PACAP Promotes Neural Stem CellProliferation in Adult Mouse Brain

Alex Mercer,1* Harriet Ronnholm,1 Johan Holmberg,2 Hanna Lundh,1

Jessica Heidrich,1 Olof Zachrisson,1 Amina Ossoinak,1 Jonas Frisen,2

and Cesare Patrone1*1NeuroNova AB, Stockholm, Sweden2Department of Cell and Molecular Biology, Medical Nobel Institute, Karolinska Institute, Stockholm, Sweden

In recent years, it has become evident that neural stemcells in the adult mammalian brain continuously generatenew neurons, mainly in the hippocampus and olfactorybulb. Although different growth factors have been shownto stimulate neurogenesis in the adult brain, very little isknown about the role of neuropeptides in this process.Pituitary adenylate cyclase-activating polypeptide(PACAP) is a neuropeptide with pleiotropic effects actingthrough three receptors to which it has high affinity,namely, PACAP receptor 1 (PAC1), vasoactive intestinalpeptide (VIP) receptor 1, and VIP receptor 2. We showthat PAC1 is expressed in the neurogenic regions of theadult mouse brain, namely the ventricular zone of thelateral ventricle and the hippocampal dentate gyrus. Cul-tured neural stem cells isolated from the lateral ventriclewall of adult mice express PAC1 and proliferate in vitro inresponse to two PAC1 agonists, PACAP and Maxadilan,but not VIP at physiologic concentrations, indicatingPAC1 as a mediator of neural stem cell proliferation.Pharmacologic and biochemical characterization ofPACAP-induced neural stem cell proliferation revealedthe protein kinase C pathway as the principal signalingpathway, whereas addition of epidermal growth factorsynergistically enhanced the proliferating effect ofPACAP. Further in vitro characterization of the effect ofPACAP on neural stem cells showed PACAP capable ofstimulating ex novo in vitro formation of multipotent neu-rospheres with the capacity to generate both neuronaland glial cells. Finally, intracerebroventricular infusion ofPACAP increases cell proliferation in the ventricular zoneof the lateral ventricle and the dentate gyrus of the hip-pocampus. We conclude that PACAP, through PAC1, isa potent mediator of adult neural stem cell proliferation.© 2004 Wiley-Liss, Inc.

Key words: neural stem cells; PAC1; subventricularzone; dentate gyrus

The brain has been considered for many years theonly tissue absent of regenerative potential. It has becomeaccepted recently, however, that throughout adult life, thebrain of rodents, primates, and humans retain neural stemcells to constantly generate new neurons and glia (for

review see Kuhn and Svendsen, 1999; Momma et al.,2000). Neurogenesis occurs in discrete regions of the adultbrain including the subventricular zone of the lateral ven-tricles and the subgranular layer of the hippocampal den-tate gyrus. Evidence suggests, however, that neural stemcells in these two areas show cell heterogeneity, in terms ofmultipotency (Seaberg and van der Kooy, 2002) and dif-ferential responses to several growth factors (Kuhn andSvendsen, 1999).

To date, treatment of various neurologic diseases hasbeen addressed mainly by attempts to prevent neuronalcell death or the use of fetal tissue for brain transplantationwith its inherent ethical issues and capacity limitations(Bjorklund and Lindvall, 2000; Horner and Gage, 2000).The accumulating evidence of neurogenesis in the adult,however, raises the alternative possibility that endogenousneural stem/progenitor cells could be activated and maybe recruited to generate new functional neurons to repairthe damaged central nervous system (CNS). This concepthas been proven valid in only a few cases (Nakatomi et al.,2002; Sun et al., 2003); thus, the therapeutic value ofneuronal replacement strategies through neurogenesishave not yet been evaluated extensively. For this reason,the identification of new factors able to increase neuronalprogenitor numbers is important from different perspec-tives: the attempt to fine-tune neurogenesis for repairingspecific brain damage, or in vitro expansion of specificsubsets of neurons for transplantation where these neuronsare lacking (e.g., substantia nigra in Parkinson’s disease).

Contract grant sponsor: Swedish Research Council; Contract grant spon-sor: Karolinska Institute; Contract grant sponsor: Swedish Cancer Society;Contract grant sponsor: Goran Gustafsson Foundation for Research inNatural Sciences and Medicine; Contract grant sponsor: Foundation forStrategic Research.

*Correspondence to: Alex Mercer, NeuroNova AB, Fiskartorpsvagen 15-AD, 114 33 Stockholm, Sweden. E-mail: alex.mercer@neuronova.comand Cesare Patrone, NeuroNova AB, Fiskartorpsvagen 15 A-D, 114 33Stockholm, Sweden. Email: cesare.patrone@neuronova.com

Received 8 July 2003; Revised 13 October 2003; Accepted 5 November2003

Published online 15 March 2004 in Wiley InterScience (www.interscience.wiley.com). DOI: 10.1002/jnr.20038

Journal of Neuroscience Research 76:205–215 (2004)

© 2004 Wiley-Liss, Inc.

The pituitary adenylate cyclase-activating polypep-tide (PACAP) is a member of the vasoactive intestinalpeptide (VIP)/secretin/glucagon family of peptides andexists in two amidated forms, PACAP38 and PACAP27,that share an identical 27-amino acid (aa) N terminus andare processed alternatively from the 176-aa precursor pre-proPACAP (Arimura, 1998; Vaudry et al., 2000). Theprimary structure of PACAP38 has been highly conservedthrough evolution from protochordates to mammals(Arimura, 1998; Vaudry et al., 2000). In mammals,PACAP is expressed by neurons where it acts as a pleio-tropic neuropeptide via three heptahelical G protein-linked receptors, one PACAP-specific (PAC1) receptor,and two receptors that it shares with VIP (VIPR1 andVIPR2). PAC1 is expressed at very high levels in ventric-ular zones through the neuroaxis during development ofrodents (Jaworski and Proctor, 2000). In addition to theembryonic enrichment in proliferating zones, PAC1 ex-pression is maintained in areas of neurogenesis in the adultrat CNS, whereas its ligand PACAP is expressed in theneighboring parenchyma (Jaworski and Proctor, 2000).This suggests a role for PAC1 in adult neurogenesis. Incontrast to rat, adult mouse PAC1 expression is less wellcharacterized.

We show that PAC1 is expressed in the neurogenicventricular zone and hippocampal dentate gyrus of theadult mouse. Furthermore, we show that PAC1 is alsoexpressed in neural stem cells isolated from the lateralventricle wall and cultured in vitro. In vitro application ofPACAP to adult neural stem cells induce a pronouncedproliferating response elicited through PAC1, an effectthat can be enhanced synergistically with epidermalgrowth factor (EGF). Furthermore, in vivo functional dataindicate PACAP as an adult neural stem cell proliferatingagent, suggesting this factor as a new important player inadult neurogenesis.

