Depletion of Alpha-Melanocyte-Stimulating Hormone Induces Insatiable Appetite and Gains in Energy Reserves and Body Weight in Zebrafish

The functions of anorexigenic neurons secreting proopiomelanocortin (POMC)/alpha-melanocyte-stimulating hormone (α-MSH) of the melanocortin system in the hypothalamus in vertebrates are energy homeostasis, food intake, and body weight regulation. However, the mechanisms remain elusive. This article reports on zebrafish that have been genetically engineered to produce α-MSH mutants, α-MSH−7aa and α-MSH−8aa, selectively lacking 7 and 8 amino acids within the α-MSH region, but retaining most of the other normal melanocortin-signaling (Pomc-derived) peptides. The α-MSH mutants exhibited hyperphagic phenotypes leading to body weight gain, as observed in human patients and mammalian models. The actions of several genes regulating appetite in zebrafish are similar to those in mammals when analyzed using gene expression analysis. These include four selected orexigenic genes: Promelanin-concentrating hormone (pmch), agouti-related protein 2 (agrp2), neuropeptide Y (npy), and hypothalamic hypocretin/orexin (hcrt). We also study five selected anorexigenic genes: Brain-derived neurotrophic factor (bdnf), single-minded homolog 1-a (sim1a), corticotropin-releasing hormone b (crhb), thyrotropin-releasing hormone (trh), and prohormone convertase 2 (pcsk2). The orexigenic actions of α-MSH mutants are rescued completely after hindbrain ventricle injection with a synthetic analog of α-MSH and a melanocortin receptor agonist, Melanotan II. We evaluate the adverse effects of MSH depletion on energy balance using the Alamar Blue metabolic rate assay. Our results show that α-MSH is a key regulator of POMC signaling in appetite regulation and energy expenditure, suggesting that it might be a potential therapeutic target for treating human obesity.


Fish Husbandry
All zebrafish were maintained in a controlled environment with a 14/10-h light-dark cycle at 28 • C. They were fed twice daily with brine shrimp and commercial fish food pellets. All animal experiments were conducted in accordance with the guidelines and approval of the Institutional Animal Care and Use Committee (IACUC) of the National Yang Ming Chiao Tung University, Taiwan.

Quantitative Reverse Transcription Polymerase Chain Reaction (RT-qPCR)
Total RNA was extracted using TRIzol reagent (Thermo Fisher Scientific, USA) or the RNeasy Mini Kit (QIAGEN, Hilden, Germany) and reverse transcribed using the firststrand cDNA synthesis kit (K1691; Thermo Fisher Scientific). Real-time RT-qPCR was performed using the StepOne Real-Time PCR System (Applied Biosystems, Foster City, USA). The genes and their corresponding primer sequences are listed in Supplementary  Table S1. These genes, including those encoding growth hormone (gh), growth hormone receptor b (ghrb), insulin-like growth factor 2b (igf2b), insulin-like growth factor binding protein 1b (igfbp1b), insulin-like growth factor binding protein 5b (igfbp5b), and insulin-like growth factor binding protein 6b (igfbp6b) have been validated in zebrafish [33].

In-Situ Hybridization (ISH)
The gene-specific probes were cloned by PCR into a pGEM ® -T Easy TA cloning vector (Promega, Madison, WI USA) using the primers listed in Supplementary Table S2. Antisense probes were synthesized by in vitro transcription using the DIG RNA Labelling Kit (SP6/T7) (Roche Applied Science, Mannheim, Germany). ISH was performed as described previously [34].

Quantification of Food Intake
Our procedure was adapted from that described previously [35]. The larval fish food was fluorescently labeled paramecia prepared using the lipophilic tracer 4-(4-(Didecylamino)styryl)-N-Dethylpyridinium iodide (4-Di-10-ASP; Invitrogen, Carlsbad, CA, USA). We conducted feeding of 7 dpf zebrafish in 6-well plates to allow for free swimming. At 1.5 h after feeding, the larvae were anesthetized. After two washes to remove residual paramecia, groups of five larvae were transferred into a 96-well roundbottom black plate in an anesthetic solution. The intra-abdominal fluorescent signal was measured using the Synergy™ HT Multi-Detection Reader (BioTek Instruments, Winooski, VT) in fluorescence area scan mode 11 × 11 multipoint/well, 0.1 s/point using a fluorescein filter set (excitation wavelength, 485 nm; emission wavelength, 590 nm).

