Corrections and Retraction

CORRECTIONS

CELL BIOLOGYCorrection for “A small molecule accelerates neuronal differ-entiation in the adult rat,” by Heiko Wurdak, Shoutian Zhu,Kyung Hoon Min, Lindsey Aimone, Luke L. Lairson, JamesWatson, Gregory Chopiuk, James Demas, Bradley Charette,Eranthie Weerapana, Benjamin F. Cravatt, Hollis T. Cline, EricC. Peters, Jay Zhang, John R. Walker, Chunlei Wu, JonathanChang, Tove Tuntland, Charles Y. Cho, and Peter G. Schultz,which appeared in issue 38, September 21, 2010, of Proc NatlAcad Sci USA (107:16542–16547; first published September 7,2010; 10.1073/pnas.1010300107).The authors note that Rajkumar Halder should be added

to the author line between Bradley Charette and EranthieWeerapana. Rajkumar Halder should be credited with contrib-uting new reagents/analytic tools. The online version has beencorrected. The corrected author and affiliation lines, and authorcontributions appear below.HeikoWurdaka,1, Shoutian Zhua,1, KyungHoonMina,2, Lindsey

Aimoneb, Luke L. Lairsona, James Watsonb, Gregory Chopiukb,James Demasc,3, Bradley Charettea, Rajkumar Haldera, EranthieWeerapanaa, Benjamin F. Cravatta, Hollis T. Clinec, Eric C.Petersb, Jay Zhangb, John R. Walkerb, Chunlei Wub, JonathanChangb, Tove Tuntlandb, Charles Y. Chob, and Peter G. Schultza,4

aThe Skaggs Institute of Chemical Biology and Department of Chemistry,The Scripps Research Institute, La Jolla, CA 92037; bGenomics Institute ofthe Novartis Research Foundation, San Diego, CA 92121; and cDepartment ofCell Biology, The Scripps Research Institute, La Jolla, CA 92037

Author contributions: H.W., S.Z., L.L.L., B.C., B.F.C., T.T., C.Y.C., and P.G.S. designedresearch; H.W., S.Z., K.H.M., L.A., L.L.L., J.W., G.C., E.W., E.C.P., J.Z., and J.C. performedresearch; K.H.M., G.C., R.H., and H.T.C. contributed new reagents/analytic tools; H.W., S.Z.,L.L.L., J.D., B.C., E.C.P., J.R.W., C.W., T.T., and P.G.S. analyzed data; and H.W., S.Z., andP.G.S. wrote the paper.

www.pnas.org/cgi/doi/10.1073/pnas.1016908108

GENETICSCorrection for “SETDOMAINGROUP2 is themajor histoneH3lysie 4 trimethyltransferase in Arabidopsis,” by Lin Guo, YanchunYu, Julie A. Law, and Xiaoyu Zhang, which appeared in issue 43,October 26, 2010, of Proc Natl Acad Sci USA (107:18557–18562;first published October 11, 2010; 10.1073/pnas.1010478107).The authors note that, due to a printer’s error, the manuscript

title “SET DOMAIN GROUP2 is the major histone H3 lysie 4trimethyltransferase in Arabidopsis” should instead appear as“SET DOMAIN GROUP2 is the major histone H3 lysine 4 tri-methyltransferase in Arabidopsis.” The online version has beencorrected.

www.pnas.org/cgi/doi/10.1073/pnas.1016620107

RETRACTION

MEDICAL SCIENCESRetraction for “Wnt-5a signaling restores tamoxifen sensitivity inestrogen receptor-negative breast cancer cells,” by Caroline E.Ford, Elin J. Ekström, and Tommy Andersson, which appearedin issue 10, March 10, 2009, of Proc Natl Acad Sci USA(106:3919–3924; first published February 23, 2009; 10.1073/pnas.0809516106).The authors wish to note the following: “During efforts to

extend this work, we re-examined the laboratory records for allfigures and found that the Excel files on which Fig. 4C was basedcontained serious calculation errors; the first author of the papertakes full responsibility for these inaccuracies. Considering theimportance of this figure for the conclusions drawn, the authorshereby retract the work. We apologize for any inconvenience thismay have caused.”

Caroline E. FordElin J. Ekström

Tommy Andersson

www.pnas.org/cgi/doi/10.1073/pnas.1017549108

22360 | PNAS | December 21, 2010 | vol. 107 | no. 51 www.pnas.org

A small molecule accelerates neuronal differentiationin the adult ratHeiko Wurdaka,1, Shoutian Zhua,1, Kyung Hoon Mina,2, Lindsey Aimoneb, Luke L. Lairsona, James Watsonb,Gregory Chopiukb, James Demasc,3, Bradley Charettea, Rajkumar Haldera, Eranthie Weerapanaa, Benjamin F. Cravatta,Hollis T. Clinec, Eric C. Petersb, Jay Zhangb, John R. Walkerb, Chunlei Wub, Jonathan Changb, Tove Tuntlandb,Charles Y. Chob, and Peter G. Schultza,4

aThe Skaggs Institute of Chemical Biology and Department of Chemistry, The Scripps Research Institute, La Jolla, CA 92037; bGenomics Institute of the NovartisResearch Foundation, San Diego, CA 92121; and cDepartment of Cell Biology, The Scripps Research Institute, La Jolla, CA 92037

Contributed by Peter G. Schultz, July 20, 2010 (sent for review June 8, 2010)

