Neuroendocrine transcriptional programs adapt dynamically to the supply and demand for neuropeptides as revealed in NSF mutant zebrafish
- 10k Downloads
Regulated secretion of specialized neuropeptides in the vertebrate neuroendocrine system is critical for ensuring physiological homeostasis. Expression of these cell-specific peptide markers in the differentiating hypothalamus commences prior to birth, often predating the physiological demand for secreted neuropeptides. The conserved function and spatial expression of hypothalamic peptides in vertebrates prompted us to search for critical neuroendocrine genes in newly hatched zebrafish larvae.
We screened mutant 5 days post-fertilization zebrafish larvae that fail to undergo visually mediated background adaptation for disruption in hypothalamic pomc expression. To our surprise, the ATPase N-ethylmaleimide sensitive factor (nsf) was identified as an essential gene for maintenance of neuroendocrine transcriptional programs during the embryo-to-larva transition. Despite normal hypothalamic development in nsfst 53mutants, neuropeptidergic cells exhibited a dramatic loss of cell-specific markers by 5 days post-fertilization that is accompanied by elevated intracellular neuropeptide protein. Consistent with the role of NSF in vesicle-membrane fusion events and intracellular trafficking, cytoplasmic endoplasmic reticulum-like membranes accumulate in nsf-/- hypothalamic neurons similar to that observed for SEC18 (nsf ortholog) yeast mutants. Our data support a model in which unspent neuropeptide cargo feedbacks to extinguish transcription in neuropeptidergic cells just as they become functionally required. In support of this model we found that gnrh3 transcripts remained unchanged in pre-migratory, non-functional gonadotropin-releasing hormone (GnRH) neurons in nsf-/- zebrafish. Furthermore, oxytocin-like (oxtl, intp) transcripts, which are found in osmoreceptive neurons and persist in mutant zebrafish, drop precipitously after mutant zebrafish are acutely challenged with high salt.
Our analyses of nsf mutant zebrafish reveal an unexpected role for NSF in hypothalamic development, with mutant 5 days post-fertilization larvae exhibiting a stage-dependent loss of neuroendocrine transcripts and a corresponding accumulation of neuropeptides in the soma. Based on our collective findings, we speculate that neuroendocrine transcriptional programs adapt dynamically to both the supply and demand for neuropeptides to ensure adequate homeostatic responses.
KeywordsBotulinum Toxin Transcriptional Program Hypothalamic Neuron Zebrafish Larva Olfactory Placode
in situ hybridization
N-ethylmaleimide sensitive factor
synaptic vesicle protein 2
Terminal deoxynucleotidyltransferase-mediated dUTP nick end labeling
visually mediated background adaptation.
The hypothalamus participates in the maintenance of homeostasis through the synthesis and release of neuropeptides. The neuroendocrine system appears to be highly conserved between mammals and teleosts, as evidenced by shared programs found for pituitary development [1, 2, 3, 4] and the spatial patterns and functional roles of hypothalamic neuropeptides [5, 6, 7, 8, 9, 10, 11]. In teleosts, much of the neuroendocrine system becomes functional between 48 and 72 hours post-fertilization (hpf), when the embryo hatches from its chorion and becomes fully exposed to an external environment . By 5 days post-fertilization (dpf) this newly hatched embryo has developed into a free-swimming larva that must respond and adapt to external and internal cues. Thus, this embryo-to-larva transition (3 to 5 dpf) is a distinct developmental stage, somewhat analogous to mammalian birth when the newborn integrates sensory input and initiates regulated-secretion to maintain homeostasis.
While the demand for neuroendocrine signaling and regulated neuropeptide secretion increases dramatically when a newborn is exposed to an external environment, cell-specific transcriptional programs appear to initiate much earlier [13, 14, 15, 16]. Indeed, birthdating experiments in rodents show that hypothalamic neurons originate from the third ventricular neuroepithelium beginning at embryonic day (E)12 and are largely differentiated by E16 [17, 18], as judged by expression and translation of cell type-specific neuropeptide transcripts and protein, respectively. That the onset of hypothalamic transcriptional programs predates function is perhaps best illustrated by the ontogeny of gonadotropin-releasing hormone (GnRH) neurons (also known as luteinizing hormone-releasing hormone (LHRH) neurons). In both mammals and zebrafish this subpopulation of neurons migrates from the olfactory placode into the hypothalamus and eventually orchestrates germ cell maturation at the onset of sexual maturity [6, 19, 20, 21, 22, 23]. However, during embryonic development and prior to their migration, immature GnRH neurons express their signature neuropeptide [6, 21, 22]. These and other studies suggest that the neuroendocrine system develops early in the fetus and remains primed to secrete neuropeptides at later time points [24, 25].
