THE JOURNAL OF COMPARATIVE NEUROLOGY 487:428 – 440 (2005)
A Three-Dimensional Atlas of Pituitary
Gland Development in the Zebrafish
SUSAN C. CHAPMAN,1* ARLEEN L. SAWITZKE,1,2 DOUGLAS S. CAMPBELL,3
1
AND GARY C. SCHOENWOLF
1
Department of Neurobiology and Anatomy, and Children’s Health Research Center,
University of Utah School of Medicine, Salt Lake City, Utah 84132-3401
2
Division of Natural Sciences, Salt Lake Community College, Salt Lake City, Utah 84123
3
Department of Neurobiology and Anatomy, University of Utah School of Medicine,
Salt Lake City, Utah, 84132-3401
ABSTRACT
The pituitary gland is unique to Chordates, with significant variation within this group,
offering an excellent opportunity to increase insight into phylogenetic relationships within
this phylum. The structure of the pituitary in adult Teleosts (class: Osteichthyes) is quite
different from that in other chordates and is also variable among members of the class.
Therefore, a complete description of the structure and development of the pituitary in
members of this class is a critical component to our overall understanding of this gland. An
obvious teleost model organism is the zebrafish, Danio rerio, as a significant amount of work
has been done on the molecular control of pituitary development in this fish. However, very
little work has been published on the morphological development of the pituitary in the
zebrafish; the present study aims to fill this void. The pituitary develops from cells on the
rostrodorsal portion of the head and reaches its final position, ventral to the hypothalamus,
as the cephalic flexure occurs and the jaws and mouth form. The pituitary placode is
juxtaposed to cells that will form the olfactory vesicles, the stomodeum, and the hatching
gland. The volume of the pituitary is greatest at 24 hours post fertilization (hpf). From 24 to
120 hpf, the pituitary decreases in height and width as it undergoes convergent extension,
increasing in length with the axis. The adenohypophysis is a morphologically distinct structure by 24 hpf, whereas the neurohypophysis remains indistinct until 72 hpf. The findings of
this study correlate well with the available molecular data. J. Comp. Neurol. 487:428 – 440,
2005. © 2005 Wiley-Liss, Inc.
Indexing terms: adenohypophysis; neurohypophysis; forebrain; morphology; anatomy; stomodeum
The pituitary gland, present in all vertebrates, is a
structure unique to the Chordate phylum (Gorbman,
1995; Norris, 1997). In higher vertebrates, the pituitary,
under the control of the hypothalamus, is referred to as
the master gland of the body because of its varied effects
on growth, metabolism, reproduction, and water balance.
The gland has a dual origin and functions as a bridge
between the nervous/sensory and endocrine systems of the
body. Although anterior and posterior are often used when
discussing the pituitary gland, these terms are not developmentally or functionally consistent. The terms adenohypophysis (AH), consisting of the pars distalis (rostral
and proximal), pars intermedia, and pars tuberalis, and
the neurohypophysis (NH), consisting of the median eminence and pars nervosa, provide more consistency. The
NH arises from neural tissue (ventral diencephalon/
hypothalamus) (Baker and Bronner-Fraser, 2001),
© 2005 WILEY-LISS, INC.
whereas the AH forms from the anterior neural ridge
(ANR) of teleost, amphibian, avian, and mammalian embryos (Couly and Le Douarin, 1985, 1987; Eagleson et al.,
1986; el Amraoui and Dubois, 1993; Eagleson et al., 1995;
Glasgow et al., 1997; Whitlock and Westerfield, 2000;
Grant sponsor: National Institutes of Health; Grant number: DK066445;
Grant sponsor: European Molecular Biology Organisation long-term fellowship (to D.S.C.).
*Correspondence to: Susan C. Chapman, Department of Neurobiology
and Anatomy and Children’s Health Research Center, University of Utah
School of Medicine, 20 N. 1900 East, Room 401 MREB, Salt Lake City,
Utah, 84132-3401. E-mail: [email protected]
Received 6 October 2004; Revised 22 December 2004; Accepted 11 February 2005
DOI 10.1002/cne.20568
Published online in Wiley InterScience (www.interscience.wiley.com).
ZEBRAFISH DEVELOPING PITUITARY ATLAS
429
Kouki et al., 2001; Kawamura et al., 2002). The presence,
function, spatial relationship, and regulation of the structures within the AH and NH vary among the different
classes of chordates (Norris, 1997). These variations presumably reflect the divergent environments encountered
by the chordates after the origin of the pituitary as a
regulatory structure (see Discussion).
The widely used model organism, the zebrafish (Danio
rerio), is an obvious teleost model in which to investigate
AH patterning in both molecular/functional terms and at
the level of morphological development. Understanding of
the molecular control of zebrafish AH development is far
more complete (Herzog et al., 2003, 2004b; Liu et al., 2003;
Sbrogna et al., 2003; Nica et al., 2004) than that of its
morphologic development; thus, the current study was
undertaken to fill the gap in the morphological description
of the developing pituitary in the zebrafish and provide an
atlas for future research.
MATERIALS AND METHODS
Adult wild-type zebrafish were housed and mated at the
University of Utah centralized zebrafish animal resource.
Fertilized eggs were collected, cleaned, and placed into E3
medium (5 mM NaCl, 0.17 mM KCl, 0.33 mM CaCl2 䡠
2H2O, 0.33 mM MgSO4 䡠 7H2O, 0.0002% methylene blue)
within 2 hours of lights on in the facility. The embryos
were then incubated in 100-mm Petri dishes at 28.5°C.
The medium was changed to 1:1 E3:0.2 mM PTU/E3 after
the embryos had reached 70% epiboly. PTU delays pigmentation of melanocytes and thus allows better visualization of desired structures. Embryos were collected and
fixed in 4% paraformaldehyde/phosphate-buffered saline
(PBS) at 18, 24, 36, 48, 72, 96, and 120 hours post fertilization (hpf). Staging was done according to Kimmel et al.
(1995).
