The Central Roles of Keratinocytes in Coordinating Skin Immunity

Abstract

The function of keratinocytes (KCs) to form a barrier and produce cytokines is well-known, but recent progress has revealed many different roles for KCs in regulation of skin immunity. In this review, we provide an update on the current understanding of how KCs communicate with microbes, immunocytes, neurons, and other cells to form an effective immune barrier. We catalog the large list of genes and metabolites of KCs that participate in host defense and discuss the mechanisms of immune crosstalk, addressing how KCs simultaneously form a physical barrier, communicate with fibroblasts, and control immune signals. Overall, the signals sent and received by KCs are an exciting group of therapeutic targets to explore in the treatment of dermatologic disorders.

INTRODUCTION

Immune defense is primarily concerned with host protection, and keratinocytes (KCs) are clearly a critical cell type for defense against external challenges. In this review, we discuss the current understanding of how KCs are remarkably multifunctional and serve many essential roles by providing physical, biochemical, and immune defense. Progress in understanding how KCs communicate with microbes, immune cells, and nonimmune cells reveals the biological significance of these cells to immunity. Although this review seeks to compile and organize the large list and breadth of known KC functions, it does not purport to be a complete catalog of all we know about these complex cells. The goal of this review is to provide a resource for better understanding of the critical functions of KCs in coordinating the host immune responses.

BARRIER FUNCTION IS A FORM OF IMMUNITY

Formation of a physical barrier

The most obvious role of KCs is to form a barrier to protect the body from mechanical injury, dehydration, UVR, pathogens, and other environmental dangers. Terminally differentiated KCs in the stratum corneum are strengthened by a rigid keratin–FLG matrix and the cornified cell envelope, a tough, crosslinked protein structure that lines the interior of the cell membrane. Beneath these cornified cells in the stratum granulosum, intercellular space is sealed by tight junctions (Brandner et al, 2002), and KCs produce several proteins, lipids, and carbohydrates to form a continuous, robust barrier (Brandner et al, 2002; Selby, 1957). The unique brick and mortar structure of corneocytes and lipids allows the epidermis to limit evaporative transepidermal water loss as well as provide an innate antimicrobial and antioxidative barrier. Specialized KCs in the follicular unit have unique functions and must control a higher density of microbes and in this specialized environment (Adachi et al, 2015). Indeed, recognizing the surface area within follicular structures makes the skin the largest area of the human body for communication with the microbiome (Gallo, 2017).

The physical properties of the epidermis are influenced by calcium signaling. A concentration gradient of calcium ions regulates formation of desmosomes, adherens junctions, tight junctions, inositol triphosphate (IP₃), and other products necessary for differentiation. An ongoing cycle perpetuates this process as IP₃ stimulates the release of more calcium ions from intracellular stores, thus inducing differentiation in nearby KCs as they migrate out toward the stratum corneum (Jaken and Yuspa, 1988; Xie and Bikle, 1999). Recent work has defined corneoptosis, a process through which KCs undergo intracellular acidification and cell death, losing their nuclei and other organelles (Matsui et al, 2021; Moore and Greco, 2021). Granular KCs have been described to take on a unique, flattened hexagonal shape known as Kelvin’s tetrakaidecahedron (Yokouchi et al, 2016). This unique shape contributes physically to the mechanism of cell turnover (desquamation), and it enables KCs to migrate upward while simultaneously maintaining a tight permeability barrier. Together, corneoptosis and desquamation make the epidermis better equipped to resist physical damage, UVR, or other environmental stresses than if permanent, living cells found at other epithelial surfaces were located on the outermost surface. The cornified layer of the epidermis also maintains an acidic environment (average pH of 5.4–5.9) in what is known as the acid mantle, which limits pathogen survival (Braun-Falco and Korting, 1986). Adding to the barrier functions of the epidermis, the products of sebaceous and eccrine glands in the skin regulate the composition of the microbiome and have critical roles in temperature regulation (Swaney et al, 2023). The acidic epidermal environment also restrains the activity of proteases such as kallikreins 5 and 7, which in neutral pH would reduce barrier integrity (Gunathilake et al, 2009). Several lipids are critical components of the epidermal barrier but will not be further discussed in this review.

Defense by antimicrobials from KCs

KCs actively produce a range of molecules with direct antimicrobial activity that controls the survival of microbes on the surface and defends against pathogens (Gallo and Hooper, 2012). These antimicrobials include a wide array of antimicrobial peptides (AMPs) and proteins and antimicrobial fatty acids (AFAs). Table 1 provides an extensive list of AMPs and lipids produced by KCs as well as many of the other classes of molecules that are expressed by KCs to influence immune defense.

Table 1.