To identify genes that might be critical for hypothalamic development during this embryo-to-larva transition, we re-screened mutant 5 dpf zebrafish larvae that fail to undergo visually mediated background adaptation (VBA) for disruption in neuropeptide transcription. VBA is a well characterized neuroendocrine reflex that occurs in fish, reptiles, and amphibians (reviewed in ). When placed on a dark, light-absorbing background, pituitary melanotropes release α-melanocyte-stimulating hormone; this neuropeptide mediates dispersion of black pigment granules in dermal melanophores and causes a darkening of skin color in fish. Conversely, when placed on a light background, a drop in α-melanocyte-stimulating hormone levels results in the aggregation of melanin and a loss of this 'dark' phenotype or an overall blanching. Because α-melanocyte-stimulating hormone is directly regulated by hypothalamic inputs, we reasoned that critical developmental neuroendocrine genes might be discovered in VBA-mutant zebrafish  with altered proopiomelanocortin (pomc) expression.
From our screen, one mutant line showed a complete and specific loss of pomc expression in the presumptive arcuate nucleus of the hypothalamus. To our surprise, this mutated gene was identified as N-ethylmaleimide sensitive factor (nsf, SEC18), which is an ATPase and functions in membrane fusion events as well as acting as a structural chaperone (reviewed in [28, 29]). These functional roles of NSF are important for regulated secretion [30, 31, 32] and for intracellular trafficking . Our analysis of nsf mutant zebrafish led us to examine how hypothalamic developmental programs respond to a major secretory defect. We report that nsf-deficient hypothalamic neurons exhibit a marked accumulation of neuropeptide proteins with a corresponding stage-dependent loss of mRNA, suggesting that neuropeptide stores in the cytoplasm feedback to regulate transcriptional programs in the nucleus.
Loss of hypothalamic pomc expression in nsf mutant zebrafish
Multiple complementation crosses revealed that s364 represents a novel loss-of-function allele of nsf. Similar to s364, both st53 and the s364/st53 and s364/st25 transheterozygotes showed disrupted pomc staining in the arcuate nucleus of the hypothalamus in mutant zebrafish, whereas expression was maintained in the anterior and posterior pituitary (Figure 1B). Low, but persistent pomc expression was observed in the arcuate nucleus of the missense nsf allele st25 and in the s364/st25 transheterozygotes (Figure 1B). Sequencing of the exons and the exon-intron junctions of s364 revealed no obvious mutations, suggesting that this nsf allele contains a non-coding mutation (data not shown). Because the genetic lesion generating the s364 allele remains unknown, we used nsfst 53mutant zebrafish for all subsequent analyses so that mutant embryos could be identified by genotyping, prior to the appearance of the 'dark' phenotype.
Disruption of nsf is partially compensated by upregulation of nsfb
The zebrafish genome contains duplicates of an ancestral nsf gene, annotated as nsf and nsfb . nsf is ubiquitously expressed in the central nervous system throughout development  and in early larval stages (Additional file 1 Figure S3A). As might be predicted by the nonsense mutation in nsfst 53, nsf transcript levels are nearly absent in nsfst 53zebrafish (Figure 1C). However, quantitative PCR revealed that nsfb is upregulated nearly two-fold in the heads of nsf mutant zebrafish (Figure 1C). Thus, compensation by nsfb may account for the survival of nsfst 53zebrafish to 8 dpf given that combined nsf and nsfb morpholino injections result in prominent and early lethality (Additional file 1 Table S1). Interestingly, nsfst 53zebrafish were noted to be longer than wild-type siblings (3.8 mm versus 3.6 mm; Figure 1D), and while the molecular basis of this phenotype is unclear, others have found that disruption of POMC-melanocortin signaling increases linear growth in both zebrafish and mice [35, 36].
