pc3 overexpression affects the pattern of cell division of rat cortical

PC3 overexpression affects the pattern of cell division of ratcortical precursors

Paolo Malatestaa, Magdalena GoÈtzb, Giuseppina Barsacchia, Jack Pricec, Roberto Zoncua,Federico Cremisia,d,*

aDipartimento di Fisiologia e Biochimica, Sezione di Biologia Cellulare e dello Sviluppo, UniversitaÁ di Pisa, via Carducci 13, 56010 Ghezzano (Pisa), ItalybMax-Plank -Institute fuÈr Neurobiologie, Am Klopferspitz 18a, 82152 Planegg-Martinsried, MuÈnchen, Germany

cInstitute of Psychiatry, Denmark Hill, London 5E5 8AF, UKdScuola Normale Superiore di Pisa, piazza dei Cavalieri 7, 56100 Pisa, Italy

Received 2 August 1999; received in revised form 31 August 1999; accepted 1 September 1999

Abstract

The PC3 gene is transiently expressed during neurogenesis in precursor cells of the telencephalic ventricular/subventricular zone, and is

rapidly downregulated before cell migration and differentiation. It is thought to have a role in controlling cell proliferation, but its precise

function is not known. Here we present evidence that PC3, when overexpressed in vitro by retroviral-mediated gene transfer, acts by

interfering with the normal pattern of cell division. Firstly, we report evidence that PC3 overexpression reduces the rate of cell proliferation

in both NIH 3T3 cells and embryonic precursor cells from the rat cerebral cortex. Secondly, when studying the pattern of BrdU dilution in

clones of cortical precursors, we observe that clones transduced with PC3 show an asymmetric pattern of BrdU dilution more frequently than

clones transduced with a control vector. We discuss the hypothesis that the higher number of PC3 transduced clones showing an asymmetric

pattern of BrdU dilution may be due to an increase in asymmetric cell divisions. q 2000 Elsevier Science Ireland Ltd. All rights reserved.

Keywords: Neurogenesis; Asymmetric cell divisions; Cell proliferation; Rat cortical precursors; Retroviral vectors; BrdU labelling

1. Introduction

Cortical development is a complex process that leads to

the formation of a six layered structure from a simple

neuroepithelium. This process requires precursor cells

dividing in the ventricular zone (VZ) that are able to

produce both more dividing precursors and post-mitotic

cells (see McConnell, 1995). One question still to be

answered is how the number of both progenitors and post-

mitotic neurons is regulated. It has long been thought that

most post-mitotic neurons are generated by a series of asym-

metric cell divisions by precursors in the VZ. Such divisions

generate one dividing daughter cell and one that is post-

mitotic (Rakic, 1972; Price and Thurlow, 1988; Reid et

al., 1995). Moreover, it has been proposed that the transition

from a symmetric towards an asymmetric pattern of cell

divisions would both mark the onset of neurogenesis and

specify a pool of neurogenetic precursor cells, which keep

on producing post-mitotic neurons (Chenn and McConnell,

1995). Nonetheless, both the functional relationship

between asymmetric divisions and cortical neurogenesis,

and the nature of the genes controlling the pattern of cell

divisions during cortical development, remain poorly under-

stood. The relationship between cell proliferation control

and mode of cell division is also presently unde®ned.

The PC3/Tis21/BTG2 gene (Bradbury et al., 1991;

Fletcher et al., 1991; Rouault et al., 1992; Rouault et al.,

1996), called PC3 hereafter, is transiently expressed in the

VZ during CNS development and was shown to be a marker

of neuronal cell birthday (Iacopetti et al., 1994). In fact, at

the onset of neurogenesis, its expression identi®es single

neuroepithelial cells that switch from proliferative to

neuron-generating division (Iacopetti et al., 1999). In addi-

tion, PC3 overexpression in cell lines exerts an antiproli-

ferative effect (Montagnoli et al., 1996; Rouault et al.,

1996). Taken together, these observations suggest that

PC3 may regulate cell proliferation during neurogenesis.

Nonetheless, the precise function of the PC3 protein is

presently unknown.

Up to now, a PC3 antiproliferative effect has been demon-

strated only in cell lines (Montagnoli et al., 1996; Rouault et

al., 1996). PC3 is transiently expressed in PC12 cells

Mechanisms of Development 90 (2000) 17±28

0925-4773/00/$ - see front matter q 2000 Elsevier Science Ireland Ltd. All rights reserved.

PII: S0925-4773(99)00224-5

www.elsevier.com/locate/modo

* Corresponding author. Tel.: 139-050-878-356; fax: 139-050-878-

486.

E-mail address: [email protected] (F. Cremisi)

induced to differentiate by NGF treatment, suggesting that

its function may be associated with the exit of cells from the

cell cycle occurring before cell differentiation. This is

consistent with the ®ndings that in vivo, PC3 mRNA is

transcribed transiently during the G1 phase of the cell

cycle, and that PC3 protein is found in post-mitotic cells

during the very ®rst phase of neuronal differentiation (Iaco-

petti et al., 1999). However, cell lines might not faithfully

re¯ect in vivo neurogenesis in regard to mitotic control

mechanisms. In order to study the effects of PC3 gene

expression on the mode of cell division of non cell line-

derived neuroblasts, we applied the retroviral-mediated

gene transfer to overexpress PC3 in primary cultures of

cortical precursors. This study reports evidence that PC3

overexpression decreases the cell proliferation rate of both

NIH 3T3 cells and E15 rat cortical precursors. In addition,

the retroviral-mediated gene transfer approach allowed us to

carry out a clonal analysis of cortical precursors transduced

with either PC3 or control vectors. When analysing the

pattern of BrdU dilution in the progeny of cortical clones

transduced with PC3, we observed a higher number of

clones displaying asymmetric BrdU cell dilution with

respect to control clones. Among the possible mechanisms

accounting for such asymmetric dilution, we propose that

post-mitotic cells are generated together with dividing

daughters in an asymmetric lineage of transduced progeny.

In addition, the PC3 transduced clones display clonal sizes

smaller than those of control clones. Finally, the higher

frequency of clones showing an asymmetric BrdU cell dilu-

tion appears to be related to in vitro conditions supporting

neurogenesis rather than cell proliferation. We discuss the

hypothesis that PC3 gene may be involved in pushing corti-

cal precursors towards a neurogenic, asymmetric mode of

cell division.

2. Results

2.1. In vitro PC3 expression

In vitro PC3 gene expression was investigated by in situ

hybridisation. NIH 3T3 ®broblasts were transduced with a

retroviral vector coding for PC3 protein (pPC3c-i-nZ, Fig.