MATERIALS AND METHODS

In Situ Hybridization

Sections (14 �m) of whole adult (3 months old) mousebrain were cut on a cryostat at �17°C, thawed onto microscopeslides (Superfrost Plus; BDH, UK) and fixed in 4% formalde-hyde for 5 min, deproteinated for 15 min in 0.2 M HCl, treatedin 0.25% acetic anhydride in 0.1 M triethanolamine buffer (pH8.0) for 20 min and dehydrated in an ascending series of ethanolconcentrations including a 5-min chloroform step before hy-bridization. To detect PAC1 mRNA, an antisense cRNA probespecific for all known isoforms was transcribed from a plasmid(pGEM-Teasy) containing cDNA corresponding to bases 1181–1475 of the coding sequence of the PAC1 and concurrently[�-35S]UTP-labeled. Sections were incubated with the probe at55°C for 16 hr in a hybridization buffer containing 52% form-amide, 10% dextran sulfate, 208 mM NaCl, 2% 50� Denhardt’ssolution (1% Ficoll, 1% polyvinylpyrrolidone, 1% bovine serumalbumin [BSA]), 10 mM Tris pH 8.0, 1 mM EDTA, 500 ng/mlyeast tRNA, 10 mM dithiothreitol (DTT) and 20 � 106 cpmprobe per ml buffer. After hybridization, the sections weretreated with RNase A, 10 �g/ml in 0.5 M NaCl, at 37°C for

30 min and washed in 4� saline sodium citrate (SSC; 1� SSCis 0.15 M sodium chloride and 0.015 M trisodium citrate, pH7.0) for 20 min, 2� SSC for 10 min, 1� SSC for 10 min, and0.5� SSC for 10 min at room temperature. A high stringencywash was carried out at 70°C for 30 min in 0.1� SSC. All washsteps included the addition of 1 mM DTT. The sections weredehydrated in an ascending series of ethanol concentrations,dried overnight and mounted in cassettes with autoradiographicfilms (Beta-max, Amersham) laid on top for 3 weeks. The filmswere developed in Kodak D-19 developer, fixed in KodakRA-3000 diluted 1:3, rinsed, and dried. Sections were thendipped in Kodak NTB-2 nuclear track emulsion diluted 1:1,exposed for 6 weeks, developed in Kodak D-19 for 3 min, fixedin Kodak RA-3000 fixer, and counterstained with cresyl violet.The specificity of the hybridization was tested using a senseprobe transcribed from the same plasmid. No hybridizationsignal was obtained under this condition. The emulsion-dippedsections were analyzed using a Nikon E600 microscope.

Adult Mouse Neural Stem Cell Culture

The lateral wall of the lateral ventricle of 5–6-week-oldmice was enzymatically dissociated in 0.8mg/ml hyaluronidaseand 0.5 mg/ml trypsin in Dulbecco’s modified Eagle medium(DMEM) containing 4.5 mg/ml glucose and 80 U/ml DNase at37°C for 20 min. The cells were gently triturated and mixedwith three volumes of neurosphere medium (DMEM/F12, B27supplement, 12.5 mM HEPES, pH7.4) containing 20 ng/mlEGF (unless otherwise stated), 100 U/ml penicillin and100 �g/ml streptomycin. After passing through a 70-�mstrainer, cells were pelleted at 160 � g for 5 min. The superna-tant was subsequently removed and the cells resuspended inneurosphere medium supplemented as above, plated in culturedishes, and incubated at 37°C. Neurospheres were ready to besplit approximately 7 days after plating.

To split neurosphere cultures, neurospheres were col-lected by centrifugation at 160 � g for 5 min. Neurosphereswere resuspended in 0.5 ml Trypsin/EDTA in HEPES bufferedsaline solution, incubated at 37°C for 2 min and trituratedgently to aid dissociation. After a further 3-min incubation at37°C and with trituration, three volumes of ice-cold neuro-sphere medium containing EGF were added. The cells werepelleted at 220 � g for 4 min, resuspended in fresh neurospheremedium supplemented with 20 ng/ml EGF and 1 nM basicfibroblast growth factor (bFGF) plated out and incubated at37°C.

RT-PCR

The anterior lateral ventricle wall of 5–6-week-old micewas dissected for either RNA isolation or neurospheres prepa-ration as stated above. The neurospheres were harvested 3 daysafter the first split. Total RNA was isolated using Qiagen(Hilden, Germany) RNeasy Mini Kit according to the manu-facturer’s instructions and DNase treated with DNase I (Am-bion) according to protocol. A One-Step RT-PCR Kit (LifeTechnology, Gaithersburg, MD) was used to detect the presenceof PAC1 mRNA. Briefly, 12.5 ng of total RNA was used ineach 30-cycle reaction, with primers 5�-CCTGTCGGTGAA-GGCCCTCTACACA-3� (sense) and 5�-CCCAGCCCAAG-CTCAAACACAAGTC-3� (antisense) at an annealing temper-

206 Mercer et al.

ature of 55°C, generating a 801-base pair (bp) PCR product. Toensure that genomic DNA contamination did not give rise tofalse positive results, an identical reaction in which the RT-Taqpolymerase mix was replaced by Taq polymerase alone and runin parallel with the experimental RT-PCR. Reactions wereresolved on a 1.5% agarose gel containing ethidium bromide andthe bands visualized under UV light. Bands corresponding to theestimated length of PCR products of the desired genes werecloned into the cloning vector pGEM-Teasy and sequenced toverify their identity.

Intracellular ATP and Lactate Dehydrogenase Assays

Intracellular ATP levels have been shown previously tocorrelate to cell numbers (Crouch et al., 1993). Mouse neuro-spheres, cultured as described above, from passage 2, wereseeded in DMEM/F12 and supplemented with B27 into a96-well plate as single cells (10,000 cells/well). To the neuro-spheres, substances to be measured were added at the concen-trations indicated. After 3 days incubation, intracellular ATP wasmeasured using the ATP-SL kit (BioThema, Sweden) accordingto the manufacturer’s instructions. PACAP and VIP were pur-chased from Bachem. Maxadilan was a kind gift from RichardG. Titus, of the Department of Microbiology, Immunology,and Pathology, College of Veterinary Medicine and BiologicalSciences, Colorado State University.

