Histological and microscopic examination of vernix caseosa found ovid or polygonal cells without nuclei or organelles while some had nuclear ghosts (Agorastos et al., 1988; Pickens et al., 2000). The cellular acid phosphatase activity was variable, from none to very high. The cytoplasm and cell membranes showed no alkaline phosphatase activity. The cells were thin (1–2 μm) and differed from regular to irregular with 5-6 sides with microvilli projections on the surface. The degree of keratinization varied, suggesting that they were from the outermost fetal stratum corneum and deeper levels, that is, less mature keratinocytes. Alternatively, the cells may originate during the transition from periderm to keratinized epidermis. There were few keratin filaments, lacking in orientation, and no evidence of desmosomes. In micrographs, the lipids between cells were generally amorphous, with occasional lamellae (Pickens et al., 2000). The cell diameter was ∼40 μm, larger than stratum corneum cells, perhaps due to absorption of water from amniotic fluid, and individual cellular hydration varied (Pickens et al., 2000).

Vernix caseosa is an amorphous, white waxy mixture of water-containing cells covered by a mixture of lipids (Pickens et al., 2000; Rissmann et al., 2006). It may appear on the fetal eyebrows at gestational week seventeen. Over time, it covers the fetal skin surface, advancing from head to toe and back to front (Visscher et al., 2005). Placental or hypothalamic corticotropic-releasing factors (CRF) may signal the pituitary gland to release adrenocorticotropic hormone (ACTH), causing the adrenal gland to release androgenic steroids (Zouboulis et al., 2003). They become active androgens and function within the sebaceous gland. Hair follicles have a local hypothalamic-pituitary-adrenal-like axis (Ito et al., 2005) that may be involved in vernix formation. Several vernix lipid types are also produced by the sebaceous glands, namely triglycerides, wax esters, and squalene (Nicolaides et al., 1972; Rissmann et al., 2006). Fetal cells possibly originate from the hair follicles (Kurokawa et al., 2009), mix with sebaceous lipids, extrude through the hair shaft, and continue to form and spread over the interfollicular epidermis during latter gestation (Hardman et al., 1998). Vernix cells may also come from the infundibular portion of sebaceous glands (Rissmann et al., 2006). Vernix films (i.e., spread on a porous substrate) in vitro are hydrophobic, due to the lipid cover on the hydrated cells (Youssef et al., 2001).

Vernix lipids cover the hydrated vernix cells to create a hydrophobic coating during latter gestation, thereby protecting the underlying fetal epidermis from exposure to amniotic fluid (Youssef et al., 2001). Vernix films are non-occlusive and permit water vapor transport through them (Tansirikongkol et al., 2007a). In utero, cornification of the fetal epidermis is incomplete thereby permitting a high water flux potential driven by osmotic gradients. Water gradients occur in skin homeostasis. Specifically, the SC exhibits a water gradient with higher hydration in the lower layers and decreased hydration at the skin surface (Warner et al., 1988; Caspers et al., 2003; Verdier-Sevrain and Bonte, 2007). The ability of the SC barrier to recover after the damage is due in part to the transepidermal water gradient and subsequently increased synthesis of DNA and lipids (Proksch et al., 1991; Denda et al., 1998a; Denda et al., 1998b; Fluhr et al., 1999). Vernix may serve as a semi-regulated barrier and/or physiological gradient for transepidermal water and nutrients in utero. This process, in turn, prompts epidermal cornification by increasing the synthesis of DNA and lipids.

As full-term gestation approaches, the mature fetal lungs secrete phospholipid surfactants that cause some of the vernix to detach from the skin surface (Narendran et al., 2000). This process causes the amniotic fluid to become cloudy. The infant swallows the amniotic fluid and, thereby, vernix provides nutrients to prepare the intestine for extra-utero feeding.

