Amphibian Hatching Gland Cells

8/12/2019 Amphibian Hatching Gland Cells

1/16

Tissue and Cell 39 (2007) 225240

Amphibian hatching gland cells: Pattern and distribution in anurans

M. Nokhbatolfoghahai a,, J.R. Downie b

a Biology Department, Faculty of Sciences, Shiraz University, Shiraz, Iranb Division of Environmental and Evolutionary Biology, Graham Kerr Building, University of Glasgow, Glasgow G12 8QQ, Scotland, UK

Received 25 September 2006; received in revised form 14 April 2007; accepted 26 April 2007

Available online 21 June 2007

Abstract

The hatching gland (HG) is a transient organ, found in most anuran embryos and early larvae, and located on the dorsal side of the head.The enzymes secreted by hatching gland cells (HGCs) aid the embryos to escape from their enveloping coats. Analysis of HG morphology

and distribution in 20 anuran species from six families using scanning electron microscopy revealed small differences in the shape and pattern

of the gland particularly in the length and width of the posterior mid-dorsal extension of the gland. The four species of foam-nest making

leptodactylids examined had HGs of a somewhat different shape to the others, but otherwise, there was little sign of a relationship between

HG shape and taxonomic position. In the single Eleutherodactylusspecies examined, cells with the appearance and location of HGCs were

transiently present long before the active stage of hatching. No sign of HGCs was seen on the head surface of one species, Phyllomedusa

trinitatis. It seems possible that in this species, hatching is achieved by a mechanical rather than an enzymatic mechanism. The microvilli

characteristic of the surfaces of HGCs were quite variablein density and length from species to species, and at different stages. HGCs remained

at the surface of the embryo for some time after hatching and the possibility of a post-hatching function is briefly discussed.

2007 Elsevier Ltd. All rights reserved.

Keywords: Hatching gland cell; Anuran embryos; Microvilli; SEM

1. Introduction

In the oviparous vertebrates, the egg is surrounded by

a set of enveloping structures formed in the ovary and in

the oviduct. In amphibians, the enveloping structures are the

vitelline membrane (formed in the ovary) and a set of jelly

coats (secreted by the oviduct). The jelly coat varies in com-

plexity and thickness, relating to taxonomic group and the

environment into which the eggs are released (Duellman and

Trueb, 1994; Salthe, 1963).According toMartin (1999),the

end of the embryonic phase and the start of larval life maybe defined by the stage of hatching, though other authors

(e.g. Balon, 1984) regard hatching as occurring at such varied

stages in terms of morphological and functional development

Correspondingauthor at: Biology Department, Faculty of Sciences, Shi-

raz University, Shiraz 71454, Iran. Tel.: +98 7112280916;

fax: +98 7112280916.

E-mail addresses:[email protected], [email protected]

(M. Nokhbatolfoghahai).

that it is preferable to define theonset of exogenousfeeding as

the start of the larval phase. Hatching is the process by which

the developing animal frees itself of its enveloping structures

and fully enters the external environment. The stage at when

this happens is generally regardedas specific to any particular

species but in some amphibians, thestage of hatching canvary

and is responsive to environmental factors (Martin, 1999). For

example, terrestrial embryos of the salamander Amphiuma

meanshatch in response to inundation and the stage at which

this happens can vary considerably (Gunzburger, 2003);in

several anuran species, hatching can occur prematurely inresponse to attack by predators or pathogens (Warkentin,

1995; Vonesh, 2005; Touchon et al., 2006).

Whatever the precise timing, hatching in amphibians is

mediated primarily by the activity of a transient popula-

tion of late embryonic epidermal cells, the hatching gland

cells (HGCs). The hatching gland (HG), sometimes known

as the frontal gland, has generally been described as a Y-

shaped array of cells on the head of the embryo, the arms

of the Y pointing anteriorly (Altig and McDiarmid, 1999).

0040-8166/$ see front matter 2007 Elsevier Ltd. All rights reserved.

doi:10.1016/j.tice.2007.04.003
mailto:[email protected]:[email protected]://localhost/var/www/apps/conversion/tmp/scratch_2/dx.doi.org/10.1016/j.tice.2007.04.003http://localhost/var/www/apps/conversion/tmp/scratch_2/dx.doi.org/10.1016/j.tice.2007.04.003mailto:[email protected]:[email protected]


2/16

226 M. Nokhbatolfoghahai, J.R. Downie / Tissue and Cell 39 (2007) 225240

The activity of these cells is the release of a set of pro-

teolytic enzymes whose role is to weaken or breakdown

the enveloping structures (Yoshizaki and Katagiri, 1975;

Yoshizaki, 1991).Fan and Katagiri (2001)identified two dif-

ferent proteolytic molecules (40 and 60 kDa, respectively)

whichactedco-operatively to hydrolyse the embryo envelope

inXenopus laevis.In many amphibians, hatching is a two-phase process. In

most urodeles, the vitelline membrane breaks down at neu-

rulation, but hatching from the remaining enveloping layers

is much later (Duellman and Trueb, 1994).In anurans, three

basic patterns were describedby Duellman and Trueb (1994).