Adult neurogenesis occurs in mammals and provides a mechanismfor continuous neural plasticity in the brain. However, little is knownabout the molecular mechanisms regulating hippocampal neuralprogenitor cells (NPCs) and whether their fate can be pharmacolog-icallymodulated to improveneuralplasticity and regeneration.Here,we report the characterization of a small molecule (KHS101) thatselectively induces a neuronal differentiation phenotype. Mecha-nism of action studies revealed a link of KHS101 to cell cycle exit andspecific binding to the TACC3 protein, whose knockdown in NPCsrecapitulates the KHS101-induced phenotype. Upon systemic ad-ministration, KHS101 distributed to the brain and resulted ina significant increase inneuronaldifferentiation invivo.Ourfindingsindicate that KHS101 accelerates neuronal differentiation by in-teraction with TACC3 and may provide a basis for pharmacologicalintervention directed at endogenous NPCs.

adult neurogenesis | neural progenitor cell | Tacc3

The adult hippocampal dentate gyrus (DG) is a major neuro-genic region that harbors self-renewing neural progenitor cells

(NPCs). NPCs arise within the subgranular layer (SGL) of the DGand primarilymigrate into the adjacent granule cell layer (GCL) todifferentiate and mature into neurons (1). Although adult NPCshold substantial promise for neural repair after brain injury, theirmobilization is hampered by the lack of knowledge about the mo-lecular mechanisms controlling adult neurogenesis under normaland pathophysiological conditions. Pleiotropic signalingmolecules[e.g., retinoic acid (RA), Wnt, and brain-derived growth factor(BDNF)] as well as antidepressant and anticonvulsant drugs (e.g.,fluoxetin and sodiumvalproate) have been implicated in regulatingadult NPC proliferation, differentiation, or survival in vivo (2–8).Therefore, it may be possible to develop neurogenic agents forpharmacological intervention specifically directed at the endoge-nous NPC pool. Here, we describe the small molecule KHS101,which specifically induces neuronal differentiation in vitro and invivo upon systemic administration. Additionally, mechanism ofaction studies revealed a critical role for transforming acidic coiled-coil–containing protein 3 (TACC3) in the regulation of NPCmaintenance and differentiation.

ResultsKHS101 Induces Neuronal Differentiation in Cultured NPCs. We pre-viously reported a phenotypic screen that identified a group ofsynthetic 4-aminothiazole compounds, termed neuropathiazols,which induce neuronal differentiation of cultured rat hippocampalNPCs (9). Synthesis of several analogs of the original neuro-pathiazol structure and a focused structure–activity relationship(SAR) study afforded a molecule (KHS101) that possessedincreased activity and improved pharmacokinetic properties com-pared with the original compound (Fig. 1A, Fig. S1, Scheme S1,Table S1). KHS101 was further studied in vitro using establishedprotocols for the isolation, propagation, and differentiation ofadult rat hippocampalNPCs (3, 8–10).KHS101 increasedneuronal

differentiation of adherently cultured rat NPCs in a dose-de-pendent fashion (EC50 ∼ 1 μM) as assessed by quantitative reversetranscription (RT)-PCR for the neurogenic transcription factorNeuroD (11) and immunostaining for the panneuronal markerTuJ1 (Fig. 1 B–D, Fig. S1). KHS101-induced neuron formation(40–60% TuJ1+ cells at 1.5–5 μM KHS101) was also observedunder neurosphere-forming conditions (12) in secondary neuro-spheres derived fromboth the hippocampus and the subventricularzone (SVZ) of adult rats (Fig. S1). Moreover, hippocampal NPCstreated with KHS101 and cultured adherently on microelectrodearrays for 12 d exhibited neuronal morphologies as well as spon-taneous spiking activity, hence indicating the presence of func-tional, maturing neurons (Fig. S2).

KHS101 Suppresses Astrocyte Formation in Cultured NPCs. To furtheranalyze the effect ofKHS101 on lineage-specific differentiation, weexposed rat NPCs to the astrocyte-inducing cytokine bone mor-phogenetic protein 4 (BMP4), which is expressed in the adultneurogenic niches and has been reported to inhibit hippocampalneurogenesis (13–16). As expected, BMP4 (50 ng/mL) treatmentfor 4 d increased differentiation to astrocytes (from 0.5 to 10%)withinNPCculturesasassessedby immunostaining for theastrocytemarker glial fibrillary acidic protein (GFAP). Addition of 2 μMRAdid not significantly alter BMP4-induced astrogenesis. However,KHS101 (5 μM) treatment decreased BMP4-induced astrocytedifferentiation>4-fold, while increasing neuronal differentiation ina dose-dependent fashion (Fig. 1 E–G). Thus, KHS101 promotesspecifically neuronal differentiation and can override the astrocyte-promoting BMP signal.

KHS101 Negatively Affects Cell Cycle Progression and Proliferation ofNPCs. To elucidate genes and pathways affected by KHS101, wecarried out microarray mRNA profiling experiments in adherentlycultured rat hippocampal NPCs that were treated with KHS101,the inactive derivativeKHS92,orDMSOcontrol. In addition to theexpected up-regulation of NeuroD, pathway analysis of differen-tially regulated genes revealed that KHS101 treatment primarily

Author contributions: H.W., S.Z., L.L.L., B.C., B.F.C., T.T., C.Y.C., and P.G.S. designed re-search; H.W., S.Z., K.H.M., L.A., L.L.L., J.W., G.C., E.W., E.C.P., J.Z., and J.C. performedresearch; K.H.M., G.C., R.H., and H.T.C. contributed new reagents/analytic tools; H.W.,S.Z., L.L.L., J.D., B.C., E.C.P., J.R.W., C.W., T.T., and P.G.S. analyzed data; and H.W., S.Z.,and P.G.S. wrote the paper.

The authors declare no conflict of interest.

Data deposition: The data reported in this paper have been deposited in the Gene Ex-pression Omnibus (GEO) database, www.ncbi.nlm.nih.gov/geo (accession no. GSE23668).1H.W. and S.Z. contributed equally to this work.2Present address: College of Pharmacy, Chung-Ang University, Dongjak-Gu, Seoul 156-756, South Korea.

