Crucial role of zebrafish prox1 in hypothalamic catecholaminergic neurons development

Background Prox1, the vertebrate homolog of prospero in Drosophila melanogaster, is a divergent homeogene that regulates cell proliferation, fate determination and differentiation during vertebrate embryonic development. Results Here we report that, in zebrafish, prox1 is widely expressed in several districts of the Central Nervous System (CNS). Specifically, we evidenced prox1 expression in a group of neurons, already positive for otp1, located in the hypothalamus at the level of the posterior tuberculum (PT). Prox1 knock-down determines the severe loss of hypothalamic catecholaminergic (CA) neurons, identified by tyrosine hydroxylase (TH) expression, and the synergistic prox1/otp1 overexpression induces the appearance of hypothalamic supernumerary TH-positive neurons and ectopic TH-positive cells on the yolk epitelium. Conclusion Our findings indicate that prox1 activity is crucial for the proper development of the otp1-positive hypothalamic neuronal precursors to their terminal CA phenotype.


Background
The catecholaminergic neurons of the CNS of vertebrates participate in a wide variety of tasks, including motor coordination, mood regulation, and cognitive function, among others. Neurotransmitters catecholamines (CA), namely Dopamine (DA), Adrenaline (AD), and Noradrenaline (NA), are neuroactive molecules that exert strong influence on vertebrates behavior [1] and serve a variety of central and peripheral functions [2].
Embryological studies indicate that several extracellular signals, as Hedgehog and FGF, are vital to define the development of the prosencephalic CA neurons [3][4][5][6][7]. The homeodomain transcription factor Orthopedia (Otp), regulated by such signaling pathways [8], is crucial in restricting the fate of multiple classes of secreting neurons in the neuroendocrine hypothalamus of vertebrates [9,10]. Specifically, Otp is required for the correct differentiation of the CA neurons positioned in the zebrafish Posterior Tuberculum (PT) and hypothalamus [8,11]. Despite all these evidences, the role of specific transcription factors leading to the proper differentiation of the hypothalamic CA neurons remains largely unclear [7].
Prox1 homeobox gene is the vertebrate homologous of prospero in Drosophila melanogaster. During Drosophila embryonic development, prospero is expressed in neuronal precursors and determines the neuronal/glial fate of sibling cells [12,13]. prospero/Prox1's high level of homology pinpoints possible functional conservation through evolution, suggesting Prox1 involvement in vertebrate cell fate determination. Indeed, also during murine brain development, Prox1 is expressed in most of the locations in which neurogenesis and glial formation occur during middle and late prenatal and postnatal stages, as the subventricular zone, several regions of the prethalamus and hypothalamus, the cerebellum, and the hippocampus [14].
Here, we demonstrated that, in zebrafish, prox1 is widely expressed in the developing CNS, and one of its expression domains is located in the area corresponding to the ventral part of the PT and the adjacent hypothalamic district, the area hosting the cluster of CA neurons positive for otp1 expression [8,11]. Moreover, we took advantage of the zebrafish animal model to investigate the in vivo influence of prox1 on hypothalamic CA neuronal development by means of morpholino-and mRNA-loss and gain of function methodologies.
We provide evidence that prox1 is required for the development of hypothalamic neuronal progenitors into mature CA neurons.

Results and discussion
Homeobox genes are expressed in a temporal and spatial restricted manner and play crucial roles for cell type specification [15,16]. Zebrafish prox1 is a divergent homeodomain transcription factor whose homologues in Drosophila and mice regulate cell proliferation, fate determination and differentiation in CNS and sensory tissues [17][18][19][20][21]. Noteworthingly, during murine brain development, Prox1 is also expressed in the hypothalamus [14], where several CA neurons differentiate.

