The helix-loop-helix protein Id1 controls stem cell proliferation during regenerative neurogenesis in the adult zebrafish telencephalon

STEM CELLS 2014;00:00‐00 www.StemCells.com ©AlphaMed Press 2014

TISSUE‐SPECIFIC STEM CELLS 1Karlsruhe Institute of Technology, Institute of Toxicology and Genetics, Karlsruhe, Germany; 2 Institute of Neu‐robiology A. Fessard, CNRS UPR3294, Gif‐sur‐Yvette, France 6correspondence: Uwe Strähle, Karlsruhe Institute of Technology, Insti‐tute of Toxicology and Genetics, Eg‐genstein, Germany, [email protected]; 3current ad‐dress: Centre for Pulmonary Hyperten‐sion Thoraxclinic and Institute of Hu‐man Genetics, University of Heidel‐berg, Im Neuenheimer Feld 366, Ger‐many; 4current address: Laboratoire de Biochimie et Génétique Moléculaire, Groupe d'Etude sur l'Inflammation Chronique et l'Obésité, Plateforme CYROI, Université de La Réunion, 97715 Saint Denis de La Réunion Cedex 09, France; 5current address: Depart‐ment of Pediatrics, Medical Faculty Mannheim, Heidelberg University, Theodor‐Kutzer‐Ufer 1‐3, 68167 Mannheim, Germany Received May 21, 2014; accepted for publication September 27, 2014; available online without subscription through the open access option. ©AlphaMed Press 1066‐5099/2014/$30.00/0 This article has been accepted for publication and undergone full peer review but has not been through the copyediting, typesetting, pagination and proofreading process which may lead to differences between this version and the Version of Record. Please cite this article as doi: 10.1002/stem.1883

The Helix‐Loop‐Helix Protein Id1 Controls Stem Cell Proliferation During Regenerative Neuro‐genesis in the Adult Zebrafish Telencephalon

Rebecca Rodriguez Viales1, 3, Nicolas Diotel1, 4, Marco Ferg1,

Olivier Armant1, Julia Eich1, Alessandro Alunni2, Martin März1,5, Laure Bally‐Cuif2, Sepand Rastegar1, 6 and Uwe Strähle1, 6

Key words. id1 • transcription regulators • reactive neurogenesis • expres‐sion atlas • telencephalon • zebrafish

ABSTRACT

The teleost brain has the remarkable ability to generate new neurons and to repair injuries during adult life stages. Maintaining life‐long neurogene‐sis requires careful management of neural stem cell pools. In a genome‐wide expression screen for transcription regulators (TRs), the id1 gene, encoding a negative regulator of E‐proteins, was found to be up‐regulated in response to injury. id1 expression was mapped to quiescent type I neu‐ral stem cells in the adult telencephalic stem cell niche. Gain and loss of id1 function in vivo demonstrated that Id1 promotes stem cell quiescence. The increased id1 expression observed in neural stem cells in response to injury appeared independent of inflammatory signals, suggesting multiple an‐tagonistic pathways in the regulation of reactive neurogenesis. Together, we propose that Id1 acts to maintain the neural stem cell pool by counter‐acting neurogenesis‐promoting signals. STEM CELLS 2014; 00:000–000

INTRODUCTION

While the adult mammalian brain has a rather limited ability to generate new neurons and to repair injured nervous tissue, adult zebrafish form new neurons abundantly and repair brain lesions most effectively [1–3]. Many regions of the adult zebrafish brain harbor neurogenic niches. In the adult telencephalon, for ex‐ample, the entire ventricular zone produces new neu‐rons that integrate into existing neural tissue [4–9]. A

wound inflicted to the telencephalon leads to increased cell proliferation in the ventricular region within 2 to 3 days. The wound is healed in three weeks without re‐maining traces of the traumatic impact [10–12]. Although limited proliferation of oligodendrocytes and invading immune cells can be observed at the site of lesion [10, 12], new neurons are born distantly at the periventricular domain of the telencephalon [11]. These regions are occupied by the cell bodies of radial glial cells (RGCs), which extend long processes, reaching

Id1 controls neural stem cell proliferation

www.StemCells.com ©AlphaMed Press 2014

2

through the parenchyma to the pial surface. RGCs ex‐press markers such as Glial acidic fibrillary protein (GFAP), Nestin and S100β, which are well‐known mark‐ers for neural stem cells in the murine telencephalon [6, 9, 13, 14]. Most of the RGCs are normally quiescent (type I cells). During constitutive neurogenesis a small proportion of these RGCs (type II cells) express prolif‐eration markers such as proliferating cell nuclear anti‐gen (PCNA) and incorporate desoxythymidine analogues [6, 7, 9, 13, 15–17]. They can divide symmetrically or asymmetrically [18], thereby self‐renew and give rise to neuronal precursors (type III cells) [4, 7, 8]. Upon injury, many RGCs enter the cell cycle [10–12, 19]. Reactive neurogenesis in the zebrafish telencephalon can also be induced by inflammatory signals such as cysteinyl leu‐kotrienes [20], accompanied by up‐regulation of the zinc finger transcription factor Gata3 [21]. Maintaining the neural stem cell pool during adulthood requires a fine‐tuned balance between differentiation and self‐renewal. Activation of the Notch pathway con‐fers quiescence to RGCs during constitutive neurogene‐sis [15, 22]. Notch signaling has also been implicated in the retention of neural stem cells in the mouse brain [19, 23–25]. In the adult mouse, several lines of evi‐dence suggest that Id genes also play a role in mainte‐nance of the neural stem cell pool [26–28]. Id proteins contain a helix‐loop‐helix domain and are negative regulators of basic helix‐loop‐helix transcription factors (bHLH). They bind to the HLH domain of E class bHLH proteins and prevent heterodimer formation with Tcf3 (E12/47, E2A) [29]. Mouse Id1, Id2, Id3 proteins regu‐late redundantly Rap1GAP and thereby control the ad‐hesive architecture of the stem cell niche [28]. Whether zebrafish constitutive and reactive neurogenesis differ mechanistically remains to be thoroughly examined. Chemical inhibition of Notch signaling causes increased proliferation of RGCs in the pallium of the uninjured brain [15]. In contrast Notch signalling is up‐regulated concomitantly with cell proliferation in the subpallium of the wounded telencephalon [30] suggesting opposite functions of Notch signaling in constitutive and reactive neurogenesis. Also, Gata3 is not required for constitu‐tive neurogenesis [21]. Little is known about how stem cells are maintained after injury. We systematically analysed the expression of transcription regulators (TR) in the intact and injured telencephalon. We identified 279 TR genes whose ex‐pression is altered by injury, providing a comprehensive expression map of the injured telencephalon. Among these genes, id1 was up‐regulated specifically in the ventricular zone of injured brains, and functional analy‐sis of id1 suggested a role in maintaining the stem cell pool in this context. The elevation of id1 levels is inde‐pendent of inflammatory signals. Our results argue for multiple and antagonistic pathways in the control of reactive neurogenesis.

