Human Molecular Genetics, 2003, Vol. 12, No. 22
DOI: 10.1093/hmg/ddg325
2931–2940
Inducible mEDA-A1 transgene mediates
sebaceous gland hyperplasia and differential
formation of two types of mouse hair follicles
Chang-Yi Cui1, Meredith Durmowicz1, Chris Ottolenghi1, Tsuyoshi Hashimoto1,
Bradley Griggs2, Anand K. Srivastava2 and David Schlessinger1,*
1
Laboratory of Genetics, NIH/National Institute on Aging, Baltimore, MD 21224, USA and 2JC Self Research
Institute of Human Genetics, Greenwood Genetic Center, Greenwood, SC 29646, USA
Received June 25, 2003; Revised and Accepted September 15, 2003
EDA splice isoforms EDA-A1 and EDA-A2 belong to the TNF ligand family and regulate skin appendage
formation by activating NF-kB- and JNK- promoted transcription. To analyze their action further, we
conditionally expressed the isoforms as tetracycline (‘Tet’)-regulated transgenes in Tabby (EDA-negative)
and wild-type mice. Expression of only the mEDA-A1 transgene had two types of effects during
embryogenesis: (1) determinative effects on sweat glands and hair follicles. In Tabby mice, one type of
hair follicle (‘guard hair’) was restored, whereas a second type, the dominant undercoat hair follicle (‘zigzag’)
was not; furthermore, the transgene sharply suppressed zigzag hair formation in wild-type mice, with the
overall numbers of back hair follicles remaining the same; and (2) trophic effects on sebaceous and
Meibomian glands. Marked hyperplasia resulted from expansion of the sebocyte-producing zone in
sebaceous glands, with particularly high expression of the transgene and the replication marker PCNA,
and correspondingly high production of sebum. The phenotypic effects of mEDA-A1 on sebaceous glands,
but not on hair follicles, were reversed when the gene was repressed in adult animals. The results thus reveal
both initiating and trophic isoform-specific effects of the EDA gene, and suggest a possible balance of
isoform interactions in skin appendage formation.
INTRODUCTION
As in many developmental processes, the formation of skin
appendages involves a tissue-specific route to the activation of
major signaling pathways. High specificity for skin appendage
formation is based on ectodysplasin, the protein encoded by
the X-linked EDA gene (1,2). It intervenes in skin appendage
development as a novel TNF family ligand, activating NF-kB
transcription factors through EDAR/XEDAR receptors, an
EDARADD receptor adaptor and TRAF signal proteins (3–10).
Consistent with the first steps in a coherent pathway, mutations
in EDA cause the human disorder anhidrotic ectodermal
dysplasia (‘EDA’; OMIM305100) and its mouse equivalent,
‘Tabby’, and mutations in the EDA receptor/adaptor genes
cause similar phenotypes (6–14). Thus, the pathway is pivotal in
skin appendage formation, but its details remain largely unknown.
In the complete absence of EDA expression, sweat glands are
totally missing, some teeth are rudimentary or aberrant, and
two types of back hair follicles, ‘guard’ and ‘zigzag’ are not
formed. However, the third type of back hair, ‘medium’
(including awls and auchenes), is still formed in Tabby, and is
even more numerous compared to wild-type littermates (15–17).
Recent evidence suggests that a complex role for EDA
derives in part from the action of several isoforms. We found
that moderate levels of one of the longest isoforms, mEDA-A1,
can largely restore missing appendages (15), with the
accompanying selective augmentation of expression of the
downstream NF-kB and JNK pathways and a number of newly
identified downstream target genes (18). However, it was not
clear when mEDA-A1 action was critical during mouse
development (or thereafter). Also, because some appendages
were not restored, it seemed possible that mEDA-A1 action was
dose-dependent, or that formation of other appendages might
require other isoforms.
To determine the timing, dose-dependence, and reversibility
of EDA action, we generated both Tabby and wild-type mice in
*To whom correspondence should be addressed at: Laboratory of Genetics, NIH/National Institute on Aging, 333 Cassell Dr., Suite 3000, Baltimore,
MD 21224, USA. Tel: þ1 4105588337; Fax: þ1 4105588331; Email: [email protected]
2932
Human Molecular Genetics, 2003, Vol. 12, No. 22
which transgenic mEDA-A1 or mEDA-A2 isoforms could be
conditionally expressed up to high levels from a tetracyclineregulated promoter. Consistent with recent reports (19,20) (see
below), we saw no effect of mEDA-A2 on its own; but mEDAA1 alone had two kinds of effects on particular appendages: it
commited precursor cells to form sweat glands and the longest
(‘guard’) hair, and repressed the alternative pathway to form
zigzag hair follicles; and in a trophic role, it provoked the
proliferation of sebaceous gland progenitors. The sebaceous
gland hypertrophy, but not the determination of hair follicles,
was reversed when the gene was shut down after birth.
RESULTS
To facilitate the recognition of time- and isoform-dependent
actions of EDA in embryonic and adult mice, we have used a
controllable, double transgenic model.
Mice transgenic for Tet-regulated EDA/EDA
pathway genes
The two longest EDA isoforms, EDA-A1 and EDA-A2, are the
most common, and the only ones that include a TNF-ligand
domain. Though differing by only two amino acids, they have
distinct corresponding receptors, EDAR and XEDAR (7–9).
Transgene constructs for the two mouse isoforms are
diagrammed in Figure 1A, and were confirmed in animals by
PCR genotyping. Expression of these constructs is negatively
regulated by the Tet promoter, that is, EDA is repressed by the
addition of doxicycline to cells or when fed to animals as a
dietary supplement, and is expressed in doxicycline’s absence.
Appropriate regulation was confirmed by transient transfection
assays in cell culture. Regulation of mRNA levels was
confirmed in transgenic mice (Fig. 1B) by real-time RT–PCR
with RNAs from skin samples (see Materials and Methods).
Their level was increased about 1000-fold in the mEDA-A1
transgenic mice in either Tabby or wild-type background, and
returned to basal levels when the mice were given food
containing doxicycline (200 mg/kg). In comparable experiments, transgenic mEDA-A2 expression was 50-fold higher
than in wild-type mice (Fig. 1B, A2 on) and also returned
to basal levels in the presence of doxicycline. As shown in
Figure 1C, for mEDA-A1, protein expression was apparent in
the ‘Tet-’ condition (lanes 2 and 3), and was completely
inhibited after the addition of 2 mg/ml doxicycline to culture
medium (lanes 4 and 5). Thus, the gene can be selectively
turned on or off at any stage of development or adulthood.
mEDA-A1 transgene restored guard hair, but
not zigzag hair, in Tabby mice
Tabby mice bearing the mEDA-A1 transgene under tetracycline
control showed striking phenotypes in the ‘on’ state (in the
absence of doxicycline; see Materials and Methods). Table 1
summarizes the effects. Some partial restorative effects
observed previously with lower-level expression of the isoform
(15) were more complete with the higher levels attained here.