Vernix is composed of ∼80% water, associated with the cells, 10.3% protein, and 9.7% lipids (Pickens et al., 2000; Hoath et al., 2006). The non-lamellar lipid mixture that covers the flattened vernix cells comprises both non-polar, as the predominant fraction, and polar lipids, including fatty acids, ceramides, squalene, cholesterol, wax esters, and triglycerides (Rissmann et al., 2006). Epidermal barrier lipids, that is, cholesterol, fatty acids, and ceramides compose 10–30% of the vernix lipid fraction (Hoeger et al., 2002; Rissmann et al., 2006), with ceramide lipids comprising 4.9% of the total vernix lipid fraction (Rissmann et al., 2006). Vernix ceramide profiles (weight percent of total ceramides) compared to adult stratum corneum (SC) ceramides are in Figure 1A (Rissmann et al., 2006). Ceramide AH (AH contains α-hydroxy acids and sphingosines) was the most abundant in vernix, followed by ceramide NS (NS contains non-hydroxy fatty acids and sphingosines), ceramide AS/NH (AS contains α-hydroxy fatty acids and sphingosines and NH contains non-hydroxy fatty acids and 6-hydroxysphingosines) and ceramide EOS (EOS contains ester-linked fatty acids, ω-hydroxy fatty acids, and sphingosine). The distributions were similar while relative ceramide levels were higher in vernix compared to adult SC, except for ceramides AP (AP contains α-hydroxy fatty acids and phytosphingosine) and NP (NP contains non-hydroxy fatty acids and phytosphingosine) that were lower in vernix. The ceramide profiles in vernix, fetal SC (16–20 weeks GA), mid-gestational SC (23–25 weeks GA), infant SC (1–11 months), and child 1-6 years were compared, shown in Figure 1B (Hoeger et al., 2002) and the relative abundance of ceramides was similar to that of Rissmann (Rissmann et al., 2006). The vernix ceramide AH level was higher and vernix NP and NS levels were lower than those of infants aged 1–11 months (p < 0.05). Ceramide levels in vernix and premature infant stratum corneum at 23–25 gestational weeks were comparable, except for ceramide AP that was higher in vernix (p < 0.05). Collectively, these results reveal the uniqueness but also the similarities between vernix and SC, as well as the dynamic, variable nature of SC ceramide composition over time.

The fatty acid profile of vernix lipids includes branched-chain fatty acids (BCFA), a species that is not present in epidermal barrier lipids, as well as saturated, mono-unsaturated, and poly-unsaturated fatty acids (Table 1). Additionally, the fatty acid distribution in vernix lipids differed significantly by GA. Premature infants (29–36 wks GA) had significantly higher saturated and poly-unsaturated FAs and lower BCFA and mono-unsaturated FAs than full-term infants (≥37 wks GA) (Li et al., 2021).

Understanding the composition of vernix, which the infant ingests before birth, could have important pathophysiological consequences. For example, 20–30% of premature infants are affected by necrotizing enterocolitis, a potentially fatal intestinal inflammatory condition. In an animal model, dietary supplementation with 20% of vernix-type BCFAs reduced necrotizing enterocolitis by 50%, increased the intestinal microbiota diversity, and increased IL-10 three-fold versus the control whose diet was lacking in BCFA, indicating a protective role for this fatty acid in newborn intestines (Ran-Ressler et al., 2011). Additionally, data from an in vitro model demonstrated that induction of inflammation in intestinal cells with lipopolysaccharide (LPS) was associated with, a 20% reduction in cell viability was observed (Yan et al., 2018). However, when cells were treated with either vernix monoacylglycerides or vernix free fatty acids, the cell viability was restored. This study showed that the intestinal cells assimilated the BCFAs after treatment with vernix lipids. A putative role in prevention was suggested by experiments in which cells that were pretreated with vernix monoacylglycerides or vernix free fatty acids, followed by LPS exposure, expressed lower levels of IL-8 and NF-kB, suggesting that pretreatment with BCFA attenuated LPS-induced inflammation.

How epidermal lipids might mediate skin inflammation and immune function is unknown, but the mechanisms could include keratinocyte production of antimicrobial compounds, fibroblast migration, regulation of the rate of wound healing, and/or regulation of dendritic cells, for example, antigen uptake and activation of T cells (Kendall and Nicolaou, 2013). The impact of gestational age and gender has been studied with vernix samples from 156 infants in 3 GA categories, that is, 36–38 weeks, 39–40 weeks, and 41–42 weeks, revealing 54 lipid mediators (coefficient of variation <30% and in >70% of samples (Checa et al., 2015). Three classes of lipids were identified, namely, sphingolipids (n = 23), oxylipins (n = 43) and endocannabinoids (n = 14), and gender differences were noted (Checa et al., 2015). Sphingolipids are of interest for their potential role in skin barrier integrity and function, particularly to facilitate skin maturation and immunity in very premature infants (i.e., <28 weeks GA) who lack exposure to vernix. Within the sphingolipids, sphingomyelins increased with gestational age. The ceramide/sphingomyelin ratio (corrected for gender and maternal lifestyle) was significantly higher with increasing gestational age for chain lengths 12:0, 16:0, 18:0, 18:1, 24:0, and 24:1.