Spatio-temporal expression of prox1 during embryogenesis
Previously published immunostaining analysis of the Prox1 expression pattern revealed that the gene is active in several zebrafish embryonic districts [22]. To explore the role of prox1 during zebrafish CNS development, we first performed a more detailed characterization of prox1 expression during embryogenesis and in adult organs by means of RT-PCR (Fig. 1). We detected the presence of prox1 transcript at all stages analyzed, including the zygote, indicating that prox1 is also maternally expressed (Fig. 1A). Furthermore, we report prox1 expression in the adult brain, eyes, and in non-neuroectodermal territories (testis, ovary, gills, gut, liver) (Fig. 1B). The spatial and temporal distribution of prox1 transcripts was then examined by whole mount in situ hybridization (WISH), following standard protocols, with digoxigenin-and fluorescein-UTP-labeled probes [23]. At all stages analyzed, from 1-2 cell stage to 5 days post fertilization (dpf), zebrafish prox1 expression analysis confirms and improves previous immunostaining results [22]. From an evolutionary point of view, the striking resemblance of the prox1 expression pattern among vertebrates strongly suggests a conserved role for the gene during evolution. Although RT-PCR revealed the presence of maternal and zygotic transcripts, prox1 mRNA is first detectable through WISH around the 2 somites (s) stage in the ectodermic region corresponding to the otic placode ( Fig. 1C). At 15 s stage, prox1 is also expressed in the lens placode ( Fig.  1D) and in the first formed somites (Fig. 1D, inset). At 24 hours post fertilization (hpf), prox1 expression persists in the lens (Fig. 1E,F) and in the adaxial cells (see Additional file 1) that will later differentiate in slow muscle fibers [24]. Moreover, two distinct bilateral prox1 signals appear on each side of the midline, at the rostral end of the neural tube, in a region corresponding to the hypothalamus ( Fig.  1E). At the same stage, strong prox1 expression signals define the pituitary, the pretectal segment (prosomere 1, according to Rink and Wullimann [25]), and each hindbrain neuromeric segment (rhombomeres), where prox1 expression is visible in segmentally arranged clusters of cells (Fig. 1E). Additionally, prox1 signals are detectable in the liver and the posterior lateral line primordium (PLLP) (Fig. 1G). Starting from 48 hpf, further signals appear in the retina, pancreas (Fig. 1H), and in the cephalic ganglia (see Additional file 2). At 7 dpf, when the retinal's layers are fully differentiated, prox1 signal is detectable specifically in the inner nuclear layer, as previously shown in other vertebrates [20], and in the pretectal nuclei (Fig. 1I). In this work, we focused our attention on prox1 role during zebrafish hypothalamic development.

prox1 is required for the development of a group of hypothalamic CA neurons
The determination of the neurotransmitter phenotype is an important aspect of neuronal differentiation. Degeneration of substantia nigra DA neurons in humans is a hallmark of Parkinson's disease, and the malfunction of CA neurons in other brain regions is implicated in psychiatric disorders and neuroendocrine dysregulation [26][27][28].
In zebrafish, a detailed characterization of the CA neurotransmitter pathway makes this organism a favorite model to address the ontogeny of the vertebrate neurosecretory system [8,11,[29][30][31]. TH-expressing CA neurons are primarily located in the anterior dorsal telencephalon and hypothalamus of the developing forebrain, with a few additional neurons present near the postoptic commissure and pretectum region [29]. 36 hpf embryos hybridized with prox1 mRNA probe and immunostained with TH antibody show that prox1 transcript is present in close proximity to the most caudal posterior tubercular and the adjacent hypothalamic TH-expressing cells, with partial overlap of the two signals ( Fig. 2A,E,I). To determine whether prox1 is required for CA neuron development, we knocked-down the protein level by injecting 4 ng/embryo of a specific ATG-targeted morpholino oligonucleotide (prox1 MO) [32,33]. Abrogation of Prox1 function leads to a severe loss of neurons in the hypothalamic portion of the PT/hypothalamic CA cluster (at 36 hpf, 70% of the embryos showing no or few TH-positive cells in this area, n = 150) ( Fig. 2B,C,F,G). This defect is already evident at 24 hpf (see Additional file 3) and persists later during development, as shown by TH immunostaining at 48 hpf (see Additional file 3). The overall architecture of the ventral diencephalon was not affected in prox1 MO injected embryos, as suggested by the normal expression of shh (see Additional file 4). Therefore, we concluded that the decrease in the number of CA cells is not determined by an alteration in the patterning of the hypothalamus. We also analyzed the th expression levels in prox1 MO injected embryos by means of quantitative real time RT-PCR. The th specific mRNA level was about five-fold decreased in prox1 MO injected embryos when compared to the th expression in control embryos injected with the standard control morpholino oligonucleotide (stdr MO) (Fig. 2L), confirming the immunohistochemical analysis reported above. To demonstrate that the reduction of TH positive cells in the hypothalamus of the embryos is specifically caused by the MO-induced abrogation of Prox1 function, prox1 temporal and spatial expression pattern analyzed by RT-PCR and in situ hybridization we performed a rescue experiment by coinjecting the embryos with 4 ng/embryo of prox1 MO and 400 pg/ embryo of prox1 mRNA (Fig. 2D,H). 80% of the embryos at 36 hpf (n = 120) rescued the normal phenotype and displayed a proper number of hypothalamic TH-positive neurons (Fig. 2D,H). On the other hand, overexpression of prox1 alone does not lead to supernumerary CA neurons in the ventral diencephalon, nor determines ectopic CA neuron formation, confirming that prox1 functions are required for the proper development of the TH phenotype in a subpopulation of hypothalamic neurons, but are not sufficient to determine the appearance of supernumerary or ectopic TH-expressing cells. Interestingly, prox1 overexpression induces a slight increase of th mRNA levels, as detectable by means of quantitative real time RT-PCR only (data not shown); further investigations are necessary to elucidate this aspect that might reflect the ability of prox1 to directly modulate th expression in those few cells where the two genes are coexpressed.
In order to address whether prox1 is involved in neurogenesis processes, we analyzed the expression pattern of the proneural gene ngn1 [34]. ngn1 expression domains resulted unaffected in prox1 MO injected embryos, allowing us to conclude that loss of CA neurons in the hypothalamus of prox1 MO injected embryos is not caused by alteration in neurogenesis (see Additional file 5). Moreover, in order to address this issue, we analyzed the development of other neurotransmitter-producing neurons. Specifically, the neighboring serotonergic neurons prox1 is required for the development of CA neurons in the hypothalamus  only slightly affected by Prox1 ablation in comparison to the most relevant effects we observed in the CA population (see Additional file 6). However, our description of prox1 effects on CA neuron development cannot rule out its potential involvement in the differentiation or fate determination of other neuronal types.