MATERIAL AND METHODS Fish maintenance and surgery Fish were bred and maintained as described [31]. The data were acquired from analysis of the wild‐type zebrafish strain AB2O2.Stab wounds were intro‐duced as described by [10]. All zebrafish husbandry and experimental procedures were performed in accor‐dance with the German law on Animal Protection and were approved by Regierungspräsidium Karlsruhe. Cloning, protein‐DNA Interaction and RNA sequencing Standard procedures were used for cloning and protein DNA interaction studies [32, 33] (suppl. Material and Methods). RNA sequencing and transcriptome analysis was carried out as described [32]. In situ analysis In situ hybridization on 50 μm vibratome sections was performed as described [4, 34]. In situ hybridization with digoxigenin and dinitrophenol‐11‐UTP (DNP‐11‐UTP) labeled probes was performed using tyramide amplification (TSA Plus Cyanine 3 System, Perkin Elmer, Boston, MA) [10, 13]. Immunostainings were performed on free‐floating vibratome‐sections [4]. Primary anti‐bodies: chicken anti‐GFP (1:1,000, Aves Labs), mouse anti‐PCNA (1:250, DAKO), rabbit anti‐S100 (1:500, DAKO), rat anti‐BrdU (1:400, Serotec), mouse anti‐glutamine‐synthase (1:500, Millipore), mouse anti‐PSA‐NCAM (1:500, Chemicon, Temecula, CA). Secondary antibodies: anti‐mouse, ‐rat, ‐rabbit and ‐chicken Alexa 488 or 594 conjugated antibodies (,1:1000, Invitrogen). Pictures were acquired with a Leica compound micro‐scope (DM5000B) or a laser scanning confocal micro‐scope Leica TCS2 SP5.For quantifications, 3 sections per brain were analysed. Cells were counted in 3 different areas of each section, the dorso‐medial, the dorsal and the dorsolateral ventricular zone. Mean fluorescent signal intensities of id1:EGFP expression were measured in the region of interest using ImageJ 1.44a. A paired t‐test of two datasets was performed to calculate the p‐value (equal variances). Significance is indicated in the figures by asterisks: *: 0.01 ≤ p < 0.05, **: 0,01 ≤ p < 0,001, ***: p < 0,001. Lipofection, injections and chemical treat‐ments Fish were anesthetized with Tricaine. Using a syringe needle (30 G) a hole was introduced into the skull at the telecephalic‐diencephalic‐junction. The solution was injected into the ventricular fluid with a glass capillary. For Lipofection a mix of 5 μl OptiMem medium, 5 μl Lipofectamine 2000 (Invitrogen) and 10 µl 1 µg/µl DNA diluted in OptiMem medium were used. In knock‐down experiments, 500 nl of vivo‐morpholinos (Genetools, Ltd) were injected at 250 μM. Id1 morpholino:



3

TAGGTCCCACAACTTTCATTTTGGC;5 bp mismatch mor‐pholino: TAGCTGCCAGAACTTTGATTTTCGC. For Zymosan A induction, 100 nl 10 mg/ml Zymosan A Bioparticles (Alexa‐488 conjugate, Invitrogen, Cat. No: Z‐23373) were injected [20]. BrdU was injected intrap‐eritoneally [35] 4, 5, 6 and 7 days after the vivo‐morpholino injection followed by a 3 weeks chase be‐fore fish were killed and further processed as described [13]. Fish were treated with 10 μM LY‐411575 or 0.1% DMSO in the swimming water at 28°C. Fish were kept in 15 mg/L dexamethasone or 0.1% methanol alone for 7 days. Stab wounds were inflicted after 2 d treatment followed by expression analysis 5 d after injury. RESULTS

Transcription regulatory genes regulated by injury of the telencephalon Differential expression of transcription regulators (TRs) is a hallmark of neurogenesis during early life stages [32, 36]. We reasoned that this may also be the case for adult regenerative neurogenesis. We employed in‐depth RNASeq to identify TR genes regulated by injury in the adult telencephalon. Wounds were inflicted by inserting a needle through the skull into the right hemi‐sphere of the telencephalon [10]. Brains were prepared 5 days post‐lesion (dpl), when the proliferation marker PCNA was maximally expressed [10]. Since the prolifera‐tive response was previously shown to be confined en‐tirely to the injured hemisphere [10], we used the unin‐jured side to determine the baseline of RNA expression. In each of the 3 sequencing repeats, 3 wounded hemi‐spheres and 3 uninjured hemispheres were pooled in test and control samples, respectively. We generated more than 800 Million 100 nt paired‐end reads from wild type and regenerating adult telencephala (suppl. Tab. 1) and assessed the expression of 28391 genes in control and wounded hemispheres, providing a com‐prehensive view into the transcriptome of the regener‐ating zebrafish brain. Among the misregulated genes, 279 are TR genes based on InterPro protein domain predictions [32]. We found that most of these TR genes were induced upon injury (252 TR) while expression of 27 TR genes was decreased compared to the control hemisphere (suppl. Tab. 2). We searched for signaling pathways and cellular processes known to converge on the 279 TR genes. The most prominent ones were im‐mune response (jun, runx3, nfatc2, stat4, gata3, nfkb2, stat6, stat1b), apoptosis (irf3, pkz, ddit3, tp53, nfkb2, atf3, mych) and Wnt signaling (tcf7l2, vim, pparda, tcf7l1b, dvl2, dvl3). To identify factors specifically expressed at the ventricu‐lar zone that may thus have a role in the regulation of neural stem cells, we investigated the expression of these TR genes by ISH on cross‐sections through the telencephalon 5 dpl on the left telencephalic hemi‐

sphere. As a source of TR cDNA probes for expression mapping, we used a collection of 2200 zebrafish TR cDNAs [32]. In the in situ analysis, we included also probes of genes that were not significantly regulated in the RNASeq analysis but had an expression in the em‐bryonic telencephalon. More than 60 TR genes were up‐regulated and 1 gene (neurod6b) was down‐regulated at the ventricular zone upon injury (suppl. Tab. 3). Amongst those, we identified inhibitor of DNA binding 1 (id1), doublesex and mab‐3 related transcription factor like family A2 (dmrta2), leucine zipper, putative tumor suppressor 2a (lzts2a), notch homolog 3 (notch3), the bHLH domain‐encoding hairy‐related 4.5 (her4.5), achaete‐scute complex‐like 1a (ascl1a), MAD homolog 5 (smad5), POU class 3 homeobox 3a (pou3f3a), empty spiracles homeobox 3 (emx3), SRY‐box containing gene 2 (sox2), 9a (sox9a) and19b (sox19b) genes to be up‐regulated only at the ventricular zone upon lesioning (Fig. 1, suppl. Tab. 3). The expression of the remaining genes was also affected in other regions in addition to the ventricular zone (suppl. Tab. 3). The injury‐regulated genes are enriched for particular classes of TRs such as the HMG box containing Sox (sox3, sox9a, sox19b) or bHLH domain containg factors (IPR011598; mycn, ascl1a, tfeb, her4.3, id1, nhlh2, tal2, neurod6b, BX088691.1). A number of genes are linked to Notch signaling. These include the Notch receptor 3 as well as downstream bHLH genes known to be regulated by Notch signaling (her4.3, her 4.5, ascl1a). Id1 is maximally expressed at 5 days post‐lesion In mammals, Id1 interacts with Hes bHLH proteins and is involved in maintenance of adult neural stem cells as opposed to neurogenesis [27, 37]. Thus, the measured up‐regulation of id1 in the injured zebrafish telencepha‐lon, under conditions where neurogenesis is stimulated, was rather unexpected. We focused our subsequent analysis on id1. With the exception of the rostral migra‐tory stream (arrows, Fig. 2A), id1 mRNA was expressed throughout the ventricular domains of the telencepha‐lon (Fig. 2A‐F): High levels were detected in the dor‐somedial (Dm, Fig. 2A, D) as well as the dorsolateral ventricular zone (Dl, Fig. 2A, F) of the pallium. Expres‐sion of id1 in the ventricular zone of the subpallium (Vv, Fig. 2A, B) and the dorsal region of the pallium (Dd, Fig. 2A, E) was less abundant. To assess the kinetics of increased id1 expression upon injury, we performed ISH with id1 antisense probe on wounded brains at different time points. No change in expression was detected at 1 dpl in comparison to the uninjured site (Fig. 2G). At 3 dpl, a slight up‐regulation of id1 was detectable on the injured side (Fig. 2H), which became clearly obvious by 5 dpl (Fig. 2I). id1 ex‐pression then decreased at 7 dpl (Fig. 2J) and reached baseline levels at 14 dpl relative to the uninjured hemi‐sphere (Fig. 2K). Thus, in comparison to Gata3 induc‐tion, which is up‐regulated already at 1 dpl [21], the