It is noteworthy that vernix from healthy full-term infants contained cytokines TNFα, IL8, IL1α, IL1β, IL6, MCP1, and IP10 (Narendran et al., 2010). The levels were substantially lower than in skin surface (stratum corneum) samples from premature infants, full-term infants, and adults (Narendran et al., 2010). This finding is consistent with the reduction of IL-8 and NF-kB in LPS-mediated intestinal cells (Yan et al., 2018). IL1α from the vernix covering may accelerate SC barrier maturation after birth (Jiang et al., 2009).

Qiao et al. investigated the effects of vernix lipids (n = 10 infants) on the expression of a critically important skin protein, filaggrin (FLG), and markers of inflammation in normal human epidermal keratinocytes in vitro (Qiao et al., 2019). Inflammation, evidenced by increased amounts of cytokines tumor necrosis factor alpha (TNFα) and thymic stromal lymphopoietin (TSLP) was provoked by exposure of keratinocytes to polyinosinic:polycytidylic acid (poly I:C), a synthetic double-stranded RNA. This resulted in a dose-dependent reduction in cell viability (Qiao et al., 2019). Exposure to vernix lipids attenuated TNFα and TSLP levels, further supporting the antiinflammatory potential of vernix. In this work, keratinocytes treated with Poly I:C decreased FLG expression while the addition of 25, 50, and 100 μg/ml of vernix lipids increased filaggrin (FLG) expression relative to cells that were not treated with Poly I:C. (Qiao et al., 2019). In contrast, FLG expression decreased in keratinocytes that were treated with poly I:C. This work may have broader relevance to newborn infants as FLG is a precursor of the Natural Moisturizing Factor. FLG mutations are implicated in atopic dermatitis. Further investigation of the effect of vernix on atopic dermatitis is warranted.

Analysis of vernix proteins with 2D gel electrophoresis identified 41 proteins, including 16 associated with innate immunity (Tollin et al., 2006). They were: UBB (ubiquitin), S100A8, S100A9, S100A7, LYZ, NGAL, H2AC11, H2BC1, RNASE7, SLPI, CAMP (LL-37), MUC7, BPIFA1, PSMB2, ARG1, and SOD1. Additionally, the first 12 genes (UBB to MUC7) demonstrated antimicrobial properties (Tollin et al., 2006). Vernix contains antimicrobial proteins lysozyme and lactoferrin, localized in “granules” that may facilitate “quick release” in the presence of infectious agents (Akinbi et al., 2004). Vernix decreased specific perinatal pathogens, namely group B Streptococcus, K pneumoniae, and L monocytogenes (Akinbi et al., 2004). Holm, et al., analyzed 34 individual vernix samples using liquid chromatography tandem mass spectrometry and identified 203 proteins (Holm et al., 2014). Their analysis with multivariate and classification methods revealed 25 functional classes. Hydrolases, proteases, and enzyme modulators encompassed 29, 22, and 22 proteins, respectively, with 11 proteins classified as immunity/defense and generally consistent with Tollin et al., (2006); Holm et al., (2014). The 34 vernix samples were from 16 infants who had developed atopic dermatitis by 2 years of age and 18 non-atopic healthy controls. A comparison of the proteomic data found significantly reduced levels of both UBC (polyubiquitin-C) and CALM5 (calmodulin-like protein 5) in vernix of infants who later developed atopic eczema versus vernix from infants who did not develop eczema, shown in Figure 2 (Holm et al., 2014). Furthermore, investigation to determine whether these biomarkers are early indicators of atopic disease is clearly warranted, given the increased incidence and morbidity associated with this condition.

Vernix has demonstrated multiple “protective” functions. Evidence of these actions includes the following. 1) Vernix was spread on a highly permeable fiber substrate to create films of known thicknesses in vitro. The vernix films impeded exogenous chymotrypsin transport and maintained the native enzyme activity that is necessary for epidermal development (Tansirikongkol et al., 2007b). 2) Normal adult skin was treated with native vernix and common skin creams petrolatum, Aquaphor and Eucerin, and an untreated control. Vernix-treated skin had a significantly higher peak water sorption value than all the cream treatments and the control, indicating that it binds exogenous water to the skin (Bautista et al., 2000). 3) In parallel cohorts of full-term infants, vernix was retained on the skin of one group and removed from the other group. The skin covered with vernix was significantly more hydrated, less erythematous, and had a lower surface pH than skin where the vernix was removed (Visscher et al., 2005). These differences were observed immediately after birth and 24 h later. 4) The SC from the vernix retained and vernix removed groups was sampled 24 h after birth and analyzed for the free amino acid (FAA) component of the natural moisturizing factor. Free amino acid levels were significantly higher for infants with vernix retained versus those with vernix removed where FFAs were extremely low or undetectable (Visscher et al., 2011b). The FFA appeared to originate from the vernix that was retained on the skin after birth. That is, native vernix contained FFAs. 5) Skin barrier damage was created by repeatedly tape stripping the SC in the hairless mouse model. The damaged skin, treated with vernix, demonstrated a significantly increased rate of SC barrier repair compared to untreated, damaged control skin (Oudshoorn et al., 2009a). In the same study, treatment of damaged skin with petrolatum also significantly increased the SC barrier repair rate versus the untreated control, but the skin was more erythematous and thickened compared to the vernix treated skin (Figure 3). 6) Wounds that were produced with 25 microns of laser energy (animal model) showed an increased rate of barrier recovery after 2 days of treatment with either vernix or a petrolatum-based cream compared to a wounded, untreated control (Visscher et al., 2011a).

