Maintenance of segment and appendage primordia

by the Tribolium gene knodel

Hilde Wohlfrom a, Johannes Benno Schinko b, Martin Klingler a, Gregor Bucher b,*

a Institut fur Biologie, Friedrich-Alexander-Universitat Erlangen, Staudtstrasse 5, 91058 Erlangen, Germanyb Johann-Friedrich-Blumenbach-Institut fur Zoologie und Anthropologie, Georg-August-Universitat Gottingen, Abteilung Entwicklungsbiologie,

Justus-von-Liebig-Weg 11, 37077 Gottingen, Germany

Received 2 November 2005; received in revised form 11 April 2006; accepted 26 April 2006

Available online 5 May 2006

Abstract

For homeotic and segment-polarity genes in Drosophila, a switch in gene regulation has been described that distinguishes patterning and

maintenance phases. Maintenance of segment and organ primordia involves secondary patterning and differentiation steps, as well as survival

factors regulating proliferation and organ size. In a screen for embryonic lethal mutations in the flour beetle Tribolium castaneum, we have

recovered two alleles of the knodel gene, which result in short, bag-like embryos. These embryos have severely reduced appendages and

differentiate a cuticle that lacks most overt signs of segmentation. In addition, they lack bristles and display defects in the nervous system. Early

patterning in knodel mutant embryos is normal up to the extended germ band stage, as indicated by the formation of regular even-skipped (Tc’eve)

and wingless (Tc’wg) stripes. Afterwards, however, these patterns degenerate. Similarly, proximo-distal growth and patterning of limbs are nearly

normal initially, but limb primordia shrink, and proximo-distal patterns degenerate, during subsequent stages. knodel could be a segment polarity

gene required for segment border maintenance in both trunk and appendages. Alternatively, it may have a more general role in tissue or organ

maintenance.

q 2006 Elsevier Ireland Ltd. All rights reserved.

Keywords: Tribolium; knodel; Naked cuticle; Segmentation; Segment polarity; Maintenance; leg development

1. Introduction

The principles underlying the formation of Drosophila body

segments are well understood (Pankratz and Jackle, 1990; St

Johnston and Nusslein-Volhard, 1992). Equally important, but

less understood, is the maintenance of established segment and

organ primordia. During pattern formation the embryo is

subdivided in ever smaller units as for instance seen by the

expression domains of gap and pair rule genes. These subdivi-

sions are transient in nature and are not directly transformed

into morphological traits. Only the last level of the segmenta-

tion hierarchy, the segment polarity pattern, provides direct

cues for morphogenesis, i.e. compartment boundaries, cuticle

differentiation or muscle attachment sites (Martinez Arias et al.,

1988; Sanson, 2001; Larsen et al., 2003). Pattern maintenance,

in contrast, requires stable cues that are able to maintain borders

and regulate the size of morphological units. Only few of these

0925-4773/$ - see front matter q 2006 Elsevier Ireland Ltd. All rights reserved.

doi:10.1016/j.mod.2006.04.003

* Corresponding author. Tel.: C49 551 395426; fax: C49 551 395416.

E-mail address: gbucher1@uni-goettingen.de (G. Bucher).

cues are known. Mutual activation of wingless and engrailed

expression is crucial for the maintenance of parasegmental

boundaries during germ band extension. Afterwards, their

expression domains become independent from each other

(Heemskerk et al., 1991; Martinez Arias, 1993; DiNardo et

al., 1994). Engrailed expression then comes under cell autono-

mous control involving the action of polycomb group genes

(Moazed and O’Farrell, 1992). In the case of the segment

polarity gene gooseberry, different enhancer elements are

responsible for early and late functions (Li et al., 1993).

Adjustment of parasegment size involves cell death and

delamination of miss-specified cells (Hughes and Krause,

2001). In wing imaginal discs, maintenance of parasegmental

borders involves hedgehog regulated cell sorting (Dahmann and

Basler, 2000). It has been assumed that differential cell

adhesion may be involved but the molecules responsible for

the hypothetical differential cell adhesion remains elusive

despite specific screening efforts (Dahmann and Basler, 1999).

All conclusions drawn from Drosophila research bear the

drawback that this species develops through a highly derived

Mechanisms of Development 123 (2006) 430–439

www.elsevier.com/locate/modo

H. Wohlfrom et al. / Mechanisms of Development 123 (2006) 430–439 431

mode of embryogenesis (long germ embryogenesis) that is not

typical for insects (Tautz et al., 1994; Davis and Patel, 2002).

Hence, statements on general properties of insect development

have to be based on studies of insects with more typical

features. We have chosen the red flour beetle Tribolium

castaneum as model system, because it is readily manipulated

by genetic, transgenic and RNAi approaches (Maderspacher et

al., 1998; Berghammer et al., 1999; Brown et al., 1999; Bucher

et al., 2002; Klingler, 2004). As in most insects, in Tribolium

only the anteriormost segments are patterned in the blastoderm

stage while all following segments are formed one by one from

a posterior growth zone (short germ embryogenesis). In

addition, Tribolium displays a fully developed larval head,

larval appendages and extraembryonic tissue. The morpho-

logical conservation suggests that also the genetic control

reflects the ancestral state more accurately than it does in

Drosophila. Indeed, it has been shown that early patterning

in Tribolium does not involve bicoid and that the abdominal

gap gene orthologs have drastically changed their function

(Brown et al., 2001; Bucher and Klingler, 2004; Cerny et al.,

2005). In contrast, on the level of segment polarity genes

studied so far, both expression and function appear to be

conserved to great extent (Brown et al., 1994; Nagy and

Carroll, 1994; Oppenheimer et al., 1999).

In a screen for embryonic lethal mutations in Tribolium, we

have recovered knodel (kno), a mutant lacking cuticular

segmentation. We show that segmentation itself is not compro-

mised but the maintenance of the pattern is impeded. We

speculate that kno may be a late acting segment polarity

gene, or a gene required for adhesion. The phenotype of kno

mutant embryos shows that also in short germ embryos, initial

patterning of the growing germ band is followed by a distinct

maintenance phase.

