Parakeratotic corneocytes play a unique role in human skin wound healing.

TO THE EDITOR Upon revisiting our large library of archived human skin wound specimens, we observed that the stratum lucidum interfollicular parakeratotic corneocytes appear to have an active role in epidermal wound healing by expanding, migrating, and bifurcating to interact directly with and secure the wound scab in place. Our observations are based on the morphologic analysis of 240 acute, normal, full-thickness, incisional, 1- to 21-day human upper arm and lower-leg skin wounds obtained from 30 healthy volunteers, with an average age of 64±8 years (mean±SD). Wound tissue was either ½ Karnovsky’s-fixed, PolyBed embedded, and Richardson’s-stained or Carnoy-fixed, paraffin-embedded, and hematoxylin and eosin–stained. All wounds were obtained in accordance with the Declaration of Helsinki Principles and the University of Washington Human Subjects Institutional Review Board. In the interfollicular epidermis, terminal differentiation begins in the granular layer of the epidermis, resulting in the formation of a cornified stratum corneum (SC), often described as the “dead” outer layer of the skin (Candi et al., 2005). The SC, however, has been shown to be a dynamic and metabolically interactive tissue acting as a biosensor to regulate metabolic responses in the underlying nucleated cell layers (Elias, 1996). The SC appears to be composed of three structurally (Brody, 1962) and functionally (Richter et al., 2004) identifiable sublayers: (1) the outermost layer composed of corneocytes that undergo desquamation (stratum disjunctum); (2) an intermediary zone of uniform-sized, tightly-compacted anucleate corneocytes (stratum compactum); and (3) the zone immediately adjacent to the stratum granulosum populated with parakeratotic corneocytes (stratum lucidum) (Kligman, 1964). Stratum lucidum parakeratotic corneocytes retain rod-shaped nuclei, ribosomes, lysosomes, mitochondria, Golgi apparatus, numerous granules, fibrillar structures (Ebling and Rook, 1972), and have “fragile” rather than “rigid” cornified envelopes identified in the upper layers of the SC (Haftek et al., 2011). The stratum lucidum is ~1–2 cell layers thick in normal interfollicular skin and is more prominently visible (~8–10 layers) in palmoplantar skin and lip (Ebling and Rook, 1972). Stratum lucidum corneocytes in interfollicular skin have been distinguished from the upper SC corneocytes by the use of unique tissue fixations and histochemical stains (Montagna et al., 1992) but are poorly discernable in routine hematoxylin and eosin–stained tissue. Organelles are only sporadically identifiable in corneocytes in this transition zone even using transmission electron microscopy (Montagna et al., 1992). Teleologically, a transient zone must exist where major cellular changes (loss of nuclei, organelles, and keratohyalin granules) take place; however, the difficulty of identifying the zone lies in the fact that this metamorphosis occurs with great speed (Kligman, 1964). Upon injury to the skin, the clotting cascade is initiated in order to stop the hemorrhaging of damaged vessels (Singer and Clark, 1999). A clot filled with fibrin, blood components, wound debris (collagen and elastin fragments), and glycoproteins (Singer and Clark, 1999) provides a “sticky” plug that covers and fills the wound bed. As wound healing progresses, the clot appears to compartmentalize into two identifiable regions, with the upper region of the clot separated from the lower region by a “polyband” layer of polymorphonuclear leukocytes (Jonkman et al., 1988). This upper clot region desiccates, forming a scab, crust, or eschar that sloughs eventually during tissue repair, whereas the lower fibrin-rich region of the clot serves as the wound bed in which granulation tissue forms (Singer and Clark, 1999). In this study, we show that in acute human skin wounds a unique population of parakeratotic corneocytes interacts directly with scabs, appearing to secure a temporary barrier to cover a wound until the underlying epidermis fully epithelializes, terminally differentiates, and restores a permanent skin barrier. Along the wound margin of a 6-μm tissue section of a 1-day wound immunolabeled with a pan-cytokeratin antibody (Dako, Carpenteria, CA), parakeratotic corneocytes above granular layer corneocytes expand in number and begin to migrate laterally toward the wound exudate (Figure 1a). In the 2-day wound (Figure 1b), the parakeratotic corneocyte population then bifurcates with a second population of parakeratotic corneocytes appearing to stream ventrally down along the underlying epidermis. In the 3-day wound (Figure 1c), the ventral branch parakeratotic corneocytes travel beyond the migrating tip (arrowhead) of the underlying nucleated migrating epidermis. By 7 days (Figure 1d), a mature scab has formed above an epithelialized new wound epidermis, showing parakeratotic corneocytes interacting with the upper region of the scab and undermining the base of the scab. Figure 1e shows a schematic illustration of the relationship between parakeratotic corneocytes and scabs as we observed. Forty-four percent (106/240) of the wounds appeared to have identifiable scabs (many immature wounds not yet forming scabs), of which 93% (98/106) of the wounds with scabs showed a pattern of scab/parakeratotic