In experiments examining signaling pathways, dissociatedcells were plated as above. PACAP (100 nM) was co-incubatedwith 10 �M protein kinase A (PKA) inhibitor H89 (AlexisBiochemicals), or 1 �M of the protein kinase C (PKC) inhibitorGo6976 (Sigma-Aldrich). Cells were incubated for 4 days beforemeasurement of ATP.

In the experiment analyzing the combinatory effect ofPACAP and EGF, PACAP (100 nM) was co-incubated withEGF (3 nM) for 3 days.

To assay for cell death in mouse neurosphere cultures,neurospheres were cultured as described above and lactate de-hydrogenase (LDH) released by dying cells was quantified usingan enzymatic assay (CytoTox96 cytotoxicity assay, Promega).

Thymidine Incorporation

To determine thymidine incorporation into DNA, neu-rospheres were dissociated to a single-cell suspension. Cells wereplated in neurosphere medium as single cells in 96-well plates(10,000 cells/well). Substances to be analyzed were added inquadruplicate and cells incubated at 37°C for 3 days. 3H-thymidine (Perkin-Elmer) was present the last 24 hr(10 �Ci/ml). Cells were harvested on to a filter paper andradioactivity was measured.

Self-Renewal and Multipotency

The lateral walls of the lateral ventricle of 10 mice weredissociated as described above, and the cells resuspended inneurosphere medium without EGF. For self-renewal experi-ments, the cell suspension was divided in 9 wells of a 24-wellplate to which vehicle, PACAP, or EGF were added in triplicateat the concentrations mentioned above. After 7 days, the result-ant neurospheres were analyzed for size and morphology andphotographed. To compare the number of neurospheres gen-erated in each treatment, 10 random fields (20� magnification)

in each well were counted. For multipotency experiments, thecell suspension was grown in neurosphere medium for 7 daysand the spheres were then replated, avoiding dissociation, inneurosphere medium supplemented with 1% fetal calf serum(FCS; Gibco) and 100 nM PACAP (Bachem) onto poly-D-lysine plates. The spheres adhered to the poly-D-lysine plates andthe medium was changed to neurosphere medium supple-mented with 100 nM PACAP without FCS. The cells werecultured for a further 3 days, after which they were washedtwice in phosphate-buffered saline (PBS; Gibco) and fixed for15 min at room temperature with 4% formaldehyde (Sigma) andpermeabilized for 20 min at room temperature in 0.1% TritonX-100 (Sigma) in PBS. After fixation and permeabilization, cellswere labeled with mouse monoclonal anti-�-III tubulin (1:1,000; Promega), rabbit anti-GFAP (1:500; Sigma), and mouseanti-O4 (1:500; Sigma). Primary antibodies for �-III tubulinand GFAP were visualized using the secondary antibodies anti-mouse IgG Texas-Red-conjugated (1:500; Vector Labs) andanti-rabbit IgG FITC-conjugated (1:500; Vector Labs), respec-tively. All antibodies were diluted in PBS with 0.1% TritonX-100.

Implantation of Osmotic Pumps and PACAP/Bromodeoxyuridine Infusion

Ten-week-old male mice (C57B), maintained on a 12-hrlight/dark cycle with food and water ad lib were infused in theright lateral ventricle with PACAP38 (Bachem) or vehicle, usingAlzet pumps (1007D), for 3.5 or 7 days at a dose of 31 ng/day(600 nM PACAP pump concentration infused at a rate of0.5 �l/hr). Bromodeoxyuridine (BrdU) (50 mg/ml) was alsoincluded in the infusion vehicle (0.9% saline containing1 mg/ml mouse serum albumin; Sigma) to enable measurementof proliferation by quantitation of BrdU incorporation in theDNA. Animals treated for 3.5 days were sacrificed for subven-tricular zone analysis, whereas animals treated for 7 days wereallowed to survive for a further 10 days before sacrifice fordentate gyrus analysis. Animals were perfused with PBS, and thebrains were removed and frozen at �70°C before sectioning forimmunohistochemical analysis.

Immunohistochemistry

Brains were cut in 14-�m coronal sections using a cryo-stat. Sections were thawed onto pretreated slides and fixed in 4%(wt/vol) paraformaldehyde/PBS for 10 min. After washing inPBS, the sections were treated with 2 M HCl at 37°C for 30 minto increase accessibility of the anti-BrdU antibody to the cellnuclei. The sections were rinsed in PBS and transferred toblocking solution (0.1% Tween and 10% goat serum in PBS)overnight at 4°C. Primary antibody (rat anti-BrdU; Harlan SeraLabs) was applied at 1:100 in blocking solution for 90 min atroom temperature. After washing in 0.1% Tween in PBS for3 � 30 min, secondary biotinylated antiserum (goat anti-rat,Vector Labs) was added at a 1:200 dilution in blocking solutionfor 60 min at room temperature. The sections were washed for2 hr before treatment with Vectastain Kit (Vector Labs) accord-ing to the manufacturer’s protocol. After 1 hr of washing, theBrdU-antibody complex was detected using 0.05% diamino-benzidine with 0.01% H2O2, and counterstained with hema-toxylin. Sections were dehydrated in a graded series of ethanol

PACAP Promotes Neural Stem Cell Proliferation 207

concentrations, followed by xylene and 99% ethanol, andmounted in Pertex. Sections were visualized using a NikonEclipse E600 microscope and pictures taken with a Spot InsightCCD camera.

Quantification and Statistical Analysis

For BrdU labeling experiments, 11 vehicle and 10PACAP-treated animals were sectioned for analysis of the sub-ventricular zone, and 8 vehicle and 11 PACAP-treated animalsfor the dentate gyrus. Three to six sections per animal wereanalyzed. For hippocampus, the quantification was carried out asdescribed previously (McCabe et al., 2001). Briefly, sectionsdivided between the anterior, middle, and posterior portions ofthe dorsal hippocampus were analyzed in an area encompassingthe entire granule cell layer (superior and inferior blades) in-cluding the subgranular zone, which was defined as extending amaximum of two cell widths into the hilus region. Based onanatomic landmarks, equivalent sections from control and ex-perimental animals were chosen and coded by one of the au-thors, and remained concealed to the examiner throughout thestudy. The number of BrdU-labeled cells per area of dentategranule cell layer was counted manually. For the lateral ventriclewall, sections were collected posterior to the genus of corpuscallosum and anterior to the closing of anterior commissura,corresponding to coronal plates number 22–29 of the Atlas overthe mouse brain (Paxinos and Franklin, 2001). As above, equiv-alent sections from control and experimental animals were cho-sen and coded by one of the authors, and remained concealed tothe examiner throughout the study. BrdU-positive cells werecounted along 2 � 250 �m strips of the lateral ventricle wall.Area measurements of the dentate granule cell layer and the“counted” region of the lateral ventricle wall (including theependymal cell layer and subventricular zone) were made fromeach slide used for the cell counts. The mean value for theexperimental group was compared to that of the control group.Results of the dentate gyrus BrdU counts are expressed as theaverage number of BrdU-positive cells per area (mm2) for eachindividual animal and reported as the mean � SEM. Datagenerated analyzing lateral ventricle wall sections come fromtwo independent experiments. The number of BrdU-positivecells per area (mm2) are expressed as percentage of the mean �SEM of the control group. Differences between means weredetermined by Student’s t-test.