Embryos for plastic sectioning were prestained with
0.5% toluidine blue/sodium borate (Sigma, St. Louis, MO),
washed twice with PBS, dehydrated (10 minutes each: 50,
70, 2⫻ 95, 3⫻ 100% ethanol), infiltrated (30 minutes in 1:1
Immunobed:100% ethanol, 3⫻ 1 hour in 100% Immunobed [Polysciences, Warrington, PA, #17324]), and embedded in Immunobed. Serial sections (5 ␮m) were made on a
Leica-Jung RM 2055 microtome using a glass blade. The
sections were stained for 30 seconds with 0.25% toluidine
blue/sodium borate and coverslipped with Pro-Texx
mounting medium (Lerner Laboratories, Pittsburgh, PA
#137635).
DiI/CRSE (1,1⬘-dioctadecyl-3,3,3⬘-tetramethylindocarbocyanine perchlorate/5-carboxytetramethylrhodamine,
succinimidyl ester) was used to inject the presumptive AH
cells at 18 hpf according to our standard protocol for fate
mapping, followed by immunocytochemistry to develop a
permanent marker that could be identified after plastic sectioning (Lawson and Schoenwolf, 2003; Lopez-Sanchez et al.,
2004).
Photographs were taken with a Qimaging micropublisher 5.0 camera on a Nikon microscope (SMZ1500). For
volume measurements and reconstructions, serial micrographs were taken, and the sections were aligned in Adobe
Photoshop 7.0 and then analyzed as stacks in Image J
(NIH) or as image sequences in Volocity (Improvision,
Lexington, MA). The AH and the NH were selected in
Adobe Photoshop by using the magnetic lasso to follow the
contrast of the cell borders. These selections were then
filled with contrasting color to allow for selection in Volocity software. The selections were also saved as separate
files to allow analysis in Image J. The areas of each section
were determined with the “analyze particle” feature in
Image J. Three-dimensional reconstructions were made
by using Volocity software. In Volocity, the colorized areas
of the AH in the image sequence were selected by using
the magic wand tool. These selected areas were then displayed in three dimensions in the XYZ panel.
The locations of the AH and NH at each of the stages
reported in this paper were determined by using distinguishable morphological borders, by comparison with previously published studies (Waterman and Kao, 1982; Verbeek et al., 2000; Sprague et al., 2001; http://www.zfin.org;
http://bio-imaging.liacs.nl/index.html) and by comparison
Abbreviations
III
IV
a-hy
ah
cc
ce
cf
ch
d
e
ec
en
ep
er
et
ha
hg
ht
hpf
hy
l
m
ma
mc
med
third ventricle
fourth ventricle
anterior hypothalamus
adenohypophysis
cranial cavity
cerebellum
cephalic flexure
ceratohyal cartilage
diencephalon
epithelium
ectoderm
endoderm
epiphysis
epithelial roof of oral cavity
ethmoid plate
hyoidal arch
hatching gland
heart
hours post fertilization
hypothalamus
lens vesicle
mesencephalon
mandibular arch
Meckel’s cartilage
medulla
my
n
nh
o
oc
on
or
os
ot
ov
pa
pc
pcc
p-hy
pi
ppd
r
rpd
sot
st
t
tb, tc
te
tg
ys
myelencephalon
notochord
neurohypophysis
olfactory placode
oral cavity
optic nerve
optic recess
optic stalk
optic tract
otic vesicle
pharyngeal arches
posterior commissure
parachordal cartilage
posterior hypothalamus
pars intermedia
proximal pars distalis
sensory layer of the optic cup (future retina)
rostral pars distalis
superior optic tract
stomodeum (presumptive)
telencephalon
trabecular cartilage
tectum
tegmentum
yolk sac
430
S.C. CHAPMAN ET AL.
TABLE 1. Comparison of the Micrographs in the Current Study with Previously Published Micrographs1
Fig. no. in present paper (hpf)
Herzog et al., 2004b
Sbrogna et al. 2003
1G (20)
2C, G (24)
4M, N (18, 19) [lim3]
2K, L (24)
4O (24) [lim3]
3B,C (36)
3H–J, O (36)
3L, M (36)
5A, B (30) [lim3]
3M, N (36)
4L, M (42, 48)
5A–D (96)
1
Herzog et al. 2003
Nica et al. 2004
Herzog et al. 2004a
3A–C (24–26) [prl,
lim3]
2D (24) [pit1] 3C,
D (24) [lim3] 6A
(24) [pit1, prl]
3H (24) [fgf8] 3J M
(26**) [lim3, fgf3]
3D (26) [lim3] 6A
(22) [shh, lim3]
3H (24) [fgf8]
2F(32) [pit 1]
3H (32) [prl, pomc]
4B, C (32, 40)
[nk2.1, pomc]
1A (32) [lim3]
3O (32) [lim3, fgf3]
3N (32) [lim3, fgf3]
1 (36) [lim3],
morphology 4P (30)
[lim3]
4D, E (40, 45) [pomc]
2A–D (72) [prl, pomc,
gh, tsh]
The hours post fertilization (hpf) for each figure are shown in parentheses. The molecular markers used in the various studies are shown in brackets.
with previously published in situ hybridization of the lim3
and pit1 probes (Glasgow et al., 1997; Sbrogna et al., 2003;
Herzog et al., 2004a, b; Nica et al., 2004). Correlations of
figures in the present paper with previously published
figures are presented in Table 1.
RESULTS
At 18 –20 hpf, cells fated to become the AH become
morphologically distinct. Figure 1 shows four transverse
sections of the head, in rostral to caudal order (Fig. 1A–D).
The levels of these sections are indicated on a whole
mount embryo (Fig. 1E) and a sagittal section (Fig. 1F) for
orientation. Based on the location of lim3 expression in
whole mount embryos at this stage (Glasgow et al., 1997;
Herzog et al., 2003; Sbrogna et al., 2003) and our unpublished data from areas rostral and ventral to the optic
recess, the cells indicated by the arrowheads in Figure 1B,
F, and G give rise to the AH and to the anlagen of the
olfactory placode located lateral to the AH (Fig. 1D). The
fates of this cell population were confirmed by fate mapping. The site of injection of presumptive AH and olfactory
cells with a mixture of DiI/CRSE dye at 18 hpf is shown in
Figure 1H. Figures 2L (inset) and 3O show that the dye
injected at 18 hpf is located in the AH at 24 and 36 hpf.