3Present address: Physics Department, St. Olaf College, Northfield, MN 55057.4To whom correspondence should be addressed. E-mail: schultz@scripps.edu.

This article contains supporting information online at www.pnas.org/lookup/suppl/doi:10.1073/pnas.1010300107/-/DCSupplemental.

16542–16547 | PNAS | September 21, 2010 | vol. 107 | no. 38 www.pnas.org/cgi/doi/10.1073/pnas.1010300107

affects cell cycle regulatory networks in NPCs (Fig. 1H, Fig. S3).Several factors required for cell cycle progression (e.g., cyclins)were down-regulated, whereas microarray mRNA profiling andquantitative RT-PCR experiments showed a marked up-regulation(5-fold at 1.7 μMKHS101) of the negative cell cycle regulatorCdkn1(Fig. 1H and I). This result is consistentwith the notion thatCDKN1(also termed p21) plays a key role in regulating the balance betweenNPC proliferation and differentiation (8, 17–20). Analysis of theproliferation marker Ki67, the mitotic marker phospho-histone H3(P-HH3), and the NPC marker (sex-determining region Y)-box 2(SOX2), as well as secondary neurosphere formation assays, con-firmed that in thepresenceofKHS101 thevastmajorityofNPCs stopproliferating within 72 h, become mitotically inactive, and lose Sox2expression, thus shifting toward terminal differentiation (Fig. 1 J –Land Figs. S1 and S4). Interestingly, KHS101 decreased the pro-liferation of rat oligodendrocyte precursor cells from the optic nerve(21), but failed to induce myelin basic protein expression and oli-godendrocyte differentiation in a phenotypic screen, suggesting thatKHS101-induced differentiation depends on the cellular context.

KHS101 Physically Interacts with TACC3 Protein. To determine thespecific target ofKHS101 inNPCs, we generated a derivative of theparent compound containing the photocrosslinking benzophenonemoiety and an alkyne substituent (KHS101-BP, Fig. 2A andScheme S5). Upon irradiation this derivative is expected to form

a covalent bond between KHS101 and its target protein, allowingisolation of protein–compound complexes after labeling with re-porter tags such as biotin-azide. NPC lysate was incubated withKHS101-BP and proteins were separated by 2D SDS/PAGE afterUV irradiation and biotin-tag labeling. Western blot analysisidentified a distinct labeled protein, whose level was significantlyreduced by competition with a 50-fold excess of free KHS101 inindependent experiments (Fig. 2B). Mass spectrometry revealedthe 80-kDa protein to be TACC3 and its identity was confirmed byWestern blotting using a TACC3-specific antibody (Fig. 2C). Wefurther confirmed direct physical interaction between TACC3 andKHS101 in independent pulldown experiments using purifiedrecombinant rat TACC3 protein and a biotinylated KHS101 de-rivative (Fig. 2D, Table S1).TACC3was originally identified as amember of a protein family

with a highly conserved C-terminal coiled-coil domain and isthought to function as an important structural component of thecentrosome and mitotic spindle apparatus (22). Moreover, pre-vious expression studies of Tacc3 as well as loss-of-function studiesin mice, hematopoietic stem cells, and embryonic neural stem cellspoint to a role for Tacc3 in controlling progenitor cell expansionand terminal differentiation during development (23–27). Post-natally, TACC3 expression becomes restricted to the remainingproliferative tissues such as spleen, thymus, gastrointestinal (GI)tract, and cerebral areas including the hippocampus (23, 26).

Fig. 1. KHS101 specifically induces neuronal differentiation in rat NPCs. (A) Chemical structure of KHS101. (B) Real time RT-PCR analysis of NPCs treated withKHS101 for 24 h showing a dose-dependent induction of NeuroD mRNA expression normalized to the DMSO control. (C and D) TuJ1 staining (green) ofDMSO- (0.1%, C) and KHS101-treated (5 μM, D) NPCs after a 4-d differentiation period (see also Figs. S1 and S2). (E and F) GFAP (green) and TuJ1 staining (red)of NPCs cultured under astrocyte-inducing conditions (50 ng/mL BMP4) in presence of DMSO (0.1%, J) or KHS101 (5 μM, K). (G) GFAP+ and TuJ1+ cell per-centages show that KHS101 significantly suppresses BMP4-induced astrogenesis in NPCs, while significantly increasing neurogenesis. (H) Ingenuity pathwayanalysis reveals KHS101-induced up-regulation (red) of negative cell cycle regulators (Cdkn1 and Gadd45) and down-regulation of positive cell cycle regu-lators at 24 h of treatment (e.g., Ccnb1 and Ccne1; see also Fig. S3). (I) Real time RT-PCR analysis of NPCs treated with KHS101 for 24 h showing a dose-dependent induction of Cdkn1 mRNA expression, whereas inactive derivatives (KHS91, KHS92, and KHS103) fail to up-regulate Cdkn1 mRNA levelsnormalized to DMSO. (J and K) Ki67 (green) and P-HH3 staining (red) of DMSO- (0.1%, G) and KHS101-treated (5 μM, H) NPCs. (L) Ki67+ and P-HH3+ cellpercentages indicate a significant reduction in proliferation and mitotic activity in KHS-treated NPCs over time. (Scale bars: 20 μm.) Error bars, SDs (threeindependent experiments; biological replicates in B and I); statistical significance (t test), *P < 0.05, **P < 0.01; nuclei were visualized with DAPI (blue).

