Tumour necrosis factor-α impairs neuronal differentiation but not proliferation of hippocampal neural precursor cells: Role of Hes1

Molecular and Cellular Neuroscience 43 (2010) 127–135

Contents lists available at ScienceDirect

Molecular and Cellular Neuroscience

j ourna l homepage: www.e lsev ie r.com/ locate /ymcne

Tumour necrosis factor-α impairs neuronal differentiation but not proliferation ofhippocampal neural precursor cells: Role of Hes1

Aoife Keohane, Sinead Ryan, Eimer Maloney, Aideen M. Sullivan, Yvonne M. Nolan ⁎Department of Anatomy and Neuroscience, University College Cork, Ireland

⁎ Corresponding author. Fax: +353 21 427 3518.E-mail address: [email protected] (Y.M. Nolan).

1044-7431/$ – see front matter © 2009 Elsevier Inc. Adoi:10.1016/j.mcn.2009.10.003

a b s t r a c t

Article history:Received 27 April 2009Revised 27 September 2009Accepted 9 October 2009Available online 17 October 2009

Keywords:Tumour necrosis factor-alphaNeurogenesisNeural precursor cellHippocampusHes1

Tumour necrosis factor-α (TNFα) is a pro-inflammatory cytokine, which influences neuronal survival andfunction yet there is limited information available on its effects on hippocampal neural precursor cells(NPCs). We show that TNFα treatment during proliferation had no effect on the percentage of proliferatingcells prepared from embryonic rat hippocampal neurosphere cultures, nor did it affect cell fate towardseither an astrocytic or neuronal lineage when cells were then allowed to differentiate. However, when cellswere differentiated in the presence of TNFα, significantly reduced percentages of newly born and post-mitotic neurons, significantly increased percentages of astrocytes and increased expression of TNFαreceptors, TNF-R1 and TNF-R2, as well as expression of the anti-neurogenic Hes1 gene, were observed. Thesedata indicate that exposure of hippocampal NPCs to TNFα when they are undergoing differentiation but notproliferation has a detrimental effect on their neuronal lineage fate, which may be mediated throughincreased expression of Hes1.

© 2009 Elsevier Inc. All rights reserved.

Introduction

The hippocampus is one of a small number of regions in themammalian brain that are known to have neurogenic potential inadulthood. Here, the formation of new granule cells in the dentategyrus of the hippocampus can be affected by a large number ofphysiological and pathological stimuli including age, stress, hormonesand diet (Taupin, 2005). Inflammation also contributes to themicroenvironment of the neurogenic ‘niche’ where neural precursorcells (NPCs) reside and has been shown to have a negative impact onadult hippocampal neurogenesis (Monje et al., 2003). Indeed, pro-inflammatory cytokines have been shown to be involved in numeroushippocampal pathologies (reviewed by McGeer and McGeer, 2004).Due to the fact that the overwhelming majority of hippocampalneurogenesis occurs pre-natally, however, it is important to investi-gate the mechanisms by which pro-inflammatory cytokines affectembryonic neurogenesis and how they determine the fate of NPCs.This information may then be used to develop transplantationtherapies or indeed to promote activation of endogenous NPCs toresist diseases of the adult hippocampus.

The pro-inflammatory cytokine TNFα has been shown topromote neuronal death in the hippocampus (Zhao et al., 2001).It is a member of the TNF ligand superfamily and the majority ofits activity is mediated through its 17 kDa soluble form (reviewed

ll rights reserved.

by MacEwan, 2002). The main cell type, which secretes TNFα, isthe activated monocyte, but in the brain TNFα has been shown tobe produced by both resident microglia and infiltrating macro-phages (Renno et al., 1995), by astrocytes (Lieberman et al., 1989)and by neurons (Knoblach et al., 1999; Liu et al., 1994). It isprimarily implicated in immunity and inflammation though it hasalso been shown to be involved in cellular processes such asapoptosis, survival and proliferation (reviewed by Gaur andAggarwal, 2003). TNFα exerts these effects by activating tworeceptor subtypes; TNF receptor type-1 (TNF-R1), which containsan intracellular death domain and mediates TNFα's death-inducingactivity, and TNF receptor type-2 (TNF-R2), which contributes tocell survival (reviewed by MacEwan, 2002).

TNFα has been shown to negatively affect embryonic and adultneurogenesis in vitro. Ben-Hur et al. (2003) showed that treatment ofNPCs obtained from newborn rat striatum with TNFα inhibited theproliferation of NPCs in these cultures but did not promote apoptoticcell death. It has been shown byWong et al. (2004) that TNFα inhibitsthe proliferative ability of neural stem cell lines derived from thesubventricular zone (SVZ) (a neurogenic region) of adult mice. Whenthe effect of TNFα on differentiation was examined by Liu et al.(2005), they demonstrated that TNFα reduced the number of NPCsadopting a neuronal phenotype under differentiating conditions inembryonic rat whole-brain neurosphere cultures. Conversely, Ber-nardino et al. (2008) have recently reported that exposure of NPCsderived from the SVZ of neonatal mice to TNFα, induced differenti-ation to a neuronal phenotype. They also demonstrated that exposureof these cells to a low dose of TNFα induced cell proliferation whileexposure to a higher dose induced apoptotic cell death.

128 A. Keohane et al. / Molecular and Cellular Neuroscience 43 (2010) 127–135

Hairy-enhancer-of-split 1 (Hes1) is a nuclear target of the Notchreceptor and belongs to the basic helix–loop–helix family oftranscription factors. It is highly expressed by NPCs and it negativelyregulates neurogenesis by antagonising pro-neural genes and thusrepressing the commitment of NPCs to a neuronal fate (Tanigaki et al.,2001; reviewed by Teng et al., 2009). It has been shown that there isreduced expression of Hes1 in embryonic rat hippocampal neuronsduring differentiation and that its expression inhibited neuriteoutgrowth (Castella et al., 1999). It has also been demonstrated thatactivation of a component of the receptor for the pro-inflammatorycytokine interleukin-6 increases Hes1 expression in forebrain neuralstem cells (Chojnacki et al., 2003).

