ABSTRACT
Because zebrafish continue to grow throughout their adult life, they are constantly
generating and differentiating new neural tissue. The periventricular zone (PVZ) of the optic
tectum, given its generalized, undifferentiated cell morphology, is a possible source of neural
precursor cells whose migration into the tectum and differentiation into several different neural
fates may be the source of tectal growth. GFP lipofection of brain cross-slices indicated that at
least some of these cells are dividing much of the time. We developed a novel procedure for the
isolation of PVZ cells and demonstrated growth of neurospheres in culture from these
progenitors when driven under various growth factors (Insulin, IGF, Retinoic acid, BFGF).
Additionally, preliminary antibody staining suggests that at least a subset of these PVZ cells
express the Notchl receptor in vivo, suggesting the possibility of using this receptor as a specific
marker for neural progenitor cells.
INTRODUCTION
The control and regulation of neural tissue regeneration is an area of research that has
immediate relevance for the treatment of human conditions of neural degeneration such as
Parkinson’s, Alzheimer’s, Creutzfeldt-Jakob, and spinal cord injury. Such research generally
involves the search for neural stem cells: cells capable of indefinite division that possess the
ability to differentiate into several neuronal fates. Zebrafish provide a useful model for division
and differentiation in adult neural tissue as they continue to grow throughout their adult lives,
unlike mammals which generally stop growing with the onset of sexual maturity. The zebrafish
brain is constantly generating new neural tissue as it increases in size and has the regenerative
capacity to repair extensive neural damage (S. Thompson, unpublished data).
The optic tectum of the zebrafish is a large, striated, complex structure of the dorsal
zebrafish brain consisting of four zones subdivided into 15 layers (Wullimann et al., 1996). The
tectum receives nonvisual neuronal input from many parts of the brain (Northcutt, 1982) and
probably functions analogously to the mammalian prefrontal cortex with tasks such as multi-
sensory processing and integration. However, the periventricular zone of the optic tectum,
composed of round, undifferentiated cells, has a very different morphology from the rest the
tectum and presumably serves a different function. Tectal growth may be a result of the division.
migration, and differentiation of neural precursors from a reservoir of such cells in the PVZ. A
migration/differentiation model of neural regeneration has been demonstrated by Lois and
Alvarez-Buylla (1994) for mouse neural precursors which migrate from the lateral ventricles to
the olfactory bulb and differentiate into neurons. This study focuses on cell morphology in the
tectum and PVZ using biolistics to introduce a lipophilic dye into the membranes of scattered,
individual cells. Furthermore, GFP lipofection was used as an assay for cell division in the
tectum and PVZ given that only a cell going through the latter stages of mitosis (no nuclear
envelope) can express introduced DNA, having reformed the nuclear envelope around the new
gene.
Although in vivo analysis is instrumental to the identification of potential neural
progenitors, demonstration of division and differentiation can be done most easily in vitro. This
study describes a novel tectal print method for the isolation and amplification of PVZ cells in
culture using insulin, IGF, bFGF, and retinoic acid in defined medium. Neurosphere sizes and
concentrations were evaluated for each growth factor treatment and cultured cell size was
compared to parental PVZ cell size.
A marker specific for neural precursors would be inherently useful for any
characterization of the regulation of their maintenance and differentiation. The Notch family of
receptors, although active in a wide variety of differentiation pathways, may provide such a
marker. Notchl pathway signaling has been implicated in the inhibition of neuronal
differentiation (Faux et al., 2001) and in the maintenance of mammalian neural stem cell state
(Hitoshi et al., 2002). We used immunohistochemistry to explore the Notchl receptor as a
potential marker for neural progenitor cells of the PVZ.
MATERIALS AND METHODS
PVZ and Tectal Morphology by DiIOC6 Biolistics
A zebrafish brain was removed in dissection media (80% PBS with 10mM HEPES and
5% Penstrep adjusted to pH 7.4) and placed on a piece of millipore paper. The paper was placed
onto a cold chopping block and the brain cut into cross-slices. The cross-slices were placed on a
glass slide and bombarded with DilOC6-coated gold bullets of .6um diameter using a Helios
Gene Gun (Biorad). The slices were then viewed using a confocal microscope.
