Scholarly article on topic 'Retinoid Acid Specifies Neuronal Identity through Graded Expression of Ascl1'

Retinoid Acid Specifies Neuronal Identity through Graded Expression of Ascl1 Academic research paper on "Biological sciences"

Share paper
Academic journal
Current Biology
OECD Field of science

Abstract of research paper on Biological sciences, author of scientific article — John Jacob, Jennifer Kong, Steven Moore, Christopher Milton, Noriaki Sasai, et al.

Summary Cell diversity and organization in the neural tube depend on the integration of extrinsic signals acting along orthogonal axes. These are believed to specify distinct cellular identities by triggering all-or-none changes in expression of combinations of transcription factors [1]. Under the influence of a common dorsoventral signal, sonic hedgehog, and distinct anterior-posterior (A-P) inductive signals [2, 3], two topographically related progenitor pools that share a common transcriptional code produce serotonergic and V3 neurons in the hindbrain and spinal cord, respectively [4–7]. These neurons have different physiological properties, functions, and connectivity [8, 9]. Serotonergic involvement in neuropsychiatric diseases has prompted greater characterization of their postmitotic repertoire of fate determinants, which include Gata2, Lmx1b, and Pet1 [10], whereas V3 neurons express Sim1 [4]. How distinct serotonergic and V3 neuronal identities emerge from progenitors that share a common transcriptional code is not understood. Here, we show that changes in retinoid activity in these two progenitor pools determine their fates. Retinoids, via Notch signaling, control the expression level in progenitors of the transcription factor Ascl1, which selects serotonergic and V3 neuronal identities in a dose-dependent manner. Therefore, quantitative differences in the expression of a single component of a transcriptional code can select distinct cell fates.

Academic research paper on topic "Retinoid Acid Specifies Neuronal Identity through Graded Expression of Ascl1"

Current Biology 23, 412-418, March 4, 2013 ©2013 Elsevier Ltd All rights reserved http://dx.doi.Org/10.1016/j.cub.2013.01.046


Retinoid Acid Specifies Neuronal Identity

through Graded Expression of Ascll

John Jacob,1'3'* Jennifer Kong,4'5 Steven Moore,1 Christopher Milton,1 Noriaki Sasai,1

Rosa Gonzalez-Quevedo,2 Javier Terriente,2 Itaru Imayoshi,6

Ryoichiro Kageyama,6 David G. Wilkinson,2

Bennett G. Novitch,4,5 and James Briscoe1,*

1 Division of Developmental Biology

2Division of Developmental Neurobiology

MRC National Institute for Medical Research, The Ridgeway,

Mill Hill, London NW7 1AA, UK

3National Hospital for Neurology and Neurosurgery,

Queen Square, London WC1N 3BG, UK

4Department of Neurobiology

5Broad Center of Regenerative Medicine and Stem Cell Research

David Geffen School of Medicine at UCLA, 610 Charles E. Young Drive E, Los Angeles, CA 90095, USA 6Laboratory of Growth Regulation, Institute for Virus Research, Kyoto University, Shogoin-Kawahara, Sakyo-ku, Kyoto 606- 8507, Japan


Cell diversity and organization in the neural tube depend on the integration of extrinsic signals acting along orthogonal axes. These are believed to specify distinct cellular identities by triggering all-or-none changes in expression of combinations of transcription factors [1]. Under the influence of a common dorsoventral signal, sonic hedgehog, and distinct anterior-posterior (A-P) inductive signals [2, 3], two topographically related progenitor pools that share a common transcriptional code produce serotonergic and V3 neurons in the hindbrain and spinal cord, respectively [4-7]. These neurons have different physiological properties, functions, and connectivity [8, 9]. Serotonergic involvement in neuro-psychiatric diseases has prompted greater characterization of their postmitotic repertoire of fate determinants, which include Gata2, Lmxlb, and Pet1 [10], whereas V3 neurons express Sim1 [4]. How distinct serotonergic and V3 neuronal identities emerge from progenitors that share a common transcriptional code is not understood. Here, we show that changes in retinoid activity in these two progenitor pools determine their fates. Retinoids, via Notch signaling, control the expression level in progenitors of the transcription factor Ascl1, which selects serotonergic and V3 neuronal identities in a dose-dependent manner. Therefore, quantitative differences in the expression of a single component of a transcriptional code can select distinct cell fates.


Through the inductive effects of signaling molecules such as sonic hedgehog [4] and fibroblast growth factors [3], hindbrain serotonergic (5HT) progenitors acquire a Foxa2+Nkx2.2+

'Correspondence: (J.J.), (J.B.)

