/ . Embryol. exp. Morph. Vol. 58, pp. 131-142, 1980Printed in Great Britain © Company of Biologists Limited 1980

Cell behaviour and cleft palate in themutant mouse, amputated

By O. P. FLINT1

From the Department of Zoology, University of Glasgow

SUMMARYCleft palate with a genetic origin normally arises because of a failure of the palatal shelves

to elevate or fuse. Until now attention in studies of palatal development has been focusedon two critical phases, those of elevation and fusion. In the mutant mouse, amputated,however, cleft palate arises because of a failure of the palatal shelves to make any significantoutgrowth between the 12th day after conception when the palatal shelves are first observedand the 14th day when elevation and fusion normally occur. When cell proliferation (mitoticindex) was measured in the palatal shelves on days 12, 13 and 14 no difference was foundbetween mutant and normal. The failure of the mutant palate to grow cannot be accountedfor on grounds of reduced cell proliferation. For this reason the palatal mesenchyme in12-5-day and 14-5-day normal and amputated mice has been studied with the scanningelectron microscope. This work shows that the mesenchymal cells in the mutant palate areclumped together and have much greater areas of cell contact than are observed in the normalpalate. The abnormal cell behaviour described in mutant palatal mesenchyme is typical ofamputated embryonic mesenchyme in general, and in other cases has been shown to causeabnormal morphogenesis. We propose that aberrant cell behaviour causing an aggregationthrough increased cell adhesion inhibits palatal outgrowth in the mutant, and for this reasonthe palatal shelves subsequently fail to elevate and fuse.

INTRODUCTION

Abnormal cell behaviour causing the cells to clump together and inhibiting cellmovement affects the development of the mouse mutant, amputated, at allstages involving morphogenetic cell movements. The length of the embryo isreduced because node and streak regression is retarded (Flint, Ede, Wilby &Proctor, 1978). Clumping of the somite sclerotome cells may partly contributeto fusions and distortions later found among the vertebral cartilages (Flint,\911a). Snout outgrowth is strongly reduced because increased cell adhesionand cell clumping retards a critical early stage of facial development when thefacial mesenchyme grows by expansion of the extracellular volume, whichnormally forces the cells to move apart (Flint, 1977b; Flint & Ede, 1978a). Inin vitro culture, amputated cells clump together and cell movement can beobserved in time-lapse films to have been retarded (Flint, 1977a and Flint, inpreparation).

1 Author's address; ICI Pharmaceuticals Division Mereside Alderley Park MacclesfieldCheshire SK10 4TG, U.K.

132 O.P.FLINT

Another developmental system involving cell movement is the palate. Cleftpalate is found in all mouse embryos homozygous for the single recessive geneamputated. Its development is described as part of a programme of researchinto the cellular basis of abnormalities in mutant development. We find thatamputated palates fail to rotate and therefore to fuse because palatal shelfoutgrowth is retarded through anomalous cell behaviour similar to that alreadydescribed as affecting development in other parts of the embryo. Palate develop-ment has been reviewed by Greene & Pratt (1976) who describe two phases ofpalate development: (1) elevation (including rotation) of the palatal shelves and(2) palatal fusion. A failure of elevation or fusion can cause cleft palate. In themouse mutant shorthead (Fitch, 1961) elevation of the palatal shelves is arrested,by a tongue which is too large for the abnormally small buccal cavity. In thechick, on the other hand, rotation occurs but there is no fusion when the palatalshelves meet, resulting in the type of cleft palate which is normal in avianembryos (Greene & Pratt, 1976). Cleft palate arising out of the earliest stage ofpalatal shelf outgrowth, as in amputated, has not so far been described, and thisearliest stage of palatal morphogenesis has been neglected.

METHODS

The mice are a CBA/101 hybrid intercross carrying the single recessive geneamputated. Conception was taken to coincide with the midpoint of the darkperiod just prior to noticing a vaginal plug (Snell, Fekete, Hummel & Law,1940). Matings were designed to produce litters with homozygous amputatedembryos. The techniques of fixation and critical-point drying of mouse embryosfor the scanning electron microscope have been described by Flint & Ede(19786). After fixation and while in 70 % ethanol the heads were dissected underthe dissecting microscope with micro-surgery scalpels (Moria-France) toremove the mandibles and tongue, revealing the roof of the buccal cavity. Insome cases the heads were then cut transversely to produce sections of thepalatal shelves. It was found that the effect of cutting like this was not to crushthe tissue but to fracture it in the plane of cutting. The stages of palatal develop-ment we describe are not so highly resolved as those described by Walker &Fraser (1956) or Walker & Crain (1960) since we were not specifically interestedin palatal fusion and also since the earliest stage we describe (12-5 days) is twodays younger than the embryos described by these authors.