The dye is also in the adjacent cells of the stomodeum and
the olfactory anlagen at 36 hpf (Fig. 3O and data not
shown). The adjacent localization of the AH and olfactory
placode cells is consistent with published in situ hybridization studies and with the origins of these structures
from the anterior neural ridge (Whitlock and Westerfield,
2000; Herzog et al., 2003; Sbrogna et al., 2003). The presumptive pituitary cells are elongated and aligned, compared with the more rounded, randomly oriented cells of
the dorsally located telencephalon and the more rostroventrally located stomodeal cells (indicated in sagittal
sections in Fig. 1F and G). The stomodeal cells form the
caudal border of the mouth as development continues
(Waterman and Kao, 1982). The AH cells are directly
dorsal to the hypothalamus, which gives rise to the neurohypophysis (Fig. 1B,F). The AH cells are approximately
50 ␮m rostral to the mideye region shown in Figure 1D.
Figure 1C shows the telencephalon and diencephalon separated by the optic recess. At 18 hpf, there is no observable
cephalic flexure, and the most rostral structure is the
hatching gland, which has developed from the polster.
Careful examination of Figure 1F and G confirms that the
hatching gland and the stomodeal cells are separated from
the neural tissue of the head.
The cephalic flexure forms by 20 hpf, and increased cell
proliferation and/or growth occurs in the telencephalon
and diencephalon, including the hypothalamus, leading to
forward rotation of the rostral head onto the yolk and
lifting of the more caudal embryo (compare Figs. 1G and
2K). A layer of cuboidal hypothalamic cells (double arrow
in Fig. 1G) has formed just dorsal to the stomodeal cells,
and the third ventricle is now visible (Fig 1G). As the
angle of flexure increases and growth continues, the
dorsal-ventral relationship between AH and olfactory placode becomes more pronounced. The presumptive AH cells
are still rostral to the cells of the diencephalon and will
remain closely apposed to the diencephalon throughout
development. This juxtaposition of the AH and the diencephalon from the earliest stages of development is unique
to the teleosts.
By 24 hpf the cells of the AH are identifiable as the
pituitary anlage (arrowheads in Fig. 2C,G,L). A distinct
group of cuboidal cells, first seen at 20 hpf (double arrow
in Fig. 1G), now separate the pituitary from the hypothalamus (double arrow in Fig. 2L). The AH cells are loosely
grouped together and occupy a volume of 25,224 ␮m3
(Table 2).The pituitary anlage is now immediately rostroventral to the hypothalamus and immediately caudoventral to the stomodeal cells (Fig. 2C,D,F,G). The olfactory placode (Fig. 2B) and the epiphysis (Fig. 2I) are
both identifiable. The increased angle of the cephalic flex-
Fig. 1. Transverse sections (A–D, dorsal up), lateral view (E, dorsal up, anterior to left, low-magnification orientation figure), and
midsagittal sections (F–H dorsal up, anterior to left) of 18 hpf (A–F,H)
and 20 hpf (G) embryos. H is an embryo immediately after injection of
DiI/CRSE into the adenohypophyseal placode and adjacent neural
and epidermal cells. Lines in E and F indicate the orientation and
position of the transverse sections. Arrowheads (B,F,G) indicate the
site of the presumptive adenohypophysis outlined by white dots. Double arrow in G indicates cuboidal cells of the hypothalamus. Numbers
in upper right corners (A–D) indicate the distance of each section from
the starting point of that series. Scale bars ⫽ 100 ␮m in A–D; 200 ␮m
in E; 80 ␮m in F–H.
Figure 1
432
S.C. CHAPMAN ET AL.
Fig. 2. Frontal sections (A–E, anterior to left), transverse sections
(F–I, dorsal up), lateral view (J, dorsal up, anterior to left, lowmagnification orientation figure), and midsagittal sections (K,L, dorsal up, anterior to left) of 24 hpf embryos. L is an enlargement of K.
M is a 3-D reconstruction of the adenohypophyseal area (white),
anterior coming out of the plane of the paper. Inset in L is of an
embryo 6 hours after injection of DiI-rhodamine into the adenohypophyseal placode at 18 hpf. Arrowheads (C,G,K,L,L inset) identify
the adenohypophysis, with white dots defining the outline in C, K, L,
and L inset and black dots in G. Lines in J and K indicate the
orientation and position of the transverse and frontal sections. Double
arrows in L indicate cuboidal cells of the hypothalamus. Numbers in
upper right corners (A–E,F–I) indicate the distance of each section
from the starting point of that series. Arrow (I) probably indicates the
saccus vasculosus. K shows the relationship of the third ventricle to
the optic recess. *, anterior commissure; **, postoptic commissure. For
abbreviations, see list. Scale bars ⫽ 100 ␮m in A–C,F–I,M; 50 ␮m in
D,E; 60 ␮m in K; 50 ␮m in L, inset L.
ure and the continued growth of the forebrain have
pushed the dorsal portion of the head forward while the
ventral hypothalamus has remained attached to the yolk
sac. Thus, morphological movement, rather than active
migration of the AH cells, has brought the cells of the
pituitary into their present position between the hypothalamus and the stomodeal cells (Fig. 2L). The cells of the
hatching gland have migrated away from the head to the
periphery of the yolk sac (Fig. 2J). (Compare with hgg1
expression at http://cegs.standord.edu; CEGS ID number
466.) At this stage, the dorsal/ventral height and the
rostral/caudal length of the pituitary anlage are approximately equal, whereas the left/right width is approximately 3.5 times that of the height/length (Table 2). The
ZEBRAFISH DEVELOPING PITUITARY ATLAS
Fig. 3. Frontal sections (A–F, anterior to left), transverse sections
(G–J,O, dorsal up), lateral view (K, dorsal up, anterior to left, lowmagnification orientation figure), and sagittal sections (L,M, dorsal
up, anterior to left) of 36 hpf embryos. M is an enlargement of L. N is
a 3-D reconstruction of the adenohypophyseal area (white), anterior
coming out of the plane of the paper. F–H are from a single embryo, I
is from a second, and J is from a third embryo. O is an embryo 18
hours after injection of DiI/CRSE into the adenohypophyseal placode
433
and adjacent neural and epidermal cells at 18 hpf. Arrowheads
(B,C,H–J,M,O) identify the adenohypophysis, with white dots defining the outline. Lines in L indicate the orientation and position of the
first transverse and frontal sections of each series. Numbers in upper
right corners (A–F,G,H) indicate the distance of each section from the
starting point of that series. *, anterior commissure; **, postoptic
commissure. Fpr abbreviations, see list. Scale bars ⫽ 100 ␮m in
A,B,G,H,L,N,O (bar in B also applies to C–F); 10 ␮m in I,J,M.