Ascll + transcriptional code. Topographically related V3 neural progenitors in the spinal cord express the same three factors (see Figure S1A available online). All three factors have been shown to be necessary for the generation of 5HT neurons, and Nkx2.2 and Ascll are critical for V3 neurogenesis [4, 5, 11, 12]. Consistent with its expression in spinal cord p3 progenitors, Foxa2 is also required for the generation of V3 neurons (Figures S1B, S1D, and S1E) [13]. This raises the question of how hindbrain and spinal p3 progenitors, designated p3 [5HT] and p3[V3], respectively, generate different cell types.

Retinoids Regulate p3 Progenitor Identity

We analyzed p-gal expression in the hindbrain and spinal cord of RARE-LacZtransgenic reporter mice [14]. The hindbrain has a segmental organization and is composed of eight compartments termed rhombomeres [15]. p-gal immunostaining revealed low or absent RA signaling in p3[5HT] progenitors at multiple rhombomeric levels, but high levels of activity in p3 [V3] cells (Figure S2A). This is consistent with other assays of retinoid signaling in amniotes (for example, see [16, 17]) and suggested that retinoid signaling could be involved in p3 progenitor fate determination. To test this, we took advantage of a constitutively active retinoid receptor a derivative, VP16RAR, which has previously been shown to activate reti-noic acid receptor target genes independently of RA [18]. Forced expression of VP16RAR by in ovo electroporation reduced 5HT neuronal differentiation by40%-50% throughout the A-P extent of the hindbrain (Figures 1A-1C). Concomi-tantly, occasional ectopic V3 neurons, marked by expression of the postmitotically expressed transcription factor Sim1, were produced in the caudal hindbrain (data not shown) [4]. Conversely, blockade of retinoid signaling in p3[V3] progenitors, either by misexpression of Cyp26B1, a member of the cytochrome P450 class of enzymes that degrades retinoic acid, or RAR403, a dominant negative version of the human RA receptor a [19], resulted in ectopic 5HT neuronal generation (Figure 1D). These neurons expressed the postmitotic serotonergic determinants Gata2 [20] and Lmx1b [21, 22], but not Pet1 [23], and they also expressed the terminal differentiation marker Tph2 [24] (Figure 1E). Therefore, ectopic spinal 5HT neurons share the same differentiation pathway as hindbrain 5HT neurons, of which approximately one-third do not depend on Pet1 for their generation [25, 26].

The ventral restriction of ectopic spinal 5HT neurons suggested their derivation from presumptive p3[V3] progenitors, and blockade of Nkx2.2 or Foxa2 activity using a dominant interfering derivative of Nkx2.2 (Nkx2.2-VP16) [27] or Foxa2 (FkhA2-EnR) [5], respectively, prevented RAR403 induction of ectopic 5HT neurons (Figure S2B). Consistent with these data, forced expression of Foxa2, Nkx2.2, and RAR403 in the dorsal neural tube led to extensive ectopic generation of 5HT neurons, whereas the omission of RAR403 led to a failure to induce ectopic 5HT neurons (Figure 1F). This result also suggested that the small number of ectopic 5HT neurons generated by p3[V3] in ovo electroporation with RAR403 is likely to be a consequence of the restricted overlap of endogenous Foxa2 and Nkx2.2 expression and mosaic misexpression of RAR403. Furthermore, as quantified below, at brachial level


Graded Ascl1 Expression and Control of Neural Fate

the p3 domain normally only generates relatively small numbers of V3 neurons, suggesting that low neurogenic potential might further limit ectopic 5HT neurogenesis.

V3 neuronal differentiation was suppressed by 60% upon RAR403 misexpression (V3 neurons in control = 10.7 [mean] ± 4.1 [SD]; V3 neurons following RAR403 misexpression = 4.7 ± 2.2; p = 7.18 x 10"9, unpaired Student's t test) (Figure 1E). This was not due to cell death or to a fate transformation to other ventral neuronal subtypes (Figure S2C). Furthermore, ectopic visceral motor (VM) neurons that precede the generation of 5HT neurons in the hindbrain from p3 progenitors [6] (Figure S1A) were not detected (Figure S2C). Altogether, these data indicate that high-level, versus low-level or absent, retinoid signaling in p3[V3] and p3[5HT], respectively, determines the corresponding neuronal identities.

Retinoids Regulate Ascl1 Expression in p3 Progenitors via the Notch Pathway

The profile of RA signaling in the ventral neural tube correlates with the longitudinal expression of Ascl1 in p3 progenitors (Figures S1C and S2A). The significance of this

Figure 1. Retinoid Signaling Determines p3 Progenitor Fate

(A) Upregulation of RA signaling in hindbrain p3[5HT] progenitors by misexpression of VP16RAR-IRES GFP reduces Pet1 expression. Scale bar represents 25 mm. (B and C) Effect of VP16RAR misexpression on Pet1+ neurons in the rostral (B) or caudal (C) hindbrain at E5. *p = 0.025 (B, blue bar), *p = 0.026 (B, red bar), *p = 8.11 x 10"6 (C, black bar) (unpaired Student's t tests). Error bars represent SD in all graphs. r, rhombomere.