prox1 functions are required for proper otp1 and TH phenotypes in the hypothalamic area
To further investigate the roles of prox1 in the hypothalamic CA neurons development, we showed that prox1 is also able to influence the otp1 (otpb) phenotype in hypothalamic neurons. As previously reported by our group [8], zebrafish otp1 contributes to the specification and differentiation of DA diencephalic neurons in the PT and the hypothalamus. In search of possible relationships between prox1 and otp1 in determining the TH phenotype, we performed a double WISH that evidenced the coexpression of the two genes in the hypothalamus (Fig.  3A,B). We also demonstrated that some of these prox1/ otp1-positive cells are also positive for TH (Fig. 3B,B'), supporting the hypothesis that the coexpression of the two genes influences the development of the final TH phenotype. Injection of the embryos with otp1 MO or otp1 mRNA does not result in significant changes in prox1 expression pattern at 24 and 48 hpf (data not shown). On the other hand, a MO-mediated reduction of Prox1 levels determines significant modifications in otp1 expression pattern in the hypothalamic area, while otp1 expression in the rhombomeres resulted unperturbed. Specifically, the injection of prox1 MO causes the decrease of otp1-positive neurons in the hypothalamus at 48 hpf in comparison with embryos injected with the stdr MO (Fig. 3C,D). On the contrary, prox1 overexpression does not modify otp1 hypothalamic domains, nor determines otp1 ectopic expression (data not shown). Our findings indicate that prox1 is necessary, but not sufficient, to the proper otp1 phenotype in the hypothalamic area. Interestingly, otp1 appearance in the hypothalamus [8] precedes the onset of prox1 expression in this area, suggesting that prox1 is not involved in the activation of otp1 transcription. Rather, prox1 activity might be vital to control the switch of the otp1-positive cells towards their final TH fate. The hypothesis is supported by the evidence that prox1 MO and otp1 synthetic mRNA coinjection did not restore the normal TH phenotype (see Additional file 7). Thus, lack of Prox1 might impede further neuronal differentiation, causing apoptosis or cell misspecification, with consequent loss of prox1 functions are required for proper otp1 and TH phenotypes in the hypothalamic area otp1 and th expression. In order to discriminate between these two possibilities, we performed a TUNEL assay on prox1 MO injected embryos at 24 hpf (see Additional file 8) and at 36 hpf (data not shown). The level of apoptosis is not increased by Prox1 knock-down, suggesting that the lower number of otp1-positive hypothalamic neurons might be determined by misspecification events rather than apoptosis. Interestingly, otp1 is expressed in those cells already switched towards a more differentiated state, such as early postmitotic DA precursors, as well as newly specified and mature DA cells [8,11]. Thus, according to the wealth of literature data pinpointing prox1 as a key player in the passage from proliferation to differentiation [20,21,35], prox1 in the hypothalamus might drive the cells towards differentiative processes, leading to the terminal TH phenotype of those precursors committed to a CA phenotype by the expression of otp1.