First, early vitelline membrane breakdown followed by later

release from the outer capsules. In the direct-developing

eleutherodactylids, final release is mediated by the egg tooth,

rather than by hatching enzymes. Second, in many aquatic

developing species, the outermost layer of the jelly capsule

ruptures first, probably as a result of differential swelling of

the capsular layers. Vitelline membrane and inner layers are

later brokendowntogetherby hatching gland activity(Carrolland Hedrick, 1974).Third, in others, vitelline membrane and

capsular layers are broken down together with no preliminary

stage.

Detailed descriptions of the hatching process are scarce in

the literature.Bless (1905)account of hatching in X. laevis

indicates that the morphology of the hatching gland may be

important to the process but also that both physical activity

and the cement gland have roles. Bles (1905)noticed that a

mucus thread secreted by the cement gland anchors the head

of the late embryo to a particular point on the inner vitelline

membrane surface. The embryo turns over every 10 min or

so, touching a localised point on the vitelline membrane withits frontal gland (site of HGCs), and presumably releasing the

glands secretion at this point. This leads to a localisedsoften-

ing of the vitelline membrane so that it progressively bulges

outwards under the high fluid pressure within the capsule.

Eventually, the membrane ruptures and the larva is shot out

into thesurroundingwater,remaining attached to thenow col-

lapsed vitelline membrane by its mucus thread.Bles (1905)

confirmed the need for the localisation of the frontal glands

secretion by an experimental manipulation and showed that

a similar process occurred in a Hylaspecies.

Three methods have been employed previously to demon-

strate the distribution of HGCs in amphibian embryos:

immunohistochemistry, histology and scanning electron

microscopy. Drysdale and Elinson (1991) stained whole

mounts ofX. laevis immunohistochemically using an anti-

body to tyrosine hydroxylase. Fan and Katagiri (2001)

employed thesame basic techniquewith an antibody to hatch-

ing enzyme. The only multi-species investigation we can find

of HGC distribution (Meyer et al., 1969)used fixed embry-

onic skin whole mounts with HGCs stained by the periodic

acidSchiff (PAS) method, in order to assessthe HGCpattern

in 40 anuran species.Periodic acidSchiff detects the location

of secretory granules in the apical region of HGCs. Yoshizaki

and Yamamoto (1979) used scanning electron microscopy

(SEM) to follow the appearance and loss of HGCs at the epi-

dermal surface ofRana japonica embryos, identifying HGCs

by their dense array of apical microvilli. Yoshizaki (1991)

made similar observations on HGC appearance inX. laevis

noting particularly changes in the free cell surface area of

HGCs, compared to neighbouring common epidermal cells.

Transmission electron microscopy shows HGCs to beelongated bottle-shaped cells bearing apical microvilli and

containing abundant secretory granules in the apical cyto-

plasm (Fox, 1986).After hatching, the HGCs regress, losing

their position at the epidermal surface, and eventually dis-

appearing, presumably deleted by apoptosis (Yoshizaki and

Yamamoto, 1979).

In the work reported here, we have used SEM to map

the definitive pattern of HGCs in 20 anuran species from

six families. Our aims were to assess whether the pat-

tern of HGCs might have value as a taxonomic character,

and whether the pattern could help in understanding the

mechanism of hatching in the different species, following

Bless (1905)observations demonstrating the importance oflocalised secretion by HGCs.

2. Materials and methods

2.1. Egg collection, incubation, and fixation

All spawn was collected from field sites in Trinidad, Scot-

land and Iran as soon after spawning as was practicable.

Spawn from the many Trinidad species was identified with

the aid ofKenny (1969)andMurphy (1997).The exception

wasX. laevis, provided from a captive population maintainedat the University of St. Andrews.

The 20 species studied are shown inTable 1.Spawn was

incubated at temperatures appropriate to the normal habitat

(Trinidad 2729 C; Scotland 1517 C; Iran 2527 C; X.

laevis, 25 C) either in water (most species) or in air (foam-

nesting species; terrestrial spawners). At appropriate times,

samples were taken for fixation in order to examine hatching

gland morphology. We examined mostGosner (1960)stages

from approximately neurulation (stage 14) to early feed-

ing tadpoles (stage 27). The normal stage at which hatching

occurred in each species was also noted. All specimens were

fixed in 2.5% glutaraldehyde in 0.1 M (pH 7.4) phosphate

buffer for 25 h, then rinsed and stored in the same buffer at

5 C, until further preparation for electron microscopy.

2.2. Specimen staging

The jelly coat and vitelline membrane were removed from

stored pre-hatching specimens using forceps or filter paper.

The specimens were then staged using Gosners (1960) crite-

ria. We use Gosner staging throughout this paper, and when

referring to stages using other systems in the literature have

converted them to Gosner, using the conversion table in

McDiarmid and Altig (1999). The exception is the direct-


3/16

M. Nokhbatolfoghahai, J.R. Downie / Tissue and Cell 39 (2007) 225240 227

Table 1

Classification, country of origin, habitat and spawning place of the 20 species of anurans used in this study