Hindbrain Ventricle Injection of Zebrafish Larvae
Briefly, anesthetized fish were placed on a 3% agarose plate with water containing 0.05% MS222. The skulls were impaled with a 0.53 mm diameter needle attached to a syringe in the midline at the telencephalon-diencephalon border. The fish were injected intracranially with 4.6 nL of sterile water, an α-MSH analog (100 µM in sterile water; M4135, Sigma-Aldrich), or Melanotan II (MTII, 100 µM in sterile water; M8693, Sigma-Aldrich) into the cranial cavity by a heat-pulled glass capillary micropipette attached to microinjection equipment. After administration, the cut heads were sampled, and at least 30 larvae were pooled to yield enough total RNA.

Alamar Blue Metabolic Rate Assay
Our metabolic rate assay was adapted from an established method [38]. The assay buffer was supplemented with egg water containing 0.1% DMSO, 1% Alamar Blue (A50101, Thermo Fisher Scientific), and 4 mM sodium bicarbonate. Zebrafish were rinsed with 0.22 µm filtered egg water and pipetted into a 96-well plate (three specimens per well). Fluorescence of the plate was immediately read on the Infinite 200 PRO multimode plate reader (Tecan Group Ltd., Switzerland) with excitation at 530 nm and emission monitored at 590 nm (fluorescence area scan mode 3 × 3 multipoint/well, 10 times/point per reading). The plate was incubated in the dark at 28 • C for 24 h and then read again. Any wells containing dead fish were excluded from the analyses. The change in fluorescence from time 0 to 24 h was then calculated. The data were corrected by setting the average of the wild-type (WT) controls to 1.

Whole-Mount Oil Red O Staining
Zebrafish larvae were fixed in 4% paraformaldehyde solution in PBS overnight at 4 • C. Equal numbers of control and mutant larvae were transferred to 1.5 mL Eppendorf tubes and rinsed three times (5 min each) with PBS. After removing the PBS, the larvae were prestained with a mixture of 60% isopropyl alcohol for at least 1 h. Then, fresh 0.5% Oil Red O solution was added at 4 • C for 1 h, and larvae were washed in PBS. They were stored in 70% glycerol and imaged using a bright-field dissecting microscope (Stemi 305, Carl Zeiss AG, Oberkochen, Germany).

Growth Rate
The growth rate was recorded monthly starting at two months postfertilization (mpf) and ending at 12 mpf. Twenty fish per diet (mixed sex) were selected randomly. Body length (BL) was measured using a standard metric ruler and determined the distance from the snout to the caudal peduncle. BL (cm) and BW (mg) were used to calculate the body mass index (BMI).

Morphological and Morphometric Studies: Analysis of Zebrafish Fat Tissues and Adipocytes
These studies were carried out on histological sections according to Mon-Talbano et al. [39].

Statistical Analysis
All data are presented as the mean ± standard error of the mean (SEM). Statistical analysis was performed using analysis of variance (ANOVA) followed by Bonferroni posthoc tests. All analyses were performed using GraphPad Prism software (v. 8.0; GraphPad, San Diego, CA, USA). Differences were considered statistically significant at p < 0.05.

Generation of α-MSH Depletion Lines in Zebrafish
The TALEN-based genome editing technique was performed for the depletion of α-MSH. According to the sequence information, the targeted site of TALENs was located in the third exon of the pomca gene locus. Both arms of the designed binding arms were 18 bp long. The spacer between the two arms was 17 bp. The AgeI restriction digestion site within the space region was used for genotyping ( Figure 1A). Three pomca mutants were generated here. Two independent α-MSH-depleted mutant lines, α-MSH −7aa and α-MSH −8aa were generated with 21-bp and 24-bp bp in-frame deletions, respectively. One independent pomca mutant line, POMCa 141a , was generated with a 13-bp frameshift deletion ( Figure 1A,B). The α-MSH −7aa and α-MSH −8aa mutations resulted in in-frame deletions that produced truncated proteins of 7 amino acids (aa) and 8 aa within the α-MSH regions, respectively ( Figure 1C,D). No significantly decreased expression pattern of pomca transcripts was observed in the pituitary gland or hypothalamus of the α-MSH mutants, except for the POMCa 141a mutant compared with the WT controls at 5 dpf ( Figure 1E). The α-MSH mutants caused deficiencies of only the α-MSH peptide hormones ( Figure 1F), whereas the POMCa 141a mutant resulted in premature stop codons that produced truncated proteins of 141 aa ( Figure 1C,D), which caused a deficiency in the majority of melanocortin peptides derived from POMCa ( Figure 1F). In addition, three pomca mutants showed attenuated melanosome dispersion in dark conditions, and a significant decrease in melanosomes was observed in the POMCa 141a mutant ( Figure 1G).