Cell proliferation and cell density were measured in normal and amputatedpalatal shelves and in the mesenchyme to one side of the base of the palatal shelf(control area) at 12-5, 13-5 and 14-5 days. Cell proliferation was recorded asmitotic index; the number of mitotic figures as a percentage of total cell number.Statistical analysis of the figures was carried out by the standard analysis ofvariance procedure.

Cell behaviour and palatal morphogenesis 133

Palatal development

Gross development changes:Normal. The earliest stage at which the palatal shelves can be seen is at

12-5 days, when they appear as a pair of parallel ridges growing down fromeither side of the roof of the buccal cavity (Fig. 1 a). That these are ridges canbe seen at their posterior end where the knife has cut a glancing transversesection through the tissue. By the 14th day of development (14-5 days) the ridgeshave grown well out and away from the palatal roof (Figs. lc,3a, c) and rotationcan be seen to have begun in the anterior half (Fig. 3 a) of at least one shelf. Thepalatal rugae (Figs. \c, 3a, c) begin to make their appearance in an antero-posterior sequence at this stage before the palatal shelves meet and fuse. Eachpair of rugae appears at the same level of the palate so that after palatal fusion(Fig. 3 b) they will form almost continuous ridges across the roof of the mouth.At the end of the 14th day of development (approximately 15-0 days) fusion hasbegun along the meeting central portions of the two palatal shelves (Figs. 2 a, 3 b).

Amputated. As in the normal embryo, mutant palatal shelves begin theiroutgrowth during the 12th day of development (Fig. Ib). There are two ana-tomical features which differ from the normal embryo. First, the width of thehead is greater and the snout shorter (Flint & Ede, 1978 a), so that the palatalshelves form at a greater distance apart. Secondly, the opening to Rathke'spouch persists to this late stage in amputated, but not in the normal embryo(Fig. 1 a, b). By 14-5 days, though, the overall shape of the head in amputatedremains blunter anteriorly and broader than normal, the distance between thepalatal shelves being the same in normal and mutant (Fig. 1 c, d). There is littleor no downgrowth of the palatal shelves away from the roof of the mouth(Fig. 1 d, 3d), when compared to the normal (Figs. 1 c, 3a, c) at the same stage,or the amputated embryo at 12-5 days (Fig. \b). As in the normal embryo, thefirst palatal rugae make their appearance at 14-5 days (Fig. 1 d) and by the endof the 14th day of development several more have appeared (Fig. 2b), but thereis no further outgrowth of the mutant palatal shelves. In the normal embryogrowth of the mandibles and the cranial base enlarges the buccal cavity so thatthe tongue drops out of the way of the elevating palatal shelves. In amputated,snout outgrowth is retarded (Flint, 19776 and Flint & Ede, 1978a). But growthof the tongue is not so strongly inhibited and its impression can be seen in thepalatal shelves in Fig. 2b.

The epithelial surface of the palate

Normal. The aboral epithelial surface of the palate shows some interestingregional variation at 15-0 days associated with the appearance of the rugae andthe point of fusion. Between the rugae numerous microvilli are seen on thesurfaces of the epithelial cells (Fig. 4 c). These are especially dense at the borders

134 O. P. FLINT

FIGURE 1

Scanning electron micrographs of the heads of normal (o, c) and amputated (b, d)mice at 12-5 {a, b) and 14-5 (c, d) days of development. The mandibles and tonguehave been dissected away so that a view of the roof of the buccal cavity is obtainedin each case, in, internal nares; pp, primary palate including median process;pr, palatal ruga; ps, palatal shelf; rp, opening to Rathke's pouch.

Cell behaviour and palatal morphogenesis 135

FIGURE 2

Scanning electron micrographs of normal (a) and amputated (b) mouse heads at150 days. The preparation is the same as that described in Fig. l.fp, point of fusionbetween the palatal shelves; pp, primary palate and median process; /?r, palatal ruga;ps, palatal shelf; tg, tooth germ (upper incisor) exposed in amputated because thebroadening and foreshortening of the cranium draws the lips apart.

between cells. This is typical of an epithelial surface in contact with a fluidmedium, in this case the amniotic fluid, and has been observed on the chicklimb epidermis (Ede, Bellairs & Bancroft, 1974) and on the naso-frontal regionof the hamster face (Waterman & Meller, 1973), as well as on the mouse(Waterman, Ross & Meller, 1973) and human (Waterman & Meller, 1974)palate. There is a transition between inter-rugal epithelium and epitheliumassociated with the rugae. Rugal epithelium is comparatively bare of microvilliand the cell surfaces are much rougher (Fig. 4a). At the point of palatal fusionall boundary distinction between cells is lost and numerous elongate cell pro-jections are produced. A large quantity of debris also accumulates on the surface.