EGFP Transfection of Organotypic Culture
A zebrafish brain was removed in sterile dissection media (80% PBS with 10mM HEPES
and .5% Penstrep adjusted to pH 7.4) and placed on a piece of millipore paper. The paper was
placed onto a cold chopping block and the brain was sliced once through the tecta into anterior
and posterior sections. These were placed in a petri dish, still on the filter paper, and left under a
drop of dissection media. Lipofection took place according to the CellFECTIN protocol (Life
Technologies) using either a GFP gene with a generalized promoter and a Clonetch EGFP gene
with a CMV promoter (pEGFP-CI Vector, Catt 6084-1). 6uL of CellFECTIN reagent was
diluted into 10OuL of 80% DMEMF12 in an Eppendorf tube and 2ug of DNA was similarly
diluted into 100uL in a different Eppendorf tube. Neither Penstrep nor serum were used in
media as they may limit the efficiency of transfection. Both tubes were mixed by inverting
several times and then combined and left shaking for 15 minutes. An additional 800uL of 80%
DMEMF12 was added to the solution before it was introduced to the petri dish containing slices.
The slices were incubated overnight before the media was replaced with a solution of 80%
DMEM/F12 with .05% Penstrep and 10uM insulin and 25ngmL IGF. The organotypic cultures
were left for 2 to 7 days in the incubator before imaging on the confocal microscope.
Tissue Print Method for Isolating Periventricular Zone Cells
The tecta were dissected out of the brain of an adult zebrafish under sterile conditions in a
solution of 80% PBS with 10mM HEPES and 5% Penstrep adjusted to pH 7.4. Each tectum was
placed on a dry patch of millipore paper with the PVZ side in contact with the millipore paper,
then gently scraped off and the millipore paper was suspended upside down in a solution of 80%
DMEM/F12 and .5% Penstrep and placed in an incubator. As two tecta may be harvested from
each dissection, a dual-well glass-bottomed slide was used for each pair of prints with different
growth factors added to each well. The growth factors used were insulin at 1OuM, IGF at
25ng/mL, BFGF at 3uM, or Retinoic Acid at 20uM. After several days, the pieces of millipore
paper were removed from the cultures and placed in new media to seed new cultures. Tissue
prints could yield at least secondary cultures and sometimes tertiary cultures, resulting in as
many as 6 different cell cultures from a single fish. Neural ball counts and size estimates were
made periodically during several weeks of growth in culture.
Notch1 Immunohistochemstry
A zebrafish brain was removed in dissection media (80% PBS with 10mM HEPES and
5% Penstrep adjusted to pH 7.4) and placed on a piece of millipore paper. The paper was placed
onto a cold chopping block and the brain cut into cross-slices, which were fixed in 4%
paraformaldahyde overnight at 4’C, washed with 25%, 50%, and 75% methanolPBS and left in
100% methanol at -20’C until staining (at least 4 hours). The slices were washed again using the
reversed methanol/PBS series and rinsed several times with PBS-Tw. Each slice was placed in
300uL blocking solution (horse serum, kindly provided by the B. Block lab) and left for 10
minutes shaking gently. 3uL of polyclonal goat antiNotchl IgG (sc-6015, Santa Cruz Biotech)
was added to each slice which was incubated shaking at room temperature for 2 hours. The
slices were then rinsed 3 times and washed an additional 4 times for 15 minutes in PBS-D-Tw
and placed back in 300uL of blocking solution (horse serum) for another 10 minutes shaking.
ZuL of FITC-labeled bovine antigoat IgG was added (sc-2348, Santa Cruz Biotech) to each slice
which was incubated rocking at room temperature for 2 hours. The slices were then rinsed 3
times and washed an additional 4 times for 15 minutes in PBS-D-Tw, rinsed in PBS-D, and
placed back in PBS. Finally, the slices were imaged using the confocal microscope.
