The zebrafish as a model system for forebrain GnRH neuronal development

General and Comparative Endocrinology 164 (2009) 151–160

Contents lists available at ScienceDirect

General and Comparative Endocrinology journal homepage: www.elsevier.com/locate/ygcen

Review

The zebrafish as a model system for forebrain GnRH neuronal development Eytan Abraham a, Ori Palevitch b, Yoav Gothilf b, Yonathan Zohar a,* a b

Center of Marine Biotechnology, University of Maryland Biotechnology Institute, Baltimore, MD, USA Faculty of Life Sciences, Tel-Aviv University, Tel-Aviv, Israel

a r t i c l e

i n f o

Article history: Received 8 October 2008 Revised 31 December 2008 Accepted 20 January 2009 Available online 31 January 2009 Keywords: Zebrafish GnRH Development Transgenic CNS

a b s t r a c t Development and function of the forebrain gonadotropin-releasing hormone (GnRH) neuronal system has long been the focus of study in various vertebrate species. This system is crucial for reproduction and an important model for studying tangential neuronal migration. In addition, the finding that multiple forms of GnRH exist in the CNS as well as in non-CNS tissues, coupled with the fact that GnRH fibers project to many CNS regions, implies that GnRH has a variety of functions in addition to its classic reproductive role. The study of the GnRH system and its functions is, however, limited by available model systems and methodologies. The transgenic (Tg) GnRH3:EGFP zebrafish line, in which GnRH3 neurons express EGFP, allows in vivo study of the GnRH3 system in the context of the entire animal. Coupling the use of this line with the attributes and molecular tools available in zebrafish has expanded our ability to study the forebrain GnRH system. Herein, we discuss the use of the Tg(GnRH3:EGFP) zebrafish line as a model for studying forebrain GnRH neurons, both in developing larvae and in sexually mature animals. We also discuss the potential use of this line to study regulation of GnRH3 system development. Published by Elsevier Inc.

1. Introduction GnRH is a decapeptide neurohormone that is crucial for reproduction. To date, 15 variants of GnRH have been described in vertebrates (Lethimonier et al., 2004; Kavanaugh et al., 2008). All vertebrate species studied have two or three paralogous forms of GnRH. The hypophysiotropic GnRH neurons are located primarily in the preoptic area (POA)–hypothalamus, but in most species are spread in a loose continuum from the olfactory region to the POA–hpothalamus (Campbell, 2007). Axons extend from the POA–hypothalamic GnRH neurons to innervate the hypophyseal blood portal system in higher vertebrates or pituitary in fish. GnRH decapeptide acts in the pituitary to induce the biosynthesis and release of gonadotropins from the adenohypophysis into the blood stream. The midbrain GnRH neuron population is located in the midbrain tegmentum (Herbison, 2006; Kah et al., 2007). Over the years, several GnRH nomenclature schemes have been developed. The recent consensus is the GnRH1/2/3 terminology (White et al., 1998). However this terminology is divergently used, either based on GnRH function/location, or based on GnRH gene phylogeny. The function/location methodology may however be challenging to use given that additional forms of GnRH will probably be found and the fact that not all functions of each GnRH are known. Herein the phylogenetic-based

* Corresponding author. E-mail address: [email protected] (Y. Zohar). 0016-6480/$ - see front matter Published by Elsevier Inc. doi:10.1016/j.ygcen.2009.01.012

GnRH1/2/3 terminology is used. According to this nomenclature, the zebrafish hypothalamic form is GnRH3. However in sea bream (Sparus aurata), for example, the hypophysiotropic form is GnRH1, and the terminal nerve (TN) form is GnRH3 (Lethimonier et al., 2004; Kuo et al., 2005). In mammals, the hypophysiotropic form is GnRH1. Recently, a fourth branch of the GnRH phylogenetic tree has been proposed, this branch includes lamprey GnRHs (Kavanaugh et al., 2008). In teleost species that have two forms of GnRH, the hypothalamic form is located in the olfactory bulbs (OB)–TN and POA– hypothalamus (Amano et al., 1991; Steven et al., 2003; Lethimonier et al., 2004; Abraham et al., 2008). In these species, GnRH3 has been shown to elicit gonadotropin-releasing activity and is thought to assimilate non-redundant functions of GnRH1 (Amano et al., 1995; Okubo and Nagahama, 2008). In fish species that have three forms of GnRH, the third form is located primarily in the olfactory and terminal nerve (TN) regions, although some overlap between the two forebrain forms exists (Gonzalez-Martinez et al., 2004; Okubo et al., 2006). GnRH neurons, either GnRH3, GnRH1, or both, depending on the species, extend a complex network of fibers that innervates many CNS regions. The multiple forms and locations of GnRH in the CNS as well as in extra-CNS regions, and the extensive innervation of the CNS by GnRH neurons, suggests that GnRH plays numerous roles throughout development and maturity. The classical view of GnRH as a gonadotropin-releasing hormone continues expanding to encompass additional roles.

152

E. Abraham et al. / General and Comparative Endocrinology 164 (2009) 151–160

During development, hypothalamic GnRH neurons conduct an extremely long tangential migration, originating in the olfactory region and reaching the POA-hypothalamus. This migrational path has been well documented in mammals (Wierman et al., 2004), however, as discussed below, it is still disputed with regard to teleosts (Whitlock et al., 2003; Gopinath et al., 2004; Wu et al., 2006). Two GnRH forms, GnRH2 and GnRH3, have been identified in the zebrafish brain using HPLC analysis (Powell et al., 1996). Several years later, two genes encoding these forms of GnRH were isolated and characterized (Torgersen et al., 2002; Steven et al., 2003). It appears that in zebrafish and several other cyprinids, the GnRH1 gene was lost during evolution (Kuo et al., 2005; Kah et al., 2007). In adult zebrafish, GnRH2 is localized to the midbrain tegmentum as was described in all other investigated jawed vertebrates, while GnRH3 is localized to the OB– TN as well as to the POA–hypothalamus (Steven et al., 2003). Additionally, levels of GnRH3 peptide in the adult zebrafish pituitary were shown to be 3- to 4-fold higher than those of GnRH2. Therefore, GnRH3 is considered to be the hypophysiotropic form (Powell et al., 1996; Steven et al., 2003). Our detailed investigations in zebrafish, using whole-mount ISH, a transient promoter–reporter expression system and the transgenic fish line suggest that hypothalamic GnRH3 neurons emerge from the olfactory area and migrate posteriorly to the hypothalamus. Thus, the OB–TN and POA–hypothalamic GnRH neuronal populations share the same embryonic origin. These findings are in accordance with results in other fish species that have only two molecular GnRH forms. However, these findings differ from other studies proposing that zebrafish hypothalamic GnRH neurons have a different origin. According to this hypothesis, the TN GnRH population originates from the cranial neural crest. In contrast, the hypothalamic population, which reportedly expresses a different form of immunoreactive GnRH, emerges from the adenohypophyseal region of the developing anterior neural plate (Whitlock et al., 2005, Wu et al., 2006), and thus does not migrate to its final hypothalamic location. The use of zebrafish for studying tangential neuron migration is also of importance in the context of human development and disease, especially with regard to better understanding GnRH-specific tangential migration developmental diseases such as Kallmann syndrome. This syndrome is caused by abnormal migration of GnRH neurons, resulting in idiopathic hypogonadotropic hypogonadism, and anosmia (Zenteno et al., 1999; MacColl et al., 2002; Sato et al., 2004). Gaining a better understanding of forebrain GnRH soma migration and the process that leads to aberrant migration serves to elucidate how reproductive abnormalities emerge and their basic etiology and biology. The other benefit of studying GnRH neuron migration is to expand our basic understanding of long range tangential neuronal migration. There is ample evidence regarding the importance of tangential migration of neurons to human development and health. One example is the tangential mode of migration of inhibitory interneurons (Tanaka et al., 2003). A defect in tangential migration of some inhibitory GABAergic neurons has been shown to be a factor in development of Tourette syndrome in children. The rostral migratory stream (migration of neurons from the subventricular zone to the olfactory bulb) that has been shown to occur in adult humans and is potentially crucial for maintenance and function of the adult CNS (Doetsch et al., 1997; Curtis et al., 2007) is another example. This minireview will focus on the advantages of using zebrafish to study development of the hypophysiotropic GnRH neurons, and summarize our recent findings regarding GnRH3 using the Tg(GnRH3:EGFP) zebrafish line.

