Keratinocyte
Keratinocytes are the cells that make up almost all of the epidermis, differentiating from the basal layer upward and dying in a programmed way to build the stratum corneum barrier.
A keratinocyte begins in the basal layer expressing keratins 5 and 14, then commits to differentiation and switches to keratins 1 and 10 in the spinous layer. In the granular layer it produces keratohyalin granules containing profilaggrin, and lamellar bodies loaded with ceramide precursors, cholesterol, and free fatty acids. At the transition to the stratum corneum it extrudes those lipids, crosslinks involucrin and loricrin into a cornified envelope, and destroys its own nucleus, becoming a corneocyte. A rising calcium gradient drives the sequence.
Transit from basal division to shedding takes roughly a month and lengthens with age. The importance of the programme shows in its failures: loss-of-function mutations in filaggrin, derived from profilaggrin, are the strongest known genetic risk factor for atopic dermatitis, and disordered proliferation with incomplete differentiation underlies psoriatic plaque. Keratinocytes are also immunologically active, releasing interleukin-1 alpha and antimicrobial peptides including cathelicidin LL-37 and beta-defensins.
Identifying a target as keratinocyte rather than fibroblast changes what evidence is relevant and what has to be crossed. Barrier repair, desquamation, and pigment transfer are epidermal processes testable in keratinocyte culture or reconstructed epidermis, and a product acting on them never needs to reach the dermis. That makes such claims more plausible than dermal ones and easier to test honestly.
The framing to resist is that accelerating cell turnover is inherently good. Faster desquamation without matching lipid synthesis gives a thinner, leakier barrier, the mechanism behind over-exfoliation. Likewise, a keratinocyte proliferation assay reported as evidence of renewal is measuring the same variable that runs out of control in psoriasis, so proliferation alone is not a quality worth maximising.