4

kinetics of id1 induction is slower and follows the profile of PCNA activation in RGCs [10, 12, 30]. Id1 is predominantly expressed in quiescent type I stem cells We next asked which cells express id1 along the telen‐cephalic ventricular zone. We performed ISH with id1 antisense RNA probe in combination with immunohis‐tochemical staining for the stem cell marker S100β and PCNA labeling cells in cycle, thereby distinguishing be‐tween type I (PCNA‐, S100β+) and type II stem cells (PCNA+, S100β+) [6]. id1 is predominantly expressed in S100β+/PCNA‐ type I cells (85.1% ± 4.9%, n=7 animals), while only 14.9% ± 4.9% of the S100β+/PCNA+ type II cells are positive for id1 mRNA (Fig. 3A‐E). We did not detect id1 expression in PCNA+ and S100β‐ type III neuroblasts (data not shown). Thus, id1 expression in the ventricular zone of the telencephalon is predomi‐nantly associated with non‐proliferating RGCs. We next asked whether this pattern of expression is changed in response to injury. To this end, we inflicted a wound in the telencephalon and stained sections 5 dpI with id1 antisense probe and antibodies against S100β and PCNA. Overall, there were more id1‐positive cells in the injured hemisphere (suppl. Fig 1G) in com‐parison to the control side (suppl. Fig 1B). id1‐expressing cells co‐expressed S100β but few of the id1+/S100β+ cells expressed also PCNA (suppl. Fig 1B‐J). Thus, id1 was predominantly expressed in PCNA‐

/S100β+ non‐cycling RGCs also in the injured brain (Fig. 3E, suppl. Fig 1F). These results suggest that injury leads to a transient expansion of quiescent (S100β+/PCNA‐) stem cells. The proportion of id1+ cycling and quiescent RGCs remains, however, constant in injured and unin‐jured hemispheres (compare Fig. 3E with suppl. Fig 1F). To confirm these observations, we generated a trans‐genic line (Tg(id1:EGFP)) using a BAC clone of the id1 locus to drive enhanced green fluorescent protein (EGFP) expression. EGFP faithfully reports expression of the endogenous id1 gene in the ventricular zone of the telencephalon (suppl. Fig. 2). Immunohistochemistry showed that 96.4% ± 2.2% of the EGFP‐positive cells express S100β while only 7.4% ± 2.4% of the EGFP‐positive cells express PCNA (Fig. 3F‐J) confirming that id1 is predominantly expressed in quiescent RGCs. Moreover, injury induced an up‐regulation of EGFP in the wounded hemisphere (Fig. 3K‐N) and did not change the predominant expression of id1 in quiescent S100β+/PCNA‐ RGCs (Fig. 3X). Expression of id1 was only found at a low level in a small proportion of dividing type II cells (S100β+/PCNA+) (yellow arrow, Fig. 3O‐R), while most type II cells lacked id1 expression (white arrows, Fig. 3O‐R, Fig. 3X). Taken together, these data fully support the notion that id1 is predominantly ex‐pressed in quiescent type I (S100β+/PCNA‐) stem cells during both constitutive and reactive neurogenesis. Another crucial question is whether the increase in id1 mRNA at the ventricular zone is only due to an overall

increase in id1‐expressing cells or also to an increase of the expression levels in individual cells. Type I cells ex‐press higher levels of id1 mRNA in injured brains (suppl. Fig. 1A). We also found that injury induced higher levels of id1:EGFP expression in individual cells of the injured hemisphere (Fig. 3W, Z) in addition to the significant increase in the number of EGFP‐positive cells upon in‐jury (Fig. 3Y). These results demonstrate that both the number of id1‐expressing cells as well as the level of id1 expression in individual cells was higher in the injured hemisphere. id1 confers quiescence to stem cells The selective expression of id1 in type I cells suggests that it may confer quiescence to RGCs. To test this hy‐pothesis, an id1 expression construct (CMV:id1‐T2A‐EGFP) was lipofected into the cells lining the telen‐cephalic ventricle. The id1 coding sequence was fused to a self‐cleaving T2A peptide [38, 39] and EGFP under the control of the CMV promoter. The cycling activity of transfected RGCs was analyzed 24 h later by co‐staining with EGFP, S100β and PCNA antibodies (Fig. 4A–D’). A more than 10‐fold lower proportion of PCNA‐positive cells was observed among S100β/Id1‐T2A‐EGFP double positive cells (0.4% ± 0.6, n=8 animals) compared to cells transfected with GFP alone (4.4% ± 1.6, n=8 ani‐mals; Fig. 4O). This suggests that id1 increases the oc‐currence of quiescent cells within the RGC population. Knock‐down of id1 induces proliferation of RGCs and increases neurogenesis To determine whether Id1 could be necessary for RGC quiescence, we asked whether loss of id1 function would lead to increased proliferation of RGCs. We knocked‐down id1 by ventricular injection of a vivo‐morpholino targeting id1 (id1 MO). The knock‐down of id1 led to an overall increase of PCNA‐positive cells (557 ± 44 cells per section, n=6 animals; Fig. 5A‐C, G) in com‐parison to a control morpholino (Ctrl MO; 320 ± 57 cells, n=6 animals; Fig. 5D‐F, G) and to uninjected con‐trols (305 ± 37 cells, n=6 animals; Fig. 5G) at 5 dpl. Also, we could detect a 1.9‐fold higher number of PCNA‐positive type II cells, while the overall RGC number re‐mained unchanged (Fig. 5G). Thus, it appears that knock‐down of id1 shifts RGC from a quiescent into a proliferating state. Morpholinos can cause cell death unspecifically [40]. Apoptotic cell death was not in‐creased after ventricular injection of morpholinos, as assessed by TUNEL staining (data not shown). In sum‐mary, these data support the notion that id1 promotes the quiescence of RGCs in the zebrafish telencephalon. To test if the activation of RGCs after id1 knock‐down leads to increased neurogenesis, we first investigated the expression of PSA‐NCAM, a marker for neuronal precursors (type III cells). We detected a significant in‐crease of PSA‐NCAM in the periventricular region 5 days after id1 knock‐down (suppl. Fig. 3A, B). To assess whether this leads also to an increased production of