2. Experimental procedures

2.1. Identification of the knodel mutant strain

kno was identified independently in two genetic backgrounds. In a screen

for embryonic lethal mutations (Maderspacher et al., 1998), the allele 13h8 was

isolated in an EMS-treated SB wild-type background. Independently, we found

that our copy of a Tiw-1 wild-type strain harbored a large number of

chromosomes carrying a kno allele: 39 of approximately 400 inter se back-

crosses segregated embryos indistinguishable from 13h8. Complementation

analysis showed that the two mutants (kno13h8 and knoTiw1) are allelic. knodel

(dumpling) is a dish typical for southern Germany and consists of amorphous

balls made of potato flour or bread that resembles the knodel phenotype.

2.2. Complementation study

One knoTiw1 male was crossed to three kno13h8 females and the offspring

checked for mutant offspring (nZ8) and vice versa (nZ10). Mutant offspring

was detected in 3/8 and 7/10 of the crosses, respectively. Therefore, the

mutations do not complement and are alleles of the same locus.

2.3. Histology

Whole-mount in situ hybridizations were performed according to estab-

lished protocols (Tautz and Pfeifle, 1989). For initial double stainings,

fluorescein- and digoxigenin-labeled probes were detected using alkaline

phosphatase and beta-galaktosidase, the latter after signal enhancement via

biotin deposition (Prpic et al., 2001). For the remaining double in situs both

stainings were detected successively by alkaline phosphatase using NBT/BCip

and the fluorescent FastRed reaction (Sigma), respectively. The false color

pictures were produced by pasting the inversed bright field image into the red

and blue channels of the image of the fluorescent staining. Detailed protocols

are available from the authors. The stained embryos were dissected free from

the yolk and embedded in 50% glycerol. Cuticles were embedded in Hoyer’s

medium mixed with lactic acid (1:1) and analyzed using standard dark field

illumination. In addition, confocal microscopy of the auto fluorescent cuticle

was used (excitation: argon laser, 488 nm; emission filter: long pass 505 nm)

2.4. Embedding in low refraction medium

In glycerol, the yolk granules disturb optical observation. Therefore,

embryos usually have to be dissected from the yolk. In order to observe

embryos within the egg, we have developed an embedding method for making

yolk transparent for normal bright field microscopy. Embryos are dehydrated in

alcohol (ethanol or methanol, 25, 50, 75, 100%). They are then embedded in a

solution of high refractory index, ‘benz-mix’, a 4.3:1 mixture of benzyl

benzoate: benzyl alcohol for embryos dehydrated in ethanol (3:1 for embryos

dehydrated in methanol). This technique was applied to the whole-mount

Tc’even-skipped staining.

2.5. Scanning electron microscopy

Embryos were dehydrated by 10 min incubations in raising acetone

concentrations (70%, 80%, 2!98%, 2!100%). They were then immediately

transferred to HMDS (1,1,1,3,3,3 hexadimethyldisilazane; Merck-Suchardt,

Darmstadt) and incubated for 30 min. HMDS was then removed and embryos

transferred to an exsiccator loaded with silica gel balls (Merck-Suchardt,

Darmstadt). The exsiccator was evacuated in order to avoid contamination

with water (which causes shrinkage), and evaporation was performed over

night. Embryos were then transferred to an aluminium carrier (Plano, Wetzlar)

that had been prepared with a double-sided adherent disc (Plano, Wetzlar). The

embryo was then sputtered with gold (BIO-RAD SC 510, Munchen) at 200 V

for 120 s. Scans were performed with a Philips XS-20 scanning electron

microscope at 20 kV.

3. Results

3.1. knodel embryos lack segment boundaries and bristles

kno mutant larvae differentiate as nearly amorphous cuticu-

lar sacs that lack most bristles, proper appendages and any

sign of abdominal segmentation (Fig. 1A–C). The cuticle of

kno embryos appears to be shaped by the egg shell because

the cuticle closely lines the vitelline membrane. Accordingly,

the proportions of kno embryos differ strongly from wild-type:

they are about half as long, but approximately 30% more wide

than wild-type 1st instar (L1) larvae. The anterior posterior

axis is well established, however, as in all larvae the anterior

pole can be recognized. In the head, several rudimentary

appendages are usually formed, although it is often difficult

to assign their identity unequivocally. The labrum is

commonly discernable as are shortened antennae that lack

the flagellum. Mandibles, maxillae and labium are strongly

reduced to cuticular humps that sometimes carry distal

structures like reduced palps or sclerotized tips (mandible).

In most cases, remnants of most but not all gnathal segments

can be found. Legs are either missing or discernable as

Fig. 1. Cuticle phenotype of kno mutant embryos. White lines indicate the ventral midline, (B 0) and (C 0) are red–green representations of the embryos in (B) and (C),

respectively. (A) Wild-type larva displaying head, thorax with three pairs of legs, and eight discernible abdominal segments. (B) In weak kno phenotypes, the labrum

(open arrowhead), antenna (open arrow) and maxilla (star) are clearly visible. The labium is not present in this case, but in other specimens, the labium is present

while other gnathal structures may be missing. The only indication of trunk segmentation is provided by the remnants of thoracic legs (white arrowheads). Arrows

point to four nests of cuticular structures (those marked with D and E are shown in the respective panels). (C) Even in strong phenotypes, the labrum is usually visible

(open arrowhead) and indicates the anterior end of the embryo. Only few remnants of gnathal and thoracic appendages can be distinguished (open arrow: antenna,

white arrowheads: thorax). The lengths of wild-type and kno cuticles are to scale. The width of the cuticles cannot be compared, however, because only the wild-type

has been flattened in order to display all structures in one plane. Actually, kno cuticles are approximately 33% wider than cuticles of wild-type siblings. (D)–(E 0)

Close-up of ‘nests’ of cuticular structures of the cuticle shown in B. These structures presumably derive from malformed and fused bracts normally associated with

dorsal bristles. (D) This type of cuticular structure resembles an elevation with several sclerotized peaks pointing outwards. (E), (E 0) Two focal planes of another type

of bract nest. This structure is sunk below the surrounding cuticle and the blunt bracts point towards the center. An identical structure has been described for the

Tribolium cobbled mutant.

