**4. Discussion**

In the present survey, we characterized NUCB2B/Nesf-1 in the brain and in the mammalian stomach equivalent structure of *N. furzeri* and reported, for the first time, its regulation upon aging in a teleost species.

In fish, two isoforms of NUCB2 (NUCB2A and NUCB2B) exist, which presumably arose due to the teleosts-specific whole genome duplication known as 3R (third round of genome duplication) [50]. Also, in *N. furzeri* [34], two paralogues exist: NUCB2A and NUCB2B. Here, we determined that NUCB2A and NUCB2B gene sequences in killifish are highly similar to NUCB2 from other fish and mammals. This suggests that NUCB2 and its paralogues are highly conserved genes; therefore, Nesf-1 (a product of proteolytic processing of NUCB2 protein) has a similar aminoacid sequence among vertebrates. In fish, the proposed prohormone convertase cleavage site (Lys-Arg) is ubiquitously conserved in both NUCB2A and NUCB2B genes, suggesting that the putative Nesf-1 peptide can be cleaved from the larger NUCB2 precursor [14,15]. We analyzed the distribution of NUCB2B mRNA in the whole brain and the transcript was mainly detected in the hypothalamic nuclei, as previously reported in mammals [3,6,51–54], as well as in other vertebrate species such as frogs [55] and goldfish [14]. In young animals, neurons expressing NUCB2B mRNA were localized in the cortical nucleus, ventro-medial thalamic nucleus, paraventricular organ, and diffuse inferior lobe of the hypothalamus. We also studied Nesf-1 protein distribution by a commercial antibody, detecting the precursor Nesf-1 (1–82). Western blot analysis on the whole brain of *N. furzeri* reveals that the protein is expressed with the expected molecular weight of about 40 kDa, as also reported in goldfish [14] and mammals [51]. In *N. furzeri*, immunoreactivity to Nesf-1 was detected in the hypothalamic area, both in neuronal perikarya and fibers of young and old animals. Some areas, i.e., paraventricular organ and the diffuse inferior lobe of the hypothalamus, displayed positive neurons in ISH and IHC. Interestingly, mRNA expression and protein distribution were also detected in non-diencephalic areas, specifically in the telencephalon, optic tectum, and semicircular tori of both young and old animals. In mammals, Nesf-1 immunoreactivity was observed in non-diencephalic areas as well, including the nucleus of tractus solitarius, another brain region implicated in the regulation of feeding [6,51,52]. Our results in the brain of *N. furzeri* agree with previous observations in goldfish, where immunohistochemical studies showed the presence of Nesf-1-like ir within the hypothalamus and preoptic areas [14]. In teleost fish, both telencephalon and optic tectum are known to be involved in the control of appetite [56]. For example, electrical stimulation of either the ventral telencephalon, secondary gustatory nucleus, or optic tectum induces enhanced feeding behavior [57], whereas feeding behavior is depressed by olfactory tract lesions [57,58]. It might be possible that the presence of NUC2B/Nesf-1 in non-diencephalic areas of *N. furzeri* brain implicates that these areas could also be involved in the regulation of feeding in this species and/or several other homeostatic systems [59].

The primary source of NUCB2/Nesf-1 is currently unknown in vertebrates. In mammals, the expression level of NUCB2 in the stomach was found to be 10-fold higher than the levels of the brain [11], suggesting a prominent role for this organ in the synthesis and secretion of NUCB2/Nesf-1. As the most common teleost fish used as an animal model, such as zebrafish [60], medaka [48], or goldfish [14], *N. furzeri* also has no proper stomach, which is different from what has been reported in a previous paper [31]. According to Smith [61], a true stomach always has mucosal glands producing hydrochloric acid juice and can be closed by a sphincter at its caudal end. In addition, the loss of the stomach phenotype is accompanied by the loss of pepsinogen and gastric proton pump genes [48]. In agastric fish, the intestine transits directly from the esophagus and usually it can be divided into three sections (anterior, mid-intestine, and caudal intestine), where the dilatation of the first section of the intestine (named the rostral intestinal bulb or J-loop) is considered homologous to the mammalian stomach [48]. For the first time, we described the rostral intestinal bulb of *N. furzeri*. The wall of the intestine showed the superposition of four tunicae, which is typical of vertebrates. Genome analysis indicates the absence of pepsinogen gene in *N. furzeri*. In the rostral intestinal bulb, NUCB2B mRNA was detected in the epithelial cells lining while Nesf-1 immunoreactivity was mainly observed in the submucosa. The morphological pattern observed in the intestine of *N. furzeri* agree with previous observations in zebrafish, where NUCB2/nesfatin-1-like ir was detected in most of the cells scattered deep within the folds of the villi of J-loop [16], and disagree with a previous study of goldfish, where Nesf-1-like ir was only detected in enteroendocrine like cells of the intestinal villi [14]. It is likely that NUCB2 concurs to regulate the high turnover of epithelial intestinal cells. Future studies are mandatory to explore the peripheral role of NUCB2B/Nesf-1in *N. furzeri*, taking into consideration that currently the receptor(s) mediating the regulatory mechanisms of actions of Nesf-1 remain unknown [47].