(D) Forced expression of Cyp26B1 or RAR403 in p3[V3] generates ectopic 5HT neurons (arrowheads) by E5. Scale bar represents 14 mm.

(E) Ectopic spinal 5HT neurons express the indicated markers (arrowheads), and Sim1 expression is reduced. Asterisk marks the floor plate. Scale bar represents 16 mm.

(F) Nkx2.2 and Foxa2 misexpression only induce ectopic spinal 5HT neurons when RA signaling is blocked by RAR403. -, control side; +, electro-porated side. See also Figure S2.

observation lies in the critical subtype-specification function of Ascl1 in p3[5HT] and p3[V3] progenitors, which is separable from its neurogenic properties [11, 12]. In the p3[5HT] domain where RA signaling is weak or absent, the expression level of chick Ascl1 appeared markedly higher than in the p3 [V3] domain where RA signaling is strong (Figure S1C). This observation prompted closer scrutiny of the expression of Ascl1 in p3 progenitors in the spinal cord and hindbrain (Figures 2A and 2B). Indeed, the mean level of Ascl1 expression was consistently greater in p3[5HT] than in p3[V3] progenitors between E9.5 and E11.5. Moreover, Ascl1 expression level was enhanced in hindbrain p3 progenitors after E9.5, corresponding to the time 5HT neurogenesis commences [5].

The Notch pathway is known to inhibit proneural gene expression, and, consistent with this regulatory relationship, in chick p3 progenitors the Hes family of Notch target genes showed reciprocal patterns of expression to Ascl1 (Figures S2D and S2E). Moreover, the Notch pathway lies downstream of retinoid signaling. Misexpression of RAR403 in the p3[V3] domain led to the downregulation of the Notch pathway by E4 (Figure 2C). Conversely, forced expression of VP16RAR in the hindbrain p3[5HT] domain upregulated Notch activity (Figure 2D). Retinoid regulation of the Notch pathway suggested Notch signaling might serve to link the A-P pattern of retinoid activity and Ascl1 expression in p3 progenitors.

To test whether the expression level of Ascl1 in p3[5HT] and p3[V3] is regulated by retinoid signaling, we analyzed Ascl1 expression in chick p3 progenitors following cell-autonomous blockade or activation of RA signaling. RAR403 markedly up-regulated Ascl1 expression in the p3[V3] domain by E5 24 hr after the downregulation of the Notch pathway (control p3[V3] expression = 126 [mean] ± 19.3 [SD] [arbitrary units];

Figure 2. Ascll Expression Level in p3 Progenitors Is Negatively Correlated with Notch Activity, which in Turn Is Regulated by Retinoids

(A and B) Ascll expression in mouse Nkx2.2+ p3 progenitors (A), with quantification (mean ± SD) shown in (B). Scale bar represents 9 mm. a.u., arbitrary units.

n = 3 sections from each of three embryos per developmental stage.

(C) Spinal RAR403 misexpression, marked by anti-GFP immunofluorescence, downregulates Hes gene expression (bracketed).

(D) Misexpression of VP16RAR upregulates hindbrain p3 Hes gene expression.

(E) Misexpression of RAR403 in p3[V3] progenitors upregulates Ascll expression (arrowhead, inset) at E5, and this is prevented by cAscll-shRNA. Inset shows a high-power image of the boxed region. Scale bar represents 18 mm.

(F) VP16RAR misexpression attenuates Ascll expression (arrowhead) in p3[5HT] progenitors. Scale bar represents 18 mm.

(G) Increased Ascll expression in brachial Nkx2.2+ p3[V3] cells (white brackets) of RbpjCKO mouse embryos is similar to Ascll expression in WT p3[5HT] cells. Scale bar represents 14 mm.

(H) Quantification of Ascll expression in presumptive p3[V3] cells misexpressing RAR403 or RAR403 and cAscll-shRNA.

(I) Misexpression of Hes1-IRES GFP in caudal hindbrain p3[5HT] reduces Ascll expression. Scale bar represents 18 mm. See also Figure S1.

RAR403 p3[V3] expression = 221.6 ± 7.8; control p3[5HT] expression = 223.1 ± 26.3). (Figures 2E and 2H; Figure S2F). Moreover, the upregulation of Ascll could be prevented by the comisexpression of a short hairpin RNA against chick Ascll (cAscll-shRNA) (control p3[V3] expression = 110.9 [mean] ± 19.4 [SD]; RAR403 + cAscll-shRNA p3[V3] expression = 107.1 ± 11.2; control p3[5HT] expression = 208.4 ± 27.5) (Figures 2E and 2H). By contrast, misexpression of VP16RAR in p3[5HT] progenitors downregulated Ascll by E4 1 day after the upregulation of the Notch pathway in these progenitors (control p3[5HT] expression = 193.1 [mean] ± 27.1 [SD]; VP16RAR p3[5HT] expression = 94.4 ± 16.9; control p3[V3] expression = 105.1 ± 16.9) (Figure 2F). In neither condition were there detectable changes in expression of the other proneural genes, Neurogl and Neurog2, or of other ventral progenitor markers or sonic hedgehog. Moreover, there was no increase in cell death (Figures S2F-S2H).