Wurdak et al. PNAS | September 21, 2010 | vol. 107 | no. 38 | 16543

CELL

BIOLO

GY

Knockdown of TACC3 Significantly Increases Neuronal Differentiationof NPCs. To address whetherTacc3 is also functionally implicated inthe regulationofadultNPCfate,weusedRNA interference (RNAi)to specifically down-regulate Tacc3 expression in adult rat hippo-campal NPCs. Two different shRNAs (shTacc3-1 and shTacc3-2)efficiently decreased Tacc3 RNA levels (60–90% reduction in in-dependent experiments) compared with the nontargeting controlshRNA (shCO) as assessed by quantitative RT-PCR and immuno-cytochemistry (Fig. 3 A–C, Fig. S4). Cdkn1 RNA levels markedlyincreased (∼4-fold, Fig. 3A) upon treatment with Tacc3-specificshRNAs, which is consistent withKHS101-induced up-regulation ofCdkn1 and with previous reports showing that TACC3 depletioncauses CDKN1-mediated cell cycle inhibition in fibroblasts (28).Furthermore, an overt neuronal differentiation phenotype anda concomitant reduction in NPCmarker immunoreactivity (Nestin,Sox2) were observed in NPC cultures ∼4 d after treatment withTacc3-specific shRNAs (Fig. 3 B–F); staining for TuJ1 and Ki67confirmed a significant increase in TuJ1+ neurons (∼4-fold) anddecrease in proliferation (≥3-fold) comparedwith shCO-expressingNPCs (Fig. 3G). In addition, Tacc3-specific shRNAs significantlysuppressed GFAP+ astrocyte formation and concomitantly pro-moted neuronal differentiation of NPCs cultured in the presenceofBMP4(Fig. 3H–J). Thus,TACC3 is required for themaintenanceof the undifferentiated NPC state in vitro and its knockdown issufficient to fully recapitulate the neurogenesis and astrocyte-suppressing phenotypes induced by KHS101. Although the contri-bution of other KHS101 binding partners cannot be excluded,TACC3 appears to be the functionally relevant target of KHS101in NPCs.

KHS101 Regulates the Nuclear Localization of the Transcription FactorARNT2. Recently, TACC3 was proposed to regulate cell fate deci-sions by controlling the subcellular localization of transcriptionfactors, including aryl-hydrocarbon receptor nuclear translocator2 (ARNT2) (24, 26). Interestingly, ARNT2 is the only knownTACC3-interacting protein that is almost exclusively expressed inneuronal brain cells throughout life (29); consistent with this reportARNT2 expression was detected within the DG by immunohisto-chemistry (Fig. S4). Moreover, the expression of several genesdownstreamofARNT2-mediated signalingwas affected byKHS101in cultured NPCs, including genes involved in aryl-hydrocarbonreceptor signaling (Fig. S3). To further investigate the putativeneurogenic mechanisms downstream of the TACC3-KHS101 in-teraction, we analyzed nuclear localization of ARNT2. Consistentwith a previous report (30), transient overexpression of rat TACC3diminished nuclear localization of ectopically expressed rat ARNT2in a concentration-dependent manner in 293T cells (Fig. 3K). Incontrast, KHS101 treatment led to increased nuclear but not cyto-plasmic levels of ARNT2 (Fig. 3L). Furthermore, treatment withKHS101 (5 μM, 12 h) orTacc3-specific shRNAs increased the levelsof endogenous nuclear ARNT2 in NPCs as quantified by confocalmicroscopy and image analysis (Fig. 3M and N, and Fig. S4). Next,we tested whether ARNT2 can regulate cell fate in adult hippo-campal NPCs by transient Arnt2 overexpression. Elevated levels ofARNT2were detected in the nuclei of NPCs but were insufficient toinduce cell cycle exit or NPC differentiation. However, ARNT2overexpression markedly favored neurogenesis and suppressed theastrocyte cell fate upon BMP4-induced differentiation (Fig. S5),suggesting that ARNT2 may provide a cell-intrinsic bias towardneuronal lineage specification of NPCs in response to glial-inducing signals.

KHS101 Significantly Increases Neuronal Differentiation in Vivo. Todetermine whether KHS101 can act as a pharmacological agent toinduce neuronal differentiation in vivo, we assessed the pharma-cokinetic properties of KHS101 including blood–brain barrierpenetration. Oral administration in vehicle resulted in very lowsystemic exposures; however, i.v. and s.c. doses of 6mg/kg KHS101resulted in reasonable plasma concentrations (>1.5 μM) witha plasma half-life of 1.1–1.4 h, and a relative bioavailability of 69%following s.c. dosing.Most importantly, the distribution ofKHS101to the brain was extensive as demonstrated by a brain-to-plasmaAUC(0–3h) ratio of∼8 (dosing: 3mg/kgKHS101, i.v.; Fig. 4A, TableS2). Consequently, to study the effect of KHS101 on neurogenesiswe injected adult rats with vehicle (5% EtOH in 15% Captisol) orKHS101 (s.c., 6 mg/kg, BID) for 14 d, while including a daily bro-modeoxyuridine (BrdU) regimen (200 mg/kg, intraperitoneal) forthe first 7 d to label dividing NPCs in the DG. We then examinedthe fate of thoseBrdU-labeled cells by assessing colocalizationwiththe astrocyte marker GFAP and the neuronal-specific nuclearprotein NeuN (Fig. 4 B and C, Fig. S6).No significant difference was detected in the percentage of

