Mash1 Is Required for Generic and Subtype ... - Oxford Journals

0888-8809/06/$15.00/0 Printed in U.S.A.

Molecular Endocrinology 20(7):1623–1632 Copyright © 2006 by The Endocrine Society doi: 10.1210/me.2005-0518

Mash1 Is Required for Generic and Subtype Differentiation of Hypothalamic Neuroendocrine Cells David E. G. McNay, Michelle Pelling, Suzanne Claxton, Franc¸ois Guillemot, and Siew-Lan Ang Divisions of Developmental Neurobiology (D.E.G.M., M.P., S.C., S.-L.A.) and Molecular Neurobiology (F.G.), Medical Research Council, The National Institute of Medical Research, London NW7 1AA, United Kingdom The neuroendocrine hypothalamus regulates a number of critical biological processes and underlies a range of diseases from growth failure to obesity. Although the elucidation of hypothalamic function has progressed well, knowledge of hypothalamic development is poor. In particular, little is known about the processes underlying the neurogenesis and specification of neurons of the ventral nuclei, the arcuate and ventromedial nuclei. The proneural gene Mash1 is expressed throughout the basal retrochiasmatic neuroepithelium and loss of Mash1 results in hypoplasia of both the arcuate and ventromedial nuclei. These defects are due to a failure of neurogenesis and apoptosis, a defect that can be rescued by ectopic Ngn2 under the

control of the Mash1 promoter. In addition to its role in neurogenesis, analysis of Mash1ⴚ/ⴚ, Mash1ⴙ/ⴚ, Mash1KINgn2/KINgn2, and Mash1KINgn2/ⴙ mice demonstrates that Mash1 is specifically required for Gsh1 expression and subsequent GHRH expression, positively regulates SF1 expression, and suppresses both tyrosine hydroxylase (TH) and neuropeptide Y (NPY) expression. Although Mash1 is not required for propiomelanocortin (POMC) expression, it is required for normal development of POMCⴙ neurons. These data demonstrate that Mash1 is both required for the generation of ventral neuroendocrine neurons as well as playing a central role in subtype specification of these neurons. (Molecular Endocrinology 20: 1623–1632, 2006)

A

neurons. The perikarya of the parvocellular neurons are located in the ARN project to the median eminence (ME), a nonneuronal structure forming the floor of the third ventricle, where they release neuropeptide hormones and dopamine regulating anterior pituitary function. Another important hypothalamic system is the feeding circuit, located in the ARN, ventromedial hypothalamic nuclei (VMH), and dorsomedial hypothalamic nuclei, which detects hormones, such as leptin produced by adipose tissue, and regulates appetite and behavior appropriately (6). Briefly, the ARN component consists of neurons which process propiomelanocortin (POMC) into ␣MSH, which inhibits appetite in opposition to neuropeptide Y (NPY) neurons, which stimulate appetite. SF1⫺/⫺ mice, in which the VMH fails to develop, mimic functional disruption of the VMH and are obese (7, 8). The dorsal neuroendocrine neurons have been the subject of a number of investigations that have elucidated the role of a number of transcription factors in their development and specification: Otp, Brn2, Sim1, Sim2, and Anrt2 (9–15). However, these studies have failed to shed much light on the development of ventral neuroendocrine neurons. At embryonic d 10.5 (E10.5), the transcription factor Otp is expressed in some postmitotic cells of the ARN and is required the expression of SS within the ARN (9). From E10.5 onward, SF1 is expressed in the retrochiasmatic hypothalamus (RCH) and is required for the condensation of VMH neurons into a nucleus rather than the specification of VMH

LTHOUGH THE KNOWLEDGE of curative therapeutic interventions and the subsequent clinical outcomes for endocrinopathies of the hypothalamus have improved greatly in recent years, our ability to detect presymptomatic endocrinopathies allowing preventative measures is poor. Similarly, whereas the field of pituitary development has grown rapidly, the field of hypothalamic development is still poorly understood. This is an unfortunate state of affairs given that mutations in known genes are relatively rare and the genetic etiology of most endocrinopathies of the hypothalamic-pituitary axis remains unknown (Refs. 1–3; and McNay, D. E. G., and M. T. Dattani, unpublished data). The rodent hypothalamus is constructed of numerous neuronal subtypes arranged in discrete nuclei (4, 5). The neuroendocrine neurons located in the arcuate nucleus (ARN) include GHRH neurons, somatostatin (SS), and dopaminergic [tyrosine hydroxylase (TH)] First Published Online February 9, 2006 Abbreviations: ARN, Arcuate nucleus; bHLH, basic helixloop-helix; BrdU, bromodeoxyuridine; E10.5, embryonic d 10.5; ME, median eminence; NPY, neuropeptide Y; POMC, propiomelanocortin; PP, posterior pituitary; RCH, retrochiasmatic hypothalamus; TH, tyrosine hydroxylase; TUNEL, terminal deoxynucleotidyl transferase-mediated biotinylated uridine triphosphate nick end labeling; TX-100, Triton X-100; VMH, ventromedial hypothalamic nucleus; WT, wild type. Molecular Endocrinology is published monthly by The Endocrine Society (http://www.endo-society.org), the foremost professional society serving the endocrine community.

