The odd thing about skin-barrier repair is that the cells capable of organising it are located beneath the structure that has failed. The stratum corneum has no blood supply, and its corneocytes cannot phone downstairs with a damage report. Yet the living epidermis responds. Stored material is released, lipid production increases and, later, the supply of future keratinocytes expands.
If you’ve ever stood in front of the mirror wondering whether post-cleanse tightness counts as barrier damage, you’ll recognise the problem from the other side. You have sensation and a reflection to go on. The keratinocytes organising repair have a different source of information: the environment the surface maintains around them.
An intact barrier restricts outward water movement and helps preserve the conditions beneath it. Disruption changes those conditions. But how do we establish that this change drives repair? Removing lipids with acetone or stripping away corneocytes exposes the tissue to chemical treatment or physical trauma. So any response afterwards could have more than one cause.
Borrow the function, reduce the response
A window doesn’t need to send a memo to your radiator when it opens. The room cools, and the heating responds to the temperature change. The proposed barrier feedback works the same way. Failure at the surface changes the environment of the living cells below.
To test that explanation, researchers needed to keep the damage while restoring something the damaged surface had stopped doing.
Controlled covering experiment in hairless mice. Every group had the same acetone damage. The mice were then left uncovered, covered with an impermeable membrane or covered with a vapour-permeable one. The researchers measured lipid synthesis, lipid return to the stratum corneum and barrier recovery. Only the impermeable covering blocked the response, which pointed to blocked water movement rather than covering itself.[1]
In 1989, Grubauer, Peter Elias and Kenneth Feingold disrupted the permeability barrier of hairless mice with acetone, then covered some sites with a water-vapour-impermeable membrane. The covering suppressed the usual rise in epidermal lipid synthesis and stopped lipids returning to the stratum corneum, and barrier function did not recover normally. But a membrane that let water vapour through did not have that effect.[1]
The disruptive acetone exposure had already happened. The covering hadn’t replaced the extracted lipids. It had borrowed one part of the barrier’s function, controlling outward water movement while the native structure remained damaged. Under that covering, the epidermis reduced its repair response.
Barrier function also regulates cell production. When Ehrhardt Proksch, Kenneth Feingold, Mao-Qiang Man and Peter Elias disrupted the barrier, acetone treatment increased epidermal DNA synthesis by 102%, and tape stripping by 127%. The peaks occurred at about 20 and 18 hours respectively, and the extra DNA synthesis after acetone treatment was limited to the basal layer. An impermeable covering largely suppressed the increase. The less permeable the covering, the stronger the suppression.[2]
The same linkage appeared in two chronic models, essential fatty acid deficiency and topical lovastatin, which damage the barrier by different routes. Together, the comparisons weakened the explanation that the tissue was responding only to the original treatment. Its response also depended on the permeability conditions afterwards. The covering experiments had exposed feedback. But they hadn’t identified what the cells sensed.
Water loss was the first suspect
The 1989 paper called transepidermal water loss (TEWL) the signal for recovery. TEWL measures water crossing the surface, but the epidermis has no access to the number on the researchers’ meter. Something accompanying the altered flow had to change cellular behaviour.
The water-flux explanation was put to a direct test in 1992. After extracting lipids from hairless-mouse skin, Seung Hun Lee and colleagues immersed the treated sites in solutions that differed in tonicity and ion content. Changing the solutions from hypotonic to hypertonic showed that water transit alone was not the decisive trigger. The ions called for closer attention.[3]
Calcium, potassium and phosphate each inhibited recovery modestly on their own. Calcium and potassium together had a synergistic effect: controls recovered about 50% of their barrier function over two and a half hours, while recovery in the combined-ion conditions fell to between 0% and 11%. The ions also prevented the usual rise in activity of HMG-CoA reductase, the rate-limiting enzyme in cholesterol synthesis, and blocked the return of lipids.[3]
Water movement remained part of the explanation, but the experiment showed it wasn’t enough on its own. Lee’s group proposed that increased water flux carries these ions out of the upper epidermis, so that a fall in their concentrations signals the need for repair. They didn’t demonstrate that route. What they did show was that the ionic environment affected both the metabolic response and the recovery that followed.