Up to now, this is the first study reporting the regulation of NUCB2/Nesf-1 during aging in a vertebrate model. Interestingly, we document the increase of NUCB2B during aging, with the highest expression at the peripheral level. This could be related to an energy drive failure leading to anorexic phenotype, which indeed is well described and typically occurs during aging [62]. There are enough demonstrations available to indicate that nucleobindin and its encoded peptides have a pleiotropic role in cell biology, such as inhibition of proliferation and enhancing apoptosis [63]. In this context, it is relevant to specify that in the current genome annotation of *N. furzeri*, neither ghrelin nor leptin b were annotated, which are highly conserved during vertebrate evolution. With the exclusion of any annotation error, it is likely that that they are evolutionarily lost. This will be in line with the presence of an anorexic phenotype in old *N. furzeri* and it would elucidate a new role of nesfatin as an independent anorexigenic player in this short-living species.

These results, obtained in a phylogenetically distant vertebrate, highlight the growing evidence in support of a role for Nesf-1 as a novel brain-gut regulatory peptide. The communication between the central nervous system and the gastrointestinal tract plays a fundamental role in the regulation of food intake and energy balance, modulating short-term satiety and hunger responses. This axis also has a role in the regulation of blood glucose levels and adipocyte function [64]. The gut-brain axis has both neuronal and humoral components that convey information to the key brain regions involved in the homeostatic regulation of feeding, located mainly in the hypothalamus (including the arcuate nucleus) and brainstem (including the nucleus tractus solitarius) [65]. A number of gut hormones have been identified in the gastrointestinal system, including Nesf-1 [66], which are released from enteroendocrine cells. In addition to the coordination of the digestive process, they also relay information regarding the current state of energy balance, exerting endocrine effects on other organ systems, particularly the brain, where some have also been found to exist as neurotransmitters [67].

#### **5. Conclusions**

In conclusion, the present work provides the first evidence for the occurrence of NUCB2B/Nesf-1 in *N. furzeri* and to the best of our knowledge, it is the first description of aging regulation reported in any vertebrate. We consider that the experimental modulation could contribute to validate the

African turquoise killifish in clinical and preclinical studies. Indeed, in previous work [28], we used 96 h of fasting as a paradigm to better evaluate the regulation of neuropeptides involved in food intake. In that experimental design, 96 h of fasting represents a metabolic stimulus in this species, since we have demonstrated that it is able to activate neurons (pS6 marker), although in the brain of old animals, very few neurons were activated, suggesting that the metabolic stimulus needs to be more intense in old organisms with physiologically low metabolic rates. Based on these evidences, we are planning a similar experiment with a longer starvation period to evaluate the activated metabolic pathways at central and peripheral levels.

**Supplementary Materials:** The following are available online at http://www.mdpi.com/2077-0383/9/1/103/s1, Figure S1: NESF-1 positive cells in the submucosa of the rostral intestinal bulb of young and old animals. Cell count was carried out manually on 7 consecutive sections by using an open source image-processing program (ImageJ). Cells were identified on the basis of their morphological aspect. The graphical analysis was produced by Excel and did not reveal any significance difference between the two age points studied, Table S1: Ensemble Species Code, Table S2: Summary of primer pair sequences used for Real Time PCR (RT- PCR) and in situ hybridization (ISH).

**Author Contributions:** Conceptualization, A.M., M.B., and L.D.; Data curation, A.P.; Formal analysis, A.M., A.L., and L.D.; Funding acquisition, P.d.G.; Investigation, A.M., E.D.F., and A.L.; Project administration, M.B. and L.D.; Supervision, P.d.G. and M.B.; Validation, A.M. and E.D.F.; Visualization, E.D.F. and L.D.; Writing—original draft, E.D.F.; Writing—review & editing, C.L., P.S., and M.B. All authors have read and agreed to the published version of the manuscript.

**Funding:** This work was supported by a grant from the University of Naples Federico II (DR/2017/409-Project F.I.A.T.).

**Acknowledgments:** We are grateful to Alessandro Cellerino for their critical revision of the work. We thank Sabine Matz for their technical work and Antonio Calamo for their imaging assistance.

**Conflicts of Interest:** The authors declare no conflict of interest.