1623

1624 Mol Endocrinol, July 2006, 20(7):1623–1632

neurons (16). Gsh1, which is widely expressed in the developing hypothalamus, is required for the expression of GHRH in the ARN (17), although this may be due to a requirement of Gsh1 for ARN expression of the GHRH gene rather than the specification of the GHRH lineage (18). Hmx2/Hmx3 double-mutant mice fail to express Gsh1 at E18.5 and thus fail to express GHRH in the ARN (19). We have undertaken an investigation of the role of proneural basic helix-loop-helix (bHLH) transcription factor Mash1 in the development of basal RCH because this gene has been shown to function both in general neurogenesis and neuronal subtype specification (20). In addition to its neuronal role (21, 22), Mash1 is expressed in and required for the development of endocrine cells of the lung (23), adrenal (24), and thyroid (25). Mash1 is expressed throughout the basal RCH neuroepithelium and the complete loss of Mash1 results in a failure of neurogenesis and apoptosis of both the ARN and the VMH with a subsequent reduction in all neuronal lineages. Loss of a single copy of Mash1 (21) results in an expansion of both TH⫹ and NPY⫹ neurons and a reduction in SF1⫹ neurons, whereas POMC⫹ and GHRH⫹ neurons are unaffected. However, not all lineages of the ARN are equally affected, with TH⫹ and NPY⫹ neurons less severely affected than GHRH⫹ and POMC⫹ neurons, indicating that, in addition to a reduction in the pool of ARN neurons, the proportion of each neuronal lineage within this reduced pool is modified in a manner resembling that resulting from the loss of a single copy of Mash1. In addition, ectopic Ngn2 under the control of the Mash1

McNay et al. • Role of Mash1 in the Hypothalamus

promoter (20) was used to rescue general neurogenesis, and elucidate the specific functions of Mash1. Although ectopic Ngn2 rescues neurogenesis, it is unable to restore the normal differentiation of the ARN and VMH. This analysis demonstrates that Mash1, although required for the normal development of all lineages, is only absolutely required for the specification of GHRH⫹ neurons.

RESULTS AND DISCUSSION The Architecture of the Ventral Hypothalamus Is Identifiable at E10.5 At E10.5, the posterior pituitary (PP) is identifiable as a pouch derived from and connected to the hypothalamic midline (26). At this stage, POMC⫹ neurons (identified using an antibody raised against ␥MSH) form a V-shaped region within the Nkx2.1⫹ RCH basal plate such that they occupy the midline rostral to the Fgf8⫹ PP and are located dorsal to the PP at their caudal limit, marking the location of the ARN (27) (Fig. 1, B–E). Between the Shh⫹ basal plate and the PP, there exists a nonneurogenic region, as indicated by the absence of Dll1 expression, occupying the location of the ME at later stages (Fig. 1, F and G). Dorsal to the POMC⫹ neurons, the VMH can be detected by the VMH marker SF1 (28) (Fig. 1C). However, double labeling of POMC and SF1 indicates that SF1 is also expressed in the POMC lineage. Thus the PP, ME, ARN, and VMH are identifiable at E10.5.

Fig. 1. Mash1 Is Expressed in Progenitors Adjacent to the Nascent Arcuate and Ventromedial Nuclei A, Diagram showing location of E10.5 coronal sections (adapted from Ref. 45). B, Flat-mounted E10.5 head viewed with the neuroepithelium uppermost, stained with anti-␥MSH showing the location of the POMC neurons within the RCH. C and D, Both SF1⫹ VMH neurons and POMC ARN neurons are located within the RCH Nkx2.1⫹ E10.5 basal plate. E–G, Between the Fgf8⫹ PP and the Shh⫹ neurogenic basal plate lies a nonneurogenic region corresponding to the location of ME. H–K, Mash1 is expressed throughout the basal plate and ventral alar plate; Ngn1 and Ngn2 are limited to more dorsal regions, whereas Ngn3 is expressed in a smaller domain within the nascent ARN/VMH. L, Expression of the bHLH proneural genes remains similar over the E10.5–E12.5 period. Bar, 100 ␮m.


bHLH Proneural Gene Expression in the Hypothalamic Neuroepithelium Mash1 is expressed throughout the neuroepithelium of the RCH basal plate as well as the ventral alar plate (Fig. 1H). Neither Ngn1 nor Ngn2 is expressed within the RCH basal plate, being limited to more dorsal neuroepithelium (Fig. 1, I and J). Ngn3, on the other hand, is expressed within a small region of the RCH basal plate in the region of ARN/VMH (Fig. 1K). These four genes are expressed in a similar pattern at E12.5 (Figs. 1L and 2I; data not shown). Given that Mash1 represents the major bHLH proneural gene expressed in the RCH basal plate, we investigated the role of Mash1 in the development of neuronal lineages of the ARN and VMH. Mash1 Regulates Neurogenesis within the Ventral RCH As is the case in the ventral telencephalon (22), loss of Mash1 leads to failure of neurogenesis in the ventral diencephalon, indicated by the loss of SCG10⫹ neurons in the RCH in Mash1⫺/⫺ embryos (Fig. 2, D and D⬘). This loss of SCG10 expression occurs alongside a loss of the bHLH transcription factors NeuroD, Nhlh2, and Ngn3 (Fig. 2, E–F⬘, H, and H⬘). Notch-mediated lateral inhibition is largely absent in the ventral RCH, as indicated by the uniform expression of the Mash1 locus in Mash1⫺/⫺ mutants compared with the speckled pattern in wild-type (WT) mice and the absence of Dll1 and Hes5 expression (Fig. 2, A–C⬘). The failure of neurogenesis does not lead to an increase in the proliferating progenitor pool. A 30-min BrdU pulse labels 29 ⫾ 5% (n ⫽ 3) of all cells within the nascent ARN/ VMH and adjacent neuroepithelium in WT embryos compared with 33 ⫾ 2% (n ⫽ 3, P ⫽ 0.3) in Mash1⫺/⫺. However, terminal deoxynucleotidyl transferase-mediated biotinylated uridine triphosphate nick end labeling (TUNEL) staining (Fig. 2, G and G⬘) indicates that there is an increase in apoptosis in Mash1⫺/⫺ (1438 ⫾ 571, n ⫽ 3) compared with Mash1⫹/⫹ (224 ⫾ 97, n ⫽ 3, P ⫽ 0.01). At E10.5, neurogenesis is not completely absent, because a small domain of neurogenesis is variably apparent at the dorsal margin of the RCH coinciding with Ngn3 expression (Fig. 2, C⬘, F⬘, and H⬘). In addition, Ngn3 expression recovers in Mash1⫺/⫺ embryos by E12.5, coinciding with the generation of a substantial population of VMH neurons (Fig. 2, I and I⬘). These results suggest that Ngn3 is able to facilitate neurogenesis in at least a subset of RCH neurons. Further investigation into the specific role played by Ngn3 is currently being undertaken. The failure of neurogenesis was confirmed by the use of mice carrying a knock-in of Ngn2 into the Mash1 locus (Mash1KINgn2/KINgn2). Ngn2 rescues neurogenesis as indicated by the presence of SCG10⫹ neurons in the region of the nascent ARN similar to that seen in WT embryos (Fig. 2D⬙). These neurons

