DEVELOPMENTAL DYNAMICS 238:241–248, 2009

PATTERNS & PHENOTYPES

Osterix-mCherry Transgenic Medaka for In

Vivo Imaging of Bone Formation
Joerg Renn and Christoph Winkler*

Intramembranous and chondral bone formation by osteoblasts is found in all vertebrates. The genetic
network controlling osteoblast differentiation is highly conserved and regulated by a small number of key
factors, including the zinc-finger transcription factor Osterix. Expression analysis of osterix in the teleost
model medaka revealed a highly restricted expression in skeletal regions. For in vivo imaging, we generated
transgenic medaka expressing mCherry under control of the osterix promoter. We show that the transgene
becomes expressed in early osteoblasts, which have not yet mineralized bone matrix, and remains high in
matured and mineralizing osteoblasts. Life imaging of transgenic larvae provided insight into the
appearance and behavior of early osteoblasts during development of the teleost cranium, vertebrae, and
caudal fin. In summary, osterix-mCherry transgenic medaka enable us to analyze osteoblasts during
different maturation phases in vivo and represent a unique tool to study osteoblast behavior in vertebrate
embryos and adults. Developmental Dynamics 238:241–248, 2009. © 2008 Wiley-Liss, Inc.

Key words: bone formation; osteoblast; osterix; medaka; transgenic fish; life imaging

Accepted 15 November 2008

INTRODUCTION
Vertebrate bone formation is charac-
terized by two distinct modes: chon-
dral and intramembranous ossifica-
tion. During chondral bone formation,
cartilage is formed initially and is ei-
ther replaced by bone (endochondral
ossification) or becomes surrounded
by bone (perichondral ossification). In-
tramembranous bone, on the other
hand, is formed directly without any
cartilage scaffold. In general, chondral
and intramembranous bone formation
is found in all vertebrate species. In
both cases, ossification is performed
by specialized bone matrix producing
cells, the osteoblasts. These differen-
tiate from mesenchymal precursor
cells on a pathway that is tightly con-
trolled by a complex interplay of tran-
scription factors, hormones, growth
factors, and cell– cell and cell–matrix
factors. Several factors important for
bone formation in human and other
mammals have also been identified in
teleosts, such as medaka and ze-
brafish. The corresponding orthologs
share significantly similar sequences
and expression patterns. These find-
ings indicate that the genetic net-
works controlling bone formation are
highly conserved in all vertebrates
(Fisher et al., 2003; Flores et al., 2004;
Yan et al., 2005; Renn et al., 2006b;
Nemoto et al., 2007).
For investigating molecular mecha-
nisms underlying bone development, in
vivo models have clear advantages over
cell culture systems. To study develop-
ment of cells and tissues in vivo in living
organisms, transgenic animals express-
ing fluorescent reporters under the con-
trol of cell or tissue-specific promoters
have become important research tools.
Especially teleost species, such as ze-
brafish and medaka, represent attrac-
tive model systems due to features like
external development, large offspring
numbers, transparency of embryos and
larvae, short generation time, and
availability of mutants and whole ge-
nome sequences. Excellent examples of
how, for example, transgenic zebrafish
have helped to address a wide range of

Additional Supporting Information may be found in the online version of this article.
Department of Biological Sciences, National University of Singapore, Singapore
Grant sponsor: European Space Agency (ESA); Grant number: 15452/01/NL/SH; Grant sponsor: A-STAR BMRC; Grant number: 07/1/21/
19/544.
*Correspondence to: Christoph Winkler, Department of Biological Sciences, National University of Singapore, 14 Science
Drive 4, Block S2, Level 4, Singapore 117543. E-mail: dbswcw@nus.edu.sg
DOI 10.1002/dvdy.21836
Published online 17 December 2008 in Wiley InterScience (www.interscience.wiley.com).

© 2008 Wiley-Liss, Inc.