1) and hybridised with an antisense PC3 probe, as described

in Section 4. The expression of PC3 mRNA driven by the

viral promoter in transduced NIH 3T3 clones was compared

with PC3 basal expression of non-transduced NIH 3T3 cells

(Figs. 2A,B). Clones of transduced cells analysed 5 days

after viral transduction showed a high level of labelling

(Fig. 2B), while a very faint labelling was detected in the

control (Fig. 2A).

PC3 transgenic expression of NIH 3T3 transduced cells

was compared with endogenous PC3 expression of E15

cortical precursors (Figs. 2B,D). Cortical cells were disso-

ciated, cultured for 48 h in serum-supplemented medium

(see Section 4) and hybridised as for NIH 3T3 cells. The

level of PC3 mRNA expression is not homogeneous in the

cortical cell population: some cells are strongly labelled

P. Malatesta et al. / Mechanisms of Development 90 (2000) 17±2818

Fig. 1. Structures of the vectors used in this study. All the replication-

incompetent retroviral vectors used to produce recombinant retroviral parti-

cles derive from 1704 plasmid. Long Terminal Repeat (LTR) and Csequence are from Rous Sarcoma Virus. All vectors lead lacZ translation

by means of the Internal Ribosome Entry Site (IRES) of the Encephalo-

myocarditis Virus (EMC). PC3c: 565 bp long PC3 coding sequence,

¯anked upstream by 65 bp of 5 0 untranslated leader sequence. Nts, nuclear

translocation sequence.

Fig. 2. PC3 in situ hybridisation. A and B show NIH 3T3 cells, C and D

show E15 rat cortical precursors. Cells were hybridised with antisense

(A,B,D) and sense (C) digoxigenin-labelled PC3 probes. In B, a clone of

PC3-transduced cells shows a high level of labelling compared to non

transduced cells (A). Endogenous PC3 expression of E15 rat cortical

cells is shown in D. Arrowheads in D point to cells whose labelling is

comparable with the labelling of control cells hybridised with sense

probe (C). Arrow in D points to the most labelled cell, in which PC3

expression is comparable with transgenic expression of NIH 3T3 trans-

duced cells (B). Scale bars 50 mm.

(Fig. 2D, arrow), some others show a level of labelling

comparable to that of cells hybridised with sense probe

(Fig. 2D, arrowheads, C), and many of them show an inter-

mediate level of labelling. The expression driven by retro-

viral vector in NIH 3T3 cells (Fig. 2B) is comparable to

endogenous expression of the most labelled cortical cells

(arrow in Fig. 2D).

2.2. PC3 constitutive expression driven by retroviral vectors

decreases the cell proliferation rate of both NIH 3T3 cells

and cortical precursors

In order to assay the ability of PC3 overexpression to

affect the cell proliferation rate of cortical precursor cells,

we established a proliferation assay using retroviral-

mediated gene transfer. The assay compared the cell prolif-

eration rates of two cell populations transduced with differ-

ent replication-incompetent retroviral vectors, one encoding

PC3 and lacZ, and the second (the control) encoding lacZ

only. In order to distinguish the two types of lacZ-labelled

cells in the same culture dish, one construct carried the nts

(nuclear translation signal)-lacZ and the other the cytoplas-

mic lacZ (Figs. 1,4).

We ®rst applied this assay to NIH 3T3 cells, a cell line

where an antiproliferative effect of PC3 had already been

reported (Montagnoli et al., 1996; Rouault et al., 1996).

Cells were transduced with vectors 1703, 1726 or PC3c-

i-nZ. After 2 days, co-cultures were established with either

1726 1 1703 (control) or 1726 1 pPC3c-i-nZ (Section 4).

The relative growth of the two labelled populations in each

culture was analysed over the following ®ve passages by

calculating the relative size of each population (Fig. 3A

and Table 1; see Section 4). In the control experiment,

there was no signi®cant change in the value of Dpro1703/

1726 � {(1703 transduced cells 2 1726 transduced cells)/all

transduced cells}. In the experimental condition, however,

there was a signi®cant decrease of DproPC3/1726 � {(pPC3c-

i-nZ transduced cells 2 1726 transduced cells)/all trans-

duced cells} over ®ve passages, indicating that the popula-

tion expressing PC3 underwent a smaller expansion in

comparison to the control population. Nonetheless, PC3-

expressing cells had expanded and were still present in the

cultures after ®ve passages in vitro. These data show that

PC3 overexpression exerts an antiproliferative effect on

NIH 3T3 cells without actually blocking their division

per se.

We applied the same analysis to E15 rat cortical precur-

sor cells grown in 10% FCS (Table 1). As with the NIH

3T3 cells, the control populations showed no relative

difference in expansion up to six passages in culture (Fig.

3B). In the experimental condition, however, the reduction

in DproPC3/1726 was dramatic after just two passages. As

with the NIH 3T3 cells, cortical precursor cells overexpres-

sing PC3 failed to expand to the degree shown by the

control population.

We reasoned that the observed PC3 antiproliferative

effect could have a number of plausible explanations (see

Section 3). PC3 expression could slow down the rate of cell

divisions, effectively increasing the cell cycle length; or it

could alter the mode of division by pushing symmetrically

dividing precursor cells into an asymmetric mode of divi-

sion; or both.

2.3. In vitro PC3 overexpression affects the pattern of cell

division of rat cortical precursors

We developed a BrdU dilution assay in order to deter-

mine the degree of asymmetric division in cortical cultures.

Since NIH 3T3 cells can be assumed to divide symmetri-

cally, we initially used these cells to validate the assay. NIH

3T3 cells were transduced with control vector (1703) and

labelled with BrdU for 36 h, by which time all the cells were

P. Malatesta et al. / Mechanisms of Development 90 (2000) 17±28 19

Fig. 3. Proliferation assay on NIH 3T3 cells (A) and E15 rat cortical cells

(B). The graphics show the Dpro variation over the time (A) and passages in

culture (B). The decrease of DproPC3/1726 means a general antiproliferative

effect of PC3 overexpression. Notation: DproPC3/1726 � {(pPC3c-i-nZ trans-

duced cells 2 1726 transduced cells/all transduced cells}; Dpro1703/1726 �{(1703 transduced cells 2 1726 transduced cells)/all transduced cells}.

Data from two different experiments were plotted on the graphic in B. At

early passages, reporter activity of transduced cortical cells was better

detected by immuno¯uorescence. For this reason, the ®rst set of data

(passage 1±3) was obtained by immuno¯uorescence analysis, while the

second set (passage 3±6) comes from X-gal staining. Details in the text

(see also Table 1).