Mol Endocrinol, July 2006, 20(7):1623–1632 1625

express NeuroD, Nhlh2, and Ngn3 normally (Fig. 2, E⬙, F⬙, and H⬙). In addition, Ngn2 is able to reduce the level of apoptosis in the ventral RCH as indicated by TUNEL staining (454 ⫾ 263, n ⫽ 3, P ⫽ 0.04) (Fig. 2G⬙). Mash1 Is Required for the Differentiation of GHRHⴙ Neurons Although GHRH is not expressed at E12.5 (data not shown), the three transcription factors known to regulate GHRH expression, Gsh1, Hmx2, and Hmx3 (17, 19), are widely expressed in the hypothalamus (Fig. 3, A–C). Mash1⫺/⫺ embryos fail to express Gsh1 within the RCH basal plate (Fig. 3A⬘), although expression in both the zona limitans intrathalamica and the RCH alar plate is unchanged (supplemental Fig. 1, published as supplemental data on The Endocrine Society’s Journals Online web site at http://mend.endojournals.org). On the other hand, expression of Hmx2 and Hmx3 is not extinguished in the RCH basal plate of Mash1⫺/⫺ embryos, indicating that in general the loss of Gsh1 is independent of Hmx2 and Hmx3 (Fig. 3, B⬘ and C⬘). Neither endogenous Ngn3 (from E12.5 onwards) nor ectopic Ngn2 (Fig. 3, A⬘ and A⬙) is able to rescue Gsh1 expression, indicating that the expression Gsh1 requires Mash1. These results indicate that the relationship between Mash1 and Gsh1 in the hypothalamus is similar to that of Mash1 and Gsh2 in the chick spinal cord where Mash1 activates Gsh2 (29). In particular, the maintenance of Gsh1 expression in the Mash1⫹/Shh⫺ RCH alar plate in concert with the loss of Gsh1 expression in the Mash1⫹/Shh⫹ basal plate indicates that Mash1 is specifically required for Gsh1 expression only in the region corresponding to the ARN and VMH. As expected, loss of Gsh1 expression is correlated with an absence of GHRH neurons in the ARN of Mash1⫺/⫺ embryos at E17.5 (Fig. 4, A and C). No difference in GHRH expression in Mash1⫹/⫺ embryos is apparent (n ⫽ 3, P ⫽ 0.54) (Fig. 4B). Mash1KINgn2/KINgn2 embryos fail to express GHRH in the ARN, whereas Mash1KINgn2/⫹ embryos express GHRH normally (n ⫽ 3, P ⫽ 0.14) (Fig. 4, D and E). Given that TH⫹, POMC⫹, and NPY⫹ neurons are present (see below), whereas no GHRH⫹ neurons are detectable in the absence of Mash1, these data indicate that Mash1 is absolutely required for GHRH expression in the ARN, whereas Ngn2 does not affect GHRH neuropeptide production (Fig. 5A). The requirement for Mash1 to facilitate Gsh1 expression in the ARH and VMH suggests that the loss of GHRH neurons is due to a failure of Gsh1 expression. Although Mash1 Is Required for the Neurogenesis of A12 THⴙ Neurons, Heterozygous Loss of Mash1 Leads to an Expansion of TH Expression At E12.5, only one or two TH⫹ neurons can be found in the ventral RCH region of Mash1⫺/⫺ embryos compared with a substantial population in WT embryos and Ngn2 is unable to rescue this expression in



Fig. 2. Block in Neurogenesis in Mash1⫺/⫺ Embryos Is Rescued by Ectopic Ngn2 Expressed in the RCH of Mash1KINgn2/KINgn2 and Endogenous Ngn3 at E12.5 A–H⬙, Coronal sections through the RCH at E10.5. I–I⬙, Coronal sections through the RCH at E12.5. A–C⬘, The salt-and-pepper pattern of expression of Dll1 and Hes5 is largely missing in Mash1⫺/⫺ embryos, whereas Mash1 appears more uniform in mutant compared with WT embryos. D–F⬙, Loss of postmitotic neurons expressing NeuroD, SCG10, and Nhlh2 in Mash1 mutants is recovered in Mash1KINgn2/KINgn2 embryos. H–I⬙, At E10.5, Ngn3 expression in the RCH basal plate is lost in Mash1⫺/⫺ mutants compared with WT, but is rescued by Ngn2. Ngn3 expression recovers in Mash1⫺/⫺ mutants by E12.5. G–G⬙, Loss of Mash1 results in an increase in apoptosis within the nascent ARN/VMH, which is rescued by ectopic Ngn2. Arrowheads indicate the region of neurogenesis coinciding with Ngn3 expression. Bar, 100 ␮m; ⌬, Mash1-null allele; KI, Mash1KINgn2 allele.

Mash1KINgn2/KINgn2 embryos (Fig. 3D–D⬙). By E17.5, TH⫹ neurons are significantly reduced in Mash1⫺/⫺ embryos with an average of 1234 ⫾ 202 (n ⫽ 3) TH⫹

neurons at E17.5 compared with an average of 5504 ⫾ 67 (n ⫽ 3, P ⬍ 0.01) in WT mice (Fig. 4, F and H). The remaining TH⫹ neurons are restricted to a thin layer at