To date, there have been relatively few studies that have examinedthe effect of TNFα on hippocampal NPCs. Cacci et al. (2005) haveshown that TNFα compromises the survival of newly formedembryonic rat hippocampal cells in the progenitor cell line, HiB5.Studies in TNF-R1−/− mice demonstrated increased proliferation ofneuronal NPCs in the subgranular zone (SGZ) of the hippocampus,implying that TNF-R1 is a negative regulator of adult hippocampalneurogenesis (Iosif et al., 2006). Thus, although TNFα has been shownto affect proliferation and lineage fate of NPCs, little is known aboutthe effects of TNFα on hippocampal NPCs before or after they adopt aneuronal or astroglial phenotype. Furthermore, there are no reportsshowing that stimulation of NPCs with a pro-inflammatory cytokinealters Hes1 expression. The aim of the present series of experimentswas to investigate the effect of TNFα on the proliferation anddifferentiation of both neuronal and astroglial lineages derived fromembryonic rat hippocampal NPCs.

Results

Embryonic rat hippocampal cells were cultured in ‘proliferationmedium’ as neurospheres for 7 DIV and subsequently plated as singlecells in ‘differentiation medium.’ To determine whether TNFα has amodulatory effect on neuronal and astroglial differentiation, differ-entiating cells were exposed to TNFα (1–100 ng/mL) for 7 DIV.Immunofluorescent staining performed after 7 days of differentiationshowed that the percentage of NPCs (nestin-positive cells) in thecultures was significantly higher in the presence of all concentrationsof TNFα compared to control conditions (pb0.01; ANOVA) (Fig. 1A).TNFα (10–100 ng/mL) significantly decreased the percentage ofnewly born (DCX-positive) neurons (Fig. 1B) and post-mitotic (βIII-tubulin-positive) neurons (Fig. 1C) compared to those in the controlcultures (pb 0.01; ANOVA). The percentage composition of astrocytes(GFAP-positive cells) increased significantly after incubation withTNFα (10–100 ng/mL) compared to control cultures (pb 0.01;ANOVA) (Fig. 1D). There was no significant difference in the totalnumber of cells (as determined by bisbenzimide staining) betweenany of the TNFα-treated and untreated cultures (data not shown).

Morphological analysis of DCX-positive cells during the differen-tiation phase revealed that exposure to TNFα significantly increasedthe mean cell somal area per DCX-positive cell (pb 0.01; ANOVA)(Fig. 2A). The total number of neurites was significantly decreased byTNFα treatment (pb 0.01; ANOVA) (Fig. 2B). Analysis of the degreeof branching demonstrated that the number of primary, secondaryand tertiary neurites was significantly decreased by 10 ng/mL TNFα(pb 0.01; ANOVA) while the higher dose of TNFα (100 ng/mL)significantly increased primary and secondary branch numbers butdid not have a significant effect on the number of tertiary branches(Figs. 2C–E).

To examine the effect of TNFα on cell proliferation, BrdUincorporation was examined in neurospheres that were expandedfor 7 DIV in the presence of TNFα at a concentration of 10 ng/mL. Thisdose was chosen as an optimal dose from results of the differentiationstudies. Cells were exposed to BrdU for the final 12 h in vitro. TNFαhad no effect on the number of BrdU-positive cells (Fig. 3A).When the

percentage of nestin-positive, DCX-positive and GFAP-positive cellswas assessed, it was determined that TNFα had no significant effecton NPCs (Fig. 3B) or on newly born neurons (Fig. 3C) but that itsignificantly decreased the percentage of astrocytes (pb 0.01;Student's t-test) (Fig. 3D), suggesting that TNFα influences thedevelopment of astrocytes rather than immature neurons underproliferative culture conditions. Following 7 days of proliferation,there was no effect of TNFα on the mean total cell number in tissueculture flasks (data not shown).

Proliferating cells were exposed to TNFα and the differentiation ofthese cells towards a neuronal or astroglial fate was also examined.After 7 days of differentiation, there was no significant effect of TNFαon the percentage of DCX-positive cells (Fig. 4A) or GFAP-positivecells (Fig. 4D) compared to untreated cultures. These results indicatethat exposure to TNFα during proliferation does not affect the numberof cells proceeding to either a neuronal or astrocytic lineage underdifferentiation conditions. This is in contrast to TNFα treatment underdifferentiation conditions, where it has a detrimental effect onneuronal development and a stimulatory effect on astroglial devel-opment. TNFα treatment during proliferation had no significant effecton the mean total cell number following 7 days of differentiation, asdetermined by bisbenzimide staining (data not shown).

The gene expression of the TNFα receptors, TNF-R1 and TNF-R2, indissociated cells from hippocampal neurospheres was examined byRT-PCR. Fig. 5A shows that both receptor types were expressed inuntreated (lane 1) and TNFα-treated (lane 2) cultures following7 days of proliferation, with TNF-R2 expression present at a higherintensity. To examine whether cells expressed TNF-R1 and TNF-R2during differentiation, RNA was isolated from control and TNFα-treated culture preparations following 7 days of differentiation. RT-PCR analysis demonstrated the expression of both receptors in controlcells (lane 1) and at a higher intensity in TNFα-treated cells (lane 2)(Fig. 5B).

As the Hes1 gene is an inhibitor of neurogenesis, and may mediatethe anti-neurogenic effect of TNFα on hippocampal NPCs, we decidedto examine its expression in NPCs that had been treated with TNFαduring proliferation or differentiation. RT-PCR analysis demonstratesthat TNFα treatment during proliferation has no effect on theexpression of Hes1 when examined after 7 days (Fig. 5A; lane 1(control), lane 2 (TNFα)) whereas treatment during differentiationincreases the expression of Hes1 after 7 days (Fig. 5B; lane 1 (control),lane 2 (TNFα)).