2. Zebrafish as a model for early vertebrate development Many models have been used to study early development of vertebrates. An especially challenging area of research has been the field of CNS development. Zebrafish have proven over the past two decades to be an extremely powerful model for studying vertebrate developmental biology, as well as for research related to human diseases and potential targets for therapeutic intervention. This stems from several valuable traits found in zebrafish as well as the many molecular tools available for use in this species. Specifically, beneficial traits that zebrafish possess include (1) large number of embryos in each spawn, (2) fast development from fertilization to the phylotipic stage (24 h post-fertilization), (3) transparent embryos and larvae during development allowing for visualization of processes in vivo, (4) external fertilization and development, (5) the ability to raise hundreds of zebrafish inexpensively and in a small space, and (6) the relatively short time in which zebrafish reach reproductive maturity. In terms of available molecular tools: (a) the zebrafish sequencing project is close to completion, (b) forward genetics via mutagenesis is employed resulting in a plethora of well defined zebrafish mutants, (c) reverse genetics methods enable the creation of zebrafish transgenic lines in which markers or other proteins are specifically expressed, (d) gene knockdown (KD) and over-expression (OE) methods are readily available (Bradbury, 2004). Many transgenic zebrafish lines and mutant lines can be obtained from the zebrafish international resource center (ZIRC). The available methodologies used in zebrafish are continually being expanded and improved. Two examples of recent progress in genetic manipulation of zebrafish are the advent of Tol2 transposon element usage enabling faster and easier creation of transgenic lines (Kawakami et al., 1998; Kawakami, 2007), and the recent development of the synthetic zinc finger nuclease-mediated knockout (Ekker, 2008). With regard to developmental events, in zebrafish it is possible to endogenously express markers in specific tissues thus enabling observation of tissue/organ development in the whole animal. When this is coupled to KD or OE of specific genes, it allows in vivo analysis of downstream morphological, developmental and physiological effects. In contrast to the use of ISH and ICC, in which development is only seen as a snapshot in time and by which sensitivity is limited, endogenous expression of proteins such as EGFP or RFP is highly specific and temporally sensitive. In comparison to transgenic mammals, zebrafish afford the advantage of optic transparency, hence the ability to fully take advantage of endogenous marker expression. When utilizing reporter expressing transgenic lines, the use of phenylthiourea (PTU) which delays pigmentation is recommended as it allows prolonged observation of reporter expression. However, when exposing larvae to PTU one must take into consideration possible effects that this compound may have on thyroid function (Elsalini and Rohr, 2003) . One of the main issues in studying GnRH neurons and their development is the fact that only a modest number (700–1200 in mammals) of neurons are scattered along a continuum in an adult CNS (Rubin and King, 1994; Tsai et al., 2005), this has made studying GnRH neuron properties and development a challenge. With the advent of transgenesis and promoter–reporter protein expression, the use of animals in which GnRH neurons express florescent proteins has become possible, vastly improving visualization of GnRH neurons. Indeed, there are several lines of GnRH-GFP transgenic mice in use (Spergel et al., 1999; Suter et al., 2000; Han et al., 2004). Compared with these lines, use of a transgenic GnRH3:EGFP zebrafish allows clear observation of GnRH neurons not only in tissue slices, but also in the intact live animal, resulting in real-time

153

E. Abraham et al. / General and Comparative Endocrinology 164 (2009) 151–160

high-resolution analysis of developmental events, and of the factors effecting these events in a whole-animal context. Similar to the case in mammals, and unlike many other fish species, zebrafish has two GnRH forms (GnRH3 and GnRH2). In addition, the migratory path of hypophysiotropic GnRH neurons in teleosts is very similar to mammals. Combining the use of a Tg(GnRH3:EGFP) zebrafish line with the various traits of zebrafish as well as the molecular tools and mutants available, has greatly aided in a better understanding of GnRH system development, as well as the factors that control this process. 3. Development of the GnRH3:EGFP transgenic line The Tg(GnRH3:EGFP) line was developed in view of gaining a better understanding of GnRH3 system development. Using this line, one can not only follow GnRH3 neuronal development in detail in vivo, but also observe the specific effects of KD and OE of various factors on the GnRH3 system. The Tg(GnRH3:EGFP) line was developed using the GnRH3 zebrafish promoter (Palevitch et al., 2007; Abraham et al., 2008) (Fig. 1). The GnRH3:EGFP construct was first used to induce transient expression, this expression was found to be accurate based on ISH as well as previous studies in teleosts (Palevitch et al., 2007). This result indicated that the promoter region chosen encompassed all the necessary regulatory elements for correct EGFP expression in GnRH3 neurons. The Tg(GnRH3:EGFP) line is highly stable (currently at F8), and the EGFP expression pattern is robust and specific to GnRH3 somata and fibers (Abraham et al., 2008). 4. Development of GnRH3 fibers The distribution of hypophysiotropic GnRH fibers within the CNS of mature animals has been examined in multiple species such as masu salmon, European sea bass, rat, and mouse (Merchenthaler et al., 1980; Amano et al., 1997; Skynner et al., 1999; GonzalezMartinez et al., 2004). In addition to innervating the pituitary or median eminence in fish and mammals respectively, GnRH fibers innervate many additional regions of the mature CNS. These regions include the olfactory bulbs, telencephalon, preoptic region, hypothalamus, optic tectum, cerebellum, medulla oblongata, and spinal cord. However, the spatiotemporal nature of this innervation as it occurs during early development has been elusive. Using the Tg(GnRH3:EGFP) line, we have documented this developmental aspect of the GnRH3 system. GnRH3 neurons begin to extend fibers shortly after initiation of GnRH3 transcript expression at approximately 24 hpf. These fiber tracts have a clear directionality likely mediated by various factors in the olfactory region. By 36–48 hpf, five major tracts are formed: (a) two tracts connect the bilateral GnRH3 somata at the midline, these tracts form in the pallium and subpallium, dorsally and ventrally to the GnRH3 somata clusters, (b) a tract apposed to the anterior commissure, (c) two tracts extend along the optic nerve, cross at the midline and innervate the retina, (d) a tract