Hypothalamic cell markers are absent in nsfst 53mutant larvae
Loss of hypothalamic markers is not due to apoptosis
Hypothalamic neuropeptides accumulate in nsfst 53mutants
Transiently blocking secretion in wild-type zebrafish mimics the nsfst 53phenotype
A physiological challenge silences persistent oxtl transcripts in nsfst 53mutants
We hypothesized that the selective loss of neuropeptide transcripts in nsfst 53zebrafish might reflect activation of these particular neurosecretory cells during the embryo-to-larval transition. This 'use-dependent' transcriptional hypothesis predicts that gnrh3 and oxtl neurons are functionally silent at 3 to 5 dpf. If true, then activation of either gnrh3 or oxtl neurons during this developmental stage should diminish their respective neuropeptide transcripts. Although activating gnrh3 neurons is technically difficult, we reasoned that activating oxtl cells by a high-salt challenge might be feasible given that these neurons are osmoreceptive, and together with arginine vasopressin-like (avpl; formerly known as vsnp) neurons are required to maintain electrolyte homeostasis [38, 39]. Indeed, we found oxtl expression significantly diminished after a hyperosmolar acute challenge in nsfst 53zebrafish, but not in wild-type zebrafish (Figure 7A). This result was confirmed by quantitative PCR showing a nearly 50% reduction of oxtl transcripts after a 24 h salt challenge (Figure 7B). We also observed TUNEL-positive cells around the gills in salt-challenged nsfst 53but not wild-type zebrafish (Figure 7A). Based on these observations, we suggest that the demand for oxytocin-like neuropeptide during a high salt challenge overloads the defective secretory pathway in nsfst 53mutants, resulting in downregulation of oxtl transcripts.
Our results show that loss of hypothalamic cell markers occurs in all nsf alleles examined, well after neuroendocrine neurons initiate expression of their signature markers. Indeed, expression patterns of all assayed neuropeptides and transcription factors were indistinguishable in wild-type and nsfst 53zebrafish, beginning as early as 24 hpf; a similar phenotype was also noted in the nsfs364 allele (data not shown). This is surprising considering the general role of NSF in membrane fusion and its ubiquitous expression throughout development, and given that Drosophila dNSF1 mutants exhibit impaired neurogenesis . It is possible that normal hypothalamic development in nsfst 53may result from compensation by nsfb, maternally deposited nsf transcripts or residual activity of NSF encoded by the nsfst 53allele. Clearly, at least some functional NSF is required for survival as evidenced by the fact that co-injection of nsf-MO and nsfb-MO into nsfst 53embryos was lethal (Additional file 1 Table S1). The persistence of gnrh3 in immature, pre-migratory nsfst 53neurons, coupled with the ability to dampen oxtl transcripts following an acute salt challenge in wild-type zebrafish, implies that sub-populations of neuroendocrine cells escape the consequences of disrupted NSF function. Thus, we speculate that increased physiological demand and secretion of neuropeptide cargo during the embryo-to-larva transition accounts for this late-stage loss of neuroendocrine transcripts in nsf mutants.
The neuroendocrine phenotype exhibited by the nsfst 53zebrafish is partially mimicked in munc18-1 mutant mice, which exhibit a selective and stage-dependent (E18) decrease primarily in neuropeptide transcript expression . Munc18-1 is essential for regulated exocytosis. The authors of this study posited that transcripts encoding 'consumables' such as secreted products are selectively downregulated in the munc18-1 synaptically silenced neurons. Although neuropeptide protein levels and localization were not assessed in their study, our findings predict that neuropeptides would accumulate in munc18-1 hypothalami. We can now extend their hypothesis by suggesting that unspent neuropeptide consumables feed back to silence neuroendocrine transcriptional programs, as shown in our model figure (Figure 8).