BrdU/GFAP double-positive cells upon KHS101 treatment.However, the majority of these cells (>80%) were nonstellate andlocated within the SGL, therefore likely representing a NPC sub-type (GFAP+ type I progenitors) (1), rather than differentiatedastrocytes (Fig. S6). Notably, we found a significant increase in thepercentage of BrdU/NeuN double-positive cells from∼20 to∼40%upon KHS101 dosing, indicating increased neuronal differentia-tion. Consistent with this result, the number of Ki67- and BrdU-positive cells significantly decreased within the SGL, indicatingreduced proliferation of NPCs in KHS101 compared with vehicle-treated animals (Fig. 4D–F andH). The remaining Ki67 and BrdUimmunoreactivity within the SGL of treated animals suggests thatthe KHS101 dosing regimen did not result in the exhaustion of theself-renewing progenitor pool. Moreover, KHS101 administrationdid not alterKi67 immunostaining in nonneuralTACC3-expressingtissues such as spleen and gut (Fig. S7), suggesting either a cell type-

Fig. 2. KHS101 specifically interacts with TACC3 protein. (A) Structure ofthe benzophenone-containing alkyne-tagged KHS101 conjugate (KHS101-BP) used for target identification. (B) Representative two-dimensional SDS/PAGE and Western blotting of NPC cell lysates (2 mg/mL) detecting protein-KHS101-BP complexes after photocrosslinking (1 h) and biotin-tag labeling(25 μM biotin-azide, Left). Unlabeled KHS101 served as a competitor forspecific KHS101-BP–protein binding (50-fold excess, Right). Independentexperiments identified a protein spot that was reproducibly competed byunlabeled KHS101 (arrowheads). Mass spectrometry revealed the 80-kDaprotein to be TACC3. (C) Western blot analysis of NPC lysate using a TACC3-specific antibody confirmed TACC3 identity after pulldown with KHS101-BP.(D) Recombinant rat TACC3 binds KHS101. Purified protein was incubatedwith biotinylated KHS101 (Table S1 and Schemes S3 and S4) in presence/absence of unlabeled compound, precipitated with streptavidin-coatedagarose beads, and then detected by silver staining of SDS/PAGE gels.

16544 | www.pnas.org/cgi/doi/10.1073/pnas.1010300107 Wurdak et al.

specific effect of KHS101 or different levels of exposure or clear-ance in different organs. Finally, we did not observe signs of leth-argy, weight loss, or other indicators of sickness in KHS101-treatedanimals during the study period and apoptosis was not alteredwithin theDGofKHS101- compared with vehicle-injected animalsas assessed by cleaved caspase 3 staining. Overall, these results areconsistent with our in vitro data and indicate a significant KHS101-induced acceleration of neuronal differentiation in vivo.

DiscussionEndogenous pools of NPCs are a potential source for tissue repairand regeneration in the central nervous system (CNS). NPCs areable to self-renew throughout life in a location-specificmanner andtheir fate may be modulated in response to pharmacological in-tervention with small molecules. NPC-mediated neurogenesisinvolves at least three different processes: NPC proliferation, dif-ferentiation, and the survival of NPCs committed toward a neuro-nal fate. Small molecules have been shown to modulate each ofthese distinct processes. For example, the antidepressant agent

fluoxetin has been shown to increase NPC proliferation in vivo (5),and the anticonvulsant drug valproate has been shown to increaseneuronal differentiation of NPCs in the adult rat dentate gyrus (3).A very recent study describes the identification of a proneurogenicsmall molecule that protects newborn hippocampal neurons fromapoptosis (31).Moreover, a series of experimental small moleculesthat control the self-renewal and differentiation of stem and pro-genitor cells in vitro have been described (3, 9, 32–36).This study shows that small molecules such as KHS101, isolated

fromphenotypic screens of chemical libraries, can bepowerful toolsto elucidate and modulate the biology of endogenous progenitorcells both in vitro and in vivo. KHS101 specifically promoted neu-ronal differentiation of NPCs, concomitantly suppressing pro-liferation. The levels of KHS101-inducedNPC differentiation werequalitatively comparable to those induced by the known neurogenicfactors RA and BDNF under adherent and sphere-forming con-ditions (Fig. S1), although the optimal concentrations varied (2–5 μM for KHS101, 2 μM for RA). However, unlike most knownneurogenic factors, KHS101 can override astrocyte-inducing cues

Fig. 3. Tacc3-specific shRNA recapitulates the neurogenic effect of KHS101 in rat NPCs. (A) Representative real time RT-PCR showing that Tacc3 RNA levelsare markedly decreased in NPCs after electroporation with Tacc3-specific shRNA constructs (shTacc3-1, shTacc3-2). Concomitantly, Cdkn1 mRNA levels areelevated compared with the nontargeting control (shCO). (B and C) Nestin (green) and TACC3 (red) immunopositivity is observed in shCO-expressing NPCs(B, note centrosomal TACC3 localization) and strongly reduced upon TACC3 knockdown (C) (Fig. S4). (D) Tacc3-specific shRNA causes a TuJ1+ neuronalphenotype in NPCs. (E and F) Staining for Ki67 (red) and TuJ1 (green) in shCO- (E) and shTacc3-expressing NPCs (F) at 4 d after shRNA electroporation. (G) TuJ1+/Ki67− and Ki67+ cell percentages indicate significantly increased neurogenesis and significantly decreased proliferation upon Tacc3 RNAi in NPCs. (H and I)GFAP (green) and TuJ1 (red) staining in shCO- (H) and shTacc3-expressing (I) NPCs cultured with BMP4 (50 ng/mL) for 4 d. (J) GFAP+ and TuJ1+ cell percentagesindicate that BMP4-induced astrogenesis is significantly reduced, whereas neurogenesis is significantly increased in NPCs upon Tacc3 RNAi. (K and L) Ectopicrat Tacc3 and Arnt2 cDNA overexpression in 293T cells and Western blot analysis. Increased expression of TACC3 is associated with decreased levels of nuclearARNT2 (K). KHS101 treatment (0–15 μM) elevates nuclear but not cytoplasmic ARNT2 after 24 h of exposure (L). (M and N) Staining for ARNT2 reveals in-creased nuclear localization of ARNT2 in NPCs upon 5 μM KHS101 treatment for 12 h (Fig. S4). (Scale bars: 20 μm.) Error bars, SDs (three independentexperiments; biological replicates in A); statistical significance (t test), *P < 0.05, **P < 0.01; nuclei were visualized with DAPI (blue).