5

new neurons, we performed BrdU lineage‐tracing to‐gether with HuC/D staining specific for differentiated neurons. We injected BrdU intraperitoneally at 4, 5, 6 and 7 days after id1 MO injection and sacrificed the fish 3 weeks later. More BrdU+/HuC/D+ cells were detected in id1 MO injected brains (43.7 ± 4.9 cells, n=6 animals) in comparison to the Ctrl MO (27.5 ± 3.3 cells, n=6 ani‐mals) and PBS injected brains (25.0 ± 3.0 cells; Fig. 5H and suppl. Fig. 3C‐H). Thus, knock‐down of id1 leads to a net increase in neurogenesis. Together, these results indicate that id1 is involved in maintaining quiescence of RGCs. Thus, the induction of id1 in response to injury may serve to limit the prolif‐erative response of RGCs in the repair process. To ad‐dress this, the id1 MO was injected into the ventricular fluid at 3 dpI. At 6 dpI, the proliferation state of RGCs was assessed by immunohistochemistry with PCNA and S100β antibodies. The percentage of type II cells among the total RGC number (41.9% ± 7.0% cells, n=6 animals) was 2.4‐fold higher in the id1 knock‐down in compari‐son to the Ctrl MO injected animals (16.5% ± 1.5%, n=6 animals, Fig. 5I, suppl. Fig. 3I‐N) or PBS injection con‐trols (17.7% ± 5.0%, n=6 animals, Fig. 5I). These results show that id1 also limits the activation of RGCs during reactive neurogenesis. Id1 and Gata3 are induced by distinct path‐ways in response to injury Inflammatory signals induce reactive neurogenesis and activation of Gata3 in the zebrafish telencephalon [20, 21]. We thus tested whether id1 expression is con‐trolled by inflammatory signals upon injury. To this aim, we assessed id1 expression following injection of Zymo‐san A. id1 levels were not altered in Zymosan A‐injected brains (n=6 animals, Fig. 6A, B) relative to solvent injec‐tion (n=6 animals, Fig. 6C, D), as visualized using in situ hybridization. In contrast, gata3 mRNA levels were clearly elevated (Fig. 6E‐H). Suppression of an inflam‐matory response by exposure to dexamethasone abol‐ished the induction of gata3 after lesion [20]. Dexa‐methasone treatment, however, did not reduce the increase of id1 expression in the wounded hemisphere (n=6 animals, Fig. 6I, J). In contrast, injury‐induced gata3 expression (n=5 animals) was almost completely abolished by dexamethasone treatment (Fig. 6K, L). The effectiveness of dexamethasone treatment was con‐firmed by qPCR of gata3 (3 dpl) and the dexa‐methasone‐inducible FKBP5 mRNA (5 dpl) (data not shown) [41, 42]. Hence, increased expression of id1 in type I stem cells in response to injury does not rely on inflammatory signals. Activation of the Notch pathway induces quiescence in the uninjured brain [15]. Hence, we tested whether chemical inhibition of Notch signaling with LY‐411575 (LY) [15, 18] would affect id1 expression. Bath treat‐ment of Tg(id1:EGFP) zebrafish in 10 µM LY did not change the number of id1:EGFP‐positive cells in com‐parison to DMSO controls (Fig. 6M). Also, its restriction to quiescent type I cells is not changed in LY‐treated fish

(Fig. 6M). However, the same treatment led to a strong down‐regulation of her4.1 mRNA, demonstrating that the LY treatment was efficient (Fig. 6N, O, arrows). The same unresponsiveness was noted for expression of endogenous id1 mRNA after 48 h treatment (suppl. Fig. 4A‐D). id1 mRNA expression, however, appeared de‐creased after 5 days of treatment (suppl. Fig. 4A‐D), likely as a secondary consequence of RGC activation or division. As expected, LY triggered a massive increase in the number of proliferating type II and a reduction of quiescent type I cells after 48 hours and 5 days of treatment (suppl. Fig. 4E‐P) [22]. We also tested whether the induction of id1:gfp upon injury of the brain involves Notch signalling. Inhibition of Notch did not alter the induction of id1:gfp 5 days after injury of the telencephalon (data not shown). Thus, id1 expres‐sion appears to be controlled only indirectly by Notch signalling. Together, the results suggest multiple, dis‐tinct pathways in the regulation of regenerative neuro‐genesis. Interaction of id1 with bHLH transcription factors Id1 is a negative regulator of basic helix‐loop‐helix pro‐teins by forming heteroduplexes, preventing their bind‐ing to target DNA. We selected potential Id1 interaction partners using the data generated by the ISH screen (suppl. Tab. 3) [43]. bHLH factors her6, her9, her4.1, her4.5 and ascl1a were expressed in the ventricular zone of the telencephalon, while neuroD expression was mainly detected in the pallium, in the parenchyma and along the periventricular layer. Tcf3, an E12/47/E2A‐like protein known to interact with bHLH factors is ubiquitously expressed. We investigated whether Id1 protein can interact with these transcrip‐tional regulators by pull‐down studies. We synthesized Id1‐GFP protein and the bHLH transcriptional regulators fused to GST in vitro. Id1‐GFP was strongly pulled down by Tcf3, Her4.1, Her4.5, Her6 and Her9 (Fig. 7A). How‐ever, Ascl1a and NeuroD interacted only very weakly with Id1‐GFP, and β‐Actin as negative control did not pull down Id1‐GFP at all (Fig. 7A). The GFP tag was not responsible for the pull‐down, as it did not reveal an interaction with the bHLH proteins (Fig. 7B). Thus, Id1 interacts with Her4.1, Her4.5, Her6, Her9, and most strongly with Tcf3. However, only very weak interac‐tions were observed with the neuronal specification proteins Ascl1 and NeuroD. DISCUSSION

Changes in the expression of TR genes are a hallmark of neurogenesis during early life stages. Here, we ana‐lysed, at a transcriptome level, the response of TR genes to injury of the telencephalon in the adult zebraf‐ish. We identified 279 TR genes that responded to injury by changes in expression levels, representing about one tenth of all TR genes encoded in the zebrafish genome



6

[32]. Among these, 64 TR genes were altered in their expression at the ventricular zone and are thus candi‐date regulators of stem cell function in response to wounding and subsequent regeneration. We focused our further analysis on one of these factors, the HLH factor Id1. We provide evidence that Id1 limits the number of stem cells entering the cell cycle and thus indirectly also the number of new‐born neurons in re‐sponse to injury. We also show that id1 induction upon injury is independent of inflammation, which was re‐cently identified as a major driver of the neurogenesis response in this context. Thus, instead of a simple re‐orientation of progenitors towards regeneration upon injury, our data rather suggest that the balance be‐tween multiple and antagonistic regulatory inputs, which are controlled by independent pathways, fine‐tunes stem cell proliferation and fate during regenera‐tion. Transcriptional regulatory programmes of constitutive and reactive neurogenesis Given the substantial injury caused by stabbing the brain with a syringe needle, it is not too surprising that this triggers a massive immune response [20, 44], which is reflected in the type of TR genes induced by wound‐ing. In addition, inflammatory signals activate reactive neurogenesis [20]. We also observed linkage of some of the induced TR genes to the Wnt and PDGF pathways. With a Wnt reporter line [45], Wnt signaling was de‐tected in blood vessels in the injured and uninjured brain but could not be linked to neurogenesis (unpub‐lished). Another pathway over‐represented among the identified TR genes is indicated by genes linked to Notch signaling. These genes include the up‐regulation of the Notch3 receptor and genes downstream of it. The role of id1 in the ventricular zone of the adult zebrafish telencephalon In the injured telencephalon, the time course of in‐crease of id1 expression succeeded induction of gata3 [21] and roughly accompanied the described increase in proliferation at the ventricular zone [10–12, 30]. This suggests that the response of id1, unlike that of gata3, is not an immediate early response. In constitutive neu‐rogenesis, id1 is almost exclusively expressed in quies‐cent type I (S100β+, PCNA‐) and at low level in few type II RGCs (S100β+, PCNA+). Intriguingly, this specificity was maintained with a similar ratio upon injury in spite of an increase in the number of RGCs. Our results indicate that the increase of type I cells after injury is the result of type II cells being pushed back by Id1 into quies‐cence. This suggests the existence of effective mecha‐nisms to retain this ratio during reactive neurogenesis. Our gain‐ and loss‐of‐function experiments argue for a role of id1 in maintaining this balance by promoting RGC quiescence. In addition, upon Id1 abrogation, we observed an increase of newly born post‐mitotic neu‐rons which was roughly proportional to the number of

