Is the Corneal Contour Influenced by a Tension in the Superficial Epithelial Cells? A New Hypothesis
Background: The corneal epithelium has a limited capacity to smooth stromal irregularities by compensatory variations in epithelial thickness.
Methods: Cuts were made in fixed and nonfixed human and rabbit corneal epithelium with an excimer laser, and undulations were produced in the anterior stromal lining of rabbit corneas by inducing hypotony of the eye balls.
Results: A retraction of the nonfixed epithelium at the edges of the cuts, and immediate small variations in epithelial thickness partly compensating for the induced irregularities in the stroma were observed.
Conclusions: A tension between the superficial epithelial cells could explain these phenomena. The balance between the pressure exerted by the superficial cell layer under tension and the epithelial growth pressure might be the factor determining the thickness of the epithelium an any point of the cornea in steady state conditions. A preliminary mathematical model and consequences for refractive surgery are presented. [Refract Corneal Surg 1992;8:54-59.)
The corneal epithelium is a layer of uniform thickness covering Bowman's layer. It is well known that this epithelium has a limited capacity to smooth stromal irregularities. The characteristic "sawtooth" appearance of the epithelium in ReisBückler's dystrophy is a manifestation of this feature.1 In small stromal defects, an epithelial plug fills the stromal gap and restores the original surface. In larger lamellar keratectomy wounds, a hyperplasia over the depression tends to restore the original contour.2·3 Eventually, the dimensions (depth, diameter, and contour) of the wound will determine to what extent the original contour of the cornea will be restored.
This leveling potential of epithelium is an important element in new refractive techniques as epikeratophakia or excimer photorefractive keratectomy.4,5 To be able to predict and eventually influence the reaction of the epithelium, it is necessary to understand the mechanism of compensatory hyperor hypoplasia of the epithelium. The hypothesis proposed in this article is that the superficial cells exert a traction upon each other, probably by means of their microfilaments and junctional complexes. This results in a tension in the superficial epithelial cell sheet. To verify this hypothesis, two experiments were performed.
MATERIALS AND METHODS
Two human eyes enucleated because of a malignant melanoma were used. The corneal epithelium was scraped off half the surface by means of a surgical knife while the other half was left intact. Trenches 0.2 mm wide and 0.1 mm deep were cut in the corneal surface extending in both the denuded and the epithelialized part of the cornea. These cuts were made with an excimer laser beam (Questek 2620, Billerica, Mass; beam homogenizer from Exitech, Oxford, UK) limited by a mask with a slit, 0.2 mm wide and 4 mm long. The mask was held a millimeter above the corneal surface. The cuts produced were observed by biomicroscopy (Pig 1) and pictures were taken. After this procedure, the eyes were immersed for 3 hours in B5-fixative (formaldehyde 4%, mercury chloride 0.2 M, buffered in sodium acetate 0.15 M). The corneas were excised and prepared for paraffin sections.
Figure 1: (A) An excimer laser cut In the living human epithelium presents a rounded border (big arrow). The edge where the knife had removed the epithelium Is also rounded (small arrow). (B) An excimer laser cut In the fixed human epithelium presents a nearly right angle (big arrow). The edge where the knife has removed the epithelium is also rectangular (small arrow.)
As a control, the same procedure with an inverted sequence was performed on two human donor eyes not suitable for corneal transplantation. They were first fixed in B5-solution for 3 hours. Then the epithelium was scraped off over half the surface, while the other half was left intact. Afterwards, they were subjected to the same excimer laser cut as described above and pictures were taken. The corneas were excised and further prepared for light microscopy.
The entire experiment was repeated on rabbit eyes. Two Dutch belted rabbits were killed by means of an intravenous air embolus after sedation with 2 mL of Hypnorm (fluanizone 1 mg/mL; fentanyl 0.2 mg/mL). The eyes were enucleated. On two eyes, the laser cut was made on the living epithelium; on the other two eyes, it was done on the fixed epithelium.
Eight eyes were obtained from four Dutch belted rabbits as described above. One eye of each rabbit was made hypotonic with a small sclerectomy at the optic nerve, causing a partial collapse of the globe. This resulted in a wrinkling of the cornea. Two minutes after this intervention, the eyes were immersed in B5-fixative for 3 hours.
The second eye of each rabbit served as a control. The globe was kept intact so that the original shape Of the cornea was maintained prior to and during fixation. After fixation, the corneas were excised and prepared for light microscopy with conventional paraffin sections.