Wurdak et al. PNAS | September 21, 2010 | vol. 107 | no. 38 | 16545

CELL

BIOLO

GY

such as BMP4 in favor of neuronal differentiation. Affinity-basedpurification methodology revealed physical interaction of KHS101with TACC3, although binding of KHS101 to other cellular pro-teins cannot be excluded. Importantly, knockdown of TACC3caused a strong neuronal differentiation phenotype and alsoblocked BMP-induced astrocyte formation in cultured NPCs in acomparable fashion to KHS101 treatment, suggesting that TACC3is likely the most relevant biological target of KHS101. Inter-estingly, KHS101 treatment and knockdown of TACC3 led to in-creased nuclear localization of the nervous system-specifictranscription factor ARNT2, and overexpression of ARNT2markedly favored neuronal differentiation over the astrocyte cellfate upon BMP-induced differentiation.Taken together, these findings indicate that KHS101 specifically

accelerates neuronal differentiationby interactionwith theTACC3protein and support a functional link between KHS101 and theTACC3-ARNT2 axis. Our findings suggest that shRNA- orKHS101-mediated interference with TACC3 accelerates neuro-genesis through negative regulation of the cell cycle and concom-itant activation of a neuronal differentiation program in NPCs. Abody of literature suggests that TACC3 is playing a crucial role inprogenitor cell maintenance and is down-regulated upon differ-entiation (23–27, 37), although the molecular mechanisms down-stream of TACC3 have yet to be determined. As proposed ina previous study (24), TACC3 may sequester distinct transcriptionfactors to the cytoplasmor to the centrosome, therefore preventingthem from binding to transcriptionally active promoters. Alterna-tively, TACC3 may modulate the proteolytic turnover of its bind-

ing partners by providing a target for proteasomal degradation.Overall, it is conceivable that the regulation of TACC3 and itsdownstream interaction partners strongly depends on the cellularcontext (e.g., tissue type and state of the cell cycle) and that specificTACC3binding transcription factors, such asARNT2,mayactivelycontribute to lineage-specific progenitor cell differentiation asa result of TACC3 modulation.It is a commonnotion thatNPCproliferation increases after injury

to the CNS (38). Thus, differentiation-inducing agents such asKHS101 may lead to new therapeutics that act by enhancing thecontribution of newborn cells to CNS repair. Moreover, KHS101may affect the self-renewal potential of malignant progenitor-likecells that can fuel aggressive brain cancer such as glioblastomamultiformeandhavebeen shown to share several functional featuresas well as in vitro growth conditions with normal neural progenitorcells (39–41). Future researchwill shed light on themolecular natureof the KHS101-TACC3 interaction, the cell type- and age-specificroles of TACC3, and a potential therapeutic effect of KHS101 inanimal models of neurodegeneration and CNS malignancies.

MethodsHippocampal NPC Culture. Rat NPCs were derived and cultured as describedpreviously by others (10). After hippocampal cell isolation, the number of dis-sociated cells was determined and ∼5 × 105 cells were plated in 60-mm un-coated plates. After overnight incubation (37 °C, 5% CO2, and 95% humidity),the medium was changed and the cells were expanded and maintained inan undifferentiated state on polyornithine- (10 μg/mL in water; Sigma) andlaminin-coated (5 μg/mL in PBS; Invitrogen) dishes in DMEM/F12 (Invitrogen)supplemented with N2 (Invitrogen) and basic fibroblast growth factor (bFGF,

Fig. 4. KHS101 significantly increases neuronal differentiation in rats in vivo. (A) Pharmacokinetic profile of KHS101 in brain and plasma after single ad-ministration (3 mg/kg, i.v.) to Sprague–Dawley rats. (B and C) Immunohistochemistry of BrdU (red) and NeuN (green). White arrowheads mark BrdU-positivenuclei in the subgranular layer (SGL) of vehicle- and KHS101-treated animals. The yellow arrowhead marks a BrdU/NeuN double-positive nucleus, indicative ofneuronal differentiation in the granule cell layer (GCL). (Scale bar: 20 μm.) (D and E) Ki67 (black arrowhead) and cleaved caspase 3 (brown arrowhead) double-stained DG sections (including the Hilus area: H) of vehicle- and KHS101-treated animals (E). (Scale bar: 50 μm.) (F –H) Percentage of BrdU+/NeuN+ cellsindicates significantly increased neurogenesis (F), whereas the number of Ki67-positive cells (G) significantly decreases in the DG upon KHS101 dosing; thenumber of cleaved caspase 3+ cells within the DG was not altered. (H) Quantification of BrdU+ cells localized in SGL, GL, and H of vehicle- and KHS101-treatedanimals (≥10 sections representative for the DG were counted for each animal). Note a significant reduction of BrdU+ cells in the SGL and a slight increase ofBrdU+ cells in the GCL upon KHS101 administration. Error bars, SDs (six animals per group); statistical significance (t test), **P < 0.01.

16546 | www.pnas.org/cgi/doi/10.1073/pnas.1010300107 Wurdak et al.

20 ng/mL; Invitrogen). For KHS101 and shRNA-induction experiments, earlypassage cells (passaged no more than six times after hippocampal isolation)were trypsinized and plated at a density of ∼1,000 cells/cm2 into N2 medium(DMEM/F12 supplemented with N2) containing KHS analogs (e.g., KHS101,KHS92, andNP; SI Text) at different concentrations (0.5–5 μM)orDMSO (0.1%),RA (1–2 μM), BDNF (100 ng/mL), and/or BMP4 (50–100 ng/mL) for 4 d.