Defective α-MSH Increased Food Intake in Zebrafish
To evaluate the pattern of food intake at the larval stage of the zebrafish mutants, we used a previously described feeding protocol to prepare fluorescent food composed of larval fish food with lipophilic tracer 4-Di-10-ASP-labeled paramecia [35] (see Methods). The larvae ingested this fluorescent food, which led to an accumulation of fluorescent signals within the intestine that could be visualized readily by fluorescence microscopy (Figure 2A).
The α-MSH −7aa and α-MSH −8aa larvae exhibited obvious hyperphagia at 7 dpf ( Figure 2B), when food intake significantly increased by 2.28-and 2.03-fold, respectively, compared with control WT zebrafish ( Figure 2C). However, no obvious differences between POMCa 141a and WT controls were observed in terms of food intake at 7 dpf (data not shown).

Defective α-MSH Enhanced Somatic Growth and Decreased Energy Expenditure Concomitant with Liver Steatosis in Zebrafish Larvae
Next, we tested whether growth and energy budget were proportional to food intake in the POMC mutants, and simultaneous observation was necessary to estimate these parameters. Predictably, the α-MSH mutants significantly increased in BL at 8 dpf ( Figure 3A). Somatic growth is predominantly controlled by regulating the GH/insulin-like growth

Defective α-MSH Enhanced Somatic Growth and Decreased Energy Expenditure Concomitant with Liver Steatosis in Zebrafish Larvae
Next, we tested whether growth and energy budget were proportional to food intake in the POMC mutants, and simultaneous observation was necessary to estimate these parameters. Predictably, the α-MSH mutants significantly increased in BL at 8 dpf ( Figure 3A). Somatic growth is predominantly controlled by regulating the GH/insulin-like growth factor (IGF) axis in fish [40,41]. Increased mRNA expression of genes controlling the GH/IGF axis accompanied by increased BL was detected in the pomc mutants compared with WT controls at 8 dpf ( Figure 3B). Using whole-mount ISH analysis, a significantly increased expression pattern of gh was observed in pomc mutants compared with WT at 5 dpf ( Figure 3C). Thus, these results indicate that the observed growth activation could be attributed to the upregulation of the GH/IGF axis.  To assess the potential roles of α-MSH in the metabolism of the pomc mutants for further studies of energy expenditure that might influence obesity, BW, and appetite regulation, we performed the Alamar Blue metabolic rate assay in larvae at 6-8 dpf. The assay showed increased signals of the zebrafish mutants with incubation time compared with the WT controls ( Figure 3D). All pomc mutants larvae exhibited significantly lower metabolic rates, with 53% to 84% between the 6 and 7 dpf stages. Both the α-MSH −7aa and α-MSH −8aa larvae exhibited lower metabolic rates, with decreases of 84% and 81%, respectively, compared with control WT zebrafish ( Figure 3D). However, the POMCa 141a larvae exhibited a slightly higher metabolic rate, with an increase of 37% compared with control WT zebrafish at 8 dpf ( Figure 3D). Furthermore, hepatic steatosis in the α-MSH mutants increased markedly compared with that in the WT larvae ( Figure 3E), thus confirming that lipid accumulation as an energy reserve is reflected in hepatic steatosis.