Fig. 3. Neurogenesis Recovers in Mash1⫺/⫺ Embryos by E12.5 Giving Rise to Mainly VMH Neurons A–D⬙ and F–F⬙, Coronal sections through the RCH at E12.5. E–E⬙, Coronal sections through the RCH at E10.5. A–F⬙, The neurons generated by E12.5 consist mainly of SF1⫹ VMH neurons with few neurons expressing arcuate nucleus neuronal markers: TH, POMC, or Gsh1 in Mash1⫺/⫺ embryos. A–C⬙, Expression of Gsh1 requires Mash1, and this loss of Gsh1 does not occur via a loss of Hmx2/Hmx3. D–D⬙, Ectopic Ngn2 is unable to rescue expression of TH. E–F⬙, There is a severe reduction of SF1⫹/POMC⫺ and SF1⫹/POMC⫹ neurons in both Mash1⫺/⫺ and Mash1KINgn2/KINgn2 embryos at E10.5. By E12.5, expression of SF1 recovers in Mash1⫺/⫺, but POMC expression does not. On the other hand, POMC recovers in Mash1KINgn2/KINgn2, but SF1 expression does not. Green arrowheads indicate SF1⫹/POMC⫹ neurons, red arrowheads indicate SF1⫹/POMC⫺ neurons, and white arrowheads indicate ␥MSH⫹ fibers. Bar, 100 ␮m; ⌬, Mash1-null allele; KI, Mash1KINgn2 allele.

the base of the diencephalon. Mash1⫹/⫺ embryos carrying one WT copy of Mash1 have significantly more 8299 ⫾ 1412 (P ⫽ 0.018, n ⫽ 3) TH⫹ neurons than WT embryos. The ectopic TH expression in Mash1⫹/⫺ occurs in neurons located in the medial-ventral portion of the arcuate nucleus (Fig. 4G). Ngn2 fails to rescue TH⫹ expression with an average of 57 ⫾ 49 (n ⫽ 3, P ⬍ 0.01) A12 TH⫹ neurons in Mash1KINgn2/KINgn2 embryos (Fig. 4I). Indeed, this represents a significant (P ⬍ 0.01) reduction in the number of TH⫹ neurons beyond that resulting from the loss of Mash1 alone. Mash1 KINgn2/⫹ embryos expressing both Mash1 and Ngn2 have significantly fewer (3690 ⫾ 1109, n ⫽ 3, P ⫽ 0.03) TH⫹ neurons than WT embryos (Fig. 4J). Thus, whereas Mash1 is required for neurogenesis of ARN neurons, including the A12 dopaminergic neurons, Mash1 (and Ngn2) suppresses TH expression (Fig. 5B). Given that the majority of GHRH⫹ neurons are TH⫹ in adults, these data also suggest that, in addition to GHRH and TH expression being independently regulated during the postnatal period (6), GHRH and TH expression are regulated by separate pathways during development. Although Mash1 Is Required for the Neurogenesis of ARN NPYⴙ Neurons, Heterozygous Loss of Mash1 Leads to an Expansion of NPY Expression NPY⫹ neurons are detectable in the ARN of WT embryos at E17.5 (3488 ⫾ 174, n ⫽ 3) (Fig. 4K). Loss of Mash1 results in a severe reduction of NPY⫹ neurons

compared with WT (368 ⫾ 202, n ⫽ 3, P ⬍ 0.01) (Fig. 4M). Similar to TH⫹ neurons, NPY⫹ neurons are expanded in Mash1⫹/⫺ compared with Mash1⫹/⫹ (4126 ⫾ 161, n ⫽ 3, P ⬍ 0.01) (Fig. 4L). However, the increase of NPY⫹ neurons (18%) is not as large as the increase in TH⫹ neurons (51%). The rescue of neurogenesis by Ngn2 in Mash1KINgn2/KINgn2 leads to only a modest increase in the number of NPY⫹ neurons compared with the absence of Mash1 alone (1150 ⫾ 160, n ⫽ 3, P ⬍ 0.01) and the number of NPY⫹ neurons remains significantly lower than that in the WT (P ⬍ 0.01) (Fig. 4N). The number of NPY⫹ neurons in Mash1KINgn2/⫹ (2614 ⫾ 365, n ⫽ 3) (Fig. 4O) is lower than that in the Mash1⫹/⫹ (P ⫽ 0.01), indicating that Ngn2 suppresses NPY. Thus, whereas Mash1 is required for neurogenesis of ARN neurons, including the NPY⫹ neurons, Mash1 (and Ngn2) suppresses NPY expression (Fig. 5C). Mash1 Is Required in a Dose-Dependent Manner for the Development of the SF1ⴙ Neurons of the Ventromedial Nucleus SF1⫹ neurons are largely absent within the RCH basal plate of Mash1⫺/⫺ embryos at E10.5 (Fig. 3, E and E⬘). Expression of SF1 recovers at E12.5 alongside the recovery of Ngn3 expression (Fig. 3, F and F⬘). These results correlate with the failure of neurogenesis of both the ARN and VMH in Mash1⫺/⫺ embryos at E10.5 with a minor recovery of neurogenesis in the VMH by E12.5. Although Ngn2 is able to rescue neurogenesis at E10.5, SF1 expression is not rescued (Fig. 2E⬙).

1628

Mol Endocrinol, July 2006, 20(7):1623–1632


Fig. 4. Mash1 Is Required for the Neurogenesis of All Neuronal Subtypes of the ARN as Well as the Specification of Ghrh Neurons A–E, Ghrh expression requires Mash1 and is unaffected by Ngn2. F–J, TH expression is repressed by both Mash1 and Ngn2. K–O, NPY expression is repressed by both Mash1 and Ngn2. P–T, POMC expression does not require Mash1. U–Y, SF1⫹ VMH neurons show a dose-dependent requirement for Mash1. Bar, 100 ␮m; ⌬, Mash1-null allele; KI, Mash1KINgn2 allele.