242 RENN AND WINKLER
fundamental biological questions are
the gata1-eGFP and fli1-eGFP trans-
genic lines (Long et al., 1997, Lawson
and Weinstein, 2002).
For bone research, twist-eGFP and
osteocalcin-DsRed transgenic medaka
lines have recently been established.
These lines contributed significantly
to our understanding of the dynamic
processes during early vertebrae forma-
tion (Inohaya et al., 2007). Twist ex-
pression is a marker for mesenchymal
cells, which differentiate into several
cell types, including adipocytes, chon-
drocytes, and osteoblasts (Komori,
2006). Osteocalcin, on the other hand, is
a marker for mature bone matrix pro-
ducing cells in mammals as well as te-
leost species (Aubin and Liu, 1996;
Pinto et al., 2003; Gavaia et al., 2006).
However, because twist-positive mesen-
chymal precursor cells still have to com-
plete several intermediate stages of dif-
ferentiation before becoming active
osteoblasts, a gap remains between
twist and osteocalcin-positive cells. This
gap needs to be filled to gain more de-
tailed insight into the early processes of
bone formation.
Osterix is a zinc finger transcription
factor, which plays important regula-
tory roles during differentiation of
preosteoblasts to mature osteoblasts
(Nakashima et al., 2002). It is sug-
gested that Osterix drives osteoblast
differentiation while concomitantly
inhibiting chondrocyte development.
In the present study, we analyzed the
expression of osterix during medaka
embryogenesis by whole-mount in situ
hybridization. Medaka osterix is ex-
pressed in regions of early bone forma-
tion and remains highly expressed in
mature osteoblasts that mineralize
bone matrix. This makes this factor an
ideal candidate to label various differ-
entiation stages between twist-positive
precursors and osteocalcin-positive ma-
ture osteoblasts. We, therefore, estab-
lished a transgenic medaka line that
expresses mCherry under the control of
the osterix (osx) promoter. mCherry ex-
pression is restricted to the forming
skeleton and recapitulates endogenous
osterix expression both spatially and
temporally. Because the transgene re-
mains expressed in mature osteoblasts,
we can analyze osteoblasts over a long
time period. In summary, this study
demonstrates that osx-mCherry trans-
genic medaka represents a unique in
vivo tool to analyze osteoblast formation
and function in a living organism.
RESULTS
Osterix Is Exclusively
Expressed in Regions of
Skeletogenesis in Medaka
To analyze the expression pattern of
medaka osterix, we performed RNA
whole-mount in situ hybridization in
embryos at different developmental
stages. Osterix expression was first
detected in single cells located in a
position anterior–laterally to the first
pair of somites (data not shown). Ex-
pression in this region, at the future
position of the cleithrum, becomes
more prominent during further devel-
opment and extends into all following
embryonic stages (Fig. 1A–C). At
stage 28 (2.5 days postfertilization
[dpf], days postfertilization), osterix is
expressed in the ventral head, as well
as laterally to the notochord at the
transition from head to trunk (Fig.
1A). During the next 2 days, osterix
expression is found in regions of in-
tramembranous bone formation, such
as the forming parasphenoid, cle-
ithrum, operculum, first pair of bran-
chiostegal rays, and upper and lower
pharyngeal jaws (Fig. 1B, and data
not shown). Expression in these re-
gions further expands within the fol-
lowing days of development. At later
stages, osterix becomes also expressed
in those regions of the cartilage that
undergo perichondral ossification, such
as the occipital arch and, 1 day later,
the hyosymplectic (Fig. 1C). After
hatching, osterix is still mainly ex-
pressed in intramembranous bones, but
transcription rapidly increases in re-
gions of cartilage bone formation, such
as the ceratobranchials, ceratohyal, and
basihyoid (Fig. 1C). The position of osx
transcripts in the head overlaps with
the ossification centers surrounding
these cartilage bones at 20 dpf (S2A, B;
Supp. Fig. S2, which is available on-
line).
In the axial skeleton, first osterix
expression is found at 6 dpf surround-
ing the notochord. There, it demar-
cates the regions of vertebrae forma-
tion and remains highly expressed
until the embryo hatches (Fig. 1D,E).
In free swimming larvae, transcripts
also appear in the developing caudal
fin rays (Fig. 1F), the two hypurals
(data not shown), and both neural and
hemal arches (Fig. 1G).
Generation of osterix-
mCherry Transgenic Medaka
We amplified a 4.1-kb fragment (in-
cluding a putative Runx2 binding site)
upstream of the osterix (osx) transla-
tional start site from genomic DNA.
This fragment was cloned in front of
mCherry into a vector with flanking
I-SceI meganuclease restriction sites.
Below, this construct is referred to as
osx-mCherry (Fig. 2A). Co-injection of
osx-mCherry and I-SceI meganuclease
into one-cell stage medaka embryos
resulted in mosaic expression of
mCherry in the early head and axial
skeleton, including the cleithrum,
parasphenoid, operculum, branchio-
stegal rays, and, after hatching, in the
neural arches (data not shown). A to-
tal of 34 of 331 injected embryos
showed expression at least in the cle-
ithrum before hatching. These em-
bryos were selected and grown to
adulthood. Founder fish (G0) were
identified by crossing them to wild-
type fish. The resulting heterozygous
offspring (F1) were used to maintain
the line and to analyze transgene ex-
pression. In total, we identified 13
founder fish transmitting the trans-
gene in a Mendelian pattern suggest-
ing single loci integration into the ge-
nome. Expression of the transgene
varied slightly in onset and strength
of expression in these founders. All
data presented here refer to heterozy-
gous offspring of one identified
founder female (G0). Fifty percent of
analyzed F1 offspring showed early
and strong expression of mCherry,
while the remaining embryos were
negative, indicating an early germ
line integration event.
Bone-Specific mCherry
Expression in Stable osx-
mCherry Transgenic Medaka
Fish
First mCherry expression was de-
tected in the epidermal fold of the pec-
toral fins (Fig. 2B,C) and the tail at 2.5
dpf (data not shown). Additionally,
transgene expression was identified in
the anterior head region (Fig. 2C) and