extending dorsally towards the region of the pineal gland, (e) a tract extending posteriorly towards the hypothalamus and POA. The bilateral tracts that enter the hypothalamus make a ventral turn towards the ventral hypothalamus and pituitary, although this ventral turn is significantly less pronounced than in mammals. A subset of fibers from this tract do not make the ventral turn and continue posteriorly, along to the spinal cord, and innervate the trunk (Fig. 2). While these are the major fiber tracts, additional fibers innervate the midbrain and hindbrain, as well as other regions (Abraham et al., 2008). Several interesting observations can be made based on the nature of this fiber development. The first fiber tracts to form are the ventral and dorsal commissures between the two GnRH3 somata clusters. The early establishment of these commissures is interesting and establishes that there is a morphological connection between the GnRH3 contralateral neurones. It is also important to note that this extensive fiber scaffold which is formed by 36–48 hpf and is in place well before the GnRH3 somata begin migrating. The tract that extends through the telencephalon and into the POA–hypothalamus, which appears to be apposed to the vomeronasal and terminal nerves (VNN/TN), serves as a scaffold for axophilic GnRH3 soma migration. Additionally, by 48 hpf the hypothalamic tract innervates the pituitary, suggesting that GnRH3 may have a modulatory role in early development of the HPG axis well before GnRH3 soma reach their final location in the POA–hypothalamus. Lastly, the extensive GnRH innervation that occurs during development and persists to maturity reflects the fact that GnRH3 plays multiple additional roles, possibly as a neuromodulator/neurotransmitter. Examples of such functions include the hypothesized role of GnRH in the retina as a neuromodulator of sensory processing (Maruska and Tricas, 2007), as well as the possible role of GnRH1/3 in the midbrain as a component of GnRH2 regulated sexual behavior and feeding (Barnett et al., 2006; Hoskins et al., 2008). When comparing the fiber distribution between fish that have two forms versus three forms of GnRH, it is evident that GnRH3 fibers in zebrafish (two form model) innervate the same CNS region innervated by both GnRH1 (primarily the pituitary) and GnRH3 (other CNS regions) in fish that possess three forms of GnRH. This raises two points: (1) that GnRH3 in zebrafish has taken over roles played by both GnRH1 and GnRH3 in three-form fish and (2) that these roles are sufficiently important to be conserved despite the loss of one GnRH gene in two form fish.

1. Pallium commissure

Olfactory organ 2. Anterior commissure

GnRH3 soma 3. Pineal region

5. Hypothalamic innervation

Exon II

Exon I GnRH3 promoter

Intron I EGFP 4. Optic chiasm

GnRH3 fibers

-1 Fig. 1. GnRH3:EGFP expression vector. A 2.4-kb sequence upstream of the gonadotropin-releasing hormone (GnRH3) decapeptide coding region including 1.3 kb of 50 flanking sequence, exon I, intron I and part of exon II was inserted into an expression vector p-EGFP1 upstream to a EGFP coding sequence.

Fig. 2. Schematic illustration of major fiber tracks extending from GnRH3 somata in the olfactory region at 48 hpf. Five major tracts are depicted from a ventral view: (1) pallium and subpallium commissures, (2) anterior commissure, (3) innervation of the pineal region, (4) innervation of the retina via the optic tract, and (5) bilateral innervation of the hypothalamus (Palevitch et al., 2007; Abraham et al., 2008).

154

E. Abraham et al. / General and Comparative Endocrinology 164 (2009) 151–160

5. Migration of GnRH3 somata

6. Localization of GnRH3 somata and fibers in mature zebrafish

Migration of GnRH somata is one of the better studied neural migrations that occurs during development (Wray, 2001, 2002; Cariboni et al., 2007), and yet much is unknown about this process. The lineage of forebrain GnRH neurons is unclear in both mammals and fish. Some evidence points to the olfactory placode as the source of GnRH neurons, although they may diverge from the olfactory cells during early placode differentiation (Wray, 2002). Other lines of evidence point to the boundry region between the respiratory epithelial area and the olfactory placode as the origin of GnRH neurons (Kramer et al., 2000). Given that the Tg(GnRH3:EGFP) model allows for direct observation of the GnRH3 neurons only from 24 hpf and on, it does not supply us with much direct information regarding the specific placodial lineage of GnRH neurons that immerge in the olfactory region. An alternative hypothesis is that GnRH3 neurons in zebrafish arise not in the olfactory region, but rather in the anterior limits of the neural fold from the same region that gives rise to the anterior pituitary (Whitlock, 2004). Contrary to the ‘olfactory region origin hypothesis’, a pituitary adjacent origin would mean that mature hypophysiotropic GnRH neurons first appear in the hypothalamic region. While this is certainly not the case in mammals, several reports in zebrafish suggest the presence of GnRH3-immunoreactive cells in the hypothalamus as early as 28 hpf, which would support the ‘anterior neural fold origin hypothesis’ (Whitlock et al., 2003; Gopinath et al., 2004; Wu et al., 2006). In our observations, GnRH3 neurons first appear at approximately 24 hpf and are limited to the medial boundary of the olfactory organs (Fig 3). ISH analysis as well as observations of the Tg(GnRH3:EGFP) have not detected GnRH3 neuron immergence in the hypothalamic region (Palevitch et al., 2007; Abraham et al., 2008) (Fig 3). This finding can not be attributed to a basic deficiency of the Tg(GnRH3:EGFP) line since GnRH3-EGFP expressing neurons in mature Tg(GnRH3:EGFP) fish are found both in the olfactory region and in the hypothalamus (Figs. 4 and 5). Moreover, our preliminary results show that early developmental ablation of olfactory region GnRH3 neurons results in a lack of GnRH3 neurons in the POA–hypothalamus of mature fish, strongly suggesting that the POA–hypothalamic GnRH3 neurons originate in the olfactory region. Migration of GnRH3 somata from the olfactory region begins at approximately 3 dpf. Prior to this point there are both differentiation and positional alterations that are a consequence of tissue remodeling in the olfactory region. Soma migration takes place in an axophilic manner along the preexisting GnRH3 fiber tract that innervates the hypothalamus. Presumably, this migration path occurs apposed to the preexisting VNN/TN, as shown in mammals (Schwarting et al., 2001, 2004). Using living Tg(GnRH3:EGFP) larvae, we observed continuous migration of GnRH3 somata between 3 and 15 dpf (later observations are unfeasible due to loss of transparency). This migration occurs as a bilateral stream of neurons moving from the olfactory region through the TN ganglion and ventral telencephalon to the ventral hypothalamus (Figs. 3 and 6). The first somata reach the hypothalamus at approximately 12–13 dpf. This developmental profile is in agreement with what has been seen when conducting ISH experiments during development and in adults (Steven et al., 2003; Palevitch et al., 2007; Abraham et al., 2008). The early stages of this migration pattern (24–72 hpf) are in agreement with descriptions by Whitlock et al. (2005) and Sherwood and Wu (2005), with the exception of the aforementioned hypothalamic GnRH3 expression. Interestingly, the zebrafish gonad is established and primordial germ cells become mitotically active at about 10 dpf (Braat et al., 1999), coinciding with the first GnRH3 soma entering the hypothalamus.