434
S.C. CHAPMAN ET AL.
TABLE 2. Volumes and Dimensions of the Adenohypophysis at 24, 36, 48,
96, and 120 Hours Post Fertilization (hpf)
Age (hpf)
Volume
(␮m3)
Rostralcaudal
length (␮m)
Average
(maximum)
width (␮m)
Average
(maximum)
height (␮m)
24
36
48
96
120
25,224
16,932
17,941
20,449
19,956
24
26
40
45
65
68 (86)
41 (47)
38 (46)
51 (63)
41 (63)
17 (20)
16 (19)
13 (19)
10 (13)
8.6 (12)
pituitary anlage is wider than the hypothalamus (Fig.
2G). Ventral extension of the third ventricle (arrow in Fig.
2I) may represent the infundibulum homolog, a structure
known in teleosts as the saccus vasculosus. The saccus
vasculosus has been identified in several teleost species;
however, a true infundibulum has not (Norris, 1997). The
postoptic commissure (POC; double asterisk in Fig. 2K)
and the pituitary are in a dorsal-ventral relationship separated by approximately 15 ␮m.
The AH cells are more compactly arranged at 36 hpf,
reducing pituitary volume by almost 30%, with convergent
extension occurring (Table 2). Interestingly, the height of
the rostral portion of the AH is almost double that of the
caudal portion (Fig. 3M). Notice that the width of the
pituitary is now slightly less than the width of the hypothalamus (Fig. 3H). Significant growth has occurred in the
forebrain and the angle of cephalic flexure has continued
to increase, moving the hypothalamus directly dorsal to
the stationary pituitary (Fig. 3B,C,M). The cells of the
hypothalamus do not have the cuboidal organization that
they had at 20 and 24 hpf, nor are they as tightly apposed
to the yolk sac (compare Fig. 2K,L with Fig. 3L,M). Because of growth and movement, the front of the head now
has a rounded rather than a pointed appearance (compare
Fig. 2J with Fig. 3K), with the indentation between the
head and yolk sac indicating the formation of the stomodeum. This stomodeal cavity will enlarge and ultimately
fuse with the foregut to form the oral cavity. Immediately
rostral to the pituitary, the cells of the stomodeum that
will eventually form the oral membrane can be identified
(Fig. 3D,M). These cells are marked by the DiI/CRSE that
was injected into the AH/olfactory placode at 18 hpf (Fig.
3O).
The olfactory cells are also marked with DiI/CRSE at 36
hpf (data not shown). The mandibular arch has now
formed (Fig. 3E,J). Thus, the beginnings of the oral cavity
have just become distinguishable even though the pituitary has been an identifiable structure, both morphologically and functionally, for several hours (Herzog et al.,
2003; Liu et al., 2003; Sbrogna et al., 2003; Herzog et al.,
2004b; Nica et al., 2004). In some embryos at this stage,
the pericardium and heart are located ventral and slightly
caudal to the pituitary (Fig. 3E,I,J,L). In other embryos of
this stage, variation in development occurs, with the epithelial ectoderm of the head still resting upon the yolk sac
and the heart in a slightly more caudal position (Fig. 3H).
In either case, the rudiments of the mouth and jaw are in
close proximity to the pituitary (Fig. 3C–E,J). The optic
nerve is approximately 20 ␮m rostral to the pituitary
(data not shown). The notochord is caudal to the mandibular arch (Fig. 3F) and approximately 20 ␮m caudal to the
pituitary. The distance between the POC and the pituitary
is still approximately 15 ␮m (Fig. 3L).
Lifting of the head from the yolk sac, formation of the
mouth, and growth of the jaws are some of the major
events that occur between 36 and 60 hpf. During this time,
the head-to-tail angle (HTA) decreases from 75 to 35 degrees (Sprague et al., 2001), resulting in significant
changes in the positional relationships of numerous structures (compare Fig. 4F–H with Fig. 4I,J). These changes
ultimately position the AH dorsal to the roof of the oral
cavity and ventral to the hypothalamus. As expected from
previous studies on pituitary development in teleosts
(Norris, 1997), no structure similar to Rathke’s pouch can
be identified during this or any other time period. The
volume of the AH changes only slightly as its rostralcaudal length increases and its dorsal-ventral height decreases (Table 2).
At 42 hpf, cells of the stomodeum are directly rostral to
the AH (Fig. 4L), whereas by 48 hours, stomodeal cells
have proliferated and formed a layer ventral to the rostral
AH (Fig. 4D,M). The cells of the mandibular arch are
visible in Figure 4B, F–I, and M. By 48 hpf, the hypothalamus is directly caudal to the tract of the POC (Fig.
4B,F,G,L) and is easily distinguished from the diencephalon in transverse sections (Fig. 4I). A body of loosely
grouped cells are now positioned between the AH and the
hypothalamus (arrow in Fig. 4N). Cells of the AH are still
a fairly compact mass with no clear morphological differences among them. The height of the AH is greatest in the
middle, with tapering to the cephalic and caudal ends (Fig.
4N). The distance from the tract of the POC and optic
nerve to the AH has increased to approximately 20 ␮m
(Fig. 4F–H). The pituitary is 15 ␮m rostral to the notochord at 48 hpf.
By 55 hpf the pharyngeal region has undergone significant patterning, with the cells of the AH remaining stationary, immediately ventral to the hypothalamus. The
notochord is now only 10 ␮m caudal to the AH, the heart
is ventral to the mandible, and the remains of the stomodeal cells are 90 ␮m rostral to the AH. Interestingly, there
is a layer of cells running from the rostral edge of the AH
to the remaining stomodeal cells (double arrow in Fig. 4O)
that could be the buccohypophyseal canal described in
other bony fish (Norris, 1997). The hypothalamus is now
subdivided into anterior and posterior portions (Fig. 4O).