in cells of the tail tip (data not shown).
Expression in these regions is tran-
sient and disappears with further de-
velopment. At 3 dpf, the transgene be-
comes expressed in the cleithrum and
in two spots dorsally to the auditory
capsula (Fig. 2B). One day later,
mCherry is highly expressed in the
parasphenoid and the onsets of oper-
culum (Fig. 2C). At 5 dpf, expression is
found in the first pair of the branchio-
stegal rays, the pharyngeal jaws, and
the occipital arches (Fig. 2D, data not
shown). During the following days, ex-
pression becomes stronger and ex-
pands into the regions of future bone
formation (Fig. 2E–G). In the axial
skeleton, mCherry becomes first ex-
pressed in the neural arches. At 8 dpf,
fluorescence is detected in the most
anterior neural arches and in the first
rays of the caudal fin (Fig. 2E, and
data not shown). At 7 days after
hatching (dph), the larvae show ex-
pression in all neural and hemal
arches and in the centrae, with strong
expression at their edges, but also in
single cells in the middle of the cen-
trae (Fig. 2H,I). Additionally, strong
fluorescence is found in the fin rays
and hypurals of the caudal fin (Fig.
2H,J). Detailed analyzes by laser
scanning microscopy showed layers of
single cells along bony structures,
such as the branchiostegal rays, oper-
culum, centrae, neural arches and hy-
purals, and fin rays of the caudal fin
(Fig. 2K–N; videos with three-dimen-
sional [3D] projections are available
from the authors on request). In adult
fish, the transgene remains expressed
in mature bones, for example, in the
operculum, rib bones, vertebrae, fin
rays, and in scales (Fig. 2O–Q, and
data not shown). Double in situ hy-
bridization with osx and mcherry
probes clearly showed that both, en-
dogenous gene and transgene, are ex-
pressed in the same cells (S4A, B).
Expression of mCherry in
Osteoblasts Precedes
Ossification
To determine the position of mCherry-
positive osteoblasts in the spatial con-
text of ossifying matrix, we stained
calcifying bone matrix in transgenic
fish with calcein and analyzed miner-
alization by fluorescence microscopy
with green fluorescent protein (GFP)
Osterix:mCherry TRANSGENIC MEDAKA 243
filter settings. In the head of hatched
larvae, mCherry-positive osteoblasts
are located at the surface of already
ossified bones, such as the operculum,
branchiostegal rays, and hyosymplec-
tic (Fig. 3A–C). In the axial skeleton,
osteoblasts appear mainly at the
anterior and posterior edges of the
mineralized centrae and as single
cells in the middle region (Fig. 3D–F).
mCherry-positive cells surround the
mineralized neural and hemal arches
but are also found at more distal posi-
tions in the not yet ossified parts of
the arches (data not shown). Medaka
larvae show ossification of the caudal
fin rays and around the cartilage of
the hypurals already at hatching. The
transgenic analysis shows that osteo-
blasts are already located in the fin
rays that are not yet mineralized and
are found more distal to the ossified
border of the hypurals (Fig. 3G–I).
This strongly suggests that mCherry
expression precedes the mineraliza-
tion of these skeletal elements. To con-
firm this, we analyzed embryos that
are double transgenic for osx-mCherry
and enhanced GFP (eGFP) under con-
trol of the osteocalcin promoter (osc-
eGFP). In the double transgenic em-
bryos, we detected early mCherry-
positive osteoblasts, for example, at
the edges of the operculum or the
newly formed branchiostegal rays,
which initially were eGFP-negative.
More differentiated osteoblasts in the
center of the operculum or branchio-
stegal rays in contrast started to ex-
press eGFP (S3A–C).
In Vivo Analysis of Early
Osteoblasts by Life Imaging
The parasphenoid, a planar rhombic
bone that is located medially between
the eyes, is the first intramembranous
(dermal) bone of the cranial skeleton in
medaka (Langille and Hall, 1987). Us-
ing the mCherry transgenic medaka,
we analyzed the dynamics of parasphe-
noid development during embryogene-
sis. For this, we mounted medaka em-
bryos at 4 dpf in low melting point
agarose and followed development us-
ing a live imaging system over the next
24 hr. At 4 dpf, no transgene expression
in the parasphenoid is detectable.
Transgenic mCherry expression in
early osteoblasts became detectable 2
hr after the onset of analysis in two
centers within the presumptive para-
sphenoid region. Over time, these cen-
ters expand anteriorly and posteriorly
and eventually fuse before they expand
laterally and establish the shape of the
mature parasphenoid (Fig. 4A1–A6).
Ossification of the parasphenoid, how-
ever, is found only later at 5 dpf (data
not shown). Time-lapse analysis of the
forming vertebrae starting at 5 dpf
showed that mCherry-positive osteo-
blasts appear first in the proximal part
of the neural arches and over time in
more distally located regions (Fig. 4B1–
B6). Subsequent onset of expression
was detected in a continuous manner
along the anterior to posterior axis. In
the caudal fins, the first osteoblasts
were detected in a single fin ray at the
border to the developing hypural.
Again, onset of mCherry expression ap-
pears sequentially in the adjacent fin
rays (Fig. 4C1–C6). While osx-mCherry
is clearly visible early, ossification of
neural arches and fin rays takes place
after a significant delay at 5 pdf and 9
dpf, respectively (data not shown; for
video sequences, see Supp. Fig. S1A–C).
DISCUSSION
In this study, we report the expression
of osterix during embryonic develop-
ment in medaka. We have chosen the
medaka for this analysis, as bone for-
mation is more rapid in this model
compared with zebrafish. Appearance
of first alizarin red–positive bone ma-
trix in the vertebrae, for example, is
detected 5 days after fertilization in
medaka (Inohaya et al., 2007), while
the same structures become visible
only at 9 days in zebrafish (Gavaia et
al., 2006). Osterix is a key regulator
driving the differentiation of mesen-
chymal progenitor cells into the osteo-
blast lineage in mammals (Na-
kashima et al., 2002). In medaka, we
show that osterix is expressed in a
highly restricted pattern in early re-
gions of bone formation suggesting a
conserved role in osteoblast differenti-
ation. Its onset of expression precedes
the mineralization of intramembra-
nous and cartilage bones and makes
osterix an ideal marker for early os-
teoblasts, also in teleost fish. Subse-
quently, its expression continues in
mature osteoblasts, which have
started to produce and ossify bone
matrix.
244 RENN AND WINKLER