The Tg(GnRH3:EGFP) line has proven to be an excellent model not only for studying GnRH3 system development, but also for establishing localization of GnRH3 somata and fibers in sexually mature zebrafish. Using the Tg(GnRH3:EGFP), these observations can be done either via cryosections or simply by observing EGFP expression in the intact brain. GnRH3 somata are comprised of several distinct populations in the mature CNS (Fig. 7). The first population is in the OB/ TN, These neurons are located bilaterally in the ventral OB and extend in the form of intermittent soma clusters into the ventral nucleus of the ventral telencephalic area (Vv) (Fig. 4). These somata form in clusters, express high EGFP levels, and are connected by a robust fiber tract (Figs. 4 and 5). A second population of somata is the POA population. These somata are located in the anteroventral parvocellular preoptic nucleus (PPa) apposed to the anterior commissure and extend posteriorly to reach the posteroventral parvocellular preoptic nucleus (PPp) (Wullimann and Reichert, 1996). A third smaller population of GnRH3 somata is the ventral hypothalamic population located in the ventral periventricular hypothalamus (Hv) just above the adenohypophysis (Fig. 5). GnRH3 somata in these two regions make up the POA–hypothalamic population that is the hypophysiotropic population which innervates the pituitary. The POA–hypothalamic populations are distinctly different from the OB/TN population in that they are well dispersed within the ventral POA and seem to express lower levels of EGFP (Figs. 4 and 5). The difference in the level of EGFP expressed by somata belonging to these two populations may reflect differential control over GnRH3 transcription. Such a difference would be warranted given that these two populations play different roles while being controlled by the same GnRH3 promoter. This suggests that control over GnRH3 transcription is mediated by trans-acting promoter binding elements that are GnRH3 population-specific. Finally, a fourth population of GnRH3 somata are located in the facial lobe (not shown). These findings are in agreement with multiple previous studies of GnRH in teleosts (Lethimonier et al., 2004). The locations of GnRH3 in mature zebrafish encompass the location of both GnRH1 and GnRH3 in fish that have three forms of GnRH. This organization of GnRH3 neurons is similar to what is seen in Masu salmon (Oncorhynchus masou) (Amano et al., 1991), a teleost that also has two GnRH forms. The small somata population in the facial lobe is in agreement with previous ISH-based data (Steven et al., 2003), these cells are most probably associated with GnRH3-expressing cells found in the trigeminal ganglia (Abraham et al., 2008), as both these regions have been shown to be connected (Kerem et al., 2005; Xue et al., 2006). Location of GnRH somata in the facial lobe/trigeminal ganglia indicates possible involvement of GnRH3 neurons in smell, taste and tactile stimuli (Kiyohara and Caprio, 1996; Kerem et al., 2005). The general localization of forebrain GnRH neurons in zebrafish, as in other teleost species, is relatively conserved and comparable to localization of forebrain GnRH neurons in mammals. However, the terminology for CNS regions in teleosts somewhat differs from that in mammals. Comparable terminology of relevant CNS regions in both classes can be found in Table 1. These findings clearly demonstrate that in zebrafish both the OB/TN GnRH population and the hypophysiotropic POA–hypothalamus populations that innervate the pituitary are indeed GnRH3expressing populations as shown previously (Steven et al., 2003) and are controlled by the same promoter. GnRH3 fibers are located widely in the CNS (Figs. 4, 5, 7 and 8). Areas in which GnRH3 fibers are localized include the olfactory

E. Abraham et al. / General and Comparative Endocrinology 164 (2009) 151–160

155

Fig. 3. GnRH3 spatiotemporal expression profile as seen via use of ISH and the Tg(GnRH3:EGFP) line at 1–5 dpf. (A–C) Sagittal sections of whole-mount ISH specimens (MB, midbrain; Tel, telencephalon), anterior towards the left. mRNA expression in neurons as dark purple signal (arrowheads). (D–F) Dorsal view whole-mount ISH of GnRH3 mRNA (arrows), anterior towards the top. (G) Frontal view. (H and I) Dorsal view of Tg(GnRH3:EGFP) larvae. Regions framed by unbroken line represent the olfactory organs, oval represents hypothalamic region, dashed line represents the ventral telencephalon. Chronological progression of the larval samples are as follows: (A, D, and G) 2 dpf, (B, E, and H) 3 dpf, (F and I) 5 pdf, (C) 6 dpf. Scale bars: 50 lm.

bulbs, telencephalon, optic tectum, cerebellar corpus, hypothalamus, and hindbrain. Particularly robust are fibers extending from the GnRH3 clusters in the OB towards the telencephalon and midbrain, as well as fibers extending towards the pituitary from the PPa and the Hv (Fig. 8). These fibers clearly take two tracts, one turning in a ventral direction towards the pituitary, and another continuing posteriorly (Fig. 8). These findings are in agreement with previous accounts regarding GnRH fiber localization in mammals as well as fish. 7. Studying factors involved in GnRH3 system development A plethora of factors have been shown to influence and guide GnRH neuron migration and fiber targeting during early development. These factors include secreted molecules such as GABA and Netrin1, cell surface proteins such as various receptors and adhesion molecules, and nuclear factors (Kramer and Wray, 2000b; Wray, 2001; Schwarting et al., 2004; Tobet and Schwarting, 2006; Cariboni et al., 2007). Given the complexity of GnRH neuron migration, and the diversity of regions that these neurons encounter, it is reasonable to assume that different factors control different stages of migration in different regions of the migratory route.

One important distinction to make, especially given the findings from the Tg(GnRH3:EGFP) line, is that GnRH3 system development is divided into two distinct phases. Phase one consists of early fiber development. As discussed, these fibers innervate multiple CNS regions and some also serve as a scaffold for subsequent soma migration. The second phase begins only after the first phase ends and consists of GnRH3 soma migration along one of their own fiber tracts from the olfactory region into the POA-hypothalamus. It is clear that each of these two phases is controlled either by specific factors, or by the same factors that are expressed in a spatiotemporally distinct manner. Although much work has been published regarding factors controlling hypophysiotropic GnRH neuronal migration (Cariboni et al., 2007), reference to the difference in control over the two developmental phases has been lacking, presumably due to the difficulty in visualizing early development of forebrain GnRH fibers. As has been noted throughout this text, the zebrafish is an excellent model for studying early development of the GnRH3 systems. The attributes of zebrafish larvae, coupled with the availability of molecular biology tools, make this an excellent model system. Moreover, the availability of the Tg(GnRH3:EGFP) line enables the use of generic zebrafish tools to study development of the GnRH3 system.

156

E. Abraham et al. / General and Comparative Endocrinology 164 (2009) 151–160

Fig. 4. GnRH3 neurons in a mature Tg(GnRH3:EGFP) zebrafish intact brain. (A) Ventral view of the ON, OB, Tel and PPa. GnRH3 somata are located in a continuum from the anteroventral OB to the Vv and are connected by fiber tracts. (B) Enlarged image of upper boxed area in (A), GnRH3 soma clusters and connecting fibers can be seen in the ventral OB and extending to the Vv. (C) Enlarged fluorescent image on lower boxed area in (A). GnRH3 somata are scattered in the PPa apposed to the anterior commissure. Scale bars: (A and C) 100 lm and (B) 50 lm. ON, olfactory nerve; OB, olfactory bulb; Tel, telencephalon; PPa, parvocellular preoptic nucleus; Vv, ventral nucleus of the ventral telencephalon.

