that a false-negative result is unlikely.
Can the experimental setup explain the difference in growth phenotypes?
We noticed two key differences in the experimental setup among the published Lep/Lepr LOF studies: the breeding scheme and the way animals were raised. One set of studies reported crossing heterozygous parents and raising the offspring in tanks of mixed genotype and no growth phenotype (Michel et al. 2016, Ahi et al. 2019). The other reported breeding from a WT line and a LOF line and raising in separate tanks and found a phenotype (Chisada et al. 2014; X Wang, personal communication) or an unreported cross raised separately (Audira et al. 2018). As far as we can see, each study represents one cohort of animals.
The breeding scheme can have a large effect on body growth. We found significant growth differences in WT animals not only among individuals within the same clutch but also among clutches of different parental pairs (personal observation). This is likely due to the significant variation found in zebrafish laboratory stocks which are highly outbred (Parichy et al. 2009, LaFave et al. 2014). This stands in contrast to reasonably uniform growth among individuals in mouse research, where mouse lines have been bred to be free of significant variation (LaFave et al. 2014). Background variation would lead to differences in growth among clutches of WT and lof fish that are not consistent among cohorts.
Raising genotypes separated from each other lead to a loss of competition between WT and lof animals which may uncover a growth phenotype. One could hypothesize a different stress sensitivity between genotypes for example (Deck et al. 2017) or competition for a scarce resource such as food.
When we raised Lepa and Lepr LOF alleles as siblings (from a heterozygous incross) in mixed genotype tanks no difference in body growth among genotypes was observed. To test whether background genetics or competition could have affected the different results across labs we also raised WT and LOF crosses separately while very carefully ascertaining that density and access to food were the same across genotypes and tanks. To our surprise, we found significant growth differences between genotypes in this setup. However, there was no consistent obesity phenotype in lof animals and instead, we found that in some cases, lof animals showed decreased and in others increased growth (Table 2). This is particularly evident in male fish, where we expect weight fluctuations could be more evident as energy is primarily used for muscle or storage, while female fish carry up to 25% of their body weight in egg mass which undergoes constant flux (Leibold & Hammerschmidt 2015). This observation is consistent with the hypothesis that inbreeding of the WT and the lof allele separately can exacerbate cryptic variation which in turn can lead to differences in growth, particularly in male fish.
These results stand in contrast to the role of leptin in mammalian biology, where leptin has a high penetrance in obesity. The original studies in mice report an increase of body adiposity from 9% body fat to up to 42.3% body fat in LEPR and LEP LOF mice on the background of a >40% weight increase without effect on length (Coleman 1978). The reported body growth phenotype of Lepr or Lep LOF in fish in contrast is relatively small (Chisada et al. 2014, Fei et al. 2017, Audira et al. 2018).
Toward the biological role of leptin in fish
If the reported body growth phenotype for Lepr/Lep LOF fish can be partially or fully attributed to experimental differences, what then is the biological role of leptin in fish, and does leptin still play a role in metabolic processes?
The observation that different Lepr or Lep LOF alleles do not lead to a reproducible obesity phenotype in fish argues against the role of leptin as the fish adipostat. Leptin is not even expressed in adipose tissue of fish (Londraville et al. 2014, Deck et al. 2017). In mammals, loss of leptin signaling leads to the inability of the brain to perceive peripheral fat stores resulting in severe hyperphagia and morbid obesity (Barsh & Schwartz 2002). This suggests that the role of leptin as the main adipostat in mammals could be a neofunctionalization of the leptin gene.