In summation, the literature suggests that vernix protects the infant throughout fetal development and at birth, supporting its role in the innate immune function of the epidermis. Vernix appears during the last 10 weeks of gestation. Consequently, premature infants, particularly those <29 weeks GA at birth, lack exposure to significant amounts of vernix, raising these questions. What is the effect of exposure to vernix caseosa during gestation on the development of the innate immune system? How might the presumed positive effects of vernix be implemented to facilitate innate immune system development in very premature infants?

The dramatic transition from aqueous in utero conditions to a dry, gaseous environment at birth initiates changes in the skin that are required for the full-term infant to survive and thrive. Remarkably, the epidermal barrier is intact and functional, despite submersion in amniotic fluid. This is in marked contrast to skin maceration and SC lipid disruption with prolonged water exposure in older children and adults (Warner et al., 1999; Ogawa-Fuse et al., 2019). Within minutes after birth, full-term skin hydration changes and varies due to the presence of vernix, infant care practices, for example, exposure to radiant warming, and body site (Visscher et al., 1999; Visscher et al., 2005). Despite prolonged exposure to water during gestation, a rapid decrease in hydration occurs consistently within the first day, followed by an increase over the first 2 weeks and suggesting SC adaptation to the drier environment (Visscher et al., 1999; Visscher et al., 2000; Visscher et al., 2005; Fluhr et al., 2012). The low transepidermal water loss (TEWL) observed in full-term newborn skin indicated a well-functioning epidermal barrier (Hammarlund et al., 1979; Fluhr et al., 2012; Ludriksone et al., 2014). A rapid humidity decrease (hairless mice) lead to increased DNA synthesis, lower free amino acid levels, dry skin, and lower filaggrin immunoreactivity, due to decreased epidermal keratohyalin granules (Scott and Harding, 1986; Denda et al., 1998a; Visscher et al., 1999; Katagiri et al., 2003).

Full-term skin pH is nearly neutral at birth, decreases significantly by day 4 (Visscher et al., 2000) and then gradually continues to decrease over the next few months. An acidic skin pH is important in establishing the skin barrier as it promotes the effective functioning of enzymes required for SC development and integrity, that is, lipid metabolism, bilayer structure formation, ceramide synthesis, lipid bilayer formation, and desquamation (Rippke et al., 2002; Schmid-Wendtner and Korting, 2006). The skin pH reduction after birth is due to multiple mechanisms, including 1) filaggrin proteolysis to amino acids, pyrrolidone carboxylic acid, and urocanic acid; 2) secretory phospholipase hydrolysis to FFA; 3) acidification in the lower SC by a Na⁺H⁺ antiporter mechanism (NHE1); 4) melanin granule dispersion to release H+; and 5) cholesterol sulfate production of H⁺(Elias, 2017).

Full-term skin microbiota colonization begins at birth (Capone et al., 2011; Cuenca et al., 2013) and is populated by Lactobacillus, Propionibacterium, Streptococcus, and Staphylococcus, differing by body site at 6 weeks of life (Chu et al., 2017). Skin microbiota contributes to innate immunity by regulating antimicrobial peptides, including cathelicidins and β-defensins, and responding to inflammation via IL1α (Naik et al., 2012). S. epidermidis and hominis produce antimicrobial peptides that are noxious to S. aureus (Chen et al., 2018a; b). Skin bacteria and yeasts hydrolyze sebaceous gland triglycerides to glycerin and free fatty acids (Yosipovitch et al., 2000) that, in turn, have antimicrobial properties and contribute to skin surface acidity (Eyerich et al., 2018).