Fig. 1.

Fig. 2.
 Osterix:mCherry TRANSGENIC MEDAKA 245

During osteoblast differentiation in eral, osterix promoter driven transgene around the notochord also at earlier
mammals, Osterix transcription is expression very well recapitulates the stages.
regulated by Runx2, which binds to endogenous osterix pattern in regions Using double labeling of mineral-
the sequence “AGTGGTT”. This motif of embryonic bone formation. How- ized tissue with calcein in transgenic
is conserved in mouse, human, and rat ever, surprisingly, we also detected medaka, we were able to compare the
Osterix promoters (Nishio et al., weak fluorescence in a few regions appearance of osteoblasts with the
2006). The promoter region that we where we do not see osterix transcrip- stage and position of ossification. In
chose to generate transgenic medaka tion by RNA in situ hybridization. the head skeleton, first mCherry-pos-
lines also contains the identical bind- This includes the epidermal fin folds itive osteoblasts appear in intramem-
ing site, indicating a regulatory func- of the pectoral fins and tail (see Fig. branous bone regions, before any cal-
tion of Runx2 for osterix also in 2B), the anterior-most head region cification takes place. The same was
medaka. Intriguingly in medaka, a (see Fig. 2B,C) and elongated cells, observed for the caudal fin (hypurals
second putative Runx2 binding site most likely muscle fibers, in the pos- and fin rays). Furthermore, in osx-
with identical sequence is located in terior trunk during early develop- mCherry (osterix) and osc-eGFP (os-
the unique intron (data not shown). ment. Expression in these regions dis- teocalcin) double transgenic medaka,
Two transgenic twist-eGFP and osteo- appears during further development. early osteoblasts expressing mCherry
calcin-DsRed medaka lines have recently It seems unlikely that these ectopic are osteocalcin negative initially, but
been reported for the analysis of bone for- sites are caused by positional effects become positive during later differen-
mation (Inohaya et al., 2007). Unfortu- at the transgene insertion site, as sev- tiation. These findings clearly show
nately, these lines do not cover the inter- eral independent lines showed identi- that osterix promoter activity is found
mediate stages when twist-positive cal patterns. We rather speculate in early osteoblasts, which have not
mesenchymal progenitors convert into os- that, due to the significant stability of yet produced calcified bone matrix.
teocalcin-expressing mature osteoblasts. the brightly fluorescent mCherry pro- Importantly, promoter activity is still
Therefore, we decided to establish a tein, expression sites are revealed found in mature and bone matrix pro-
transgenic medaka line that expresses that are characterized by only weak ducing osteoblasts. Therefore, this
mCherry under the control of the osterix and transient activation of the endog- transgenic line allows the analysis of
promoter (osx-mCherry). The heterozy- enous promoter, which escapes detec- osteoblasts in vivo across a wide range
gous transgenic fish described here show tion by RNA in situ hybridization. of stages. Moreover, we observe
early and strong expression of mCherry Similarly, we were not able to detect mCherry expression also in bones and
in bone forming regions. In the parasphe- osx transcription around the noto- scales of adult animals possibly sug-
noid, for example, mCherry fluorescence chord at 2 dph using RNA in situ hy- gesting de novo osteoblast differentia-
is detectable approximately 12 hr after bridization under the experimental tion and activity in adult bones and