Discussion

The present study explored the possibility that the pro-inflam-matory cytokine TNFα exerts different effects on the proliferation anddifferentiation of embryonic hippocampal NPCs, dependent onwhether they are exposed to it during their proliferative or diffe-rentiation phase. Consistent with the findings of other groups, ourdata suggest that TNFα prevents the birth of neurons (Cacci et al.,2005; Iosif et al., 2006; Liu et al., 2005). We show that TNFαtreatment of hippocampal NPCs during differentiation rather thanproliferation affects cell lineage determination (Figs. 6B and C).Furthermore, TNFα decreased the percentage of cells proceedingtowards an astrocytic lineage but not towards a neuronal lineage,when administered during the proliferative phase (Fig. 6A). We alsoshow that TNF-R1 and TNF-R2 are expressed on both proliferatingand differentiated hippocampal NPCs, and that their expression isupregulated on differentiated but not proliferating NPCs in thepresence of TNFα. When we examined the expression of Hes1, ananti-neurogenic gene, we observed that it was upregulated on NPCsafter treatment with TNFα during differentiation. Thus it is possiblethat TNFα may exert its inhibitory effects on neuronal differentiationvia activation of either or both TNF-R1 and TNF-R2 and upregulationof Hes1 expression.

Fig. 1. Effect of TNFα during differentiation on lineage fate of hippocampal NPCs. Percentage of nestin-positive cells (A), DCX-positive cells (B), βIII-tubulin-positive neurons (C) andGFAP-positive astrocytes (D) out of the total cells in untreated and TNFα-treated (1–100 ng/mL) cultures after 7 DIV under differentiation conditions. Data are expressed asmeans±SEM. ⁎⁎ pb 0.01 vs. control; ANOVAwith post hoc Dunnett's test; n=3. Representative photomicrographs of cells stained for nestin (red) and DCX (green) in untreated (E), 1 ng/mL(F), 10 ng/mL (G), 20 ng/mL (H), 50 ng/mL (I) and 100 ng/mL (J) TNFα-treated cultures, and of cells stained for βIII-tubulin (red) and GFAP (green) in untreated (K), 1 ng/mL (L),10 ng/mL (M), 20 ng/mL (N), 50 ng/mL (O) and 100 ng/mL (P) TNFα-treated cultures after 7 DIV. All cells were counterstained with bisbenzamide (blue). Scale bar=50 μm.

129A. Keohane et al. / Molecular and Cellular Neuroscience 43 (2010) 127–135

Fig. 2. Effect of TNFα during differentiation on cell somal area and neurite branching of hippocampal newly born neurons. Somal area (μm2) (A), total number of neurites (B) andnumber of primary (C), secondary (D) and tertiary (E) neurites were measured in untreated and in TNFα-treated (10 and 100 ng/mL) DCX-positive cells after 7 DIV underdifferentiation conditions. Data are expressed as means±SEM. ⁎pb 0.05, ⁎⁎pb 0.01 vs. control; ANOVAwith post hoc Dunnett's test; n=3. Representative photomicrographs of cellsstained for DCX in untreated (F), 10 ng/mL (G) and 100 ng/mL (H) TNFα-treated cultures. Scale bar=50 μm.

130 A. Keohane et al. / Molecular and Cellular Neuroscience 43 (2010) 127–135

TNFα has previously been implicated in modulating hippocampalneurogenesis; however, the exact profile of its effects on proliferationand differentiation of hippocampal NPCs has not been clearly defined.Liu et al. (2005) have demonstrated that TNFα exerts detrimentaleffects on differentiation of NPCs, derived from embryonic rat wholebrain into a neuronal lineage as assessed by βIII-tubulin immuno-reactivity. Our results show a similar effect of TNFα on hippocampalNPCs, when treated under differentiation conditions. We have alsoextended this result by establishing that several doses of TNFα inhibitdifferentiation of NPCs to cells that express the early neuronal lineagemarker DCX, and promote differentiation of NPCs to an astroglial fate.The percentages of cells immunoreactive for nestin, an intermediatefilament protein predominantly expressed by neural stem cells, werealso increased after TNFα treatment of NPCs during differentiation. Asthere is an overlap in the expression profile of nestin and GFAP (Krumand Rosenstein, 1999), this result suggests that TNFαmaintains thesecells in an undifferentiated or astrocytic state, or indeed induces anupregulation of GFAP expression in glial progenitors, while simulta-neously preventing direction of the cells towards a neuronal fate.Because reactive astrogliosis is a characteristic response of astrocytesto inflammation in the central nervous system, and nestin expressionis re-induced in reactive astrocytes (Lin et al., 1995), it is also likelythat the increased nestin expression in these cells is due to TNFα-

induced astrogliosis. In support of this theory, we observed thatnestin-positive cells that had been exposed to TNFα, especially athigher doses, appeared to have enlarged soma and thickenedprocesses, indicative of a reactive astrocytic phenotype.

In an attempt to explore the negative impact of TNFα on neuronaldifferentiation, we examined the morphology of hippocampal newlyborn neurons. We observed that TNFα increased the cell somal areaand reduced the neuritic branching in these cells. Another study haspreviously implicated TNFα in the regulation of cellular morphologyduring hippocampal development. TNFα-treated embryonic murinehippocampal neurons cultured in the presence of glial cells displayeda reduction in length and branching of neurites, and this effect wasabsent in hippocampal neurons cultured from TNF receptor-deficientmice (Neumann et al., 2002). We propose that the reactiveastrogliosis, which results from TNFα treatment in our cultures ofNPCs, induces an increase in cell somal area and a decrease in neuriteoutgrowth of newly born neurons. This notion is supported by studiesthat demonstrated that GFAP-null astrocytes provided a favourableenvironment for neurite growth and survival (Menet et al., 2000) andthat depletion of reactive astrocytes promoted neurite outgrowthafter injury in a transgenic model (Bush et al., 1999).

Having shown that TNFα affected differentiation of hippocampalNPCs when they were exposed to TNFα under differentiation

Fig. 3. Effect of TNFα on NPCs under proliferative culture conditions. NPCs in hippocampal neurospheres were treated with TNFα (10 ng/mL) for 7 DIV under proliferationconditions and BrdU+ cells (A), nestin+ cells (B), DCX+ cells (C) and GFAP+ cells (D) as a percentage of total cells were calculated. Data are expressed as means±SEM. ⁎⁎pb 0.01;Student's t-test; n=3.