To determine whether Notch signaling regulates Ascl1 expression in p3 progenitors, we conditionally deleted recombination signal binding protein for immunoglobulin kappa j region (Rbpj), an obligate component of the Notch signaling pathway [28], using Olig2-cre [29]. At E10.5, mice lacking Rbpj (RbpjCKO) expressed higher levels of Ascl1 in the spinal cord p3 domain, compared to either mice that have only one conditional allele or mice that lacked Cre (WT p3[V3]

expression: 50.5 [mean Ascl1 expression] ± 28.0 [SD]; Rbpj«°x'+ p3[V3] expression: 49.3 ± 29.0; RbpjCKO p3[V3] expression: 78.1 ± 50.4; WT p3[5HT] expression: 90.3 ± 56.8) (Figure 2G). Conversely, misexpression of Hes1, an effector of Notch signaling, in the caudal hindbrain of the chick lowered the expression of Ascll in p3[5HT] progenitors to levels comparable to the expression of Ascl1 in spinal cord p3 progenitors (control p3[5HT] = 221.9 [mean] ± 20.9 [SD]; Hes1 electroporation p3[5HT] = 123.4 ± 8.7; control p3[V3] = 119.8 ± 22.1) (Figure 2I). The expression of Foxa2 and Nkx2.2 were unaffected (Figure S2I). Together, these gain-and loss-of-function data suggested that differential expression of Ascl1 in the two p3 progenitor pools is determined by retinoid signaling acting via the Notch pathway. Furthermore, our findings raised the possibility that differences in the expression level of Ascl1 might determine p3 progenitor fate.

Ascl1 Expression Level Discriminates p3[V3] from p3[5HT] Progenitors

We exploited the regulatory relationship between Notch signaling and proneural gene expression to test the effect of boosting Ascl1 expression in the spinal cord on p3 progenitor fate. Strikingly, although 5HT neurons are normally never found in this region of the neural tube, ectopic spinal 5HT+GFP+ neuronal cell bodies were observed in the cervical

Graded Ascii Expression and Controi of Neurai Fate

cords of nine out of nine Rbpj conditional null mice and two out of five Cre/+; Rbpflox/+ mice (Figure 3A). Moreover, analysis of embryos lacking Hairy and Enhancer-of-Split (Hes) family member Hes5, which is a target of Notch signaling broadly expressed in the spinal cord [30] (Figure 3B), revealed ectopic 5HT neurons in the ventral spinal cord at forelimb levels in three out of four Hes5+/~ embryos and four out of four Hes5 null embryos. These neurons were more abundant than in Rbpj mutants: the later deletion of Rbpj in neural cells by Olig2-cre (~E9.5) might explain the smaller number of 5HT neurons.

To further explore the role of Ascll expression level in regulating p3 progenitor fate, we first increased Ascll levels in the p3[V3] domain by direct overexpression in ovo (Figure 3C). Boosting the level of Ascll in the p3[V3] domain was sufficient to generate ectopic 5HT neurons cell autonomously and suppress V3 neuronal differentiation (the number of Sim1+ cells on the electroporated side was 1.59 [mean] ± 1.6 [SD], and on the control side there were 8.12 ± 2.3, p = 2.80 x 10"10; n = 18 sections from five embryos), indicating that upre-gulation of Ascl1 reprograms the p3[V3] domain to a p3[5HT] identity. Moreover, there was no evidence that cell death, as revealed by activated Caspase3 immunostaining, could

Figure 3. Altering Ascll Expression Level in p3 Progenitors Produces Corresponding Changes in p3 Progenitor Identity

(A and B) Ectopic 5HT+GFP+ neurons (yellow cells, arrowheads) in the cervical cord of Rbpj mutants (A) and 5HT+ cells in the brachial cord of Hes5 mutants (B). Insets show high-power images of 5HT+GFP+ neurons in boxed regions. Scale bar represents 25 mm in (A) and 13 mm in (B).

(C) Forced Ascii expression in cervicobrachial p3[V3] cells induces ectopic 5HT neurons (arrows) and reduces V3 differentiation. Cell death, assayed by activated Caspase3 expression, is not detected in the ventral spinal cord. Ascii, but not Neurog2, substitutes for RAR403 in combined electroporation with Foxa2 and Nkx2.2 to induce ectopic spinal 5HT neurons. Scale bar represents 13 mm.