From 55 to 96 hpf, the pharyngeal region continues to
mature, and the distance between the optic nerve/rostral
tract of the POC and the pituitary increases greatly (95
␮m). In frontal sections, the caudal extension of the tract
of the POC is just rostral to the pituitary (Fig. 5A). During
this maturation, the pituitary maintains its position directly ventral to the hypothalamus (Fig. 5K,M). Cartilage
of the ethmoid plate (Fig. 5A), trabeculae (Fig. 5C,F,M),
and branchial arches (Fig. 5F,M) has developed by 96 hpf.
The notochord is now directly (5 ␮m) caudal to the pituitary (Fig. 5J). The heart is now caudal to, rather than
only ventral to, the pituitary (compare Figs. 4M with 5K).
By 96 hpf the posterior hypothalamus is distinctive, and
the NH is now identifiable (Fig. 5B,C,I,N). The cells of the
AH are positioned ventrally and laterally to the NH (Fig.
5H). The cells of the AH also extend more rostrally than
those of the NH (Fig. 5C,D). A separation between the
pituitary and the epithelium of the oral cavity can be seen
in Figure 5I and N (double arrowheads). At 96 hpf, the
pituitary is much narrower than the anterior hypothalamus (Fig. 5G,H), whereas the posterior hypothalamus and
pituitary are of approximately equal width (Fig. 5I). A
ZEBRAFISH DEVELOPING PITUITARY ATLAS
Fig. 4. Frontal sections (A–D, anterior to left), a 3-D reconstruction (E, anterior out of the paper) of the adenohypophyseal area
(white), transverse sections (F–J, dorsal up), lateral view (K, dorsal
up, anterior to left, low-magnification orientation figure), and sagittal
sections (L–O, dorsal up, anterior to left) of 42 (L), 48 (A–K,M), and 55
(N,O) hpf larvae. N is an enlargement of O. Arrow (N) indicates
mesenchyme between the hypothalamus and adenohypophysis. Arrowheads (C,D,H,J,L–O) identify the adenohypophysis, with white
dots defining the outline. Double arrow (O) indicates possible bucco-
435
hypophyseal canal. Lines in K indicate the orientation and position of
the first transverse (lines labeled F and I indicate sections F and I,
respectively) and frontal (line labeled A) sections of each series. Numbers in upper right corners (A–D,F–H,I–J) indicate the distance of
each section from the starting point of that series. Section in M is
parasagittal, and L, N, and O are midsagittal. *, tract of the anterior
commissure; **, tract of the postoptic commissure. For abbreviations,
see list. Scale bars ⫽ 100 ␮m in A,B,E–K,M,O; 10 ␮m in C,D,L,N.
436
Fig. 5. Frontal sections (A–E, anterior to left), transverse sections
(F–I, dorsal up), a 3-D reconstruction (J, anterior out of the paper) of
the adenohypophyseal area (white), and sagittal sections (K–N, dorsal
up, anterior to left) of 72 hpf (K,L) and 96 hpf (A–J,M,N) larvae. L and
N are enlargements of K and M, respectively. Arrowheads (D,G,K–M)
identify the adenohypophysis, with white dots defining the outline
(D,G-I,K–N).White dashes (C,H,I,N) outline the neurohypophysis.
Double arrowheads in I and N indicate separation between the ade-
S.C. CHAPMAN ET AL.
nohypophysis and oral cavity. Lines in M indicate the orientation and
position of sections A and F. Numbers in upper right corners (A–
E,F–I) indicate the distance of each section from the starting point of
that series. *, tract of the anterior commissure; **, tract of the postoptic commissure. For abbreviations, see list. Scale bars ⫽ 100 ␮m in
A,B,F,J (C–E same magnification as B; K and M same magnification
as A); 10 ␮m in G (also applies to H,I,L,N).
ZEBRAFISH DEVELOPING PITUITARY ATLAS
very thin layer of cells can still be seen extending rostrally
from the pituitary (Fig. 5M).
The morphologies of the 120 hpf pituitary and hypothalamus are very adult-like. The rpd, ppd, and pi are indicated based on their position in adult fish. These locations
were originally determined by immunocytochemical experiments labeling the hormones secreted by terminally
differentiated cells (Peter, 1990, Norris, 1997). Many of
the features seen at 96 hpf are more defined by 120 hpf.
The medial region of the AH has flattened (Fig. 6J,N);
however, the lateral edges of the AH continue to surround
the NH (Fig. 6B,C,J) for approximately 20 ␮m. The threedimensional reconstruction of the pituitary region (Fig.
6O) only presents the most posterior section of the NH so
as not to obscure the three-dimensional view of the AH.
The notochord is directly caudal to but is not in contact
with, the pituitary (Fig. 6B,M). The AH continues to rest
on a layer of epithelial cells (Fig. 6F); however, a cavity
now separates the brain from the roof of the oral cavity
(labeled as cc in Fig. 6I–L,M). This cavity has been labeled
the cranial cavity in the atlas at the zebrafish information
network (http://www.zfin.org), but the significance of this
cavity is not discussed. This cavity may be the rostral
extension of the swim bladder, a structure unique to the
subclass Actinopterygii. Cells of this swim bladder, or
other nearby structures, may be important in detecting
osmolarity changes in the water. These osmolarity
changes result in secretion of oxytocin from the NH. By
120 hpf, the AH is subdivided into the rostral pars distalis,
the proximal pars distalis, and the pars intermedia (Fig.
6N). The NH can also be identified as the ventral extension of the hypothalamus in Figure 6C, J, and N.
Quantification of the size of the pituitary at each time
period is summarized in Table 2. Note that the volume of
the AH is greatest at 24 hpf. At this stage, the AH is not
only wider but also taller than it will be by 120 hpf. The
volume of the AH changes little from 24 to 120 hpf; however, the length almost doubles, whereas the height is
decreased by half with convergent extension.