Family Species N Country Habitat Spawning place

Bufonidae Bufo beebeiGallardo 20 Trinidad S Temporary pools

B. bufoLinnaeus 25 United Kingdom WT Pools, ponds

B. marinusLinnaeus 20 Trinidad R, E, S Pools, ponds, rivers

B. viridisLaurenti 30 Iran WT Pools, ponds

Pipidae Xenopus laevisDaudin 30 Africa S Ponds

Hylidae Hyla crepitansa Wied-Neuwied 10 Trinidad E, S Pools, ponds

H. boansa (Linnaeus) 15 R Pools at river edge

H. geographicaa Spix 16 R, E Edge of streams

H. microcephala miserab Fouquette 19 S Temporary pool

H. minutab Peters 10 R, E, S Forest and road-side pools

H. minusculab Rivero 10 S Ponds, swamps

Phyllomedusa trinitatisMertens 20 R, E, S Leaves above water

Phrynohyas venulosac (Laurenti) 30 E, S Temporary pools

Leptodactylidae Eleutherodactylus urichid (Boettger) 10 R, E Leaves on ground

Leptodactylus fuscus(Schneider) 40 S Burrow in ground*

L. validusGarman 15 R, E, S On water surface, hidden*

L. bolivianusBoulenger 10 S On water surface, hidden *

Physalaemus pustulosus(Cope) 30 S On water surface, open*Microhylidae Elachistocleis ovalis(Schneider) 10 S/R Temporary pool, ditches

Ranidae Rana temporariaLinnaeus 30 United Kingdom WT Pools, ponds

Habitat codes: R: rainforest; E: forest-edge; S: Savanna; WT: widespread temperate; (*): in foam. Superscript letters refer to taxonomic changes made byFrost

et al. (2006).N= number of specimens examined.a New genusHypsiboas.b New genusDendropsophus.c New genusTrachycephalus.d New family Brachycephalidae.

developing species, Eleutherodactylus urichi where staging

is byTownsend and Stewart (1985).

2.3. Specimen preparation and examination

Glutaraldehyde-fixed specimens were post-fixed in 1%

osmium tetroxide, stained in 0.5% aqueous uranyl acetate,

dehydrated using an acetone series, then critical-point-dried

and coated with gold using a Polaron SC 515. Post-fixation

and staining is used as a standard procedure in our laboratory

to enhance the conductivity of specimens in SEM. Specimens

were examinedusinga Phillips 500SEM over a magnification

range of 1003200and images recorded by Imageslave for

Windows (Meeco Holdings, Australia). SEM photographs at

low and high magnification were taken from the dorsal side

of the head and trunk with focus on the hatching gland cells.

The stages when the HGCs appeared or disappeared at the

surface were distinguished. The stage when the HGCs were

close together and highest in number was determined. A dis-

tribution map of HGCs on the surface was drawn for each

species. The distribution and length of microvilli on HGC

surfaces were measured from SEM photographs (at magni-

fication between 1200 and 5000). At the each stage for

each species, we examined two specimens when they were

available. We assessed the length of microvilli on the HGC

surfaces, by choosing five of the cells randomly, measuring

two microvilli from each cell. Microvilli density, based on

the ratio of microvillated area of the cell surface to non-

microvillated area was classified into three categories as

below. In the data presentation table, we used a star system

to denote the different density categories, as shown.

(1) Dense: the ratio of microvilli area to non-microvillatedarea is greater than or equal to 3:1 (***).

(2) Intermediate: the ratio is between 3:1 and more than 1:1

(**).

(3) Dispersed: the ratio is less than 1:1 (*).

3. Results

3.1. Overview

Several aspects of HGC patterns such as timing of HGC

development, HGC distribution on the surface, HGC shape

(including apical microvilli) as shown by SEM, for each

species investigated in this study, are described inTable 2.

We were unable to collect every stage for every species. For

consistency, these are shown as stage not available in the

Table. At their maximum extent, HGCs tended to form a con-

tinuous patch or line, with no other cells intervening; as the

hatching gland regressed, HGCs became separated from one

another and smaller in individual cell surface area. The stages

when these changes occurred are recorded inTable 2.