They included fully restored hair behind the ears, and darker
and thicker but somewhat ‘scruffy’ disordered coat hair
Figure 1. Constructs, expression, and regulation of transgenic mEDA-A1 and
mEDA-A2. (A) The coding regions of mEDA-A1 and mEDA-A2 cDNAs were
inserted between TRE/PminCMV promoter and SV40 poly A region (see
Materials and Methods). The tetracycline-controlled transactivator (tTA) binds
to TRE region of mEDA transgenes and activates transcription. (B) Quantitative
RT–PCR shows elevated expression of mEDA-A1 and mEDA-A2 in transgenic
mouse skin, and shut-down of transgene expression by doxicycline chow
administration. Note exponential scale of expression; the Tabby mutation is a
single base change, so that Ta and WT mRNA levels are equal. (C) EDA
protein expression from transient transfection of mEDA-A1 into MCF-7 Tetoff cell line. Tet- condition (lanes 2 and 3) and shut-down in Tetþ condition
(lanes 4 and 5). Lane 1, transfected with pTRE vector only and cultured
under Tet condition. Two bands about 46 kDa are glycosylated EDA protein (40).
[compare WT(mEDA-A1) and Ta(mEDA-A1) with WT and
Ta in (A–D) of Fig. 2].
The analysis of transgene effects was extended to quantitate
hair follicles and subtypes of hair. Tabby male mice lack both
guard (‘G’) and zigzag (‘Z’) hair, although the other major
class, medium-length (‘M’) hair, is present at higher levels
than in wild-type mice (15,16) (see examples in Fig. 3C). To
assess possible differential effects, hair was harvested by
shaving equivalent areas of back skin of wild-type, Tabby
and mEDA-A1 transgenic mice and counting the numbers of
G, M and Z hairs under a stereomicroscope. Wild-type mice
have a population of hairs containing about 5.3 0.4% G,
27.0 0.9% M and 67.7 0.5% Z; in agreement with other
recent studies (16,21), Tabby male mice have no G or Z hair,
but more M hair is formed in Tabby than in wild-type mice, so
that the overall number of hair follicles is the same (Fig. 3A).
In male mEDA-A1 transgenic Tabby mice, the number of M
hairs remains comparable to Tabby, but G hair is restored to
wild-type levels (6.6% of total hair). In contrast, however, no Z
Human Molecular Genetics, 2003, Vol. 12, No. 22
2933
Table 1. Gross phenotypes of mEDA-A1 transgenic mice in Tabby or wild-type
background
Restoration of sweat but not preputial
glands by mEDA-A1
Gross phenotypes
WT
Ta
Ta(A1)
WT(A1)
Ta(A2)
Hair coat
Dark,
shiny,
ordered
Yellowish,
short, thin
Dark,
long,
angled
Yellowish,
short, thin
þ
Dark,
long,
angled
þ
þ
þ
þ
þ
þ
þ
þ
þ/
þ
þ
þ
þ
þ
þ
þ
þ
þ
þ
þ
3
100
2–3
77
3
80
3
101
2–3
82
Like coat- and ear-associated hair, sweat glands were more fully
developed with the mEDA-A1 transgene under tetracycline
control. Positive results with the starch-iodine test (15) were
confirmed in histological sections, exemplified in Figure 4
(‘S’ panels). Tabby footpad sections showed only dermis and
expanded muscle tissue, but sections from Tabby animals
bearing the mEDA-A1 transgene showed the restoration of
clearly defined sweat glands, very similar to wild-type mice
with or without the transgene.
In contrast, the supragonadal preputial gland, which secretes
pheromones (22), is absent in Tabby mice, which retain only an
associated fat pad, and is not at all restored by the mEDA-A1
transgene. Nor does an added transgene change the appearance
of the gland of wild-type animals [Fig. 4 (‘P’ panels) shows
both hematoxylin–eosin staining of sections and photographs of
the dissected surface of the gland regions, with the fat pads
outlined in black]. Thus, as in the case of hair follicles, mEDAA1 has a selective determinative effect on some exocrine glands.
Bald patch
behind ear
Tail hair
Tail skin ridges
Vibrissae
Sweat glands
Preputial glands
Meibomian glands
Blindness by
9 months
Teeth (molars)
Weight (%)a
þ
(þ) Yes or normal; () no or missing; (þ/) partial.
a
Four male mice [three for Ta-(A2)], 5 months old, were weighed for each strain.
hairs at all were found on the backs of the six transgenic male
mice from two independent transgenic lines (Fig. 3B).
Interestingly, in these animals as well, although the composition of hair types was dramatically changed, the total hair
numbers were unchanged in transgenic mouse skin, either in
Tabby or in wild-type background (Fig. 3A and B).
Other subtle differences between wild-type and mEDA-A1
transgenic mice were also seen. Tail hairs were reduced in
number and length, and the tail-specific ridges on skin surfaces,
which are prominent in wild-type mice but absent in Tabby
mice, are not restored in transgenic Tabby mice (Table 1). One
other measured phenotype is, like Z hair, completely unrestored: Tabby mice are smaller than wild-type littermates, and
male transgenic mice were as small as Tabby. At 5 months, the
weights of transgenic and Tabby mice were about 80.0 1.7
and 76.9 1.8% of wild type, respectively (Table 1).
Zigzag hair formation was suppressed in wild-type
mice carrying an mEDA-A1 transgene
The failure of Z hair follicle formation in mEDA-A1 transgenic
Tabby mice was not simply a result of a lack of capacity of the
A1 isoform to initiate those follicles. Greater complexity was
revealed when the transgene was introduced into wild-type
mice (which were identified by SSCP) (15). In the presence of a
fully intact EDA gene, overall skin, tail, and sweat gland
function were indistinguishable from mice lacking the
transgene (Fig. 2). However, the distribution and numbers of
hair follicle types on the backs of the transgenic animals were
markedly affected. G hair numbers were somewhat increased
(to 8.7% compared with 5–6% in wild-type); but strikingly, Z
hair was reduced to about 6% (range 0–11%), at least 6-fold
less than in wild-type (Fig. 3B). These results suggest that
mEDA-A1 is not only unable to initiate Z hair follicles, but also
suppresses their formation (see Discussion). Furthermore, Z
hair was not restored when the A1 transgene was turned off in adult
wild-type transgenic skin (data not shown). This is again consistent
with a critical irreversible role for EDA in embryonic stages.