BrdU labelled. The cells were then re-plated in fresh

medium without BrdU and cultured until con¯uent. We

stained the cells with anti-BrdU and quanti®ed the staining

in each clone. For each clone, we calculated the ratio of

labelling in the least labelled cell in a clone to that of the

most labelled cell in that same clone (%MIN/MAX). For

NIH 3T3 cells, the mean %MIN/MAX value is 66% (Table

2, Figs. 5C,F,I). If two cells in a clone were an entire cell

cycle apart, the clone would have a %MIN/MAX value of

50%, so the NIH 3T3 cells of a clone have on average less

than one cell cycle between them. This ®ts with the expec-

tation for symmetrically dividing cells.

The BrdU dilution assay was then applied to E15 cortical

precursor cells. Cultures were transduced in vitro with either

1703 (control) or PC3c-I-nz, labelled with BrdU for 36 h,

then re-plated in new wells without BrdU. Cells were

cultured with serum (as in the proliferation assay) for 7

days. Under these conditions, cortical precursor cells prolif-

erated and most of them maintained nestin expression (see

Section 2.4). Clones were analysed as for the NIH 3T3 cells,

and classi®ed as either À' (Asymmetric) clones having

%MIN/MAX , 50%, or as `S' (Symmetric) clones having

%MIN/MAX $ 50% (Fig. 5, Table 2). The proportion of

À' clones was dramatically increased in PC3 clones

compared with control (70%, n � 116 vs. 24%, n � 125;

Fig. 8A). Moreover, PC3 clones were smaller than control

clones (7.65 cells per clone, n � 156 vs. 9.55, n � 163).

2.4. Different cell culture conditions affect both cell

differentiation and pattern of cell division of E15 cortical

precursors

In all the experiments described so far, the cortical

precursor cells were cultured in medium with serum.

Serum-free medium has previously been reported to encou-

rage neurogenesis, and thereby (it might be supposed) asym-

metric division (Williams et al., 1991; Williams and Price,

1995). We wanted therefore to compare the effect of PC3 in

the presence and absence of serum. Firstly, we wanted to

con®rm the previous ®ndings that neurogenesis was

increased in the absence of serum. We grew primary cortical

cultures either in serum-free or serum-containing medium,

then stained them with antibodies that recognise either

undifferentiated precursor cells (anti-nestin), neurons

(anti-b -tubulin), or astrocytes (GFAP). In the presence of

serum, cultures had a considerably higher proportion of

undifferentiated cells (67 vs. 21%), fewer neurons (17 vs.

40%), and roughly the same proportion of astrocytes (Figs. 6

and 8B), conforming to previously reported data (Williams

et al., 1991; Williams and Price, 1995).

We next considered the effect of serum on the incidence

of `S' clones versus À' clones. In the absence of serum,

clones labelled with the control vector have a dramatically

increased incidence of À' clones (62 vs. 24% in the

presence of serum: Fig. 8A). This correlates with the

increased neurogenesis in these cultures. In PC3-transduced

clones cultured without serum, there is an increase (albeit

not signi®cant) of À' clones (78 vs. 70% in the presence of

serum: Fig. 8A). Thus, in the absence of serum, a greater

proportion of cortical clones are À' clones, and PC3 makes

only a modest impact on this overwhelmingly asymmetric

pro®le.

The increased frequency of À' clones induced by PC3

overexpression could be expected to parallel an increase of

post-mitotic cells that differentiate as neurons in vitro. We

directly assayed this point by counting the number of PC3

and control transduced cells showing b-tubulin immuno-

reactivity after a week in culture (Fig. 7). The percentage

of PC3 overexpressing neurons is higher than the percentage

of control transduced neurons cultured without serum (54

vs. 44%, Fig. 8C). Conversely, the percentage of PC3 over-

expressing neurons did not signi®cantly change with respect

to the control in FCS containing cultures (not shown).

3. Discussion

3.1. PC3 overexpression decreases the cell proliferation

rate and affects the pattern of cell division of rat cortical

precursors

In this study, we sought to understand the function of PC3

in the process of neurogenesis by transducing PC3 into rat

cortical neural precursor cells in vitro. We used retroviral

vectors expressing PC3 to see how the gene in¯uenced


Fig. 4. NIH 3T3 ®broblasts (A and B) and E15 rat cortical cells (C and D),

transduced with IRES containing vectors. A and B show NIH 3T3 cells

transduced either with 1726 or PC3c-i-nZ vectors, respectively, at the third

passage of the proliferation assay described in Section 2. X-gal staining

allows to distinguish between 1726 transduced cells, in which the reporter

activity stains the cytoplasm (A), and PC3c-i-nZ transduced cells, with a

nuclear localisation of the reporter activity (B). C and D show cortical cells

transduced with 1726 and PC3c-i-nZ vectors, respectively, at the second

passage in vitro. At early passages, it was easier to detect the difference

between cytoplasmic (C) and nuclear (D) reporter activity of cortical cells

by immunostaining. Cells were immunostained with a polyclonal antibody

to b -galactosidase (see Section 4). Scale bars 25 mm.

proliferation of these embryonic precursor cells. We also

used the retroviral vectors in conjunction with BrdU label-

ling to assay the degree of symmetry of division of precursor

cells overexpressing PC3. We discovered that in compari-

son with control clones, clones that carry PC3 proliferate

less, are smaller with regard to the number of cells per clone,

and are more asymmetric in that the BrdU is less equally

diluted by different members of a clone. We knew from

previous studies that in ®broblasts, PC3 was anti-prolifera-

tive, but those data gave us very little idea of the precise

biological role that PC3 might have during neurogenesis.

Our studies now suggest that in neural precursor cells the

apparent anti-proliferative effect is actually an increased

tendency towards asymmetric divisions that acts to drive

the production of post-mitotic neurons.

Before accepting this conclusion, it is important to

consider alternative explanations of the data. The anti-

proliferative effect and the smaller clone size could concei-

vably be explained by cell death. If PC3 overexpression

caused a greater proportion of the cells to die compared


Table 1

Analysis of the proliferation assays on NIH 3T3 and E15 cortical cellsa

Passage PC3/1726 cultures 1703/1726 cultures

#PC3 #1726 Dpro SD #1703 #1726 Dpro SD

NIH 3T3 cells

1 5813 4174 0.164 0.068 10420 7404 0.169 0.068

2 4163 4491 20.038 0.070 12427 8511 0.187 0.068

3 1939 4770 20.422 0.058 6945 4215 0.245 0.066

4 1194 3187 20.455 0.056 3350 3382 20.005 0.070

5 452 2583 20.702 0.035 4751 1696 0.474 0.054

Cortical cells

1 597 317 0.31 0.065 357 553 0.22 0.067

2 89 64 0.16 0.068 184 279 0.21 0.067

3 246 243 0.006 0.07 265 414 0.22 0.067

3 1423 973 0.19 0.048 310 521 0.25 0.047

4 711 706 0.004 0.05 992 1542 0.22 0.048

5 580 602 20.019 0.05 476 750 0.22 0.048

6 340 350 20.014 0.05 209 340 0.24 0.047

a The cell proliferation assay was carried out on cell cultures containing a mixed population of transduced cells. #PC3, #1726 and #1703 are the numbers of