Major Keratinocyte Receptors and Products that Participate in Immune Responses

Products		References
Antimicrobial peptides, proteins, and lipids		Antimicrobial peptides and proteins
	α-MSH	Donnarumma et al, 2004; Luger et al, 1993
	β-Defensin 1	Abiko et al, 2003; Mathews et al, 1999; Seo et al, 2001
	β-Defensin 2	Mathews et al, 1999; Seo et al, 2001
	β-Defensin 3	Abiko et al, 2003; Pernet et al, 2003
	Cysteine-rich tail protein 1	Niehues et al, 2023
	LCE3A, LCE3B, LCE3C	Archer et al, 2017; Niehues et al, 2017
	Lipocalin-2	Bando et al, 2007; Wolk et al, 2017
	LL-37 (Cathelicidin)	Dorschner et al, 2001; Frohm et al, 1997
	RNase 7	Harder and Schroder, 2002
	SPRR1A (Cornifin-A)	Sark et al, 1998; Zhang et al, 2022
	SPRR2A	Fischer et al, 1996; Zhang et al, 2022
	S100A7 (Psoriasin)	Madsen et al, 1991; Wolf et al, 2011; Zhu et al, 2013
	S100A7A (Koebnerisin)	de Castro et al, 2014; Gudjonsson et al, 2010; Wolf et al, 2011
	S100A8/9 (Calprotectin)	Eversole et al, 1993; Mørk et al, 2003; Zhu et al, 2013
		Antimicrobial lipids
	Arachidonic acid	Dubos, 1947; Ponec et al, 1988
	Dihydro-sphingosine	Fischer et al, 2012; Łuczaj et al, 2020; Stewart and Downing, 1995
	Palmitic acid	Grayson and Elias, 1988; Kabara et al, 1972; Ponec et al, 1988
	Phytosphingosine	Fischer et al, 2012; Łuczaj et al, 2020; Mizutani et al, 2004
	Sapienic acid (human)	Drake et al, 2008; Fischer et al, 2012; Ponec et al, 1988
	Sphinganine	Bibel et al, 1992; Kim et al, 2011
	Sphingosine	Bibel et al, 1992; Fischer et al, 2012; Kim et al, 2011
	Receptors	Potential Ligands	References
Microbe and danger recognition receptors		Inflammasome family
	AIM2	Cytosolic DNA	Dombrowski et al, 2011; Kopfnagel et al, 2011
	NLRP1	dsDNA, dsRNA, bacterial lethal toxin, poly(dA:dT)	Burian and Yazdi, 2018; Sand et al, 2018; Zhou et al, 2023b
	NLRP3	Wide range of PAMPs, DAMPs, and particles	Cho et al, 2012; Que et al, 2024; Watanabe et al, 2007
	NOD1	Peptidoglycan fragments, iE-DAP	Harder and Núñez, 2009
	NOD2	Peptidoglycan fragments, muramyl dipeptide	Voss et al, 2006
		TLR family
	TLR1/2	Triacyl lipopeptides, peptidoglycans	Baker et al, 2003; Kawai et al, 2002
	TLR2	Glycolipids, lipoteichoic acid, peptidoglycans	Kawai et al, 2002
	TLR2/6	Diacyl lipopeptides, Malp-2	Büchau et al, 2008; Kawai et al, 2002; Vu et al, 2010
	TLR3	dsRNA, Poly(I:C)	Lebre et al, 2007; Mempel et al, 2003
	TLR4	LPS	Lebre et al, 2007; Song et al, 2002
	TLR5	Flagellin	Baker et al, 2003; Lebre et al, 2007
	TLR9	Unmethylated CpG DNA	Lebre et al, 2007; Mempel et al, 2003
	TLR10	Unknown	Köllisch et al, 2005
		Other microbe and danger recognition receptors
	CD14	LPS	Köllisch et al, 2005
	CD280	Mannose, N-acetylglucosamine, and fucose residues on glycans	Szolnoky et al, 2001
	Dectin-1	β-glucan	Kobayashi et al, 2009
	Ly-96	LPS, diacyl lipopeptides	Kollisch et al, 2005
	MDA5	dsRNA	Prens et al, 2008
	RIG-I	Viral dsRNA	Kalali et al, 2008
	STING	Cytosolic DNA	Baird et al, 2017; Nishikawa et al, 2018
Cytokines and GFs		CC-motif chemokines
	CCL2	Barker et al, 1991
	CCL5	Li et al, 1996; Yamada et al, 1997
	CCL11	Valdebran et al, 2019; Wakugawa et al, 2000
	CCL17	Dillon et al, 2004; Vestergaard et al, 1999; Yu et al, 2002
	CCL20	Charbonnier et al, 1999; Homey et al, 2000
	CCL22	Rottman et al, 2001; Xiao et al, 2003
	CCL26	Igawa et al, 2006; Kagami et al, 2005b
	CCL27	Kakinuma et al, 2003; Morales et al, 1999
	CCL28	Kagami et al, 2005a; Kagami et al, 2006
		CXC-motif chemokines
	CXCL1	Richmond and Thomas, 1988; Tettelbach et al, 1993
	CXCL2	Enk et al, 1993; Nukui et al, 2008; Wetzler et al, 2000
	CXCL3	Farley et al, 2006; Nograles et al, 2008
	CXCL5	Boniface et al, 2005; Nograles et al, 2008
	CXCL6	Kim et al, 2009; Nograles et al, 2008
	CXCL8	Larsen et al, 1989; Nograles et al, 2008; Takematsu et al, 1990b
	CXCL9	Albanesi et al, 2000; Tensen et al, 1999
	CXCL10	Albanesi et al, 2000; Gottlieb et al, 1988; Tensen et al, 1999
	CXCL11	Albanesi et al, 2000; Tensen et al, 1999
	CXCL14	Meuter and Moser, 2008; Sleeman et al, 2000
	CXCL16	Scholz et al, 2007; Tohyama et al, 2007
		Colony-stimulating factors
	G-CSF	Arlian et al, 2003; Denburg and Sauder, 1986; Takematsu et al, 1990a
	GM-CSF	Kupper et al, 1986b; Kupper et al, 1988a; Kupper et al, 1988b
	IL-3	Danner and Luger, 1987; Gallo et al, 1991; Luger et al, 1985
	IL-34	Greter et al, 2012; Wang et al, 2012
	M-CSF	Chodakewitz et al, 1990
		IFNs
	IFN-β	Zhang et al, 2016
	IFN-γ	Howie et al, 1996
	IFN-κ	LaFleur et al, 2001; Scarponi et al, 2006
		IL-1 family
	IL-1α	Kupper et al, 1986a; Luger et al, 1981
	IL-1β	Kupper et al, 1986a; Luger et al, 1981
	IL-1Ra	Bigler et al, 1992
	IL-18	Mee et al, 2000
	IL-18BP	Mee et al, 2000; Mühl et al, 2000
	IL-33	Balato et al, 2012; Moussion et al, 2008
	IL-36α	Johnston et al, 2011; Williams and Kim, 2022
	IL-36β	Boutet et al, 2016; Johnston et al, 2011
	IL-36γ	Johnston et al, 2011; Lian et al, 2012
	IL-36Ra	Boutet et al, 2016; Johnston et al, 2011
	IL-37	Fujita et al, 2013; Muhr et al, 2011; Smithrithee et al, 2015
	IL-38	Boutet et al, 2016; Lachner et al, 2017
		IL-6/IL-12 superfamily
	IL-6	Grossman et al, 1989; Oxholm et al, 1991; Yoshizaki et al, 1990
	IL-12	Müller et al, 1994; Yawalkar et al, 1996
	LIF	McKenzie et al, 1996; Paglia et al, 1996
	OSM	Canady et al, 2013)
		IL-10 family
	IL-10	Grewe et al, 1995; Nickoloff et al, 1994
	IL-19	Bao et al, 2014; Sun et al, 2013
	IL-20	Cornelissen et al, 2012; Hunt et al, 2006
	IL-24	Cornelissen et al, 2012; Poindexter et al, 2010
		Other ILs
	IL-7	Dalloul et al, 1992; Heufler et al, 1993; Sieling et al, 1995