onset of endogenous osterix transcription. conditions used. In osx-mCherry scales.
This suggests that mCherry expression is transgenics at 7 dph, however, We analyzed the appearance of os-
only slightly delayed due to maturation of mCherry is expressed in all centrae, teoblasts in the parasphenoid, verte-
newly translated reporter protein. In gen- suggesting endogenous osx expression brae, and caudal fin by life imaging.
Our analyzes showed for the first time
that osteoblasts in the parasphenoid
anlage differentiate from two juxta-
posed centers, which grow together
Fig. 1. Expression of osterix in skeletal elements of medaka embryos analyzed by whole-mount
RNA in situ hybridization. A–C: Ventral views of the head at 2 days postfertilization (dpf; A; stage
and expand laterally and in anterior
28), 5.5 dpf (B; stage 35), and after hatching (C; stage 40). Arrows in B and C indicate pharyngeal direction to form the rhombic shape of
jaws (B) and hyosymplectic (C), respectively. Arrowheads in C mark ceratobranchials. D,F,G: the parasphenoid.
Lateral views of trunk regions and caudal fin at 6 dpf (D; stage 36) and after hatching (F,G; stage One possible limitation of our trans-
40). E: Transverse section through trunk at 6dpf (stage 36). bs, branchiostegal rays; cl, cleithrum;
gene is that it apparently does not
d, dental; ha, hemal arches; na, neural arches; p, parasphenoid; op, operculum.
recapitulate endogenous osterix ex-
Fig. 2. Transgenic medaka fish expressing mCherry under control of the osterix promoter. A: pression found around the notochord
osx-mCherry construct used to generate transgenic fish. Numbers refer to the nucleotide position from 6 dpf until hatching. At present,
upstream of the ATG translation start. B–Q: osx-mCherry stable transgenic larvae and adult fish. B: we can only speculate that the ab-
Lateral view, expression dorsal to the auditory capsula (arrow). C: Dorsal view, parasphenoid
(arrowhead). D: Lateral view, operculum (black arrow), occipital arches (white arrow), branchioste-
sence of mCherry expression in the
gal rays (arrowhead). Note that expression in the epidermal fin fold has disappeared (compare with early axial skeleton is due to missing
B,C). E: Dorsal view, neural arches (arrowhead). F,G: Dorsal and ventral views of the head, regulatory elements in the used pro-
respectively, in a hatchling (arrowhead, quadrate). H–J: Lateral view of larvae after hatching moter region. At this point, we can
(approximately 7-mm body length). I,J: Details of H at higher magnification. B–J: Overlays of images
exclude that we miss possible regula-
taken with brightfield, FRITC and green fluorescent protein (GFP) settings. K–N: Laser scanning
analyzes (projections) of hatched larvae (approximately 7-mm body length), anterior to the left. K: tory elements in the first intron. We
Ventral view, single cell layers at hyosymplectic (arrow) and branchiostegal rays (arrowheads). L: have established additional lines,
Lateral view, operculum, large cells at ventral edges of the operculum (arrowheads). M: Lateral which carry the complete sequence of
view, single cells at anterior edge of a centrum (arrow), elongated cells at neural arches (arrow- the only intron and the first exon in
head). N: Lateral view of caudal fin, single elongated cells at surface of a fin ray (arrow) and hypurals
(arrowheads). O–Q: Adult transgenic medaka (male, 2.6-cm body length), expression in adult bones
addition to the same upstream genomic
and scales (see P and Q for higher magnification). O–Q: Overlays of images taken with FRITC and region of medaka osterix. When com-
GFP filter settings. pared with the pattern described in this
Fig. 3.