131A. Keohane et al. / Molecular and Cellular Neuroscience 43 (2010) 127–135

conditions, we then assessed whether TNFα modulated proliferationand differentiation when cells were treated with TNFα underproliferation conditions. We observed that TNFα treatment did notinduce a significant difference in the percentage of total BrdU-positivecells, nestin-positive cells or DCX-positive cells, after proliferation ofthese cultures. The percentage composition of GFAP-positive cells inour cultures was decreased as a result of TNFα exposure duringproliferation, suggesting that TNFα prevents NPCs from proceedingtowards an astrocytic lineage but has no effect on neuronaldevelopment during the proliferative phase. As a substantial percent-age of GFAP-positive cells are likely to also be positive for nestin, but asnestin-positive cells are not affected by TNFα, we speculate that GFAP-positive/nestin-negative cells only are susceptible to TNFα duringproliferation. However, investigation of TNF receptor expression onthese cell types would be necessary to clarify this hypothesis. Thesedata differ to those presented by Ben-Hur et al. (2003) who found thatTNFα significantly reduced the percentage of BrdU-positive cellsexpanded as neurospheres from newborn rat striatum. Contrary tothese and to our present results, Widera et al. (2006) has implicatedTNFα as a positive regulator of neurogenesis. Widera and co-workersfound that TNFα treatment of proliferating neurospheres from adultrat SVZ increased BrdU incorporation and volume of the neurospheres.Likewise, Bernardino et al. (2008) recently explored the effect of TNFαon proliferation and differentiation in cultures of NPCs derived fromneonatal mice SVZ. They found that TNFα treatment during prolifer-ation increased cell proliferation and differentiation of cells towards aneuronal phenotype. The apparent discrepancies between thesestudies may be explained by brain regional or age differences insusceptibility to the effects of TNFα. When we allowed the cells thathad been treated with TNFα under proliferation conditions todifferentiate (in the absence of TNFα), we found that TNFα did notaffect the lineage fate of the cells towards either an astrocytic or aneuronal fate. This is in contrast to our previous experiment, wherecells treated with TNFα during differentiation displayed a decrease inneurogenesis and an increase in astrogenesis. The dose of TNFαused inthis second experiment (10 ng/mL) produced a maximal response

when administered during differentiation (Fig. 1) and so aninadequate concentration of TNFα is unlikely to have been the reasonfor this result. These findings are in agreement with those of Ben-Huret al. (2003) who found that exposure of newborn rat striatal cells toTNFα did not affect lineage fate if exposed prior to differentiation.Taken together, these results show that TNFα decreases thepercentage of cells proceeding towards a neuronal fate and increasesthe percentage of cells proceeding towards an astrocytic fate whenNPCs are exposed during differentiation, but has no effect when theyare exposed during proliferation. This finding emphasises the im-portance of regulation of the inflammatory environment duringdifferent phases of NPC development.

We quantified total cell numbers following 7 days of TNFα-treatment during proliferation, following 7 days of differentiationafter TNFα-treatment during proliferation, and following 7 days ofTNFα-treatment during differentiation. We observed that there wasno difference in total cell numbers between untreated and TNFα-treated cells after any of these protocols, suggesting that TNFα doesnot affect the viability of hippocampal NPCs, although cell deathassays would be required to confirm this observation. Similarly, Ben-Hur et al. (2003) treated newborn rat striatal neurospheres with TNFαfor 48 h during proliferation and found that TNFα had no effect on thepercentage of apoptotic cells or on cell viability. Furthermore, thefindings of Wong et al. (2004) are of interest here, as they found thatTNFα was not toxic to an adult mouse-derived neural stem cell linewhen exposed to cells for 24 h during differentiation. However, it hasalso been shown that exposure to TNFα for 36 h reduced the numberof hippocampal progenitor cells in an embryonic rat-derived cell line(Cacci et al., 2005). Therefore, it is likely that differences in cellularpreparations, brain regions, species, dose, age and exposure time toTNFα may have an impact on the response of the cells to TNFα.

As TNFα has an effect on hippocampal neural differentiation, it islikely that it acts through one of its receptors. It has been shown thatNPCs derived from neonatal rat striatum express TNF-R1 (Ben-Hur etal., 2003). Both TNF-R1 and TNF-R2 have been detected on culturedhuman fetal brain (Sheng et al., 2005) and on NPCs prepared from

Fig. 4. Effect of TNFα during proliferation on lineage fate of hippocampal NPCs. NPCs in hippocampal neurospheres were treated with TNFα (10 ng/mL) for 7 DIV under proliferationconditions. Cells were then allowed to differentiate untreated for 7 DIV and the percentages of DCX-positive neurons (A) and GFAP-positive astrocytes (D) out of the total cells inuntreated and TNFα cultures were calculated. Data are expressed as means±SEM;. n=3. Representative photomicrographs of cells stained for DCX (green) and GFAP (green) inuntreated (B, E) and TNFα-treated (C, F) cultures. All cells were counterstained with bisbenzamide (blue). Scale bar=50 μm.