Hyperplastic trophic effect of transgenic mEDA-A1 on
sebaceous and Meibomian glands
Histological studies revealed a hypertrophic effect of mEDAA1 on several types of sebaceous glands. One, the Meibomian
gland, is a variant sebaceous gland that is attached to the inner
surface of the eyelid and produces lipid-rich secretions (22). It
was lacking in Tabby mice and was not restored by an mEDAA1 transgene (Fig. 4, panel M), which may account for the
progressive blindness of the aging mice (noted in Table 1).
However, the prominent gland of wild-type mice was
consistently larger in the presence of an active mEDA-A1
transgene (Fig. 4, panel M).
An equivalent or even more marked change was seen in
sebaceous glands of transgenic mouse skin, and was analyzed
further. Sebaceous glands were 200–300% larger than in wildtype or Tabby mice. The size of individual sebocytes was the
same (Fig. 5A–C), but the numbers were increased, with a
correspondingly excessive production of sebum (Fig. 5D–F).
Normal G follicles have two associated sebaceous glands; M
follicles have only one (21). Hyperplasia was obvious in both
glands associated with each de novo G follicle (Fig. 5), and also
in the single sebaceous gland associated with each M follicle.
Thus, the effect is exerted distal to the commitment to
formation of hair follicle subtypes. Furthermore, because
functional sebaceous glands are formed on M hair follicles in
the absence of active EDA in Tabby mice (Fig. 5B and E), the
effect of the mEDA-A1 transgene is apparently trophic rather
than morphogenetic.
The larger sebaceous glands could result from either a slower
turnover of sebocytes or an increased formation and flux of
sebocytes through their life cycle. An effect on the formation of
sebocytes was consistent with increased numbers of multiple
immature sebocytes, which are sparse in wild-type or Tabby,
but are numerous in the transgenic animals (Fig. 5C, circled,
diagrammed in Fig. 6B). [Interestingly, the mEDA-A1
transgene is highly expressed in the bottom generative layer
of sebaceous glands (Fig. 6A).]
2934
Human Molecular Genetics, 2003, Vol. 12, No. 22
Figure 2. Appearance of mEDA-A1 transgenic mice. A1 transgenic mice in Tabby background [Ta(mEDA-A1)] showed darker, longer, disoriented hair and
restored hair patches behind ears. Transgenic mice in wild-type background were similar. Photographs by Olympus C-700 digital camera.
Immunostaining with the proliferation marker PCNA further
supported increased formation of mature sebocytes as the basis
for the hypertrophy. PCNA is, like mEDA-A1, highly
expressed in the generative layer, and is down-regulated when
transgene expression is shut off (circled region in Fig. 6C–F).
Thus, the large sebaceous glands in transgenic mice probably
result from EDA activation of hyperproliferation of sebocyte
progenitors located at the base of mouse sebaceous glands. In
good agreement with this scenario, sebum production,
visualized by Oil Red O staining, gradually increases from
the lower, proliferating layer to the upper differentiated zone of
transgenic sebaceous glands, i.e. along the proliferation/
maturation axis (Figs 5F and 6B).
The reversible Tet-regulation of the transgene permitted us to
adduce more evidence regarding a trophic effect. When
transgene function was shut down through the addition of
200 mg/kg doxicycline to the normal diet of 3-month-old
transgenic mice, we found that the hyperplasia of sebaceous
glands was reversed (Fig. 5C). We confirmed relatively
complete shutdown of transgene expression in transgenic skin
by quantitative real-time RT–PCR (Fig. 1B, A1 off).
Histological examination showed that hyperplasia of sebaceous
glands was partially reversed within one month of the shutoff
of the gene, and after 3 months effects were more pronounced;
the remaining sebaceous glands decreased in size and were
comparable to the size of those in wild-type mice (insert in
Fig. 5C). This suggests a requirement for continuous expression of the mEDA-A1 isoform for sebaceous gland hyperplasia.
Although sebaceous glands diminished in size after doxicycline administration, the number of guard hairs remained
unchanged (data not shown). This suggests that the A1 isoform
exerts its primary effect on guard hair synthesis, with little or
no influence on its maintenance. However, the upright, rather
perpendicular follicular angle to the skin’s surface in transgenic
animals is completely reverted to the wild-type angle (insert in
Fig. 5C). It is speculatively possible that the enlarged
sebaceous glands distort the angle of the hair shaft.
Induction of mEDA-A1 expression in adult Tabby mice
did not rescue Tabby phenotypes
The time at which mEDA-A1 affects various skin appendages
could also be tested through controlled expression of the
transgene. When mice were given doxicycline throughout
pregnancy, male transgenic progeny were essentially identical
to Tabby littermates. In particular, they showed the same lack
of sweat glands, tail hair, guard hair and zigzag hair;
subsequent withdrawal of doxicycline from food (at 6 weeks
of age) and resumption of A1 expression showed that these
mice remained like their Tabby littermates after 2, 4, 6, 8 or 10
weeks of re-expression (data not shown). These results are
consistent with requisite EDA function in early embryonic life.
mEDA-A2 transgene failed to complement
Tabby phenotypes
Because the mEDA-A2 isoform also activates NF-kB
transcription factors through its own receptor, XEDAR (8),
during skin appendage development, we expected that it
might function analogously to EDA-A1 as a transgene (15).
Instead, Tabby mice bearing the mEDA-A2 transgene were
indistinguishable from transgene-less Tabby littermates; and
transgenic wild-type mice were like wild-type mice (Table 1).
mEDA-A2 transgene expression was 50-fold higher in
transgenic skin compared with wild-type (Fig. 1B). The
corresponding level of protein, which is low in wild-type and
Human Molecular Genetics, 2003, Vol. 12, No. 22
2935
possibly the absence of overt toxic effects). Thus we generated
Tet-EDA-A1 transgenic mice and, for comparison, Tet-mice with
transgenic A2, -B and -C isoforms. The Tet-EDA transgenic
system provides high levels of expression of EDA isoforms in the
skin epithelium that are sharply controlled by doxicycline.
Correspondingly sharp phenotypic changes, again without overt
toxicity, were observed with mEDA-A1 alone. We focus here on
two types of observed effects, determinative and trophic.
Positive for sweat but not preputial glands, and
G but not Z follicle formation
Figure 3. Total hair number and ratio of hair subtypes in mEDA-A1 transgenic
mice. (A) Hair number in equivalent areas of back skin. Three independent
mice for each strain, and three independent areas for each mouse, were counted.
(B) Relative proportions of G, M and Z hair. (C) Three major hair subtypes
from wild-type mice. Tabby male mice lack G and Z hairs.
transgenic animals (20), was not directly measured, so that
inferences are indirect; in ongoing work, real time RT–PCR
assays showed that the EDA-A2 XEDAR receptor was slightly
higher in transgenic skin, and the findings listed in Table 1 are
consistent with recently reported negative results (19,20).