cells transduced with the retroviral vectors PC3c-i-nZ, 1726 and 1703, respectively, that were counted at each passage in culture. In mixed cultures containing

cells transduced either with PC3 vector or 1726 (control) vector (PC3/1726 cultures), the number of PC3-transduced cells decreased with respect to the number

of 1726-transduced cells. Since cell counting was not exhaustive at each passage, the total number of cells counted varies between passages. Nonetheless, cell

counting was representative of the actual ratio between the two types of transduced cells present in the whole culture at each passage, thus it did not in¯uence

the ®nal analysis. The decrease of the relative number of PC3-transduced cells in culture is better expressed by the decrease of the value PC3/1726Dpro �(pPC3c-i-nZ transduced cells 2 1726 transduced cells)/all transduced cells. In the same experimental conditions, no dramatic change of the relative cell

proliferation rate is detectable in control cultures (1703/1726 cultures). Values refer to graphics in Fig. 4. The values of passages 1±3 and 3±6 of the

proliferation assay on cortical cells are from two different experiments. SD, standard deviation. Other explanations in Section 2 and Section 4.

Table 2

BrdU dilution analysisa

NIH 3T3 clones

# aa3 aa33 ab31 aa6 ab4 an2 an8 ab27 aa25 an27 aa20 an5 ab7 an11 an14 aa32 an23 an20 an30 an36

% 88 99 50 99 60 60 78 66 79 90 63 41 83 22 24 77 11 76 54 93

Control cortical clones

# zo7 zo10 zo13 zo16 zo 19 zo22 zo25 zo28 zo31 zo34 zl9 zl12 zl18 zl21 zl24 zl27 zl30 zl33 zl36 zm4

% 72 66 72 61 43 66 70 80 51 56 10 55 22 52 52 21 33 17 40 16

PC3 cortical clones

# zi3 zi6 zi9 zi12 zi16 zi19 zi22 zi25 zm33 zm36 r4 r7 r10 r13 r16 r19 r22 r25 r28 r31

% 36 34 70 31 50 12 25 11 28 42 14 22 52 40 20 35 42 44 18 35

a The table shows BrdU dilution analysis of 60 clones transduced with 1703 control vector (NIH 3T3 and Control cortical clones) or PC3 carrying vector

(PC3 cortical clones). These clones are representative of larger samples of clones analysed (see Fig. 8 and text). All NIH 3T3 clones and Cortical clones from

zo7 to zo34 and from zi3 to zm33 were cultured in vitro in the presence of FCS (see text), while the others were cultured in serum-free medium. #, the name of

the clone, % represents the ratio between the highest and the lowest BrdU cell content in the clone, expressed as % of the highest value.

with control, this would indeed lead to smaller clones.

Nonetheless, we tend to exclude this possibility for the

following reasons. Firstly, we failed to detect any increase

of picnotic nuclei in PC3 transduced cells (not shown).

Secondly, a decrease of cell proliferation induced by PC3

in cell lines was reported by using different experimental

approaches, in which PC3 overexpression does not induce

cell death (Montagnoli et al., 1996). Thirdly, an increase of

cell death due to PC3 overexpression would not be the

correct explanation because it fails to explain the BrdU

dilution effect. We can see no way whereby increased

death among members of a clone could cause the surviving

members to have diluted their BrdU label to different

extents. This increased differential labelling must mean

that members of a clone divided to differing extents.

The analysis of BrdU dilution suggests that PC3 over-

expression does not simply act to slow down the cell

cycle of all the clonal progeny to the same extent. The

BrdU dilution assay, as it has been applied in this study,

highlights patterns of BrdU inheritance in a clonal progeny

which do not conform to a symmetric lineage. In principle, a

number of mechanisms could explain how asymmetric

BrdU dilution is generated. One possible mechanism

could be an increase in asynchrony of division. Two daugh-

ter cells from one division might both still be mitotic, but

one might divide earlier than the other. Consequently at the

point of analysis, one has divided while the other has not yet

divided. However, in vitro time-lapse analysis of cortical

lineages did not highlight any selective lengthening of the

cell cycle of sublineages in a lineage (Qian et al., 1998). A

further possible mechanism could be the exit from the cell

cycle of sibling cells in sublineages of a clone. Such circum-

stance was shown to exist in in vitro cortical lineages; in

fact, both sibling cells that stop dividing and asymmetric

cell divisions are often observed in neurogenetic lineages

(Qian et al., 1998). An additional mechanism could be the

increase of asymmetric cell divisions, here de®ned as divi-

sions generating a daughter cell that stop dividing and a


Fig. 5. BrdU analysis of cell divisions of E15 cortical cells (A,D,G; B,E,H) and NIH 3T3 ®broblasts (C,F and I). Photographs show parts of larger clones

generated by single cells transduced with 1703 control vector (A,D,G; C,F,I) and pPC3c-i-nZ vector (B,E and H), respectively. Nuclei were stained with

Hoechst No. 33258 in A, B and C. The nuclear reporter activity driven by both 1703 and pPC3c-i-nZ vectors was immunodetected with a polyclonal antibody

to b-galactosidase in D,E and F. G,H and I show the BrdU content as evaluated by immunostaining with a monoclonal antibody. Cells were transduced with

replication-incompetent retroviral vectors, labelled with BrdU and then maintained in FCS containing medium without BrdU for 7 days, to allow the analysis of

BrdU dilution in the cell progeny of a clone (see Section 2). NIH 3T3 cells belonging to the same clone (identi®ed by red nuclei in F) always showed very low

amount of BrdU (barely detectable in I). The majority of cortical clones (76%) transduced with 1703 control vector and cultured with FCS showed the same

pattern of BrdU dilution as NIH 3T3 cells. BrdU staining in G is easily detectable only in a large nucleus (asterisk) which does not belong to the clone (compare

with red nuclei in D). Conversely, most of the pPC3c-i-nZ transduced clones (70%) cultured in the same conditions showed very high differences of BrdU

content in their cell progeny. Arrow and arrowhead in E and H point to the most (MAX) and to the least (MIN) BrdU labelled cells (see BrdU staining in H),

respectively, of a pPC3c-i-nZ cortical clone. The percentage of BrdU staining of the least labelled cell with respect to the most labelled cell in the clone

(%MIN/MAX) was evaluated in order to distinguish between `S' and `A' clones (see Section 2). Scale bars 50 mm.

daughter cell which goes on cycling. Both these mechan-

isms contribute to generate neurogenetic asymmetric

lineages (Qian et al., 1998) and could account for the

observed asymmetric BrdU dilution.