Biomicroscopy shows that the edges of the troughs in Bowman's layer and in the anterior stroma formed a nearly right angle. In the epithelium, however, the edges of the cut formed a rounded border. This could clearly be seen immediately after ablation by observing the specular reflection of light on the corneal surface. Fixation of the cornea afterwards did not change this rounded appearance of the wound edges (Fig IA: big arrow). Note that the same rounding was seen at the edge of the cut produced by the knife (Fig LA: small arrow).
In the control eyes that had been subjected to fixation before laser treatment, the epithelial edges of the cuts appear square-edged (Fig IB: big arrow); the edge produced by the knife was rectangular, as well (Fig IB: small arrow).
Although the rabbit corneal epithelium is thinner, similar observations were made. The denuded anterior stromal surface was less smooth because of the absence of Bowman's layer.
Light microscopy of paraffin sections yield rounded edges, confirming the above biomicroscopic observations (Fig 2A). These rounded edges were in sharp contrast to the almost square edges in the epithelium fixed before cutting (Fig 2B). The rounding extended over about 50 µ.. The outer lining of the edge consisted of a single layer of collapsed cells. Until about the 10th row of the basal cells lining the cut, the cell bodies were inclined, leaning away from the cut. This was not found when the epithelium had been fixed prior to the laser cut. The collagen fibers of the superficial stroma were bending slightly upward at their cut end.
Figure 2: (A) The edge of the laser cut on the living epithelium is rounded. It shows collapsed cells and skewing of the basal cells away from the cut (original magnification, × 160). (B) The edge of the laser cut on the fixed epithelium is nearly rectangular. It shows no collapse of the cut cells and no skewing of the basal cells (original magnification, × 160).
The eyes that were made hypotonic prior to fixation showed undulations of the stromal collagen lamellae. The anterior stromal surface showed the same undulations. However, the outer surface of the epithelium was always smoother due to compensatory variations of the epithelial thickness (Fig 3). Where the epithelium was thicker, the number of wing cells increased and the basal cells appeared slightly more elongated.
The control eyes had a normal appearance with parallel collagen lamellae and an epithelium of uniform thickness throughout the section (not shown).
These experiments were performed to substantiate the hypothesis that the superficial epithelial cells exert a traction on each other. The observed rounding of the cut in the living epithelium was not an artifact of the ablation, as demonstrated on the epithelium fixed before the laser treatment. Nor was it an artifact of fixation because the phenomenon is biomicroscopically visible in the living epithelium.
Retraction of the epithelium and the rounded appearance of the epithelial cell borders have been described previously.6-9 Berns et al10 describe a "ridging" of the surface epithelium along the incision. Kuwabara et al6 explain that at the wound edge the cut epithelial cells lose their content. Consequently, the damaged cells collapse and the wound edge of the epithelium becomes rounded. However, (1) the rounding of the epithelium extended further than could be expected from the collapse of one row of cells lining the cut, and (2) this collapse would not explain the skewing of the basal cells. Therefore, it would appear that the hypothetical retraction of the epithelial cells is a contributing factor in explaining the observed phenomena. Marshall et al8 also suggest that the rounded margins of the troughs may have arisen through the release of elastic tension within the epithelium.
Figure 3: The anterior stromal lining of this rabbit cornea shows undulations due to hypotony of the globe prior to fixation. The epithelial contour invariably has a smoother appearance due to compensatory changes in epithelial thickness (original magnification, × 40).
Equally interesting is the observation of compensatory changes in epithelial thickness occurring almost instantaneously after inducing irregularities of the anterior stroma. This phenomenon has most likely been observed during many cornea research experiments, but no attempts seem to have been made toward explaining it. These compensations must have been formed in the short interval between the manipulation and the fixation. This is compatible with the concept of an epithelium build-up of flexible cells that have some degree of freedom. A superficial cell layer seems to be tended over the wing and basal cells so that the elastic recoil of these cells limits the exposed surface. This causes a redistribution of the wing cells, compensating for the small irregularities in the anterior stromal lining. This elastic recoil of the superficial epithelial cells can also be assumed in the case where the epithelium is isolated according to a technique described by Gipson and Grill.11 The isolated epithelium spontaneously curls and forms a roll so that the outside of this roll is formed by the basal cells. The superficial epithelial cells would appear to have a greater capacity of elastic recoil than the basal cells, causing curling of the epithelial sheet as described above.