DISCUSSION
In the transgenic Tabby mice that we studied originally (15), in
spite of the use of the ubiquitous CMV-promoter, mEDA-A1
partly rescued the Tabby phenotype without causing detectable
abnormalities in the skin or elsewhere. It seemed possible that
other isoforms of EDA might be involved in rescue of some
features, or, alternatively, because the CMV-mEDA-A1 isoform
expression was lower than wild-type levels, low expression of
the transgene might account for the limited rescue (and
Tet regulation permits control of EDA expression over a wide
range. Many phenotypic effects of mEDA-A1 expressed up to
1000-fold greater than wild-type levels were qualitatively
comparable to those seen earlier with levels about 60–90% of
wild-type (e.g. the restoration of sweat glands [Fig. 4, top
panels, and cf. Table 1 in Srivastava et al. (15)]. In general,
determinative effects were selective at all levels of expression.
Thus, unlike sweat glands, exocrine preputial glands were not
restored even at high levels of EDA-A1. However, quantitation
and additional histological and timed observations revealed
additional features of mEDA-A1 action. Strikingly, in the
presence or absence of high levels of the isoform, the absolute
number of hair follicles per square cm of back skin was
the same in Tabby, wild-type, and transgenic mice; but the
proportions of subtypes were changed (Fig. 3). In the presence
of mEDA-A1, G hair was fully restored to Tabby mice, and
relatively high numbers of M follicles were still retained, but
the dominant undercoat hair, Z, was still missing.
Overexpression of the mEDA-A1 transgene in a wild-type
background revealed that it not only failed to induce Z hair
follicles, but also could repress their formation. [The low
expression from the CMV promoter showed no comparable
effect (15).] The reduction of Z follicles was compensated by
an increase of M follicles to maintain the constant overall
number of follicles. Thus, mEDA-A1 at high levels may block
a transition at E17, when Z follicles are formed. Perhaps EDAA1 alters the commitment of follicles to Z form, or a negative
trophic effect on follicles causes them to revert to a ‘default M
pathway’. Additional information should come from examination of embryonic stages in the Tet-mEDA-A1 transgenic mice,
although the experiments are cumbersome, because only a few
male progeny of crosses contain the Tet gene along with the
Tabby mutation and the transgene.
It is notable that expression of mEDA-A1 did not change
overall hair follicle number and had no effect on follicle
formation or structure after birth. Nor did the expression of
mEDA-A1 (or mEDA-A2) lead to tumorigenesis, as has been
reported for hair follicle morphogens such as transgenic Wnt,
SHH, Notch and FGFs (23–25). Mutant and knockout mouse
studies indicate that different types of hair have different
requirements for the Wnt and EDA pathways as well as for
BMPs, Noggin and SHH (11,12,26–32). Anomalies in these
mutants might reflect redundant roles for Wnt and EDA
pathways, or alterations of a single gene network with various
compensatory mechanisms active at different times during
development. However, the restriction of activity of the EDA
pathway, but not the Wnt pathway, to embryo-fetal action,
coupled with the lack of tumorigenicity of EDA, suggests that
2936
Human Molecular Genetics, 2003, Vol. 12, No. 22
Figure 4. Differential effects of mEDA-A1 on three types of gland. (S) Determinative effect restoring sweat glands to Tabby [Ta(mEDA-A1)] mice with no additional effect in a wild-type [WT(mEDA-A1)] background. Hematoxylin-eosin staining. (P) Lack of restoration of preputial gland to Tabby mice, with no additional
effect in wild-type. Paired subpanels show hematoxylin-eosin stained sections (left) and digital photographs of the surface of the gland region (right). The dashed
black line outlines the overlying fat pad. (M) Lack of restoration of Meibomian gland to Ta, but trophic effect on gland in wild-type mice. Scale bars, 0.1 mm.
the pathways are not redundant, but affect distinct features of
hair follicles in the context of a single gene network.
Trophic effect on sebaceous glands
A more clearly trophic, and positive, effect of mEDA-A1 was
seen in sebaceous glands. Meibomian glands provide an
instance in which determinative and trophic effects of
mEDA-A1 are clearly separated. The transgene did not restore
the gland to Tabby animals, but it increased the size of glands
in wild-type mice (Fig 4, ‘M’ panels). In a comparable way, the
glands associated with G follicles obviously depend secondarily for their formation on mEDA-A1, but the augmentation of
expression from the Tet-regulated promoter led to proliferation
of sebocytes in wild-type as well as Tabby animals. It is of
interest that, although M follicles and their single associated
sebaceous glands form in Tabby mice in the complete absence
of EDA, the glands nevertheless respond to high levels of
mEDA-A1 with hyperplasia. Strikingly, multiple proliferating
cells in transgenic sebaceous glands, which were sparse in wild
or Tabby mice, formed a ‘tail-like’ structure in the bottom layer
where the A1 transgene is most highly expressed (Fig. 6). We
infer that the layer contains progenitor or partially committed
stem cells for sebocytes.
Consistent action of mEDA-A1 provided as
transgene or protein ligand
Two publications that appeared during the preparation of this
report used either injection of recombinant mEDA-A1 protein,
containing a heterologous multimerization domain, into Tabby
mice (19) or promotion of mEDA-A1 from a skin-specific
keratin 14 (K14) promoter in wild-type mice (20) to study its
effects. Possibly resulting from the different modes of provision
of the protein, there are some inconsistencies in the results.
For example, free ligand—in contrast to mEDA-A1 transgene
(Fig. 4, ‘M’ panels)—supported the restoration of Meibomian
glands, and numbers of molars were somewhat variable
between approaches, but other phenotypes that were scored
in different studies (Table 1) were generally in accord.
The results with soluble mEDA-A1 over a range of
administered doses have practical implications for possible
intervention (19), because the lack of toxic effects and the
correction of the Tabby phenotype were very similar to that
reported for the CMV-promoted isoform (15). Also, in the
study with K14 promotion in a wild-type background, which
would be expected to magnify any toxic effects, high expression
in the epidermal basal layer produced only mild anomalies
(occasional supernumerary molars or nipples, enamel hypoplasia, absence of zigzag hair, and enlarged sebaceous glands).