Amputated. Late in the 14th day of development the inter-rugal epitheliumis very similar in mutant and normal palates (cf. Fig. 4c, d). But the rugae,far from lacking microvilli, are even more profusely covered, with many moremicrovilli on the surface as well as at the junctions between cells (Fig. 4 b).

136 O. P. FLINT

FIGURE 3

Low-power scanning electron micrographs of normal embryos at 145 (a) and 150 (6)days to show the palatal shelves elevating (a) and fusing (b). Higher powers of sections,at 14-5 days, of the right palatal shelf in a normal embryo (c) and the left palatal shelfin an amputated embryo (d) are also shown to compare the outgrowth of the palate inmutant and normal, in, internal nares; pp, primary palate and median process; pr, palatalruga; ps, palatal shelf.

Sectioned palatal shelves

Normal. The section through the raised palatal shelf (Fig. 3 c) is reminiscentof a section through a developing limb bud, but there is no apical ectodermalridge specialization of the distal ectoderm. At 12-5 days cells are spaced wellapart and are connected by numerous fine and very fine filopodia (Fig. 5 a).By 14-5 days the cells are still well spaced but, in addition to the filopodia, theyare coated with a dense fibrous and globular matrix of intercellular materials(Fig. 5 c, e). On the basis of the work of Hassell & Orkin (1976), who studied theproduction of extracellular materials in the palate, we would identify the long

Cell behaviour and palatal morphogenesis 137

FIGURE 4

The epithelial surface of the palatal shelves at 150 days in normal {a, c, e) andamputated {b, d) mice. The normal series shows a, the surface of a ruga and thetransition from rugal to inter-rugal surface (at the bottom of the figure); c, theinter-rugal surface and e, the point of fusion between two palatal shelves. Themutant series shows b, the rugal surface and d, the inter-rugal surface. A sectionthrough the fusing palatal shelves of a normal mouse at 150 days is shown in/.

138 O. P. FLINT

FIGURE 5

The mesenchyme of the palatal shelves in normal (a, c, e) and amputated (b, dj)mice at 12-5 days (a, b) and 14-5 days (c, d, e,f) just prior to elevation.

nbryo

NamNam

12-5 days

l-94±013210±0141-44 ±0122041015

13-5 days

l-81±0143-21 ±0-85l-38±012l-47±013

14-5 days

O-98±O13l-86±O130-21 ±0080-72 ±003

Cellscounted

98364029

100615947

Cell behaviour and palatal morphogenesis 139

Table 1. Mitotic index {mitoses per 100 cells) in the palatal shelves and adjacentcontrol areas of normal and amputated mouse embryosValues are given as mean plus or minus the standard deviation.

Tissue

Palate

Control

fibrous intercellular material as collagen and the globular material as a productof the basement membrane which can extend far into the palate matrix. At thepoint where the two palatal shelves meet a zone of continuity is formed, wherethe breakdown of the epithelial interface occurs and the two palatal mesen-chymes become one (Fig. 4/). The scanning electron microscope evidence isconsistent with transmission electron microscope studies on the palatal mesen-chyme (Babiarz, Allenspach & Zimmerman, 1975 and Innes, 1978).

Amputated. Whereas normal cells at 12-5 days give a well-spaced structure tothe palatal tissue, amputated cells appear to have aggregated together (Fig. 5 b).The filopodia adhere in a tangled web over all the cell surfaces. This is exactlysimilar to preparations of the somite sclerotome in the mutant embryo at9*5 days (Flint & Ede, 1978&). By 14-5 days there has been no overall change tothis aggregated or clumped appearance (Fig. 5d). But the intercellular spaces,as in the normal palatal tissue, have become filled with globular and filamentousextracellular material (Fig. 5d,f).

Cell proliferation in the palatal shelves

Values of mitotic index measured from sections of palatal shelves on 12-5,13-5 and 14-5 days after conception are given in Table 1.