(D) Ectopic Lmxlb and Gata2 expression (arrows) upon combined Foxa2, Nkx2.2, and Ascii electroporation at brachial level. Scale bar represents 14 mm.

(E) Hesl-IRES GFP misexpression in caudal p3[5HT] cells reduces 5HT neuronal differentiation marked by Lmxlb (circled) and Peti and induces ectopic V3 neurons (arrowheads).

account for the reduced number of V3 neurons (Figure 3C). Next, we misex-pressed the three p3-defining transcription factors, Nkx2.2, Foxa2, and Ascii, in the dorsal spinal cord. The widespread and cell-autonomous generation of ectopic 5HT neurons that resulted shows that Ascii can substitute for RAR403 and supports the view that high-level Ascl1 expression is sufficient to select a p3[5HT] identity over a p3 [V3] identity (Figure 3C). Ectopic 5HT neurons expressed postmitotic markers of hindbrain 5HT neurons Lmx1b and Gata2, but not Peti (Figure 3D and data not shown), suggesting that the differentiation of these neurons largely recapitulates the serotonergic differentiation pathway. Importantly, neither overexpression of Foxa2 and Nkx2.2 alone nor replacement of Ascii by another bHLH gene, Neurog2, led to ectopic 5HT neuronal generation (Figures 1F and 3C). Therefore, increasing the expression of Ascii converts p3[V3] to a p3[5HT] identity.

We then asked whether reducing the expression of Ascii in presumptive p3[5HT] progenitors would cause a reciprocal switch to a p3[V3] identity. Following Hesi in ovo electroporation into presumptive p3[5HT] progenitors, significant reductions in 5HT neuronal differentiation were observed, using Lmxlb and Peti as markers (control side 5HT neurons = 8.2 [mean] ± 2.8 [SD]; Hesi -electroporated side = 2.5 ± 2.1, p = 8.2 x 10"8), and small numbers of Simi-expressing ectopic V3 neurons were consistently detected in the caudal hindbrain (Figure 3E). This is consistent with the low neurogenic potential of the caudal hindbrain p3[5HT] domain. The block in generation of ectopic V3 neurons by comisexpression of Nkx2.2-VPi6 confirmed the derivation of these Hesi-induced ectopic V3 neurons from the presumptive p3[5HT] domain (Figure S3A). To further test the idea that Ascll expression level is important for p3[5HT] identity, we measured the expression of Ascll in

cAscll-shRNA hindbrain

A'sclï -1 v" -jt i RFP 1 / ; : d sr_

Pet1\ / . >. .......'.......4

Lmxlb Lmxlb RFP

.-A + ü

cAscH-shRNA + rAscii hindbrain



RAR403 + cAscH-shRNA

spinal cord

Sim1 RFP

- -. i^^iÎ^jl.

Figure 4. Knockdown of Chick Ascll Alters p3 Progenitor Identity

(A) Misexpression of cAscll-shRNA IRES RFP decreases Ascll expression in hindbrain p3 progenitors (bracket) and reduces Pet1 (arrowhead) and Lmxlb expression (short arrows). Dotted white line demarcates Lmxlb expression corresponding to the floor plate. RFP expression on the control side is confined to decussating axons. Scale bar represents 18 mm.

(B) Coelectroporation of cAscll-shRNA and rAscll rescues Pet1 expression (arrowhead).

(C) cAscll-shRNA restores Sim1+ V3 neurons (arrowhead) and blocks ectopic spinal 5HT expression induced by RAR403. Scale bar represents 13 mm.

(D and E) Quantification of Pet1+ neurons in the caudal hindbrain at E5 following forced expression of cAscll-shRNA (D) or cAscll-shRNA and rAscll (E). Red asterisk, p = 0.028; blue asterisk, p = 0.002. See also Figure S3.

o 1.4 Eg 1.2 •

¡I 1 "g ë 0-8 £ A 0.6 -0.4 0.2 0

Control side


Control side

Ascll heterozygote mice, in which 5HT neuronal differentiation is intact [11]. No significant change in Ascll expression was detected (WT p3[5HT] expression: 151.1 [mean] ± 53.2 [SD]; Ascl1+/~ p3[5HT] expression: 149.3 ± 39.7), which implies Discussion that allele number is not the main determinant of the level of Ascl1 expression (Figure S3B). Together, these data support the case for the critical role of Ascl1 expression level in the selection of 5HT over V3 neuronal fate.