Figure 4: The superficial cell layer can be considered the thin wall of a sphere. The tension T generated In this wall exerts a slight pressure PT on the underlying cells. This is balanced by the growth pressure Px of the basal cells.
It seems that the anatomical substrate for a tension in the superficial cell layer is present.
In normal epithelium, a web of actin filaments is seen in the region beneath the microplieae of the superficial cells.12 This web of microfilaments seems to be related to the smoothness of the apical cell wall,13 comparable to the basal cortical microfilamentous mat.14 We know that the actin filaments in a cell invariably organize parallel to the lines of tension.15 Myosin and actin have been shown to be essential elements in corneal epithelial cells.16 They constitute the cell muscles.17 The tension may be transmitted from one cell to another through their junctional complexes. A comprehensive study of the ultrastructural basis of this hypothesis is currently underway.
Preliminary Mathematical Model
The superficial intercellular tension may have an influence on the epithelial thickness in small keratectomy wounds, and also in large area corneal reprpfiling. For photorefractive keratectomy in particular it is important to be able to predict the degree of compensatory hypo- or hyperplasia of the epithelium. The following preliminary mathematical model is based on the supposition that this intercellular tension really exists.
Because this superficial cell layer is tended over a spherical surface, it will exert a slight pressure on the underlying epithelial layers, the basal and wing cells. For this pressure, the symbol Pt will be used (Fig 4).
In steady state conditions, this pressure must be equal and opposite to the pressure from the underlying and upward moving cells. This pressure is called the "growth pressure." This term has been used previously18 to indicate the Y-component of the epithelial proliferation.19 In this article, it is used to indicate the X-component, with the symbol Px. In a normal cornea, Pt is equal to PX, and the epithelial thickness does not change.
Tension is related to the radius of curvature and described in the formula for the tension in thinwalled spheres:20
T = Px R/2e (eq 1), wherein:
T = tension in the wall, P = the pressure in the sphere or the pressure from wall on its content, R = the radius of curvature of the sphere, and e = the thickness of the wall.
The tension-generating layer in the corneal epithelium is analogous to the wall of a sphere. This tension-generating layer might be the entire superficial cell layer, or only the actin web in the superficial cells. The content of the sphere consists of a viscous mass of basal and wing cells. The stroma is to be considered as a very large and rigid core.
It is not to be expected that this intercellular tension plays a role with regard to the intraocular pressure. Neither should this "epithelial surface tension* be confused with the real surface tension of the tearfilm. The forces between the superficial cells are probably from the same order of magnitude as the forces that make the cell migrate.
For the corneal epithelium, the equation for the tension in thin-walled spheres becomes:
T = PT × R/2e (eq 2) or
PT = T × 2e/R (eq 3) wherein:
T = tension in the superficial layer, PT = the pressure from this superficial cell layer on the deeper layers of the epithelium, R = the radius of curvature of the cornea, and e = the thickness of the tension generating layer.
Only R is easily measured. Values for PT, Px, T, and e are actually unknown.
What happens if the radius of curvature is changed? Looking back to equation 3, we see that P is inversely proportionate to R. In other words: the larger the radius of curvature, the smaller the pressure from the superficial on the basal cells (Fig 5). Suppose a myopic excimer photorefractive keratectomy of 10.00 D is performed on a cornea. In the optical zone, the cornea will be flattened: an initial radius of curvature of 7.8 mm will increase to 10.1 mm, or 1.3 × the original radius.
Assuming (1) that the tension between healthy superficial epithelial cells is a constant feature, and not a variable one as it is in muscle cells, and (2) that the thickness of the tension generating layer e is also constant, the product PxR will be constant. In the example, the pressure on the underlying cells PT would decrease with a factor 1.3. If at the same time, the "growth pressure" Px remains the same, the epithelium will tend to increase in thickness. A new steady state will be reached.
Figure 5: In an area with a larger radius of curvature (R1 > R0), and assuming a constant tension T, the Inward pressure of the superficial epithelium PT is smaller. The epithelium will tend to get thicker.
Figure 6: Over a deep lamellar keratectomy the epithelium forms a concave surface (R0 > 0 > R1). The tension in the superficial cell layer will cause an outward pressure component. This will help the epithelium to level the defect.