Real Time RT-PCR. Total RNA was purified using the RNeasy kit (Qiagen) andcDNA was produced using the High Capacity cDNA Reverse Transcription kit(Applied Biosystems). TaqMan probes for rat NeuroD, Tacc3, and Gapdh geneexpression assayswere purchased fromApplied Biosystems and used accordingto the manufacturer’s instructions.

Immunocytochemistry. Cells were fixedwith 10% (vol/vol) formalin solution atroom temperature (RT) for 10 min, permeabilized with 0.5% (vol/vol) TritonX-100 (Sigma) in PBS for 10 min, and then blocked with PBS containing 0.3%(vol/vol) Triton X-100, 10% FCS, 1% (wt/vol) BSA at RT for 1 h. Cells wereincubated with primary antibodies (SI Text) in a PBS solution containing 0.1%(vol/vol) Triton X-100 and 5% FCS at 4 °C overnight.

Affinity-Based Target Identification. NPC lysate was prepared by sonication inPBS and protein samples were prepared at a concentration of 2 mg/mL. Thebenzophenone-KHS101 compound (KHS101-BP, 5 μM; SI Text) was added to50 μL of the proteome reaction with and without unlabeled compound(250 μM). Irradiation was for 1 h using a hand-held UV lamp at long wave-length (365 nm), and subsequently a copper-catalyzed azide-alkyne cycload-dition reaction was performed (SI Text). After incubation for 1 h at RT, proteinswere precipitated using trichloroacetic acid and resuspended in isoelectric fo-cusing sample buffer. 2D SDS/PAGE was performed using ReadyStripe IPGstripes (Bio-Rad) following the manufacturer’s protocol.

NPC Electroporation. Electroporation of plasmid DNA (SI Text) into hippo-campal NPCs was performed using the rat neural stem cell Nucleofector Kit(Lonza) according to the manufacturer’s protocol. All shRNA and cDNA ex-pression vectors harbored a puromycin resistance gene allowing for the se-lection of cells expressing plasmid DNA.

Animal Experiments. To investigate the pharmacokinetic properties of KHS101,male Sprague–Dawley rats were administered 3mg/kgKHS101 i.v. or s.c. One ratwas killed per time point at 5min, 40 min, 1 h, and 3 h after dosing, and samplesof blood (100 μL) andwhole brains were collected. In a separate study, rats wereadministered 6 mg/kg KHS101 i.v. or s.c. Five blood samples of 100 μL each werecollected serially via a jugular vein catheterat 2min (i.v. only), 0.5 h (s.c. only), and1, 3, 7 and24hafterdosing. Plasmaandhomogenizedwholebrain sampleswereanalyzed by liquid chromatography tandem mass spectrometry (LC-MS/MS). Tostudy neuronal differentiation upon KHS101 administration in vivo, adult Fisher344 rats (∼10wkold) received s.c. injections of 6mg/kg KHS101 or vehicle control(5% ethanol in 15% Captisol). All rats received one daily i.p. injection of 200mg/kg BrdU for 6 consecutive days after the first day. After 14 d, the animals werekilled and perfusion fixed, and the brains were removed and subjected to im-munohistochemical analysis (SI Text).

Detailed experimental procedures can be found in SI Text.

ACKNOWLEDGMENTS. We thank M. Spooner, N. Gaylord, E. Miller, R. Halder,V. Deshmukh, J. S. Lee, M. Ruiz, B. Chandler, and C. Wright for excellenttechnical assistance. We thank E. Remba and V. Seely for excellent administra-tive assistance. L.L.Lwas supported by a fellowship from theCanadian Institutefor Health Research. H.W. was funded by the European Molecular BiologyOrganization andAmericanAssociation for Cancer Research fellowships. P.G.S.was supported by The Skaggs Institute for Chemical Biology.

1. Zhao C, Deng W, Gage FH (2008) Mechanisms and functional implications of adult

neurogenesis. Cell 132:645–660.2. Chan JP, Cordeira J, Calderon GA, Iyer LK, Rios M (2008) Depletion of central BDNF in

mice impedes terminal differentiation of new granule neurons in the adult

hippocampus. Mol Cell Neurosci 39:372–383.3. Hsieh J, Nakashima K, Kuwabara T, Mejia E, Gage FH (2004) Histone deacetylase

inhibition-mediated neuronal differentiation of multipotent adult neural progenitor

cells. Proc Natl Acad Sci USA 101:16659–16664.4. Kuwabara T, et al. (2009) Wnt-mediated activation of NeuroD1 and retro-elements

during adult neurogenesis. Nat Neurosci 12:1097–1105.5. Malberg JE, Duman RS (2003) Cell proliferation in adult hippocampus is decreased by

inescapable stress: Reversal by fluoxetine treatment. Neuropsychopharmacology 28:

1562–1571.6. Perera TD, et al. (2007) Antidepressant-induced neurogenesis in the hippocampus of

adult nonhuman primates. J Neurosci 27:4894–4901.7. Rossi C, et al. (2006) Brain-derived neurotrophic factor (BDNF) is required for the

enhancement of hippocampal neurogenesis following environmental enrichment. Eur J