132 A. Keohane et al. / Molecular and Cellular Neuroscience 43 (2010) 127–135

adult rat SVZ (Widera et al., 2006). In this study, RT-PCR analysis ofproliferating embryonic hippocampal NPCs showed expression ofTNF-R1 and, to a higher extent, of TNF-R2 in both control and TNFα-treated cultures. Similar results have been obtained by Iosif et al.(2006) in adult hippocampal NPCs obtained from both rats and miceand by Cacci et al. (2005) in an embryonic rat hippocampus-derivedprogenitor cell line. However, neither of these groups has investigatedthe effect of ligand binding on receptor expression. To determine ifTNF-R1 and TNF-R2 gene transcripts are present following 7 days ofdifferentiation in the presence or absence of TNFα, RT-PCR wascarried out on differentiated neural cells. While expression of bothreceptor types appeared absent in control cultures, expression wasinduced following TNFα treatment, which indicates that ligation ofTNFα to either TNF-R1 or TNF-R2 or both, may act to suppresshippocampal neuronal differentiation or promote astroglial differen-tiation. Knockout studies on adult mice have shown that TNF-R1 is anegative regulator of hippocampal neurogenesis (Iosif et al., 2006),while the actions of TNFα through TNF-R2 have been suggested to beneuroprotective (Heldmann et al., 2005). Even when TNFα is present

at high concentrations, TNF-R2 has a lower affinity for TNFα than TNF-R1 in cultures of primary hippocampal neurons (Yang et al., 2002), soit is likely that the anti-neurogenic activity of TNFα detected in ourstudies is mediated through TNF-R1.

To determine the contribution of Hes1 to transcriptional regulationof TNFα-induced impairment in neuronal differentiation, we investi-gated the expression of Hes1, a transcription factor that inhibits theadaptation of a neuronal phenotype in proliferating NPCs and is anegative regulator of neuronal differentiation (Feder et al., 1993;Tanigaki et al., 2001).Wemeasured the transcript levels of this gene incells following 7 days under proliferation conditions and following 7days under differentiation conditions in the presence or absence ofTNFα. Hes1 represses the commitment of undifferentiated cells to aneuronal fate (Nakamura et al., 2000) and so the increased expressionof Hes1 we observed in TNFα-treated cells under differentiationconditions is likely to promote the nestin-positive cells present inthese cultures to remain in an undifferentiated state. Under differen-tiation conditions, the downregulation of Hes1 is necessary forautonomous hippocampal differentiation (Castella et al., 1999); our

Fig. 5. Gene expression of TNF-R1, TNF-R2 and Hes1 on proliferating and differentiatedhippocampal NPCs. RT-PCR analysis of the expression of TNF-R1, TNF-R2, Hes1 and 18SRNA (A) on hippocampal NPCs proliferated for 7 DIV in untreated (lane 1) and TNFα-treated (10 ng/mL) (lane 2) cultures and (B) on hippocampal NPCs differentiated for 7DIV in untreated (lane 1) and TNFα-treated (10 ng/mL) (lane 2) cultures. Lane L is thesize marker. PCR products of 244, 264, 318 and 359 bp size indicate TNF-R1, TNF-R2,Hes1 and 18S RNA expression, respectively.

133A. Keohane et al. / Molecular and Cellular Neuroscience 43 (2010) 127–135

data intimate that TNFα treatment inhibited neuronal differentiationdue to its promotion ofHes1 expression. Previous studies on the role ofHes1 in the regulation of neuronal differentiation have focused on itsinhibitory effect on nerve growth factor-induced neurite outgrowth(Castella et al., 1999; Salama-Cohen et al., 2006). To date, only onestudy has suggested that a pro-inflammatory cytokine may exert ananti-neurogenic effect by increasing expression of Hes1; Chojnacki etal. (2003) demonstrated that activation of glycoprotein 130, acomponent of the interleukin-6 receptor, stimulates the Notch1pathway and increases the expression of Hes1 in forebrain neuralstem cells. It has been recognised for many years that inflammatorystimuli activate the transcription factor nuclear factor kappa B (NFκB)in the hippocampus (Kaltschmidt et al., 1993; Kelly et al., 2003) and ithas been demonstrated that inhibition of NFκB activity decreasedHes1mRNA levels in embryonic mouse hippocampal neuronal cultures(Salama-Cohen et al., 2005). Aguilera et al. (2004) have also shownthat IκBα, the inhibitor of NFκB, participates in the regulation of Hes1in mouse embryonic fibroblasts in response to TNFα. Our resultsprovide further support for an interaction between NFκB and Notchsignalling in an inflammatory environment, specifically in hippocam-pal NPCs.

Fig. 6. Schematic representation of the impact of TNFα on neural cell lineage determinatidecrease in the percentage of cells proceeding towards an astrocytic lineage but not towardexpression of TNF-R1, TNF-R2 and Hes1 on these cells. Treatment of hippocampal NPCs wdetermination, and expression of TNF-R1, TNF-R2 and Hes1.

This study provides experimental evidence that treatment ofhippocampal NPCs with TNFα during differentiation but not prolif-eration affects neuronal cell lineage determination and the morphol-ogy of newly born neurons, possibly through the involvement of TNF-R1 and Hes1. It is likely therefore that therapeutic interventionstargeting the inflammatory process aremost important when cells areundergoing differentiation. The possibility of increasing survival ofnewly formed neurons, or of increasing the percentage of cellsproceeding towards a neuronal lineage, is of utmost importance in thetreatment of age-related neurological disorders such as Alzheimer'sdisease, where both hippocampal neuronal loss and inflammation areknown to occur.

Experimental methods

Preparation of rat hippocampal NPCs

Embryonic day (E) 18 rat embryos were obtained by laparotomyfollowing anaesthesia with halothane. Hippocampi were removedusing a dissecting microscope, collected in Hank's balanced saltssolution (HBSS) (Sigma) and finely chopped. The tissue wascentrifuged for 5 min at 700×g and the supernatant removed. Thepelletwas incubatedwith 0.1% trypsin-EDTA (0.1%; Sigma) for 5min at37 °C to enzymatically dissociate the cells. Trypsin activity wasquenched by adding soybean trypsin inhibitor (0. 5 mg/mL) (Sigma)and the solution was triturated through a flame-polished Pasteurpipette. After centrifugation for 5 min at 700×g, the pellet wasresuspended in 1 mL of pre-warmed ‘basic medium’ (Dulbecco'smodified eagles's medium (DMEM)-F12 (Sigma); 1% antibiotic-antimycotic solution (Sigma); 200 mM L-glutamine (Sigma); 33 mMD-glucose (Sigma)). The cells were seeded in T-25 culture flasks at adensity of 2×106 cells per flask in 10 mL of ‘proliferation medium’