α-MSH Mutant Adults Develop Characteristic Melanocortin-Related Obesity
Homozygous α-MSH mutant larvae developed normally and were slightly larger than the WT larvae ( Figure 3A,E). Surprisingly, by measuring the growth rate of POMC mutants and WT zebrafish, a dramatic increase in BL ( Figure 4A,B) and BW ( Figure 4C) was detected in both α-MSH mutant adults compared with WT adults 12 months post-fertilization (mpf). The POMCa 141a adults showed the same BL as WT adults after 12 mpf and a slight increase in BW compared with WT adults until 9 mpf. ( Figure 4C). There were also statistically significant increases in the BMI at 10 mpf for the α-MSH mutant adults ( Figure 4D).
The BMI showed the same trend observed for BL and BW ( Figure 4D). Examination of the BW data showed a significant increase in viscera and visceral fat contents in both the α-MSH −7aa and α-MSH −8aa groups at the 12 mpf stage ( Figure 4E). Morphometric analysis of adipose tissues showed a dramatic difference in development between the α-MSH mutant adults and the WT controls. Both the α-MSH −7aa and α-MSH −8aa adult fish had larger amounts of adipose tissue, with increases of 8-and 12-fold, respectively, compared with control WT zebrafish ( Figure 4F,G). Indeed, the average area of both visceral and subcutaneous adipose tissue showed a significant increase in α-MSH mutant adults compared with WT controls, indicating that an increase in food intake and decreased energy expenditure determined the growth of an abundant layer of adipose tissue, resulting in obesity.

Pre-and Postprandial Expression of Appetite-Related Genes in α-MSH Mutants
As several neuronal peptides are involved in regulating food intake and energy balance in zebrafish, as in mammals [42][43][44], appetite-related gene expression levels were investigated in whole α-MSH mutant brains using RT-qPCR. To compare the expression pattern of appetite-related genes between the WT and the pomca mutants, we examined several important factors, including four selected orexigenic genes: Promelanin-concentrating hormone (pmch), agouti-related protein 2 (agrp2), neuropeptide Y (npy), and hypothalamic hypocretin/orexin (hcrt) ( Figure 5A). We also measured the expression levels of five selected anorexigenic genes: Brain-derived neurotrophic factor (bdnf ), single-minded homolog 1a (sim1a), corticotropin-releasing hormone b (crhb), thyrotropin-releasing hormone (trh), and prohormone convertase 2 (pcsk2) ( Figure 5B). Although the analyses showed that the orexigenic mRNA levels were slightly downregulated in POMC mutants compared with those in WT controls at 7 dpf ( Figure 5A), the anorexigenic mRNA levels were dramatically downregulated in pomca mutants compared with those in WT controls at 7 dpf ( Figure 5B). Using whole-mount ISH analysis, significantly decreased expression patterns of bdnf and sim1a were observed in pomca mutants compared with WT at 5 dpf ( Figure 5C). In addition, the intestinal orexigenic gene, ghrelin (ghrl), was dramatically upregulated in pomca mutants compared with that in WT adults ( Figure 5D). Thus, functional α-MSH signaling is clearly required for transcriptional suppression to some extent of anorexigenic and orexigenic genes and interrupts the balance of hunger and satiety, which in turn possibly increases appetite under normal feeding conditions.