Below, we briefly discuss three factors that are of interest with regard to GnRH3 system development and how zebrafish techniques and lines can be combined to further study these factors.

crine modulation of nestin levels by GnRH3 may be involved in control over GnRH3 neuron proliferation. 7.2. Netrin1a/Dcc

7.1. GnRH3 Expression of GnRH transcripts and GnRH receptors in fish, as well as in mammals, begins at early stages of development, prior to establishment of the HPG axis (Romanelli et al., 2004; Whitlock et al., 2006; Palevitch et al., 2007). This expression coincides with GnRH neuronal differentiation and fiber development. In vitro studies in GnRH cell lines have shown that exposure of these cell lines to GnRH can cause cytoskeletal remodeling and growth cone extension abnormalities (Krsmanovic et al., 2003; Romanelli et al., 2004). To examine this phenomenon in vivo, morpholino-mediated KD of GnRH3 was conducted in Tg(GnRH3:EGFP) zebrafish. The resulting effects included abnormalities to early GnRH3 fiber development, as well as abnormal localization and proliferation of GnRH3 somata (Abraham et al., 2008). GnRH3 KD in zebrafish preformed in another study (Wu et al., 2006), revealed general CNS and eye deformaties. The variation in results may be due to differences in the amount of MO injected, MO sequence and the ability to detect abnormalities in the GnRH3 system. However, our findings are in agreement with the aforementioned in vitro studies and show that GnRH3 itself is a factor that acts in an autocrine fashion to regulate early development of GnRH3 somata and fibers. Enhanced proliferation of GnRH3 neurons as a result of GnRH3 KD may well be connected to the finding that in vitro exposure of the FNC-B4 GnRH-producing cell line to GnRH induces a decrease in nestin (Romanelli et al., 2004). Nestin, a marker for stem cells and precursor cells in the developing embryonic nervous system, is expressed by early GnRH neurons and localized to active neurogenesis regions in adult zebrafish (Dahlstrand et al., 1995; Kramer and Wray, 2000a; Mahler and Driever, 2007). Thus, auto-

Netrin is a diffusible molecule that was first described as a factor that induces commissural axon outgrowth and is required to guide commissural neuron axons to the midline (Serafini et al., 1994, 1996; Charron et al., 2003). In netrin1 mutant mice, commissural axons fail to enter the spinal cord and are misguided (Serafini et al., 1996). The Netrin1 receptor, Dcc, mediates chemoattraction to Netrin1 while expression of both Dcc and Unc5 receptors is thought to mediate chemorepulsion to netrin (Hong et al., 1999; Round and Stein, 2007). In mice, a deficiency in Netrin1- or Dcc-induced abnormal targeting of GnRH1 neurons (Schwarting et al., 2001, 2004). Zebrafish has four forms of Netrin, Netrin1a/1b/2/4 (Park et al., 2005). Our studies have focused on Netrin1a. The early spatiotemporal expression pattern of Netrin1a and Dcc in zebrafish in conjuncture with early development of GnRH3 fibers, suggests a role for Nertin1a in early GnRH3 fiber development and attraction towards the midline (Fig. 9). Indeed, our preliminary results show that when conducting KD of either Netrin1a or Dcc, development of GnRH3 fibers that target or cross the midline is disrupted. These fiber tracts include the pallium and subpallium commissures, anterior commissure, and innervation of the retina via the optic chiasm. These preliminary findings suggest that Netrin1a in zebrafish is involved in mediating midline attraction of GnRH3 fibers. 7.3. Cxcl12 and Cxcr4 Stromal cell derived factor-1, Cxcl12 (previously SDF-1), is a chemokine that has been shown to be involved in fiber targeting and soma migration (Tran and Miller, 2003; Stumm et al., 2007). In mice, SDF-1 and its receptor CXCR4 are involved in GnRH1

E. Abraham et al. / General and Comparative Endocrinology 164 (2009) 151–160

157

Fig. 5. GnRH3 neurons are located in the preoptic area and periventricular hypothalamus. (A, B, D, and E) Sagittal cryosections of Tg(GnRH3:EGFP) mature zebrafish brains, anterior towards the left. (A) GnRH3 somata are located in the Vv (arrow) and PPa (boxed area), GnRH3 fibers are located in various regions (arrowhead). (B) Enlarged image of boxed area in (A), GnRH3 somata in the PPa (arrows). (C) Ventral view of PPa, anterior towards the top (whole brain, disconnected telencephalon), circles surround clusters of GnRH3 somata in the PPa. (D) GnRH3 somata are located in the Hv (boxed area). (E) Enlarged image of boxed area in (C), GnRH3 somata in the Hv (arrow). Scale bars: (A and D) 250 lm and (B and E) 50 lm. Tel, telencephalon; TeO, optic tectum; PPa, parvocellular preoptic nucleus; Vv, ventral nucleus of the ventral telencephalon; Hv, ventral zone of the periventricular hypothalamus.

Fig. 6. GnRH3 spatiotemporal expression profile as seen via use of ISH and the Tg(GnRH3:EGFP) line at 6, 12 and 25 dpf. (A–C) Dorsal view, anterior towards the top. (A) At 6 dpf, migration continues along the lateral aspects of the telencephalon (arrows). The dashed line represents the telencephalon border. (B) At 12 dpf, the first soma reach the hypothalamus (arrow). (C) Representation of the square region in (B) (different animals). At 25 dpf, GnRH3 localization remains similar to 12 dpf. mRNA is localized to the TN (arrowheads) caudally to the olfactory bulb (OB, dotted line) and along the lateral aspects of the telencephalon (Tel, arrows), and cells are seen in the preoptic region (POA, asterisk). Scale bars: 100 lm.

migration from the olfactory region to the forebrain (Schwarting et al., 2006). An additional line of evidence pointing to Cxcl12 is the involvement of this factor in organization of the olfactory plac-

ode and targeting of olfactory sensory neurons during zebrafish development (Miyasaka et al., 2007). Based on these findings, Cxcl12 and its receptor Cxcr4 are candidates for study of GnRH3

158

E. Abraham et al. / General and Comparative Endocrinology 164 (2009) 151–160

CCe TeO Tel OB

Vv

LVII

PPa Hv Pit

Fig. 7. A sagittal schematic of a mature zebrafish brain. Green dots represent GnRH3 soma green lines represent GnRH3 fibers. Four distinct GnRH3 populations are located within the CNS (a) in the anteroventral OB extending towards the Vv and presumed TN, (b) in the PPa, (c) in the Hv and (d) in the LVII. GnRH3 fibers are located in many brain regions including the OB, Tel, Vv, TeO, Hv, midbrain and hindbrain. Especially prominent tracks extend out of the OB towards the Tel and TeO, as well as from the PPa making a ventral turn into the Hv and pituitary. OB, olfactory bulb; Tel, telencephalon; TeO, optic tectum; Vv, ventral nucleus of the ventral telencephalon; PPa, parvocellular preoptic nucleus; Hv, ventral zone of the periventricular hypothalamus; CCe, cerebellar corpus; LVII, facial lobe; TN, terminal nerve ganglion; Pit, pituitary. (For interpretation of the references to color in this figure legend the reader is referred to the web version of the article.)