(10mL of ‘basic medium’; supplementedwith 2% B-27 (Gibco), 20 ng/mL of epidermal growth factor (EGF; Sigma) and 20 ng/mL of basicfibroblast growth factor (bFGF) (Chemicon). Cells in the neurosphereswere allowed to proliferate for 7 days in vitro (DIV) under a humidifiedatmosphere containing 5% CO2 at 37 °C. Half of the ‘proliferationmedium’ was replaced every 2 to 3 days. After proliferation for 7 DIV,the neurospheres were enzymatically andmechanically dissociated toa single cell suspension as described above. After re-suspension in‘basic medium,’ a portion of the cell suspension was centrifuged at13,000×g for 2 min, the supernatant was removed and the pellet wasfrozen at −80 °C until required for RT-PCR analysis. The remaining

on. (A) Hippocampal NPCs treated with TNFα during the proliferative phase causes as a neuronal lineage. It has no effect on the percentage of proliferating cells or on theith TNFα during differentiation (C) rather than proliferation (B) affects cell lineage

134 A. Keohane et al. / Molecular and Cellular Neuroscience 43 (2010) 127–135

cells were seeded at a density of 5×104 cells/poly-D-lysine-coated13 mm glass coverslip in 24-well tissue culture plates and wereincubated for 1 h at 5%CO2/37 °C to allowattachment to the coverslips.For experiments examining the effects of TNFα on cell differentiation,wells were flooded with 500 μL of ‘differentiating medium’ (‘basicmedium’ supplemented with 1% B-27). The cells were allowed todifferentiate for 7 DIV under a humidified atmosphere containing 5%CO2 at 37 °C. Half of the ‘differentiation medium’ was replaced everysecond day.

Treatment of rat hippocampal NPCs

To determine the effect of TNFα treatment during differentiationon cell lineage determination, recombinant rat TNFα (1–100 ng/mL;R&D Systems) was added to the ‘differentiation medium’ and the cellswere allowed to differentiate for 7 DIV following 7 days ofproliferation. To determine the effect of TNFα treatment onproliferation, recombinant rat TNFα (10 ng/mL) was added to the‘proliferation medium’ and the cells were allowed to proliferate for 7DIV. 5′-Bromo-deoxyuridine (BrdU) (final conc. 0. 2 μM) (Sigma) wasadded to the medium for the last 12 h before collecting theneurospheres for evaluation of proliferation (Caldwell et al., 2005).To determine the effect of TNFα treatment during proliferation on celllineage determination, recombinant rat TNFα (10 ng/mL) was addedto the ‘proliferation medium’ and the cells were allowed to proliferatefor 7 DIV. Dissociated cells were then allowed to differentiate for 7 DIVbefore assessment of cell phenotypes.

RT-PCR

Total cellular RNA was extracted from hippocampal neurospheresor from differentiated NPCs that were cultured for 7 DIV in thepresence or absence of TNFα (10 ng/mL) using anRNeasy kit (Qiagen).RNA (0. 5 μg) was reverse-transcribed in a reaction containing oligo(dT) (500 ng), random primers (500 ng), MgCl2 (25 mM), dNTPs (0.5 mM each), RNase inhibitor (20 U), reaction buffer and reversetranscriptase (Promega) in a final volume of 20 μL. The reaction wasperformed at 40 °C for 1 h, followed by 15 min incubation at 70 °C toinactivate the reverse transcriptase. Two microliters of the cDNA wasamplified in a PCRmix containing PCR buffer, MgCl2 (1. 5mM), 10mMof each primer, dNTP mix (1.25 mM) and Taq DNA polymerase (1 U)(Promega) in afinal volumeof 25 μL. The following primerswere used:TNF-R1, 5′-CCCAGGACTCAGGTACTGCCGT-3′ (+), 5′-CCCA-GAGTGGGGTTGAAGCCGG-3′ (−); TNF-R2, 5′-GATGACAAATCCCAG-GATGCAATAGG-3′ (+), 5′-TGCTACAGACGTTCACGATGAAGG-3′ (−);Hes1, 5′-ATGCCAGCTGATATAATGGAG-3′ (+), 5′-CACGCTCGGGTCTGTGCTGAGAGC-3′ (−); 18S RNA, 5′-TCAAGAACGAAAGTCGGAGG-3′ (+), 5′-GGACATCTAAGGGCATCACA-3′ (−). PCR was performedwith the following temperatures; an initial denaturation for 2 min at95 °C (TNF-R1, TNF-R2 and 18S RNA) or 94 °C (Hes1), amplification for30 cycles (TNF-R1: 95 °C for 40 s, 54 °C for 40 s and 72 °C for 30 s; TNF-R2: 95 °C for 1min, 59 °C for 1 min and 72 °C for 1 min; Hes1: 94 °C for30 s, 57 °C for 30 s and 72 °C for 30 s; 18S RNA; 95 °C for 30 s, 57 °C for30 s and 72 °C for 30 s). The final extension was at 72 °C for 5 min. Tocontrol for genomic DNA contamination, RT-PCR was performed inwhich the cDNA synthesis reaction was carried out in the absence ofreverse transcriptase. The amplification products were electrophor-esed on a 1% agarose gel containing ethidium bromide and visualisedon an ultraviolet transilluminator.

Immunocytochemistry

Cells attached to coverslips were fixed with ice-cold methanol at−20 °C for 10 min and 4% paraformaldehyde for 15 min at roomtemperature (RT). Following three 5min washes with 0.02% Triton-X-100/phosphate buffered saline (PBS), non-specific binding was

blocked by incubating the cells overnight at 4 °C with 5% horseserum/20% Triton-X in PBS. Antibodies directed against nestin (1:200;goat polyclonal; Santa Cruz Biotechnology), βIII-tubulin (1:300;mouse monoclonal; Promega), doublecortin (DCX) (1:200; goatpolyclonal; Santa Cruz Biotechnology) and glial fibrillary acidicprotein (GFAP) (1:300; rabbit polyclonal; Dako) were used to identifyneural precursor cells, neurons, newly born neurons and astrocytes,respectively. In the proliferation study, BrdU-treated cells wereincubated with mouse monoclonal anti-BrdU antibody (undiluted;GE Healthcare). Following three 5 min washes, the cells wereincubated for 1 h at RT with the appropriate secondary antibodies:Alexa Fluor 488 donkey anti-goat IgG, Alexa Fluor 488 donkey anti-rabbit IgG and Alexa Fluor 594 donkey anti-mouse IgG (all 1:2000;Molecular Probes). Cells were counterstained with Hoechst 33258(1:2500; Sigma) for 5 min at RT to identify the nuclei. The cells werewashed with PBS and the coverslips were mounted onto microscopeslides using Dakocytomation fluorescent mounting medium (Dako).