reactivated RGCs. This strongly supports a specific role for Id1 in the control of the neural stem cell quies‐cence/activation cycle. Taken together these results indicate that Id1 limits the proliferative response of RGCs both during constitutive and regenerative neuro‐genesis. It has been suggested in mouse that increased neural stem cell division was eventually followed by exhaustion [46, 47]. Thus, it is tempting to speculate that the function of Id1 may overall be relevant to maintaining a neural stem cell pool and homeostasis of the telencephalic germinal zone. Multiple signals contribute to reactive neuro‐genesis A central question is how stem cells are regulated to endow the zebrafish brain with the remarkable ability to completely regenerate lesions. Acute inflammatory signals like cysteinyl leukotriene or Zymosan A were sufficient for the initiation of proliferation in reactive neurogenesis [20]. Suppression of an inflammatory re‐sponse by glucocorticoids blocked proliferation of RGCs and expression of Gata3, an immediate early response gene in regeneration [20, 21]. To our surprise, id1 in‐duction in response to injury could not be suppressed by anti‐inflammatory drugs. Also in contrast to gata3, id1 was not induced by sterile inflammation. Thus, the increase of id1 expression upon wounding> is regulated by distinct mechanisms. Notch signaling via the Notch receptor 3 has been im‐plicated in the induction of quiescence in RGCs during constitutive neurogenesis in the telencephalon [15, 22]. This function seems conserved in the regenerating spi‐nal cord [48]. We observed up‐regulation of Notch sig‐naling components in the injured telencephalic hemi‐sphere such as notch3, and her4 [22]. Notch signaling was shown in different contexts to regulate id1 expres‐sion [49–51], and a simple explanation could thus be that id1 expression and function are downstream of Notch signaling. Upon treatment with a gamma‐secretase inhibitor for 48 h, id1 expression was however unaffected, while the known Notch target gene her4.1 was strongly suppressed and NSC activation was mas‐sive. After 5 days of Notch inhibition id1 expression was reduced. However, this is most likely the indirect con‐sequence of the many RGCs having entered the cell cycle in the inhibitor treated animals. Thus, Id1 and Notch3 appear to belong to independent pathways con‐trolling the proliferative state of RGCs in the telen‐cephalon. They may nevertheless converge on the ex‐pression or function of Her factors. In the search for alternative inducers of id1 expression in the injured telencephalon, we tested BMP [52, 53], and pros‐taglandin E2 [54], known inducers of id1 expression in other tissues. None of these signals appeared to con‐tribute to the activation of id1 in the injured telen‐cephalon (unpublished). Irrespective of the precise signals, our data are consis‐tent with at least three pathways being implicated in the regulation of RGCs, comprising one stimulatory



7

pathway (inflammation) and two inhibitory pathways (Notch, Id1). Thus, rather than a linear series of events activated upon injury to stimulate neural stem cells for brain repair, our data suggest the existence of inde‐pendent negative feedback loops that moderate the neural stem cell response and may protect them from exhaustion. The relevance of these loops in vertebrate species where adult neurogenesis and regeneration are less efficient will be an interesting issue to address. Mechanism of Id1 action A central question is the mechanism by which Id1 pro‐motes neural stem cell quiescence. Triple knock‐down of the Id1, Id2 and Id3 genes in the mouse resulted in loss of anchorage of neural stem cells to their niche, loss of stemness and premature differentiation of neu‐ral stem cells [28]. It remains to be seen whether the stem cell niches are under similar control under reactive neurogenesis in the injured zebrafish telencephalon. Our in vitro binding studies show that zebrafish Id1 in‐teracts with Her factors and the ubiquitous co‐factor Tcf3. Although zebrafish Id1 interacted only very weakly with the proneural protein Ascl1a or the neural differ‐entiation protein NeuroD, it can still block activity of these proteins by binding to their common co‐factor Tcf3. Our data support the notion that zebrafish Id1 blocks the activity of bHLH proneural proteins and the bHLH neurogenesis inhibitors of the Her/Hes class ex‐pressed in the ventricular zone [37]. The function of Her factors in the adult zebrafish telencephalon has not been tested. In mouse fibroblasts in culture, Hes1 is a key component of the core quiescence programme [55, 56]. In the zebrafish adult telencephalon, her4 and her15 expression are downstream of Notch3 signaling ([22] and unpub.). Although these observations may seem at odds with the function of Id1 demonstrated here, we note that the interaction of Id1 with Hes1 in

mouse neural stem cells specifically affects the negative auto‐regulation of Hes1 on its own promoter, but not its interaction with other targets [37]. It is possible, therefore, that Id1 enhances the function of some Her factors in the adult zebrafish telencephalon. At these high levels of Id1 in RGCs in the injured hemisphere, Id1 could alternatively exert a general block of bHLH factors overriding all Notch signalling. ACKNOWLEDGMENTS

We thank N. Borel, M. Rastegar, C. Lederer, T. Beil, I. Baader, I. Foucher, M. Coolen, T. Dickmeis for technical support, advice or discussion. FINANCIAL DISCLOSURE

This work was supported by EU‐IP ZF‐Health, NeuroX‐sys, Interreg NSB‐Upper Rhine and EraSysBioPLUS to US, and by ERC AdG 322936 to LBC. The funders had no role in study design, data collection and analysis, decision to publish or in the preparation of the manuscript. COMPETING INTEREST

The authors declare no competing financial interests. AUTHOR CONTRIBUTIONS

S.R.: and U.S.: designed the experiments and supervised the work; R.R.V., N.D., M.F., J.E., A.A. and M.M.: con‐ducted the experiments; R.R.V., N.D., L.B., S.R. and U.S.: analyzed the data; O.A.: provided the RNA sequencing data and performed the bioinformatic analysis; R.R.V., S.R.: and U.S.: wrote the manuscript.

REFERENCES

1 Schmidt R, Strähle U, Scholpp S.

Neurogenesis in zebrafish ‐‐ from embryo to adult. NEURAL DEV 2013;8(1):3. Available at: http://www.ncbi.nlm.nih.gov/pubmed/23433260. Accessed February 28, 2013.

2 Kizil C, Dudczig S, Kyritsis N, et al. The chemokine receptor cxcr5 regulates the regenerative neurogenesis response in the adult zebrafish brain. NEURAL DEV 2012;7:27. Available at: http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=3441421&tool=pmcentrez&rendertype=abstract. Accessed March 11, 2013.

3 Zupanc GKH, Sîrbulescu RF. Adult neurogenesis and neuronal regeneration in the central nervous system of teleost fish. EUR J NEUROSCI 2011;34(6):917–29. Available at: http://www.ncbi.nlm.nih.gov/pubmed/21929625. Accessed October 16, 2012.

4 Adolf B, Chapouton P, Lam CS, et al. Conserved and acquired features of adult neurogenesis in the zebrafish telencephalon. DEV BIOL 2006;295(1):278–93. Available at: http://www.ncbi.nlm.nih.gov/pubmed/16828638. Accessed July 14, 2012.

5 Lindsey BW, Darabie A, Tropepe V. The cellular composition of neurogenic periventricular zones in the adult zebrafish forebrain. J COMP NEUROL 2012;520(10):2275–316. Available at: http://www.ncbi.nlm.nih.gov/pubmed/22318736. Accessed September 13, 2012.

6 März M, Chapouton P, Diotel N, et al. Heterogeneity in progenitor cell subtypes in the ventricular zone of the zebrafish adult telencephalon. GLIA 2010;58(7):870–88. Available at: http://www.ncbi.nlm.nih.gov/pubmed/20155821. Accessed August 2, 2012.

7 Pellegrini E, Mouriec K, Anglade I, et al. Identification of aromatase‐positive radial glial cells as progenitor cells in the ventricular layer of the forebrain in

zebrafish. J COMP NEUROL 2007;501(1):150–67. Available at: http://www.ncbi.nlm.nih.gov/pubmed/17206614. Accessed July 2, 2013.

8 Grandel H, Kaslin J, Ganz J, et al. Neural stem cells and neurogenesis in the adult zebrafish brain: origin, proliferation dynamics, migration and cell fate. DEV BIOL 2006;295(1):263–77. Available at: http://www.ncbi.nlm.nih.gov/pubmed/16682018. Accessed July 21, 2012.

9 Ganz J, Kaslin J, Hochmann S, et al. Heterogeneity and Fgf dependence of adult neural progenitors in the zebrafish telencephalon. GLIA 2010;58(11):1345–63. Available at: http://www.ncbi.nlm.nih.gov/pubmed/20607866. Accessed October 12, 2012.