	IL-15	Blauvelt et al, 1996; Mohamadzadeh et al, 1995; Rückert et al, 2000
	IL-16	Frezzolini et al, 2004; Shiraki et al, 2006
	IL-17C	Johansen et al, 2010; Ramirez-Carrozzi et al, 2011
	IL-25	Bao et al, 2013; Leyva-Castillo et al, 2020; Wang et al, 2007
		TNF superfamily
	LT-α	Fania et al, 2020; Middel et al, 2000
	TNF	Köck et al, 1990
FasL	Leverkus et al, 1997; Maeda et al, 2001
TRAIL	Kothny-Wilkes et al, 1998
TWEAK	Chen et al, 2011
APRIL	Chen et al, 2011; Mhawech-Fauceglia et al, 2006
BAFF	Chen et al, 2011
TL1A	Bamias et al, 2011
	VEGF/PDGF family GFs
PDGF-AA, AB, BB	Ansel et al, 1993; Zhang et al, 1995
PlGF	Failla et al, 2000
VEGFA, B, C	Bae et al, 2015; Detmar et al, 1995; Frank et al, 1995; Larcher et al, 1998; O-charoenrat et al, 2001; Trompezinski et al, 2004
	Other cytokines and GFs
Amphiregulin	Cook et al, 1991; Kobayashi et al, 1998; Shoyab et al, 1989
CTGF	Camerer et al, 2000; Dazard et al, 2003; Rittié et al, 2011
Endothelin-1	Imokawa et al, 1992; Yohn et al, 1993
FGF2	Halaban et al, 1988; Itoh et al, 2019
FGF5	Camerer et al, 2000
FGF-BP	Beer et al, 2005
HB-EGF	Hashimoto et al, 1994; Kobayashi et al, 1998
IGF-I	Rudman et al, 1997
β-NGF	Di Marco et al, 1991; Tron et al, 1990
Semaphorin 3A	Fukamachi et al, 2011; Kurschat et al, 2006
TGF-α	Coffey et al, 1987; Elder et al, 1989; Kobayashi et al, 1998
TGF-β1, TGF-β2	Elder et al, 1989; Kane et al, 1990; Xia et al, 2006)
TSLP	Demehri et al, 2009; Li et al, 2006; Soumelis et al, 2002
Wnt-3a	Reyes et al, 2019; Sun et al, 2021
	Receptors	Potential Ligands	References
Cytokine and GF receptors		CC-motif chemokine receptors
	CCR1	CCL2-7,14-16,23	Kroeze et al, 2012
	CCR3	CCL5,7,11,13,15,24,26,28	Petering et al, 2001; Wakugawa et al, 2001
	CCR4	CCL3,5,17,22	Fujimoto et al, 2008; Kroeze et al, 2012
	CCR6	CCL20	Kroeze et al, 2012
	CCR9	CCL25	Bünemann et al, 2018
	CCR10	CCL27,28	Fujimoto et al, 2008; Kroeze et al, 2012
		CXC-motif chemokine receptors
	CXCR1	CXCL1,2,3,5,8	Kondo et al, 2000; Kroeze et al, 2012
	CXCR2	CXCL1,2,3,5,8	Kondo et al, 2000; Kroeze et al, 2012
	CXCR3	CXCL9,10,11	Kroeze et al, 2012; Kurata et al, 2010
	CXCR4	CXCL12	Avniel et al, 2006; Kroeze et al, 2012
		IL-1 receptor family
	IL-1R1/IL-1R3	IL-1α, IL-1β, IL-1Ra	Blanton et al, 1989
	IL-1R2	IL-1α, IL-1β, IL-1Ra	Groves et al, 1994
	IL-18R1/IL-1R7	IL-18, IL-37	Mee et al, 2000
	IL-1RL1 (ST2)/IL-1R3	IL-33	Shimizu et al, 2005; Zeng et al, 2021
	IL-36R/IL-1R3	IL-36α, IL-36β, IL-36γ, IL-36Ra	Hernández-Santana et al, 2020; Johnston et al, 2011
		IL-2/common gamma chain receptor family
	γC/IL-2Rα/IL-2Rβ	IL-2	Yasumura et al, 1994
	γC/IL-4Rα	IL-4	Junghans et al, 1996; Progneaux et al, 2023
	γC/IL-9R	IL-9	Hong et al, 2015; Kamiya et al, 2022
	γC/IL-15Rα/IL-2Rβ	IL-15	Ruckert et al, 2000
		IL-6/IL-12 receptor superfamily
	IL-6Rα/IL-6Rβ	IL-6	Grossman et al, 1989; Yoshizaki et al, 1990
	IL-12Rβ1/IL-12Rβ2	IL-12	Kanda and Watanabe, 2008
	IL-12Rβ1/IL-23R	IL-23	Kanda and Watanabe, 2008
	IL-6Rβ/IL-27Rα	IL-27	Kanda and Watanabe, 2008
	IL-6Rβ/LIF-R	OSM, LIF	Paglia et al, 1996)
	IL-6Rβ/OSM-Rβ	OSM	Boniface et al, 2007
	OSM-Rβ/IL-31Rα	IL-31	Dillon et al, 2004
		IL-10 receptor family
	IL-10Rα/IL-10Rβ	IL-10	Michel et al, 1997; Müschen et al, 1999
	IL-10Rβ/IL-20Rα	IL-26	Itoh et al, 2019; Kamijo et al, 2020
	IL-10Rβ/IL-22Rα1	IL-22	Boniface et al, 2005
	IL-20Rα/IL-20Rβ	IL-19, IL-20, IL-24	Blumberg et al, 2001; Wang et al, 2002
	IL-20Rβ/IL-22Rα1	IL-20, IL-24	Boniface et al, 2005; Wang et al, 2002
		Other IL receptors
	IL-4Rα/IL-13Rα1	IL-4, IL-13	Akaiwa et al, 2001; Junghans et al, 1996
	IL-13Rα2	IL-13	David et al, 2001; Lü et al, 2009
	IL-17RA/IL-17RC	IL-17A, IL-17F	Albanesi et al, 1999
	IL-17RA/IL-17RE	IL-17C	Ramirez-Carrozzi et al, 2011
		IFN receptors
	IFN-R1/IFN-R2	Type 1 IFNs (ie, 13 IFN-α subtypes, IFN-β, IFN-ε, IFN-κ, IFN-ω)	Gui et al, 2016; Stadler et al, 1986; Wittling et al, 2020
	IFN-γR1/IFN-γR2	IFN-γ	Dustin et al, 1988; Nickoloff, 1987
	IFN-λR1/IL-10Rβ	IFN-λ1,2,3	Goel et al, 2020; Zhang et al, 2011
		TNF receptor superfamily
	TNF-R1	TNF	Dustin et al, 1988; Kondo and Sauder, 1997
	LT-βR	LT-α/LT-β, LIGHT	Krutmann et al, 1990; Symington, 1989
	CD40	CD40L	Denfeld et al, 1996; Gaspari et al, 1996
	FasL-R	FasL	Leverkus et al, 1997; Sayama et al, 1994
	DcR1, DcR2	TRAIL	Kothny-Wilkes et al, 1998; Maeda et al, 2001
	DcR3	FasL, LIGHT, TL1A	Maeda et al, 2001; Wu et al, 2013
	DR4, DR5	TRAIL	Kothny-Wilkes et al, 1998; Maeda et al, 2001
	TWEAK-R	TWEAK	Jin et al, 2004
	HVEM	LIGHT	Herro et al, 2018
	NGF-R	β-NGF, BDNF, NT-3, NT-4	Botchkarev et al, 1999; Di Marco et al, 1993; Pincelli et al, 1994
		Other cytokine and GF receptors
	EGF-R	EGF, TGF-α, HB-EGF, amphiregulin, epiregulin, betacellulin, epigen	Cohen, 1962; Foroughi et al, 2019; O’Keefe et al, 1982; Tsuboi et al, 1992
	FGFR1-IIIb	FGF1, 2, 3, 7, 10, 22	Bottaro et al, 1990; O’Keefe et al, 1988; Takenaka et al, 1997; Takenaka et al, 2002
	FGFR2-IIIb (KGF-R)	FGF1, 3, 7, 10, 22	Stanulis-Praeger and Gilchrest, 1986; Takenaka et al, 2002
	FGFR3-IIIb	FGF1, 9, 16, 20	Avivi et al, 1993; Takenaka et al, 2002; Tsuboi et al, 1992
	HGF-R	HGF	Sato et al, 1995; Tsuboi et al, 1992
	IGF-1R	IGF-I, IGF-II, insulin	Rudman et al, 1997; Tsuboi et al, 1992
Neuropilin-1	Various ligands (eg, PlGF, VEGFA/B, semaphorin 3A)	Kurschat et al, 2006; Man et al, 2006; Zhu et al, 2012
Neuropilin-2	Various ligands (eg, TGF-β, semaphorin 3C/F, VEGFA/B)	Man et al, 2006; Rushing et al, 2012
TGF-R1, TGF-R2, TGF-R3	TGF-β1, TGF-β2, TGF-β3	Hebda, 1988; Shipley et al, 1986; Tam et al, 1998