Quantification of cellular phenotypes and measurement of neuronalmorphology

Fluorescent images of immunopositive cells were viewed under anOlympus Provis AX70 upright microscope. For each treatmentcondition, photomicrographs were taken at ×40 magnification usingan Olympus DP50 digital camera and Studiolite™ software. Thenumber of total cells, the numbers of each cell type and the numbers ofBrdU+ cells were counted by an observer who was blind to theexperimental treatments in 5 randomly chosen fields of view percoverslip. For each condition, cells from 4 coverslips were stained andcounted, and each experiment was repeated 3 times. The cell somalarea and number of primary, secondary and tertiary neurites of DCX-positive neurons were calculated by a blinded observer using AnalysisD software and established stereological methods (Mayhew, 1991).The formula used to calculate cell somal area was somal area=n×Bwhere n is the number of times the neurites intersect the grid linesand B is the area associated with each point (taking the magnificationinto account). The extent of neurite branching was determined bycounting the numbers of processes per cell, where a primary process isconsidered to be a branch from the cell body, secondary processes areconsidered branches from primary processes and tertiary processesare considered branches from secondary processes.

Statistical analyses

ANOVA with a post hoc Dunnett's test or an unpaired Student's t-test was performed as appropriate, to determine which conditionswere significantly different from each other. Results were expressedas means with SEM and deemed significant when pb 0.05.

Acknowledgments

We thank Dr. Kieran McDermott, University College Cork forproviding us with Hes1 primers.

References

Aguilera, C., Hoya-Arias, R., Haegeman, G., Espinosa, L., Bigas, A., 2004. Recruitment ofIkappaBalpha to the hes1 promoter is associated with transcriptional repression.Proc. Natl. Acad. Sci. USA 101, 16537–16542.

Ben-Hur, T., Ben-Menachem, O., Furer, V., Einstein, O., Mizrachi-Kol, R., Grigoriadis, N.,2003. Effects of proinflammatory cytokines on the growth, fate, and motility ofmultipotential neural precursor cells. Mol. Cell. Neurosci. 24, 623–631.

Bernardino, L., Agasse, F., Silva, B., Ferreira, R., Grade, S., Malva, J.O., 2008. Tumornecrosis factor-alpha modulates survival, proliferation, and neuronal differentia-tion in neonatal subventricular zone cell cultures. Stem cells 26, 2361–2371.

Bush, T.G., Puvanachandra, N., Horner, C.H., Polito, A., Ostenfeld, T., Svendsen, C.N.,Mucke, L., Johnson, M.H., Sofroniew, M.V., 1999. Leukocyte infiltration, neuronaldegeneration, and neurite outgrowth after ablation of scar-forming, reactiveastrocytes in adult transgenic mice. Neuron 23, 297–308.

135A. Keohane et al. / Molecular and Cellular Neuroscience 43 (2010) 127–135

Cacci, E., Claasen, J.H., Kokaia, Z., 2005. Microglia-derived tumor necrosis factor-alphaexaggerates death of newborn hippocampal progenitor cells in vitro. J. Neurosci.Res. 80, 789–797.

Caldwell, M.A., He, X., Svendsen, C.N., 2005. 5-Bromo-2′-deoxyuridine is selectivelytoxic to neuronal precursors in vitro. Eur. J. Neurosci. 22, 2965–2970.

Castella, P., Wagner, J.A., Caudy, M., 1999. Regulation of hippocampal neuronaldifferentiation by the basic helix–loop–helix transcription factors HES-1 andMASH-1. J. Neurosci. Res. 56, 229–240.

Chojnacki, A., Shimazaki, T., Gregg, C., Weinmaster, G., Weiss, S., 2003. Glycoprotein 130signaling regulates Notch1 expression and activation in the self-renewal ofmammalian forebrain neural stem cells. J. Neurosci. 23, 1730–1741.

Feder, J.N., Jan, L.Y., Jan, Y.N., 1993. A rat gene with sequence homology to the Droso-phila gene hairy is rapidly induced by growth factors known to influence neuronaldifferentiation. Mol. Cell. Biol. 13, 105–113.

Gaur, U., Aggarwal, B.B., 2003. Regulation of proliferation, survival and apoptosis bymembers of the TNF superfamily. Biochem. Pharmacol. 66, 1403–1408.

Heldmann, U., Thored, P., Claasen, J.H., Arvidsson, A., Kokaia, Z., Lindvall, O., 2005. TNF-alpha antibody infusion impairs survival of stroke-generated neuroblasts in adultrat brain. Exp. Neurol. 196, 204–208.

Iosif, R.E., Ekdahl, C.T., Ahlenius, H., Pronk, C.J., Bonde, S., Kokaia, Z., Jacobsen, S.E., Lindvall,O., 2006. Tumor necrosis factor receptor 1 is a negative regulator of progenitorproliferation in adult hippocampal neurogenesis. J. Neurosci. 26, 9703–9712.

Kaltschmidt, C., Kaltschmidt, B., Baeuerle, P.A., 1993. Brain synapses contain inducibleforms of the transcription factor NF-kappa B. Mech. Dev. 43, 135–147.

Kelly, A., Vereker, E., Nolan, Y., Brady, M., Barry, C., Loscher, C.E., Mills, K.H., Lynch, M.A.,2003. Activation of p38 plays a pivotal role in the inhibitory effect oflipopolysaccharide and interleukin-1 beta on long term potentiation in rat dentategyrus. J. Biol. Chem. 278, 19453–19462.