HGCs were identifiable by their dense clusters of

microvilli.Table 3provides data on microvillous length and

density. The other two cell types found on the epidermal


4/16


5/16


Table 3

Length and density of microvilli of the HGCs in 16 species of anurans

Family Species Stages Length (m, meanS.D.) Density

Bufonidae B. beebei 17 1.75 0.25 ***

B. bufo 19 1.4 0.11 ***

20 1.4 0.11 ***B. marinus 18 1.8 0.32 ***

19 1.75 0.25 ***20 1.7 0.21 **

22 0.8 0.29 **

B. viridis 16 0.6 0.12 **

18 1.1 0.29 ***

20 1 0.2 ***

21 1.1 0.1 ***

24 0.9 0.32 ***

26 0.94 0.13 ***

Pipidae X. laevis 18 2 0.65 **

24 2.1 0.55 **

Hylidae H. crepitans 17 3.6 0.49 **

18 3.2 0.75 **

19 2.7 0.59 **

20 4.1 0.67 **

21 2.6 0.31 **22 3 1 **

H. geographica 17 2.65 0.43 ***

18 2.4 0.49 ***

19 1.74 0.49 ***

20 2.9 0.3 **

22 2.24 0.56 **

H. microcephala 18 1.6 0.14 **

19 1.4 0.36 *

20 1.24 0.19 **

P. venulosa 16/17 0.73 0.25 ***

18 2 0.41 ***

19 1.84 0.44 ***

20 1.7 0.64 ***

21 2.3 0.67 ***

22 2.23 0.39 ***Leptodactylidae E. urichi 6 1 0.14 ***

L. fuscus 18 1.7 0.14 **

20 1.6 0.45 *

21 1.5 0.5 *

22 1.2 0.29 *

23 1.27 0.44 *

L. validus 20 2 0.71 **

23 1.3 0.25 **

L. bolivianus 21 1.94 0.24 ***

22 2 0.38 ***

23 2.5 0.5 **

P. pustulosus 16 1.4 0.48 **

17 1.8 0.13 **

20 1.12 0.32 **

Microhylidae E. ovalis 18 1.15 0.39 *

19 1.5 0.29 *

Ranidae R. temporaria 20 1.7 0.23 **

The staging inE. urichiis based onTownsend and Stewart (1985); the remainder onGosner (1960).

surface in the region of the hatching gland were common

epidermal cells (CEC), generally with microridged surfaces

and secretory micropores; and ciliated cells, easily distin-

guishable by their long cilia.

Observation at low magnification showed that there are

several general HGC patterns and distributions among the

different species of anurans. Examples of the HGC patterns

found are shown inFig. 1.The shape of the HGC apical pro-

files at the epidermal surface changed over the developmental

stages. Different species in a genus or family may have simi-

lar shapes of hatching gland but this was not always the case.

Fig. 2shows the overall pattern of HGCs on the dorsal sur-

face of the various species of anuran embryos we report on

here.Fig. 3gives a frontal view for two species.


6/16


3.2. Species descriptions, ordered by family

3.2.1. Bufonidae

The details of HGC pattern were considered for four

bufonid species: Bufo beebei (Figs. 1a and 2a), B. Bufo

(Fig. 1b,c; 2b), B. marinus (Figs. 1d; 2c; 3b)and B. viridis

(Fig. 2d). Common features of all four species were: at aboutstage 16, HGCs appeared on the dorsal side of the head. In all

species, the HGCs were distributed on the front of the head

more widely than on the back of the head or trunk. The cells

were all close together in a strand, distributed from the tip of

the head to the posterior end of the head, except B. marinus,

where all cells were concentrated as a patch on the front of

the head.

Features which differed between the species were: HGCsextended along the trunk as a continuous line inB. bufoand

Fig. 1. Examples of distribution and morphology of HGCs on the mid-dorsal side of the head in anuran embryos/larvae in different species and at different

stages: (a)B. beebei, stage 17; (b) B. bufo, stage 19 low magnification; (c) B. bufo, stage 19 high magnification; (d) B. marinus, stage 20; (e)B. viridis, stage

26; (f) X. laevis, stage 17/18; (g) H. crepitans, stage 21; (h) H. geographica, stage 17; (i) H. geographica, stage 22; (j) H. microcephala, stage 18; (k) H.

microcephala, stage 20; (l) H. microcephala, stage 21; (m) H. minuta, stage 20; (n)P. venulosa, stage 19; (o)L. fuscus, stage 18; (p) L. fuscus, stage 23; (q)L.

bolivianus, stage 22 low magnification; (r) L. bolivianushigh magnification, stage 22; (s) L. validus, stage 19 low magnification; (t) L. validus, stage 19 high

magnification; (u)P. pustulosus, stage 17; (v)E. urichi, stage 8/9; (w)E. ovalis, stage 18; (x)E. ovalis, stage 19. Scale bars: a, d, f, g, h, m,v, w, and x = 15m;

c, e, j, k and n = 8m; e, l= 60m; i, o, p, and u = 35m; q and s= 19m; r= 6m; t= 3m; b= 250m. Arrow head: HGC; arrow: ciliated cell; (*): common

epidermal cell.


7/16


8/16


Fig. 1. (Continued)

B. viridis. InB. beebei, a few HGCs were scattered individ-

ually on the trunk, and in B. marinus, HGCs were present

only on the head. At hatching stage, HGCs were very densly

packed inB. beebei andB. bufo especially on the head, and in

B. viridisand B. marinusthe density of HGCs was interme-

diate. Microvilli were very densely packed on the HGCs of

B. beebeiandB. bufoand dense inB. viridisandB. marinus.

HGCs remained at the surface up to stage 25/26 inB. viridis,


9/16


Fig. 1. (Continued)


10/16


Fig. 1. (Continued)

while they disappeared at stage 22 in B. bufoandB. marinusand at stage 23 inB. beebei. The shape of the HGCs changed

with stage, but we described the shape when HGCs reached

their maximum size. In B. beebei HGCs were round, in B.

bufo oval and in B. viridis elongated. HGCs in B. marinus

were four-sided or round.

3.2.2. Pipidae (X. laevis only)

X. laevis (Fig.1f and 2e): at stage 1718, most HGCs were

in a straight band, 36 cells wide, along the median line of

the dorsal side of the head and extending back much of the

length of the trunk. Two branches of cells extended from a

large patch at the tip of the head to the anterior-rostral sideof the head. The apical surfaces of the HGCs were highly

microridged at this stage. At stage 18 (hatching), microvilli

had grown on the surfaces of the HGCs especially on the

dorsalside of thehead.At stage 20,the HGCs were separating

from each other and at stage 22, only a few traces of HGCs

could be seen.