Administration of a Synthetic α-MSH Analog Rescued Hyperphagic Phenotypes in α-MSH Mutants
We investigated whether hindbrain ventricle injections of an α-MSH analog and MTII into α-MSH mutant larvae could reverse the hyperphagic phenotypes ( Figure 6A). The feeding evaluation results after the administration of α-MSH analog ( Figure 6B, left) and MTII ( Figure 6C, left) were obvious in both the α-MSH −7aa and α-MSH −8aa specimens at 7dpf, respectively. Both α-MSH −7aa and α-MSH −8aa fish could be rescued by the α-MSH analog and MTII-similar to normal phenotypes-although their anorexigenic effects were stronger than in WT controls ( Figure 6B,C).
To clarify whether the appetite-related gene expression levels were modulated directly by administration of the α-MSH analog in α-MSH mutant larvae, the mRNA levels of the orexigenic genes npy, agrp2, and hcrt were significantly lower than the mRNA levels of the anorexigenic genes bdnf, sim1a, chrb, and trh at 1.5 h after intracranial administration compared with the α-MSH mutant larvae ( Figure 6D,E). Values are means ± SEM. * p < 0.05, ** p < 0.01, and *** p < 0.001 compared with WT control groups. and MTII ( Figure 6C, left) were obvious in both the α-MSH and α-MSH specimens at 7dpf, respectively. Both α-MSH −7aa and α-MSH −8aa fish could be rescued by the α-MSH analog and MTII-similar to normal phenotypes-although their anorexigenic effects were stronger than in WT controls ( Figure 6B,C).
To clarify whether the appetite-related gene expression levels were modulated directly by administration of the α-MSH analog in α-MSH mutant larvae, the mRNA levels of the orexigenic genes npy, agrp2, and hcrt were significantly lower than the mRNA levels of the anorexigenic genes bdnf, sim1a, chrb, and trh at 1.5 h after intracranial administration compared with the α-MSH mutant larvae ( Figure 6D,E). Values are means ± SEM. * p < 0.05, ** p < 0.01, and *** p < 0.001 compared with WT groups. (D,E) RT-qPCR analysis was used to measure the mRNA levels of three selected orexigenic genes, agrp2, npy, and hcrt (D), and four selected anorexigenic genes, bdnf, sim1a, crhb, and trh (E) in the WT and α-MSH mutant larvae at 1.5 h after feeding. We used relative fluorescent change and relative transcriptome expression for quantification three times (n = 50) independently.
The depletion of α-MSH in Pomc-null mice resulted in obesity, whereas the blockade of neuronal melanocortin signaling resulted in a decrease in response to centrally administered leptin [53,54]. Moreover, the central administration of α-MSH suppressed food intake and reduced BW gain in rodent models [55,56]. A deficiency of α-MSH in a Pomc tm1/tm1 mouse produced exacerbated hyperphagia and obesity when feeding with a high-fat diet (HFD) [12]. Our results-consistent with mouse models-showed significant increases in food intake in the α-MSH mutants ( Figure 2B,C). Because no α-MSH peptide was produced in α-MSH mutant fish, a significant increase in food intake was found in their larval stages ( Figure 2B,C). Increased BW and obesity in adult α-MSH mutants were also developed under regular feeding and overfeeding conditions ( Figure 4A,C). However, a zebrafish pomc mutant, POMCa 141a , which was also α-MSH deficient, showed enhanced somatic growth without increasing food intake. This suggests that some aspects in addition to regulating appetite could be the key cause(s) for BW gain in our zebrafish model of POMCa 141a .
Hyperphagia alone did not promote fish growth. In medaka and zebrafish models, leptin-receptor deficiency also elicited an increased food intake pattern with a normal growth rate [57,58]. Because the energy homeostasis controlling system is highly complicated, simply increasing food intake might not explain the causal relationship with an acceleration of the growth rate or with weight gain [59][60][61]. In the Alamar Blue metabolic assay, we observed that the α-MSH mutant decreased energy expenditure by decreasing its metabolic rate ( Figure 3D). Thus, the deletion of α-MSH in zebrafish induced hyperphagia and lower metabolism levels, which resulted in obesity. We found not only increases in food intake but also increases in the growth rate, including the standard BL, BW, and obesity, in the α-MSH mutant fish compared with the WT controls. The growth rate was determined by energy absorption and conversion efficiency, as well as the individual metabolic rate.
We selected appetite regulatory genes that have been demonstrated to take part in regulating food intake and to serve as indicators of controlling the anorexigenic and orexigenic functions and compared their expression patterns during preprandial stages between the pomc mutant fish and WT control. We observed that the anorexigenic genes (sim1a, crhb, trh, and bdnf ) were significantly downregulated in the α-MSH mutant fish at 1.5 h after feeding. To further confirm the relationship between α-MSH and appetite regulation, we used intracranial administration of an α-MSH analog and MTII. During 1.5 h after administration, we discovered that fluorescence intensities were repressed significantly in the hypothalamus of both WT and α-MSH mutant fish. Moreover, the bdnf and sim1a expression levels in the α-MSH −/− and α-MSH analog/MTII administration groups were significantly enhanced compared with the WT and saline administration control groups. Lack of α-MSH leads to suppression of anorexigenic genes expression (bdnf, sim1a, crhb, and trh) results in a reversible imbalance between food intake and energy homeostasis in α-MSH mutant fish ( Figure 7A). Most of these associated features might be regulated by α-MSH [19,45], suggesting their critical functions in controlling metabolism observed in our α-MSH mutant zebrafish ( Figure 7B). In line with previous reports, most metabolic defects observed in fish [62], mammalian models [4,63,64], and humans [65][66][67] have been attributed to the impairment of functional MC4R signaling in various pomc mutants.

Conclusions
Together with the above findings in our study, the BW gain (or obesity) and increase in body lipid contents were associated with increased appetite and decreased energy expenditure rates in this zebrafish model. In summary, we characterized a critical role for melanocortin signaling for appetite suppression. Therefore, our findings may help clarify underlying mechanisms between α-MSH signaling and obesity.

Data Availability Statement:
The data presented in this study are available upon request from the corresponding author.