H. Wohlfrom et al. / Mechanisms of Development 123 (2006) 430–439432

cuticular bulges. Segmentation of gnathum and thorax is only

detectable from the reduced appendages since no segmental

folds are evident. Correspondingly, in the abdomen no signs

of segmentation whatsoever can be found. The posterior

terminal urogomphi are visible in a portion of kno mutant

cuticles, while the pygopods are never detected. Most differ-

entiated kno embryos lack spiracles—only in few cases a

single spiracle was found. Sometimes up to seven irregularly

distributed bristles are present, some of which split at the

base. In some embryos, however, one or two types of unusual

‘nests’ of cuticular structures can be found (Fig. 1D and E).

One type of nests has a distinct roundish circumference and

contains several bract-like structures that point to the center of

the nest (Fig. 1E). The other type consists of sclerotized tips

pointing upwards (Fig. 1D). These structures lack a clear

border and appear situated on cuticular elevations. In some

embryos we find 3–5 such nests while they are absent in

others. We have not found any regularity regarding their

position apart from the fact that they occur in the abdomen

only.

H. Wohlfrom et al. / Mechanisms of Development 123 (2006) 430–439 433

In order to test for penetrance of the mutation, we crossed a

single F1 male (probability for being a mutation carrier: 50%)

with three F1 females (cumulative probability that at least one

of them is a mutation carrier: 87.5%). Mutant phenotypes could

be detected in 21 of 48 such crosses (43.8%), which coin-

cidently, is exactly the expected proportion (50% * 87.5%Z43.8%). In order to test for heterozygous effects of the

mutation, we crossed mutation carrier males with wild-type

females and tested the offspring for cuticle defects. As no

defects could be found we conclude that the mutation is strictly

recessive.

In summary, the gene affected in kno is recessive lethal and

affects segmentation. The presence of rudimentary head and

thorax appendages indicates that segmentation is not totally

impeded and that some proximal-distal axis formation is

performed. In addition, the mutation also interferes with

bristle development. The kno gene may therefore also be

required for peripheral sensory organ development.

3.2. During germ band growth, even-skipped expression

is normal

To determine the moment when segmentation is disturbed in

kno mutant embryos, we stained the offspring of heterozygous

parents for the product of the pair-rule gene Tc’even-skipped

(Tc’eve). Like in Drosophila, this gene is initially expressed in

every other segment but resolves into a segmental pattern later

on. We did not detect any disturbance of the pair-rule pattern or

the subsequent segmental Tc’eve pattern in kno mutant

embryos during segmentation (not shown). We also find no

overt changes in the morphology of elongating embryos.

Together, these data indicate that growth and the segmentation

process are not affected in kno embryos.

After segmentation is completed, Tc’eve becomes expressed

in a segmentally reiterated pattern. Ventrally, Tc’eve

Fig. 2. Expression of Tc’eve wild-type (A,B) and kno embryos (C,D). The embryo

views with the anterior towards left (posterior parts are out of focus). (A,B) Tc’eve

clusters most likely mark mesodermal cells that contribute to the larval heart and dor

(A). (C) A kno mutant embryo at approximately the same age as that in (A). Staining

the latter comprise more cells and come into close proximity. (D) In older mutant

number of expressing cells. Some clusters have fused reducing the overall number o

(see text). Note that we do neither find changes in Tc’eve expression in younger st

expression presumably marks a set of neural precursor cells

(Fig. 2A and B). This pattern is conserved within insects and

crustaceans (Duman-Scheel and Patel, 1999). In kno mutant

embryos, also the neural Tc’eve expression initiates normally

but becomes increasingly disorganized during retraction stages

(Fig. 2C and D). Eventually, the pattern is strongly disturbed in

embryos that have finished retraction. In Drosophila, eve-

expressing cell clusters arise segmentally also in the dorsal

mesoderm and contribute to the larval heart and dorsal somatic

muscles. Among the regulators necessary for proper activation

of these cell clusters are several segmentation genes: wingless,

hedgehog and sloppy paired (Lawrence et al., 1995; Riech-

mann et al., 1997; Liu et al., 2006). In Tribolium, we find

Tc’eve positive cell clusters that resemble those in Drosophila

with respect to the onset of expression, their number and dorsal

position (Fig. 2A, black arrowhead). We therefore assume that

these clusters are heart and dorsal somatic muscle precursors.

In kno mutant embryos, the staining in these dorsal cell clusters

becomes more prominent than in wild-type (Fig. 2C, black

arrowhead). This is partly due to the greater number of cells in

each cluster but might in addition reflect Tc’eve up-regulation.

In addition, some clusters appear to fuse with their segmental

neighbors. Accordingly, the number of clusters is reduced in

kno mutant embryos as compared to wild-type. In the stages

examined, the number is not reduced to half as would be

expected for a pair rule phenotype. Intriguingly, this pattern is

almost identical to what happens in Drosophila when wingless

is evenly expressed in embryos (Lawrence et al., 1995). This

raises the possibility that knodel may be a component of the

wingless signaling pathway.

Together, the Tc’eve staining suggests that segmentation

and early neurogenesis are not affected while development

becomes disturbed during retraction. This indicates that kno

probably is involved in maintenance of segment primordia

rather than their establishment.

s are embedded in low refraction medium without removing the yolk. Ventral

is expressed in a specific and segmentally repeated set of neurons. Dorsal cell

sal somatic muscles (black arrowhead). The embryo in B is older than the one in

in ventral cell clusters fades while it intensifies in the dorsal clusters. In addition,

embryos, the lateral clusters have further gained in strength of expression and

f clusters. This phenotype is similar to wingless over expression in Drosophila

ages undergoing segmentation nor hints for a pair rule defect.