10 März M, Schmidt R, Rastegar S, et al. Regenerative response following stab injury in the adult zebrafish telencephalon. DEV DYN 2011;240(9):2221–31. Available at: http://www.ncbi.nlm.nih.gov/pubmed/22016188. Accessed September 13, 2012.



8

11 Kroehne V, Freudenreich D, Hans S, et al. Regeneration of the adult zebrafish brain from neurogenic radial glia‐type progenitors. DEVELOPMENT 2011;138(22):4831–41. Available at: http://www.ncbi.nlm.nih.gov/pubmed/22007133. Accessed July 13, 2012.

12 Baumgart EV, Barbosa JS, Bally‐Cuif L, et al. Stab wound injury of the zebrafish telencephalon: a model for comparative analysis of reactive gliosis. GLIA 2012;60(3):343–57. Available at: http://www.ncbi.nlm.nih.gov/pubmed/22105794. Accessed September 13, 2012.

13 Lam CS, März M, Strähle U. Gfap and Nestin Reporter Lines Reveal Characteristics of Neural Progenitors in the Adult Zebrafish Brain. DEV DYN 2009;238(2):475–86. Available at: http://www.ncbi.nlm.nih.gov/pubmed/19161226. Accessed September 3, 2012.

14 Ludwin SK, Kosek JC, Eng LF. The topographical distribution of S‐100 and GFA proteins in the adult rat brain: an immunohistochemical study using horseradish peroxidase‐labelled antibodies. J COMP NEUROL 1976;165(2):197–207. Available at: http://www.ncbi.nlm.nih.gov/pubmed/1107363. Accessed August 12, 2013.

15 Chapouton P, Skupien P, Hesl B, et al. Notch activity levels control the balance between quiescence and recruitment of adult neural stem cells. J NEUROSCI 2010;30(23):7961–74. Available at: http://www.ncbi.nlm.nih.gov/pubmed/20534844. Accessed July 30, 2012.

16 Topp S, Stigloher C, Komisarczuk AZ, et al. Fgf signaling in the zebrafish adult brain: association of Fgf activity with ventricular zones but not cell proliferation. J COMP NEUROL 2008;510(4):422–39. Available at: http://www.ncbi.nlm.nih.gov/pubmed/18666124. Accessed August 1, 2013.

17 Diotel N, Page Y Le, Mouriec K, et al. Aromatase in the brain of teleost fish: expression, regulation and putative functions. FRONT NEUROENDOCR. 2010;31(2):172–92. Available at: http://www.ncbi.nlm.nih.gov/pubmed/20116395. Accessed October 6, 2012.

18 Rothenaigner I, Krecsmarik M, Hayes J a, et al. Clonal analysis by distinct viral vectors identifies bona fide neural stem cells in the adult zebrafish telencephalon and characterizes their division properties and fate. DEVELOPMENT 2011;138(8):1459–69. Available at: http://www.ncbi.nlm.nih.gov/pubmed/21367818. Accessed July 13, 2012.

19 Imayoshi I, Sakamoto M, Yamaguchi M, et al. Essential roles of Notch signaling in maintenance of neural stem cells in developing and adult brains. J NEUROSCI 2010;30(9):3489–98. Available at: http://www.ncbi.nlm.nih.gov/pubmed/20203209. Accessed August 10, 2012.

20 Kyritsis N, Kizil C, Zocher S, et al. Acute inflammation initiates the regenerative response in the adult zebrafish brain. SCIENCE 2012;338(6112):1353–6. Available at:

http://www.ncbi.nlm.nih.gov/pubmed/23138980. Accessed March 21, 2014.

21 Kizil C, Kyritsis N, Dudczig S, et al. Regenerative Neurogenesis from Neural Progenitor Cells Requires Injury‐Induced Expression of Gata3. DEVCELL 2012:1–8. Available at: http://linkinghub.elsevier.com/retrieve/pii/S1534580712004777. Accessed November 16, 2012.

22 Alunni A, Krecsmarik M, Bosco A, et al. Notch3 signaling gates cell cycle entry and limits neural stem cell amplification in the adult pallium. DEVELOPMENT 2013;140(16):3335–47. Available at: http://www.ncbi.nlm.nih.gov/pubmed/23863484. Accessed July 31, 2013.

23 Basak O, Giachino C, Fiorini E, et al. Neurogenic subventricular zone stem/progenitor cells are Notch1‐dependent in their active but not quiescent state. J NEUROSCI 2012;32(16):5654–66. Available at: http://www.ncbi.nlm.nih.gov/pubmed/22514327. Accessed March 6, 2013.

24 Ables JL, Decarolis NA, Johnson MA, et al. Notch1 is required for maintenance of the reservoir of adult hippocampal stem cells. J NEUROSCI 2010;30(31):10484–92. Available at: http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=2935844&tool=pmcentrez&rendertype=abstract. Accessed February 28, 2013.

25 Ehm O, Göritz C, Covic M, et al. RBPJkappa‐dependent signaling is essential for long‐term maintenance of neural stem cells in the adult hippocampus. J NEUROSCI 2010;30(41):13794–807. Available at: http://www.ncbi.nlm.nih.gov/pubmed/20943920. Accessed August 10, 2013.

26 Jung S, Park R‐H, Kim S, et al. Id proteins facilitate self‐renewal and proliferation of neural stem cells. STEM CELLS DEV 2010;19(6):831–41. Available at: http://www.ncbi.nlm.nih.gov/pubmed/19757990. Accessed April 23, 2013.

27 Nam H, Benezra R. High levels of Id1 expression define B1 type adult neural stem cells. CELL STEM CELL 2009;5(5):515–26. Available at: http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=2775820&tool=pmcentrez&rendertype=abstract. Accessed October 16, 2012.

28 Niola F, Zhao X, Singh D, et al. Id proteins synchronize stemness and anchorage to the niche of neural stem cells. NAT CELL BIOL 2012;14(5):477–87. Available at: http://www.ncbi.nlm.nih.gov/pubmed/22522171. Accessed October 9, 2012.

29 Gribble SL, Kim H‐S, Bonner J, et al. Tcf3 inhibits spinal cord neurogenesis by regulating sox4a expression. DEVELOPMENT 2009;136(5):781–9. Available at: http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=2685945&tool=pmcentrez&rendertype=abstract. Accessed August 13, 2013.

30 Kishimoto N, Shimizu K, Sawamoto K. Neuronal regeneration in a zebrafish model of adult brain injury. DIS MODEL

MECH 2012;5(2):200–9. Available at: http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=3291641&tool=pmcentrez&rendertype=abstract. Accessed October 9, 2012.

31 Westerfield M. The Zebrafish Book. A guide for the laboratory use of zebrafish (Danio rerio). 5th ed. Univ. of Oregon Press, Eugene; 2007.

32 Armant O, März M, Schmidt R, et al. Genome‐wide, whole mount in situ analysis of transcriptional regulators in zebrafish embryos. DEV BIOL 2013;380(2):351–362. Available at: http://www.ncbi.nlm.nih.gov/pubmed/23684812. Accessed May 21, 2013.

33 Green M, Sambrook J. Molecular Cloning: A Laboratory Manual. 4the editi. Cold Spring Harbor Laboratory; 2012:2028.

34 Schmidt R, Beil T, Strähle U, et al. Stab wound injury of the zebrafish adult telencephalon: a method to investigate vertebrate brain neurogenesis and regeneration. J VIS EXP 2014;(90). Available at: http://www.ncbi.nlm.nih.gov/pubmed/25146302. Accessed September 10, 2014.

35 Chapouton P, Adolf B, Leucht C, et al. Her5 Expression Reveals a Pool of Neural Stem Cells in the Adult Zebrafish Midbrain. DEVELOPMENT 2006;133(21):4293–303. Available at: http://www.ncbi.nlm.nih.gov/pubmed/17038515. Accessed September 13, 2012.