It is, however, unclear whether results can be generalized across fish or whether fish families have adapted differently due to the large diversity in the fish orders (Ronnestad et al. 2017). A recent study in trout adipocytes, for example, found that leptin can be secreted in cell culture (Salmeron et al. 2015). In mammalian lof models, dysfunction can be seen in blood sugar control, reproduction, osteogenesis, thermogenesis, angiogenesis, immune function, hematopoiesis, and arterial pressure control, and it is not easy to tease specific and local effects apart from morbid obesity and lipotoxicity (Friedman 2014). Indeed, in LEPR LOF db/db mice, severe hyperglycemia precedes the obesity phenotype (Hummel et al. 1966). In fish, a role for leptin in glucose mobilization was identified in tilapia (Baltzegar et al. 2014, Mankiewicz et al. 2021) and zebrafish (Michel et al. 2016, Fei et al. 2017, Audira et al. 2018, He et al. 2021). Dysregulated glucose levels could in turn lead to a later adiposity phenotype and authors report an elevation of visceral fat in Lepr LOF medaka (Chisada et al. 2014) and a trend toward an increased whole-body triglyceride level in Lepr LOF animals at 1 year of age (Michel et al. 2016) as well as increased adiposity at an advanced age in Lepb LOF animals (He et al. 2021). We did not observe an increase in body adiposity in this study even at 1.5 years of age, but it would be interesting to elucidate the exact interplay between Lepa, Lepb and Lepr in the etiology of glucose dysregulation and possible late adiposity changes.
This study provides evidence that growth conditions need to be very carefully monitored throughout fish life in order to study metabolism and obesity, and that fish do not show signs of complete adipose dysregulation after the loss of leptin signaling that is comparable to the role of leptin in mice. The lack of a gross bodyweight phenotype up to at least a year of age is a strength of fish LOF models in the study of the biological role of leptin as the level of adiposity in mouse LOF models significantly confounds mammalian studies. It is, therefore, feasible in fish to study the role of leptin on single organs and paracrine/endocrine systems.
Declaration of interest
The authors declare that there is no conflict of interest that could be perceived as prejudicing the impartiality of the research reported.
Funding
This work did not receive any specific grant from funding agencies in the public, commercial or not-for-profit sectors.
Acknowledgements
The authors would like to thank Matthias Hammerschmidt for his support as well as the Hammerschmidt and Cone labs, particularly Joy Armistead, for fruitful discussions.
References
Ahi EP, Brunel M, Tsakoumis E & Schmitz M 2019 Transcriptional study of appetite regulating genes in the brain of zebrafish (Danio rerio) with impaired leptin signalling. Scientific Reports 9 20166. (https://doi.org/10.1038/s41598-019-56779-z)
Audira G, Sarasamma S, Chen JR, Juniardi S, Sampurna BP, Liang ST, Lai YH, Lin GM, Hsieh MC & Hsiag CD 2018 Zebrafish mutants carrying leptin a (lepa) gene deficiency display obesity, anxiety, less aggression and fear, and circadian rhythm and color preference dysregulation. International Journal of Molecular Sciences 19 4038. (https://doi.org/10.3390/ijms19124038)
Baltzegar DA, Reading BJ, Douros JD & Borski RJ 2014 Role for leptin in promoting glucose mobilization during acute hyperosmotic stress in teleost fishes. Journal of Endocrinology 220 61–72. (https://doi.org/10.1530/JOE-13-0292)
Barsh GS & Schwartz MW 2002 Genetic approaches to studying energy balance: Perception and integration. Nature Reviews: Genetics 3 589–600. (https://doi.org/10.1038/nrg862)
Brand M, Granto M & Nusslein-Volhard C 2002 Keeping and raising zebrafish. In Zebrafish-A Practical Approach. Eds Dahm RN-V Oxford, UK: Oxford University Press.
Chisada S, Kurokawa T, Murashita K, Ronnestad I, Taniguchi Y, Toyoda A, Sakaki Y, Takeda S & Yoshiura Y 2014 Leptin receptor-deficient (knockout) medaka, Oryzias latipes, show chronical up-regulated levels of orexigenic neuropeptides, elevated food intake and stage specific effects on growth and fat allocation. General and Comparative Endocrinology 195 9–20. (https://doi.org/10.1016/j.ygcen.2013.10.008)
Coleman DL 1978 Obese and diabetes: two mutant genes causing diabetes-obesity syndromes in mice. Diabetologia 14 141–148. (https://doi.org/10.1007/BF00429772)
Coleman DL 2010 A historical perspective on leptin. Nature Medicine 16 1097–1099. (https://doi.org/10.1038/nm1010-1097)
Deck CA, Honeycutt JL, Cheung E, Reynolds HM & Borski RJ 2017 Assessing the functional role of leptin in energy homeostasis and the stress response in vertebrates. Frontiers in Endocrinology 8 63. (https://doi.org/10.3389/fendo.2017.00063)
Denver RJ, Bonett RM & Boorse GC 2011 Evolution of leptin structure and function. Neuroendocrinology 94 21–38. (https://doi.org/10.1159/000328435)
Doran AG, Wong K, Flint J, Adams DJ, Hunter KW & Keane TM 2016 Deep genome sequencing and variation analysis of 13 inbred mouse strains defines candidate phenotypic alleles, private variation and homozygous truncating mutations. Genome Biology 17 167. (https://doi.org/10.1186/s13059-016-1024-y)
Farooqi IS & O’Rahilly S 2014 Twenty years of leptin: human disorders of leptin action. Journal of Endocrinology 223 T63–T70. (https://doi.org/10.1530/JOE-14-0480)
Fei F, Sun SY, Yao YX & Wang X 2017 Generation and phenotype analysis of zebrafish mutations of obesity-related genes lepr and mc4r. Sheng Li Xue Bao: Acta Physiologica Sinica 69 61–69.