Finally, to test directly whether the level of Ascl1 determines p3 progenitor fate, we used cAscll-shRNA. Forced expression of cAscll-shRNA in the ventral region of the caudal hindbrain reduced the expression of Ascll in hindbrain p3 progenitors (control p3[5HT] expression: 213.6 [mean] ± 22.8 [SD]; cAscll-shRNA p3[5HT] expression: 123.9 ± 13.3; control p3[V3] expression: 119.5 ± 19.5), and this was accompanied by a ~40% reduction in 5HT neuronal differentiation (Figures 4A and 4D). Consistent with the retinoid and Notch pathway gain-of-function experiments in the caudal hindbrain in ovo, occasional ectopic V3 neurons were also detected (Figure S3C). We then attempted to rescue these neural patterning defects by comisexpression of rat Ascl1 (rAscl1), which is predicted to be resistant to this shRNA (Figure 4B). rAscll restored serotonergic neurogenesis and prevented the ectopic generation of any V3 neurons by cAscll-shRNA (Figures 4B and 4E; Figure S3D). Transfection of a scrambled version of the chick Ascll shRNA (cAscl1-shRNA-SCR) did not alter the expression of Ascl1 in p3[5HT] progenitors (control p3[5HT] expression = 185.9 [mean] ± 30.1 [SD]; cAscl1-shRNA-SCR p3[5HT] expression = 176.0 ± 38.5), nor did it alter 5HT neurogenesis or induce any ectopic V3 neurons (Figures S3E and S3F). Furthermore, when the increased expression of Ascll in p3[V3] progenitors elicited by RAR403

was prevented by comisexpression of cAscll-shRNA (Figure 2E), V3 neuronal differentiation was rescued (V3 neurons in control = 11 [mean] ± 4.6 [SD]; V3 neurons following RAR403 + cAscll-shRNA misexpression = 10.8 ± 5.1; p = 0.92, not significant), and ectopic 5HT neuronal generation was blocked (Figure 4C). Taken together, these data indicate that Ascl1 recapitulates the effect of retinoids on p3 progenitor identity and that retinoids regulate the neuronal fates of p3 progenitors in an Ascll-dependent manner.

cAscll-shRNA + rAscll

We have shown how different profiles of retinoid signaling in p3 progenitors that share a common transcriptional code generate distinct neuronal identities at spinal cord and hind-brain levels. Binary differences in retinoid activity result in graded shifts in expression of a pivotal transcription factor, Ascl1, which is critical for the selection of alternate progenitor identities. This mechanism of progenitor fate specification is distinct from the conventional combinatorial model, because allocation to V3 or 5HT neuronal identity is determined not by qualitative but by quantitative differences in the transcriptional code [1]. Moreover, this model does not exclude the possibility that other, as yet uncharacterized, intrinsic determinants function in the same pathway as Ascl1 and could discriminate p3[5HT] from p3[V3].

In invertebrates, graded activity of transcription factors has been shown to control differential gene expression and thereby diversify cell fates [31]. By contrast in vertebrates, in general, transcription factor gradients appear to refine neural identity rather than instruct distinct cell identities in a dose-dependent manner [32-36]. Consistent with their closely similar genetic identities, the manipulations of retinoid signaling and Ascl1 expression demonstrate that spinal p3 progenitors possess a cryptic bipotency. The induction of ectopic V3 neurons in the hindbrain is less robust, which implies that hindbrain p3 progenitors are not fully interconvertible by graded Ascl1 expression alone. Nevertheless, it is clear that Ascll expression must be tightly regulated to regionally

Graded Ascl1 Expression and Control of Neural Fate

constrain 5HT neurogenesis. Importantly, failure to appropriately regulate the number of 5HT neurons is associated with neuropsychiatric disease states, for example Smith-Lemli-Opitz syndrome, in which hypermorphic serotonergic differentiation occurs [37].

Supplemental Information

Supplemental Information includes three figures and Supplemental Experimental Procedures and can be found with this article online at http://dx.doi. org/10.1016/j.cub.2013.01.046.


We thank F. Guillemot and E. Kutejova for comments on the manuscript and P. Dolle and V. Ribes for RARE-LacZ mice. This work was funded by the Medical Research Council (MRC) (UK). J.K. and B.G.N. were supported by the Broad Center for Regenerative Medicine and Stem Cell Research and by grants to B.N. from the National Institute of Neurological Disorders and Stroke (NS053976) and the March of Dimes Foundation (06-FY10-296). Olig2-Cre; ROSA26-GFP; Rbpjflox'+ mice were maintained in accordance with UCLA Chancellor's Animal Research Committee husbandry guidelines. Ascl1*'~ mice were maintained in the animal husbandry facility at the MRC National Institute for Medical Research in accordance with standards set by the Animals (Scientific Procedures) Act, 1986. Hes5*'~ mice were handled in accordance with the Kyoto University Guide for the Care and Use of Laboratory Animals.

Received: July 17, 2012 Revised: November 1, 2012 Accepted: January 15, 2013 Published: February 14, 2013


1. Jessell, T.M. (2000). Neuronal specification in the spinal cord: inductive signals and transcriptional codes. Nat. Rev. Genet. 1, 20-29.