Abbreviations: BP, binding protein; CTGF, connective tissue GF; DAMP, damage-associated molecular pattern; DcR, decoy receptor; DR, death receptor; dsDNA, double-stranded DNA; dsRNA, double-stranded RNA; EGF, epidermal GF; FGF, fibroblast GF; GM-CSF, granulocyte macrophage colony stimulating factor; HB-EGF, heparin-binding epidermal GF; HGF, hepatocyte GF; HVEM, herpesvirus entry mediator; iE-DAP, γ-D-glutamyl-meso-diaminopimelic acid; IGF, insulin-like GF; KGF, keratinocyte GF; LIF, leukemia inhibitory factor; LPS, lipopolysaccharide; LT, lymphotoxin; MSH, melanocyte-stimulating hormone; NGF, nerve GF; NT, neurotrophin; OSM, oncostatin M; PAMP, pathogen-associated molecular pattern; Poly(dA:dT), poly(deoxyadenylic-deoxythymidylic) acid; Poly(I:C), polyInosinic:polycytidylic acid; STING, stimulator of IFN genes; TLR, toll-like receptor.

The table shows the list of keratinocyte products organized by class. References cited represent descriptions of gene products found in human keratinocytes. Molecules described only in nonhuman species or transformed cells have been excluded as well as those found only in nonvalidated transcriptional datasets. Molecules such as lipids and proteins involved in physical barrier properties of the epidermis are not listed unless described to have direct antimicrobial properties.

Among the most abundant and impactful antimicrobials produced by KCs are cathelicidin (LL-37), β-defensins, and S100 family proteins. The function of most AMPs relies on their cationic and amphiphilic nature, which enables them to disrupt relatively electronegative bacterial membranes (ShaiMatsuzaki-Huang model). Other antimicrobials, such as some S100 family proteins and LCN2, act by limiting bacterial acquisition of necessary ions such as zinc or iron (Wolf et al, 2011; Wolk et al, 2017). In addition to killing bacteria, some antimicrobials (eg, CCL28, LCN2, LL-37, S100A7) boost chemotactic activity (Agerberth et al, 2000; Schroll et al, 2012; Wolf et al, 2008; Yung et al, 2011). Specific structures of AMPs and AFAs provide varying selectivity and potency in killing gram-positive or gram-negative bacteria, viruses, and fungi (Dorschner et al, 2001; López-García et al, 2005). Basal levels of AMP production help to regulate the proliferation of microbiota that inhabit the skin, but production of the most potent AMPs is strictly regulated. Certain KC AMPs require either indirect induction by immune signals such as IL-17A or TNF or direct induction through pathogen-associated molecular pattern (PAMP) recognition receptors such as toll-like receptors (TLRs) 2 or 3 (Lai et al, 2009; Liu et al, 2006; Miller and Modlin, 2007).

Triggering an inflammatory response

Another important function of KCs is to sound the alarm through cytokine production and release. KCs express many different cytokines, such as ILs (eg, IL-1, IL-6, IL-36), chemokines (eg, CXCL8, CCL20), TNF, IFNs (eg, IFN-β, IFN-γ, IFN-κ), and TSLP (Table 1). These secreted proinflammatory signals recruit specialized leukocytes to the site of injury or infection and are coordinated to address the specific form of danger. Some cytokines are stored within the cell and rapidly released, whereas others are transcriptionally regulated. Thus, inflammatory responses of KCs can be triggered by passive release from ruptured cells (Bernard and Gallo, 2011; Cordier-Dirikoc et al, 2022; England et al, 2014; Kouzaki et al, 2011), detection of PAMPs (Miller and Modlin, 2007; Olaru and Jensen, 2010), or signals from other cells in the skin (Kamijo et al, 2020; Oka et al, 2017). KCs also receive inflammatory signals through a wide variety of cytokine and chemokine receptors (Table 1). A key to advancing our understanding of innate immunity in the skin is understanding the interconnected feedback mechanisms between KCs and a variety of cell types. In the following sections of this review, we delve into many of the specific interactions between KCs and other cells, and we address some important gaps in knowledge that we hope future studies can fill.

KCs COORDINATE IMMUNE NETWORKS

KC–microbiome interactions

KCs have a vast network of cell communication systems, and considering this network from the outside in, the first interactions to discuss are those between KCs and the microbes that inhabit the skin surface. Figure 1 illustrates the relationship of microbes and KCs while summarizing the major functional concepts discussed in this review. The skin is home to a diverse microbiome (Grice et al, 2009; Saheb Kashaf et al, 2022; Schommer and Gallo, 2013) containing bacteria, fungi, viruses, and parasites that have adapted to a seemingly hostile environment with an acidic pH and low nutrient levels (Swaney et al, 2023). Although DNA-sequencing methodologies may not accurately reflect the viable microbes that function as part of the human skin holobiome (Acosta et al, 2023; Burger and Gallo, 2023), the skin microbiome is essential for normal immune defense (Nakatsuji et al, 2021). These microbes feed on phospholipids, sphingolipids, cholesterol, sugars, and proteins produced by KCs. Beneficial commensal bacteria compete with pathogens and pathobionts for nutrients and even secrete their own AMPs and AFAs (Almoughrabie et al, 2023; Claesen et al, 2020). Although the corneocyte protein meshwork and lamellar-lipid membranes of the stratum corneum prevent many bacteria from penetrating the skin (Feingold and Elias, 2014), diverse microbes have been found deep in subepidermal tissue, capable of directly interacting with other skin cells below the basement membrane (Nakatsuji et al, 2013).

Figure 1. Multiple signaling mechanisms for keratinocytes in immunity.

(a) List of multiple antimicrobial molecules made by keratinocytes and the receptors involved in recognizing microbial and environmental danger signals. (b) Keratinocytes have been described to produce an extensive array of cytokines, chemokines, ILs, and GFs that communicate with surrounding cells, including fibroblasts, adipocytes, melanocytes, sensory neurons, macrophages, monocytes, dendritic cells, Langerhans cells, innate lymphoid cells, T cells, neutrophils, eosinophils, basophils, mast cells, and endothelial cells. Many receptors present on keratinocytes alter their function in response to signals sent by these adjacent cell types and/or other keratinocytes. Table 1 provides details for receptor ligands and references. MSH, melanocyte-stimulating hormone; TLR, toll-like receptor. Illustration assistance provided by Jan Ruvido Stebbins, Ruvido Medical Illustration, Dexter, MI.