36 Ferg M, Armant O, Yang L, et al. Gene transcription in the zebrafish embryo: regulators and networks. BR. FUNCT GENOMICS 2014;13(2):131–43. Available at: http://www.ncbi.nlm.nih.gov/pubmed/24152666. Accessed March 24, 2014.

37 Bai G, Sheng N, Xie Z, et al. Id sustains Hes1 expression to inhibit precocious neurogenesis by releasing negative autoregulation of Hes1. DEV CELL 2007;13(2):283–97. Available at: http://www.ncbi.nlm.nih.gov/pubmed/17681138. Accessed October 16, 2012.

38 Provost E, Rhee J, Leach SD. Viral 2A peptides allow expression of multiple proteins from a single ORF in transgenic zebrafish embryos. GENESIS 2007;45(10):625–9. Available at: http://www.ncbi.nlm.nih.gov/pubmed/17941043. Accessed August 12, 2013.

39 Hans S, Freudenreich D, Geffarth M, et al. Generation of a non‐leaky heat shock‐inducible Cre line for conditional Cre/lox strategies in zebrafish. DEV DYN 2011;240(1):108–15. Available at: http://www.ncbi.nlm.nih.gov/pubmed/21117149. Accessed October 12, 2012.

40 Robu ME, Larson JD, Nasevicius A, et al. p53 activation by knockdown technologies. PLOS GENET 2007;3(5):e78. Available at: http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=1877875&tool=pmcentrez&rendertype=abstract. Accessed March 1, 2013.

41 Weger BD, Weger M, Nusser M, et al. A chemical screening system for glucocorticoid stress hormone signaling in an intact vertebrate. ACS CHEM BIOL



9

2012;7(7):1178–83. Available at: http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=3401037&tool=pmcentrez&rendertype=abstract. Accessed April 24, 2013.

42 Scharf SH, Liebl C, Binder EB, et al. Expression and regulation of the Fkbp5 gene in the adult mouse brain. PLOS ONE 2011;6(2):e16883. Available at: http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=3036725&tool=pmcentrez&rendertype=abstract. Accessed April 4, 2013.

43 Chapouton P, Webb KJ, Stigloher C, et al. Expression of hairy/enhancer of split genes in neural progenitors and neurogenesis domains of the adult zebrafish brain. J COMP NEUROL 2011;519(9):1748–69. Available at: http://www.ncbi.nlm.nih.gov/pubmed/21452233. Accessed September 13, 2012.

44 März M, Schmidt R, Rastegar S, et al. Expression of the transcription factor Olig2 in proliferating cells in the adult zebrafish telencephalon. DEV DYN 2010;239(12):3336–49. Available at: http://www.ncbi.nlm.nih.gov/pubmed/20981834. Accessed July 24, 2012.

45 Moro E, Ozhan‐Kizil G, Mongera A, et al. In vivo Wnt signaling tracing through a transgenic biosensor fish reveals novel activity domains. DEV BIOL 2012;366(2):327–40. Available at: http://www.ncbi.nlm.nih.gov/pubmed/22546689. Accessed August 7, 2013.

46 Encinas JM, Sierra A. Neural stem cell deforestation as the main force driving the age‐related decline in adult hippocampal neurogenesis. BEHAV BRAIN RES. 2012;227(2):433–9. Available at:

http://www.ncbi.nlm.nih.gov/pubmed/22019362. Accessed February 25, 2014.

47 Encinas JM, Michurina T V, Peunova N, et al. Division‐coupled astrocytic differentiation and age‐related depletion of neural stem cells in the adult hippocampus. CELL STEM CELL 2011;8(5):566–79. Available at: http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=3286186&tool=pmcentrez&rendertype=abstract. Accessed February 25, 2014.

48 Dias TB, Yang Y‐J, Ogai K, et al. Notch signaling controls generation of motor neurons in the lesioned spinal cord of adult zebrafish. J NEUROSCI 2012;32(9):3245–52. Available at: http://www.ncbi.nlm.nih.gov/pubmed/22378895. Accessed October 15, 2012.

49 Wang H‐C, Peng V, Zhao Y, et al. Enhanced Notch activation is advantageous but not essential for T cell lymphomagenesis in Id1 transgenic mice. PLOS ONE 2012;7(2):e32944. Available at: http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=3290631&tool=pmcentrez&rendertype=abstract. Accessed May 14, 2013.

50 Reynaud‐Deonauth S, Zhang H, Afouda A, et al. Notch signaling is involved in the regulation of Id3 gene transcription during Xenopus embryogenesis. DIFFERENTIATION 2002;69(4‐5):198–208. Available at: http://www.ncbi.nlm.nih.gov/pubmed/11841478. Accessed May 14, 2013.

51 Wang H‐C, Perry SS, Sun X‐H. Id1 attenuates Notch signaling and impairs T‐cell commitment by elevating Deltex1 expression. MOL CELL BIOL 2009;29(17):4640–52. Available at: http://www.pubmedcentral.nih.gov/artic

lerender.fcgi?artid=2725715&tool=pmcentrez&rendertype=abstract. Accessed April 22, 2013.

52 Hollnagel A, Oehlmann V, Heymer J, et al. Id genes are direct targets of bone morphogenetic protein induction in embryonic stem cells. J BIOL CHEM 1999;274(28):19838–45. Available at: http://www.ncbi.nlm.nih.gov/pubmed/10391928. Accessed May 14, 2013.

53 Katagiri T, Imada M, Yanai T, et al. Identification of a BMP‐responsive element in Id1, the gene for inhibition of myogenesis. GENES CELLS 2002;7(9):949–60. Available at: http://www.ncbi.nlm.nih.gov/pubmed/12296825. Accessed May 14, 2013.

54 Subbaramaiah K, Benezra R, Hudis C, et al. Cyclooxygenase‐2‐derived prostaglandin E2 stimulates Id‐1 transcription. J BIOL CHEM 2008;283(49):33955–68. Available at: http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=2662219&tool=pmcentrez&rendertype=abstract. Accessed February 22, 2013.

55 Sang L, Coller HA, Roberts JM. Control of the reversibility of cellular quiescence by the transcriptional repressor HES1. SCIENCE (80‐. ). 2008;321(5892):1095–100. Available at: http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=2721335&tool=pmcentrez&rendertype=abstract. Accessed February 4, 2014.

56 Sang L, Coller HA. Fear of commitment: Hes1 protects quiescent fibroblasts from irreversible cellular fates. CELL CYCLE 2009;8(14):2161–7. Available at: http://www.ncbi.nlm.nih.gov/pubmed/19587546. Accessed February 25, 2014.

See www.StemCells.com for supporting information available online. STEM

CELLS ; 00:000–000



10

Figure 1. In situ hybridization on sections of the adult zebrafish telencephalon 5 days after stab injury. (A) inhibitor of DNA binding 1 (id1), (B) doublesex and mab‐3 related transcription factor like family A2 (dmrta2), (C) leucine zipper, putative tumor suppressor 2a (lzts2a), (D) notch homolog 3 (notch3), (E) hairy‐related 4.3 (her4.3), (F) achaete‐scute complex‐like 1a (ascl1a), (G) MAD homolog 5 (smad5), (H) POU class 3 homeobox 3a (pou3f3a), (I) empty spiracles homeobox 3(emx3), (J‐L) SRY‐box containing gene sox2 (J), sox19b (K), sox9a (L). Arrows indicate up‐regulation of gene expression. Injury always in left hemisphere. Scale bar: 100 μm.