Fig. 4.
 Osterix:mCherry TRANSGENIC MEDAKA 247

report, the intron containing trans- also appearing in the middle of the mCherry double transgenic lines to make
genes show the identical mCherry pat- centrae. Inohaya et al. (2007) sug- this point.
tern, including the lack of notochord ex- gested that twist-eGFP– expressing In summary, osx-mCherry trans-
pression (data not shown). cells differentiate into four distinct genic fish represent a unique model to
In a further study, the medaka pro- cell types (I–IV). Two of these cell give detailed insight into the develop-
moter construct described here has types (I, IV) are located at the outer ment of early and mature osteoblasts
also been used to generate transgenic surface of the edges and the middle in vivo. Double or even triple trans-
zebrafish (Spoorendonk et al., 2008). part of the centrae and later become genic fish carrying osterix, as well as
Similar expression patterns of this osteocalcin-positive. These two cell twist or osteocalcin promoters can be
transgene in the analyzed late larval types most likely correspond to the used in the future to study osteoblast
zebrafish suggest that the regulatory osx-mCherry-positive cells in our differentiation at a cellular level in
control of osterix expression is con- transgenic lines. We did not observe vivo in a whole organism.
served in both teleost species. Fur- mCherry expression between two cen-
thermore, injection of a zebrafish BAC trae, as was described for the two ad- EXPERIMENTAL
clone carrying GFP inserted in frame ditional twist-positive cell populations PROCEDURES
in the osterix locus gives the identical (II, III), which do not become osteocal-
pattern. As in medaka, the transgene cin-positive (Inohaya et al., 2007). This Analysis of Gene Expression
does not result in expression around shows that osx-mCherry expression en- by RNA Whole-Mount In
the notochord during early develop- ables to distinguish early between Situ Hybridization
ment (K. Spoorendonk and S. Schulte- twist-positive cells that differentiate
Medaka embryos were obtained from
Merker, personal communication). This into osteoblasts and those that do not.
matings of wild-type fish of the Golden
strongly suggests that the absence of It, furthermore, opens the possibility
strain kept as closed colony stocks in
reporter expression around the noto- that osterix-negative cells are responsi-
our institute since many generations.
chord is not a limitation of the used ble for the early calcification of the ver-
Embryos were morphologically staged
promoter region. Future studies have to tebrae. Yasutake et al. (2004) showed according to Iwamatsu (2004) and fixed
show whether, for example, migration that twist-eGFP–positive cells migrate in 4% paraformaldehyde at desired
of osterix RNA-positive precursor cells around the notochord into the neural stages. The nucleotide sequence of
from the center (as evident by in situ arch region. Before onset of mineraliza- medaka osterix, which is available in
hybridization) to the edges of the form- tion, the eGFP fluorescence disappears the medaka genome sequence database
ing vertebrae (as evident shortly later in the neural arches. The first, osx- (http://www.shigen.nig.ac.jp/medaka/
by transgene expression) is a possible mCherry-positive cells are found in the genome/top.jsp; Ensembl Transcript ID:
explanation for this observation. proximal part of the neural arch region ENSORLT00000006574) has been used
Of interest, in the axial skeleton, and expression expands distally. To- to design primers 5⬘-ACTCCTACACT-
initial calcification of the centrae of gether, this suggests that twist- GGCTCCTTCAC-3⬘ and 5⬘-GTCTTTC-
the vertebrae occurs before mCherry- positive progenitor cells migrate CAGCTCCTGACAGTT-3⬘. We ampli-
expressing cells appear. mCherry flu- first into the neural arch region and fied a partial sequence of 672 bp and
orescence is first detectable in the subsequently differentiate into osterix- cloned it into the pMBL vector
neural arches, as shown by life imag- positive osteoblasts. However, more de- (Genaxxon) to obtain the plasmid pMB-
ing, and only later at the edges of the tailed analyses are required in the future, Losterix. For preparation of osterix and
centrae with additional single cells for example, using twist-EGFP and osx- mCherry antisense riboprobes, the plas-
mids pMBPosterix and pCSmcherry
were digested with HindIII and