Furthermore, ectopic Ngn2 suppresses SF1 expression within the early VMH, because SF1 is not detectable at E12.5 in Mash1KINgn2/KINgn2 embryos (Fig. 3F⬙). Indeed, no SF1 expression occurs at this stage in Mash1KINgn2/⫹ embryos, although expression does occur at E10.5 in presumptive POMC⫹ neurons (supplemental Fig. 2). The VMH in Mash1⫺/⫺ embryos is significantly reduced by E17.5 with 18,344 ⫾ 9,086 (n ⫽3, P ⬍ 0.01) SF1⫹ neurons compared with 78,632 ⫾ 1,847 (n ⫽ 3) SF1⫹ neurons in WT embryos (Fig. 4, U and W). The VMH of Mash1⫹/⫺ embryos (Fig. 4V) is significantly reduced with 63,100 ⫾ 6,633 SF1⫹ neurons (P ⫽ 0.04, n ⫽ 3). Ngn2 is only partially able to rescue the loss of VMH neurons. Fewer SF1⫹ neurons are found in Mash1KINgn2/KINgn2 embryos compared with WT (35,476 ⫾ 5,745, n ⫽ 3, P ⬍ 0.01) (Fig. 4X), although the number of SF1⫹ VMN neurons is greater than Mash1⫺/⫺ embryos (P ⫽ 0.04). Thus, although Ngn2 is able to rescue SF1⫹ VMH neurons, it is a poor mediator of VMH development. Mash1KINgn2/⫹ embryos have a greater number of SF1⫹ VMH neurons at E17.5 than Mash1KINgn2/KINgn2 embryos (59,716 ⫾ 7,232 n ⫽ 3, P ⬍ 0.01) (Fig. 4Y), although the number of SF1⫹ neurons in Mash1KINgn2/⫹ embryos is not different from that in Mash1⫹/⫺ embryos (P ⫽ 0.54). Thus, Mash1 is required for the development of SF1⫹ VMH neurons in a dose-dependent manner, and although Ngn2 is a weak mediator of VMH develop-

ment, it does not interfere with Mash1 activity in this lineage (Fig. 5D). Mash1 Is Not Required for the Specification of POMCⴙ Neurons POMC⫹ neurons are present in the ARN from E10.5 onward. Initially, these neurons do not process POMC into ␣MSH, indicated by a lack of anti-ACTH immunoreactivity (data not shown), but are identifiable by anti-␥MSH immunoreactivity (Fig. 3E). At E12.5, both anti-ACTH and anti-␥MSH colabel cells in the RCH and pituitary (Fig. 3F, inset). At E10.5, all POMC⫹ neurons express SF1 and Mash1⫺/⫺ embryos fail to express either POMC or SF1 except for a few scattered cells (Fig. 3, E and E⬘). By E12.5, expression of POMC is not recovered with 60 ⫾ 118 (n ⫽ 3, P ⬍ 0.01) POMC⫹ neurons in Mash1⫺/⫺ embryos compared with 3278 ⫾ 378 (n ⫽ 3) POMC⫹ neurons in WT embryos (Fig. 3, F and F⬘). The number of POMC⫹ neurons is greatly reduced at E17.5 in Mash1⫺/⫺ embryos (Fig. 4, P and R). Ectopic Ngn2 is unable to rescue POMC⫹ neurons at E10.5 with only sparse POMC/SF1 expression in Mash1KINgn2/KINgn2 embryos (Fig. 3E⬙). By E12.5, POMC expression recovers in Mash1KINgn2/KINgn2 embryos, albeit the number of POMC⫹ neurons (2108 ⫾ 574 n ⫽ 3 P ⫽ 0.02) is significantly less than WT mice and these POMC⫹ neurons fail to express SF1 (Fig. 3F⬙). By E17.5, POMC



Fig. 5. Comparison of Neuroendocrine Populations in Mutant Strains and Total Body Mass of Mash1KINgn2/⫹ A, Mash1 is absolutely required for GHRH expression but not in a dose-dependent manner. B, Although both Mash1 and Ngn2 suppress TH expression, Mash1 is not as potent as Ngn2. C, Although both Mash1 and Ngn2 suppress NPY expression, Mash1 is not as potent as Ngn2. D, Mash1 is required in a dose-dependent manner for differentiation of the SF1⫹ neurons of the VMH. E, Mash1 is not required for specification of POMC neurons. However, the early requirement of Mash1 for POMC expression results in an initial reduction in the number of POMC⫹ neurons in Mash1KINgn2/KINgn2 compared with WT embryos. F, Mash1KINgn2/⫹ mice display a transient reduction in total body mass at 6 wk (*, significant difference compared with WT; **, significant difference compared with both WT and Mash1⫺/⫺; bar, 100 ␮m).

expression is expanded in Mash1KINgn2/KINgn2 embryos (Fig. 5, F and H). This is not due to ectopic NeuroD expression stimulating the bHLH response element in the pituitary-specific POMC enhancer (Ref. 30 and data not shown). The late expression of POMC in Mash1KINgn2/KINgn2 embryos indicates that the reduced number of POMC⫹ neurons at E12.5 reflects a

failure to express POMC rather than a reduction in POMC neurons. At E12.5, no difference in the number of POMC⫹ neurons is apparent in either Mash1⫹/⫺ (2821 ⫾ 481 n ⫽ 3, P ⫽ 0.18) or Mash1KINgn2/⫹ embryos (3755 ⫾ 562, n ⫽3, P ⫽ 0.18), although Mash1KINgn2/⫹ embryos fail to express SF1 within the POMC⫹ neurons at


this stage (supplemental Fig. 2). At E10.5 Mash1KINgn2/⫹ embryos express SF1 ectopically within the neuroepithelium as well as in POMC⫹ neurons (supplemental Fig. 2). Together, these results identify two specific phases of POMC expression. Initially, POMC expression is dependent on Mash1 and correlates with SF1 expression (E10.5) and subsequently POMC expression becomes largely independent of either Mash1 or SF1 (E12.5) in agreement with the presence of two independently functioning POMC neural enhancers, only one of which (nPE2) contains conserved putative AD4 (SF1) and Nkx2.1 binding sites (Refs. 31–33 and data not shown). These data indicate that, whereas Mash1 is required for neurogenesis of ARN neurons, Mash1 is not required for the specification of POMC⫹ neurons (Fig. 5E), although early POMC expression requires Mash1. Although Mash1 is expressed in the pituitary, POMC expression this region appears unchanged in either Mash1⫺/⫺ or Mash1KINgn2/KINgn2 embryos, although detailed analysis was not carried out (data not shown). Total Body Mass Is Transiently Reduced in Mash1KINgn2/ⴙ Given the effects of ectopic Ngn2 on the ventral hypothalamus, in particular the reduction in NPY⫹ neurons alongside an expansion in POMC⫹ neurons, it is surprising that no growth phenotype has been described in Mash1KINgn2/⫹ mice (20). On closer inspection (Fig. 5F), a small reduction in total body mass is seen at 6 wk after birth (Mash1⫹/⫹, 23.8 ⫾ 0.7 g, n ⫽ 46; vs. Mash1KINgn2/⫹, 22.6 ⫾ 1.2 g, n ⫽ 29, P ⫽ 0.02), although no difference was seen at weaning (3 wk, P ⫽ 0.97), 9 wk (P ⫽ 0.38), or 12 wk (0.53). Given Mash1 expression in the thyroid, pituitary, adrenal, lung, and central nervous system, it is impossible to identify the cause underlying this phenotype without further investigation.