Hence, this model predicts that the epithelium will tend to thicken in areas with a flatter curvature until the growth pressure is balanced by the pressure resulting from the tension in the superficial cell layer. The balance between these two opposed pressures determines the thickness of the epithelium at any point of the cornea.
This mechanism would explain the clinical finding of regression of the refractive effect in corrections of myopia with photorefractive keratectomy,21,22 the epithelial component of the so-called myopic shift. The epithelial hyperplasia found in the center of a myopia epikeratophakia23-24 or on keratomileusis lenticules25,26 could be caused in the same way, as well as the epithelial hyperplasia after myopic excimer photorefractive keratectomy in rabbit,27 monkey,28-31 and human corneas.32
A second situation where the model can be applied is a lamellar keratectomy with abrupt edges (Fig 6). After initial covering of the defect, the epithelial surface will be concave in the wound area, at least for a brief time period. The tension will have a pressure component outward, instead of inward. This will help to level the defect. If the defect is small, the original contour will be restored entirely. If the diameter of the defect is larger, the hyperplasia at the edge will tend to get thinner toward the center. This differential thickness of epithelium will make the cornea flatter. This is an alternative explanation for the hyperopic shift observed in therapeutic excimer laser superficial keratectomies (ELSK) with a single axial zone of ablation.33,34
Figure 7: An intrastromal ring causes a bulging of the surface with a small radius of curvature (R1 < R0). The tension causes a large inward pressure component PT. The covering epithelium will tend to get thinner.
In many other situations, this model provides an explanation for the behavior of the epithelium. A third and last example is the one of a bulging caused by an intrastromal ring (Fig 7). Suppose the local radius of curvature on top of the bulging is 0.8 mm, or one tenth of the radius of curvature of the cornea. If the product PxR would be constant, the PT will increase tenfold, the epithelium will tend to get thinner, and one or two basal cell layers will remain. This is compatible with the "epithelial atrophy" observed over the shoulder of intracorneal rings.35
1. Perry HD, Pine BS, Caldwell D. Reis-Bücklers dystrophy: a study of eight cases. Arch Ophthalmol. 1979;97:664-670.
2. Tuft S, Marshall J, Rothery S. Stromal remodeling following photorefractive keratectomy. Lasers in Ophthalmology. 1987;1:177-183.
3. Hanna KD, Pouliquen Y, Waring GO, et al. Corneal stromal wound healing in rabbits after 193 nm excimer laser surface ablation. Arch Ophthalmol. 1987;107:895-901.
4. Gipson I. Corneal epithelial and stromal reactions to excimer laser photorefractive keratectomy: concerns regarding the response of the corneal epithelium to excimer laser ablation. Arch Ophthalmol. 1990;108:1539-1540.
5. Tuft SJ, Zabel RW, Marshall J. Corneal repair following keratectomy. Invest Ophthalmol Vis Sci. 1989;30:1769-1777.
6. Kuwabara T, Perkins DG, Cogan DG. Sliding of the epithelium in experimental corneal wounds. Invest Ophthalmol Vis Sci. 1976;15:4-14.
7. Brewitt H. Sliding of the epithelium in experimental corneal wounds: a scanning electron microscopic study. Acta Ophtalmologica. 1979;57:945-958.
8. Marshall J, Trokel S, Rothery S, Schubert H. An ultrastructural study of corneal incisions induced by an excimer laser at 193 nm. Ophthalmology. 1985;92:749-758.
9. Haik BG, Zimny ML. Scanning electron microscopy of corneal wound healing in rabbit. Invest Ophthalmol Vis Sci. 1977;16:787-796.
10. Berns MW, Liaw LH, Olivia A, Andrews JJ, Rasmussen RE, Kimel S. An acute light and electron microscopic study of ultraviolet 193 nm excimer laser corneal incisions. Ophthalmology. 1988;95:1422-1433.
11. Gipson JK, Grill SM. A technique for obtaining sheets of intact rabbit corneal epithelium. Invest Ophthalmol Vis Sci. 1982;23:269.
12. Gipson IK, Anderson RA. Actin filaments in normal and migrating epithelial cells. Invest Ophthalmol Vis Sci. 1977;16:161-166.
13. Pfister RR, Burstein NL. The normal and abnormal human corneal epithelial surface: a scanning electron microscope study. Invest Ophthalmol Vis Sci. 1977;16:614-622.
