Effect of the X-linked gene Tabby (Ta) on eyelid opening and incisor eruption in neonatal mice is opposite to that of epidermal growth factor

Epidermal growth factor (EGF), a polypeptide of low relative molecular mass (6054), was first isolated from mouse submandibular salivary gland (Cohen, 1962). EGF has been found in virtually all murine and human body fluids (Bynny et al. 1974; Ladda et al. 1979; Shinoda et al. 1988). Among the first effects noted for EGF was its ability to induce precocious eyelid opening and incisor eruption in wildtype newborn mice (Cohen, 1962).

Tabby (Ta), an X-linked gene in the mouse, was first identified by Falconer (1952, 1953). In hemizygous and homozygous states, Tabby produces a characteristic syndrome, the features of which include narrowed palpebral fissures, reduced number of facial vibrissae (Falconer, 1953), reduction in size and number of teeth (Gruneberg, 1966; Heller and Blecher, 1982) and abnormal texture of the coat (Kindred, 1967). Furthermore, the development of certain exocrine glands including the Meibomian, lacrimal, and Harderian glands is impaired (Gruneberg, 1971).

Blecher et al. (1982, 1983) pointed out that many of the traits of Ta correspond to those of human X-linked hypohidrotic ectodermal dysplasia (HED), that in both cases many of these traits occur in target organs of EGF, and that Ta markedly reduces the number of EGF-producing cells of the mouse submandibular gland. They proposed the possibility that Ta and HED may be homologous, and that the effects of these genes could be explained by EGF deficiency at molecular target sites. The present study was designed to determine the effect of Ta on postnatal timing of eyelid opening and incisor eruption and to study the effects of EGF on these integumental processes in mice affected by the Ta gene. We report here that in the hemizygous and homozygous state Ta delays the timing of eyelid opening and incisor eruption. The delays are reversed by exogenous EGF. These observations support the suggestion that the Ta gene exerts its effect by influencing EGF production or metabolism.

Mice used in this study were from a partially inbred Ta-bearing congenie line of the strain DCH (Blecher and Kirkeby, 1979; Kirkeby and Blecher, 1981). Animals were housed in our breeding colony at the University of Guelph and maintained under standard conditions with a 14 h photoperiod, and with free access to food (Purina Lab Chow) and water. Adult (2–5 month old) Ta/Y or +/Y males were mated by monogamous pairing with their Ta/+ or +/+ littermate sisters. These mating schemes produce Ta/Y, +/Y, Ta/Ta, Ta/+ and +/+ offspring. Ta reduces viability, and Ta/Ta offspring survive less often than others.

The first phase of this study involved examination of 216 neonatal mice comprising 62 wildtype males (+/Y), 32 wildtype females (+/+), 35 hemizygous Tabby males (Ta/Y), 30 homozygote Tabby females (Ta/Ta) and 57 heterozygote Tabby females (Ta/+). This phase was primarily designed to determine postnatal timing of eyelid opening and incisor eruption in the various genotypes of the DCH-Ta strain. In the second phase, 44 EGF-treated mutant and wildtype mice and 42 saline-treated controls were studied to examine for interactions between EGF and Ta in respect of the traits studied.

The same examination procedures were followed in the normal timing studies and the EGF-injection experiments. In all cases daily pup weighing and routine examinations commenced on the day of birth and continued until puberty. In postnatal studies the day of birth was taken as day zero.

Cage rounds were conducted daily between 09:00h and 12:00h to examine for pregnancy and births. Adult female mice were examined daily for the presence of vaginal plugs, for determination of the date of conception (day zero of gestation) and hence prediction of the day of birth.

Pups in each litter were identified by clipping a toe at birth (Dubin, 1968; Kumar, 1979). Genotypes were determined from genotypes of the mating pair and sex and vibrissa scoring of the pups.

The pups were weighed on a Mettler AE 160 compact electronic microbalance (Mettler Instruments Corporation, Highstown, NJ, USA) to an accuracy of 0.01g. It was then examined under good lighting, with the dorsum up, from the head region to the trunk and tail, to determine vibrissa number, eyelid opening, degree of coat hair eruption and kinking of the tail. To assess vibrissa number the head of the animal was viewed at 10 × magnification through a CRC magnifying lens (Roxter Corp., Cleveland, Ohio, USA). In wildtype mice this number is constantly two supraorbitals and one postorbital on each side (Dun, 1959), and the total complement emerges by postnatal day 3. In Ta /Y and Ta/Ta mutants, only the supraorbital vibrissae emerge at all, usually by day 3 but often on day 4–5.

Eyelid opening was deemed to be present when breakdown of the line of fusion was sufficient to render the underlying cornea visible. Generally, eyelid opening proceeds from the centre of the line of fusion to the canthi.