Change the calcium, leave the barrier intact
In undisturbed epidermis, calcium is relatively low in the basal layer and rises towards the stratum granulosum. There, the last living keratinocytes prepare to become corneocytes and deliver the lipid material that will surround them. This calcium gradient helps regulate those processes.
After barrier disruption, its distribution changes. That observation alone leaves a question: does the change help trigger secretion, or does it accompany something else that does?
Gopinathan Menon and colleagues separated those possibilities in 1994. They used high-frequency sonophoresis, an ultrasound-assisted method, to alter calcium conditions across hairless-mouse epidermis without altering barrier function. With an ion-free solution, extracellular calcium in the upper epidermis shifted downwards and lamellar-body secretion accelerated. With a calcium-containing solution, secretion stayed at its basal rate.[4]
No prior breach was required. A change in the calcium gradient was enough to trigger release of the stored barrier material.
The older explanation described barrier failure as a loss of the calcium gradient. In 2011, Martin Behne and colleagues used fluorescence-lifetime imaging on unfixed hairless-mouse skin biopsies and found a more complicated redistribution. Calcium moved between regions and compartments, with a large amount mobilised from intracellular stores shortly after the insult. The pattern didn’t behave like one extracellular reservoir draining away.[5]
That raises a different question. Which calcium pool is changing?
The endoplasmic reticulum stores calcium inside cells. Anna Celli and colleagues depleted this store using thapsigargin, which inhibits the pump that normally returns calcium to it. In mouse experiments, the treatment induced lamellar-body secretion and aspects of terminal differentiation. Depleting an intracellular store reproduced parts of the repair response while the barrier remained intact.[6]
Their later work brought the response into human tissue. In 2016, Celli, Debra Crumrine, Jason Meyer and Theodora Mauro observed an early rise in free calcium after disrupting human skin explants, followed by later depletion. In separate experiments using reconstructed human epidermis, an indicator targeted to the endoplasmic reticulum recorded calcium leaving that store within minutes of tape stripping, then being actively taken back up.[7]
Those results explain why “calcium falls” is incomplete. When a store releases calcium into another compartment, a fall in one place comes with a rise in another. The evidence points towards the changing distribution, rather than any single average calcium value, as the information the cells read.
Repair arrives in overlapping waves
The first response uses what is already packed. Lamellar bodies in granular keratinocytes carry lipid material and processing enzymes. By releasing these contents where the granular layer meets the stratum corneum, the epidermis begins supplying the lipid matrix between corneocytes before it has replenished its stores.
Menon’s group observed extensive secretion within 15 to 30 minutes of disruption in mice. While a vapour-permeable wrap allowed secretion to proceed, an impermeable membrane suppressed it. The released material still needed to be processed and organised: new lamellar bilayer units appeared in the lower stratum corneum between one and three hours. Yet fresh lamellar bodies had begun reappearing in the cells by 30 minutes. Delivery, assembly and replenishment were already overlapping.[8]
But a stored supply is finite. Sustained reconstruction requires synthesis.
Switching an enzyme on. Many enzymes, HMG-CoA reductase among them, switch between active and inactive forms when phosphate groups are added or removed. That change takes minutes and needs no new enzyme. Making more enzyme takes longer, which is why activity rose in two steps.[9]
Proksch, Elias and Feingold found that the activation state of HMG-CoA reductase changed within 15 minutes of acute disruption. Regulating the enzyme already present gave the epidermis a rapid metabolic response. Total enzyme activity began increasing after about an hour and a half. The second wave had started while the first was still under way.[9]
An increase in synthesis fits the repair explanation. Does recovery actually depend on it?
Blocking the enzyme answered it. After disrupting the mouse barrier, Kenneth Feingold and colleagues inhibited cholesterol synthesis with topical lovastatin, and recovery was impaired. Supplying mevalonate, the immediate product of the inhibited enzyme, restored recovery. Supplying cholesterol restored it too.[10]
The second rescue sharpened the first. Mevalonate feeds several biochemical routes, so its rescue alone would leave other downstream products in contention. Cholesterol is the pathway’s end product, so its rescue tied the delay more closely to the missing cholesterol.