RESULTS

PAC1 Is Expressed in the Subventricular Zone andDentate Gyrus of the Adult Mouse

It has been shown previously by Jaworski and Proc-tor (2000) that PAC1 is expressed in the rat subventricularzone and in dentate gyrus of the hippocampus. Interest-ingly, the data reported PAC1 to be upregulated duringearly rat developmental stages and early postnatal life butdeclining to lower expression levels in adult life. To ad-dress the potential role of PACAP and PAC1 in adultmouse neurogenesis, we carried out in situ hybridizationfor PAC1 mRNA in adult mouse brain sections, analyzingin particular neurogenic regions, namely the subventricu-lar zone of the lateral ventricle and the hippocampus. As

shown in Figure 1, PAC1 expression was clearly detectedin both the subventricular zone (Fig. 1B,E) and dentategyrus of the hippocampus (Fig. 1A,D). These results in-dicate that PAC1 is expressed at substantial levels in adultmouse brain areas where neurogenesis occurs.

PAC1 Is Expressed in Cultured Adult Neural StemCells

As described above, PAC1 is expressed in neurogenicareas of the developing and adult nervous system. There isvery little information, however, describing a potential roleand mechanism of action for PACAP in adult neurogenesis.To address whether PACAP through its receptor PAC1could potentially influence adult neural stem cells, we lookedfor expression of PAC1 in cultured adult neural stem cells,derived from the lateral ventricle wall, using RT-PCR.Bands corresponding to PAC1 were detected (Fig. 2) andthese sequences verified indicating PAC1 is indeed expressedby adult neural stem cells. Together with in situ expression ofPAC1 in neurogenic regions, these data imply that PACAPmay modulate adult neurogenesis.

PACAP Stimulates Proliferation of Adult MouseNeural Stem Cells

To address whether PACAP may modulate adultneurogenesis, we first investigated the effect of PACAP onneural stem cell proliferation in vitro. Adult mouse neuralstem cells derived from the lateral ventricle wall wereexpanded as neurospheres using EGF. Neurospheres weredissociated and supplemented with varying concentrationsof PACAP in the absence of EGF for 3 days. To analyzewhether there was a difference in number of treated cellsrelative to control cells, we employed an assay measuringintracellular ATP levels, shown previously to correlatewith cell number (Crouch et al., 1993). The data in Figure3A shows a statistically significant increase in intracellularATP levels and hence cell number in response to PACAPin a dose-dependent manner. The ability of PACAP tobind with essentially equal affinity to PAC1 and VIPR1and 2 poses the possibility that PACAP is eliciting itsmitogenic effects through the VIPRs rather than throughPAC1. To address this question, VIP, specific only forVIPR1 and 2 at high affinity and with a 100–1,000-foldlower affinity for PAC1, was applied to the neural stemcells in parallel with PACAP. Interestingly, VIP was inef-fective in stimulating proliferation unless at nonphysi-ologic concentration (Fig. 3B), thus indicating that PAC1but not VIPR is involved in neural stem cell proliferation.To test further whether PAC1 is indeed the sole PACAPreceptor responsible for the proliferating effect, neuralstem cells were grown as above and treated with thespecific PAC1 agonist, Maxadilan (Lerner et al., 1991;Lerner and Shoemaker, 1992; Moro and Lerner, 1997), atvarying concentrations. Figure 3C shows that Maxadilaninduces a proliferating effect to that of PACAP at similarnanomolar concentrations. To confirm that the PACAP-treated cells were indeed proliferating, incorporation oftritiated thymidine was used to assess DNA synthesis.Greater incorporation of tritiated thymidine was observed

208 Mercer et al.

with all PACAP treatments relative to controls, indicatingthat PACAP was eliciting a proliferative response in neuralstem cells (Fig. 3D). Comparable release of LDH, a sen-sitive marker for cell death, in control and PACAP-treated

neurospheres ruled out the possibility that the increase incell number in response to PACAP was through a survivaleffect (data not shown). These data indicate that PAC1activation can trigger adult neural stem cell proliferationand that PACAP can evoke this effect.

The Proliferative Effect of PACAP on AdultNeural Stem Cells Is Mediated Through a PKC-Dependent Pathway

Six intracellular splice variants have been reportedfor PAC1, varying in their propensity to signal throughphospholipase C� (PLC�) via PKC or through adenyl-ate cyclase via protein kinase A (PKA) (Waschek, 2002).To address which biochemical pathway mediatesPACAP neural stem cell proliferation, we blocked PKAand PKC using specific inhibitors. PKC (Go6976) in-hibition resulted in complete arrest of PACAP-inducedneural stem cell proliferation, whereas inhibition ofPKA (H89) was ineffective (Fig. 4). These results indi-cate that the PLC�/PKC signaling pathway rather thanadenylate cyclase/PKA signaling is responsible for pro-liferation.

PACAP and EGF elicit a Positive Synergistic Effecton Adult Mouse Neural Stem Cell Number In Vitro

To characterize further the effect of PACAP onneural stem cells, we investigated the possible interplaybetween PACAP and EGF on adult neural stem cellproliferation. EGF is one principle factor used to support

Fig. 2. PAC1 gene is expressed in cultured adult mouse neurospheres(NS) and lateral ventricle wall (LVW) tissue of adult mice. RT-PCRwas carried out on total RNA prepared from NS, LVW tissue, andwhole brain using primer pairs specific for the PAC1 gene. Arrows,bands corresponding to bands of desired size that were cloned andsequenced to verify that they represented the correct product. Controlreactions (actin primers and Taq polymerase) were carried out to ruleout false positive bands due to genomic contamination.

Fig. 1. PAC1 gene is expressed in neurogenic regions of the adult mouse brain as detected by in situhybridization. Low magnification sagittal (A) and coronal (B) and high magnification coronal (D, E)micrographs showing PAC1 gene expression in the hippocampal granule cell layer (gcl) (A, D) and thesubventricular zone (svz) of the lateral ventricle (lv) wall (B, E). No specific signals were seen with thesense probes (C, F).