RESULTS
Biolistics micrographs emphasized the morphological differences between the tectum and
the PVZ (see Fig. 1). Cross-sections of the tectum showed mostly longitudinal striations of
cellular processes, whereas the PVZ consisted of round cell bodies without processes. However.
the boundary is not very clearly delineated and some linear protrusions appear to descend from
the tectal striated layers into the PVZ.
The GFP transfected cultures demonstrated the occurrence of cellular division and the
presence of post-mitotic cells in the tectum and PVZ area (see Fig. 2). The CMV promoted
EGFP and the GFP with a generalized promoter seemed to work equally well. Some cells were
caught in the act of dividing (see Fig. 3). It is not known whether these cells are in the tectum or
in the PVZ.
The tectal prints demonstrated cellular division in culture and neurosphere formation (see
Fig. 4). An IGF/insulin combination in defined medium drove the growth of neurospheres to a
density 3 times greater than insulin alone and produced considerably larger neurospheres after a
12 day incubation (see Fig. 5a). IGF and insulin also out-performed retinoic acid and BFGF by
encouraging larger neurosphere size and higher neurosphere concentration after a 34 day
incubation (see Fig. 6). The average size of an individual neurosphere cell was 4.2um whereas
the average size of the parental PVZ cells was 6.3um (see Fig. 7a). The range in neurosphere
size (as measured by standard deviation) went from 33% of average size to 57% of average size
from day 12 to day 34 (se Fig
Cross-sections stained for the Notchl demonstrated that blood vessels in the brain (not
shown) as well as a small subset of cells in the PVZ (see Fig. 8) expressed a Notch family
receptor in high enough quantities and with enough homology to the mammalian Notchl to stain
with a polyclonal anti-human Notch-1.
DISCUSSION
Some processes crossing the boundary into the PVZ (see Fig. 1) suggest that there may
be tectal control over the fate of PVZ cells including division, migration, and differentiation.
Conversely, the control may be in the way of PVZ cells sending processes into the tectum itself.
Either way, there is likely signaling occurring between distant cells that organizes the movement
between layers. An injection of lipophilic dye into the tectal layers may determine the location
of the cell bodies associated with these processes. An early study with horseradish peroxidase
used a similar method to determine the location of processes afferent to the tectum from other
parts of the brain (Northcutt, 1984).
GFP lipofection was successful in demonstrating cell division subsequent to dissection in
the PVZ and tectum (see Fig. 2 and 3). Given the slicing technique used (one slice dividing the
brain into anterior and posterior sections), the assignment of post-mitotic cells into only striated
tectum or only PVZ was challenging as discrimination of adjacent tissue layers under
fluorescence is difficult in deep tissue. However, the use of relatively large tissue sections had
merit in that it left large areas of tissue far from the blade path and relatively undisturbed. The
presence of dividing cells in tissue away from the traumatized edge (see Fig 2c and 2d) suggests
that the observed division is not solely an over-reactive glial response to injury. The exact
location of these post-mitotic cells (PVZ versus tectum) may be clarified by a whole brain
lipofection and incubation, followed by slicing into thin cross-sections immediately prior to
visualization. Using different incubation times, it might even be possible to identify division in
the PVZ and successive migration into the tectum.
Tectal printing demonstrated the expansion of PVZ neural progenitors into neurospheres
over a period of several weeks when driven with growth factors (see Fig. 4). IGF and insulin
combined proved to be a better promoter of neurosphere growth than insulin alone or retinoic
acid and BFGF (see Fig. 5 and 6). Because there was a variation in seeding density from each
tectal print, neurosphere size was examined along with neurosphere concentration. Remaining
tissue remnants also presented a complication as they were at times difficult to discern from
growing neurospheres and may have had some effect on the averaged neurosphere size of the
different growth factor treatments. The use of dual tectal prints from a single fish in growth
factor trials controlled for genetic differences in the expansion of paired cultures and may be
responsible for the higher concentration of neurospheres in the 12 day IGF/insulin incubation
over the 35 day IGF/insulin incubation.