There is no significant overall difference between normal and amputated(P > 0-10). But there is a significant overall difference in mitotic index measuredin the palatal shelves and in the adjacent control areas (0-05 > P > 0-025).This difference largely rests on a drop in mitotic index in the control area between13-5 and 14-5 days. The only other significant difference occurs at 14-5 dayswhen in palate and adjacent control areas mitotic indices are higher in ampu-tated than in normal (P < 0-05). No explanation for this difference emergesfrom the current work.

Cell density

Values of cell density measured at the same time as mitotic index are givenin Table 2. There is no significant overall difference between normal andamputated (P > 0-10). But there is a significant overall difference in the cell

IO EMB 58

NamNam

1408 ±1-5215-44±l-56l5-24±3-1215-24 ±1-72

16-08 ±1-4415-28 ±1-6015-40 ±1-9615-68 ±1-22

ll-92±l-441116±l-8O21-76±l-5418-68 ±0-98

140 O. P. FLINT

Table 2. Cell density (cells per 1OZ ju,m2) in the palatal shelves and adjacent controlareas of normal and amputated mouse embryos (same number of cells counted asin Table 1)

Values are given as mean plus or minus the standard deviation.

Tissue Embryo 12-5 days 13-5 days 14-5 days

Palate

Control

density measured in the palatal shelves and adjacent control areas. This dif-ference rests largely on a drop in cell density in the palatal shelves between13-5 and 14-5 days.

DISCUSSION

Wherever genetically-caused cleft palate has been described in the mouse thecritical phase at which palate development has been inhibited has been that ofelevation and rotation (i.e. during the 14th day of development).

But there is no significant change in the outgrowth of mutant palatal shelvesfrom the time when they are first observed at 12-5 days up to early in the 14thday of development (Figs. 1 b, 3 d). Normal specimens show considerable down-growth of the palatal shelves on either side of the tongue and away from theroof of the buccal cavity during this time (Figs. \a, c, 3 a, c).

In spite of abnormal morphogenesis the differentiation of palatal shelf cellsproceeds normally, so that not only is the extracellular matrix material secretedat the appropriate time accompanied by a parallel reduction in mutant andnormal cell density (see Table 2) but also the palatal rugae make their appearanceon cue (Fig. 1 d, 2b). Between the rugae the normal palatal epithelium is coveredwith microvilli (Fig. 4 c), but the rugae have lost their coating of microvilli(Fig. 4a). Microvilli are most probably a mechanism of conserving cell mem-brane (O'Neill & Follett, 1970). The development of the elevated ridge-likerugae in the normal embryo on the expanding surface of the palatal shelvesinvolves a stretching of the epithelial surfaces and consequently a loss of themicrovilli. In the mutant there is no expansion of the palatal shelves. Theformation of the rugae involves a concentration of the available epithelium,and consequently an increase in the number of microvilli (Fig. 4 b), rather thana decrease.

Outgrowth of the naso-frontal region, like that of the palatal shelves, isinhibited at the earliest stage in the amputated mouse (Flint, 19776; Flint &Ede, 1978 a). This inhibition cannot be accounted for on the grounds that cellproliferation is reduced. Similarly mitotic index is the same in mutant and

Cell behaviour and palatal morphogenesis 141

normal palatal shelves throughout the earliest stages from the 12th to the 14thday of development. In the case of the naso-frontal region we have found thatincreased cell adhesion between the cells in the mutant causes clumping similarto that found elsewhere in the embryo and this retards the tissue expansionwhich occurs in the normal embryo by secretion of extracellular material, andtherefore retards morphogenesis from the 10th day of development. Thisclumping of cells is typical of the mutant mesenchyme both in vivo (Flint,1977 a, b; Flint & Ede, 1978/7, b) and in vitro (Flint, 1977 a; Flint, in preparation).Increased cell contacts and areas of cell-cell adhesion causing an aggregationof the palatal mesenchyme exactly similar to that described in other parts of theembryo persist from the 12th (Fig. 5 b) to the 14th (Fig. 5d) day of develop-ment. Since it is during this very early stage of palatal morphogenesis prior toelevation that palatal outgrowth is inhibited, and since there is such a highcorrelation in other parts of the embryo between abnormal cell behaviour andabnormal morphogenesis (see also Introduction), we ascribe the cleft palate inamputated to an anomaly of cell behaviour causing increased cell adhesionwhich inhibits downward growth of the palatal shelves. The increased celladhesion does not increase cell density in the mutant mesenchyme. There arethe same number of cells per unit volume as in normal palatal shelves, butbecause of increased cell adhesion these cells tend to clump together; they arenot so well dispersed as in the normal palate. Our measurements of the somitesclerotome in the mutant produced the same result (Flint & Ede, 1978 b).Matrix secretion causing a drop in cell density in the palatal shelves of bothmutant and normal embryos between the 13th and 14th day of developmentcomes too late to help the mutant palatal shelf catch up with the growth alreadymade by normal shelves. Comparing Figs 1 and 2, however, it can be seen thatsome expansion of mutant palatal shelves occurs late in the 14th day ofdevelopment, probably as a result of the secretion of matrix material. Thedevelopment of the palatal shelves between the 12th and the 14th day, whenelevation occurs, has hitherto received little attention, but it appears that this isa critical phase in palatal development and that disturbance of normal growthduring this stage may produce abnormal cleft palate at a later one.