3.2.3. Hylidae

The details of HGC pattern were considered for eight

hylid species: Hyla crepitans (Fig. 1g; 2f), H. boans

(Fig. 2g), H. geographica (Fig 1h, i; 2h), H. microcephala


11/16


Fig. 1. (Continued).


12/16


Fig. 2. Distribution patterns of HGCs on the dorsal surface of the different species of anuran embryos: (a)B. beebei, stage 17; (b)B. bufo, stage 18/19; (c) B.

marinus, stage 20; (d)

B. viridis, stage 18; (e)

X. laevis, stage 18/19; (f)

H. crepitans, stage 17; (g)

H. boans, stage 18; (h)

H. geographica, stage 17/18; (i)

H.

microcephala, stage 18; (j)H. minuta, stage 18; (k)H. minuscula, stage 19; (l) P. venulosa,stage 18; (m) L. fuscus, stage 18; (n)P. pustulosus, stage 18; (o)L.

validus, stage 20; (p) L. bolivianus, stage 21; (q) E. urichi, stage 7/8 (r) E. ovalis, stage 18; (s) R. temporaria, stage 17/18. Drawings made to summarize cell

patterns observed when HGCs were present at their maximum extent.

(Figs. 1j, k, l, 2i), H. minuta (Fig. lm; 2j), H. minuscula

(Fig. 2k),Phyllomedusa trinitatisand Phrynohyas venulosa

(Fig. 1n; 2l).

Of the eight species, P. trinitatis was unique: by normal

criteria, we could find no HGCs on the head surface in this

species. Stages examined were 1823.

The timing of appearance and disappearance of HGCs

was similar across the remainder of the hylid species, from

stage 1516 to stage 2225. Where hatching occurred rel-

atively late (H. microcephala, H. minuscula, stage 2021),

HGC disappearance was also relatively late (stage 23,25).

The pattern of HGCs showed similarities, but also some

differences (Fig. 2fl). At the anterior end, the cells were

in a narrow line across the head, to a variable extent; the

HGCs then extended as a line mid-dorsally backwards, also

to a variable extent: very short in H. minuta; longest in P.


13/16


Fig. 3. Distribution patterns of HGCs in a frontal view of the head in (a)L. bolivianus, stage 21; and (b) B. marinus, stage 20.

venulosa. The line was relatively straight with short lateral

extensions inH. minuscula.

There appeared to be small differences in surface morphol-

ogy of HGCs between species. For example,H. crepitans had

long microvilli compared toH. geographicaandP. venulosa

(Fig. 1g,h,n) and HGC shape ranged from essentially oval to

polygonal with sharply angled sides.

3.2.4. Leptodactylidae

The details of HGC pattern were considered for fiveleptodactylid species:Leptodactylus fuscus(Figs.1o,p; 2m),

Physalaemus pustulosus (Figs. 1u; 2n), L. validus

(Figs. 1s, t; 2o), L. bolivianus (Fig. 1q, r; 2p; 3a) and

Eleutherodactylus urichi(Fig. 1v; 2q).

Of the five species,E. urichiwas unique in having direct

development to the froglet stage, with no free-swimming

tadpole. Hatching occurs with the aid of an egg tooth at

Townsend and Stewart (1985) stage 15. However, we did

find cells with the normal morphology of HGCs at two much

earlier stages, stages 4 and 8/9; no such cells were visible

at stage 14 (we lacked stages between 9 and 14). The cells

were oval or round with their apical surfaces covered in short

microvilli (Fig. 1v), and were single, not in groups, located

on the dorsal head surface.

In L. fuscus, L. validus and L. bolivianus, HGCs were

arranged as a broad loosely connected patch on the dor-

sal surface of the head, with no posterior linear extension.

P. pustulosus had a pattern intermediate between this and

the other groups studied: a broad band of HGCs anteriorly,

narrowing posteriorly, but only extending to the back of

the head.

As in the other groups, the first appearance of HGCs well

preceded hatching time; hatching occurred around the stage

when HGCs were separated from one another; HGCs disap-

pearedfrom thesurface a few stagesafterhatching, latestinL.

fuscus. Cell shapes were similar to those of other groups, oval

to round, with surface microvilli, though rather less dense

than in some other groups (Table 3).InL. fuscusHGCs were

oval, inL. validus oval or four sided and inL. bolivianus three

or four sided.

3.2.5. Microhylidae (Elachistocleis ovalis only)

E. ovalis (Figs. 1w,x; 2r): at stage 18,HGCs were observed

on the median line of the dorsal head, from the dorsal tip ofthe head to a short distance posteriorly. They were distributed

in one or two rows along the dorsal median line. They were

concentrated on the dorsal tip of the head, with some gaps

between different parts of the line. At stage 19, the HGCs

were separating. The cell apices were dispersed, elevated,

circular or oval in shape with short microvilli on their sur-

faces. At stage 2122, no sign of HGCs could be seen on the

dorsal head, but on the tip of the head, there was one patch

with slightly microvillated cell surfaces. At stage 23,at the tip

of the head, HGCs were still attached together, with slightly

microvillated surfaces.