Neurosphere expansion might be more closely monitored using a sterile microscope stage
and a 96-well plate to isolate individual cells or neurospheres for clonal analysis, as described by
Weiss et al. (1996). Similarly, CFDASE, a lipophilic dye of great use for tracking cell division
by flow cytometry (Hasbold et al., 1999), can be used to quantify cell division in culture. This
approach was attempted in this study but was unsuccessful due to a high background
fluorescence in the media, even with staining of the entire tectum prior to printing and seeding.
Staining and printing within a well insert and movement of the insert along a series of wells to
dilute out the CFDASE-laden media may make this approach viable and may reveal
differentiation within a neurosphere.
Differentiation in culture was suggested by the 30% size reduction of neurosphere cells
from PVZ cell size of approximately 6um (see Fig. 7a). Additionally, the standard deviation in
neurosphere size increased over time in culture (see Fig. 7b), indicating that cells in some neural
balls were dividing more rapidly than cells in others. These data suggest that differentiation may
be occurring in culture. Öther studies have shown that growth factors such as retinoic acid
promote neuronal differentiation (Huang et al. 1998) whereas FGFI and FGF2 inhibit neuron
differentiation from neuroepithelial precursor cells by upregulating the expression of the Notchl
receptor (Faux et al., 2001); these and other growth factors may prove to be a potent tool for
preserving the precursor state and manipulating different lines of differentiation from tectal print-
derived neurosphere cells.
Preliminary analysis of Notchl antibody staining suggests that a Notch family receptor is
present in a subset of PVZ cells (see Fig. 8). Given that the available antiNotch IgG is a
polyclonal antibody to a human protein, it is rather surprising that in the zebrafish PVZ there is
enough homology as well as expression for any staining at all. Notch immunohistochemistry
was unsuccessful on the cells derived from tectal printing as poly-lysine proved too moderate to
keep the neurospheres stuck to a slide through the various steps of an antibody stain and the
fixing procedure dissolved both nail polish and paint pen ridges designed to provide a shallow
well for the process on a glass slide. However, acrylic ridges demarking wells on a glass slide
and a single, unmoving stage with a variable temperature setting for the entire fixing/staining
process should permit antibody staining of neurospheres. In addition, Notch ligands such as
Delta, Jagged, Serrate, and Lag2 may prove useful for exploring the controls of division and
differentiation of these neural precursor cells. Öther proteins may prove to be useful markers as
well, including the protein product of tai-ji, a gene expressed in various parts of the adult
zebrafish brain and in human neuronal precursor cell line hNT2 (Huang et al., 1998)
CONCLUSION
Tectal printing has demonstrated that cells from the PVZ are self-renewing under
IGF/insulin treatment in defined medium; differential cell size and neurosphere growth suggest
some differentiation in culture. It remains to be shown whether the cultured cells have
characteristics of either glia or neurons. GFP lipofection was useful as an assay for mitotic cells
in organotypic culture and showed the presence of divided cells in the tectum and the PVZ; a
slight change to the slicing procedure may elucidate the exact location of division and the
presence or absence of any subsequent migration. This hypothetical sequence of events may be
controlled by interactions between widely separated cell bodies suggested by biolistics imaging
of cell processes which transverse the boundary between the striated tectum and the PVZ. Such
interaction may involve the Notch receptor pathway as a Notch receptor is expressed in a subset
of PVZ cells and exists as a potential marker for the maintenance of the neural progenitor cell
state.
LITERATURE CITED
LITERATURE CITED
Faux, C., A. Turnley, R. Epa, R Cappal, and P. Bartlett. 2001. Interactions between
Fibroblast Growth Factors and Notch regulate neuronal differentiation. J. Neuroscience.
21(15):5587-5596.
Hasbold, J., A. Gett, J. Rush, and E Deenick. 1999. Quantitative analysis of lymphocyte
differentiation and proliferation in vitro using carboxyfluorescein diacetate
succinimidyl ester. Immunology Cell Bio. 77(6):516-522.