Our study is now the second to identify nsf as essential in a specific neuronal cell-type program using a forward genetic screen in zebrafish. Previously, Talbot and colleagues  unexpectedly found nsf to be essential for organization of myelinated axons, with mutant zebrafish exhibiting decreased myelin basic protein expression and altered sodium channel clustering. Here, we find that this ubiquitous protein is also essential for maintenance of neuroendocrine transcriptional programs. These distinct phenotypes illustrate how different neuronal populations adjust to deficits in membrane fusion and intracellular trafficking. We suggest that hypothalamic neurons mount a stage-dependent adaptive response to this major secretory defect by silencing transcriptional programs. For other neuronal cell types, loss of nsf results in vastly different cellular responses. Indeed, we observed prominent cell death in many sensory organs, including photoreceptors, olfactory sensory neurons, and mechanosensory cells in nsf mutants. By contrast, no measurable cell death was observed in nsf mutant hypothalami as evidenced by the normal gross organization of the nsfst 53hypothalamus, normal HuC/D staining, which marks early neuronal differentiation, and normal staining with the SV2 pan-neuronal vesicle marker. Targeted knockdown of nsf or nsfb in the mutant background also ruled out that survival of hypothalamic cells in nsfst 53was due to compensation by the duplicate gene nsfb and/or maternally deposited nsf mRNA. Taken together, hypothalamic neurons, unlike other neuronal populations, are able to survive and maintain normal cellular integrity in the face of this major secretory deficit.
The phenotypes observed in nsfst 53hypothalamic neurons, including the complete loss of neuroendocrine transcripts and the accumulation of ER-like membranes, are reminiscent of, but different from, other well-described secretory feedback signaling pathways. Regulation of nuclear targets by ER-associated events are well established for unfolded protein responsive (UPR) genes  and for sterol regulatory genes (reviewed in ). Here, we have yet to identify the precise molecular nature of the feedback signal; whether these signals arise from the buildup of ER-like membranes, from accumulation of neuropeptides in the cytoplasm, or from the inability to secrete neuropeptides at the plasma membrane remains to be determined. The build up of ER-like membrane in the cytoplasm of nsf hypothalamic neurons suggests that neuropeptidergic cells are responsive to a loss of secretion and not a general trafficking defect. Indeed, we can mimic this phenotype with a transient block on secretion in wild-type animals with botox injections. Thus, it is possible that ER-to-golgi transport in nsfst 53neurons is partially intact, but becomes overloaded when functional demands are placed on these neurosecretory neurons in the embryo-to-lava transition.
This cytoplasmic-to-nuclear signaling hypothesis assumes that phenotypes observed in nsfst 53hypothalami are cell-autonomous, as documented previously . In our case blastomere transplantation studies into the hypothalamus are technically challenging because transplanted clone sizes in this brain region are small, and because selection of the correct in situ hybridization (ISH) probe, each recognizing a very small number of cells, is difficult. Despite these inherent limitations, preliminary transplantation data revealed expression of neuroendocrine markers in the medial hypothalamus of wild-type → nsfst 53chimeras (data not shown), supporting the notion that similar to other neurons, NSF functions autonomously in specialized neuroendocrine neurons.
The unexpected role of nsf in maintenance of hypothalamic cell type markers at late developmental stages implies that hypothalamic neurons adapt to and survive a major trafficking defect by modifying their transcriptional programs. We speculate that unlike sensory neurons, which undergo apoptosis, hypothalamic neurons cope with this major cellular stress because these ancient and evolutionarily conserved neurons are inherently programmed to defend and maintain homeostatic physiological responses when confronted with changing environments.
Materials and methods
All research involving zebrafish was approved and carried out according to guidelines of the UCSF IACUC committee. Zebrafish were maintained at 28°C under 14 h light, 10 h dark as described previously . Nsfst 25and nsfst 53zebrafish were kind gifts from W Talbot (Stanford University). Adult nsf+/- zebrafish were fed a mixture of fish flakes and brine shrimp twice per day. The 5 dpf fish were maintained in egg water, consisting of 0.03% Instant Ocean→∀ sea salts (Spectrum Brands, Atlanta, GA, USA) in distilled water and not supplemented with any food. Embryos were obtained by natural crosses of identified parents and zebrafish embryo and larvae stages were determined according to . The VBA phenotype was assayed by placing 5 dpf zebrafish on a white background for >15 minutes. Fish that remained dark in color and failed to have inflated swim bladders were selected as mutant zebrafish. Zebrafish <5 dpf were genotyped as previously reported in .