Friedman J 2014 Twenty years of leptin: an overview. Journal of Endocrinology 223 T1–T8. (https://doi.org/10.1530/JOE-14-0405)
Gong N & Bjornsson BT 2014 Leptin signaling in the rainbow trout central nervous system is modulated by a truncated leptin receptor isoform. Endocrinology 155 2445–2455. (https://doi.org/10.1210/en.2013-2131)
Gremse F, Stark M, Ehling J, Menzel JR, Lammers T & Kiessling F 2016 Imalytics preclinical: interactive analysis of biomedical volume data. Theranostics 6 328–341. (https://doi.org/10.7150/thno.13624)
He J, Ding Y, Nowik N, Jager C, Eeza MNH, Alia A, Baelde HJ & Spaink HP 2021 Leptin deficiency affects glucose homeostasis and results in adiposity in zebrafish. Journal of Endocrinology 249 125–134. (https://doi.org/10.1530/JOE-20-0437)
Hodgens C, Nimchuk ZL & Kieber JJ 2017 Indcaps: a tool for designing screening primers for CRISPR/Cas9 mutagenesis events. PLoS ONE 12 e0188406. (https://doi.org/10.1371/journal.pone.0188406)
Hummel KP, Dickie MM & Coleman DL 1966 Diabetes, a new mutation in the mouse. Science 153 1127–1128. (https://doi.org/10.1126/science.153.3740.1127)
Ingalls AM, Dickie MM & Snell GD 1950 Obese, a new mutation in the house mouse. Journal of Heredity 41 317–318. (https://doi.org/10.1093/oxfordjournals.jhered.a106073)
Jao LE, Wente SR & Chen W 2013 Efficient multiplex biallelic zebrafish genome editing using a CRISPR nuclease system. PNAS 110 13904–13909. (https://doi.org/10.1073/pnas.1308335110)
Kinkel MD, Eames SC, Philipson LH & Prince VE 2010 Intraperitoneal injection into adult zebrafish. Journal of Visualized Experiments 42 2126. (https://doi.org/10.3791/2126)
Kwon O, Kim KW & Kim MS 2016 Leptin signalling pathways in hypothalamic neurons. Cellular and Molecular Life Sciences 73 1457–1477. (https://doi.org/10.1007/s00018-016-2133-1)
LaFave MC, Varshney GK, Vemulapalli M, Mullikin JC & Burgess SM 2014 A defined zebrafish line for high-throughput genetics and genomics: NHGRI-1. Genetics 198 167–170. (https://doi.org/10.1534/genetics.114.166769)
Leibold S & Hammerschmidt M 2015 Long-term hyperphagia and caloric restriction caused by low- or high-density husbandry have differential effects on zebrafish postembryonic development, somatic growth, fat accumulation and reproduction. PLoS ONE 10 e0120776. (https://doi.org/10.1371/journal.pone.0120776)
Liongue C, Sertori R & Ward AC 2016 Evolution of cytokine receptor signaling. Journal of Immunology 197 11–18. (https://doi.org/10.4049/jimmunol.1600372)
Londraville RL, Macotela Y, Duff RJ, Easterling MR, Liu Q & Crespi EJ 2014 Comparative endocrinology of leptin: assessing function in a phylogenetic context. General and Comparative Endocrinology 203 146–157. (https://doi.org/10.1016/j.ygcen.2014.02.002)
Mankiewicz JL, Deck CA, Taylor JD, Douros JD & Borski RJ 2021 Epinephrine and glucose regulation of leptin synthesis and secretion in a teleost fish, the tilapia (Oreochromis mossambicus). General and Comparative Endocrinology 302 113669. (https://doi.org/10.1016/j.ygcen.2020.113669)