In the double transgenic model that we describe here, high
levels of Tet-promoted mEDA-A1 were studied in both Tabby
and wild-type mice. In contrast to the studies with protein
ligand or K14 promotion, Tet control permitted the gene to be
selectively turned on or off at invervals in adult or embryonic
mice. High levels of embryonic expression correlated with
more complete rescue of some features of the Tabby phenotype
(e.g. sweat glands, guard hair, molars) than we had seen with
CMV-promoted isoform. As in the study with K14-promoted
expression, anomalies of phenotype were mild, but notably
Human Molecular Genetics, 2003, Vol. 12, No. 22
2937
Figure 5. Hyperplasia of sebaceous glands in mEDA-A1 transgenic mice. Panels represent hematoxylin staining, with Oil Red O staining added for (D–F), for
wild-type, Tabby, and transgenic Tabby skin. Sebaceous gland hyperplasia and follicle angle in de novo guard hair follicle were reversed by doxicycline food application in transgenic Tabby mice (C, insert). Accumulation of young sebocytes in the bottom of transgenic sebaceous glands (C, circled). Oil-Red O staining for
sebum. Scale bars, 0.05 mm.
zigzag hair was again suppressed and sebaceous glands were
enlarged. In addition, we found that the suppression of zigzag
hair occurs with the maintenance of a conserved overall
number of hair follicles, and, using the conditional promoter,
that the effects of mEDA-A1 on hair follicles required
expression in utero (19), and were thereafter irreversible when
the transgene was turned off. We also found that sebaceous
gland hypertrophy was due to increased progenitor cell
multiplication, and the trophic effect was reversible when the
transgene is later repressed.
Taken together, the various assay systems suggest that the
specificity of EDA action, and its activity across a wide range
of levels without toxicity, is probably due to regulation through
specific downstream effectors, including the receptor EDAR.
Requirement for other isoforms, and determinative
versus trophic action
How can one now rationalize the features of the Tabby
mouse that are not corrected by mEDA-A1? They include
the somewhat smaller size of Tabby mice (Table 1) and the
failure to restore preputial glands (Fig. 4, middle panels ‘P’).
The complete gene, in contrast to mEDA-A1, obviously
corrects these deficiencies and also, for example, does not
inhibit Z hair follicle formation. The differences could most
simply result from the absence of balanced expression of other
EDA isoforms when only mEDA-A1 is supplied, or when it is
in excess in wild-type mice. We (like 19,20) find no restoration
of hair follicles or other skin appendages by mEDA-A2
alone, and no pathogenetic mutations have been reported in
mEDA-A2 or its receptor (4,33,34). Thus, a pathway initiated
by EDA-A2 may have no leading role in skin appendage
formation (34). Nevertheless, mEDA-A2 itself can also
stimulate NF-kB transcription factors in vitro, and it is
suggestively selectively expressed at the time when Z hair is
produced (35). Similarly, other isoforms, mEDA-B and
mEDA-C (2), expressed under Tet control (work in progress)
have thus far themselves shown no phenotypic effects. But the
study of time- and dose-dependent expression of Tet-regulated
isoforms acting in conjunction may permit a more complete
definition of important interactions.
Finally, although studies have generally focused on other skin
appendages where EDA has a determinative effect, the
hypertrophic effect on sebaceous glands adds a possible
proliferative action to its potential roles. The rate of formation
of sebocytes can apparently be increased by EDA action (Figs 5
and 6). It is provocative that signaling through beta-catenin,
which interacts with the EDA pathway, has now been
independently implicated in sebaceous tumors and the
proliferation of undifferentiated sebocytes (36). Furthermore,
the sebocytes stimulated by mEDA-A1 are productively
making sebum. The process, presumably mediated by TNF
downstream of EDA, may be speculatively involved in acne or
other instances of altered sebum production, and could be a
target for intervention.
MATERIALS AND METHODS
Generation of transgene constructs
pTRE-mEDA-A1 construct. The construct is schematized in
Figure 1A. A 1.2 kb cDNA fragment spanning the entire open
reading frame (ORF) of mouse mEDA-A1 was generated via
PCR, using a full-length mEDA-A1 cDNA (GenBank no.
2938
Human Molecular Genetics, 2003, Vol. 12, No. 22
pTRE-mEDA-A2 construct. The mEDA-A2 isoform is only six
nucleotides shorter than the A1 isoform (35), so that we could
use high fidelity PCR amplification with a Quickchange sitedirected mutagenesis kit (Stratagene) to generate the pTREmEDA-A2 construct. The sense primer lacking the 6 nucleotides
specific for the A1 isoform was 50 -GCACCTACTTCATCTATAGTCAGGTCTACTACATCAACTTCACTG-30 , and the
antisense primer was 50 -CAGTGAAGTTGATGTAGTAGACCTGACTATAGATGAAGTAGGTGC-30 . After PCR (20 cycles
with 95 C, 0.5 min, 55 C, 1 min, 68 C, 10 min), the PCR
template pTRE-mEDA-A1 (methylated) was eliminated by DpnI
enzyme digestion, and the PCR product pTRE-mEDA-A2 was
propagated in E. coli. The pTRE-mEDA-A2 sequence was
verified by sequencing and microinjected as described above.
Cell transfection and western blotting
Figure 6. EDA expression and higher replication in sebocytes of transgenic
mice. (A) In situ hybridization shows high expression of mEDA-A1 transgene
in the lower portion of sebaceous glands [Ta(mEDA-A1)]. (B) Schematic of
sebaceous gland proliferation and differentiation in transgenic mice. (C, D)
Control PCNA staining of wild-type (C) and Tabby (D) sebaceous glands.
PCNA is elevated in sebaceous glands of Ta(mEDA-A1) mice (E), and
down-regulated when transgene is off (F). Immunostaining as in Materials
and Methods. Scale bars, 0.05 mm.
The expression competence of constructs was confirmed by
transient transfections into the MCF-7 Tet-off cell line
(Clontech). Three micrograms of construct were mixed with
30 ml of DOTAP liposomal transfection reagent (Boehringer
Mannheim, Indianapolis, IN, USA) and transfected into cells in
a 6 cm culture dish. Cells were cultured either in medium
lacking or containing 2 mg/ml doxicycline for up to 48 h, and
were then dissolved in Laemmli gel sample buffer. Proteins
were fractionated in 10% SDS/polyacrylamide gel electrophoresis and then transferred to nitrocellulose membrane. Antibody
raised to the C-terminal peptide of EDA (38) was used as a
primary antibody and the reactive bands were detected via an
Enhanced Chemiluminescence kit (ECL, Amersham Life
Sciences, Little Chalfont, Bucks, UK).
Generation of transgenic mice
AF016628) as template, and cloned into the pPCR-Script Amp
SK (þ) cloning plasmid (Stratagene, West Cedar Creek, TX,
USA). The targeted sense primer sequence for PCR
was 50 -GTAGCCACCATGGGCTACCCA GAGGTAG-30 , containing a strong Kozak consensus ribosome binding site
(underlined) (37), and the antisense primer was 50 -CATGGGCCAGGATGGAATG-30 . The amplified cDNA fragment
included the ORF and 26 bp of the 30 -untranslated region.