Because commensal and mutualistic bacteria inhabiting the skin often produce the same PAMPs as pathogenic bacteria, KCs must determine whether a potentially damaging inflammatory response is needed. Staphylococcus epidermidis, a normal inhabitant of healthy skin, secretes a unique structure of lipoteichoic acid to inhibit KC TLR3 activation (Lai et al, 2009). S epidermidis also releases extracellular vesicles that promote KC expression of β-defensin 2/3, limiting the growth of pathogenic S aureus (Zhou et al, 2023a). KC–microbe interactions are particularly important in the gaps of the physical epidermal barrier created by hair follicles. In the hypoxic conditions of a hair follicle, Cutibacterium acnes produces short-chain fatty acids (SCFAs) propionate and valerate, which inhibit histone deacetylase (HDAC)8/9 activity (Sanford et al, 2016). This hair follicle–specific C acnes activity breaks epidermal immune tolerance through increased TLR2/3 signaling and subsequent cytokine production in KCs (Sawada et al, 2021). KC CCL20 may also be involved in regional tolerance of commensal microbes because anti-inflammatory regulatory T cells are preferentially recruited by KC CCL20 and are significantly more abundant in hair follicles than in interfollicular skin (Sanchez Rodriguez et al, 2014; Scharschmidt et al, 2017). Thus, KC receive signals from commensal bacteria that assist in their capacity to tolerate some microbes while defending against others.

KC–T-cell interactions

KCs are closely intertwined with T cells as they each send and receive a variety of cytokine signals. KCs constitutively express receptors for many T-cell–derived cytokines, including IL-4, IL-13, IL-17A, IL-22, and others (Table 1 and Figure 1) (Nograles et al, 2008). T-cell–stimulated KCs can produce high levels of antimicrobials and a range of prionflammatory cytokines that activate and recruit more T cells as well as other immunocytes. In type 2 inflammation, scratching causes barrier disruption that releases IL-33 and TSLP from KCs, which recruit T helper (Th)2 cells that provide IL-4, IL-13, and IL-31 (Komai-Koma et al, 2007; Oyoshi et al, 2010; Tatsuno et al, 2015). These cytokines affect KC differentiation and trigger increased pruritus (itchiness) and continuous escalation of inflammation. In response to IL-4 and IL-13, KC produce additional IL-33 and CCL26 (Kagami et al, 2005b). Simultaneously, the IL-31 drives KC production of CCL17 and CCL22 (Dillon et al, 2004; Kagami et al, 2005b), which further amplify the chemotaxis for Th2 cells and subsequently create an inflammatory positive feedback loop. IL-4 also stimulates KC production of IL-19 (Bao et al, 2014), which helps to amplify Th2 cytokine production (Liao et al, 2004).

Outside of the classic Th2 cytokines, Th1, Th17, and Th22 cells produce IL-22, which promotes KC expression of IL-33, TSLP, and the receptor for the potent pruritogen GRP (Lou et al, 2017). IL-22 also inhibits KC differentiation (Boniface et al, 2005; Nograles et al, 2009). IL-17A, which is produced by Th17 cells, γδ T cells, and NK T cells (Sutton et al, 2009), is an important regulator of KC antimicrobials and inflammatory cytokines (Teunissen et al, 1998). Th17 cells release IL-26, which induces CXCL8, IL-1β, CCL20, IL-33, and β-defensin 2 production in KCs through the activation of signal transducer and activator of transcription (STAT)1 and STAT3 (Kamijo et al, 2020). Th1 cells have long been known as substantial producers of TNF, which can promote a variety of strong inflammatory responses in KCs by working synergistically with other cytokines. When combined with IL-17A, TNF promotes chemotaxis of Th17 cells through KC production of CCL20 (Harper et al, 2009). Alongside Th1 cell–produced IFN-γ, TNF induces KC production of IL-33 (Taniguchi et al, 2013). Similarly, MAPK and NF-κB signaling in IL-1–activated KC is enhanced by Th1 cell TNF (Komine et al, 2001). Alone, Th1 IFN-γ can promote KC expression of CXCL9, CXCL10, and CXCL11 through Jak/STAT signaling, leading to recruitment of additional T cells to the skin (Blazanin et al, 2019; Mee et al, 2007). Under normal, healthy conditions, KC utilize many of these networks to communicate with several distinct populations of resident T cells (Watanabe et al, 2015) and mobilize additional immune cells to sites of danger. Skin epithelial stem cells that are responsible for repopulation of KCs in the epidermis have also been reported to confer epigenetic immunologic memory to specific skin locations, thus potentially influencing site-specific inflammatory reactions (Naik et al, 2017).

KC–mononuclear myeloid cell interactions

Communication between KCs and mononuclear myeloid cells is critical during inflammation. KCs contain NRF2, a transcription factor that is stabilized and translocated to the nucleus under oxidative stress conditions and activates the expression of the monocyte-recruiting chemokine CCL2 (Villarreal-Ponce et al, 2020). Macrophages that are recruited to injured skin are also induced by CCL2 to produce epidermal GF (EGF), which promotes KC proliferation and initiates a positive feedback loop (Gao et al, 2016). EGF also helps to promote Granulocyte macrophage colony-stimulating factor (GM-CSF) production in KCs (Mascia et al, 2010), leading to increased macrophage infiltration of skin tissue and maturation of dendritic cells such as Langerhans cells. KCs have been known as producers of GM-CSF for decades (Kupper et al, 1988a), and KC-derived GM-CSF is critical to wound healing and inflammation (Mann et al, 2001).

In damaged skin tissue, KCs may break innate immune tolerance of dermal dendritic cells and prevent infection by opportunistic pathogens (Lande et al, 2007). Positively charged, KC-produced LL-37 binds to the negatively charged extracellular DNA released by ruptured KCs and triggers endocytosis into dendritic cells, activation of TLR9, and expression of IFNs. KCs can also activate dendritic cells with IFN-κ or IL-36γ, and dendritic cells are substantial contributors of the previously discussed TNF (Fiskin et al, 2023¹; Komine et al, 2007; Lowes et al, 2005). Various monocytes as well as activated T cells generate oncostatin M (OSM), which helps to stimulate KC β-defensin 2 (Kanda and Watanabe, 2012; Suda et al, 2002). Among KC cytokines that have also been shown to recruit monocytes are CXCL1 (Neels et al, 2009), CXCL8 (Sakai et al, 2015), and CXCL10 (Klunker et al, 2003).

KC–granulocyte interaction

Another key communication pair is between KCs and granulocytes. In addition to IL-17 from T cells, neutrophils are significant producers of IL-17A, which activates chemokine and AMP production in KCs (Liu et al, 2022; Reich et al, 2015). These KC chemokines, in turn, amplify the system as strong chemoattractants for neutrophils (Schumacher et al, 1992). Activated neutrophils recruited by KCs also contribute to host defense through the expulsion of nuclear DNA, histones, proteases, and antimicrobials in neutrophil extracellular traps (NETs)—a process known as NETosis (Brinkmann et al, 2004). NETs have been shown to stimulate KC production of β-defensins (Hu et al, 2016) and may activate KC inflammasomes (Dombrowski et al, 2011; Kahlenberg et al, 2013). Neutrophils can help to amplify the distress signals of wounded KCs by providing required proteases through NETosis or degranulation (Henry et al, 2016; Towne et al, 2011). For example, KCs release inactive proeIL-36α/β/γ, which can be cleaved by a neutrophilderived protease such as cathepsin G, elastase, or proteinase-3 to form active IL-36α/β/γ and promote an inflammatory response across diverse cell types. KCs also demonstrate an autocrine response through IL-36R, which, in conjunction with NET activation of TLR4, synergistically induces LCN2 expression, an antimicrobial with chemotaxis for neutrophils (Shao et al, 2019). Recent research indicates that this KC–neutrophil cooperation may depend on the activity of Unc-51–like autophagy activating kinase 1 in KCs (Qiu et al, 2021).