McCurley AT & Callard GV 2008 Characterization of housekeeping genes in zebrafish: male-female differences and effects of tissue type, developmental stage and chemical treatment. BMC Molecular Biology 9 102. (https://doi.org/10.1186/1471-2199-9-102)
Michel M, Page-Mccaw PS, Chen W & Cone RD 2016 Leptin signaling regulates glucose homeostasis, but not adipostasis, in the zebrafish. PNAS 113 3084–3089. (https://doi.org/10.1073/pnas.1513212113)
Montague CT, Farooqi IS, Whitehead JP, Soos MA, Rau H, Wareham NJ, Sewter CP, Digby JE, Mohammed SN & Hurst JA et al.1997 Congenital leptin deficiency is associated with severe early-onset obesity in humans. Nature 387 903–908. (https://doi.org/10.1038/43185)
Ota S, Hisano Y, Muraki M, Hoshijima K, Dahlem TJ, Grunwald DJ, Okada Y & Kawahara A 2013 Efficient identification of TALEN-mediated genome modifications using heteroduplex mobility assays. Genes to Cells 18 450–458. (https://doi.org/10.1111/gtc.12050)
Parichy DM, Elizondo MR, Mills MG, Gordon TN & Engeszer RE 2009 Normal table of postembryonic zebrafish development: staging by externally visible anatomy of the living fish. Developmental Dynamics 238 2975–3015. (https://doi.org/10.1002/dvdy.22113)
Peelman F, Zabeau L, Moharana K, Savvides SN & Tavernier J 2014 Twenty years of leptin: insights into signaling assemblies of the leptin receptor. Journal of Endocrinology 223 T9–T23. (https://doi.org/10.1530/JOE-14-0264)
Pfaffl MW 2001 A new mathematical model for relative quantification in real-time RT-PCR. Nucleic Acids Research 29 e45. (https://doi.org/10.1093/nar/29.9.e45)
Prokop JW, Duff RJ, Ball HC, Copeland DL & Londraville RL 2012 Leptin and leptin receptor: analysis of a structure to function relationship in interaction and evolution from humans to fish. Peptides 38 326–336. (https://doi.org/10.1016/j.peptides.2012.10.002)
Ronnestad I, Gomes AS, Murashita K, Angotzi R, Jonsson E & Volkoff H 2017 Appetite-controlling endocrine systems in teleosts. Frontiers in Endocrinology 8 73. (https://doi.org/10.3389/fendo.2017.00073)
Salmeron C, Johansson M, Asaad M, Angotzi AR, Ronnestad I, Stefansson SO, Jonsson E, Bjornsson BT, Gutierrez J & Navarro I et al.2015 Roles of leptin and ghrelin in adipogenesis and lipid metabolism of rainbow trout adipocytes in vitro. Comparative Biochemistry and Physiology: Part A, Molecular and Integrative Physiology 188 40–48. (https://doi.org/10.1016/j.cbpa.2015.06.017)
Shpilman M, Hollander-Cohen L, Ventura T, Gertler A & Levavi-Sivan B 2014 Production, gene structure and characterization of two orthologs of leptin and a leptin receptor in tilapia. General and Comparative Endocrinology 207 74–85. (https://doi.org/10.1016/j.ygcen.2014.05.006)
Stevenson RD & Woods Jr WA 2006 Condition indices for conservation: new uses for evolving tools. Integrative and Comparative Biology 46 1169–1190. (https://doi.org/10.1093/icb/icl052)
Volkoff H 2016 The neuroendocrine regulation of food intake in fish: a review of current knowledge. Frontiers in Neuroscience 10 540. (https://doi.org/10.3389/fnins.2016.00540)