The pup was then turned ventral side up and assessed for sex (Rugh, 1968) and incisor eruption. Incisor eruption was deemed to have occurred when the tip of the incisor just appeared above the gingival epithelium as a whitish dot.

Members of a given genotype in individual litters were distributed into experimental (EGF) and control (saline) groups. Total natural litter size varied from 6 to 12 pups. However, litter size was adjusted to 6–8 pups per litter in the injection experiments by killing excess pups. EGF is known to be present in milk (Popliker et al. 1987); reduced litter size minimizes variations in the amount of milk EGF available to pups in different litters and also decrease rates of cannibalism and neonatal death.

For a given genotype in a litter, experimental and control animals were of approximately the same weight. Treatment and control groups of Ta/Y, Ta/Ta, +/Y, +/+ and Ta/+ pups in a litter were injected subcutaneously between the shoulders with 40μg^-1 body weight per day of aseptically prepared EGF solution and equivalent volumes of physiological saline, respectively, on days 0, 1, 2, 3 and 4. (EGF was lyophilized mouse-derived CR-Epidermal growth factor from Collaborative Research Inc., Bedford, MA, USA 100μg/ 0.25 ml, cat. no. 40010, lot no. 840191, 99% pure; single band at relative molecular mass 6100 on SDS-PAGE; R. Gilbert, Collaborative Research Inc., personal communication).

Data collected by examining 216 newborn mice (62 +/Y, 35 Ta/Y, 32 +/+, 57 Ta/+ and 30 Ta/Ta) are shown in Table 1. There were significant differences in the timing of eyelid opening and incisor eruption between the hemizygote and homozygote Tabby mutants and their sex-matched wildtype counterparts (P<0.05) Generally, incisor eruption precedes eyelid opening by 3–5 days in both mutants and wildtypes.

In addition to the delay in eyelid opening of mice carrying Ta, it was observed that once open, the eyes in mutants do not remain open for long, but soon close for several days before reopening (Fig. 1). A sequential eyes open-eyes closed process follows, going through about five cycles and continuing beyond puberty.

This phenomenon was explored further by plotting on a time scale the days on which the eyes were open or closed, to determine whether the sequential process occurred at fixed postnatal time points (Fig. 2). However, as shown in this figure, there was no consistent pattern of the process even among members of the same genotype. The process was observed in all hemizygote and homozygote mutants and in over one-third of the heterozygotes.

In most Ta/Y and Ta/Ta animals both eyes were affected; however, there were unilateral instances. In almost all heterozygotes only one eye was involved, and the sequential process was not observed beyond the fourth postnatal week. In the Ta/Y and Ta/Ta animals, the process continued beyond 5 weeks.

The effect of EGF on the timing of integumental events EGF-treated pups showed accelerated eyelid opening and incisor eruption when compared with saline-treated controls (Figs 3 and 4, and Table 2). These differences were significant at the 1% level. As a general trend, eyelid opening preceded incisor eruption by one or two days (Table 2). However, in two litters incisor eruption preceded eyelid opening, and in two other litters both events occurred on the same day.

After EGF injection, the timing of opening of both eyelids occurred almost uniformly on the same day in all genotypes (Fig. 5). In saline-treated controls, however, it was not uncommon for an individual’s two eyelids to open on different days. This was particularly the case for most Ta/Y and Ta/Ta controls and some Ta/+ animals, but it was rare in the wildtype mice. There was a tendency for the homozygote and hemizygote mutants to be more sensitive to the dosage of EGF used. The average postnatal timing of eyelid opening in mutants was 5.7 days; in the wildtype animals it was 6.25 days. Table 3 shows EGF acceleration for eyelid opening and incisor eruption. Differences between Ta/Y and +/Y and between Ta/Ta and +/+ were tested by the method of designed orthogonal contrasts (Gill, 1978), and were highly significant (P<0.01).

The effect of EGF on the acceleration of integumental events in normal mice has been well documented by several previous investigators (Cohen, 1962; Blose and Fenton, 1974; Steidler and Reade, 1980; Hoath et al. 1983; Hoath, 1986; Rhodes et al. 1987; Topham et al. 1987). However, no previous studies have examined the effects of EGF in Tabby mice.

The sequential eyes open-eyes closed process was noted to affect all genotypes after EGF injection. However, fewer cycles were observed, all prior to the fourth postnatal week in the wildtype animals (Fig. 3). There is no mention of this process in wildtype mice treated with EGF in previous studies.