3.2.6. Ranidae (Rana temporaria only)

R. temporaria(Fig. 2s): at stage 17, HGCs were observed

on the median line of the dorsal head and some parts of the

trunk. The HGCs were distributed along the median line with

patches alongside the line. The boundaries between HGCs

were not distinguishable, but they were slightly microvil-

lated on their surfaces. At stage 18, HGCs were slightly

elongated with very short surface microvilli. At stage 22,

HGCs were separated from each other. Small and often four

sided HGCs with microvilli were observed on the tip of

the head. At about stage 26, HGCs disappeared from the

surface.


14/16


4. Discussion

4.1. Hatching gland cell shape and pattern

There have been few previous comparative surveys of

hatching gland morphology.Meyer et al. (1969)described

the HG of Alytes obstetricans in detail, but also examined39 other species, 13 European and 27 from elsewhere. Their

method used stained whole mount preparations of embryonic

dorsal skin. They found some relationship between HG mor-

phology and anuran taxonomic relationships, but their paper

lacks detailed figures of the different species examined. The

stained whole-mount method is laborious and is not capable

of fine resolution of individual cells. More recently whole-

mount immunohistochemistry has been used to visualise the

HG and to map molecular expression patterns in the gland

(Cheng et al., 2002),mainly in X. laevis. This method also

does not resolve the detail of individual cells.

We have followed Yoshizaki and Yamamoto (1979) in

using scanning electron microscopy to map the HG. Themethod has the advantage of allowing the complete pattern

of HG cell apices to be seen in a single specimen, but also of

allowing visualisation of individual HGC apical specializa-

tions at high magnification. Since HGCs are surface active,

SEM is also reliable in easily detecting the appearance and

disappearance of HGCs from the embryo surface. SEM does

not detect HGCs once they disappear from the embryo sur-

face, but it is likely that, once this happens, the cells are no

longer functional and are undergoing degenerative changes

(Yoshizaki and Katagiri, 1975).

Fig. 2 summarises the pattern of HGCs in the species stud-

ied andTable 2synthesises our data on the timing of HGCdevelopment and regression. The patterns show an overall

general similarity, with a concentration of cells at the dor-

sal anterior part of the head, then a row of cells extending

back along the dorsal mid-line. Because we have concen-

trated on the dorsal side, our figures do not generally show

the Y shaped pattern seen in the frontal region in Xenopus

(see, for example,Cheng et al., 2002).In most species, the

pattern of HGCs on the anterior-most part of the head shows

a T or Y shape in dorsal view (Fig. 2)and we have confirmed

this with frontal views of two species (Fig. 3). However, there

are also clearly differences, especially in the extent of the dor-

sal mid-line row. It would be of interest to investigatewhether

these differences have functional significance.Bless (1905)

investigation of hatching in X. laevis, which seems not to

be been repeated for other species, suggests that the action

of the HGCs may be highly localised, where the HG makes

contact with the overlying vitelline membrane. The ideal pat-

tern of HGCs in a particular species may therefore be related

to the geometry of this interaction. Detailed examination of

hatching in a range of species is neededto test this hypothesis.

Examination ofTable 2andFig. 2hints at a relationship

between gland cell distribution and hatching stage: those

species with long posterior extensions of the gland (Fig. 2

b,d,e,g,l,s) hatch at stages 18 or 19; those where the gland

is restricted to the head region with only a short posterior

extension(Fig.2 a,c,h,j,m,n,o,p,r) hatch over a wider range of

stages, 1720. This is small difference and possibly biased by

the relatively large number of leptodactylid species in the lat-

ter category, but there may be some relationship here between

the overall shape of the embryo at hatching, and the detailed

hatching mechanism.Meyer et al. (1969)suggested some level of relationship

between HG pattern and taxonomic position. Our survey may

be too limited to discern such relationships, but some patterns

emerge e.g. in leptodactylids, the HG is mainly a wide ante-

rior patch, with little posterior extension; in the hylids, the

posterior extension tends to be a narrow, sometimes zig-zag,

line of cells.

Table 2 shows that thoughthere is some variability in onset

and disappearance of HGCs, there is perhaps less than might

be expected, given the variability in hatching stage. SEM fea-

turesdo notnecessarilytell uswhen cells areactive. Yoshizaki

and Katagiri (1975) assessed HGC function at differentstages

by culturing embryos in medium, then measuring the prote-olytic activity of the medium, taken to be mainly derived

from HGC secretions. Activity was low in medium exposed

to stage 1719 embryos, but high after exposure to stage 21

(hatching stage) and stage 22 embryos. Unfortunately, they

did not test later stage embryos. This result at least indi-

cates that release of proteolytic enzymes persists some time

afterhatching. Yoshizaki (1991) detected intra-cellular prote-

olytic activity inXenopusHGCs over a wide range of stages

(Nieuwkoop and Faber stages 2441, equivalent to Gosner

1722) including post-hatching. Whether this has functional

significance is unclear: see discussion on L. fuscus, below.

The fact that the HG appears at the embryo surface sometime beforehatchingoccurs (by 12 Gosner stages, generally:

Table 2)and disappears well after hatching (by 34 Gosner

stages) may pre-dispose some species towards the evolution

of hatching in response to external stimuli (Martin, 1999).