Mutant and wild-type zebrafish (n = 10 each) were sorted for their dark phenotype at 5 dpf and then placed on ice to paralyze them but were kept alive. They were then transferred to a 3% agarose mold and their images were captured under an Olympus light microscope equipped with a CCD camera (Center Valley, PA, USA). The zebrafish in each image was measured from nose to tip of the tail to the nearest 0.1 cm using a standard ruler and the scale was adjusted appropriately using the scale bar present within the captured image. Statistical analyses were conducted by unpaired t-test.
Whole mount in situ hybridization
Mutant and wild-type embryos/larvae were fixed in 4% paraformaldehyde (PFA) in phosphate-buffered saline (PBS) overnight at 4°C and transferred to 100% MeOH and stored at -20°C until use. Embryos were rehydrated into PBS-Tween (PBS-Tw; 0.1% Tween-20), bleached (72 hpf to 5 dpf; 1% H2O2 + 5% formamide in PBS-Tw; 1 h), treated with proteinase K (20 μg/ml; 8 minutes), refixed (4% PFA; 20 minutes), prehybridized (65 to 70°C; 1 h) and hybridized (65 to 70°C; overnight) with hypothalamic riboprobes. Embryos were washed (50% formamide/50% 2× SSC; 2× SSC; 0.2× SSC), blocked (5% sheep serum heat inactive) and incubated overnight at 4°C with α-digoxigenin (DIG) antibody followed by washing and a second incubation with p-nitroblue tetrazolium chloride/5-bromo-4-chloro-3-indolyl phosphate (NBT/BCIP) in staining buffer (100 mM NaCl, 50 mM MgCl2, 100 mM Tris-HCl, 0.1% Tween 20). The color reaction was monitored for several hours and stopped by washing with PBS-Tw. The embryos were re-fixed and transferred to 90% glycerol for clearing. Embryos were captured using an Axiocam camera (Carl Zeiss, Thornwood, NY, USA). TUNEL-based cell death assays were carried out using this procedure with minor revisions according to manufacturer's protocol (ApopTag; Millipore Corporation, Billerica, MA, USA). DIG-cRNA probes were synthesized (Roche, Indianapolis, IN, USA) from the following generously gifted plasmids: pomc, Roger Cone (Vanderbilt University Medical Center); hcrt, Alexander Schier (Harvard University); nsf, William Talbot (Stanford); oxtl and avpl, Eric Glasgow (Georgetown University); npy, Peter Schoonheim (UCSF); emx1, Su Guo (UCSF); crh, Limor Ziv (UCSF); ff1d, Miyuki Suzawa (UCSF); nkx2.1a, Steve Wilson (University College London); gnrh3, Yonathan Zohar (Cambridge University).
Immunohistochemistry on sectioned zebrafish
Mutant and wild-type larvae were fixed in 4% PFA/PBS overnight at 4°C and cryoprotected in 30% sucrose/PBS overnight at 4°C. For horizontal sections, zebrafish were embedded in OCT (Tissue-Tek, Sakura Finetek, Tokyo, Japan) and sectioned at 12 μm (Leica Microsystems, Bannockburn, IL, USA). Sectioned zebrafish were left to dry (overnight at 25°C), then washed in PBS-Tw, incubated in block (10% Normal sheep serum/PBS; 1 h at 25°C), and left to incubate overnight with α-HuC/D (1:400; Invitrogen, Carlsbad, CA, USA), α-SV2 (1:50; Developmental Studies Hybridoma Bank, University of Iowa), α-hypocretin (1:500; Dr Kohgo, Asahikawa Medical College, Japan), α-GnRH (1:1,500; Gunma University, Japan), α-Tyrosine Hydroxylase (1:100; Chemicon, Temecula, CA, USA) and α-NPY (1:1,000; Dr Jo Harrold, University of Liverpool). After washing in PBS-Tw, sections were incubated with Alexa-488 (1:200; Invitrogen) for 4 h, 25°C. Prior to mounting, sections were stained using Hoechst 33258 (Invitrogen). Confocal images were captured using a Zeiss LSM 5 Pascal microscope and software. Confocal stacks were further processed using ImageJ software. Z-projections of a few slices were made and comparisons were made between images that were processed equivalently. Fluorescence images were adjusted in Adobe Photoshop using the brightness/contrast, levels, and curves functions in order to best represent the protein levels.