Hitoshi, S., T Alexson, and V. Tropepe. 2002. Notch pathway molecules are essential
for the maintenance, but not the generation, of mammalian neural stem cells. Genes and
Develop. 16:846-858.
Huang, S., and S. Sato. Progenitor cells in the adult zebrafish nervous system express a
Brn-1-related POU gene, tai-ji. J. Mech. Dev. 71(1):23-35.
Lois, C. and A. Alvarez-Buylla. 1994. Long-distance neuronal migration in the adult
mammalian brain. Science. 264(5162):1145-1148.
Northcutt, R.G. 1982. Localization of neurons afferent to the optic tectum in longnose
gars. J. Comparative Neurobio. 204(4):325-335.
Weiss, S., C. Dunne, J. Hewson, and C. Wohl. 1996. Multipotent CNS stem cells are
present in the adult mammalian spinal cord and ventricular neuroaxis. J.
Neuroscience. 16(23):7599-7609.
Wullimann, M., B. Rupp, and H. Reichert. 1996. Neuroanatomy of the Zebrafish Brain
Birkhauser Verlag, Basel, Switzerland.
Figure 1: Biolistics micrographs demonstrating the morphology of the tectum and the PVZ by
DilOC6 staining. Images taken with a confocal microscope using a 40x water lens. In both
cases the PVZ is on the left and the striated tectum is on the right. Arrows denote cellular
processes extending into the PVZ layer.
13
Figure 2a,b,c.d: Micrographs of GFP-lipofected organotypic culture. Post-mitotic cells
apparent in the tectum/PVZ. 2c: Micrograph from area 11 in Fig. 2d demonstrating dividing
cells in deep tectal tissue, far from blade path. 2d: representation of the posterior portion (dorsal
view) of a brain sliced crossways through the tecta. Brainstem at top, blade path at bottom, tecta
are the two large lobes on left and right.
2d


L

knas
14
Figure 3a, b,c: Micrographs of GFP-lipofected organotypic
culture revealing recent (3a, 3c) and ongoing (3b) cell division in
the tectum/PVZ. Images taken with a confocal microscope using
a 40x water lens.
Figure 4: Neurospheres formed from a tectal print following a 12 day incubation in defined
media with lOmM insulin and 25ng/mL IGF. Note the wide range in cell size and neurosphere
size suggesting differential division within and between neurospheres. Nomarsky imaging
courtesy of Stuart Thompson.
50 Microme
Figure 5: Neurosphere sizes and concentrations from dual tectal print cultures incubated for 12
days in defined medium with lOuM insulin or with lOuM insulin and 25ngmL IGF.
3500
24
H23
3000
22
L3080
2500
20
18
5 2000
E15
1500
16
1030
1000
14
Hneural
ball
500
12
conc.
Hneural
10
ball
IGF and Insulin
Insulin
sizes
Figure 6: Neurosphere sizes and concentrations from dual tectal print cultures incubated in
defined medium with 10 mM insulin. Aster 12 days in culture, 20 uM retinoic acid and ZUM
BFGF were added to one culture. Neurosphere sizes and concentrations were determined after a
total of 35 days in culture.
1600
26
H25
1400
24
□1450
1200
22
20
1000
20
800
18
600
16 ;
□ neural
ball
400
14 7
conc.
1283
12
200
Eneural
10
ball
retinoic acid and IGF and insulin
sizes
DFGF
Figure 7a: Average PVZ cell diameter and
neurosphere cell diameter. Error bars
represent standard deviation.
Figure 7b: Neurosphere size after 12 and 34
days incubation in culture. Error bars
represent standard deviation.
40
35
25
17.44
20
15
2 10
8
27
86
94
22
1
25.14
E PVZ cell diameter
Ineurosphere cell diameter
6.3
4.2
□12 days after
tectal printing
(st dev 33%)
□ 35 days after
tectal printing
(st dev 57%)
18
Figure 8a,b: Immunohistochemistr
of the PVZ by confocal microscopy. A small subset of
PVZ cells stain with antihuman Notchl IgG