2. Briscoe, J., and Ericson, J. (2001). Specification of neuronal fates in the ventral neural tube. Curr. Opin. Neurobiol. 11, 43-49.

3. Ye, W., Shimamura, K., Rubenstein, J.L., Hynes, M.A., and Rosenthal, A. (1998). FGF and Shh signals control dopaminergic and serotonergic cell fate in the anterior neural plate. Cell 93, 755-766.

4. Briscoe, J., Sussel, L., Serup, P., Hartigan-O'Connor, D., Jessell, T.M., Rubenstein, J.L., and Ericson, J. (1999). Homeobox gene Nkx2.2 and specification of neuronal identity by graded Sonic hedgehog signalling. Nature 398, 622-627.

5. Jacob, J., Ferri, A.L., Milton, C., Prin, F., Pla, P., Lin, W., Gavalas, A., Ang, S.L., and Briscoe, J. (2007). Transcriptional repression coordinates the temporal switch from motor to serotonergic neurogenesis. Nat. Neurosci. 10, 1433-1439.

6. Pattyn, A., Vallstedt, A., Dias, J.M., Samad, O.A., Krumlauf, R., Rijli, F.M., Brunet, J.F., and Ericson, J. (2003). Coordinated temporal and spatial control of motor neuron and serotonergic neuron generation from a common pool of CNS progenitors. Genes Dev. 17, 729-737.

7. Cordes, S.P. (2001). Molecular genetics of cranial nerve development in mouse. Nat. Rev. Neurosci. 2, 611-623.

8. Ray, R.S., Corcoran, A.E., Brust, R.D., Kim, J.C., Richerson, G.B., Nattie, E., and Dymecki, S.M. (2011). Impaired respiratory and body temperature control upon acute serotonergic neuron inhibition. Science 333, 637-642.

9. Zhang, Y., Narayan, S., Geiman, E., Lanuza, G.M., Velasquez, T., Shanks, B., Akay, T., Dyck, J., Pearson, K., Gosgnach, S., et al. (2008). V3 spinal neurons establish a robust and balanced locomotor rhythm during walking. Neuron 60, 84-96.

10. Deneris, E.S., and Wyler, S.C. (2012). Serotonergic transcriptional networks and potential importance to mental health. Nat. Neurosci. 15, 519-527.

11. Pattyn, A., Simplicio, N., van Doorninck, J.H., Goridis, C., Guillemot, F., and Brunet, J.F. (2004). Ascl1/Mash1 is required for the development of central serotonergic neurons. Nat. Neurosci. 7, 589-595.

12. Sugimori, M., Nagao, M., Bertrand, N., Parras, C.M., Guillemot, F., and Nakafuku, M. (2007). Combinatorial actions of patterning and HLH

transcription factors in the spatiotemporal control of neurogenesis and gliogenesis in the developing spinal cord. Development 134, 1617-1629.

13. Ribes, V., Balaskas, N., Sasai, N., Cruz, C., Dessaud, E., Cayuso, J., Tozer, S., Yang, L.L., Novitch, B., Marti, E., and Briscoe, J. (2010). Distinct Sonic Hedgehog signaling dynamics specify floor plate and ventral neuronal progenitors in the vertebrate neural tube. Genes Dev. 24,1186-1200.

14. Rossant, J., Zirngibl, R., Cado, D., Shago, M., and Giguesre, V. (1991). Expression of a retinoic acid response element-hsplacZ transgene defines specific domains of transcriptional activity during mouse embryogenesis. Genes Dev. 5,1333-1344.

15. Lumsden, A., and Krumlauf, R. (1996). Patterning the vertebrate neu-raxis. Science 274, 1109-1115.

16. MacLean, G., Abu-Abed, S., Dolle, P., Tahayato, A., Chambon, P., and Petkovich, M. (2001). Cloning of a novel retinoic-acid metabolizing cytochrome P450, Cyp26B1, and comparative expression analysis with Cyp26A1 during early murine development. Mech. Dev. 107,195-201.

17. Wilson, L.J., Myat, A., Sharma, A., Maden, M., and Wingate, R.J. (2007). Retinoic acid is a potential dorsalising signal in the late embryonic chick hindbrain. BMC Dev. Biol. 7,138.

18. Castro, D.S., Arvidsson, M., Bondesson Bolin, M., and Perlmann, T. (1999). Activity of the Nurr1 carboxyl-terminal domain depends on cell type and integrity of the activation function 2. J. Biol. Chem. 274, 37483-37490.

19. Damm, K., Heyman, R.A., Umesono, K., and Evans, R.M. (1993). Functional inhibition of retinoic acid response by dominant negative retinoic acid receptor mutants. Proc. Natl. Acad. Sci. USA 90, 29892993.