A comparison of the effect on cell behaviour of amputated with work onother mutant genes can be found in Flint & Ede (1978 a, b).

REFERENCES

BABIARZ, B. S., ALLENSPACH, A. L., ZIMMERMAN, E. F. & (1975). Ultrastructural evidence ofcontractile systems in mouse palates prior to rotation. Devi Bio I. 47, 32-44.

EDE, D. A., BELLAIRS, R. & BANCROFT, M. (1974). A scanning electron microscope study ofthe early limb-bud in normal and talpid3 mutant chick embryos. / . Embryol. exp. Morph. 31,761-785.

FITCH, N. (1961). Development of cleft palate in mice homozygous for the shortheadmutation. / . Morph. 109, 151-157.

142 O. P. FLINTFLINT, O. P. (1977a). Cell interactions in the developing axial skeleton in normal and

mutant mouse embryos. In: Vertebrate Limb and Somite Morphogenesis (ed. D. A. Ede,J. R. Hinchliffe & M. Balls), pp. 464-484. Brit. Soc. Devi. Biol. Symp. 3, CambridgeUniversity Press.

FLINT O. P. (19776). Cell interactions in facial development in the mouse. / . Anat. 124225-226.

FLINT, O. P. & EDE, D. A. (1978a). Facial development in the mouse: a comparison betweennormal and mutant (amputated) mouse embryos. J. Embryol. exp. Morph. 48, 249-267.

FLINT, O. P. & EDE, D. A. (19786). Cell interactions in the developing somite: in vivo com-parisons between amputated (am/am) and normal mouse embryos. / . Cell Sci. 31, 275-292.

FLINT, O. P., EDE, D. A., WILBY, O. K. & PROCTOR, J. (1978). Control of somite number innormal and amputated mutant mouse embryos: an experimental and theoretical analysis./ . Embryol. exp. Morph. 45, 189-202.

GREENE, R. M. & PRATT, R. M. (1976). Developmental aspects of secondary palate formation./ . Embryol. exp. Morph. 36, 225-245.

HASSELL, J. R. & ORKIN, R. W. (1976). Synthesis and distribution of collagen in the rat palateduring shelf elevation. Devi Biol. 49, 80-88.

INNES, P. B. (1978). The ultrastructure of the mesenchymal element of the palatal shelves ofthe fetal mouse. / . Embryol. exp. Morph. 43, 185-194.

O'NEILL, C. H. & FOLLETT, A. C. (1970). An inverse relation between cell density and thenumber of microvilli in cultures of BHK21 hamster fibroblasts. J. Cell Sci. 7, 695-709.

SNELL, C. D., FEKETE, E., HUMMEL, K. P. & LAW, L. W. (1940). The relation of mating,ovulation, and the estrus smear in the house mouse to the time of day. Anat. Rec. 76,39-54.

WALKER, B. E. & FRASER, F. C. (1956). Closure of the secondary palate in three strains ofmice. J. Embryol. exp. Morph. 4, 176-189.

WALKER, B. E. & CRAIN, B. (1960). Effects of hypervitaminosis A on palate development intwo strains of mice. Am. J. Anat. 107, 49-58.

WATERMAN, R. E. & MELLER, S. M. (1973). Nasal pit formation in the hamster. A trans-mission and scanning electron microscope study. Devi Biol. 34, 255-266.

WATERMAN, R. E. & MELLER, S. M. (1974). Alterations in the epithelial surface of humanpalatal shelves prior to and during fusion: a scanning electron microscope study. AnatRec. 180, 111-136.

WATERMAN, R. E., ROSS, L. M. & MELLER, S. M. (1973). Alterations in the epithelial surfaceof A/Jax mouse palatal shelves prior to and during fusion: a scanning electron microscopicstudy. Anat. Rec. 176, 361-376.

(Received 11 October 1979, revised 30 January 1980)