Walter Holleran’s group tested whether recovery needed sphingolipid synthesis, using an inhibitor of serine palmitoyltransferase, an enzyme at the start of that pathway. The inhibitor suppressed synthesis promptly, but recovery slowed only after six hours. Coapplying ceramides overcame the delay.[11]
The skin could begin recovering while this pathway was inhibited, but it couldn’t keep up the normal pace. The different timing exposed a changing dependence on lipid supply as reconstruction continued.
The third response expands cell supply. The later rise in DNA synthesis in the basal layer prepares more cells for the tissue above. Those cells still have to differentiate and move towards the surface before becoming mature corneocytes.[2] Early improvements in permeability therefore do not require the epidermis to complete an entire renewal cycle. This is why barrier repair does not share one 28-day clock.
Stored material starts the reconstruction. Increased synthesis sustains it, while additional cells replenish the tissue over a longer period. The clocks overlap: the first wave holds the breach, the second rebuilds it, and the third stands guard behind it.
the first wave holds the breach, the second rebuilds it, and the third stands guard behind it.
What the borrowed barrier does not tell us
The covering experiments make one everyday connection tempting. If an artificial barrier reduces the repair response, would an occlusive moisturiser do the same?
An impermeable experimental wrap and petrolatum expose skin to different conditions. Ruby Ghadially, Lars Halkier-Sørensen and Peter Elias tested petrolatum after acetone disruption in human volunteers in 1992. Petrolatum accelerated recovery rather than delaying it. The authors found it at all levels of the stratum corneum, where it replaced the intercellular bilayers. It did not act like an impermeable membrane placed over the surface.[12]
The proliferation result also needs its species attached. In 1995, Peter van de Kerkhof and colleagues applied hydrocolloid dressings after tape stripping in human volunteers. The dressings didn’t change the measured proliferative response.[13] The earlier mouse work the authors cite had used Duoderm too, so the dressing alone doesn’t explain the difference. Species and set-up probably do. The authors concluded that a direct effect of artificial barrier restoration on human epidermal regeneration remains unsubstantiated.
Taken together, the experiments explain how repair responds to altered permeability conditions. But they don’t diagnose a tight feeling after cleansing or assign a universal timetable to a human routine. What they reveal is a living epidermis whose surroundings change when the surface fails, and whose responses range from stored secretion to future cell supply. The skin knows because the conditions an intact barrier would usually maintain have changed.
Corneocyte. A flattened, dead cell of the stratum corneum, the outermost layer of the skin.
Keratinocyte. The main living cell of the epidermis, which eventually becomes a corneocyte at the surface.
Tonicity. A hypotonic solution holds fewer dissolved particles than the tissue, so water tends to move in. A hypertonic solution holds more, so water tends to move out. Moving between them alters the osmotic conditions for water movement.
Stratum granulosum. The granular layer: the last layer of living keratinocytes beneath the stratum corneum, where lamellar bodies are made and released.
Lamellar body. A small packet inside granular-layer keratinocytes that stores lipids and processing enzymes, then releases them into the spaces between cells.
Ceramide. A sphingolipid that makes up a large share of the lipid matrix between corneocytes.
Occlusive. Describes a covering or ingredient that slows water evaporating from the skin surface.
lamellar-body secretion. A small packet inside granular-layer keratinocytes that stores lipids and processing enzymes, then releases them into the spaces between cells.