DISCUSSION
The physiology, morphology, and molecular development of the pituitary gland have been well studied in
amphibians (Kawamura and Kikuyama, 1992; Kawamura
et al., 2002), birds (Couly and Le Douarin, 1985, 1988),
mammals (Treier and Rosenfeld, 1996; Scully and Rosenfeld, 2002), and, most recently, the lamprey (class Agnatha) (Wright, 1983; Uchida et al., 2003). However, documentation of pituitary development in the class
Osteichthyes is much more limited. The class Osteichthyes, which originated in fresh water, includes two subclasses Actinopterygii (spiny-finned) and Sarcopterygii
(lobe-finned). These two subclasses are differentiated by
the fate of a pouch off the gut. In Actinopterygii, this
pouch becomes the swim bladder, whereas in sarcopterygii, this pouch becomes an accessory breathing organ.
These two subclasses also differ in the structure of their
pituitary glands. The AH of the Sarcopterygii is more
similar to that of tetrapods than is that of the Actinopterygii (Norris, 1997).
Comprehensive knowledge of pituitary development in
the teleosts (class Osteichthyes; subclass Actinopterygii)
is critical because of the early evolutionary position and
great diversity of this infraclass. Teleosts do not form a
437
Rathke’s pouch, a hypothalamic-hypophyseal portal system, or an infundibulum. The pars intermedia and pars
nervosa are very closely associated in most fish, including
the teleosts, forming the neurointermediate lobe. Many
fish have an extension from the floor of the hypothalamus
known as the saccus vasculosus (Norris, 1997). The pituitary structures of various vertebrates are compared in
Table 3 (Norris, 1997).
Gorbman suggested, and immunocytochemical and histological studies support, the assertion that the vertebrate
pituitary evolved from a chemoreceptive olfactory structure. In amphioxus (subphylum cephalochordata),
Hatschek’s pit is thought to be functionally homologous to
the vertebrate pituitary (Gorbman, 1995; Gorbman et al.,
1999; Kawamura et al., 2002). Pituitary-like cells have
also been demonstrated in the protochordates (Powell et
al., 1996; Terakado et al., 1997; Candiani and Pestarino,
1998). The pituitary-like cells in cephalochordates and
protochordates are directly exposed to the external environment via the oral cavity and possibly regulate reproductive cycles in these animals.
Osmoregulation and reproductive regulation were also
early functions of the vertebrate pituitary; thus, the relationship between the external environment, through the
olfactory or oral cavity cells, and this gland was maintained (Peter et al., 1990; Norris, 1997). Indeed, the medial portion of the anterior neural ridge (ANR) is fated to
become the AH placode, whereas the lateral ANR develops
into the olfactory placode (Begbie and Graham, 2001).
Interestingly, the lens placode also develops from cells at
the ANR. In hedgehog mutant backgrounds, AH cells will
trans-differentiate into a lens. This spatial relationship is
conserved in all vertebrates studied (Kawamura et al.,
2002; Whitlock et al., 2003). In some teleosts, a duct connecting the pituitary to the mouth is lined with cells that
secrete prolactin in response to changes in osmolarity
(Grau and Helms, 1990; Olsson, 1990). Perhaps Rathke’s
pouch, which gives rise to the AH in amphibians, birds,
and mammals and develops from the roof of the oral
cavity, is an evolutionary remnant of the early connection
of the pituitary to the external environment (Jacobson et
al., 1979; Kimmel et al., 1995; Dubois et al., 1997).
Identification of the AH in zebrafish, based on morphological appearance, fate mapping, and in situ hybridization expression patterns (this study and Glasgow et al.,
1997; Herzog et al., 2003, Herzog et al., 2004b; Sbrogna et
al., 2003; Nica et al., 2004), allows for identification of a
morphologically separate group of AH cells by 18 –20 hpf.
Due to the speed of zebrafish development, this is comparatively earlier than when the AH cells can be morphologically identified in other vertebrates. Figure 1G (20 hpf) is
directly comparable to Figure 4J and N of Sbrogna et al.
(2003). At 19 –20 hpf, the lim3- and nk2.2-labeled cells
correspond perfectly with the cells indicated in Figure 1G
as presumptive AH cells. These figures from Sbrogna et al.
(2003) and others, as well as morphological borders, were
used to designate AH cells, hypothalamic cells, and stomodeal cells. (Glasgow et al., 1997; Herzog et al., 2003,
2004b; Sbrogna et al., 2003; Nica et al., 2004). Changes in
the shape of the pituitary from 24 to 36 hpf due to convergent extension correlate well with the change in in situ
expression patterns seen by Herzog and others (Herzog et
al., 2004a; Nica et al., 2004). The figures in this paper
correlate well with those in Herzog et al. at 24 and at 36
hpf. At 25–26 hpf, Herzog et al. demonstrate lateral strips
Figure 6
ZEBRAFISH DEVELOPING PITUITARY ATLAS
439
TABLE 3. Comparison of Pituitary Gland Structures in Various Organisms of the Chordate Phylum1
Classification
Oral pouch
NI
Olfactory hypophyseal connection
in larvae (Gorbman, 1995)
?
Pharyngeal lobe
Buccohypophyseal canal
Hypophyseal cavity
Transient cleft during development
Yes
?
Yes
Yes
Yes
Yes
Yes
Class Osteichthyes: Actinopterygii: Teleostei (salmon, trout)
No
Yes
Yes
Yes
Yes
Yes
Yes, but some axons penetrate pars
distalis
No, direct innervation of pars distalis
(Peter et al., 1990)
Class
Class
Class
Class
Class
?
Rathke’s
Rathke’s
Rathke’s
Rathke’s
Class Agnatha Petromyzontidae (lamprey)
Class
Class
Class
Class
Class
1
Chondrichthyes Selachii (sharks, rays)
Chondrichthyes Holocephali (ratfish)
Osteichthyes: Actinopterygii: Polypteri
Osteichthyes: Actinopterygii: Chondrostei (sturgeon)
Osteichthyes: Actinopterygii: Holostei (gars, bowfin)
Osteichthyes: Sarcopterygii: Dipnoi (lungfish)
Amphibia
Reptilia
Aves
Mammalia
pouch
pouch
pouch
pouch
No
No
No
No
No
Portal system
Yes, but NI under direct neural control
Yes, on surface rather than complex bed
Yes
Summarized from Norris (1997) and Gorbman (1995). NI, neurointermediate lobe.
of pit1/lim3-expressing cells extending posteriorly from
the pituitary anlage that was seen at 24 hpf. Unfortunately, we do not have sections at 25 or 26 hpf to determine whether these lateral strips could be morphologically distinguished. Our results demonstrate that length
doubles and width and height decrease between 24 and
120 hpf.