4.2. Special cases

4.2.1. E. urichi

Hatching in all eleutherodactylids involves the use of an

egg tooth which mechanically tears apart the investing jelly

coat of the egg, to allow the froglet to emerge. However, as

Salthe (1963) noted,eleutherodactylids hatchearly from their

vitelline membranes. Presumably, this process is mediated by

the HGCs we found at Townsend and Stewart stages 48/9.

4.2.2. P. trinitatis

ItispuzzlingtohavefoundnosignofHGCsinthisspecies.

Hatching occurs late in Phyllomedusa, but we would have

expected HGCs still to have a role. It may be that the cells

have different morphology, or a different location from other

species, but we were unable to find HGCs in this species

either in SEM or in sections. Kenny (1968) reported that from

about one third of the 7.5 days incubation period onwards, the

vitelline membrane becomes progressively distended with


15/16


fluid, reaching a diameter of 4 mm from an initial 2.3 mm.

Hatching can occur over a small range of stages and seems to

be stimulated by the hatching of neighbouring embryos: their

hatching leads to violent wiggling of unhatched embryos,

then sudden rupture of their vitelline membranes. It seems

possible, therefore, that in this species, there is some advan-

tage in synchronous hatching, and that this is achieved by amechanical rather than an enzymatic mechanism.

4.2.3. L. fuscus

Downie (1994) reported that hatching occurs at stage

1819 in this species and that larvae remain in the foam

nest and generate a new kind of foam. He followed HGC

regression both in larvae that remained in the foam nest and

in individuals transferred to water and allowed to feed. The

results showed that HGC regression was slower in larvae that

remained in the foam nest, and Downie suggested that the

HGCs may retain some function in these foam-dwelling lar-

vae. However, HGC presence was detected only in sections,

and it may be that the HGCs were no longer functional. How-ever, this is a possible example where post-hatching activity

of HGCs may occur. Our SEM results show that in nor-

mal development, HGC appearance at the embryo surface

remains till a relatively late stage.

4.3. Conclusion

Our previous studies on anuran embryonic and early larval

surface structure, on ciliated cell patterns and cement gland

morphology respectively (Nokhbatolfoghahai et al., 2005;

Nokhbatolfoghahai and Downie, 2005)have revealed con-

siderable diversity, related both to taxonomic position andto mode of development. Since our earlier work appeared,

Frost et al. (2006)have proposed a considerable revision of

amphibian taxonomy, some of it provisional, given the rela-

tively small number of species sampled in parts of the tree. To

allow continuity with our earlier papers, we have not changed

any the names in our descriptions, but we have noted the

changes in Table 1. Thework reported here on HG pattern and

cellsurface ultra-structure shows some fine scale diversity but

little obvious relationship to taxonomy (seeFrost et al., 2006

for family relationships) or to developmental mode, other

than some similarityin overall shape among the foam-nesting

leptodactylid species examined. The most surprising results

were the absence of HGCs from P. trinitatis and their pres-

ence at early stages inE. urichi, long before actual hatching.

P. trinitatis belongs to the same hylid sub-family as Agaly-

chnis(Duellman and Trueb, 1994),whereWarkentin (1995)

has demonstrated facultative hatching in response to exter-

nal stimuli.Warkentin (1999)reported on several aspects of

Agalychnishatchling morphology, but made no comment on

the presence or absence of HGCs.

Callery et als. (2001) investigation of the evolutionary

changes that have led to direct development in Eleuthero-

dactylus showed that remnants of the ancestral pattern of

development, such as late stage responsiveness to thyroid

hormone, remained. Our results suggest that HGC function

at an early stage is one of these remnants, allowing internal

hatching of the embryo from the vitelline membrane, even

though full hatching occurs only much later and by a different

mechanism.

Acknowledgments

We thank Margaret Mullin, Eoin Robertson and Ian Mont-

gomery for technical assistance and advice. Students in

University of Glasgow expeditions to Trinidad helped col-

lected specimens. Bernard Zonfrillo providedRana andBufo

spawn from the Glasgow area and Simon Merrywest of the

University of St Andrews providedXenopusspawn. The late

Professor Peter Bacon and colleagues at the University of

the West Indies kindly provided laboratory space and the

Trinidad Government Wildlife section provided collection

permits. J.R.D.s fieldwork in Trinidad was supported by the

Carnegie Trust. M.N. thanks the University of Shiraz, IslamicRepublic of Iran for financial support.

References

Altig, R., McDiarmid, R.W., 1999. Body-plandevelopment andmorphology.

In: McDiarmid, R.W., Altig, R. (Eds.), Tadpoles: The Biology of Anuran

Larvae. University of Chicago Press, Chicago, pp. 2451.

Balon, E.K., 1984. Patterns in the evolution of reproductive styles in fishes.

In: Potts, G.W., Wootton, R.J. (Eds.), Fish Reproduction: Strategies and

Tactics. Academic Press, London, pp. 3553.

Bles, E.J., 1905. The life history ofXenopus laevis Daudin. Trans. Royal

Soc. Edinburgh 41, 789821.Callery, E.M, Fang, H., Elinson, R.P., 2001. Frogs without polliwogs: evo-

lution of anuran direct development. Bioessays 23, 233241.

Carroll, E.J., Hedrick, J.L., 1974. Hatching in the toadXenopus laevis: mor-

phological events and evidence for a hatching enzyme. Develop. Biol.