Eosinophils, basophils, and mast cells can respond to many of the discussed KC cytokines (eg, GM-CSF, CXCL8, IL-33) (Burrows et al, 1991; Iikura et al, 2004; Taniguchi et al, 2013; Tomonaga et al, 1986). KCs can preferentially recruit eosinophils through CCR3 ligands such as CCL26 (Gaspar et al, 2013), CCL5 (McCoy et al, 2017), and CCL11 (Valdebran et al, 2019). Mast cells are known to be recruited by LL-37 (Niyonsaba et al, 2002), whereas basophil chemotaxis has been linked to IL-1β and IL-3 (Iikura et al, 2004). Eosinophils and basophils cells are each known to infiltrate the skin and produce IL-31 (Kunsleben et al, 2015; Raap et al, 2017); therefore, they likely participate in the previously discussed barrier disruption and pruritus cycle of type 2 inflammation by signaling with KCs. Many of these KC–granulocyte networks have not been thoroughly investigated, and deciphering their regulation may reveal future therapeutic targets for inflammatory diseases.

KC–innate lymphoid cell interactions

Innate lymphoid cells (ILCs), broadly distinguished from adaptive lymphocytes by their lack of rearranged antigen receptors (Spits et al, 2013), are important mediators in skin inflammation and also communicate with KCs. In healthy skin, ILCs are kept inactive by KC cadherins, which inhibit type 2 cytokines (Salimi et al, 2013). Upon skin injury, KCs release IL-25, IL-33, and TSLP, which activate ILC2s and promote the expression of IL-5 and IL-13 (Leyva-Castillo et al, 2020; Monticelli et al, 2011). In the lung epithelium, IL-33 has been shown to cause dramatic ILC2 production of OX40L (Halim et al, 2018), and as significant producers of IL-33, KCs may similarly stimulate OX40L in the skin through ILC2s. This potential KC–ILC interaction could explain the observed increase in OX40L and OX40 levels in lesional skin (Ilves and Harvima, 2013). ILC2s have also been shown to respond to TSLP with type 2 cytokines even without a requirement for IL-33 or IL-25 (Kim et al, 2013). KC-activated ILC2s contribute to the development of skin hyperplasia in inflammatory skin diseases through IL-13 production (Leyva-Castillo et al, 2020).

KC–ILC crosstalk is critical to wound healing processes. ILC2s have been found to produce EGFR ligands EGF and amphiregulin, which each stimulate KC proliferation and aid in wound repair (Alkon et al, 2022; Salimi et al, 2013). KCs also release NOTCH1 when wounded, which stimulates TNF production in nearby cells (Li et al, 2016). Together, NOTCH1 and TNF drive recruitment and activation of ILC3s. These activated ILC3s are potent sources of IL-17F, which acts on KCs in the same manner as IL-17A. Concurrently, active ILC3s recruit macrophages and encourage KC proliferation. ILC3s can also respond to IL-1β by producing IL-17A and IL-22 (Kim et al, 2014), and as previously discussed, KCs are a major source of IL-1β. IL-17A and IL-22 together activate the ROS–NLRP3–caspase-1 pathway in KCs to release additional IL-1β (Cho et al, 2012). This ILC3-derived IL-22 is a key contributor to epidermal thickening in psoriatic plaques (Teunissen et al, 2014). Because ILCs are still a poorly understood category, but their relevance to skin immunity is clear, we expect further connections to KCs to be unveiled as the field develops.

KC–sensory neuron interactions

KCs were previously seen as merely a stable structure to support intraepidermal free nerve endings (IFNEs) of sensory neurons, but recent work has uncovered a more involved role in neuroimmune communications (Liu et al, 2016; Wilson et al, 2013). KCs are important to various types of sensation in the skin, with TRPV1 responding to painful heat (Pang et al, 2015) and PIEZO1 responding to innocuous or noxious touch (Mikesell et al, 2022). KCs have also been proven to be required in signaling to sensory neurons in response to innocuous cold and heat (Sadler et al, 2020), although the involved thermoreceptors remain to be identified. In response to these stimuli, KCs use a SNARE-dependent mechanism of exocytosis (Talagas et al, 2020) to release neurotransmitters such as adenosine triphosphate and activate C- and Aδ-fibers of neurons through P2X4 and other receptors (Moehring et al, 2018).

Considering the relatively recent recognition of a KC–neuron network, it is important to look at the neuroimmune consequences of this communication. Alarmins such as IL-33 (Liu et al, 2016; Taniguchi et al, 2013) and TSLP (Wilson et al, 2013) released by stressed KCs engage with IL-1RL1 (ST2) and TSLP receptor, respectively, and are one mechanism to activate neurons and cause pruritus. A recent study also demonstrates that KC β-defensin 3 can interact with sensory neurons and contribute to pruritus (Tseng and Hoon, 2022). In healthy skin, KCs limit nerve density by preventing the extension of IFNEs past the granular KC layer with tight junctions (Takahashi et al, 2019). In individuals suffering from chronic itch, impaired barrier function allows IFNEs to further expand, and sensory neurons experience increased exposure to pruritogens from KCs and immune cells. KCs can amplify this increase of local nerve density and subsequent pruritogenic signaling with their production of nerve GF (Groneberg et al, 2005), or they can counteract it with release of the nerve repulsion factor, semaphorin 3A (Kurschat et al, 2006; Yamaguchi et al, 2008). In pruritic conditions, this KC–neuroimmune signaling creates a vicious cycle because inevitable scratching leads to barrier disruption, alarmin release, augmented inflammation, and worsened pruritus (Oetjen et al, 2017).

KC–fibroblast interactions

Epithelial–mesenchymal interactions are fundamental to development, but recent efforts have revealed an important communication network between KCs and fibroblasts in immunity. One of the most important KC–fibroblast signaling pathways may be through IL-1α/β. Fibroblasts are significantly more responsive to IL-1 than other skin cells and are uniquely sensitive to levels as low as 10 fg/ml, even in the abundance of the receptor antagonist (Cordier-Dirikoc et al, 2022). Meanwhile, KCs are a primary source of IL-1 in the skin (Kupper et al, 1986a). Both IL-1α and IL-1β bind to fibroblast IL-1R1 to activate MAPK and NF-κB (Barchowsky et al, 2000), but they are released from KCs by distinct mechanisms. Although IL-1α is already present in high levels in the cytosol of KCs, ready to be released as an alarmin upon cell death (England et al, 2014), IL-1β must be produced and secreted through inflammasome activation in response to stimulation by PAMPs or cytokines (Fenini et al, 2018). Recently, it has been shown that KCs can utilize the inflammasome to secrete IL-1α in addition to IL-1β through an integrin α3β1–dependent mechanism (Archer et al, 2019; Zheng et al, 2019). Fibroblasts can respond to IL-1 with the release of potent proinflammatory cytokines, including CXCL1 and CXCL8 (Friedrich et al, 2021; Wei et al, 2019), and the ability of activated dermal fibroblasts to recruit neutrophils and CD8⁺ T cells in vitro has been demonstrated (Xu et al, 2022; Zheng et al, 2021). IL-1 signaling through NF-κB in fibroblasts is correlated with differentiation into adipocytes (Sun et al, 2023), and KC-derived IL-1 may contribute to reactive adipogenesis and its associated LL-37 production (Dokoshi et al, 2018; O’Neill et al, 2022).