Table 1 Comparison of forebrain GnRH-relevant CNS terminology as defined in teleostei and mammals. Teleostei terminology

Mammalian terminology

Olfactory bulbs (OB) Telencephalon (Tel) Optic tectum (TeO) Ventral nucleus of the ventral telencephalon (Vv) Parvocellular preoptic nucleus (PPa) Ventral zone of the periventricular hypothalamus (Hv) Cerebellar corpus (CCe) Facial lobe (LVII)

Olfactory bulbs Telencephalon Superior colliculus Ventral diagonal band Medial preoptic nucelus Ventromedial hypothalamus Cerebellum Facial lobe

system development. This is a case in which several zebrafish-specific methods can be advantageously used to examine the role of Cxcl12/Cxcr4 in GnRH3 development. One tool is the morpholino-mediated KD of Cxcl12 and CXCR4 and another is the use of the zebrafish odysseus mutant line in which the cxcr4b gene is mutated and is thus presumably non-functional (Miyasaka et al., 2007). The use of these tools in Tg(GnRH3:EGFP) zebrafish in the form of KD or crosses between the odysseus mutant and the Tg(GnRH3:EGFP) fish to obtain a CXCR4b/–GnRH3:EGFP line,

are valuable methods of clarifying the role of Cxcl12 and Cxcr4. These methods enable real-time in vivo observations as to the effect of these specific factors on GnRH3 development. Indeed, our preliminary results, using both Cxcl12 and Cxcr4 KD as well as the odysseus mutant line, show that Cxcl12 and Cxcr4 KD/KO induces abnormal targeting of GnRH3 fibers as well as disruption to GnRH3 soma migration. An additional method that can be generically utilized to examine the role of various factors vis-avis GnRH3 is co-localization of GnRH3 somata and fibers with the factor of interest, in this case Cxcr4 and Cxcl12. If these factors are involved in GnRH3 development, one would expect to see expression of Cxcr4 by GnRH3 neurons or by elements, such as the terminal nerve, that may be essential for proper GnRH3 migration. Additionally, acting as a chemoattractant, Cxcl12 would be expected to be found in areas to which GnRH3 is neurons are targeted. 8. Conclusions Use of the Tg(GnRG3:EGFP) line has allowed detailed observations into GnRH3 system development as a continuum of events, as well as GnRH3 somata and fiber localization in mature animals. In terms of early GnRH3 development, use of the transgenic line has revealed the intricate network of GnRH3 fibers that is established during early development. This network includes multiple commissures and tracts that innervate a variety of CNS regions. These tracts can be separated into two groups: (a) tracts that target posterior regions including the POA–hypothalamic tract that targets the hypothalamus and pituitary, as well as tracts that target other CNS regions including the midbrain and hindbrain and (b) tracts that target the midline to form commissures or chiasms. These findings point to the distinction that must be made between early GnRH3 fiber network development and later GnRH somata migration. Study of factors involved in fiber targeting and somata migration should be conducted so as to clarify the differential regulation of these two distinct yet interdependent processes. Localization of GnRH3 fibers within the mature CNS is widespread, pointing to an additional role of GnRH3 throughout life as a neuromodulator/neurotransmitter. Development and migration of GnRH3 somata is captured in detail in the transgenic line. Our findings clearly indicate that GnRH3 neurons immerge in the olfactory region and migrate to the POA–hypothalamus. In mature animals, GnRH3 somata are localized to several distinct regions, the OB/TN region, the

Fig. 8. GnRH3 fibers are localized to many CNS regions. (A and B) Sagittal cryosections of Tg(GnRH3:EGFP) mature zebrafish brains. These sections of the forebrain show GnRH3 axons localized to the OB, Tel, PPa and Hv. (A) Major fiber tract (arrow) is seen exiting the ventral olfactory bulb and entering the telencephalon (Vv). Additional tracts extend from the PPa and turn ventrally towards the Hv (arrowheads). (B) Close-up view of GnRH3 fibers extending from the PPa and turning ventrally towards the Hv (arrows). A subset of fibers that do not turn ventrally target the TeO and other regions (arrowhead). Scale bars: (A) 250 lm and (B) 50 lm. Tel, telencephalon; TeO, optic tectum; PPa, parvocellular preoptic nucleus; Vv, ventral nucleus of the ventral telencephalon; Hv, ventral zone of the periventricular hypothalamus.

E. Abraham et al. / General and Comparative Endocrinology 164 (2009) 151–160

Pallium commissure

Acknowledgments

Olfactory organ Anterior commissure

GnRH3 soma

159

We thank John Stubblefield for editing this manuscript. We also thank Kathy Kight, Dr. Ulrike Klenke and Dr. Susan Wray for their guidance and assistance. This study was supported by the US–Israel Bi-national Agricultural Research and Development (BARD) Foundation (Grant 3428-03) and by grants to Y.Z. from Maryland Sea Grant (SA7528051-M) and the National Science Foundation (IOB-0548620). This publication is Contribution No. 08-196 from the University of Maryland Centre Of Marine Biotechnology. References

Dcc Optic chiasm

Netrin1a GnRH3 fibers

Fig. 9. Schematic illustration of a 36-hpf Tg(GnRH3:EGFP) non-treated larva. Netrin1a and Dcc expression localization and the GnRH3 fiber tracts that are influenced by Netrin1a and Dcc KD are depicted. Netrin1a (red) is expressed along the midline and along the optic tract (Park et al., 2005), while Dcc is expressed in the olfactory region and along the lateral aspects of the telencephalon and hypothalamus (Fricke and Chien 2005). All three fiber extensions depicted (subpallium commissure, anterior commissure and optic tract) are disrupted by Netrin1a and Dcc KD. (For interpretation of the references to color in this figure legend the reader is referred to the web version of the article.)

POA–hypothalamus region and the facial lobe/trigeminal ganglia. This distribution of GnRH3 neurons agrees with the hypothesis that all forebrain GnRH populations in zebrafish consist of one form of GnRH that originates in the olfactory region. This finding is further bolstered by data that targeted ablation of GnRH3 neurons in the olfactory region leads to loss of the POA–hypothalamic GnRH population. In zebrafish, as in mammals and some fish species, loss of a third form of GnRH occurred during evolution. GnRH3 somata and fibers are located in regions occupied by both GnRH1 and GnRH3 neurons, as reported in species that have three forms of GnRH. This suggests that the loss of a third form of GnRH was compensated for by expansion of GnRH3 distribution within the CNS, thus taking on non-redundant roles played by the lost third GnRH form. To some extent, this is probably also the case in rodents, primates and humans. The degree of ease and clarity by which GnRH3 neurons can be observed during development, coupled with the traits of zebrafish and the molecular tools available in this species, make the Tg(GnRH:EGFP) zebrafish line highly effective in studying forebrain GnRH. Use of this line in conjunction with KD, OE, co-localization, as well as crosses with existing mutant lines, will undoubtedly reveal much new information regarding the early establishment of the forebrain GnRH system, the factors controlling this complex developmental event, and its functional significance. Additional areas of future work may include use of this transgenic line in electrophysiological studies. Using this line, GnRH3 neurons can easily be targeted for patch-clamp recordings both in intact larvae as well as in mature extracted brains. Another promising area is the use of Tg(GnRH:EGFP) larvae and adults at different developmental stages to isolate GnRH3 neurons, possibly via use of flow cytometry. Isolated GnRH3 neurons from different time points can be used for microarray based and functional genomics studies that may allow further insight into topics such as neuron migration, transition to puberty, and reproductive competence.