PACAP Promotes Neural Stem Cell Proliferation 209

neural stem cells proliferation in vitro (McKay, 1997;Johansson et al., 1999). Adult neural stem cells werecultured in presence of PACAP, EGF, or a combination ofthe two factors for 4 days. PACAP stimulated a doublingof neural stem cell number compared to control, whereasthe effect of EGF was substantially greater, eliciting aseven- to tenfold increase (Fig. 5). Although both factorssignificantly increased cell proliferation alone, in combi-nation the effect was enhanced markedly, indicating asynergistic effect between the two compounds. Theseresults indicate that PACAP can serve as a mitogen forneural stem cells when administered alone; however, itsproliferative capacity is augmented when in combinationwith EGF.

PACAP Promotes Adult Mouse Neural Stem Cellsto Form Neurospheres In Vitro While RetainingTheir Potential to Differentiate Into Neurons andGlia

The identification of factors that can selectively stim-ulate neural stem cell activity, the proliferation of neuralprogenitors, and differentiation of cells into the desiredphenotype is likely to have profound consequences for thetreatment of neurologic diseases, from the perspective ofin vivo neurogenesis stimulation as well as in vitro expan-sion of neural stem cells for transplantation therapy. In thisrespect, it is important to understand whether PACAP isstimulating proliferation of neural stem cells while main-taining their multipotentiality. To address this issue, neuralstem cells from the lateral ventricle wall were prepared asdescribed above; however, EGF was either omitted orreplaced with PACAP. Neural stem cells were cultured for7 days after which time they were inspected for growth

Fig. 3. PACAP and Maxadilan stimulateadult mouse neural stem cell proliferationthrough PAC1 activation. Neural stemcells plated as single cells were grown in thepresence and absence of varying concen-trations of PACAP, VIP, and Maxadilanfor 3 days. Intracellular ATP increased in adose-dependent manner with increasingconcentrations of PACAP (A) and Max-adilan (C) but only at the maximum doseof VIP (B). 3H-thymidine incorporationincreased with PACAP treatment of neuralstem cells grown under nonadherent con-ditions (D). The data show experiments inwhich six data points were analyzed foreach individual treatment; bars represent�SEM. Significance levels of increasesabove control values were determined bypaired Student’s t-test (*P 0.05; **P 0.01; ***P 0.005).

Fig. 4. PACAP stimulates adult mouse neural stem cell proliferationthrough a PKC-dependent rather than a PKA-dependent pathway.Neural stem cells plated as single cells were grown in the presence orabsence of PACAP (100 nM), with or without the PKC inhibitorGo6976 (1 �M) and PKA inhibitor H89 (10 �M). Intracellular ATPlevels were measure after 4 days incubation. Bars represent meanpercentage � SEM. Significance levels of increases above control valueswere determined by paired Student’s t-test (**P 0.01). Significancelevels for change relative to PACAP treatment were determined bypaired Student’s t-test (##P 0.01).

Fig. 5. PACAP and EGF elicit a positive synergistic effect on adultmouse neural stem cell number in vitro. Neural stem cells plated assingle cells were grown in the presence or absence of PACAP (100 nM)and EGF alone or in combination (1 and 3 nM). Intracellular ATP wasmeasured after 4 days incubation.

210 Mercer et al.

and morphology. Figure 6 shows neural stem cells treatedwith PACAP growing in neurosphere formations substan-tially larger than that of controls (Fig. 6A–C), althoughsmaller than those generated through EGF treatment. Inaddition to size comparison, the number of neurosphereswas also evaluated (Fig. 6D). A significant twofold increase

in the number of neurospheres above that of controls wasobserved for PACAP, whereas EGF generated 10 times asmany spheres. Notably, PACAP-treated neurospherescould be passaged multiple times, showing no significantdifference when compared to EGF-treated spheres (datanot shown). To investigate further whether PACAP-treated neural stem cells retain their potential to differen-tiate into neurons and glia, first-generation neurospherescultured in the presence of PACAP were differentiated onpoly-D-lysine precoated plates. The image in Figure 6Eshows that cells grown in a single neurosphere formationcan differentiate into neurons immunoreactive for �-IIItubulin, GFAP-positive astrocytes, or O4-positive oligo-dendrocytes. This shows that PACAP-treated neural stemcells are capable of generating all cell types of the CNS.

We conclude from our in vitro studies that PACAP,via PAC1, is sufficient to promote neural stem cells pro-liferation. Furthermore, these results indicate that neuro-spheres expanded in vitro by PACAP retain their potentialto become neurons and glia.

PACAP Stimulates Proliferation in theSubventricular Zone and Dentate Gyrus in theAdult Mouse

The in vitro data described above clearly promotePACAP as a potential factor for in vivo stimulation ofneurogenesis. To address this issue, we infused into thelateral ventricle of adult mice PACAP, together withbromodeoxyuridine (BrdU) to label dividing cells, andcounted the numbers BrdU-positive cells in the forebrainsubventricular zone and dentate gyrus of hippocampus.The results in Figure 7 show a representative image of theadult mouse subventricular zone of control animals (Fig.7A) and animals treated with PACAP (Fig. 7B). After3.5 days of PACAP/BrdU treatment, the number ofBrdU-labeled cells in the subventricular zone increased byapproximately 50% relative to control animals (Fig. 7C). Asimilar increase in BrdU-positive cells was also observed inthe dentate gyrus of the hippocampus after 7 days infusionof PACAP/BrdU followed by 10 days latency period priorto sacrifice (Fig. 7D–F).

We show that the neuropeptide PACAP can stimu-late an increase in endogenous neural stem cell prolifera-tion in the two principle regions where neurogenesis isinitiated in the adult mouse brain.

DISCUSSIONWe report that the neuropeptide PACAP, via PAC1,

can stimulate adult neural stem cell proliferation in vitro.PACAP, even in the absence of any other growth factors,is sufficient to promote proliferation of neural stem cells asneurospheres while retaining their potential to differenti-ate into neurons and glia and generate new spheres. Infu-sion of PACAP into the lateral ventricle of adult micetriggered proliferation in both the subventricular zone ofthe lateral ventricle and the hippocampal dentate gyrus, asdetermined by BrdU incorporation. Taken together,present results shows that PACAP is a neural stem cell-

Fig. 6. PACAP promotes adult mouse neural stem cells to form neu-rospheres in vitro while retaining their potential to differentiate intoneurons and glia. Anterior lateral ventricle walls from adult mousebrains were dissociated, the cells resuspended in neurosphere mediumand divided into a 24-well plate corresponding to 1 brain/well. PAP-CAP (100 nM) and EGF (1 nM) were added in triplicate and thecultures incubated for 7 days, after which the resultant neurosphereswere studied for size and morphology (A, control; B, PACAP; C,EGF). To compare the number of neurospheres generated by eachtreatment, neurospheres were counted in 10 random fields observedthrough a 20� objective (D). Bars represent �SEM. Significance levelsof increases above control values were determined by paired Student’st-test (*P 0.05). PACAP-generated neurospheres were subsequentlyplated onto poly-D-lysine plates to which 1% fetal calf serum was addedand incubated overnight. Neurospheres were cultured for a further2 days in the presence of PACAP alone before fixing in 4% parafor-maldehyde and staining with antibodies against �-III tubulin (red),GFAP (blue), O4 (green), as shown in E. Scale bar 20 �m.