14. Sugrue SD, Hay ED. Response of basal epithelial cell surface and cytoskeleton to solubilized extracellular matrix molecules. J Cell Biol. 1981;91:45-54.
15. Kolega J. Effects of mechanical tension on protrusive activity and microfilament and intermediate filament organization in an epidermal epithelium moving in culture. J Cell Biol. 1986;102:1400-1411.
16. Drenckhahn D, Gröschel-Stewart U. Localization of myosin and actin in ocular nonmuscle cells. Cell Tissue Res. 1977;181:493-503.
17. Alberts B. The cytoskeleton. In: Alberts B. Molecular Biology of the Cell. 2nd ed. New York, NY: Garland Publishing, Ine; 1989:671.
18. Frieden wald JS. Growth pressure and metaplasia of conjunctival and corneal epithelium. Doc Ophthalmol. 1951;5/6:184192.
19. Thoft RA, Friend J. The X, Y1Z hypothesis of corneal epithelial maintenance. Invest Ophthalmol Vis Sci. 1983;24:1442-1443.
20. Vaughan D, Asbury T. General Ophthalmology, 10th ed. Los Altos, Calif: Lange Medical Publications; 1983:160.
21. Seiler T, Kahle G, Kriegerowski M. Excimer laser (193 nm) myopic keratomileusis in sighted and blind human eyes. Refract Corneal Surg. 1990;6:165-173.
22. Seiler T, Jean B, Pham T, Derse M, Bende T. Statistical analysis of myopic regression after excimer laser PRK. Invest Ophthalmol Vis Sci. 1991;32(suppl):721.
23. Cusumano A, Busin M, Spitznas M, Koch F. Epikeratophakia for the correction of myopia: lenticule design and related histopathological findings. Refract Corneal Surg. 199Q;6:120124.
24. Saragoussi JJ, Rayana NB, Savoldelli M, D'Hermies F, Pouliquen Y Etude de deux lenticules d'épikératoplastie myopique. JFr Ophtalmol. 1989;12:465-468.
25. Baumgarter SD, Binder PS. Refractive keratoplasty: histopathology of clinical specimens. Ophthalmology. 1985;92:16061615.
26. Jester JV, Rodrigues MM, Villaseñor RA, Schanzlin DJ. Keratophakia and keratomileusis: histopathologic, ultrastructural, and experimental studies. Ophthalmology. 1984;91:793-805.
27. Gaster RN, Binder PS, Coalwell K, Berns M, McCord MC, Burstein NL. Corneal surface ablation by 193 nm excimer laser and wound healing in rabbits. Invest Ophthalmol Vis Sci. 1989;30:90-97.
28. McDonald MB, Frantz JM, Klyce SD, Salmerón B, Beuerman RW, Munnerlyn CR. One-year refractive results of central photorefractive keratectomy for myopia in the nonhuman primate cornea. Arch Ophthalmol. 1990;108:40-47.
29. Del Pero RA1 Gigstad JE, Roberts AD, et al. A refractive and histopathologic study of excimer laser keratectomy in primates. Am J Ophthalmol. 1990;109:419-429.
30. Fantes FE, Hanna KD, Waring GO, Pouliquen Y, Thompson KP, Savoldelli M. Wound healing after excimer laser keratomileusis (photorefractive keratectomy) in monkeys. Arch Ophthalmol. 1990;108:665-675.
31. Malley DS, Steinert RF, Puliafito CA, Dobi ET. Immunofluorescence study of corneal wound healing after excimer laser anterior keratectomy in the monkey eye. Arch Ophthalmol. 1990;108:1316-1322.
32. Taylor DM, L'Espérance FA, Del Pero RA, Roberts AD, Gigstad JE, KHntworth G. Human excimer lamellar keratectomy. Ophthalmology. 1989;96:654-664.
33. Sher NA, Bowers RA, Zabel RW, et al. Clinical use of the 193 nm excimer laser in the treatment of corneal scars. Arch Ophthalmol. 1991;109:491-498.
34. Gartry D, Kerr Muir M, Marshall J. Excimer laser treatment of corneal surface pathology: a laboratory and clinical study. Br J Ophthalmol. 1991;75:258-269.
35. Hartmann C, von Ey F, D'Hermies F, Holzkämper C, Renard G. Histobiocompatibility of a refractive intracorneal PMMAring. Invest Ophthalmol Vis Sci. 1991;32(suppl):998.