The hypoplasia of the ARN and VMH in Mash1⫺/⫺ embryos (as demonstrated by the loss of the neuronal marker SCG10, notch signaling, and lateral inhibition) indicates that Mash1 regulates the size of neuronal pool via its role during neurogenesis. 2) Mash1 also regulates the specification of individual subtypes within this neuronal pool by two processes: via a nondose-dependent route as indicated by the absolute requirement for Mash1 in the expression of Gsh1 and generation of GHRH⫹ neurons, as well as a dosedependent route as shown by the expansion of both TH⫹ and NPY⫹ neurons alongside the reduction of SF1⫹ neurons in Mash1⫹/⫺ embryos. Hence, Mash1dependent regulation of the absolute number of a particular neuronal subtype would be a consequence of these two functions. For example, although the proportion of ARN neurons adopting a dopaminergic phenotype increases due to Mash1-mediated repression in a dose-dependent manner: 41% in Mash1⫹/⫹, 50% in Mash1⫹/⫺, and 74% in Mash1⫺/⫺ embryos (number of TH⫹ ARN neurons/total ARN neurons counted), the absolute number of dopaminergic ARN neurons decreases in Mash1⫺/⫺ embryos due to the overall reduction in the total number of ARN neurons.

MATERIALS AND METHODS Generation and Genotyping of Mutant Embryos and Animals Mash1⫺/⫺ mice were generated and genotyped as previously described (21). Mice carrying Ngn2 knocked into the Mash1 locus were generated and genotyped as previously described (20). E0.5 was considered to be noon of the morning in which a vaginal plug was detected. At all times, animals were handled in compliance with the Animal (Scientific Procedures) Act 1986.

Conclusions

Whole-Mount Immunohistochemistry, in Situ Hybridization of Sections, and Immunohistochemistry of Sections

Although the bHLH proneural genes play a critical role in general neurogenesis, they are involved in the development of the endocrine system. It is therefore not surprising that the major bHLH proneural gene expressed within the developing ventral hypothalamus, Mash1, is required for the development of the ventral hypothalamic neuroendocrine system and may play a role in diseases resulting from dysfunction of these regions such as growth failure or obesity. The analysis of Mash1⫹/⫺ embryos and the rescue of neurogenesis in Mash1KINgn2/KINgn2 embryos allow the separation of the general neurogenesis function of Mash1 and the elucidation of Mash1 functions in the development of specific neuroendocrine subtypes of the ventral hypothalamus. The role of Mash1 in regulating the absolute number of each neuronal subtype within the ventral hypothalamus is complex. The data are consistent with two distinct functions of Mash1. 1)

Embryos or dissected brains were fixed for 30 min (E10.5– E12.5 immunohistochemistry), 1 h (E17.5 immunohistochemistry), or overnight (in situ hybridization) at 4 C in 4% paraformaldehyde in PBS and either stored in methanol at ⫺20 C (whole-mount) or cryoprotected with 30% sucrose in PBS, embedded in OCT compound (VWR International, Poole, UK), and cryosectioned on a cryostat (CM3050S; Leica Microsystems, GmbH, Wetzlar, Germany). Section and wholemount in situ hybridization were performed as described previously (34, 35). The following antisense RNA probes have been used: Shh (36), TH (37), Dll1 (38), Mash1 (39), GHRH (40), POMC (41), Hes5 (22), Ngn3 (42), Otp (9), Ngn1 (43), Ngn2 (43), NPY (44), Nhlh2 (41), Gsh1 (17), Hmx2 (19), and Hmx3 (19). For each probe, a minimum of three control and three mutant embryos were analyzed. For immunohistochemistry, sections were incubated overnight at 4 C with the appropriate primary antibody diluted in 0.1% Tween 20 and 1% BSA in PBS. Sections were then extensively washed in PBS plus 0.1% Tween 20 and incubated 2 h at room temperature with a secondary antibody conjugated with a fluorochrome and if required TOTO-3io-


dide (Molecular Probes, Eugene, OR). Sections were then washed and mounted in Vectashield H-1000 (Vector Laboratories, Burlingame, CA). The following primary antibodies were used: rat anti-bromodeoxyuridine (BrdU) (OBT0030S, 1:10; Oxford Biotechnology, Kidlington, UK), rabbit anti-TH (AB152, 1:200; Chemicon, Temecula, CA), mouse anti-Ngn2 (1:5 Lo 2002), rabbit anti-SF1 (1:1000 Hatano 1994; kind gift from K. Morohashi, National Institute for Basic Biology, Okazaki, Japan), rabbit anti-NPY (T-4070, 1:100; Peninsula Laboratories, Torrance, CA), mouse anti-ACTH (CR1096M 1:400; Cortex Biochem, San Leandro, CA), and sheep anti-␥1MSH (1:100; kind gift from A. Bicknell, University of Reading, Reading, UK). TUNEL Cryostat sections were washed once for 5 min in PBS-0.1% Triton X-100 (TX-100), permeabilized in ice-cold 0.01 M citrate buffer and 0.1% TX-100 for 2 min, and washed again in PBS-0.1% TX-100. The enzymatic reaction was then performed at 37 C according to the protocol of the manufacturer (1 684 795; Roche Diagnostics, Mannheim, Germany). BrdU Labeling Pregnant females were injected ip with a solution of BrdU (B-5002, at 10 mg/ml in physiological serum; Sigma-Aldrich, St. Louis, MO) at 100 mg for 1 g of body weight and killed after 30 min. BrdU⫹ cells were revealed by immunohistochemistry on frozen sections. Image Processing, Cell Counting, and Statistics All images were collected on a Zeiss (Oberkochen, Germany) LSM510 microscope or Leica TCS SP2 confocal microscope and processed with Adobe Photoshop 7.0 software (Adobe Systems, San Jose, CA). All cell counts are derived from a single series per animal (counting all requisite cells within the RCH, ARN, or VMH as required) multiplied by the number of series with no correction for serial reconstruction. Three series, each from a separate animal, were counted for each genotype/marker combination. Two-way Student’s t tests were used to determine significance between groups, both for cell counts as well as total body mass measurements.