Neurosci 24:1850–1856.8. Takahashi J, Palmer TD, Gage FH (1999) Retinoic acid and neurotrophins collaborate to

regulate neurogenesis in adult-derived neural stem cell cultures. J Neurobiol 38:65–81.9. Warashina M, et al. (2006) A synthetic small molecule that induces neuronal

differentiation of adult hippocampal neural progenitor cells. Angew Chem Int Ed

Engl 45:591–593.10. Gage FH, Ray J, Fisher LJ (1995) Isolation, characterization, and use of stem cells from

the CNS. Annu Rev Neurosci 18:159–192.11. Miyata T, Maeda T, Lee JE (1999) NeuroD is required for differentiation of the granule

cells in the cerebellum and hippocampus. Genes Dev 13:1647–1652.12. Reynolds BA, Rietze RL (2005) Neural stem cells and neurospheres—re-evaluating the

relationship. Nat Methods 2:333–336.13. Bonaguidi MA, et al. (2008) Noggin expands neural stem cells in the adult

hippocampus. J Neurosci 28:9194–9204.14. Gobeske KT, et al. (2009) BMP signaling mediates effects of exercise on hippocampal

neurogenesis and cognition in mice. PLoS ONE 4:e7506.15. Lim DA, et al. (2000) Noggin antagonizes BMP signaling to create a niche for adult

neurogenesis. Neuron 28:713–726.16. Xiao Q, Du Y, Wu W, Yip HK (2010) Bone morphogenetic proteins mediate cellular

response and, together with Noggin, regulate astrocyte differentiation after spinal

cord injury. Exp Neurol 221:353–366.17. Fasano CA, et al. (2007) shRNA knockdown of Bmi-1 reveals a critical role for p21-Rb

pathway in NSC self-renewal during development. Cell Stem Cell 1:87–99.18. Kippin TE, Martens DJ, van der Kooy D (2005) p21 loss compromises the relative

quiescence of forebrain stem cell proliferation leading to exhaustion of their

proliferation capacity. Genes Dev 19:756–767.19. Siegenthaler JA, Miller MW (2005) Transforming growth factor beta 1 promotes cell

cycle exit through the cyclin-dependent kinase inhibitor p21 in the developing

cerebral cortex. J Neurosci 25:8627–8636.

20. Sun G, Yu RT, Evans RM, Shi Y (2007) Orphan nuclear receptor TLX recruits histonedeacetylases to repress transcription and regulate neural stem cell proliferation. ProcNatl Acad Sci USA 104:15282–15287.

21. Tang DG, Tokumoto YM, Apperly JA, Lloyd AC, Raff MC (2001) Lack of replicativesenescence in cultured rat oligodendrocyte precursor cells. Science 291:868–871.

22. Gergely F, et al. (2000) The TACC domain identifies a family of centrosomal proteinsthat can interact with microtubules. Proc Natl Acad Sci USA 97:14352–14357.

23. Aitola M, Sadek CM, Gustafsson JA, Pelto-Huikko M (2003) Aint/Tacc3 is highlyexpressed in proliferating mouse tissues during development, spermatogenesis, andoogenesis. J Histochem Cytochem 51:455–469.

24. Garriga-Canut M, Orkin SH (2004) Transforming acidic coiled-coil protein 3 (TACC3)controls friend of GATA-1 (FOG-1) subcellular localization and regulates theassociation between GATA-1 and FOG-1 during hematopoiesis. J Biol Chem 279:23597–23605.

25. Piekorz RP, et al. (2002) The centrosomal protein TACC3 is essential for hematopoieticstem cell function and genetically interfaces with p53-regulated apoptosis. EMBO J21:653–664.

26. SadekCM,etal. (2003) TACC3expression is tightly regulatedduringearlydifferentiation.Gene Expr Patterns 3:203–211.

27. Xie Z, et al. (2007) Cep120 and TACCs control interkinetic nuclear migration and theneural progenitor pool. Neuron 56:79–93.

28. Schneider L, et al. (2008) TACC3 depletion sensitizes to paclitaxel-induced cell deathand overrides p21WAF-mediated cell cycle arrest. Oncogene 27:116–125.

29. Drutel G, et al. (1999) ARNT2, a transcription factor for brain neuron survival? Eur JNeurosci 11:1545–1553.

30. Sadek CM, et al. (2000) Isolation and characterization of AINT: A novel ARNT interactingprotein expressed during murine embryonic development. Mech Dev 97:13–26.

31. Pieper AA, et al. (2010) Discovery of a proneurogenic, neuroprotective chemical. Cell142:39–51.

32. Chen S, et al. (2006) Self-renewal of embryonic stem cells by a small molecule. ProcNatl Acad Sci USA 103:17266–17271.

33. Kochegarov A (2009) Small molecules for stem cells. Expert Opin Ther Pat 19:275–281.34. Schneider JW, et al. (2008) Small-molecule activation of neuronal cell fate. Nat Chem

Biol 4:408–410.35. Wu X, Schultz PG (2009) Synthesis at the interface of chemistry and biology. J Am

Chem Soc 131:12497–12515.36. Zhu S, et al. (2009) A small molecule primes embryonic stem cells for differentiation.

Cell Stem Cell 4:416–426.37. Jeng JC, Lin YM, Lin CH, Shih HM (2009) Cdh1 controls the stability of TACC3. Cell

Cycle 8:3529–3536.38. Scharfman HE, Hen R (2007) Neuroscience. Is more neurogenesis always better?

Science 315:336–338.39. Nakano I, Kornblum HI (2006) Brain tumor stem cells. Pediatr Res 59:54R–58R.40. Pollard SM, et al. (2009) Glioma stem cell lines expanded in adherent culture have

tumor-specific phenotypes and are suitable for chemical and genetic screens. CellStem Cell 4:568–580.

41. Wurdak H, et al. (2010) An RNAi screen identifies TRRAP as a regulator of braintumor-initiating cell differentiation. Cell Stem Cell 6:37–47.

Wurdak et al. PNAS | September 21, 2010 | vol. 107 | no. 38 | 16547

CELL

BIOLO

GY