Knoblach, S.M., Fan, L., Faden, A.I., 1999. Early neuronal expression of tumor necrosisfactor-alpha after experimental brain injury contributes to neurological impair-ment. J. Neuroimmunol. 95, 115–125.

Krum, J.M., Rosenstein, J.M., 1999. Transient coexpression of nestin, GFAP, and vascularendothelial growth factor in mature reactive astroglia following neural grafting orbrain wounds. Exp. Neurol. 160, 348–360.

Lieberman, A.P., Pitha, P.M., Shin, H.S., Shin, M.L., 1989. Production of tumor necrosisfactor and other cytokines by astrocytes stimulated with lipopolysaccharide or aneurotropic virus. Proc. Natl. Acad. Sci. USA 86, 6348–6352.

Lin, R.C., Matesic, D.F., Marvin, M., McKay, R.D., Brustle, O., 1995. Re-expression of theintermediate filament nestin in reactive astrocytes. Neurobiol. Dis. 2, 79–85.

Liu, T., Clark, R.K., McDonnell, P.C., Young, P.R., White, R.F., Barone, F.C., Feuerstein, G.Z.,1994. Tumor necrosis factor-alpha expression in ischemic neurons. Stroke 25,1481–1488.

Liu, Y.P., Lin, H.I., Tzeng, S.F., 2005. Tumor necrosis factor-alpha and interleukin-18modulate neuronal cell fate in embryonic neural progenitor culture. Brain res.1054, 152–158.

MacEwan, D.J., 2002. TNF ligands and receptors-a matter of life and death. Br. J.Pharmacol. 135, 855–875.

Mayhew, T.M., 1991. The new stereological methods for interpreting functionalmorphology from slices of cells and organs. Exp. Physiol. 76, 639–665.

McGeer, P.L., McGeer, E.G., 2004. Inflammation and the degenerative diseases of aging.Ann. N. Y. Acad. Sci. 1035, 104–116.

Menet, V., Gimenez, Y.R.M., Sandillon, F., Privat, A., 2000. GFAP null astrocytes are afavorable substrate for neuronal survival and neurite growth. Glia 31, 267–272.

Monje, M.L., Toda, H., Palmer, T.D., 2003. Inflammatory blockade restores adulthippocampal neurogenesis. Science 302, 1760–1765.

Nakamura, Y., Sakakibara, S., Miyata, T., Ogawa, M., Shimazaki, T., Weiss, S., Kageyama,R., Okano, H., 2000. The bHLH gene hes1 as a repressor of the neuronal commitmentof CNS stem cells. J. Neurosci. 20, 283–293.

Neumann, H., Schweigreiter, R., Yamashita, T., Rosenkranz, K., Wekerle, H., Barde, Y.A.,2002. Tumor necrosis factor inhibits neurite outgrowth and branching ofhippocampal neurons by a rho-dependent mechanism. J. Neurosci. 22, 854–862.

Renno, T., Krakowski, M., Piccirillo, C., Lin, J.Y., Owens, T., 1995. TNF-alpha expression byresident microglia and infiltrating leukocytes in the central nervous system of micewith experimental allergic encephalomyelitis. Regulation by Th1 cytokines. J.Immunol. 154, 944–953.

Salama-Cohen, P., Arevalo, M.A., Meier, J., Grantyn, R., Rodriguez-Tebar, A., 2005.NGF controls dendrite development in hippocampal neurons by binding top75NTR and modulating the cellular targets of Notch. Mol. Biol. Cell 16,339–347.

Salama-Cohen, P., Arevalo, M.A., Grantyn, R., Rodriguez-Tebar, A., 2006. Notch andNGF/p75NTR control dendrite morphology and the balance of excitatory/inhibitory synaptic input to hippocampal neurones through Neurogenin 3. J.Neurochem. 97, 1269–1278.

Sheng, W.S., Hu, S., Ni, H.T., Rowen, T.N., Lokensgard, J.R., Peterson, P.K., 2005. TNF-alpha-induced chemokine production and apoptosis in human neural precursorcells. J. Leuko. Biol. 78, 1233–1241.

Tanigaki, K., Nogaki, F., Takahashi, J., Tashiro, K., Kurooka, H., Honjo, T., 2001. Notch1and Notch3 instructively restrict bFGF-responsive multipotent neural progenitorcells to an astroglial fate. Neuron 29, 45–55.

Taupin, P., 2005. Adult neurogenesis in the mammalian central nervous system:functionality and potential clinical interest. Med. Sci. Monit. 11, RA247–RA252.

Teng, F.Y., Hor, C.H., Tang, B.L., 2009. Emerging cues mediating astroglia lineagerestriction of progenitor cells in the injured/diseased adult CNS. Differentiation 77,121–127.

Widera, D., Mikenberg, I., Elvers, M., Kaltschmidt, C., Kaltschmidt, B., 2006. Tumornecrosis factor alpha triggers proliferation of adult neural stem cells via IKK/NF-kappaB signaling. BMC Neurosci. 7, 64.

Wong, G., Goldshmit, Y., Turnley, A.M., 2004. Interferon-gamma but not TNF alphapromotes neuronal differentiation and neurite outgrowth of murine adult neuralstem cells. Exp. Neurol. 187, 171–177.

Yang, L., Lindholm, K., Konishi, Y., Li, R., Shen, Y., 2002. Target depletion of distincttumor necrosis factor receptor subtypes reveals hippocampal neuron death andsurvival through different signal transduction pathways. J. Neurosci. 22,3025–3032.

Zhao, X., Bausano, B., Pike, B.R., Newcomb-Fernandez, J.K., Wang, K.K., Shohami, E.,Ringger, N.C., DeFord, S.M., Anderson, D.K., Hayes, R.L., 2001. TNF-alpha stimulatescaspase-3 activation and apoptotic cell death in primary septo-hippocampalcultures. J. Neurosci. Res. 64, 121–131.