Although the BrdU dilution assay does not permit to

reconstruct the lineage of a cell progeny, based on all

these observations we suggest that it allows to analyse the

trend of cortical precursors to generate asymmetric lineages

similar to those previously described by applying different

techniques (Qian et al., 1998). On this view, since PC3

increases asymmetric BrdU dilution, we can reasonably

assume that PC3 acts to generate asymmetric cell lineages

of cortical precursors.

If PC3 were part of the mechanism that controls the tran-

sition from symmetrically dividing, proliferating precursors

to asymmetric dividing, neurogenetic neuroblasts, so the

open question would be what the mechanisms might be by

which PC3 may act. Until recently, we had little information

on how asymmetric division might be controlled in cortical

precursor cells, but now mammalian homologues of the

Drosophila genes, Notch and Numb, have been shown to

be expressed asymmetrically in these cells (Chenn and

McConnell, 1995; Zhong et al., 1996, 1997). Hypotheses

of how PC3 might interact with these gene functions are

not clear at the present time.


Fig. 6. Effect of different cell culture conditions on cell differentiation of E15 cortical precursors. Primary cells were maintained 2 days in FCS containing

medium and then re-plated and cultured either with (A±F) or without (G±L) FCS. A,B,C,G,H and I show nuclear staining with Hoecsht No. 33258. Cells were

immunostained with antibodies to nestin (D and J), class III b -tubulin (E and K) and GFAP (F and L) as markers for neural precursors, neurons and glia,

respectively. FCS exerted a dramatic inhibition of cell differentiation (compare E and F to K and L) and supported a high number of nestin positive,

undifferentiated cells (D). Scale bars 50 mm.

Fig. 7. PC3 transduced cells differentiated as neurons after a week in serum

free culture. Primary E15 cortical cells were transduced with pPC3c-i-cz

vector and immunostained either for lacZ reporter activity (A), or for the

class III b -tubulin neuronal antigen (B). Arrowheads point to three trans-

duced neurons. Scale bar 50 mm.

3.2. Different culture conditions affect both differentiation

and pattern of cell division of cortical cells

Several soluble factors are known to affect the prolifera-

tion of telencephalic precursor cells in vitro, among them

bFGF (Ghosh and Geenberg, 1995; Kilpatrick and Bartlett,

1995; Temple and Qian, 1995), and EGF (Craig et al.,

1996). Serum has a similar effect (Kilpatrick and Bartlett,

1993, 1995), presumably via lysophosphatidic acid (LPA),

the major serum mitogen. Though cortical precursor cells

have LPA receptors (Hecht et al., 1996), their role in corti-

cogenesis is not yet clear. One effect of serum is to reduce

the level of neurogenesis in comparison with cultures grown

in serum-free media (Williams et al., 1991; Williams and

Price, 1995), and the data presented here con®rm that serum

increases the proportion of nestin-positive precursor cells,

and reduces the proportion of differentiated neurons. Our

results also suggest that the effect of serum-withdrawal

mimics the action of PC3 on the pattern of cell division,

namely on the proportion of Àsymmetric' vs. `Symmetric'

clones. This raises the possibility that serum (LPA) action is

through a mechanism that involves PC3, but this assumption

remains to be investigated. Also, at this point we cannot be

sure whether PC3 acts to drive cells to differentiate speci®-

cally towards a neuronal fate, or is also consistent with a

glial fate. Preliminary data (not shown) suggest that PC3-

transduced precursors generate GFAP-positive glial cells to

the same extent as control transduced precursors do, either

with or without serum, thus implying that PC3 expression is

not suf®cient to drive the neuronal cell fate. Nonetheless, if

PC3 increases the frequency of asymmetric cell divisions,

an increase in post-mitotic cells (presumably neurons)

would also be expected. In serum-free culture, an experi-

mental condition that is permissive to the in vitro differen-

tiation of neurons (see Section 2), the percentage of PC3

overexpressing neurons is 10% higher than the percentage

of control neurons after 1 week in culture (Fig. 8C). Such an

increase is comparable with the increasing frequency of

Àsymmetric' clones of PC3 transduced precursors with

respect to the control (16%) in serum free cultures presented

here. Even if this observation agrees with the proposed

expectation, we are not able to say whether all the PC3

overexpressing neurons derive from post-mitotic daughter

cells of asymmetric cell divisions.

3.3. A model for the role of PC3 during cortical

neurogenesis

Here we suggest that PC3 expression could induce,

directly or indirectly, a pattern of cell division which is

typical of neurogenetic cortical precursors and resembles

the neurogenetic asymmetric lineages observed in vitro

(Qian et al., 1998).

In vivo, PC3 mRNA expression is restricted to VZ cells,

and is turned off as the post-mitotic daughter cells migrate

into the intermediate zone (Iacopetti et al., 1994). PC3

protein is expressed in both a subpopulation of neuroepithe-

lial cells that increases with the progression of neurogenesis,

and in post-mitotic cells during the very ®rst stage of neuro-

nal differentiation (Iacopetti et al., 1999). Indeed, our obser-

vations could clarify the paradox created by previous

studies: if PC3 expression acts to block cell proliferation,

why should PC3 mRNA be expressed preferentially in the

dividing cells of the VZ, and be turned off when the cells

become post-mitotic and migrate? The mRNA expression

pattern would ®t more sensibly with a role in precursor cells


Fig. 8. The frequency of À' (Asymmetric) clones of cortical precursors parallels their cell differentiation in vitro. Histogram in A shows the percentage of À'

clones in different culture conditions. In A, n: total number of clones analysed in three or more sets of experiments for each cell culture condition ( 1 or 2

FCS); Control: 1703 vector; PC3: pPC3c-i-nZ vector. Histogram in B expresses the percentage of nestin, class III b-tubulin and GFAP positive cells as shown

in Fig. 6. In B, n: number of positive cells counted. Histogram in C shows the percentage of cells transduced with 1726 (Control) and pPC3c-i-cz (PC3) vectors,

which were stained by class III b-tubulin antibody after a week in serum free culture.

themselves. A hypothesis is that PC3 acts to enable VZ

precursors to divide asymmetrically in a `stem cell mode'.

On this view, PC3 expression should not contribute to

neural differentiation as such. Both our data, which disagree

with a direct role of PC3 in establishing a speci®c cell fate

(namely neuronal vs. glial), and the ®nding that PC3 protein

persists only transiently into migrating neurons and during

the very initial phase of neuronal differentiation (Iacopetti et

al., 1999; Iacopetti, pers. commun.), would agree with this

assumption.