20. Craven, S.E., Lim, K.C., Ye, W., Engel, J.D., de Sauvage, F., and Rosenthal, A. (2004). Gata2 specifies serotonergic neurons downstream of sonic hedgehog. Development 131,1165-1173.

21. Cheng, L., Chen, C.L., Luo, P., Tan, M., Qiu, M., Johnson, R., and Ma, Q. (2003). Lmx1b, Pet-1, and Nkx2.2 coordinately specify serotonergic neurotransmitter phenotype. J. Neurosci. 23, 9961-9967.

22. Ding, Y.Q., Marklund, U., Yuan, W., Yin, J., Wegman, L., Ericson, J., Deneris, E., Johnson, R.L., and Chen, Z.F. (2003). Lmx1b is essential forthe development of serotonergic neurons. Nat. Neurosci. 6,933-938.

23. Hendricks, T., Francis, N., Fyodorov, D., and Deneris, E.S. (1999). The ETS domain factor Pet-1 is an early and precise marker of central serotonin neurons and interacts with a conserved element in serotonergic genes. J. Neurosci. 19,10348-10356.

24. Zhang, X., Beaulieu, J.M., Sotnikova, T.D., Gainetdinov, R.R., and Caron, M.G. (2004). Tryptophan hydroxylase-2 controls brain serotonin synthesis. Science 305, 217.

25. Hendricks, T.J., Fyodorov, D.V., Wegman, L.J., Lelutiu, N.B., Pehek, E.A., Yamamoto, B., Silver, J., Weeber, E.J., Sweatt, J.D., and Deneris, E.S. (2003). Pet-1 ETS gene plays a critical role in 5-HT neuron development and is required for normal anxiety-like and aggressive behavior. Neuron 37, 233-247.

26. Kiyasova, V., Fernandez, S.P., Laine, J., Stankovski, L., Muzerelle, A., Doly, S., and Gaspar, P. (2011). A genetically defined morphologically and functionally unique subset of 5-HT neurons in the mouse raphe nuclei. J. Neurosci. 31, 2756-2768.

27. Muhr, J., Andersson, E., Persson, M., Jessell, T.M., and Ericson, J. (2001). Groucho-mediated transcriptional repression establishes progenitor cell pattern and neuronal fate in the ventral neural tube. Cell 104, 861-873.

28. Kopan, R., and Ilagan, M.X. (2009). The canonical Notch signaling pathway: unfolding the activation mechanism. Cell 137, 216-233.

29. Dessaud, E., Yang, L.L., Hill, K., Cox, B., Ulloa, F., Ribeiro, A., Mynett, A., Novitch, B.G., and Briscoe, J. (2007). Interpretation of the sonic hedgehog morphogen gradient by a temporal adaptation mechanism. Nature 450, 717-720.

30. Hatakeyama, J., Bessho, Y., Katoh, K., Ookawara, S., Fujioka, M., Guillemot, F., and Kageyama, R. (2004). Hes genes regulate size, shape and histogenesis of the nervous system by control of the timing of neural stem cell differentiation. Development 131, 5539-5550.

31. Lewis, J. (2008). From signals to patterns: space, time, and mathematics in developmental biology. Science 322, 399-403.

32. Chen, A.I., de Nooij, J.C., and Jessell, T.M. (2006). Graded activity of transcription factor Runx3 specifies the laminar termination pattern of sensory axons in the developing spinal cord. Neuron 49, 395-408.

33. Sansom, S.N., and Livesey, F.J. (2009). Gradients in the brain: the control of the development of form and function in the cerebral cortex. Cold Spring Harb. Perspect. Biol. 1, a002519.

34. Hamasaki, T., Leingartner, A., Ringstedt, T., and O'Leary, D.D. (2004). EMX2 regulates sizes and positioning of the primary sensory and motor areas in neocortex by direct specification of cortical progenitors. Neuron 43, 359-372.

35. Dasen, J.S., De Camilli, A., Wang, B., Tucker, P.W., and Jessell, T.M. (2008). Hox repertoires for motor neuron diversity and connectivity gated by a single accessory factor, FoxPI. Cell 134, 304-316.

36. Rousso, D.L., Gaber, Z.B., Wellik, D., Morrisey, E.E., and Novitch, B.G. (2008). Coordinated actions of the forkhead protein Foxp1 and Hox proteins in the columnar organization of spinal motor neurons. Neuron 59, 226-240.

37. Waage-Baudet, H., Lauder, J.M., Dehart, D.B., Kluckman, K., Hiller, S., Tint, G.S., and Sulik, K.K. (2003). Abnormal serotonergic development in a mouse model for the Smith-Lemli-Opitz syndrome: implications for autism. Int. J. Dev. Neurosci. 21, 451-459.