BamHI. T3 and T7 RNA polymerase
were used for in vitro transcription, re-
spectively. In situ hybridizations were
Fig. 3. In vivo analysis of ossification in transgenic medaka larvae after hatching. Ossification and
position of osteoblasts are shown by calcein staining (A,D,G) and mCherry expression (B,E,H), performed as described earlier (Renn et
respectively. C,F,I: Overlays of images showing calcein stainings and mCherry expression. A–C: al., 2006a). Stained embryos were
Ventral views, osteoblasts (arrowheads) surrounding ossified hyosymplectic (hy), branchiostegal mounted in glycerol for photography.
rays (bs), and operculum (op), e, eye. D–F: Lateral views of precaudal vertebrae, osteoblasts Sections were made manually using a
(arrowheads) at the edges of the ossified centrae (c). G–I: Lateral views of caudal fin, ossification
around notochord (white arrow in G), mCherry expression in fin rays (white arrow in H), ossified
razor blade.
border of hypural (red arrowheads in G,I), and proximal border of mCherry expression in hypural
(green arrowheads in H,I).
Osterix-mCherry and
Fig. 4. Life imaging analysis of mCherry expression in the head at 4 days postfertilization (dpf; osteocalcin-eGFP Transgenic
A1– 6), anterior vertebrae (B1– 6) and caudal fin (C1– 6) at 5 dpf. Time is given corresponding to the Medaka
start of the life imaging observation (0h: onset of imaging). A–C: Upper panels, single images
showing osx-mCherry expression only. Lower panels, overlays of mCherry images with images A 4.1-kb upstream regulatory region of
made by bright field and green fluorescent protein (GFP) filter settings for (A) and brightfield for the medaka osterix gene was amplified
(B,C). GFP filter settings reveal autofluorescence of pigment cells. A1: Expression of mCherry in two
centers (arrowheads). A2, 3: Arrowheads, single cells lateral to the initial expression areas. B1– 6:
using the primers 5⬘-TGAACATGT-
Time-lapse analysis of anterior vertebrae, asterisk marks first neural arch showing mCherry ex- CAGTGCCATCA-3⬘ and 5⬘-CGGGA-
pression. (C1– 6) Appearance of osx-mCherry-positive cells in fin rays of the caudal fin (arrow in C1). CAGTTTGGAAGAAGT-3⬘, and cloned in
248 RENN AND WINKLER
front of mCherry into an I-SceI site
containing vector. Transgenic fish
were generated by injection of circu-
lar plasmid into medaka 1-cell stage
embryos using the I-SceI meganucle-
ase approach (Rembold et al., 2006).
Founder fish were screened for
mCherry fluorescence and selected
for generation of heterozygous F1
offspring by mating to wild type
medaka. A stable transgenic osteo-
calcin-eGFP (osc-eGFP) medaka line
was generated according to Inohaya
et al. (2007). To generate double
transgenic embryos, stable osc-eGFP
carriers were crossed to osx-
mCherry fish.
Bone and Cartilage Staining
For in vivo staining of calcified bone,
chorions of embryos were removed by
treatment with hatching enzyme. De-
chorionated embryos were grown in
medium containing Calcein (0.0001%,
Sigma) until the desired developmen-
tal stage. Double staining for cartilage
(alcian blue) and bone (alizarin red) in
fixed embryos was performed as de-
scribed in Walker and Kimmel (2007).
Fluorescence Imaging in
Living Specimen
Calcification (Calcein) and mCherry
expression in embryos, larvae, and
adult fish was analyzed by fluores-
cence microscopy (Stereoscopic Zoom
Microscope SMZ1000, Nikon Japan)
using GFP and FRITC filter settings,
respectively. Pictures were taken us-
ing the NIS-element BR software (Ni-
kon, Japan). In osx-mCherry trans-
genic fish, GFP filter settings were
used to control for autofluorescence of
pigment cells. A Laser Scanning Mi-
croscope (LSM 510 Meta, Zeiss; He-
lium Neon gas laser, 543 nm; LP 560)
and LSM software (Zeiss) were used
for confocal analyzes of osx-mCherry
expression. Life imaging analysis was
performed using an inverted micro-
scope (Ti-E; Nikon, Japan) equipped
with temperature controlled on-stage
CO2 incubator (INUG2-N1; Tokai Hit,
Japan) at 30°C using a Plan Fluor
10⫻ Objective DLL (N.A. 0.45). Im-
ages were recorded with the appropri-
ate filter sets (Chroma Technology
Corp, Rockingham, VT; for mCherry:
ET Sedat Green Subset, Ex:555/25
nm; Em: 605/52 nm; for GFP: ET Se-
dat Sedat Blue subset, Ex: 490/20 nm;
525/36 nm) on a coolSNAP HQ mono-
chrome charge-coupled device (CCD)
camera (Roper Scientific, Tucson, AZ)
at intervals of 15 min. Life imaging
hardware was controlled by the NIS-