Hans G. Dierick, MD, Luc Missotten, MD, PhD
From the Excimer Study Group, Ophthalmology Clinic, University Hospital St Rafael, Katholieke Universiteit Leuven, Belgium.
Supported in part by Buchmann Optical Industries, Kapellen, Belgium, and by the Belgian Foundation for Scientific Research NFWO, Brussels, Belgium.
The authors acknowledge Dr A. Schaelenbourg, W. Van Haesendonck, Dr C. Van Mellaert, and Dr J. van den Oord for their support.
The authors have no proprietary interest in any research or materials presented within this article.
Reprint requests should be addressed to Hans G. Dierick MD, Dienst Oogheelkunde, Universitair Ziekenhuis, St Rafael, Kapucijnenvoer 33, B-3000 Leuven, Belgium.
Received: February 8, 1991
Accepted: June 16, 1991
Comment: Theo Seiler, MD, PhD, Berlin, Germany
Ronchi stated in 1926 that the outer surface of the eye must be smooth within 0.3 µp? to account for the optical resolution of the eye.1 This required smoothness was believed to be obtained by a combination of passive surface tension of the tear film (to level out small irregularities) and an active and delayed leveling potential of the epithelium (to level out gross irregularities), active and delayed because of mitotic activity. The article by Dierick and Missotten2 adds a third factor into this system: a passive surface-parallel tension located inside the superficial epithelium cells. I was surprised when I read the article for the first time. It is hard to believe that leveling of stroma irregularities occurs by an almost instantaneously hyperplasia and hypertrophy of the epithelium above the valleys. Do the cells slide into the valleys from the neighboring elevations? If this is true, what about the self-cohesion of the epithelium which allows it to be separated from Bowman's as an unbroken sheet in some cases? Aren't there interdigitations and desmosomes in between the cells?
In spite of this contradiction, the experiments and the histology of Dierick and Missotten are convincing, although there is still a possibility that the inclination of epithelial cells next to the edges of keratectomies is due to a fixation artifact (for example, a shrinkage factor different in superficial and basal epithelium cells). The authors state that the round border of a cut formed by epithelial cells could clearly be seen by observing the specular reflection of light on the corneal surface. However, when I tried to reduplicate this by performing cuts into epithelium by means of a diamond knife it was not that obvious for me. Freeze sections could help to overcome this missing link.
Any new approach in science creates some resistance which has to be understood as part of a constructive discussion. In this sense, the article of Dierick and Missotten may serve as a landmark and initiation to think about the plasticity of corneal epithelium in a new direction.
1. Ronchi V. Sulla funzione ottica del liquido lacrimale. Accad Lincei. 1926;4:476-478.
2. Dierick H, Missotten L. Is the corneal contour influenced by a tension in the superficial epithelial cells? A new hypothesis. Refract Corneal Surg. 1992;8:54-59.
Response: Dierick and Missotten
Our article does not really describe a third factor. It tries to give an explanation for the second factor: the leveling potential of the epithelium. The question is: how does the epithelium "feel" the irregularities in the stromal surface, and what is the stimulus for it to become thicker on one place and thinner on another?
On histologic sections, the epithelium has the appearance of a static sheet. Yet, we know that it is continuously regenerating. It is dynamic equiHbrium. Constantly, cells are dividing in the basal layer, moving upward, becoming flatter and desquamating. Linked by numerous desmosomes, they pull and push on each other all through their life. Our point is that these intercellular forces are not the same on every level of the epithelial sheet. We presume that in the basal layer, the intercellular forces parallel to the surface are virtually nonexistent while in the superficial cell layer they are predominant. The retraction of the epithelium that we observe at the edge of a cut and the organization of the filaments in the superficial cells support this hypothesis.
We agree with Ronchi that the leveling effect of the epithelium is active and delayed. Figures 6 and 7 in our article show the state obtained by active proliferation of the epithelium several days after the intervention. It can take weeks or even months after a refractive procedure before the epithelium reaches its final steady state. However, there is also an immediate, albeit limited, compensatory action as demonstrated in our second experiment (Fig 3). We propose that this small instantaneous compensatory action is the result of the tension in the superficial epithelial cells, and that this tension is also the stimulus for the delayed active leveling potential of the epithelium in the way as explained in our preliminary mathematical model.