PACAP Promotes Neural Stem Cell Proliferation 211

proliferating agent, suggesting a novel role for this neu-ropeptide in adult brain physiology.

PACAP has the ability to bind three receptors withhigh affinity, namely, PAC1, VIPR1, and VIPR2. Al-though PAC1 is specifically selective for PACAP, VIPR1

and VIPR2 can share both PACAP and VIP at similaraffinities (Rawlings, 1994). In the nervous system, theeffects of PACAP exerted though PAC1, are pleiotropic.Studies have documented a potent differentiation capacityon neocortical precursor cells, proliferation of sympatheticneuroblasts, regulation of neurotransmitter release, and aneuroprotective effect on hippocampal neurons (DiCicco-Bloom et al., 1998; Shioda et al., 1998; Hashimoto et al.,2001; Nicot and DiCicco-Bloom, 2001). In light of themultifaceted abilities of PACAP, we sought to investigatewhether PACAP and PAC1 may modulate adult neuro-genesis. Our in situ hybridization data showed PAC1 to behighly expressed in the two main areas of the mouse brainwhere adult neurogenesis occurs: the subventricular zoneof the lateral ventricle and the hippocampal dentate gyrus.This data corroborated previous findings by Jaworski andProctor (2000) on the localization of PAC1 in neurogenicareas of the adult rat brain. This study demonstrated PAC1expression to be at its peak during late embryonic devel-opment and early postnatal life, correlating both spatiallyand temporarily to developmental neurogenesis. Despitedownregulation of the receptor occurring as rats matureinto adulthood, expression remains abundant in the adultrodent subventricular zone and dentate gyrus. Our datashows that expression of PAC1 also persists in the adultmouse suggesting the receptors participation in adult neu-rogenesis.

To investigate the potential effect of PACAP onadult neural stem cells, we first sought to determinewhether PAC1 was expressed by lateral ventricle wall-derived adult neural stem cells grown in vitro as neuro-spheres. RT-PCR experiments showed PAC1 mRNA tobe present in adult neurospheres, suggesting the applica-bility of this in vitro model system for further investigationof the effect of PACAP. The in vitro proliferation dataidentify PACAP as a proliferative agent for adult neural

Š

Fig. 7. PACAP stimulates proliferation of neural stem cells in the adultmouse brain. Adult male mice were infused into the lateral ventriclewith PACAP or vehicle and BrdU for 3.5 or 7 days. Mice infused for3.5 days with PACAP or vehicle were sacrificed immediately thereaf-ter, the brains sectioned and the subventricular zone of the lateralventricle wall analysed for BrdU incorporation by immunohistochem-istry (A,B). Data are representative fields from two independent ex-periments. For quantification (C), 3–6 sections were taken for eachanimal and counted manually in the regions boxed. Cells immunore-active for BrdU were counted and expressed as a mean percentage �SEM (PACAP n 10 PACAP) (Vehicle n 11) of BrdU-immunopositive cells after vehicle treatment (*P 0.05 relative toVehicle). Mice infused for 7 days with PACAP or vehicle were allowedto survive a further 10 days before sacrifice, after which the brains weresectioned and the dentate gyrus analysed for BrdU incorporation byimmunohistochemistry (E,F). Data are representative fields from twoindependent experiments. For quantification of BrdU positive cells inthe dentate gyrus following PACAP and vehicle infusion (D), 3–6sections were counted manually and the results are expressed as themean � SEM number of BrdU positive cells/mm2 (PACAP n 7)(Vehicle n 8) (**P 0.005 relative to Vehicle). Abbreviations: lv,lateral ventricle; str, striatum.

212 Mercer et al.

stem cells and that this effect is through the activation ofPAC1. Two lines of evidence indicate that the PAC1 perse rather than either of the VIPRs mediate the neural stemcell response to PACAP. First, the specific PAC1 agonistMaxadilan could elicit a similar response to that inducedby PACAP, and second, VIP was ineffective when used atphysiologic concentrations. Only micromolar concentra-tions of VIP that have been reported to activate PAC1(Jaworski and Proctor, 2000) could stimulate proliferationof the neurospheres.

In the attempt to characterize the biochemical path-way involved in adult neural stem cell proliferation, wefound the PKC pathway to be critical in mediating PAC-AP’s mitogenic signal. Although PAC1 has been reportedto signal through both adenylate cyclase/PKA and PLC�/PKC pathways (Waschek, 2002), PACAP-treated adultneural stem cells seem to selectively involve PKC, to theexclusion of PKA, to proliferate. Recent studies on em-bryonic Day 14 (E14) rat cortical precursors and sympa-thetic neuroblasts suggest that the cellular response toPACAP (in this case proliferation or survival) is dictated bythe expression of specific splice forms of PAC1 that in turndetermine activation of adenylate cyclase/PKA or PLC�/PKC pathways (Dicicco-Bloom et al., 1998; Shioda et al.,1998; Hashimoto et al., 2001; Nicot and DiCicco-Bloom,2001) . Receptor isoforms differ in their third intracellularloop, a region critical for specific G-protein interactionsthat link to the adenylate cyclase and PLC second mes-senger pathways. The presence of a 28-aa “hop” cassette isthought to promote coupling to the PLC�/PKC pathway.Indeed, expression of the “hop” isoform and activation ofthe PLC�/PKC cascade was critical to the mitogeniceffect of PACAP on E14 rat cortical precursors and sym-pathetic neuroblasts. The significance of the PLC�/PKCpathway activation, to the exclusion of the adenylatecyclase/PKA pathway, in adult neural stem cell prolifera-tion thus runs parallel to the studied embryonic precursors.RT-PCR analysis indicate that the mouse adult neuralstem cells express the “hop” splice variant (data notshown) suggesting its use in promoting the PLC�/PKCpathway for mitogenic activity.