The criteria used to determine that cells rostral to the
AH anlage will form the stomodeum come from electron
micrographs in Waterman (1982) and extrapolation of Figure 3M and O. Waterman and Kao’s (1982) description of
mouth formation in zebrafish correlates well with data
presented here and serves as an excellent reference point
when studying the morphology of the pituitary. At 36 hpf,
Waterman and Kao identified the most rostral cells as
stomodeal cells but did not differentiate the two types of
cells actually present in this region. With in situ hybridization patterns, stomodeal and AH cell populations can
be easily differentiated. The current anatomical atlases
for zebrafish development, one at the zebrafish information network (http://www.zfin.org) (Sprague et al., 2001)
and another produced by the Leiden Institute of Advanced
Computer Science (LIACS; http://bio-imaging.liacs.nl/index.html; (Verbeek et al., 2000), are helpful for general
orientation; however, these atlases do not provide sufficient detail to understand fully the development of the
AH. The Neuroanatomy of the Zebrafish Brain: a Topological Atlas is an excellent resource for adult structures
(Wullimann et al., 1996).
A significant difference in AH development between the
teleosts and other chordates is the lack of an identifiable
Rathke’s pouch in teleosts. Although the formation of
Rathke’s pouch in amphibians, birds, and mammals is
often described as an evagination (Kawamura et al., 2002)
or invagination, the pouch actually forms by a folding
caused by cephalic flexure of the mesencephalon and enlargement of the prosencephalon (Jacobson et al., 1979).
The data presented here demonstrate that cephalic flexure occurs caudal to, rather than at, the central point of
the zebrafish pituitary anlage. No pouch is formed in
teleosts because of the more caudal positioning of the
cephalic flexure. Due to the folding and subsequent bending of the pituitary anlage in higher vertebrates, the cells
that were originally in an anterior-posterior orientation
are morphed into a dorsal-ventral arrangement. The cells
of the zebrafish pituitary remain in the original anteriorposterior orientation. This observation correlates very
well with our current understanding of differentiation of
cells in the zebrafish AH compared with that in amphibians, birds, and mammals (Herzog et al., 2003, 2004b;
Sbrogna et al., 2003).
This study has provided a three-dimensional description of Danio AH development from 18 to 120 hpf. The
morphological changes documented herein correlate very
well with our current understanding of pituitary development, and the micrographs of this atlas will be indispensable as we continue to probe the molecular development
and differentiation of the AH. The relationship of the AH
anlage to the position of cephalic flexure is particular
enlightening in terms of understanding the anterior/
posterior arrangement of the teleost AH as opposed to the
dorsal/ventral arrangement seen in other vertebrates.
ACKNOWLEDGMENTS
Fig. 6. Frontal sections (A–F anterior to left), transverse sections
(G–L, dorsal up), sagittal sections (M,N, dorsal up, anterior to left),
and a 3-D reconstruction (O, anterior out of the paper) of the adenohypophyseal (white) and neurohypophyseal (black) areas of 120 hpf
larvae. Only the most posterior section of the neurohypophysis is
shown in the 3-D reconstruction. N is an enlargement of M. Arrowhead identifies the adenohypophysis in M. Adenohypophysis is outlined with small black dots in B–E,H–K,N, and neurohypophysis is
outlined with black dashes in B–D,I–K,N. Lines in M indicate the
orientation and position of sections A and G. Numbers in upper right
corners (A–F,G–L) indicate the distance of each section from the
starting point of that series. *, anterior commissure; **, postoptic
commissure. For abbreviations, see list. Scale bars ⫽ 100 ␮m in A,B
(bar in A also applies to M; bar in B also applies to C–F,N); 10 ␮m in
H (also applies to I–L). G ⫽ 10 ␮m, 0 ⫽ 100 ␮m.
We thank Dr. Tatjana Piotrowski for supplying wildtype zebrafish embryos. We also acknowledge the staff at
the University of Utah Core Imaging and Electron Microscopy Facilities for their assistance. We thank Dr. Steve
Wilson for support in initiating this study.
LITERATURE CITED
Baker CV, Bronner-Fraser M. 2001. Vertebrate cranial placodes I. Embryonic induction. Dev Biol 232:1– 61.
Begbie J, Graham A. 2001. The ectodermal placodes: a dysfunctional
family. Philos Trans R Soc Lond B Biol Sci 356:1655–1660.
Candiani S, Pestarino M. 1998. Expression of the tissue-specific transcrip-
440
tion factor Pit-1 in the lancelet, Branchiostoma lanceolatum. J Comp
Neurol 392:343–351.
Couly GF, Le Douarin NM. 1985. Mapping of the early neural primordium
in quail-chick chimeras. Dev Biol 110:422– 439.
Couly GF, Le Douarin NM. 1987. Mapping of the early neural primordium
in quail-chick chimeras. Dev Biol 120:198 –214.
Couly GF, Le Douarin NM. 1988. The fate map of the cephalic neural
primordium at the presomitic to the 3-somite stage in the avian embryo. Development 103(suppl):101–113.
Dubois PM, el Amraoui A, Heritier AG. 1997. Development and differentiation of pituitary cells. Microsc Res Tech 39:98 –113.
Eagleson GW, Jenks BG, Van Overbeeke AP. 1986. The pituitary adrenocorticotropes originate from neural ridge tissue in Xenopus laevis. J
Embryol Exp Morphol 95:1–14.
Eagleson G, Ferreiro B, Harris WA. 1995. Fate of the anterior neural ridge
and the morphogenesis of the Xenopus forebrain. J Neurobiol 28:146 –
158.
el Amraoui A, Dubois PM. 1993. Experimental evidence for an early commitment of gonadotropin-releasing hormone neurons, with special regard to their origin from the ectoderm of nasal cavity presumptive
territory. Neuroendocrinology 57:991–1002.
Glasgow E, Karavanov AA, Dawid IB. 1997. Neuronal and neuroendocrine
expression of lim3, a LIM class homeobox gene, is altered in mutant
zebrafish with axial signaling defects. Dev Biol 192:405– 419.