KC–fibroblast communication is critical to maintaining healthy, nonfibrotic skin. In healing wounds, fibroblasts enhance matrix deposition and differentiate into myofibroblasts (fibrogenesis), but in excess, this process results in fibrotic diseases. Fibroblasts rely on KC signals to regulate their expression of type I collagen and matrix metalloproteinase 1, favoring either matrix deposition or matrix degradation, respectively (Zeng et al, 2020). The mediators and mechanisms of this phenomenon are still unclear because 1 recent study showed that KC TGF-β is required for enhanced matrix deposition (Dufour et al, 2020), whereas another demonstrated that KCs can promote fibrogenesis in a TGF-β-independent mechanism (McCoy et al, 2017). IL-1 has been reported as an inhibitor of TGF-β-driven matrix deposition, suggesting that the 2 KC-secreted factors are opposing regulators of fibrogenesis and adipogenesis (Shephard et al, 2004). IL-17A has also been implicated in fibrosis and scleroderma as an inhibitor of KC TGF-β production, but contradictory studies show both fibrotic and antifibrotic effects (Brembilla et al, 2013; Yang et al, 2014).

KCs, fibroblasts, and immune cells feature multilayered paracrine systems. Fibroblasts produce FGF7, which beyond being a potent stimulator of KC proliferation and migration triggers the expression of IL-19 (Sun et al, 2013). This KC IL-19 then induces fibroblast expression of additional FGF7. As was previously discussed, KC IL-19 drives immunocyte expression of Th2 cytokines, including IL-4, and KCs respond to IL-4 with additional IL-19 (Bao et al, 2014). As a result, fibroblast FGF7 production increases, KCs proliferate, and the positive feedback loop continues. Fibroblast FGF7 has also been found to initiate KC production of OSM, which in turn activates fibroblast migration (Canady et al, 2013). IL-36 is another important actor in KC–fibroblast signaling. Beyond triggering inflammatory response as previously described (Shao et al, 2019; Towne et al, 2011), KC IL-36γ mediates the alarmin function of LL-37 produced by KCs, fibroblasts, and other cells (Li et al, 2014). IL-36α production in KCs is triggered by IL-36γ released by neighboring cells, and together, IL-36α and IL-36γ induce fibroblast production of IL-6, GM-CSF, and G-CSF (Kurinna et al, 2016). Fibroblasts can also amplify the activation of KC pro–IL-36 by producing cathepsin S (Ma et al, 2023). Fibroblast responses to IL-36 signaling induce inflammation, recruitment of neutrophils, and proliferation of KCs. Fibroblasts also promote KC proliferation through the production of CXCL12 (Quan et al, 2015) and prostaglandin E2 (Arai et al, 2014).

The KC–fibroblast axis of communication is likely to involve contributions from many other factors. Fibroblast migration and angiogenesis have been shown to be enhanced by interaction with KC microvesicles (Huang et al, 2015). KC S100A8 and S100A9 trigger fibroblast activation, showing increased production of type I collagen and α-smooth muscle actin (Zhong et al, 2016). KC-derived CCL26 may be important to fibroblast migration and repair capacity (Gaspar et al, 2013). Fibroblasts are known to respond to TNF with induction of proinflammatory chemokines (Larsen et al, 1989), and KCs produce TNF in response to UVR (Köck et al, 1990). KCs have also been found to produce Wnt-3a (Reyes et al, 2019), which suppresses adipogenesis and immune activity in favor of fibrogenesis, β-catenin activation, and type VII collagen production (Sun et al, 2023). In response to androgens, fibroblasts produce FGF10, which alters KC differentiation by suppressing keratin 1 and keratin 10 (Kumtornrut et al, 2019). Recent ligand receptor analysis of single-cell RNA-sequencing data implicates a potential role for various GFs, ILs, and chemokines (eg, PDGF, VEGF, FGF2, IL-15, CXCL12, CCL19) in KC–fibroblast signaling (Ma et al, 2023). Recently, fibroblasts are coming to be understood not as a single-cell type but as a diverse family of dynamic cell populations that play distinct roles in skin biology (Mascharak et al, 2020).

Because KCs and fibroblasts are the primary components of the epidermis and dermis, their interactions would be a natural focus of research in skin immunity, but they have been woefully understudied. Recent work has provided substantial evidence that dermal fibroblast crosstalk plays a central role in coordinating immune processes (Cavagnero et al, 2024), but there are still many questions to be answered before we fully understand this essential network and its relevance to skin disease.

KC COMMUNICATIONS IN INFLAMMATION AND DISEASE

These many interwoven networks of communication predict the important roles of KCs in regulating skin immunity. Primary genetic disorders of KC function (ie, the various ichthyoses), epidermal responses to trauma or infection, and inflammatory skin diseases are all direct demonstrations of the essential roles of KCs in coordinating immunity. In the following sections, we discuss some major skin disorders that highlight important immune communications of KCs.

Atopic dermatitis

Atopic dermatitis (AD) is an inflammatory skin disease characterized by itchy and inflamed skin in up to 22% of children (Wootton et al, 2018). Progress in understanding AD has now identified that this disease is a manifestation of abnormal crosstalk between KCs, immunocytes, neurons, and microbes and may initiate from failure of KC barrier functions. The previously discussed cycle of pruritus, barrier disruption, and type 2 inflammation are central to AD. Proinflammatory Th2 cytokines such as IL-4, IL-13, and IL-31 cause KC release of TSLP, IL-33, CCL17, and CCL22. Pathobionts in the microbiome such as S aureus and S epidermidis overgrow on the skin surface owing to a suppression of KC defense functions by Th2 cytokines, and these bacteria then induce inflammation and itch (Deng et al, 2023). In AD, because epidermal levels of semaphorin 3A are repressed in skin lesions (Tominaga et al, 2008) and impaired barrier function allows the expansion of IFNEs past the granular layer (Takahashi et al, 2019), nerve density is elevated and pruritogenic signaling increases. Sensory neurons send pruritic signals to the brain and compel the patient to scratch their itchy lesions, and this action further disturbs local KCs, thus releasing more TSLP and IL-33. KC TSLP activates dendritic cells to drive T-cell differentiation into Th2 cells, resulting in more Th2 cytokines in the skin (He and Geha, 2010). Ruptured KCs also release IL-36γ, causing dendritic cells to produce neutrophil-attracting, CXC-motif chemokines (Fiskin et al, 2023).

Several signaling networks lead to impaired barrier function of KCs in AD. In lesional skin, T cells with overlapping type 2 and type 17 characteristics produce IL-13 and IL-22 (Fiskin et al, 2023), which each interfere with healthy epidermal differentiation (Boniface et al, 2005; Howell et al, 2008). In addition to its inflammatory effects, TSLP stimulates the formation of IL-33/phosphorylated STAT3 complexes in KC nuclei, which transcriptionally repress the tight junction–required claudin-1 as well as antimicrobials (Dai et al, 2022). IL-4 and IL-13 inhibit the expression of the structurally essential FLG in differentiating KCs (Howell et al, 2009). FLG expression has also been shown to be repressed by IL-17A and IL-22, (Jin et al, 2014). IL-4R activation in KCs has also been shown to reduce the expression of AMPs, selectively inhibiting antibiotic-producing, coagulase-negative Staphylococcus and allowing overgrowth of pathogenic S aureus (Nakatsuji et al, 2023). Together, these KC communications create an ineffective barrier that is vulnerable to infection and invokes continuous inflammation and pruritus.