The cDNA fragment was ligated into the pTRE vector
(Clontech, Palo Alto, CA, USA). Orientation of the cDNA
insert was confirmed by PvuII digestion and subsequent electrophoresis, followed by sequencing on an ABI Prism 377
DNA Sequencer. Primer 1 for sequencing was complementary
to the pTRE vector, 100 bp upstream of translation start site
(50 -CGCCTGGAGACGCCATCC-30 ). Primer 2 starts at the
333 bp position, in the mEDA-A1 ORF (50 -CAACAGCAGCCTTTGGAGCCG-30 ), and primer 3 begins 718 bp into the
ORF (50 -AAAACTGGAACTCGGGAAAATCAGCC-30 ). A
linear XhoI/HindIII fragment from the pTRE-mEDA-A1 construct was then purified by Elutip-D column chromatography
(Schleicher and Schuell, Keene, NH, USA), resuspended in
pre-filtered 8 mM Tris–HCl, pH 7.5, 0.1 mM EDTA buffer,
and microinjected into pronuclei of one-cell inbred C57BL/6J
mouse embryos (the genetic background of Tabbyh6Ji).
Microinjected embryos were implanted into pseudo-pregnant
female mice and DNAs from tail samples of progeny were
genotyped for presence of the transgene by PCR. Potential
founders were mated to C57BL/6J mice to identify those
passing the transgene. The transgene-positive male mice (two
independent lines for each construct) were mated with
heterozygous Tabby females (C57BL/6J-Aw-j-Ta6j strain,
Jackson Laboratory, Bar Harbor, ME, USA). Tabby females
carrying the transgene (genotyped by PCR and SSCP) were
then mated with confirmed Tet-off male mice [C57BL/6J-TgN
(MMTVtTA), Jackson Laboratory] (39) to activate the mEDAA1/mEDA-A2 transgene. Phenotypes were recorded from the
final cross’s progeny, i.e. double transgenic offspring under
Tabby or wild-type background. Doxicycline (200 mg/kg in
Purina mouse chow, Bio-Serv, Frenchtowne, NJ, USA) was
administered to transgenic progeny or final breeders to turn off
transgene expression selectively.
Genotyping and genetic background detection
Genotyping was done by PCR with two transgene-specific
primer sets, both of which were used for progeny from
both mEDA-A1 and -A2 lines. Primer set 1: sense primer,
50 -GCTACCTAGAGTTGCGGTCCG-30 ; antisense primer,
Human Molecular Genetics, 2003, Vol. 12, No. 22
50 -CCCACCTGGCCCTCCTGGTCCTCA-30 (spanning intron
I of the mEDA gene and amplify 330 bp products from the
transgene, Fig. 2E, upper panel). Cycling conditions: denaturation at 95 C for 3 min, 35 cycles of 95 C for 30 s, 60 C
for 30 s, 72 C for 1 min. Primer set 2: sense primer,
50 -CTGGAACTCGGGAAAATCAG-30 (in exon 5 of mEDA);
antisense primer, 50 -TGGTGTGCTTGCTTGCTCATATTG-30
(in exon 8, amplifying a 411 bp product only from the
transgene). Cycling conditions: denaturation at 95 C for
3 min, followed by 35 cycles of 95 C for 30 s, 60 C for 30 s,
72 C for 2 min.
To detect the presence of the Tet-off transgene, which
encodes the tetracycline-controlled transactivator (tTA), we
used the following primer set: sense primer, 50 -CTGATCTGAGCTCTGAGTG-30 ; antisense primer, 50 -GCAAAAGTGAGTATGGTGCC-30 , which amplifies a 300 bp product from
the Tet-off transgene. Cycling conditions were denaturation at
94 C for 1.5 min followed by 35 cycles of 94 C for 30 s, 52 C
for 30 s, and 72 C for 30 s.
The genetic background of transgenic progeny was
determined through the use of the Incorporation PCR SSCP
(single-strand conformational polymorphism) method (15).
Briefly, a 387 bp product was amplified in the presence of
1 mCi of [a-32P]dCTP with the primer set (sense, 50 -GCTACCTAGAGTTGCGGTCCG-30 ; antisense, 50 -AACCTGACCTGGACAACCTCT-30 ), and the 32P-labeled product was
denatured and analyzed by electrophoresis. The mutant allele
is one nucleotide shorter than wild-type, producing an
amplification product that moves correspondingly faster in
electrophoresis.
Quantitative analysis of transgene expression in
transgenic skin
One-step quantitative real-time RT-PCR with Taqman probes
and primers (ABI Prism 7700 Sequence Detection System,
Applied Biosystems, Foster, CA, USA) was performed to
detect the transcription level of transgenes and control genes
(18). Total RNAs were isolated from back skin of adult
wild-type, transgenic, and Tabby male mice with Trizol
(Gibco, Grand Island, NY, USA), and were treated with a
‘DNA free kit’ (Ambion, Austin, TX, USA), eliminating
genomic DNA contamination. Total RNAs from 19.5 d.p.c.
wild-type C57BL mouse embryos were used to generate a
standard curve. PCR was carried out according to the
manufacturer’s protocols. Reactions were normalized to
GAPDH expression levels. Probe/primer set for A1 and A2
transgenes: probe, 50 -CAGGACCGGCACCAGATGGCC-30 ;
sense primer, 50 -TTGGAGCCGGGAGAAGATC-30 , antisense
primer, 50 -CAGAATATGCCTTTTCATCAGGAA-30 . Probes
and primers for EDA pathway genes were described in Cui
et al. (18).
Analysis of hair subtypes and sebaceous
gland function
To assess transgene effects on hair type, we cut hair from
equivalent areas of back skin of mice and enumerated the
easily-distinguishable hair types (16) under a stereomicroscope.
Guard (G; Tylotrich) hair is long, straight and dark; medium
2939
size hair (M; including awls and auchenes) is yellowish, thinner
and either straight or slightly bent; zigzag hair (Z) is yellowish
and sharply angled. Hairs were counted from males including
four wild-type, three Tabby, six A1 transgenic in Tabby
background, five A1 transgenic in wild-type background, and
three A2 transgenic in Tabby background. At least 350 hairs
were counted from each mouse.
For total hair number count, we used three skin samples
from each strain. Skin samples were taken with a 6 mm
Dermal biopsy punch (Miltex Instrument, Bethpage, NY,
USA), fixed in 4% paraformaldehyde, and embedded in
paraffin. Ten-micron sections were cut parallel to the skin
surface with a microtome and stained with hematoxylin
and eosin (Sigma). Hair numbers were count from three
random areas (2.4 mm2) of each section and scored at 20
magnification.