Our in vivo results showing that intraventricularinfusion of PACAP can increase the number of BrdU-positive cells in the proliferating regions of the adult brainconfirm the in vitro data of PACAP as mitogenic for stemand progenitor cells. An approximate 50% increase inproliferation was observed in the subventricular zone ofthe lateral ventricle after 3.5 days infusion and in thedentate gyrus of the hippocampus after 7 days infusionfollowed by a 10 days latency period prior to sacrifice.Staining of sections from PACAP and vehicle brains showno observable differences in numbers of terminal deoxy-nucleotidyl transferase-mediated dUTP nick-end labeling(TUNEL)-positive cells (data not shown); indicating thatthe increase in cells that have undergone cell division afterPACAP treatment is through proliferation rather thanthrough an enhanced survival effect on the divided cells.

The identification of proliferating and differentiatingfactors for neural stem cells is likely to provide therapeuticbenefits for neuronal replacement strategies in the treat-ment of neurological diseases and disorders. Intraventric-ular infusion of growth factors such as EGF, bFGF, andvascular endothelial growth factor (VEGF) have beenshown to stimulate proliferation of the ventricle wall cellpopulation, and in the case of EGF, cause extensive mi-gration of progenitors into the neighboring striatal paren-chyma (Craig et al., 1996; Kuhn et al., 1997). A recentstudy found that intraventricular infusion of brain-derivedneurotrophic factor (BDNF) in adult rats promotes in-creases in the number of newly generated neurons in theolfactory bulb and rostral migratory stream, and in paren-chymal structures, including the striatum, septum, thala-mus, and hypothalamus (Pencea et al., 2001). In additionto growth factors, a number of cytokines and recentlyprolactin have also been identified as agents with in vivoneural stem cell-proliferating capacity, a list to whichPACAP can now be added (Hansel et al., 2001; Shingo etal., 2003). Comparing prolactin and PACAP, the effectson neural stem cell proliferation are both similar anddivergent. For example, unlike PACAP, which can gen-erate primary neurospheres with propagation abilities fromlateral ventricle wall tissue, prolactin requires EGF toevoke a mitogenic response (Shingo et al., 2003). As is thecase for prolactin and EGF, however, a synergistic prolif-erating effect was also observed between PACAP andEGF, indicating that PACAP acts in concert with othermolecules to exert its maximal potential. We speculate thatthis mechanism may represent a key mode of actionwhereby neural stem cell proliferation is augmented bycross talk between neuropeptides or hormones, such asPACAP and prolactin, and growth factors, such as EGF.These observations could be important to greatly enhanceadult neural stem cell proliferation for induction of neu-rogenesis for therapeutic gain but also for in vitro expan-sion of stem cells for transplantation. In this light, the factthat PACAP-treated neural stem cells retain their potentialto generate all cell types of the CNS, namely neurons,astrocytes, and oligodendrocytes, becomes an importantfactor to consider. Further studies evaluating the differen-tiating potential of PACAP to generate neuronal progen-itors and neurons will determine if this agent can be usednot only for neural precursor expansion purposes but alsofor the stimulation of a specific neuronal phenotype.

The finding that PACAP is a trophic factor formidbrain dopaminergic neurons and can stimulate tran-scription of the rate-limiting enzyme for the production ofdopamine (tyrosine hydroxylase) in different in vivo andin vitro systems suggests PACAP could be an importantmediator of neuronal dopaminergic differentiation andsurvival (Takei et al., 1998; Park et al., 1999; Corbitt et al.,2002). In this respect, it has been shown recently thatnewborn cells can migrate from the ventricular system notonly to the olfactory bulb and hippocampus but also toother brain areas such as the substantia nigra (Zhao et al.,2003). Not only did these newborn cells migrate to the

PACAP Promotes Neural Stem Cell Proliferation 213

substantia nigra, but they also differentiated into matureneurons expressing the dopaminergic marker tyrosine hy-droxylase. In this light, it is possible to speculate thatPACAP might be a proliferating agent, supplying new-born cells from the subventricular zone to the substantianigra where PACAP participates in promoting their finaldifferentiation as well as survival. Future studies will clarifywhether PACAP can be used as a pharmacologic agent tostimulate neurogenesis for therapeutic gain, whereas theuse of PACAP-specific antagonists will determine whetherPACAP plays a physiologic role in inducing endogenousneurogenesis.

Another interesting correlation between PACAPand neurogenesis comes from observations of PACAPregulation in the dentate gyrus after trauma (Skoglosa etal., 1999). This study showed PACAP to be stronglyupregulated in a model of brain injury, suggesting that thispeptide can exert trophic effects or perhaps induce neu-rogenesis at this level. This latter hypothesis is very in-triguing, considering that no major change in morphologyat the macroscopical level has been observed in eitherPACAP or PAC1 knockout mice (Hashimoto et al., 2000,2001), thus leaving open the possibility that PACAP playsa significant role generating newborn neurons after injury.

In conclusion, the data reported in this study dem-onstrates a novel role for PACAP in modulating neuralstem cell proliferation and provides new vistas to investi-gate this process. Further in vitro and in vivo studies willserve to elucidate more specifically the role of PACAP inneurogenesis.

ACKNOWLEDGMENTSWe thank R.G. Titus for his kind gift of Maxadilan,

and A. Andersson for skilled technical assistance.

REFERENCESArimura A. 1998. Perspectives on pituitary adenylate cyclase activating

polypeptide (PACAP) in the neuroendocrine, endocrine, and nervoussystems. Jpn J Physiol 48:301–331.

Bjorklund A, Lindvall O. 2000. Parkinson disease gene therapy movestoward the clinic. Nat Med 6:1207–1208.

Corbitt J, Hagerty T, Fernandez E, Morgan WW, Strong R. 2002. Tran-scriptional and post-transcriptional regulation of tyrosine hydroxylasemessenger RNA in PC12 cells during persistent stimulation by VIP andPACAP38: differential regulation by protein kinase A and protein kinaseC-dependent pathways. Neuropeptides 36:34–45.

Craig CG, Tropepe V, Morshead CM, Reynolds BA, Weiss S, van derKooy D. 1996. In vivo growth factor expansion of endogenous sub-ependymal neural precursor cell populations in the adult mouse brain.J Neurosci 16:2649–2658.

Crouch SP, Kozlowski R, Slater KJ, Fletcher J. 1993. The use of ATPbioluminescence as a measure of cell proliferation and cytotoxicity. J Im-munol Methods 160:81–88.

Dicicco-Bloom E, Lu N, Pintar JE, Zhang J. 1998. The PACAP ligand/receptor system regulates cerebral cortical neurogenesis. Ann N Y AcadSci 865:274–289.