H. Wohlfrom et al. / Mechanisms of Development 123 (2006) 430–439434

3.3. Segment polarity gene patterns degrades with

developmental time

In order to get more insight into the nature of the defects, we

analyzed segmental markers in kno mutant embryos. The

boundary between the expression domains of the segment

polarity genes engrailed and wingless defines the paraseg-

mental boundaries of all arthropods investigated so far

(Damen, 2002). We investigated the patterns of Tc’wingless

(Tc’wg) and Tc’engrailed (Tc’en) and did not find any

alteration in early stages up to the extended germ band stage

(not shown). Thereafter, the pattern begins to degrade

(Fig. 3D–I, M–Q). The effect of the kno mutation on the

Tc’wg pattern is particularly severe. In wild-type embryos,

Tc’wg staining evolves into a complex but regular pattern of

Tc’wg positive cells (Nagy and Carroll, 1994). In kno mutant

embryos this pattern is initially modified by under- and hyper-

regulation of certain expression spots (Fig. 3E and F compare

medial and lateral spots (red and blue arrowheads, respect-

ively) with wild-type in B,C). Later, the wild-type pattern is

lost almost completely and only a few irregularly distributed

Tc’wg positive cell clusters are maintained (Fig. 3G and I).

Similar to the Tc’eve staining, the Tc’wg expression in kno

mutant embryos becomes stronger in the lateral parts than in

median parts of the germ band. These lateral parts of the germ

band represent future dorsal tissues. Apparently, the require-

ment for kno is not equally distributed along the dorsal-ventral

axes. The changes in the Tc’en patterns are less severe

(Fig. 3M–Q, compare to wild-type J–L). In retracting

embryos, the pattern begins to fade along the ventral midline

while the stripes remain almost normal laterally (Fig. 3M and

M 0, compare to J). In older embryos, the stripes become

discontinuous and irregular laterally while the ventral tissue

remains free of Tc’en staining (Fig. 3N,O,Q, compare to K,L).

These data demonstrate that the process of segment boundary

formation is intact in kno mutant embryos but that the

maintenance of the segmental pattern is impeded.

3.4. Appendage primordia are initiated but shrink subsequently

We wondered whether the reduction of appendages is a

secondary consequence of the segment degeneration pheno-

type or whether appendages are independently affected by lack

of kno function. We examined wild-type and mutant embryos

by scanning electron microscopy (SEM) (Fig. 4). We examined

stages where segmentation is completed and appendages are

growing out (32–48 h of development at 32 8C). We find that at

these stages all appendages are present in mutants but are

shorter than in wild-type embryos at a similar stage. However,

these early appendage defects are much less severe than those

in 1st instar larvae (Fig. 1B and C). To confirm our impression

that the initiation of leg patterning is normal in kno mutants, we

stained for Tc’Distal-less (Tc’Dll) as distal marker, and

Tc’dachshund (Tc’dac) as marker for intermediate proximo-

distal positions (Beermann et al., 2001; Prpic et al., 2001).

Indeed, prior to germband retraction stages, no pronounced

difference from the wild-type expression patterns could be

observed (not shown). Even in advanced embryonic stages,

where body segmentation of mutant embryos is already visibly

disturbed, the expression of these leg genes is still strikingly

normal (Fig. 5). The almost complete deletion of leg structures

in mature embryos therefore must involve secondary degener-

ation of leg primordia, which developed quite normally during

earlier embryo stages. Similarly as in anterior–posterior

patterning, this suggests that kno mutant embryos have a

problem in maintaining proximo-distal structures rather than

in establishing them.

4. Discussion

First instar cuticles of larvae mutant for kno lack overt

abdominal segmentation and have severely reduced appen-

dages. In addition, they lack almost all bristles and display

disorganized patterns of neural precursor cells. It should be

pointed out that segmentation and appendage defects are

usually not linked in Tribolium. Segmentation mutants like

itchy, scratchy, krusty, kraken and others affect segmentation

but not proximo-distal patterning of the appendages (Sulston

and Anderson, 1996; Maderspacher et al., 1998). In godzilla

and bollig, abnormally patterned legs are observed, however,

these appendage phenotypes appear to be secondary effects of

early trunk segmentation defects. Our results indicate that the

kno phenotypes result from degeneration of well initiated

segmental and appendage primordia.

4.1. knodel shares phenotypic traits with the Tribolium

cobbled mutation

A similar combination of sac like appearance and nests of

bracts has previously been described for the mutation cobbled

(Sulston and Anderson, 1996). In contrast to kno, however, the

appendages of cobbled embryos are much better developed,

and all body segment boundaries can still be identified. Also

cobbled mutant embryos lack bristles. In contrast to kno, the

nests of bracts occur in almost all abdominal segments at the

approximate dorso-ventral position of wild-type bristle-bracts.

The combination of defects suggests that cobbled could be a

weak allele at the kno locus. Alternatively, cobbled could be a

gene involved in the same pathway as kno. Unfortunately, the

cobbled mutation has been lost, i.e. we were unable to test kno

and cobbled for complementation.