In the extensive previous literature on the Tabby syndrome, there has been no previous association of Tabby with postnatal delay of eyelid opening and incisor eruption, although other eye and incisor features of the Tabby phenotype have been described (Falconer, 1952, 1953). In this study, timing of eyelid opening and incisor eruption in wildtype controls and heterozygotes was within the normal range previously reported in the mouse, i.e. postnatal days 12-14 for eyelid opening (Gruneberg, 1943; Theiler, 1972; Pei and Rhodin, 1970; Harris and McLeod, 1982) and 8-11 for incisor eruption (Ness, 1967; Rugh, 1968; Moxham and Berkovitz, 1982). The eyes open-eyes closed process is also hitherto undescribed in the literature. These new Ta-associated findings appear to be constant features of the Tabby phenotype and may reasonably be considered as part of the syndrome.

Ta-induced delay in eyelid opening is part of a more general effect of Ta on retardation of mesenchymal-epithelial development. Studies on the coat of Tabby mutants have demonstrated suppression of formation of new hair follicles between 12.5 and 17 days of gestation and again from birth onwards (Dun, 1959). Other studies have demonstrated slow growth rate of incisor tooth germs (Sofaer, 1969) and reduced growth and differentiation of glands around the eye (Gruneberg, 1971).

Our finding that exogenous EGF accelerates eyelid opening and incisor eruption in wildtype pups is in agreement with previous studies on the effects of EGF on normal mice (Cohen, 1962; Rhodes et al. 1987). The response of the mutant mice to EGF with respect to the timing of eyelid opening and incisor eruption was similar to the control, but the hemizygote and homozygote Tabby mutants were actually more sensitive than controls (Table 3).

Although there are no doubt other possibilities we propose that the latter observation, and the intriguing findings of the eyes open-eyes closed cycle in Tabby animals and EGF injected wildtype mice, may be explained on the basis of the known concentrationdependent down-regulation of the EGF receptor by its ligand (Adamson and Warsaw, 1982; Adamson and Meek, 1984; DePalo and Das, 1988). At low concentrations EGF is known to stimulate synthesis of its own receptors (Clark et al. 1985; Kudlow et al. 1986; Earp et al. 1986; DePalo and Das, 1988), whereas higher concentrations promote down regulation, possibly resulting in reduced response of target cells to EGF (Adamson and Warshaw, 1982). Presumably, physiological levels of EGF contribute in this way to maintenance of normal levels of receptors. EGF secretion may itself be regulated by feedback of EGF on the EGF-producing cells. An autocrine role for EGF has been suggested for cancer cells (Sporn and Roberts, 1985). Feedback regulation of EGF may involve binding of EGF to receptors on the EGF-secreting cells. Ta evidently leads to deficient EGF production; if feedback regulation is as suggested above, we could not exclude the possibility that Ta acts through a primary gene effect on receptor production or function. It is conceivable that due to deficiency of EGF in Tabby mutant mice, down-regulation of EGF receptors is impaired. Therefore in the balanced state, the population of available EGF receptors may be large, so that when exogenous EGF is given it is rapidly bound to the larger number of receptors that is available in mutants than in wildtype animals. This could account for the accelerated response of Tabby animals to exogenous EGF.

The same postulates could account for the initial delay in eyelid opening and subsequent eyes open-eyes closed cycle in the mutant. In Tabby animals, because of a constant state of EGF deficiency, a longer period of time would be required than in normal mice for accumulation of sufficient EGF to trigger the integumentary event. Because of generalised redundancy of receptor, EGF binding might outpace EGF synthesis; once bound to its receptor, EGF is known to accelerate receptor-ligand internalization and intracellular degradation (Hanover et al. 1984; Dunn and Hubbard, 1984). Maintenance of open eyelids may require threshold EGF levels at a critical period, and we hypothesize that such cyclic depletion of EGF might be the cause of the eyelid closing cycle. In both mutant and normal mice, injection of exogenous EGF would not only lead to eyelid opening but also to excessive inhibition of EGF production with consequent failure of the mechanism that maintains the eyes open. This might establish a cycle of production and feedback overshoot that takes several cycles to stabilise.

It is possible that if this explanation is correct, corresponding cyclical surges in incisor eruption might also be expected. We did not assess the rate of eruption and therefore do not know whether this occurred in our experiments.

That Ta may indirectly or directly affect EGF-receptor concentrations is of interest in view of the similarity between the EGF-receptor gene and the cellular oncogene c erb-B. Further, the observation that EGF reverses Ta traits may be of therapeutic relevance with respect to human hypohidrotic ectodermal dysplasia. The status of EGF receptors in Tabby mutants needs more study before these inferences can be further developed.

We are grateful to Dr Margaret Fallding for many stimulating discussions. These studies were supported in part by a grant from the Medical Research Council of Canada to S.R.B.