Overexpression of prox1 and otp1 together leads to supernumerary CA neurons in the ventral diencephalon and TH positive cells on the yolk surface
To verify whether prox1 and otp1 together have an impact on the CA phenotype in the hypothalamus, we coinjected their specific mRNAs at a concentration of 450 and 300 pg/embryo, respectively, and stained the embryos for TH at 36 hpf. As reported above, the injection of prox1 mRNA alone did not increase the number of hypothalamic CA neurons, nor did single 300 pg/embryo otp1 mRNA injection (data not shown). Comparable results were obtained injecting 450 pg/embryo of the GFP mRNA as control (Fig. 4). The most numerous class in the group of control embryos presented 8 CA hypothalamic neurons, and only 3 embryos presented more than 11 TH hypothalamic positive cells. On the other hand, the most numerous class in the group of the overexpressed prox1/otp1 embryos (n = 64) presented 10 CA neurons, and 20 embryos showed more than 11 TH hypothalamic positive cells (Fig. 4A,B), pointing out the synergistic effect of the prox1/otp1 coexpression on the CA phenotype establishment. Remarkably, we also evidenced that prox1/otp1 coinjection induces TH positive cells on the yolk surface (Fig. 4D,E), while these ectopic cells were never observed in control (GFP) or single (prox1 or otp1) injected embryos (Fig. 4C). Noteworthy, in the CNS, the overproduction of TH positive cells induced by coinjection was detected in the hypothalamus, suggesting that prox1 and otp1 genes require additional factors, present in the ventral diencephalon, to induce the TH phenotype. Thus, we demonstrated that the coexpression of prox1 and otp1 determines a higher Synergistic prox1/otp1 overexpression induces the appearance of hypothalamic supernumerary TH-positive neurons and ectopic TH-positive cells on the yolk surface ectoderm

Conclusion
In conclusion, we highlight for the first time the role of prox1 in the proper development of the CA neurons in the ventral diencephalon. Moreover, we provide evidence of regulatory links between prox1 and otp1 genes in defining the terminal TH phenotype in the hypothalamus. The identification of prox1 as a key component in the differentiation of hypothalamic CA neurons will help in clarifying the developmental bases of several human behavioral aspects as well as pathologies such as addictions and Parkinson's disease.

Animals
Breeding wild type fish of the AB strain were maintained at 28°C on a 14 h light/10 h dark cycle. Embryos were collected by natural spawning, staged according to Kimmel and colleagues [36], and raised at 28°C in fish water (Instant Ocean, 0,1% Methylene Blue) in Petri dishes. We express the embryonic ages in somites (s), hours post fertilization (hpf) and days post fertilization (dpf).

In situ hybridization and immunohistochemistry
Whole mount in situ hybridization (WISH), was carried out as described [23] on embryos fixed for 2 h in 4% paraformaldehyde/phosphate buffered saline, then rinsed with PBS-Tween, dehydrated in 100% methanol and stored at -20°C until processed for WISH [37]. Antisense riboprobes were previously in vitro labelled with modified nucleotides (i.e. digoxigenin, fluorescin, Roche). For histological sections, stained embryos were re-fixed in 4% PFA, dehydrated and stored in methanol, wax embedded and sectioned (5 μm). For immunohistochemistry, embryos were exposed to rabbit anti-Tyrosine Hydroxilase (anti-TH) (Chemicon), or rabbit anti-Serotonin (anti-5HT) (Chemicon), then treated with biotinylated or fluorescent secondary antibody (Vector Laboratories).

Quantitative real time RT-PCR
Reverse transcriptions (RTs) were performed using 2 μg of DNase treated (DNA-free™, Ambion Inc) total RNA in presence of random hexamers (Invitrogen™) and Super-Script II reverse transcriptase (Invitrogen™). Real-time PCRs were carried out in a total volume of 15 μl containing 1× iQ SYBR Green Super Mix (BioRad), using 1 μl of the RT reaction. PCRs were performed using the BioRad iCycler iQ Real Time Detection System (BioRad Laboratories). For normalization purposes, 18S ribosomal RNA level was tested in parallel with the gene of interest. The following primers were used:

TUNEL staining
For TUNEL assay, 24 and 36 hpf embryos were fixed with 4% PFA for 2 h at room temperature. Embryos were permeabilized with methanol at -20°C and washed twice with PBC (0.001% Triton ×-100, 0.1% sodium citrate in PBS) for 10 minutes. Labeling for apoptotic cells was performed using In situ Cell Death Detection Kit (Roche). The embryos were incubated at 37°C for 1 h, washed and mounted for fluorescent microscopic imaging.