4.2. Is knodel involved in cell adhesion?

Degeneration of tissues is a hallmark of mutations that affect

epithelial cell polarity and adhesion. Drosophila embryos with

reduced armadillo/beta-catenin function for instance form a

normal blastodermal epithelium but subsequent morphogenetic

movements like germ band extension do not occur properly or

not at all. The epithelium loses its apical basal polarity and

subsequently degenerates into a disordered multilayered

mesenchyme. A contiguous cuticle is not secreted in such

degenerating embryos leading to only some residual cuticle

scrabs (Amin et al., 1999). Such ‘crumbs-like’ phenotypes are

Fig. 3. Expression of the segment polarity genes Tc’wg and Tc’en in wild-type and knodel mutant embryos. Tc’wg is shown in blue in (A)–(I) with Tc’en

counterstaining in red. Tc’en is shown in purple in (J)–(Q) with Tc’wg counterstaining in green. (M 0)–(Q 0) are brightfield images of the embryos in (M)–(Q) with

Tc’en in blue. All embryos are shown from the ventral side, anterior is to the left. White arrowheads point to the first abdominal segment in (J)–(Q). Tc’wg

expression: (A) wild-type elongated germ band. Tc’wg is expressed in transverse stripes anteriorly adjacent to Tc’en and the boundary between these genes defines

the parasegmental boundary. During segment formation, the Tc’wg stripe covers almost the complete width of the embryo (see posteriormost stripes), later it retracts

to a more ventral position (see anterior abdominal stripes). Tc’wg is also expressed in a ventral stripe in the appendages. (B) During early germ band retraction Tc’wg

becomes reduced to ventral and lateral cell clusters (red and blue arrowheads, respectively). (C) Fully retracted embryo. Each segment has a regular set of Tc’wg

expressing cell clusters. Posterior segments are formed later and therefore lag in pattern maturation. (D) kno mutant embryo slightly older than the one in (A). In the

anterior abdomen, the ventral Tc’wg domains of adjacent segments have fused. (E, H) Embryos older than the one in (B). Up-regulation of the lateral Tc’wg

expression is obvious in both embryos (blue arrowhead), while ventral miss regulation is seen in H only (red arrowhead). Both embryos have shortened appendages

as compared to wild-type. (F, I) kno mutant embryos after retraction is completed. Up-regulation of the lateral Tc’wg clusters has become stronger (blue arrowhead),

while expression in the ventral domains has become more irregular (red arrowhead). In I, the Tc’wg pattern appears to break down completely to stochastically

distributed clusters of cells. (G) Dorsal closure stage. The Tc’wg pattern has degraded even more, while some Tc’en stripes appear rather normal. Since this embryo

has not been dissected the anterior and posterior extremes are out of focus. Tc’en expression: (J–L) In wild-type embryos Tc’en (purple) expression is visible in

stripes that span the entire width of the germ band. In knodel mutant embryos, the Tc’en pattern is less severely disturbed than the Tc’wg (green) pattern. (M, M 0, P)

In retracting knodel mutant embryos the Tc’en staining first fades in ventral tissue (black arrowhead in M 0). (N, O, Q) Later, also the lateral stripes become weaker

and more irregular. Note that there are no differences in expression of Tc’wg and Tc’en up to the extended germ band stage (not shown).

H. Wohlfrom et al. / Mechanisms of Development 123 (2006) 430–439 435

Fig. 4. SEM micrographs of the anterior portion of germ bands after retraction.

(A,B) Wild-type embryos at different time points. Note the outgrowing

appendages and the external signs of segmentation in both embryos (thoracic

segments: white arrowheads, labium: open arrowhead, maxilla: star, mandible:

circle, antenna: open arrow). (C) A kno mutant embryo at a similar stage. All

segments are specified but their borders are less clear than in wild-type. The

appendages are shorter and their surface is less smooth than in wild-type.

H. Wohlfrom et al. / Mechanisms of Development 123 (2006) 430–439436

also observed in mutants of other genes that are required for

maintaining epithelial polarity and adhesion as for instance

crumbs, DE-cadherin and some innexins (Tepass and Knust,

1993; Tepass et al., 1996; Uemura et al., 1996; Lehmann et al.,

2006). Similar to kno, these phenotypes result from degener-

ation of established tissues. However, there is no indication that

the epithelial structure of the ectoderm itself is affected in kno.

Still, kno could be a hypomorphic or a complex allele that

allows for enough gene function for maintaining the integrity

of the epithelium but interferes with the expression fine-tuning

that could be essential for differential adhesion within the

epithelium.

For the maintenance of Drosophila parasegments it is

crucial that cells do not cross the parasegment boundary. The

exact mechanisms by which this cell segregation is brought

about remain unclear. But it has been suggested that the cell

populations on both sides of the border differ in their adhesion

properties, which would make them stick together rather than

invade the other cell population. One possibility is that two

different adhesion molecules are expressed in the two cell

populations to be kept apart. Alternatively, different levels of

expression of the same molecule can similarly lead to cell

segregation. Unfortunately, the responsible molecules have not

been identified so far (Dahmann and Basler, 1999, 2000). In

principle, kno could be an allele of an adhesion molecule

involved in this process. Alternatively, kno might be a

regulator that directs different expression levels of the respect-

ive adhesion molecule in different cell populations. This aspect

would be especially interesting in the light of the extensive but

so far fruitless efforts to identify the respective molecules in

Drosophila.

4.3. EGFR and Notch signaling could be affected in knodel

Assuming that the Drosophila ortholog of kno has a similar

combination of functions, do the phenotypes of Drosophila

genes hint at which gene may be mutated in kno? kno mutant

larvae resemble those of the Drosophila EGFR (DER) mutant

faint little ball. Also this mutant results in a sac-like cuticle

with few external structures. Importantly, also the faint little

ball phenotype involves degeneration of previously established

primordia, and also certain CNS patterns are affected (Price et

al., 1989; Schejter and Shilo, 1989; Shilo, 1992). However, the

developmental paths leading to this phenotype are quite

different. In faint little ball mutant embryos, ventral tissue

gets lost by cell death, leading to dorsalized embryos without

ventral denticle belts (Raz and Shilo, 1993). The rather normal

width of kno germ bands argues against extensive ventral cell

death (Fig. 3, compare E,H with B and F,I with C). In faint little

ball, germ band retraction is impeded and head tissue is

degenerating. In kno mutant embryos, in contrast, neither

retraction nor head patterning are impeded. In part, these

differences may be due different modes of development

(involuted head and reduced extraembryonic tissue in Droso-

phila versus normal head and fully developed amnion and

serosa in Tribolium). But taken together, these differences lead

us to think that kno and faint little ball are most likely not due

to mutations in the same pathway.