Acknowledgments We thank Debra Good, Scott Young, and Tom Lufkin for providing the Nhlh2, POMC, and Hmx2 and Hmx3 probes, respectively. In addition, we thank Andrew Bicknell for donating the anti-␥1MSH antibody and Ken Morohashi for donating the anti-SF1 (Ad4BP) antibody. We thank Iain Robinson for his thought-provoking conversation on the hypothalamus.

Received December 16, 2005. Accepted February 2, 2006. Address all correspondence and requests for reprints to: Siew-Lan Ang, Division of Developmental Neurobiology, Medical Research Council, The National Institute of Medical Research, Mill Hill, London NW7 1AA, United Kingdom. Email: [email protected]. This work was supported by the Medical Research Council. D.E.G.M., M.P., S.C., F.G., and S.-L.A. have nothing to declare.


REFERENCES 1. Dattani MT 2005 Growth hormone deficiency and combined pituitary hormone deficiency: does the genotype matter? Clin Endocrinol (Oxf) 63:121–130 2. Williams R, Larson R, Kronenberg H, Melmed S, Polonsky K, eds. 2003 Textbook of endocrinology. Philadelphia: WB Saunders 3. Brook C, Clayton P, Brown R, eds. 2005 Clinical pediatric endocrinology. 5th ed. Oxford: Blackwell Science 4. Swanson LW 1986 Organization of mammalian neuroendocrine systems. In: Bloom FE, ed. Handbook of physiology. Baltimore: Waverly; 317–363 5. Sawchenko PE, Imaki T, Vale W 1992 Co-localization of neuroactive substances in the endocrine hypothalamus. Ciba Found Symp 168:16–30; discussion 30–42 6. Phelps CJ, Romero MI, Hurley DL 2003 Growth hormone-releasing hormone-producing and dopaminergic neurones in the mouse arcuate nucleus are independently regulated populations. J Neuroendocrinol 15: 280–288 7. Choi S, Dallman MF 1999 Hypothalamic obesity: multiple routes mediated by loss of function in medial cell groups. Endocrinology 140:4081–4088 8. Majdic G, Young M, Gomez-Sanchez E, Anderson P, Szczepaniak LS, Dobbins RL, McGarry JD, Parker KL 2002 Knockout mice lacking steroidogenic factor 1 are a novel genetic model of hypothalamic obesity. Endocrinology 143:607–614 9. Acampora D, Postiglione MP, Avantaggiato V, Di Bonito M, Vaccarino FM, Michaud J, Simeone A 1999 Progressive impairment of developing neuroendocrine cell lineages in the hypothalamus of mice lacking the Orthopedia gene. Genes Dev 13:2787–2800 10. Wang W, Lufkin T 2000 The murine Otp homeobox gene plays an essential role in the specification of neuronal cell lineages in the developing hypothalamus. Dev Biol 227: 432–449 11. Schonemann MD, Ryan AK, McEvilly RJ, O’Connell SM, Arias CA, Kalla KA, Li P, Sawchenko PE, Rosenfeld MG 1995 Development and survival of the endocrine hypothalamus and posterior pituitary gland requires the neuronal POU domain factor Brn-2. Genes Dev 9:3122–3135 12. Nakai S, Kawano H, Yudate T, Nishi M, Kuno J, Nagata A, Jishage K, Hamada H, Fujii H, Kawamura K 1995 The POU domain transcription factor Brn-2 is required for the determination of specific neuronal lineages in the hypothalamus of the mouse. Genes Dev 9:3109–3121 13. Michaud JL, Rosenquist T, May NR, Fan CM 1998 Development of neuroendocrine lineages requires the bHLH-PAS transcription factor SIM1. Genes Dev 12: 3264–3275 14. Goshu E, Jin H, Lovejoy J, Marion JF, Michaud JL, Fan CM 2004 Sim2 contributes to neuroendocrine hormone gene expression in the anterior hypothalamus. Mol Endocrinol 18:1251–1262 15. Hosoya T, Oda Y, Takahashi S, Morita M, Kawauchi S, Ema M, Yamamoto M, Fujii-Kuriyama Y 2001 Defective development of secretory neurones in the hypothalamus of Arnt2-knockout mice. Genes Cells 6:361–374 16. Davis AM, Seney ML, Stallings NR, Zhao L, Parker KL, Tobet SA 2004 Loss of steroidogenic factor 1 alters cellular topography in the mouse ventromedial nucleus of the hypothalamus. J Neurobiol 60:424–436 17. Li H, Zeitler PS, Valerius MT, Small K, Potter SS 1996 Gsh-1, an orphan Hox gene, is required for normal pituitary development. EMBO J 15:714–724 18. Mutsuga N, Iwasaki Y, Morishita M, Nomura A, Yamamori E, Yoshida M, Asai M, Ozaki N, Kambe F, Seo H, Oiso Y, Saito H 2001 Homeobox protein Gsh-1-dependent regulation of the rat GHRH gene promoter. Mol Endocrinol 15:2149–2156