Abraham, E., Palevitch, O., et al., 2008. Early development of forebrain gonadotrophin-releasing hormone (GnRH) neurones and the role of GnRH as an autocrine migration factor. J. Neuroendocrinol. 20 (3), 394–405. Amano, M., Oka, Y., et al., 1991. Immunocytochemical demonstration of salmon GnRH and chicken GnRH-II in the brain of masu salmon, Oncorhynchus masou. J. Comp. Neurol. 314 (3), 587–597. Amano, M. et al., 1995. Salmon GnRH synthesis in the preoptic area and the ventral telencephalon is activated during gonadal maturation in female Masu Salmon. Gen. Comp. Endocrinol. 99 (1), 13–21. Amano, M., Urano, A., et al., 1997. Distribution and function of gonadotropinreleasing hormone (GnRH) in the teleost brain. Zool. Sci. 14 (1), 1–11. Barnett, D.K., Bunnell, T.M., et al., 2006. Gonadotropin-releasing hormone II stimulates female sexual behavior in marmoset monkeys. Endocrinology 147 (1), 615–623. Braat, A.K., Zandbergen, T., et al., 1999. Characterization of zebrafish primordial germ cells: morphology and early distribution of vasa RNA. Dev. Dyn. 216 (2), 153–167. Bradbury, J., 2004. Small fish, big science. PLoS Biol. 2 (5), E148. Campbell, R.E., 2007. Defining the gonadotrophin-releasing hormone neuronal network: transgenic approaches to understanding neurocircuitry. J. Neuroendocrinol. 19 (7), 561–573. Cariboni, A., Maggi, R., et al., 2007. From nose to fertility: the long migratory journey of gonadotropin-releasing hormone neurons. Trends Neurosci. 30 (12), 638– 644. Charron, F., Stein, E., et al., 2003. The morphogen sonic hedgehog is an axonal chemoattractant that collaborates with netrin-1 in midline axon guidance. Cell 113 (1), 11–23. Curtis, M.A., Kam, M., et al., 2007. Human neuroblasts migrate to the olfactory bulb via a lateral ventricular extension. Science 315 (5816), 1243–1249. Dahlstrand, J., Lardelli, M., et al., 1995. Nestin mRNA expression correlates with the central nervous system progenitor cell state in many, but not all, regions of developing central nervous system. Brain Res. Dev. Brain Res. 84 (1), 109–129. Doetsch, F., Garcia-Verdugo, J.M., et al., 1997. Cellular composition and threedimensional organization of the subventricular germinal zone in the adult mammalian brain. J. Neurosci. 17 (13), 5046–5061. Ekker, S.C., 2008. Zinc finger-based knockout punches for zebrafish genes. Zebrafish 5 (2), 121–123. Elsalini, O.A., Rohr, K.B., 2003. Phenylthiourea disrupts thyroid function in developing zebrafish. Dev. Genes Evol. 212 (12), 593–598. Fricke, C., Chien, C.B., 2005. Cloning of full-length zebrafish dcc and expression analysis during embryonic and early larval development. Dev. Dyn. 234 (3), 732–739. Gonzalez-Martinez, D., Zmora, N., et al., 2004. New insights in developmental origins of different GnRH (gonadotrophin-releasing hormone) systems in perciform fish: an immunohistochemical study in the European sea bass (Dicentrarchus labrax). J. Chem. Neuroanat. 28 (1–2), 1–15. Gopinath, A., Andrew Tseng, L., et al., 2004. Temporal and spatial expression of gonadotropin releasing hormone (GnRH) in the brain of developing zebrafish (Danio rerio). Gene Expr. Patterns 4 (1), 65–70. Han, S.K., Todman, M.G., et al., 2004. Endogenous GABA release inhibits the firing of adult gonadotropin-releasing hormone neurons. Endocrinology 145 (2), 495– 499. Herbison, A.E., 2006. Physiology of the Gonadotropin-Releasing Hormone Neuronal Network, Elsevier, New York. Hong, K., Hinck, L., et al., 1999. A ligand-gated association between cytoplasmic domains of UNC5 and DCC family receptors converts netrin-induced growth cone attraction to repulsion. Cell 97 (7), 927–941. Hoskins, L.J., Xu, M., et al., 2008. Interactions between gonadotropin-releasing hormone (GnRH) and orexin in the regulation of feeding and reproduction in goldfish (Carassius auratus). Horm. Behav. 54 (3), 379–385. Kah, O., Lethimonier, C., et al., 2007. GnRH and GnRH receptors in metazoa: a historical, comparative, and evolutive perspective. Gen. Comp. Endocrinol. 153 (1–3), 346–364. Kavanaugh, S.I., Nozaki, M., et al., 2008. Origins of gonadotropin-releasing hormone (GnRH) in vertebrates: identification of a novel GnRH in a basal vertebrate, the sea lamprey. Endocrinology 149 (8), 3860–3869. Kawakami, K., 2007. Tol2: a versatile gene transfer vector in vertebrates. Genome Biol. 8 (Suppl. 1), S7.

160

E. Abraham et al. / General and Comparative Endocrinology 164 (2009) 151–160

Kawakami, K., Koga, A., et al., 1998. Excision of the tol2 transposable element of the medaka fish, Oryzias latipes, in zebrafish, Danio rerio. Gene 225 (1–2), 17–22. Kerem, G., Yoshimoto, M., et al., 2005. Somatotopic organization of the trigeminal ganglion cells in a cichlid fish, Oreochromis (Tilapia) niloticus. Brain Behav. Evol. 65 (2), 109–126. Kiyohara, S., Caprio, J., 1996. Somatotopic organization of the facial lobe of the sea catfish Arius felis studied by transganglionic transport of horseradish peroxidase. J. Comp. Neurol. 368 (1), 121–135. Kramer, P.R., Wray, S., 2000a. Midline nasal tissue influences nestin expression in nasal-placode-derived luteinizing hormone-releasing hormone neurons during development. Dev. Biol. 227 (2), 343–357. Kramer, P.R., Wray, S., 2000b. Novel gene expressed in nasal region influences outgrowth of olfactory axons and migration of luteinizing hormone-releasing hormone (LHRH) neurons. Genes Dev. 14 (14), 1824–1834. Kramer, P.R., Guerrero, G., et al., 2000. Ectopic expression of luteinizing hormonereleasing hormone and peripherin in the respiratory epithelium of mice lacking transcription factor AP-2alpha. Mech. Dev. 94 (1–2), 79–94. Krsmanovic, L.Z., Mores, N., et al., 2003. An agonist-induced switch in G protein coupling of the gonadotropin-releasing hormone receptor regulates pulsatile neuropeptide secretion. Proc. Natl. Acad. Sci. USA 100 (5), 2969–2974. Kuo, M.W., Lou, S.W., et al., 2005. Chromosomal organization, evolutionary relationship, and expression of zebrafish GnRH family members. J. Biomed. Sci. 12 (4), 629–639. Lethimonier, C., Madigou, T., et al., 2004. Evolutionary aspects of GnRHs, GnRH neuronal systems and GnRH receptors in teleost fish. Gen. Comp. Endocrinol. 135 (1), 1–16. Wullimann, M.F., B.R., Reichert, H., 1996. Neuroanatomy of the Zebrafish Brain: A Topological Atlas. MacColl, G., Quinton, R., et al., 2002. GnRH neuronal development: insights into hypogonadotrophic hypogonadism. Trends Endocrinol. Metab. 13 (3), 112–118. Mahler, J., Driever, W., 2007. Expression of the zebrafish intermediate neurofilament Nestin in the developing nervous system and in neural proliferation zones at postembryonic stages. BMC Dev. Biol. 7, 89. Maruska, K.P., Tricas, T.C., 2007. Gonadotropin-releasing hormone and receptor distributions in the visual processing regions of four coral reef fishes. Brain Behav. Evol. 70 (1), 40–56. Merchenthaler, I., Kovacs, G., et al., 1980. The preoptico-infundibular LH–RH tract of the rat. Brain Res. 198 (1), 63–74. Miyasaka, N., Knaut, H., et al., 2007. Cxcl12/Cxcr4 chemokine signaling is required for placode assembly and sensory axon pathfinding in the zebrafish olfactory system. Development 134 (13), 2459–2468. Okubo, K., Sakai, F., et al., 2006. Forebrain gonadotropin-releasing hormone neuronal development: insights from transgenic medaka and the relevance to X-linked Kallmann syndrome. Endocrinology 147 (3), 1076–1084. Okubo, K., Nagahama, Y., 2008. Structural and functional evolution of gonadotropinreleasing hormone in vertebrates. Acta Physiol. (Oxf), 193 (1), 3–15. Palevitch, O., Kight, K., et al., 2007. Ontogeny of the GnRH systems in zebrafish brain: in situ hybridization and promoter–reporter expression analyses in intact animals. Cell Tissue Res. 327 (2), 313–322. Park, K.W., Urness, L.D., et al., 2005. Identification of new netrin family members in zebrafish: developmental expression of netrin 2 and netrin 4. Dev Dyn. 234 (3), 726–731. Powell, J.F. et al., 1996. Molecular forms of GNRH in three model fishes: rockfish, medaka and zebrafish. J. Endocrinol. 150 (1), 17–23. Romanelli, R.G., Barni, T., et al., 2004. Expression and function of gonadotropinreleasing hormone (GnRH) receptor in human olfactory GnRH-secreting neurons: an autocrine GnRH loop underlies neuronal migration. J. Biol. Chem. 279 (1), 117–126. Round, J., Stein, E., 2007. Netrin signaling leading to directed growth cone steering. Curr. Opin. Neurobiol. 17 (1), 15–21. Rubin, B.S., King, J.C., 1994. The number and distribution of detectable luteinizing hormone (LH)–releasing hormone cell bodies changes in association with the preovulatory LH surge in the brains of young but not middle-aged female rats. Endocrinology 134 (1), 467–474. Sato, N., Katsumata, N., et al., 2004. Clinical assessment and mutation analysis of Kallmann syndrome 1 (KAL1) and fibroblast growth factor receptor 1 (FGFR1 or KAL2) in five families and 18 sporadic patients. J. Clin. Endocrinol. Metab. 89 (3), 1079–1088. Schwarting, G.A., Kostek, C., et al., 2001. Deleted in colorectal cancer (DCC) regulates the migration of luteinizing hormone–releasing hormone neurons to the basal forebrain. J. Neurosci. 21 (3), 911–919.