An issue that arises is whether PC3 plays a role in driving

asymmetric division in true neural stem cells. Stem cells

that are mitotically active and generate the olfactory granule

cells are still present during the ®rst postnatal week in the

subependyma of most anterior regions of the lateral ventri-

cles (Luskin, 1993; Luskin et al., 1997; Morshead et al.,

1998). These cells, which represent the last subpopulation

of neurogenetic dividing neuroblasts of the mammalian

brain, divide asymmetrically to self-renew and give rise to

post-mitotic neurons (Morshead et al., 1998). Notably, the

most anterior part of the lateral ventricles of the rat brain is

the last region of the forebrain to express PC3 at P10 (Iaco-

petti et al., 1994). It would be interesting to test the hypoth-

esis that PC3 expression is precisely con®ned, and

functionally related, to the above mentioned stem cells

population.

4. Experimental procedures

4.1. Plasmids and retroviral vectors

All retroviral vectors carry a bacterial replication origin

and Ampr; two retroviral LTRs; the C sequence necessary

for packaging; the IRES sequence derived from EMC virus

(Ghattas et al., 1991). Recombinant retroviral vectors were

obtained starting from 1703 and 1704 plasmids that drive

the expression of the reporter gene lacZ with nuclear or

cytoplasmic localisation, respectively (see Fig. 1). A PC3

cDNA fragment of 650 bp (PC3c), containing all the PC3

coding sequence ¯anked upstream by 65 bp of 5 0 untrans-

lated leader sequence, was obtained by partially digesting

the full-length clone already described (Montagnoli et al.,

1996) with endonuclease BamHI. The PC3 coding sequence

was inserted in the Bgl II unique restriction site of the 1703

and 1704 plasmids, between the upstream LTR and the

EMC IRES sequence; the two constructs were named

pPC3c-i-nZ and pPC3c-i-cZ, respectively. The IRES

sequence allows ribosomes to entry the di-cistronic messen-

ger transcribed from the upstream LTR and translate the

downstream open reading frame coding for bacterial b -

galactosidase. In the 1726 plasmid, which drives cytoplas-

mic reporter activity and Neo expression, Neo is located

between the upstream LTR and the EMC IRES sequence.

Plasmid 1703, 1704 and 1726 were kindly provided by J.E.

Majors.

4.2. Cell cultures

BOSC 23 (kindly provided by the American Type Culture

Collection), NIH 3T3 cell lines and cerebral primary culture

cells were cultured using DMEM GIBCO 21885, with 100

units/ml penicillin, 100 mg/ml streptomycin and 10% FCS

GIBCO 10106, in humidi®ed atmosphere of 5% CO2.

For primary culture establishment, telencephalic vesicles

from E15 rat embryos were surgically removed in DMEM

supplemented with 20 mM HEPES. After removal, the

tissue was incubated in 0.5 mg/ml trypsin and 0.2 mg/ml

EDTA for 5 min, then dissociated by gently pipetting

through a pulled Pasteur. Cells were harvested by centrifu-

gation at 300 £ g and then plated on 13 mm glass coverslips

previously coated with poly-d-lysine (1 mg/ml in PBS), at

the density of 5 £ 105 cells per well. Cortical cells were

grown in either serum-free medium (DMEM supplemented

with glucose, transferrin, insulin, selenium, progesterone,

thyroxine, triiodothyridine, putrescine, bovine serum albu-

min, 100 units/ml penicillin, 100 mg/ml streptomycin, 2.5

mg/ml Amphotericin B and 0.5% FCS; Bottenstein and

Sato, 1979) or in serum-containing medium (DMEM

supplemented with 100 units/ml penicillin, 100 mg/ml strep-

tomycin, 2.5 mg/ml Amphotericin B, and 10% FCS). When

used, BrdU (10 mM) was added to the medium 12 h after

retroviral infection by a single, long pulse of 36 h, aiming to

label at least one S phase of each dividing cell in culture.

Cultures were fed with fresh medium every 2±3 days.

4.3. Retroviruses harvesting and cell transduction

In order to obtain retroviral supernatants, BOSC 23

helper-free packaging cells were transfected with calcium

phosphate (Pear et al., 1993), using 40 mg of DNA per 100

mm dish. After transfection, culture dishes were rinsed and

5 ml of fresh medium were added. Retroviral supernatants

were harvested 24 h after transfection, ®ltered with 0.45 mm

®lters, concentrated by Centricon ultra®lters (Amicon) and

then frozen in small aliquots (Cepko, 1994). Retroviral titre,

as evaluated according to Cepko (1994), ranged from 1 £106 to 3 £ 106 CFUs in the concentrated supernatants.

Supernatants were tested to exclude the presence of helper

virus.

NIH 3T3 cells were plated 24 h before the transduction at

10% of con¯uence in 60 mm culture dishes (NIH 3T3).

Transduction was performed for 1 h in 500 ml of medium

containing 5±10 ml of concentrated virus and 8 mg/ml of

polybrene (Sigma P4515). Cortical cells were transduced 12

h after plating in 200 ml of medium containing 1±2 ml of

concentrated virus and 2 mg/ml of polybrene. This proce-

dure generated, on average, ten transduced clones per well, a

clone density that is quite far from generating superimposi-

tion of two independent clones (Williams et al., 1991).

In both BrdU and differentiation analyses, cortical cells

were maintained in FCS containing medium for 48 h after


infection and then re-plated in either medium with or with-

out FCS.

4.4. In situ hybridisation

Cells cultured on poli-d-lysine-coated slides were ®xed

with 4% paraformaldehyde/PBS at RT for 20 min, washed

in PBS, dehydrated and stored at 2208C in 70% ethanol

until use. After rehydration, cells were washed three times

in PTw (PBS with 0.1% Tween-20) and then treated with

proteinase K (10 mg/ml) in PTw for 15 min at RT. After

several washes in PBS, cells were acetylated in 0.1 M

triethanolamine with 0.5% acetic anhydride for 10 min

and rinsed in PTw, post-®xed in 4% paraformaldehyde/

PBS for 20 min at RT and washed three times with PTw.

Cells were hybridised with both antisense and sense

(control) PC3 digoxygenin-labelled probes, which were

obtained by in vitro transcription of the PC3 cDNA

sequence described above, after subcloning in bluescript

vector. Hybridisation was performed o/n at 558C in the

presence of 0.5 mg/ml of probe, 50% formamide, 4£ SSC,

1 mg/ml yeast RNA, 0.1% Tween-20. Washes were

performed at 558C in 2£, 1£ and 0.5£ SSC (1 h each).

Cells were then rinsed in PTw and immunostained with

alkaline-phosphatase(AP)-conjugated anti-digoxigenin

antibody (Boehringer), according to the manufacturer's

protocol.

4.5. Histochemistry and immunostaining

For X-gal staining, cells were ®xed directly on culture

dishes with 0.1% glutaraldehyde in PBS buffer for 15

min. Culture dishes were then rinsed three times in PBS at

room temperature for 10 min. Dishes were covered with

staining solution (5 mM K3Fe(CN)6, 5 mM K4Fe(CN)6, 2

mM MgCl2, 0.01% sodium deoxycholate, 0.02% Nonidet P

40, X-gal 1 mg/ml) and incubated for 1±5 h at 378C.