In Drosophila the Notch pathway is involved in neuroblast

specification, sensory organ development and the formation of

boundaries, i.e. the dorso-ventral compartment boundary in the

wing disc, and the boundaries between appendage segments

(Lehmann et al., 1983; Artavanis-Tsakonas et al., 1995; de

Celis et al., 1998; Artavanis-Tsakonas et al., 1999). Moreover,

evidence has been put forward that this pathway is required for

ectodermal segmentation in other arthropods: segmentation in

the spider Cupiennius salei is disturbed if components of the

Notch pathway are knocked down by RNAi (Stollewerk et al.,

2003; Schoppmeier and Damen, 2005). In the grasshopper

Schistocerca gregaria, the negative regulator fringe may be

involved in maintaining segment boundaries, because it is

expressed in segmental stripes late in development (Dearden

Fig. 5. Expression of the appendage genes Tc’Distal-less (Tc’Dll) and Tc’dachshund (Tc’dac) suggest that proximo-distal pattern formation is not impeded in kno

mutant embryos. (A, B) In wild-type, Tc’Dll is expressed in the distal portion of every appendage except the mandibles. At later stages a ring appears in a more

proximal position. (D, E) In kno mutant embryos, leg morphology and Tc’Dll expression are not severely altered, even though the trunk segment pattern already has

degenerated. (C,F) Tc’dac is expressed in medial portions of the leg in both wild-type (C) and kno mutant embryos (F).

H. Wohlfrom et al. / Mechanisms of Development 123 (2006) 430–439 437

and Akam, 2000). In Drosophila, Notch is not involved in

segmentation of the ectoderm but plays a role in the segmental

patterning of the mesoderm (Tapanes-Castillo and Baylies,

2004). It is conceivable that Notch may have a role in body

segment boundary formation in Tribolium, which has been lost

in Drosophila. An involvement of kno in Notch signaling could

also explain leg phenotypes because this pathway is required

for establishing the joints of appendages in Drosophila

imaginal discs. However, the degeneration of the entire

appendages in kno argues for a function that is not restricted

to podomer boundaries. The nests of bracts found in kno could

represent a weak neurogenic phenotype. To explain the strong

late defects (i.e. PNS, segmentation and limb development) in

combination with weak early (neurogenesis) phenotypes one

might have to consider maternal contribution of the gene

product. This could specifically rescue early embryonic

requirements (i.e. neurogenesis in the CNS). This scenario is

unlikely, however, since cell proliferation in the growth zone of

short germ embryos should rapidly dilute maternal mRNAs

such that at least neurogenesis in the posterior abdomen likely

has to rely on zygotic functions.

4.4. knodel could be involved in wingless signaling

We prefer the hypothesis that knodel is a segment polarity

gene, probably involved in the wingless pathway. Indeed,

degeneration of a well-established segment pattern is a hall-

mark of segment-polarity phenotypes in Drosophila. For

example, the phenotype of the wingless antagonist naked

cuticle (nkd) (Zeng et al., 2000) resembles that of kno.

Drosophila nkd mutant larvae are much shorter than wild-

type, show no overt sign of segmentation and are devoid of

denticles. nkd is also involved in neurogenesis (Deshpande

et al., 2001). Similar to kno mutant embryos, engrailed and

wingless stripes are initiated normally in nkd mutant embryos,

but the pattern degrades after the elongated germ band stage is

reached, leading to ectopic wingless stripes. This is reminiscent

of the miss regulation of wingless in kno mutants. However, the

response of engrailed to the Drosophila nkd mutation differs

from our findings in Tribolium in that the engrailed stripes

become broader, which we have not observed in kno mutant

embryos. However, the kno mutants may not represent the

complete knockout of the gene, or the regulatory interactions in

Tribolium may be subtly different from Drosophila. Further

support for kno being involved in wingless signaling comes

from the effects of kno on the dorsal mesoderm Tc’eve positive

cell clusters: They become enlarged and partially fuse with

their neighbors (Fig. 2C and D). Exactly the same is observed

when wingless is experimentally evenly distributed in Droso-

phila embryos (Lawrence et al., 1995).

If kno represents a segment polarity gene, its appendage

defects would be a secondary consequence of the degeneration

of segmental patterns because the parasegment boundary runs

through the appendage. Alternatively, a segment polarity gene

could be mutated in kno that plays direct roles in both body

segmentation and leg patterning, as is the case with odd-

skipped in Drosophila (Hao et al., 2003). Mapping of the

mutant and comparison with the Tribolium genome sequence

H. Wohlfrom et al. / Mechanisms of Development 123 (2006) 430–439438

should clarify if this locus represents the homolog of a gene

that already has been studied in Drosophila. Alternatively, this

locus may represent a new gene only required in Tribolium, or

whose function has escaped attention in Drosophila.

kno is clearly a gene not required for the formation of

segment boundaries but for maintaining their integrity. This

shows that pattern formation and maintenance rely at least in

part on the action of different genes and/or on different

regulation of the same genes. A switch in regulation has for

instance been described for engrailed expression. Initially it is

wingless-dependent (Martinez Arias et al., 1988; Heemskerk

et al., 1991) while later maintenance of the striped pattern is

wingless independent but requires polycomb group genes

(DiNardo et al., 1988; Moazed and O’Farrell, 1992). In the

case of gooseberry, such different modes of regulation could be

assigned to separate ‘early’ and ‘late’ enhancer elements

(Li et al., 1993). By classical forward genetics, the mainten-

ance process is difficult to study because many genes are

involved also in preceding pattern formation. Their late

function is thus obscured by the early one. The availability of

the genomic sequence and the ease of reverse genetics (Brown

et al., 1999; Bucher et al., 2002; Klingler, 2004) make

Tribolium an ideal model system for studying the genes and

processes necessary for late development that are required for

maintenance of once established patterns.