Oil-Red O staining was used to test for sebaceous gland
function. Frozen sections were stained with 0.5% Oil-Red O in
2-propanol (Sigma) at 37 C for 30 min, followed by washes
with 70% 2-propanol and H2O. Sections were counterstained
with hematoxylin (Sigma).
Histology, immunohistochemistry and
in situ hybridization
Samples freshly harvested from control and experimental
animals were fixed in 4% paraformaldehyde and 1% glutaraldehyde in phosphate-buffered saline overnight at 4 C. Fixed
samples were paraffin embedded, and sections were stained
with eosin/hematoxylin (Sigma) and observed in an Axiovert
200 microscope.
For immunohistochemistry (Fig. 6), sections were heated in
10 mM citrate buffer at 121 C for 5 min, then incubated with
monoclonal antibody against proliferating cell nuclear antigen
(PCNA) (NeoMarkers, Fremont, CA, USA) overnight at 4 C
and stained with a streptavidin–biotin peroxidase system
according to the manufacturer’s recommendation (DAKO
LSABþ kit, Carpinteria, CA). Sections were counterstained
with hematoxylin.
For in situ hybridization, a 411 bp PCR product from EDA
transcript (primer set 2, above) was cloned into the pCR4-topo
vector (Invitrogen, Carlsbad, CA, USA) and was labeled with
digoxygenin-UTP (Roche, Indianapolis, IN, USA). Labeled
probes (400 ng/ml in concentration) were hybridized with 4%
paraformaldehyde fixed frozen sections at 65 C overnight.
Following washing, alkaline phosphatase-conjugated antidigoxygenin antibody (Roche) was applied for overnight
incubation. Detection was performed with NBT/BCIP solution
(Roche) for 24 h.
ACKNOWLEDGEMENTS
The authors particularly thank Eric Douglass for animal
management, Ramaiah Nagaraja, Minoru Ko and Paul Waeltz
for critical reading of the manuscript, Carol Borgmeyer and
Dave Donovan for microinjection and technical assistance, Dan
Rowley, Shengyuan Luo and Mohamed Mughal for genotyping
and Yulan Piao for providing E19.5dpc whole embryo RNA
and technical assistance.
2940
Human Molecular Genetics, 2003, Vol. 12, No. 22
REFERENCES
1. Kere, J., Srivastava, A.K., Montonen, O., Zonana, J., Thomas, N.,
Ferguson, B., Munoz, F., Morgan, D., Clarke, A., Baybayan, P. et al. (1996)
X-linked anhidrotic (hypohidrotic) ectodermal dysplasia is caused by
mutation in a novel transmembrane protein. Nat. Genet., 13, 409–416.
2. Srivastava, A.K., Pispa, J., Hartung, A.J., Du, Y., Ezer, S., Jenks, T.,
Shimada, T., Pekkanen, M., Mikkola, M.L., Ko, M.S.H. et al. (1997) The
Tabby phenotype is caused by mutation in a mouse homologue of the EDA
gene that reveals novel mouse and human exons and encodes a protein
(ectodysplasin-A) with collagenous domains. Proc. Natl Acad. Sci. USA,
94, 13069–13074.
3. Mikkola, M.L., Pispa, J., Pekkanen, M., Paulin, L., Nieminen, P., Kere, J.
and Thesleff, I. (1999) Ectodysplasin, a protein required for epithelial
morphogenesis, is a novel TNF homologue and promotes cell-matrix
adhesion. Mech. Dev., 88, 133–146.
4. Schneider, P., Street, S.L., Gaide, O., Hertig, S., Tardivel, A., Tschopp, J.,
Runkel, L., Alevizopoulos, K., Ferguson, B.M. and Zonana, J. (2001)
Mutations leading to X-linked hypohidrotic ectodermal dysplasia affect
three major functional domains in the tumor necrosis factor family member
ectodysplasin-A. J. Biol. Chem., 276, 18819–18827.
5. Ezer, S., Bayes, M., Elomaa, O., Schlessinger, D. and Kere, J. (1999)
Ectodysplasin is a collagenous trimeric type II membrane protein with a
tumor necrosis factor-like domain and co-localizes with cytoskeletal
structures at lateral and apical surfaces of cells. Hum. Mol. Genet., 8,
2079–2086.
6. Headon, D.J. and Overbeek, P.A. (1999) Involvement of a novel TNF
receptor homologue in hair follicle induction. Nat. Genet., 22, 370–374.
7. Monreal, A.W., Ferguson, B.M., Headon, D.J., Street, S.L., Overbeek, P.A.
and Zonana, J. (1999) Mutations in the human homologue of mouse dl
cause autosomal recessive and dominant hypohidrotic ectodermal dysplasia. Nat. Genet., 22, 366–369.
8. Yan, M., Wang, L.C., Hymowitz, S.G., Schilbach, S., Lee, J., Goddard, A.,
de Vos, A.M., Gao, W.Q. and Dixit, V.M. (2000) Two-amino acid molecular
switch in an epithelial morphogen that regulates binding to two distinct
receptors. Science, 290, 523–527.
9. Headon, D.J., Emmal, S.A., Ferguson, B.M., Tucker, A.S., Justice, M.J.,
Sharpe, P.T., Zonana, J. and Overbeek, P.A. (2001) Gene defect in
ectodermal dysplasia implicates a death domain adapter in development.
Nature, 414, 913–916.
10. Yan, M., Zhang, Z., Brady, J.R., Schilbach, S., Fairbrother, W.J. and
Dixit, V.M. (2002) Identification of a novel death domain-containing
adaptor molecular for ectodysplasin-A receptor that is mutated in crinkled
mice. Curr. Biol., 12, 409–413.
11. Schmidt-Ullrich, R., Aebischer, T., Hulsken, J., Birchmeier, W., Klemm, U.
and Scheidereit, C. (2001) Requirement of NF-kappaB/Rel for the
development of hair follicles and other epidermal appendices.
Development, 128, 3843–3853.
12. Naito, A., Yoshida, H., Nishioka, E., Satoh, M., Azuma, S., Yamamoto, T.,
Nishikawa, S. and Inoue, J. (2002) TRAF6-deficient mice display
hypohidrotic ectodermal dysplasia. Proc. Natl Acad. Sci. USA, 99,
8766–8771.
13. Priolo, M., Silengo, M., Lerone, M. and Ravazzolo, R. (2000) Ectodermal
dysplasias: not only ‘skin’ deep. Clin. Genet., 58, 415–430.
14. Blecher, S.R., Kapalanga, J. and Lalonde, D. (1990) Induction of sweat
glands by epidermal growth factor in murine X-linked anhidrotic
ectodermal dysplasia. Nature, 345, 542–544.
15. Srivastava, A.K., Durmowicz, M., Hartung, A.J., Hudson, J., Ouzts, L.V.,
Donovan, D.M., Cui, C.Y. and Schlessinger, D. (2001) Ectodysplasin-A1 is
sufficient to rescue both hair growth and sweat glands in Tabby mice. Hum.