References
1. Grubauer, G., Elias, P.M. and Feingold, K.R. (1989). Transepidermal water loss: the signal for recovery of barrier structure and function. Journal of Lipid Research, 30(3), pp. 323–333. https://pubmed.ncbi.nlm.nih.gov/2723540/
2. Proksch, E., Feingold, K.R., Man, M.Q. and Elias, P.M. (1991). Barrier function regulates epidermal DNA synthesis. Journal of Clinical Investigation, 87(5), pp. 1668–1673. https://doi.org/10.1172/JCI115183
3. Lee, S.H., Elias, P.M., Proksch, E., et al. (1992). Calcium and potassium are important regulators of barrier homeostasis in murine epidermis. Journal of Clinical Investigation, 89(2), pp. 530–538. https://doi.org/10.1172/JCI115617
4. Menon, G.K., Price, L.F., Bommannan, B., Elias, P.M. and Feingold, K.R. (1994). Selective obliteration of the epidermal calcium gradient leads to enhanced lamellar body secretion. Journal of Investigative Dermatology, 102(5), pp. 789–795. https://doi.org/10.1111/1523-1747.ep12377921
5. Behne, M.J., Sanchez, S., Barry, N.P., et al. (2011). Major translocation of calcium upon epidermal barrier insult: imaging and quantification via FLIM/Fourier vector analysis. Archives of Dermatological Research, 303(2), pp. 103–115. https://pmc.ncbi.nlm.nih.gov/articles/PMC4548958/
6. Celli, A., Mackenzie, D.S., Crumrine, D.S., et al. (2011). ER Ca2+ depletion activates XBP1 and controls terminal differentiation in keratinocytes and epidermis. British Journal of Dermatology, 164(1), pp. 16–25. https://doi.org/10.1111/j.1365-2133.2010.10046.x
7. Celli, A., Crumrine, D., Meyer, J.M. and Mauro, T.M. (2016). Endoplasmic reticulum calcium regulates epidermal barrier response and desmosomal structure. Journal of Investigative Dermatology, 136(9), pp. 1840–1847. https://doi.org/10.1016/j.jid.2016.05.100
8. Menon, G.K., Feingold, K.R. and Elias, P.M. (1992). Lamellar body secretory response to barrier disruption. Journal of Investigative Dermatology, 98(3), pp. 279–289. https://doi.org/10.1111/1523-1747.ep12497866
9. Proksch, E., Elias, P.M. and Feingold, K.R. (1990). Regulation of 3-hydroxy-3-methylglutaryl-coenzyme A reductase activity in murine epidermis: modulation of enzyme content and activation state by barrier requirements. Journal of Clinical Investigation, 85(3), pp. 874–882. https://doi.org/10.1172/JCI114514
10. Feingold, K.R., Man, M.Q., Menon, G.K., Cho, S.S., Brown, B.E. and Elias, P.M. (1990). Cholesterol synthesis is required for cutaneous barrier function in mice. Journal of Clinical Investigation, 86(5), pp. 1738–1745. https://doi.org/10.1172/JCI114899
11. Holleran, W.M., Man, M.Q., Gao, W.N., et al. (1991). Sphingolipids are required for mammalian epidermal barrier function: inhibition of sphingolipid synthesis delays barrier recovery after acute perturbation. Journal of Clinical Investigation, 88(4), pp. 1338–1345. https://doi.org/10.1172/JCI115439
12. Ghadially, R., Halkier-Sørensen, L. and Elias, P.M. (1992). Effects of petrolatum on stratum corneum structure and function. Journal of the American Academy of Dermatology, 26(3 Pt 2), pp. 387–396. https://doi.org/10.1016/0190-9622(92)70060-s
13. van de Kerkhof, P.C., de Mare, S., Arnold, W.P. and van Erp, P.E. (1995). Epidermal regeneration and occlusion. Acta Dermato-Venereologica, 75(1), pp. 6–8. https://doi.org/10.2340/000155557568
A short note
The sequence described here draws chiefly on acute disruption of hairless-mouse skin. The timing of secretion, enzyme activity and DNA synthesis belongs to those conditions. The three waves are a way of explaining overlapping responses, not a fixed schedule. Basal DNA synthesis does not demonstrate that new mature corneocytes have reached the surface at the measured time.[2,8,9,11]
The calcium account combines studies of different compartments and models. Tissue imaging detects redistribution, while an ER-targeted indicator in reconstructed human epidermis provides a closer measurement of store release. Human explants, cultured cells, reconstructed tissue and living volunteers answer different questions. Pharmacological depletion reproduces parts of repair, but does not identify the earliest trigger of ER release or establish that every insult uses one exclusive route.[4–7]
Much of the foundational evidence came from the same UCSF and Veterans Affairs research programme. The inspected papers declare substantial public or institutional support, including NIH, Veterans Affairs and European research funding. The cholesterol-inhibition study also records that a Merck scientist supplied the lovastatin, which is a material contribution rather than funding. The experiments test barrier mechanisms, but their rescue results do not establish the performance of a finished skincare formulation.[2,5–7,9,10]