corneocyte interaction (Figure 1f). On occasion, scabs mechanically (either from tissue processing or from handling during tissue harvest) separated from the underlying new epithelium. These scabs appeared to be attached to the SC of the wound tissue sample (Figure 1g–i). This finding suggests that scabs adhere to the wound bed not merely by their “stickiness” but by direct interaction with the SC. The earliest observation of wound keratinocyte interaction with the scab was made by Leo Loeb in 1898, who described and illustrated in great detail that the granular and “horny” (SC) layers resolved into a homogeneous multinucleated protoplasmic layer, completely independent of the underlying epidermal tongue. This protoplasmic layer, for which he saw no cellular borders, bifurcated, with the upper “arm” quickly integrating with or covering the scab, whereas the lower “arm” more slowly migrated to undermine the scab ahead of the underlying Malphigian (nucleated) keratinocytes. These protoplasmic cells adhered tightly to the scab until the scab was sloughed (Loeb, 1898). Zahir suggested that Loeb’s “upper protoplasmic layer was either coagulated exudate forming the most superficial part of the scab or layers of collagen fibers at the surface of the scab”. Zahir (1965) also described the branch extending over the upper surface of the scab as less well-developed cells with pyknotic nuclei. Rather than necrotic, pyknotic cells, our studies show that parakeratotic corneocytes appear to be viable, as indicated by their ability to migrate toward the wound. Viziam et al. stated that the expanded parakeratotic corneocyte population seen in wounds emanated from rapid differentiation of new wound suprabasal and stratum granulosum keratinocytes. They did not observe parakeratotic cells in the proximal portion of the migratory wound epithelial tongue, and concluded that the actively migrating epithelium had not yet undergone differentiation (Viziam et al., 1964). In contrast to studies by Viziam et al., we observed the presence of parakeratotic corneocytes beyond the tip of the underlying, migrating, nucleated tongue of early, 1–3-day wounds. The presence of parakeratotic corneocytes not in direct association with underlying granular layer corneocytes suggests that these parakeratotic corneocytes are not derived from an accelerated keratinocyte differentiation pathway of the new wound epidermis but appear to be derived as a much earlier response to injury. It appears that keratinocytes in the granular layer in response to injury continue to produce parakeratotic corneocytes that expand in number possibly by ceasing to differentiate into anucleate corneocytes. It is this expanded population of parakeratotic corneocytes that we believe independently interacts with the scab. The expansion of the parakeratotic population appears not to be by mitosis, as there is no evidence of Ki67 immunostaining within this cell population (Usui et al., 2005). These unique parakeratotic corneocytes may have a role in epidermal repair separate from the role of underlying differentiated suprabasal keratinocytes (keratinocytes not yet cornified) that have been postulated to actively participate in re-epithelialization by rolling onto the wound bed (Usui et al., 2005). Securely attaching a scab as a temporary barrier not only protects the wound bed, but formation of an attached transitory scab may have additional benefits in protecting the host from infection. Studies have shown that 99% of the bacterial population found in wounds is sequestered in scabs and not on or within wound beds (Barnett et al., 1986; Zhao et al., 2010). These morphologic observations necessitate additional characterization and mechanistic studies. The origin of the parakeratotic corneocytes emanating from the stratum lucidum is based strictly on our many static morphological images and is therefore hypothetical. We defined the parakeratotic keratinocytes we observed as being “corneocytes”; however, lipid membrane immaturity, fragility of cornified envelopes, and the presence of corneodesmosomes and ultrastructural components (cornified envelopes and organelles) characteristic of normal parakeratotic corneocytes must be further evaluated to clearly determine whether they are, in fact, corneocytes. In addition, studies need to be conducted to determine the mechanism by which corneocytes migrate (cytoskeletal machinery) and adhere (surface receptors) to the desiccating scab matrix. Although this was strictly a morphological, and not a mechanistic, study in which our observations show that wound parakeratotic keratinocytes interact directly with scabs, we concur with Kligman’s (1964) philosophy about morphological observations: “The usual sequence of biological knowledge is from the anatomical to the physiological.” The authors state no conflict of interest. This publication was made possible by grants from NIH AR43006, DK59221, EB004422, AR057115, NSF EEC9529161, VA R&D, RW Johnson Pharmaceutical Research Institute, and Advanced Tissue Sciences. Its contents are solely the responsibility of the authors and do not necessarily represent the official views of the NIH-NIAMS, -NIBIB, -NIDDK, or NSF. We thank The George F. Odland Endowed Research Fund, Marvin and Judy Young, John and Darcy Halloran, Peter Odland, and Peter Byers for their generous support.