Mol. Genet., 10, 2973–2981.
16. Vielkind, U. and Hardy, M.H. (1996) Changing patterns of cell adhesion
molecules during mouse pelage hair follicle development. 2. Follicle
morphogenesis in the hair mutants, Tabby and downy. Acta. Anat. (Basel),
157, 183–194.
17. Barsh, G. (1999) Of ancient tales and hairless tails. Nat. Genet., 22, 315–316.
18. Cui, C.Y., Durmowicz, M., Tanaka, T.S., Hartung, A.J., Tezuka, T.,
Hashimoto, K., Ko, M.S., Srivastava, A.K. and Schlessinger, D. (2002)
EDA targets revealed by skin gene expression profiles of wild-type, Tabby
and Tabby EDA-A1 transgenic mice. Hum. Mol. Genet., 11, 1763–1773.
19. Gaide, O. and Schneider, P. (2003) Permanent correction of an
inherited ectodermal dysplasia with recombinant EDA. Nat. Med.,
9, 614–618.
20. Mustonen, T., Pispa, J., Mikkola, M.L., Pummila, M., Kangas, A.T.,
Pakkasjarvi, L., Jaatinen, R. and Thesleff, I. (2003) Stimulation of
ectodermal organ development by Ectodysplasin-A1. Dev. Biol.,
259, 123–136.
21. Botchkarev, V.A., Botchkareva, N.V., Sharov, A.A., Funa, K., Huber, O. and
Gilchrest, B.A. (2002) Modulation of BMP signaling by noggin is required
for induction of the secondary (nontylotrich) hair follicles. J. Invest.
Dermatol., 118, 3–10.
22. Gruneberg, H. (1971) The glandular aspects of the tabby syndrome in the
mouse. J. Embryol. Exp. Morph., 25, 1–19.
23. Gat, U., DasGupta, R., Degenstein, L. and Fuchs, E. (1998) De Novo hair
follicle morphogenesis and hair tumors in mice expressing a truncated
beta-catenin in skin. Cell, 95, 605–614.
24. Oro, A.E., Higgins, K.M., Hu, Z., Bonifas, J.M., Epstein, E.H. Jr. and
Scott, M.P. (1997) Basal cell carcinomas in mice overexpressing sonic
hedgehog. Science, 276, 817–821.
25. Fitzgerald, K., Harrington, A. and Leder, P. (2000) Ras pathway signals are
required for notch-mediated oncogenesis. Oncogene, 19, 4191–4198.
26. van Genderen, C., Okamura, R.M., Farinas, I., Quo, R.G., Parslow, T.G.,
Bruhn, L. and Grosschedl, R. (1994) Development of several organs
that require inductive epithelial-mesenchymal interactions is impaired in
LEF-1-deficient mice. Genes. Dev., 8, 2691–2703.
27. Jamora, C., DasGupta, R., Kocieniewski, P. and Fuchs, E. (2003) Links
between signal transduction, transcription and adhesion in epithelial bud
development. Nature, 422, 317–322.
28. Durmowicz, M.C., Cui, C.Y. and Schlessinger, D. (2002) The EDA gene is
a target of, but does not regulate Wnt signaling. Gene, 285, 203–211.
29. Huelsken, J., Vogel, R., Erdmann, B., Cotsarelis, G. and Birchmeier, W.
(2001) Beta-catenin controls hair follicle morphogenesis and stem cell
differentiation in the skin. Cell, 105, 533–545.
30. Laurikkala, J., Pispa, J., Jung, H.S., Nieminen, P., Mikkola, M., Wang, X.,
Saarialho-Kere, U., Galceran, J., Grosschedl, R. and Thesleff, I. (2002)
Regulation of hair follicle development by the TNF signal ectodysplasin
and its receptor EDAR. Development, 129, 2541–2553.
31. Botchkarev, V.A., Botchkareva, N.V., Roth, W., Nakamura, M., Chen, L.H.,
Herzog, W., Lindner, G., McMahon, J.A., Peters, C., Lauster, R.,
McMahon, A.P. et al. (1999) Noggin is a mesenchymally derived stimulator
of hair-follicle induction. Nat. Cell. Biol., 1, 158–164.
32. Chiang, C., Swan, R.Z., Grachtchouk, M., Bolinger, M., Litingtung, Y.,
Robertson, E.K., Cooper, M.K., Gaffield, W., Westphal, H., Beachy, P.A.
et al. (1999) Essential role for Sonic hedgehog during hair follicle
morphogenesis. Dev. Biol., 205, 1–9.
33. Mikkola, M.L. and Thesleff, I. (2003) Ectodysplasin signaling in
development. Cytokine Growth Factor Rev., 14, 211–224.
34. Thesleff, I. and Mikkola, M.L. (2002) Death receptor signaling giving life
to ectodermal organs. Sci. STKE, 131, PE22.
35. Bayes, M., Hartung, A.J., Ezer, S., Pispa, J., Thesleff, I., Srivastava, A.K.
and Kere, J. (1998) The anhidrotic ectodermal dysplasia gene (EDA)
undergoes alternative splicing and encodes ectodysplasin-A with
deletionmutations in collagenous repeats. Hum. Mol. Genet., 7, 1661–1669.
36. Niemann, C., Unden, A.B., Lyle, S., Zouboulis, C.C., Toftgard, R. and
Watt, F.M. (2003) Indian hedgehog and {beta}-catenin signaling: role in the
sebaceous lineage of normal and neoplastic mammalian epidermis. Proc.
Natl Acad. Sci. USA, 100 (Suppl. 1), 11873–11880.
37. Kozak, M. (1996) Interpreting cDNA sequences: some insights from
studies on translation. Mamm. Genome, 7, 563–574.
38. Elomaa, O., Pulkkinen, K., Hannelius, U., Mikkola, M., Saarialho-Kere, U.
and Kere, J. (2001) Ectodysplasin is released by proteolytic shedding and
binds to the EDAR protein. Hum. Mol. Genet., 10, 953–962.
39. Hennighausen, L., Wall, R.J., Tillmann, U., Li, M. and Furth, P.A. (1995)
Conditional gene expression in secretory tissues and skin of transgenic
mice using the MMTV-LTR and the tetracycline responsive system. J. Cell.
Biochem., 59, 463–472.
40. Mikkola, M.L., Pispa, J., Pekkanen, M., Paulin, L., Nieminen, P., Kere, J.
and Thesleff, I. (1999) Ectodysplasin, a protein required for epithelial
morphogenesis, is a novel TNF homologue and promotes cell-matrix
adhesion. Mech. Dev., 88, 133–146.