Real-time quantitative PCR
Real-time quantitative PCR was performed as described previously . The head was dissected from the tail just prior to the swim bladder of 5 dpf mutant and wild-type zebrafish and the RNA was isolated from both head and tail using TrizolT. All data are normalized to one wild-type 'head' sample (taken to be 1) and the error bar represents the variation of the other 'head' samples (n = 4). Prior to routine use, all primer sets (Primer Express v2.0, Applied Biosystems, Foster City, CA, USA) were validated to ensure amplification of a single product with appropriate efficiency. Data obtained from the PCR reaction were analyzed using the comparative CT method (User Bulletin No. 2, PerkinElmer Life Sciences, Waltham, MA USA). Sequences are provided in Additional file 1 (Table S2).
For high salt, osmolar challenge, 4-dpf larvae were transferred to egg water composed of either 0.3% Instant Ocean→∀ (challenge) or 0.03% Instant Ocean→∀ (control) and maintained at 28°C. Larvae were collected 24 h later and fixed immediately using cold 4% PFA.
MO analyses were conducted as published previously . Briefly, MOs were designed to bind sequences surrounding the initiating methionine (Gene Tools, Philomath, OR, USA). MOs were resuspended in water to a working solution of 1 μM and injected into the yolk of one- to four-cell stage embryos (n = 100 embryos/MO; repeated 3 times). Effective doses were determined separately for each MO. Larvae (5 dpf) were collected, fixed overnight and stored in 100% MeOH at 20°C until whole mount ISH was performed. Sequences are provided in Table S2 (Additional file 1).
Tissue was fixed in 2% glutaraldehyde, 1% paraformaldehyde in 0.1 M sodium cacodylate buffer, pH 7.4, postfixed in 2% osmium tetroxide in the same buffer, and block stained in 2% aqueous uranyl acetate, dehydrated, infiltrated, and embedded in LX-112 resin (Ladd Research Industries, Burlington, VT, USA). Toluidine blue stained semi-thin sections were made to locate the area of interest, using the eyes as a guideline. Samples were ultrathin sectioned on a Reichert-Jung (Leica) Ultracut S ultramicrotome (Leica Microsystems, Bannockburn, IL USA), and stained with 0.8% lead citrate. Grids were examined on a JEOL JEM-1230 transmission electron microscope (JEOL USA, Inc., Peabody, MA, USA) and photographed using the Gatan Ultrascan 1000 digital camera (Gatan Inc., Warrendale, PA, USA).
Botulinum toxin analyses
Botulinum toxin injections and analyses were conducted as published previously . Briefly, progeny from in-crosses of heterozygous nsf53 adults were injected at the one-cell stage with an approximately 5 nl bolus of 0.1 to 1.0 ng/nl BtTxB (EMDBiosciences, Darmstadt, Germany). Larvae (5 dpf) that lacked a startle response were placed either into fix (4% PFA, overnight, 16°C) or Trizol™ (-80°C) until processed for ISH and quantitative PCR, respectively.
We especially want to thank Dr W Talbot for generously sharing the st25 and st53 nsf alleles and helpful comments. The authors also thank Dr J Kay for assistance with the screen, Dr N Gosse for help with transplant studies, and Drs M Suzawa, P Schoonheim, and L Ziv for sharing in situ probes. We thank Gunma University, Institute for Molecular and Cellular Regulation Biosignal Research Center for kindly sharing LHRH (GnRH) antibody. We thank Drs C Cheung and F Lee for careful reading of the manuscript. Support for this study includes a Giannini Foundation Fellowship to DMK, and NIH-RO1 EY013855 to HB, and NIH-RO1 DK063592 to HAI.
- 24.Swanson LW: The Hypothalamus. 1987, Amsterdam, New York, Oxford: Elsevier Science Publishers BVGoogle Scholar
- 25.Grumbach M, Gluckman P: The human fetal hypothalamus and pituitary gland: the maturation of neuroendocrine mechanisms controlling secretion of fetal pituitary growth hormone, prolactin, gonadotropins, adrenocorticotropin-related peptides, and thyrotropin. Maternal-fetal Endocrinology. Edited by: Tulchinsky D, Little AB. 1994, Philadelphia: WB Saunders, 193-261. 2Google Scholar
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.