38, 113.

Cheng, S.-M., Chen, I., Levin, M., 2002. KATP channel activity is required

for hatching in Xenopusembryos. Develop. Dyn. 225, 588591.

Downie, J.R., 1994. Developmental arrest inLeptodactylus fuscustadpoles

(Anura: Leptodactylidae). 1: descriptive analysis. Herpetol. J. 4, 2938.

Drysdale, T.A., Elinson, R.P., 1991. Development of the Xenopus laevis

hatchingglandand itsrelationshipto surfaceectodermpatterning.Devel-

opment 111, 469478.

Duellman, W.E., Trueb, L., 1994. Biology of Amphibians. The Johns Hop-

kins University Press, Baltimore, pp. 670.

Fan,T.J., Katagiri,C., 2001. Properties of thehatchingenzyme fromXenopuslaevis. Eur. J. Biochem. 268, 48924898.

Fox, H., 1986. The skin of amphibia. In: Bereiter-Hahn, J., Matoltsy, A.G.,

Richards, K.S. (Eds.), Biology of the Integument 2-The Vertebrates.

Springer-Verlag, Berlin, pp. 78149.

Frost, D.R., Grant, T., Faivovich, J., Bain, R.H., Hass, A., Haddad, C.F.B.,

De Sa, R.O., Channing, A., Wilkinson, M., Donnellan, S.C., Raxworthy,

C.J., Campbell, J.A., Blotto, B.L., Moler, P., Drewes, R.C., Nussbaum,

R.A., Lynch, J.D., Green, D.M., Wheeler, W.C., 2006. The amphibian

tree of life. Bull. Am. Mus. Nat. Hist. 297, 1370.

Gosner,K.L., 1960. A simplifiedtable for staging anuran embryosand larvae

with notes on identification. Herpetologica 16, 183190.

Gunzburger, M.S., 2003. Evaluation of the hatching trigger and lar-

val ecology of the salamander Amphiuma means. Herpetologica 54,

459468.


16/16


Kenny, J.S.,1968. Earlydevelopment and larval natural history ofPhyllome-

dusa trinitatisMertens. Carib. J. Sci. 8, 3545.

Kenny, J.S., 1969. The amphibia of Trinidad. Studies Fauna CuraCao other

Carib. Is. 29, 178.

Martin, K.L.M., 1999. Ready and waiting: delayed hatching and extended

incubation of anamniotic vertebrate terrestrial eggs. Am. Zool. 39,

279288.

McDiarmid, R.W., Altig, R., 1999. Research: materials and techniques. In:

McDiarmid, R.W., Altig, R. (Eds.), Tadpoles: The Biology of Anuran

Larvae. University of Chicago Press, Chicago, pp. 723.

Meyer, M., Aurin, H., Rummelfanger, H., 1969. Die schlupfdruse der

geburtscheferkrote (Alytes o. obstetricans (Laurenti)) und anderer

froschlurche. Ergeb. Anat. Entwicklungsgesch. 41, 364.

Murphy, J.C., 1997. Amphibians and Reptiles of Trinidad and Tobago.

Krieger Publishing Company Malabar, Florida, 245 pp.

Nokhbatolfoghahai, M., Downie, J.R., 2005. Larval cement gland of frogs:

comparative development and morphology. J. Morphol. 263, 270283.

Nokhbatolfoghahai, M., Downie, J.R., Clelland, A.K., Rennison, K., 2005.

The surface ciliation of anuran amphibian embryos and early larvae:

patterns, timing differences and functions. J. Nat. Hist. 39, 887929.

Salthe, S.N., 1963. The egg capsules in the amphibia. J. Morphol. 113,

161171.

Touchon, J.C., Gomez-Mestre, I., Warkentin, K.M., 2006. Hatching plas-

ticity in two temperate anurans: responses to a pathogen and predation

cues. Can. J. Zool. 84, 556563.

Townsend, D.S., Stewart, M.M., 1985. Direct development in Eleuthero-

dactylus coqui(Anura: Leptodactylidae): a staging table. Copeia 1985,

423436.

Vonesh, J.R., 2005. Egg predation and predator-induced hatching plasticity

in the African reed frog, Hyperolius spinigularis. Oikos 110, 241252.

Warkentin, K.M., 1995. Adaptive plasticity in hatching age: a response to

predation risk trade-offs. Proc. Natl. Acad. Sci. USA 92, 35073510.

Warkentin, K.M., 1999. Effects of hatching age on development and hatch-

ling morphology in the red-eyed tree frog, Agalychnis callidryas. Biol.

J. Linn. Soc. 68, 443470.

Yoshizaki,N., 1991. Changesin surface-ultra structureand proteolytic activ-

ity of hatchinggland cellsduringdevelopment ofXenopusembryo. Zool.

Sci. 8, 295302.

Yoshizaki, N., Katagiri, C., 1975. Cellular basis for the production and

secretion of the hatching enzyme by frog embryos. J. Exp. Zool. 192,

203212.

Yoshizaki, N., Yamamoto, M., 1979. A stereoscan study of the development

of hatching gland cells in the embryonic epidermis ofRana japonica.

Acta Embryol. Exp. (3), 339348.

Documents

Amphibian Hatching Gland Cells