Psoriasis

Another of the most prevalent skin conditions in the world is psoriasis, which is estimated to affect 2–4% of individuals (Parisi et al, 2013). Characterized by thick, scaly, erythematous patches of dry skin, psoriasis primarily involves inflammatory signaling between KCs, Th17 cells, ILC3s, neutrophils, dendritic cells, and fibroblasts. In psoriatic skin, IL-26 released by Th17 cells induces KC TGF-β, which stimulates maturation of Th17 cells and production of IL-17A and IL-22 (Fries et al, 2023). KCs further increase the production of IL-17A as well as TNF, with their release of IL-36γ (Ma et al, 2023), and they respond to IL-17A and IL-22 by producing high levels of AMPs, chemokines, and cytokines (Sønder et al, 2011; Su et al, 2019). Because S100 family proteins, β-defensins, and LL-37 are all upregulated, patients with psoriasis are less susceptible to pathogenic infection than those with AD, but these antimicrobials also contribute to inflammation. IL-17A–induced KC chemokines (eg, CXCL8, CCL20) promote chemotaxis of neutrophils, Th17 cells, and ILC3s. KCs also respond to IL-17A and IL-22 with production of IL-1β, TNF, and IL-6, which activate dendritic cell production of IL-23 to sustain Th17 activity (Lee et al, 2004). Meanwhile, KC IL-19 promotes fibroblast expression of FGF7, which alongside IL-17A, IL-22, and IL-6 drives rapid KC proliferation and creates the scaly plaques that are a hallmark of the disease (Teunissen et al, 2014).

Rosacea

The global impact of rosacea is estimated at up to 5.5% of the population. This unique skin disorder is currently characterized by a collection of diagnostic signs and symptoms that include central facial erythema, papules and pustules, telangiectasia, and phymas and may involve the eyes in addition to the skin (Gallo et al, 2018). Similar to many other inflammatory skin diseases, its pathology is closely related to aberrant KC function, but specific communications are still incompletely understood. It was discovered decades ago that rosacea lesions have elevated levels of leukocytes compared with healthy skin (Rufli and Büchner, 1984), and this immune infiltration has more recently been characterized as being dominated by Th1 cells; Th17 cells; macrophages; mast cells; and, in the case of papulopustular rosacea, neutrophils (Buhl et al, 2015).

An important mediator of rosacea is LL-37, which is uniquely activated at dramatic levels by KC KLK5 (Yamasaki et al, 2007). In a process mediated by TRPV4 and TLR2, this KC LL-37 recruits mast cells (Muto et al, 2014; Niyonsaba et al, 2002) and triggers their release of proteases, which subsequently induce KC KLK5 and amplify LL-37 activity (Mascarenhas et al, 2017). The LL-37 also recruits Th1 and Th17 cells (Agerberth et al, 2000), which provide IFN-γ and IL-17A, stimulating KC production of CXCL9, CXCL10, and other inflammatory cytokines (Buhl et al, 2015). A STAT1-mediated pathway seems to be involved in KC–macrophage–Th1 cell crosstalk specific to rosacea as well (Deng et al, 2021). An association has also been demonstrated with barrier dysfunction in facial skin (Darlenski et al, 2013) and microbiome dysbiosis involving Demodex folliculorum, S epidermidis, and C acnes (Holmes, 2013), both of which involve KC functions. Each of these aspects of rosacea will require additional research to reveal new approaches for treatments.

Acne vulgaris

Acne vulgaris is among the most common inflammatory skin diseases, impacting up to 85% of the adolescent population and decreasing in prevalence among adults (White, 1998). Its development is considered to require 4 main factors: increased sebum production, follicular hyperkeratinization, colonization of skin bacteria, and inflammation. Crosstalk between KCs, other skin cells, and commensal bacteria in the follicle is central to these phenomena. C acnes and S epidermidis each contains various TLR2 ligands (eg, peptidoglycans, lipoteichoic acid), which can activate KC NF-κB and MAPK pathways and subsequent production of proinflammatory cytokines (Takeuchi et al, 1999). The C acnes–derived heat shock protein GroEL has also been shown to activate KC TLR2 as well as CAMP Factor 1 (Graham et al, 2004; Lheure et al, 2016). Many health-associated and disease-associated strains of C acnes feature distinct secreted proteomes that contribute to the different KC inflammatory responses (Yu et al, 2016).

In acne vulgaris, proinflammatory KC cytokines (eg, IL-1α, TNF, GM-CSF, CXCL8, IL-6) drive the chemotaxis of neutrophils, macrophages, and other immunocytes into acne lesions (Kim et al, 2002; Norris and Cunliffe, 1988). C acnes has also been shown to stimulate KC IL-1β through the activation of the NLRP3 inflammasome (Li et al, 2019), and IL-1 levels in acne comedones reach proinflammatory levels (Ingham et al, 1992; Vowels et al, 1995). This involvement of IL-1 suggests a potential role for fibroblasts as IL-1–responsive cells in acne vulgaris, and this hypothesis is supported by the recent finding that fibroblasts undergo reactive adipogenesis and its associated release of LL-37 in acne lesions (O’Neill et al, 2022).

Communications unique to the biology of hair follicles also play a critical role in acne vulgaris. Biofilm produced by C acnes can accumulate inside the follicle and trap sebum, leading to the formation of a keratinaceous plug and comedone development (Burkhart and Burkhart, 2007). The hypoxic conditions of hair follicles, as we addressed previously, promote C acnes SCFAs to inhibit HDAC activity and enable the production of inflammatory cytokines in KCs through NF-κB and MAPK (Sanford et al, 2016). Follicular hyperkeratinization is promoted by fibroblast FGF10 suppressing KC keratin 1 and keratin 10 (Kumtornrut et al, 2019). A better understanding of KC responsiveness and signaling will be key to developing safer, more effective treatments for acne.

CONCLUSIONS

KCs enable protection from hazards of the environment, and they are armed with physical and biological tools to signal to a wide range of cell types within the skin. With our current understanding of the complex mechanisms that enable KCs to function normally, it has become increasingly clear that KC dysfunction is central to the pathogenesis of many diseases. The classifications of some cells as immunocytes is becoming less and less relevant with the recognition that cell types of the epithelia have many overlapping functions with classical immunocytes such as lymphocytes, macrophages, and neutrophils. Future progress in treatments for inflammatory skin diseases will depend heavily on uncovering the mechanisms through which KCs manage these networks of communication and how we can intervene when they malfunction.

Abbreviations:

AD

atopic dermatitis

AFA

antimicrobial fatty acid

AMP

antimicrobial peptide

EGF

epidermal GF

HDAC

histone deacetylase

IFNE

intraepidermal free nerve ending

ILC

innate lymphoid cell

IP₃

inositol triphosphate

KC

keratinocyte

NET

neutrophil extracellular trap

OSM

oncostatin M

PAMP

pathogen-associated molecular pattern

SCFA

short-chain fatty acid

STAT

signal transducer and activator of transcription

Th

T helper

TLR

toll-like receptor