Gorbman A. 1995. Olfactory origins and evolution of the brain-pituitary
endocrine system: facts and speculation. Gen Comp Endocrinol 97:171–
178.
Gorbman A, Nozaki M, Kubokawa K. 1999. A brain-Hatschek’s pit connection in amphioxus. Gen Comp Endocrinol 113:251–254.
Grau EG, Helms LM. 1990. The tilapia prolactin cell—twenty-five years of
investigation. Prog Clin Biol Res 342:534 –540.
Herzog W, Zeng X, Lele Z, Sonntag C, Ting JW, Chang CY, Hammerschmidt M. 2003. Adenohypophysis formation in the zebrafish and its
dependence on sonic hedgehog. Dev Biol 254:36 – 49.
Herzog W, Sonntag C, von der Hardt S, Roehl HH, Varga ZM, Hammerschmidt M. 2004a. Fgf3 signaling from the ventral diencephalon is
required for early specification and subsequent survival of the zebrafish adenohypophysis. Development 131:3681–3692.
Herzog W, Sonntag C, Walderich B, Odenthal J, Maischein HM, Hammerschmidt M. 2004b. Genetic analysis of adenohypophysis formation in
zebrafish. Mol Endocrinol.
Jacobson AG, Miyamoto DM, Mai S-H. 1979. Rathke’s pouch morphogenesis in the chick embryo. J Exp Zool 207:351–366.
Kawamura K, Kikuyama S. 1992. Evidence that hypophysis and hypothalamus constitute a single entity from the primary stage of histogenesis.
Development 115:1–9.
Kawamura K, Kouki T, Kawahara G, Kikuyama S. 2002. Hypophyseal
development in vertebrates from amphibians to mammals. Gen Comp
Endocrinol 126:130 –135.
Kimmel CB, Ballard WW, Kimmel SR, Ullmann B, Schilling T. 1995.
Stages of embryonic development of the zebrafish. Dev Dyn 203:253–
310.
Kouki T, Imai H, Aoto K, Eto K, Shioda S, Kawamura K, Kikuyama S.
2001. Developmental origin of the rat adenohypophysis prior to the
formation of Rathke’s pouch. Development 128:959 –963.
S.C. CHAPMAN ET AL.
Lawson A, Schoenwolf GC. 2003. Epiblast and primitive-streak origins of
the endoderm in the gastrulating chick embryo. Development 130:
3491–3501.
Liu NA, Huang H, Yang Z, Herzog W, Hammerschmidt M, Lin S, Melmed
S. 2003. Pituitary corticotroph ontogeny and regulation in transgenic
zebrafish. Mol Endocrinol 17:959 –966.
Lopez-Sanchez C, Garcia-Martinez V, Lawson A, Chapman SC, Schoenwolf
GC. 2004. Rapid triple-labeling method combining in situ hybridization
and double immunocytochemistry. Dev Dyn 230:309 –315.
Nica G, Herzog W, Sonntag C, Hammerschmidt M. 2004. Zebrafish pit1
mutants lack three pituitary cell types and develop severe dwarfism.
Mol Endocrinol 18:1196 –1209.
Norris DO. 1997. Vertebrate endocrinology. San Diego: Acedemic Press.
Olsson R. 1990. Evolution of chordate endocrine organs. Prog Clin Biol Res
342:272–281.
Peter RE, Yu K-L, Marchant TA, Rosenblum PM. 1990. Direct neural
regulation of the teleost adenohypophysis. J Exp Zool Suppl 4:84 – 89.
Powell JF, Reska-Skinner SM, Prakash MO, Fischer WH, Park M, Rivier
JE, Craig AG, Mackie GO, Sherwood NM. 1996. Two new forms of
gonadotropin-releasing hormone in a protochordate and the evolutionary implications. Proc Natl Acad Sci U S A 93:10461–10464.
Sbrogna JL, Barresi MJ, Karlstrom RO. 2003. Multiple roles for Hedgehog
signaling in zebrafish pituitary development. Dev Biol 254:19 –35.
Scully KM, Rosenfeld MG. 2002. Pituitary development: regulatory codes
in mammalian organogenesis. Science 295:2231–2235.
Sprague J, Doerry E, Douglas S, Westerfield M. 2001. The Zebrafish
Information Network (ZFIN): a resource for genetic, genomic and developmental research. Nucleic Acids Res 29:87–90.
Terakado K, Ogawa M, Inoue K, Yamamoto K, Kikuyama S. 1997.
Prolactin-like immunoreactivity in the granules of neural complex cells
in the ascidian Halocynthia roretzi. Cell Tissue Res 289:63–71.
Treier M, Rosenfeld MG. 1996. The hypothalamic-pituitary axis: codevelopment of two organs. Curr Opin Cell Biol 8:833– 843.
Uchida K, Murakami Y, Kuraku S, Hirano S, Kuratani S. 2003. Development of the adenohypophysis in the lamprey: evolution of epigenetic
patterning programs in organogenesis. J Exp Zool Part B Mol Dev Evol
300:32– 47.
Verbeek FJ, den Broeder MJ, Boon PJ, Buitendijk B, Doerry E, van Raaij
EJ, Zivkovic D. 2000. A standard atlas of zebrafish embryonic development for projection of experimental data. Proc SPIE 3964:242–252.
Waterman RE, Kao R. 1982. Formation of the mouth opening in the
zebrafish embryo. Scanning Electron Microsc 3:1249 –1257.
Whitlock KE, Westerfield M. 2000. The olfactory placodes of the zebrafish
form by convergence of cellular fields at the edge of the neural plate.
Development 127:3645–3653.
Whitlock KE, Wolf CD, Boyce ML. 2003. Gonadotropin-releasing hormone
(GnRH) cells arise from cranial neural crest and adenohypophyseal
regions of the neural plate in the zebrafish, Danio rerio. Dev Biol
257:140 –152.
Wright GM. 1983. Ultrastructure of the adenohypophysis in the larval
anadromous sea lamprey, Petromyzon marinus L. J Morphol 176:325–
339.
Wullimann MF, Rupp B, Reichert H. 1996. Neuroanatomy of the zebrafish
brain: a topological atlas. Boston: Birkauser.