In all immunostaining protocols, cells were ®xed in 4%

paraformaldehyde for 10 min at room temperature, washed

three times for 15 min and treated with 1% Triton in PBS

(pH 7.3) for 5 min. Both primary and secondary antibodies

were incubated in PBS (pH 7.3), 0.1% Triton and 0.5%

BSA. The following primary antibodies were used: polyclo-

nal rabbit antibody to E. coli b -galactosidase (5 0-3 0), 1:30;

monoclonal antibody to BrdU (Boehringer), 1:50; anti

nestin tissue culture supernatant, 1:4; monoclonal anti

GFAP (Boehringer), 1:10; monoclonal anti class III b -tubu-

lin (Sigma), 1:100. Secondary antibodies were used at the

following concentrations: Goat Anti Mouse (GAM)-TRITC

(Boehringer 1238), 1:10; Goat Anti Rabbit (GAR)-FITC

(Sigma F 9006), 1:200.

In double staining with anti b -galactosidase and antibo-

dies to GFAP, class III b -tubulin and nestin, both primary

antibodies were incubated for 3 h at room temperature,

followed by three washes of 30 min in PBS and incubation

in secondary antibodies at RT for 3 h. When immunodetect-

ing BrdU and b -galactosidase, ®rst we incubated the cells

with polyclonal anti b-galactosidase, then we detected the

primary antibody with GAR-FITC and, after three washes of

15 min in PBS at room temperature, we post-®xed the cells

in 4% paraformaldehyde for 20 min. After post-®xation,

three washes in PBS were followed by treatment with 2 N

HCl at RT for 30 min, three more washes in PBS, incubation

in anti BrdU antibody at RT for 3 h and immunodetection

with GAM±TRITC.

Nuclei were labelled with Hoechst No. 33258 (Sigma),

added 1:1000 during detection with the secondary antibody.

After immunostaining, cultures were mounted in glycerol/

PBS solution 8:2 and viewed under an epi¯uorescence

microscope.

4.6. Proliferation assay for NIH 3T3 and primary culture

cells

Cell cultures were transduced with each of the four

vectors PC3c-i-cZ, PC3c-i-nZ, 1703, 1726, separately.

Forty-eight hours after transduction, couples of the trans-

duced cell cultures were mixed, in order to obtain a mixed

population each time. Cell cultures were mixed in two

combinations: 1726 1 1703 (control population) and

1726 1 pPC3c-i-nZ (experimental population). Each popu-

lation contained a number of cells transduced with one of

the two vectors similar to the number of cells transduced

with the other vector. We typically obtained NIH 3T3 cell

populations made of about 1% of transduced cells vs. 99%

of wild type cells. NIH 3T3 populations were maintained in

culture for 354 h (®ve passages) in FCS containing medium.

Replica dishes of each passage were prepared in parallel for

X-gal staining. The cell populations were maintained in

culture in 60 mm Petri dishes up to con¯uence (about

1.5 £ 106 per dish) and split with a 1:10 ratio at each

passage.

Primary precursors from E15 rat cortex were transduced

as adherent cells on poly-d-lysine coated Petri dishes 12 h

after dissection. Forty-eight hours after transduction, a

period long enough to allow PC3 transgene expression,

cells were mixed and studied as previously described for

the NIH 3T3 proliferation assay, with few exceptions. Corti-

cal cells were cultured in 35-mm poly-d-lysine-treated

dishes until con¯uence and then split with 1:2 ratio at

each passage. Con¯uence required, on average, 48 h for

primary cells. Conversely, cells at later passages required

longer time to reach the con¯uence, with a maximum at the

®fth passage (7 days in culture).

Both for NIH 3T3 and primary culture cells, replica

dishes were ®xed at each passage in order to allow detection

of the reporter activity.

The number of cells showing nuclear or cytoplasmic lacZ

expression at each passage was scored (Table 1) and the

values: DproPC3/1726 � {(pPC3c-i-nZ transduced cells 21726 transduced cells)/all transduced cells} and Dpro1703/

1726 � {(1703 transduced cells 2 1726 transduced cells)/

all transduced cells} were plotted on a scattered graphic.


The variation of such values, expressing the proportion of

the two kinds of transduced cells passage after passage,

allowed to study the dynamics of cell replication rate.

Three sets of experiments were carried out for both NIH

3T3 and cortical cells. Results are summarised in Table 1.

4.7. BrdU dilution assay

E15 cortical precursors were transduced in vitro with

replication-incompetent retroviral vectors. We aimed to

analyse the pattern of BrdU dilution in the progeny

produced by a single transduced precursor endowed with

the transgenic product. Twelve hours after infection, precur-

sors were labelled with BrdU for 36 h. Such long time of

BrdU incorporation allowed the labelling of virtually all

cells dividing in culture, some of which incorporated

BrdU in more than one S-phase. Cells were then replated

in new wells with fresh medium without BrdU and cultured

for 7 days, in order to study the pattern of BrdU dilution in

the progeny of single dividing precursors. Single cortical

clones transduced with either control vector (1703) or PC3

expressing vector (pPC3c-i-nZ) were analysed by immunos-

taining. Clustered nuclei immunopositive for lacZ were

considered as belonging to a single clone and photographed

with a Zeiss III photomicroscope. The ®lms were scanned

with a NIKON SCAN apparatus, and the amount of BrdU

immunostaining of the labelled nuclei of each clone was

then analysed with the NIH 1.61 IMAGE software. MAX

was the value of the most BrdU labelled cell, MIN was the

value of the least labelled one.

In control experiments, NIH 3T3 cells were used as

experimental model of symmetrically dividing cells. Two

days after infection with 1703 control vector, NIH 3T3 cells

were split 1:20 and processed as described for the cortical

precursors. Examples of the results obtained by the analysis

of both NIH 3T3 cells and cortical precursors are shown in

Table 2.

Acknowledgements

We wish to thank John E. Majors for the generous gift of

1703, 1704 and 1726 retroviral constructs, and Felice

Tirone for having kindly provided us with the PC3

cDNA used in this work. We thank the Abiogen Pharma

of Pisa for sharing the animal house facilities and Riccardo

Bianchi for technical assistance. This work was supported

by grants from M.U.R.S.T., from Progetto ®nalizzato

C.N.R. `Ingegneria Genetica' and by E.C. (grant no.

B104-CT98-0399). F. Cremisi was supported by an ESFP

short term fellowship.

References

Bottenstein, J.E., Sato, G.H., 1979. Growth of a rat neuroblastoma cell line

in serum free supplemented medium. Proc. Natl. Acad. Sci. USA 76,

514±517.