19. Wang W, Grimmer JF, Van De Water TR, Lufkin T 2004 Hmx2 and Hmx3 homeobox genes direct development of the murine inner ear and hypothalamus and can be functionally replaced by Drosophila Hmx. Dev Cell 7:439–453 20. Parras CM, Schuurmans C, Scardigli R, Kim J, Anderson DJ, Guillemot F 2002 Divergent functions of the proneural genes Mash1 and Ngn2 in the specification of neuronal subtype identity. Genes Dev 16:324–338 21. Guillemot F 1995 Analysis of the role of basic-helix-loophelix transcription factors in the development of neural lineages in the mouse. Biol Cell 84:3–6 22. Casarosa S, Fode C, Guillemot F 1999 Mash1 regulates neurogenesis in the ventral telencephalon. Development 126:525–534 23. Borges M, Linnoila RI, van de Velde HJ, Chen H, Nelkin BD, Mabry M, Baylin SB, Ball DW 1997 An achaete-scute homologue essential for neuroendocrine differentiation in the lung. Nature 386:852–855 24. Huber K, Bruhl B, Guillemot F, Olson EN, Ernsberger U, Unsicker K 2002 Development of chromaffin cells depends on MASH1 function. Development 129:4729–4738 25. Lanigan TM, DeRaad SK, Russo AF 1998 Requirement of the MASH-1 transcription factor for neuroendocrine differentiation of thyroid C cells. J Neurobiol 34:126–134 26. Kerr T 1946 The development of the pituitary of the laboratory mouse. Q J Microsc Sci 87:3–29 27. Elkabes S, Loh YP, Nieburgs A, Wray S 1989 Prenatal ontogenesis of pro-opiomelanocortin in the mouse central nervous system and pituitary gland: an in situ hybridization and immunocytochemical study. Brain Res Dev Brain Res 46:85–95 28. Shinoda K, Lei H, Yoshii H, Nomura M, Nagano M, Shiba H, Sasaki H, Osawa Y, Ninomiya Y, Niwa O, Morohashi K, Li E 1995 Developmental defects of the ventromedial hypothalamic nucleus and pituitary gonadotroph in the Ftz-F1 disrupted mice. Dev Dyn 204:22–29 29. Kriks S, Lanuza GM, Mizuguchi R, Nakafuku M, Goulding M 2005 Gsh2 is required for the repression of Ngn1 and specification of dorsal interneuron fate in the spinal cord. Development 132:2991–3002 30. Poulin G, Turgeon B, Drouin J 1997 NeuroD1/␤2 contributes to cell-specific transcription of the proopiomelanocortin gene. Mol Cell Biol 17:6673–6682 31. de Souza FS, Santangelo AM, Bumaschny V, Avale ME, Smart JL, Low MJ, Rubinstein M 2005 Identification of neuronal enhancers of the proopiomelanocortin gene by transgenic mouse analysis and phylogenetic footprinting. Mol Cell Biol 25:3076–3086 32. Morohashi K, Honda S, Inomata Y, Handa H, Omura T 1992 A common trans-acting factor, Ad4-binding protein, to the promoters of steroidogenic P-450s. J Biol Chem 267:17913–17919


33. Fabbro D, Tell G, Pellizzari L, Leonardi A, Pucillo C, Lonigro R, Damante G 1995 Definition of the DNA-binding specificity of TTF-1 homeodomain by chromatographic selection of binding sequences. Biochem Biophys Res Commun 213:781–788 34. Schaeren-Wiemers N, Gerfin-Moser A 1993 A single protocol to detect transcripts of various types and expression levels in neural tissue and cultured cells: in situ hybridization using digoxigenin-labelled cRNA probes. Histochemistry 100:431–440 35. Conlon RA, Herrmann BG 1993 Detection of messenger RNA by in situ hybridization to postimplantation embryo whole mounts. Methods Enzymol 225:373–383 36. Echelard Y, Epstein DJ, St-Jacques B, Shen L, Mohler J, McMahon JA, McMahon AP 1993 Sonic hedgehog, a member of a family of putative signaling molecules, is implicated in the regulation of CNS polarity. Cell 75: 1417–1430 37. Grima B, Lamouroux A, Blanot F, Biguet NF, Mallet J 1985 Complete coding sequence of rat tyrosine hydroxylase mRNA. Proc Natl Acad Sci USA 82:617–621 38. Bettenhausen B, Hrabe de Angelis M, Simon D, Guenet JL, Gossler A 1995 Transient and restricted expression during mouse embryogenesis of Dll1, a murine gene closely related to Drosophila Delta. Development 121: 2407–2418 39. Guillemot F, Joyner AL 1993 Dynamic expression of the murine Achaete-Scute homologue Mash-1 in the developing nervous system. Mech Dev 42:171–185 40. Flavell DM, Wells T, Wells SE, Carmignac DF, Thomas GB, Robinson IC 1996 Dominant dwarfism in transgenic rats by targeting human growth hormone (GH) expression to hypothalamic GH-releasing factor neurons. EMBO J 15:3871–3879 41. Good DJ, Porter FD, Mahon KA, Parlow AF, Westphal H, Kirsch IR 1997 Hypogonadism and obesity in mice with a targeted deletion of the Nhlh2 gene. Nat Genet 15: 397–401 42. Gradwohl G, Dierich A, LeMeur M, Guillemot F 2000 neurogenin3 is required for the development of the four endocrine cell lineages of the pancreas. Proc Natl Acad Sci USA 97:1607–1611 43. Ma Q, Fode C, Guillemot F, Anderson DJ 1999 Neurogenin1 and neurogenin2 control two distinct waves of neurogenesis in developing dorsal root ganglia. Genes Dev 13:1717–1728 44. Higuchi H, Yang HY, Sabol SL 1988 Rat neuropeptide Y precursor gene expression. mRNA structure, tissue distribution, and regulation by glucocorticoids, cyclic AMP, and phorbol ester. J Biol Chem 263:6288–6295 45. Altman J, Bayer S 1995 Atlas of prenatal rat brain development. Boca Raton, FL: CRC Press

Molecular Endocrinology is published monthly by The Endocrine Society (http://www.endo-society.org), the foremost professional society serving the endocrine community.