Schwarting, G.A., Raitcheva, D., et al., 2004. Netrin 1-mediated chemoattraction regulates the migratory pathway of LHRH neurons. Eur. J. Neurosci. 19 (1), 11– 20. Schwarting, G.A., Henion, T.R., et al., 2006. Stromal cell-derived factor-1 (chemokine C-X-C motif ligand 12) and chemokine C-X-C motif receptor 4 are required for migration of gonadotropin-releasing hormone neurons to the forebrain. J. Neurosci. 26 (25), 6834–6840. Serafini, T., Kennedy, T.E., et al., 1994. The netrins define a family of axon outgrowth-promoting proteins homologous to C. elegans UNC-6. Cell 78 (3), 409–424. Serafini, T., Colamarino, S.A., et al., 1996. Netrin-1 is required for commissural axon guidance in the developing vertebrate nervous system. Cell 87 (6), 1001–1014. Sherwood, N.M., Wu, S., 2005. Developmental role of GnRH and PACAP in a zebrafish model. Gen. Comp. Endocrinol. 142 (1–2), 74–80. Skynner, M.J., Slater, R., et al., 1999. Promoter transgenics reveal multiple gonadotropin-releasing hormone-I-expressing cell populations of different embryological origin in mouse brain. J. Neurosci. 19 (14), 5955–5966. Spergel, D.J., Kruth, U., et al., 1999. GABA- and glutamate-activated channels in green fluorescent protein-tagged gonadotropin-releasing hormone neurons in transgenic mice. J. Neurosci. 19 (6), 2037–2050. Steven, C., Lehnen, N., et al., 2003. Molecular characterization of the GnRH system in zebrafish (Danio rerio): cloning of chicken GnRH-II, adult brain expression patterns and pituitary content of salmon GnRH and chicken GnRH-II. Gen. Comp. Endocrinol. 133 (1), 27–37. Stumm, R., Kolodziej, A., et al., 2007. Patterns of SDF-1alpha and SDF1gamma mRNAs, migration pathways, and phenotypes of CXCR4expressing neurons in the developing rat telencephalon. J. Comp. Neurol. 502 (3), 382–399. Suter, K.J., Song, W.J., et al., 2000. Genetic targeting of green fluorescent protein to gonadotropin-releasing hormone neurons: characterization of whole-cell electrophysiological properties and morphology. Endocrinology 141 (1), 412– 419. Tanaka, D., Nakaya, Y., et al., 2003. Multimodal tangential migration of neocortical GABAergic neurons independent of GPI-anchored proteins. Development 130 (23), 5803–5813. Tobet, S.A., Schwarting, G.A., 2006. Minireview: recent progress in gonadotropinreleasing hormone neuronal migration. Endocrinology 147 (3), 1159–1165. Torgersen, J., 2002. In silico and in situ characterization of the zebrafish (Danio rerio) gnrh3 (sGnRH) gene. BMC Genomics 3 (1), 25. Tran, P.B., Miller, R.J., 2003. Chemokine receptors in the brain: a developing story. J. Comp. Neurol. 457 (1), 1–6. Tsai, P.S., Moenter, S.M., et al., 2005. Targeted expression of a dominant-negative fibroblast growth factor (FGF) receptor in gonadotropin-releasing hormone (GnRH) neurons reduces FGF responsiveness and the size of GnRH neuronal population. Mol. Endocrinol. 19 (1), 225–236. White, R.B., Eisen, J.A., et al., 1998. Second gene for gonadotropin-releasing hormone in humans. Proc. Natl. Acad. Sci. USA 95 (1), 305–309. Whitlock, K.E., 2004. A new model for olfactory placode development. Brain Behav. Evol. 64 (3), 126–140. Whitlock, K.E., Wolf, C.D., et al., 2003. Gonadotropin-releasing hormone (GnRH) cells arise from cranial neural crest and adenohypophyseal regions of the neural plate in the zebrafish, Danio rerio. Dev. Biol. 257 (1), 140–152. Whitlock, K.E. et al., 2005. A role for foxd3 and sox10 in the differentiation of gonadotropin-releasing hormone (GnRH) cells in the zebrafish Danio rerio. Development 132 (24), 5491–5502. Whitlock, K.E., Illing, N., et al., 2006. Development of GnRH cells: setting the stage for puberty. Mol. Cell. Endocrinol. 254–255, 39–50. Wierman, M.E., Pawlowski, J.E., et al., 2004. Molecular mechanisms of gonadotropin-releasing hormone neuronal migration. Trends Endocrinol. Metab. 15 (3), 96–102. Wray, S., 2001. Development of luteinizing hormone releasing hormone neurones. J. Neuroendocrinol. 13 (1), 3–11. Wray, S., 2002. Molecular mechanisms for migration of placodally derived GnRH neurons. Chem. Senses 27 (6), 569–572. Wu, S., Page, L., et al., 2006. A role for GnRH in early brain regionalization and eye development in zebrafish. Mol. Cell. Endocrinol. 257C–258C, 47–64. Xue, H.G., Yamamoto, N., et al., 2006. Projections of the sensory trigeminal nucleus in a percomorph teleost, tilapia (Oreochromis niloticus). J. Comp. Neurol. 495 (3), 279–298. Zenteno, J.C., Mendez, J.P., et al., 1999. Renal abnormalities in patients with Kallmann syndrome. BJU Int. 83 (4), 383–386.