Bradbury, A., Possenti, R., Shooter, E.M., Tirone, M., 1991. Molecular

cloning of PC3, a putatively secreted protein whose mRNA is induced

by nerve growth factor and depolarization. Proc. Natl. Acad. Sci. USA

88, 3353±3357.

Cepko, C., 1994. Section III: Transduction of Genes Using Retrovirus

Vectors, Short Protocols in Molecular Biology, Greene Publishing

Assoc. and Wiley-Interscience, New York.

Chenn, A., McConnell, S.K., 1995. Cleavage orientation and the asym-

metric inheritance of Notch1 immunoreactivity in mammalian neuro-

genesis. Cell 82, 631±641.

Craig, C., Tropepe, V., Morshead, C., Reynolds, B., Weiss, S., van der

Kooy, D., 1996. In vivo growth factor expansion of endogenous sube-

pendymal neural precursor cell populations in the adult mouse brain. J.

Neurosci. 16, 2649±2658.

Fletcher, B., Lim, R., Varnum, B., Kujubu, D., Koski, R., Herschman, H.,

1991. Structure and expression of TIS21, a primary response gene

induced by growth factors and tumour promoters. J. Biol. Chem. 266,

14511±14518.

Ghattas, I.R., Sanes, J.R., Majors, J.E., 1991. The encephalomyocarditis

virus internal ribosome entry site allows ef®cient co-expression of two

genes from a recombinant provirus in cultured cells and in embryos.

Mol. Cell. Biol. 11, 5848±5859.

Ghosh, A., Geenberg, M.E., 1995. Distinct roles for bFGF and NT-3 in the

regulation of cortical neurogenesis. Neuron 15, 89±103.

Hecht, J.H., Weiner, J.A., Post, S.R., Chunn, J., 1996. Ventricular Zone

Gene (vzg-1) encodes a lysophosphatidic acid receptor expressed in

neurogenic regions of the developing cerebral cortex. J. Cell. Biol.

135 (4), 1071±1083.

Iacopetti, P., Barsacchi, G., Tirone, F., Maffei, L., Cremisi, F., 1994. Devel-

opmental expression of PC3 gene is correlated with neuronal cell birth-

day. Mech. Dev. 47, 127±137.

Iacopetti, P., Michelini, M., Stuckmann, I., Oback, B., Aaku-Saraste, E.,

Huttner, W.B., 1999. Expression of the antiproliferative gene Tis21 at

the onset of neurogenesis identi®es single neuroepithelial cells that

switch from proliferative to neuron-generating division. Proc. Natl.

Acad. Sci. USA 96 (8), 4639±4644.

Kilpatrick, T.J., Bartlett, P.F., 1993. Cloning and growth of multipotential

neural precursors: requirements for proliferation and differentiation.

Neuron 10, 255±265.

Kilpatrick, T.J., Bartlett, P.F., 1995. Cloned multipotential precursors from

the mouse cerebrum require FGF-2, whereas glial restricted precursors

are stimulated with either FGF-2 or EGF. J. Neurosci. 15, 3653±3661.

Luskin, M.B., 1993. Restricted proliferation and migration of post-natally

generated neurons derived from the forebrain subventricular zone.

Neuron 11, 173±189.

Luskin, M.B., Zigova, T., Soteres, B.J., Stewart, R.R., 1997. Neuronal

progenitor cells derived from the anterior subventricular zone of the

neonatal rat forebrain continue to proliferate in vitro and express a

neuronal phenotype. Mol. Cell. Neurosci. 8, 351±366.

McConnell, S.K., 1995. Constructing the cerebral cortex: neurogenesis and

fate determination. Neuron 15, 761±768.

Montagnoli, A., Guardavaccaro, D., Starace, G., Tirone, F., 1996. Over-

expression of the nerve growth factor-inducible PC3 immediate early

gene is associated with growth inhibition. Cell. Growth Diff. 7, 1327±

1336.

Morshead, C., Craig, C., van der Kooy, D., 1998. In vivo clonal analyses

reveal the properties of endogenous neural stem cell proliferation in the

adult mammalian forebrain. Development 125, 2251±2261.

Pear, W.S., Nolan, G.P., Scott, M.L., Baltimore, D., 1993. Production of

high-titre helper-free retroviruses by transient transfection. Proc. Natl.

Acad. Sci. USA 90, 8392±8396.

Price, J., Thurlow, L., 1988. Cell lineage in the rat cerebral cortex: a study

using retroviral- mediated gene transfer. Development 104, 473±482.

Qian, X., Goderie, S.K., Shen, Q., Stern, J.H., Temple, S., 1998. Intrinsic

programs of patterned cell lineages in isolated vertebrate CNS ventri-

cular zone cells. Development 125, 3143±3152.


Rakic, P., 1972. Mode of cell migration to the super®cial layers of foetal

monkey neocortex. J. Comp. Neurol. 145, 61±84.

Reid, C.B., Liang, I., Walsh, C., 1995. Systematic widespread clonal orga-

nization in cerebral cortex. Neuron 15, 299±310.

Rouault, J.P., Falette, N., Guehenneux, F., Guillot, C., Rimokh, R., Wang,

Q., Berthet, C., Moyret-Lalle, C., Savatier, P., 1996. Identi®cation of

BTG2, an antiproliferative p53-dependent component of the DNA

damage cellular response pathway. Nat. Genet. 14, 482±486.

Rouault, J.P., Rimokh, R., Tessa, C., Paranhos, G., Ffrench, M., Duret, L.,

Garoccio, M., Germain, D., Samarut, J., 1992. BTG1, a member of a

new family of antiproliferative genes. EMBO J. 11, 1663±1670.

Temple, S., Qian, X., 1995. bFGF, neurotrophins and the control of cortical

neurogenesis. Neuron 15, 249±252.

Williams, B.P., Price, J., 1995. Evidence for multiple precursor cell types in

the embryonic rat cerebral cortex. Neuron 14, 1181±1188.

Williams, B.P., Read, J., Price, J., 1991. The generation of neurons and

oligodendrocytes from a common precursor cell. Neuron 7, 685±693.

Zhong, W., Feder, J.N., Jiang, M., Jan, L.Y., Jan, Y.N., 1996. Asymmetric

localization of a mammalian Numb homolog during mouse cortical

neurogenesis. Neuron 17, 43±53.

Zhong, W., Jiang, M.M., Weinmaster, G., Jan, L., Jan, Y., 1997. Differen-

tial expression of mammalian Numb, Numblike and Notch1 suggests

distinct roles during mouse cortical neurogenesis. Dev. Suppl. 124,

1887±1897.


pc3 overexpression affects the pattern of cell division of rat cortical

Documents