11

Figure 2. Expression of id1during adult constitutive and regenerative neurogenesis (A) Id1 is expressed at the ventricular zone except in the region of the rostral migratory stream (arrows). (B‐F) Id1 expression at the ventricular zone of Vv (B), rostral migratory stream (C, box) Dm (D), Dd (E) and Dl (F). The ventricu‐lar region next to the rostral migratory stream is devoid of id1mRNA (box). (G‐K) In situ hybridization to id1 mRNA at different times after injury (lesion left hemisphere). Up‐regulation of id1 mRNA is indicated by arrowheads. The strongest up‐regulation was observed at 5 dpl (I). Abbreviations: Dd: Dorsal zone of the dorsal telencephalic area; Dl: lateral zone of the dorsal telencephalic area; Dm: medial zone of the dorsal telencephalic area; Vv: ventral nucleus of the ventral telencephalic area; Vd: dorsal nucleus of the ventral telencephalic area. Scale bar: 50 μm, A; 15 μm, B‐F; 120 μm, G‐K. n=10 animals for (A‐F), n=3 animals (G‐K).



12

Figure 3. Id1 is predominantly expressed in type I RGCs. (A‐D) Cross‐sections through the telencephalon were stained by fluorescent in situ hybridization with id1 antisense probe and immunostaining against PCNA and S100β. id1 is predominantly expressed in non‐proliferating type I cells (S100β+/PCNA‐, yellow arrow) and is mostly absent in proliferating type II cells (S100β+/PCNA+, white arrows). (E) Quantification of id1 expression in type I and type II cells in unlesioned brains. (F‐I, K‐W): Immunohistochemistry on Tg(id1:EGFP) brains revealing GFP, PCNA and S100β. (F‐I) The transgenic line confirms that id1 is predominantly expressed in quiescent type I cells. (J) Quantification of PCNA and S100β expres‐sion in id1:EGFP‐positive cells. (K‐N) id1 is up‐regulated upon injury in the lesioned hemisphere (left). (O‐V) High magnification images of (K‐N). The lesioned hemisphere (O‐R) shows an up‐regulation of EGFP, PCNA and S100β in comparison to the unlesioned hemisphere (S‐V). (O‐R) After stab wounding, id1 is predominantly expressed in type I cells. Only very few PCNA expressing cells co‐express EGFP (yellow arrow), while most of them are EGFP‐ (white arrows). (W) High magnification of (K). Cells in lesioned hemisphere (left) have higher EGFP expression (white arrow) in com‐parison to cells in the unlesioned hemisphere (yellow arrow). (X) Quantification of id1expression in type I and type II cells in the control and the lesioned hemisphere. The propor‐tion of id1+/S100β+/PCNA‐ type I and id1+/S100β+/PCNA+ type II stem cells is not altered in the injured hemisphere relative to the control hemisphere of the telencephalon. (Y‐Z) Quantification of EGFP‐positive cells upon injury. The number of EGFP‐expressing cells (Y) and the intensity of EGFP in single cells (Z) are both increased following injury. Bars: average ± standard deviation (for details see Material and Methods). P‐values: (Y) p = 0.00075; (Z) p = 0.01. Scale bar: 10 μm, A‐D; 30 μm, F‐I, O‐V; 90 μm, K‐N; 40 μm, W. n=7 (E) and n=4 animals (J, X‐Z).



13



14

Figure 4. Expression of id1 pushes radial glial cells into quiescence. (A‐D’) An id1‐EGFP expression construct (A‐ D) or a GFP control vector (A’‐D’) was lipofected into cells lining the ven‐tricle. Immunohistochemistry against EGFP (A, A’), PCNA (B, B’) and S100β (C, C’) and merged panels A to C (D) and merged panels A’ to C’ (D’). Type II cells are less frequent within the population of RGCs expressing id1‐EGFP than among RGCs expressing EGFP alone (D, D’, yellow arrow). (O) Quantification of the proportion of PCNA‐positive cells within lipofected RGCs. Scale bar: 25 µm. n=8 animals for each transfection (O). Columns: average ± standard devia‐tion. P‐value: p= 0.00087.



15



16

Figure 5. Knock‐down of id1 leads to increased proliferation of RGCs and birth of neurons. (A‐F) More RGCs proliferate after injection of the vivo‐morpholino targeting id1 (Id1 MO) (A‐C) in comparison to the 5 base pair mismatch control injection (Ctrl MO) (D‐F). Immunohistochemistry of telencephalic cross‐sections with an‐tibodies against PCNA (A, D) and S100β (B, E) (merged panels: C, F). (G) Quantification of knock‐down experiment shows that there are more proliferating cells (PCNA+) overall and more S100β/PCNA double positive type II cells when id1 is knocked‐down in comparison to controls. The overall number of RGCs (S100β+/Draq5+) is not increased. (H) BrdU lineage tracing shows that there are significantly more BrdU+/HuC/D+ double positive neurons three weeks after Id1 MO injection (3 wpi) in comparison to the Ctrl MO and the PBS control. Thus, knock‐down of id1 leads to the gen‐eration of an increased number of neurons. (I) Id1 knock down upon injury leads to an increased proportion of type II cells among the RGC population at 6 dpl in comparison to the Ctrl MO and PBS control. Injections were performed at 3 dpl and fish were sacrificed at 6 dpl. Scale bar: 25 µm. n=6 animals (G‐I). Columns: average ± standard deviation. P‐values: (G) PCNA columns: for uninjected vs. id1 MO injected p = 0.00028; for ctrl. MO vs. id1 MO injected p = 0.00043; for uninjected vs ctrl. MO injected p = 0,6. PCNA + S100 columns: for uninjected vs. id1 MO injected p = 0.000027; for ctrl. MO vs. id1 MO injected p = 0.00017; for uninjected vs ctrl. MO injected p = 0,4. (H) For PBS vs. id1 MO injected p = 0.00045; for ctrl. MO vs. id1 MO injected p = 0.00034; for PBS vs ctrl. MO injected p = 0,22. (I) For PBS vs. id1 MO injected p = 0.016; for ctrl. MO vs. id1 MO injected p = 0.0088; for PBS vs ctrl. MO injected p = 0,17.



17

Figure 6. Inflammation does not induce increased id1 expression. Sterile infection by zymosan A injection does not induce id1 expression (A–D), but leads to increased expression of the control gene gata3 (E–H). (A, B, E, F) Zymosan A; (C, D, G, H) PBS injection. Anti‐inflammatory treatment with dexamethasone does not decrease the up‐regulation of id1expression upon injury (J) in comparison to the methanol control (I). The expression of the control gene gata3 is decreased upon dexamethasone treatment (K, L). (M‐O) Inhibition of Notch signaling does not alter id1 expression. (M) Quantification of id1:EGFP+ cells and id1:EGFP+/PCNA+ cells upon LY‐411575 treatment. No change in cell number was observed in comparison to the 0.3% (v/v) DMSO control. (N‐O) Expression of the Notch target her4.1 is com‐pletely abolished upon Notch inhibition (arrows). Scale bar: 180 μm in A, C, E, G, I, J, K, L, N, O; 50 μm in B, D, F, H. n = 6 animals for (A‐J), 5 for (K), 4 for (L). Data are represented as average ± standard deviation. n = 3 animals for DMSO control and LY411575 treatment.



18



19

Figure 7. Id1 protein interacts with Her4.1, Her4.5, Tcf3, Her6 and Her9. In vitro pull‐down assay showing that GST‐Her4.1 (lane 1), ‐Her4.5 (lane 2), ‐Tcf3 (lane 5), ‐Her6 (lane 6) and ‐Her9 (lane 7) interact with Id1‐GFP (A) but not with GFP alone (B). GST‐β‐Actin (lane 4) does not pull down Id1‐GFP, and GST‐Ascl1a (lane 3) and GST‐NeuroD (lane 8) re‐sults in a weak Id1‐GFP pull‐down. The input proteins Id1‐GFP (A) and GFP (B) were loaded as reference in lanes 9. M: molecular weight marker in kDa (left side).

The helix-loop-helix protein Id1 controls stem cell proliferation during regenerative neurogenesis in the adult zebrafish telencephalon

Documents