Hansel DE, Eipper BA, Ronnett GV. 2001. Neuropeptide Y functions asa neuroproliferative factor. Nature 410:940–944.

Hashimoto H, Shintani N, Nishino A, Okabe M, Ikawa M, Matsuyama S,Itoh K, Yamamoto K, Tomimoto S, Fujita T, Hagihara N, Mori W,Koyama Y, Matsuda T, Nagata S, Baba A. 2000. Mice with markedly

reduced PACAP (PAC1) receptor expression by targeted deletion of thesignal peptide. J Neurochem 75:1810–1817.

Hashimoto H, Shintani N, Tanaka K, Mori W, Hirose M, Matsuda T,Sakaue M, Miyazaki J, Niwa H, Tashiro F, Yamamoto K, Koga K,Tomimoto S, Kunugi A, Suetake S, Baba A. 2001. Altered psychomotorbehaviors in mice lacking pituitary adenylate cyclase-activating polypep-tide (PACAP). Proc Natl Acad Sci USA 98:13355–13360.

Horner PJ, Gage FH. 2000. Regenerating the damaged central nervoussystem. Nature 407:963–970.

Jaworski DM, Proctor MD. 2000. Developmental regulation of pituitaryadenylate cyclase-activating polypeptide and PAC(1) receptor mRNAexpression in the rat central nervous system. Brain Res Dev Brain Res120:27–39.

Johansson CB, Svensson M, Wallstedt L, Janson AM, Frisen J. 1999. Neuralstem cells in the adult human brain. Exp Cell Res 253:733–736.

Kuhn HG, Svendsen CN. 1999. Origins, functions, and potential of adultneural stem cells. Bioessays 21:625–630.

Kuhn HG, Winkler J, Kempermann G, Thal LJ, Gage FH. 1997. Epidermalgrowth factor and fibroblast growth factor-2 have different effects onneural progenitors in the adult rat brain. J Neurosci 17:5820–5829.

Lerner EA, Shoemaker CB. 1992. Maxadilan. Cloning and functionalexpression of the gene encoding this potent vasodilator peptide. J BiolChem 267:1062–1066.

Lerner EA, Ribeiro JM, Nelson RJ, Lerner MR. 1991. Isolation of max-adilan, a potent vasodilatory peptide from the salivary glands of the sandfly Lutzomyia longipalpis. J Biol Chem 266:11234–11236.

McCabe BK, Silveira DC, Cilio MR, Cha BH, Liu X, Sogawa Y, HolmesGL. 2001. Reduced neurogenesis after neonatal seizures. J Neurosci21:2094–2103.

McKay R. 1997. Stem cells in the central nervous system. Science 276:66–71.

Momma S, Johansson CB, Frisen J. 2000. Get to know your stem cells.Curr Opin Neurobiol 10:45–49.

Moro O, Lerner EA. 1997. Maxadilan, the vasodilator from sand flies, is aspecific pituitary adenylate cyclase-activating peptide type I receptoragonist. J Biol Chem 272:966–970.

Nakatomi H, Kuriu T, Okabe S, Yamamoto S, Hatano O, Kawahara N,Tamura A, Kirino T, Nakafuku M. 2002. Regeneration of hippocampalpyramidal neurons after ischemic brain injury by recruitment of endog-enous neural progenitors. Cell 110:429–441.

Nicot A, DiCicco-Bloom E. 2001. Regulation of neuroblast mitosis isdetermined by PACAP receptor isoform expression. Proc Natl Acad SciUSA 98:4758–4763.

Park SY, Choi HJ, Hwang O. 1999. Regulation of basal expression ofcatecholamine-synthesizing enzyme genes by PACAP. Mol Cells 9:146–151.

Paxinos G, Franklin K. 2001. The mouse brain in stereotaxic coordinates.2nd ed. San Diego, CA: Academic Press.

Pencea V, Bingaman KD, Wiegand SJ, Luskin MB. 2001. Infusion ofbrain-derived neurotrophic factor into the lateral ventricle of the adult ratleads to new neurons in the parenchyma of the striatum, septum, thala-mus, and hypothalamus. J Neurosci 21:6706–6717.

Rawlings SR. 1994. PACAP, PACAP receptors, and intracellular signaling.Mol Cell Endocrinol 101:5–9.

Seaberg RM, van der Kooy D. 2002. Adult rodent neurogenic regions: theventricular subependyma contains neural stem cells, but the dentate gyruscontains restricted progenitors. J Neurosci 22:1784–1793.

Shingo T, Gregg C, Enwere E, Fujikawa H, Hassam R, Geary C, Cross JC,Weiss S. 2003. Pregnancy-stimulated neurogenesis in the adult femaleforebrain mediated by prolactin. Science 299:117–120.

Shioda S, Ozawa H, Dohi K, Mizushima H, Matsumoto K, Nakajo S,Takaki A, Zhou CJ, Nakai Y, Arimura A. 1998. PACAP protectshippocampal neurons against apoptosis: involvement of JNK/SAPK sig-naling pathway. Ann N Y Acad Sci 865:111–117.

214 Mercer et al.

Skoglosa Y, Lewen A, Takei N, Hillered L, Lindholm D. 1999. Regulation ofpituitary adenylate cyclase activating polypeptide and its receptor type 1 aftertraumatic brain injury: comparison with brain-derived neurotrophic factorand the induction of neuronal cell death. Neuroscience 90:235–247.

Sun Y, Jin K, Xie L, Childs J, Mao XO, Logvinova A, Greenberg DA.2003. VEGF-induced neuroprotection, neurogenesis, and angiogenesisafter focal cerebral ischemia. J Clin Invest 111:1843–1851.

Takei N, Skoglosa Y, Lindholm D. 1998. Neurotrophic and neuroprotec-tive effects of pituitary adenylate cyclase-activating polypeptide (PACAP)on mesencephalic dopaminergic neurons. J Neurosci Res 54:698–706.

Waschek JA. 2002. Multiple actions of pituitary adenylyl cyclase activatingpeptide in nervous system development and regeneration. Dev Neurosci24:14–23.

Vaudry D, Gonzalez BJ, Basille M, Yon L, Fournier A, Vaudry H. 2000.Pituitary adenylate cyclase-activating polypeptide and its receptors: fromstructure to functions. Pharmacol Rev 52:269–324.

Zhao M, Momma S, Delfani K, Carlen M, Cassidy RM, Johansson CB,Brismar H, Shupliakov O, Frisen J, Janson AM. 2003. Evidence forneurogenesis in the adult mammalian substantia nigra. Proc Natl Acad SciUSA 100:7925–30.

PACAP Promotes Neural Stem Cell Proliferation 215