References

Amin, A., Li, Y., Finkelstein, R., 1999. Hedgehog activates the EGF receptor

pathway during Drosophila head development. Development 126, 2623–

2630.

Artavanis-Tsakonas, S., Matsuno, K., Fortini, M.E., 1995. Notch signaling.

Science 268, 225–232.

Artavanis-Tsakonas, S., Rand, M.D., Lake, R.J., 1999. Notch signaling: cell

fate control and signal integration in development. Science 284, 770–776.

Beermann, A., Jay, D.G., Beeman, R.W., Hulskamp, M., Tautz, D., Jurgens, G.,

2001. The Short antennae gene of Tribolium is required for limb develop-

ment and encodes the orthologue of the Drosophila Distal-less protein.

Development 128, 287–297.

Berghammer, A.J., Klingler, M., Wimmer, E.A., 1999. A universal marker for

transgenic insects. Nature 402, 370–371.

Brown, S.J., Patel, N.H., Denell, R.E., 1994. Embryonic expression of the

single Tribolium engrailed homolog. Dev. Genet. 15, 7–18.

Brown, S.J., Mahaffey, J.P., Lorenzen, M.D., Denell, R.E., Mahaffey, J.W.,

1999. Using RNAi to investigate orthologous homeotic gene function

during development of distantly related insects. Evol. Dev. 1, 11–15.

Brown, S., Fellers, J., Shippy, T., Denell, R., Stauber, M., Schmidt-Ott, U.,

2001. A strategy for mapping bicoid on the phylogenetic tree. Curr. Biol.

11, R43–R44.

Bucher, G., Klingler, M., 2004. Divergent segmentation mechanism in the short

germ insect Tribolium revealed by giant expression and function. Develop-

ment 131, 1729–1740.

Bucher, G., Scholten, J., Klingler, M., 2002. Parental RNAi in Tribolium

(Coleoptera). Curr. Biol. 12, R85–R86.

Cerny, A.C., Bucher, G., Schroder, R., Klingler, M., 2005. Breakdown of

abdominal patterning in the Tribolium Kruppel mutant jaws. Development

132, 5353–63.

Dahmann, C., Basler, K., 1999. Compartment boundaries: at the edge of

development. Trends Genet. 15, 320–326.

Dahmann, C., Basler, K., 2000. Opposing transcriptional outputs of Hedgehog

signaling and engrailed control compartmental cell sorting at the Droso-

phila A/P boundary. Cell 100, 411–422.

Damen, W.G., 2002. Parasegmental organization of the spider embryo implies

that the parasegment is an evolutionary conserved entity in arthropod

embryogenesis. Development 129, 1239–1250.

Davis, G.K., Patel, N.H., 2002. Short, long, and beyond: molecular and

embryological approaches to insect segmentation. Annu. Rev. Entomol.

47, 669–699.

Dearden, P., Akam, M., 2000. A role for Fringe in segment morphogenesis but

not segment formation in the grasshopper, Schistocerca gregaria. Dev.

Genes Evol. 210, 329–336.

de Celis, J.F., Tyler, D.M., de Celis, J., Bray, S.J., 1998. Notch signalling

mediates segmentation of the Drosophila leg. Development 125, 4617–4626.

Deshpande, N., Dittrich, R., Technau, G.M., Urban, J., 2001. Successive

specification of Drosophila neuroblasts NB 6-4 and NB 7-3 depends on

interaction of the segment polarity genes wingless, gooseberry and naked

cuticle. Development 128, 3253–3261.

DiNardo, S., Sher, E., Heemskerk-Jongens, J., Kassis, J.A., O’Farrell, P.H.,

1988. Two-tiered regulation of spatially patterned engrailed gene

expression during Drosophila embryogenesis. Nature 332, 604–609.

DiNardo, S., Heemskerk, J., Dougan, S., O’Farrell, P.H., 1994. The making of a

maggot: patterning the Drosophila embryonic epidermis. Curr. Opin.

Genet. Dev. 4, 529–534.

Duman-Scheel, M., Patel, N.H., 1999. Analysis of molecular marker expression

reveals neuronal homology in distantly related arthropods. Development

126, 2327–2334.

Hao, I., Green, R.B., Dunaevsky, O., Lengyel, J.A., Rauskolb, C., 2003. The

odd-skipped family of zinc finger genes promotes Drosophila leg segmen-

tation. Dev. Biol. 263, 282–295.

Heemskerk, J., DiNardo, S., Kostriken, R., O’Farrell, P.H., 1991. Multiple

modes of engrailed regulation in the progression towards cell fate

determination. Nature 352, 404–410.

Hughes, S.C., Krause, H.M., 2001. Establishment and maintenance of para-

segmental compartments. Development 128, 1109–1118.

Klingler, M., 2004. Tribolium. Curr. Biol. 14, R639–R640.

Larsen, C.W., Hirst, E., Alexandre, C., Vincent, J.P., 2003. Segment boundary

formation in Drosophila embryos. Development 130, 5625–5635.

Lawrence, P.A., Bodmer, R., Vincent, J.P., 1995. Segmental patterning of heart

precursors in Drosophila. Development 121, 4303–4308.

Lehmann, R., Jimenez, F., Dietrich, U., Campos-Ortega, J.A., 1983. On the

phenotype and development of mutants of early neurogenesis in Drosophila

melanogaster. Wilhelm Roux Arch. Dev. Biol. 192, 62–74.

Lehmann, C., Lechner, H., Loer, B., Knieps, M., Herrmann, S., Famulok, M.,

et al., 2006. Heteromerization of innexin gap junction proteins regulates

epithelial tissue organization in Drosophila. Mol. Biol. Cell.