element AR software (Nikon, Japan).
For laser scanning and time-lapse
analyzes, embryos and larvae were
anesthetized with 0.02%– 0.05% Tric-
aine (Sigma) and mounted in 1.5% low
melting point agarose in a glass-bot-
tom petri dish.
ACKNOWLEDGMENTS
We thank Dr. Clement Khaw, Singa-
pore Bioimaging Consortium/Nikon
Imaging Center, Biopolis, for excellent
support using the Live Cell Imaging
System. We thank the NUS/DBS Bio-
imaging facility for support using the
Zeiss Confocal system. We also thank
Stefan Schulte-Merker and Kirsten
Spoorendonk for fruitful discussions
and sharing unpublished results.
REFERENCES
Aubin JE, Liu F. 1996. The osteoblast lin-
eage. In: Bilezikian JP, Raisz LG, Rodan
GA, editors. Principles of bone biology.
San Diego: Academic Press. p 51– 68.
Fisher S, Jagadeeswaran P, Halpern ME.
2003. Radiographic analysis of zebrafish
skeletal defects. Dev Biol 264:64 –76.
Flores MV, Tsang VW, Hu W, Kalev-Zylin-
ska M, Postlethwait J, Crosier P, Crosier
K, Fisher S. 2004. Duplicate zebrafish
runx2 orthologues are expressed in de-
veloping skeletal elements. Gene Expr
Patterns 4:573–581.
Gavaia PJ, Simes DC, Ortiz-Delgado JB,
Viegas CSB, Pinto JP, Kelsh RN,
Sarasquete MC, Cancela ML. 2006. Os-
teocalcin and matrix Gla protein in ze-
brafish (Danio rerio) and Senegal sole
(Solea senegalensis): comparative gene
and protein expression during larval de-
velopment through adulthood. Gene
Expr Patterns 6:637– 652.
Inohaya K, Takano Y, Kudo A. 2007. The
teleost intervertebral region acts as a
growth center of the centrum: in vivo
visualization of osteoblasts and their
progenitors in transgenic fish. Dev Dyn
236:3031–3046.
Iwamatsu T. 2004. Stages of normal devel-
opment in the medaka Oryzias latipes.
Mech Dev 121:605– 618.
Komori T. 2006. Regulation of osteoblast
differentiation by transcription factors.
J Cell Biochem 99:1233–1239.
Langille RM, Hall BK. 1987. Development
of the head skeleton of the Japanese
medaka, Oryzias latipes (Teleostei). J
Morphol 193:135–158.
Lawson ND, Weinstein BM. 2002. In vivo
imaging of embryonic vascular develop-
ment using transgenic zebrafish. Dev
Biol 248:307–318.
Long Q, Meng A, Wang H, Jessen JR, Far-
rell MJ, Lin S. 1997. GATA-1 expression
pattern can be recapitulated in living
transgenic zebrafish using GFP reporter
gene. Development 124:4105– 4111.
Nakashima K, Zhou X, Kunkel G, Zhang Z,
Deng JM, Behringer RR, de Crombrug-
ghe B. 2002. The novel zinc finger-con-
taining transcription factor osterix is re-
quired for osteoblast differentiation and
bone formation. Cell 108:17–29.
Nemoto Y, Higuchi K, Baba O, Kudo A,
Takano Y. 2007. Multinucleate oste-
oclasts in medaka as evidence of active
bone remodeling. Bone 40:399 – 408.
Nishio Y, Dong Y, Paris M, O’Keefe RJ,
Schwarz EM, Drissi H. 2006. Runx2-me-
diated regulation of the zinc finger Os-
terix/Sp7 gene. Gene 372:62–70.
Pinto JP, Conceicao N, Gavaia PJ, Cancela
ML. 2003. Matrix Gla protein gene ex-
pression and protein accumulation colo-
calize with cartilage distribution during
development of the teleost fish Sparus
aurata. Bone 32:201–210.
Rembold M, Lahiri K, Foulkes NS, Witt-
brodt J. 2006. Transgenesis in fish: effi-
cient selection of transgenic fish by co-
injection with a fluorescent reporter
construct. Nat Protoc 1:1133–1139.
Renn J, Schaedel M, Volff JN, Goerlich R,
Schartl M, Winkler C. 2006a. Dynamic
expression of sparc precedes formation of
skeletal elements in the medaka (Ory-
zias latipes). Gene 372:208 –218.
Renn J, Winkler C, Schartl M, Fischer R,
Goerlich R. 2006b. Zebrafish and
medaka as models for bone research in-
cluding implications regarding space-re-
lated issues. Protoplasma 229:209 –214.
Spoorendonk KM, Peterson-Maduro J,
Renn J, Trowe T, Kranenbarg S, Winkler
C, Schulte-Merker S. 2008. Retinoic acid
and Cyp26b1 are critical regulators of
osteogenesis in the axial skeleton. Devel-
opment 135:3765–3774.
Walker MB, Kimmel CB. 2007. A two-color
acid free cartilage and bone stain for ze-
brafish larvae. Biotech Histochem 82:23–
28.
Yan YL, Willoughby J, Liu D, Crump JG,
Wilson C, Miller CT, Singer A, Kimmel
C, Westerfield M, Postlethwait JH. 2005.
A pair of sox: distinct and overlapping
functions of zebrafish sox9 co-orthologs
in craniofacial and pectoral fin develop-
ment. Development 132:1069 –1083.
Yasutake J, Inohaya K, Kudo A. 2004.
Twist functions in vertebral column for-
mation in medaka, Oryzias latipes. Mech
Dev 121:883– 894.