Li, X., Gutjahr, T., Noll, M., 1993. Separable regulatory elements mediate the

establishment and maintenance of cell states by the Drosophila segment-

polarity gene gooseberry. Eur. Mol. Biol. Org. J. 12, 1427–1436.

Liu, J., Qian, L., Wessells, R.J., Bidet, Y., Jagla, K., Bodmer, R., 2006.

Hedgehog and RAS pathways cooperate in the anterior–posterior specifi-

cation and positioning of cardiac progenitor cells. Dev. Biol. 290, 373–385.

Maderspacher, F., Bucher, G., Klingler, M., 1998. Pair-rule and gap gene

mutants in the flour beetle Tribolium castaneum. Dev. Genes Evol. 208,

558–568.

Martinez Arias, A., 1993. The development of Drosophila melanogaster. In:

Bate, M., Martinez Arias, A. (Eds.), Development and patterning of the

larval epidermis of Drosophila, vol. 1. Cold Spring Harbour Press, Cold

Spring Harbour, NY, pp. 517–608.

Martinez Arias, A., Baker, N.E., Ingham, P.W., 1988. Role of segment polarity

genes in the definition and maintenance of cell states in the Drosophila

embryo. Development 103, 157–170.

Moazed, D., O’Farrell, P.H., 1992. Maintenance of the engrailed expression

pattern by Polycomb group genes in Drosophila. Development 116, 805–

810.

Nagy, L.M., Carroll, S., 1994. Conservation of wingless patterning functions in

the short-germ embryos of Tribolium castaneum. Nature 367, 460–463.

Oppenheimer, D.I., MacNicol, A.M., Patel, N.H., 1999. Functional conserva-

tion of the wingless-engrailed interaction as shown by a widely applicable

baculovirus misexpression system. Curr. Biol. 9, 1288–1296.

H. Wohlfrom et al. / Mechanisms of Development 123 (2006) 430–439 439

Pankratz, M.J., Jackle, H., 1990. Making stripes in the Drosophila embryo.

Trends Genet. 6, 287–292.

Price, J.V., Clifford, R.J., Schupbach, T., 1989. The maternal ventralizing locus

torpedo is allelic to faint little ball, an embryonic lethal, and encodes the

Drosophila EGF receptor homolog. Cell 56, 1085–1092.

Prpic, N.M., Wigand, B., Damen, W.G., Klingler, M., 2001. Expression of

dachshund in wild-type and Distal-less mutant Tribolium corroborates

serial homologies in insect appendages. Dev. Genes Evol. 211, 467–477.

Raz, E., Shilo, B.Z., 1993. Establishment of ventral cell fates in the Drosophila

embryonic ectoderm requires DER, the EGF receptor homolog. Genes Dev.

7, 1937–1948.

Riechmann, V., Irion, U., Wilson, R., Grosskortenhaus, R., Leptin, M., 1997.

Control of cell fates and segmentation in the Drosophila mesoderm.

Development 124, 2915–2922.

Sanson, B., 2001. Generating patterns from fields of cells. Examples from

Drosophila segmentation. Eur. Mol. Biol. Org. Rep. 2, 1083–1088.

Schejter, E.D., Shilo, B.Z., 1989. The Drosophila EGF receptor homolog

(DER) gene is allelic to faint little ball, a locus essential for embryonic

development. Cell 56, 1093–1104.

Schoppmeier, M., Damen, W.G., 2005. Suppressor of Hairless and Presenilin

phenotypes imply involvement of canonical Notch-signalling in segmenta-

tion of the spider Cupiennius salei. Dev. Biol. 280, 211–224.

Shilo, B.Z., 1992. Roles of receptor tyrosine kinases in Drosophila develop-

ment. Fed. Am. Soc. Exp. Biol. J. 6, 2915–2922.

St Johnston, D., Nusslein-Volhard, C., 1992. The origin of pattern and polarity

in the Drosophila embryo. Cell 68, 201–219.

Stollewerk, A., Schoppmeier, M., Damen, W.G., 2003. Involvement of Notch

and Delta genes in spider segmentation. Nature 423, 863–865.

Sulston, I.A., Anderson, K.V., 1996. Embryonic patterning mutants of

Tribolium castaneum. Development 122, 805–814.

Tapanes-Castillo, A., Baylies, M.K., 2004. Notch signaling patterns Droso-

phila mesodermal segments by regulating the bHLH transcription factor

twist. Development 131, 2359–2372.

Tautz, D., Pfeifle, C., 1989. A non-radioactive in situ hybridization method for

the localization of specific RNAs in Drosophila embryos reveals trans-

lational control of the segmentation gene hunchback. Chromosoma 98,

81–85.

Tautz, D., Friedrich, M., Schroder, R., 1994. Insect embryogenesis—what is

ancestral and what is derived? Dev Suppl, 193–199.

Tepass, U., Knust, E., 1993. Crumbs and stardust act in a genetic pathway that

controls the organization of epithelia in Drosophila melanogaster. Dev.

Biol. 159, 311–326.

Tepass, U., Gruszynski-DeFeo, E., Haag, T.A., Omatyar, L., Torok, T.,

Hartenstein, V., 1996. Shotgun encodes Drosophila E-cadherin and is

preferentially required during cell rearrangement in the neurectoderm and

other morphogenetically active epithelia. Genes Dev. 10, 672–685.

Uemura, T., Oda, H., Kraut, R., Hayashi, S., Kotaoka, Y., Takeichi, M., 1996.

Zygotic Drosophila E-cadherin expression is required for processes of

dynamic epithelial cell rearrangement in the Drosophila embryo. Genes

Dev. 10, 659–671.

Zeng, W., Wharton Jr., K.A., Mack, J.A., Wang, K., Gadbaw, M., Suyama, K.,

et al., 2000. Naked cuticle encodes an inducible antagonist of Wnt

signalling. Nature 403, 789–795.