The epidermal barrier is under-developed in premature infants at birth, particularly those <29 weeks GA, putting them at risk for infection and increased permeability to both internal water loss and external deleterious agents (Evans and Rutter, 1986; Cartlidge, 2000; Rutter, 2000). The skin is easily injured or torn due to deficiencies in dermal structural proteins (Eichenfield and Hardaway, 1999). Although epidermal barrier development is rapid upon exposure to a dry environment at birth (Harpin and Rutter, 1983; Okah et al., 1995; Agren et al., 2006), very premature stratum corneum is not fully competent at 1 month of life, as indicated by a considerably higher transepidermal water loss (TEWL) compared to full-term infants (Agren et al., 1998).

The preterm skin surface pH decreased following birth but the rate was slower for infants weighing less than 1,000 g than for infants weighing more than 1,000 g. The decrease was faster during postnatal week 1 versus weeks 2–4 (Fox et al., 1998). The interaction of GA and postnatal age significantly influenced the rate of pH reduction (Green et al., 1968).

Infection is one cause of premature birth. In utero exposure to infectious agents and/or to antibiotics before birth is likely to impact the microbiome soon after birth. Infants <32 weeks GA demonstrated a decrease in bacterial richness during postnatal weeks 1 and 2, followed by an increase. However, the bacterial diversity was lower in premature infants than it was in full-term infants (Pammi et al., 2017). Firmicutes and Proteobacteria were the most abundant phyla in premature infants. The implications of the skin microbiome and maternal antibiotic use are areas that warrant further investigation.

Expression patterns for biomarkers of innate immunity by GA and over time warrant further comment. Soon after birth, increased expression of the antimicrobial proteins MPO (all infants) and LTF (LPT, FT) were noted in adults. Two to three months later, the expression of biomarkers S100A8, S100A9, S100A7, and S100A11, as well as MPO, LTF, and LYZ (PTs), had increased significantly versus adults (Figures 4A–C). S100A7, S100A8, and S100A9 prompt keratinocytes to produce cytokines and chemokines. In turn, cytokines can promote the production of S100A7, S100A8, and S100A9, and, thereby, respond to threats (stressors) to facilitate immunity against pathogens (Lee et al., 2012; Lesniak and Graczyk-Jarzynka, 2015). Increased expression of S100A7, S100A8, S100A9, and S100A12 occur in inflammatory skin conditions with epidermal barrier defects, that is, atopic dermatitis and psoriasis (Oestreicher et al., 2001; Sugiura et al., 2005; Benoit et al., 2006; Wolk et al., 2006). Two to three months after birth, several clinical measures of barrier status demonstrated higher TEWL (FT) versus adults, higher visual dryness (PT, FT) and lower SC cohesion (PT, FT) (Visscher et al., 2020). Consequently, the increased S100 protein expression levels may occur in response to multiple factors, including pathogen exposure and minor barrier injury.

The increased expression of filaggrin processing biomarkers FLG, FLG2, ASPRV1, CASP14, and TGM1 for all infants compared to adults 2–3 months after birth was associated with changes in the products of filaggrin proteolysis, that is, natural moisturizing factor (NMF), histidine, proline, urocanic acid, and pyrrolidone carboxylic acid (PCA) that were quantified with reverse phase high-performance liquid chromatography and tandem mass spectrometry (Wei et al., 2016). NMF, PCA, histidine, and proline amounts were significantly higher for every infant group versus adults 2–3 months later (p < 0.05). In contrast, after birth, NMF, PCA, histidine, and proline levels were lower for all three infant groups versus adult samples (p < 0.05). The NMF increase was associated with a skin surface pH reduction for all infant groups (data not shown), and this acidification of the epidermal barrier processes is necessary to provide immunity via the promotion of colonization with effective microbiota.

All three infant groups (PT, LPT, FT) had higher levels of 9 biomarkers versus adults shortly after birth (FLG, SERPINB3, SERPINB4, PI3, MPO, CALML5, CTSC, ALB, TF, Figures 4A–C), likely indicating their importance in newborn skin development and maturation. The functions and possible implications of these proteins are discussed below. Increased FLG was associated with reduced NMF in infants, potentially due to inhibition of FLG proteolysis at high humidity described in utero (Scott and Harding, 1986). SERPINB3 and SERPINB4 are protease (e.g., serine, cysteine) inhibitors, including proteases generated by infectious pathogens (Sun et al., 2017). PI3 inhibited keratinocyte desquamation prior to terminal differentiation (Nakane et al., 2002) and kallikrein proteolysis (McGovern et al., 2017) and is a component of the corneocyte envelope (Steinert and Marekov, 1995), functions that are important to the provision of the physical epidermal barrier. While identified in the SC, MPO was produced in immune cells, for example, neutrophils, and lymphocytes (Khan A. A. et al., 2014; Liu et al., 2015), was higher in infected wounds (Gabr and Alghadir, 2019), and was higher under conditions of oxidative stress (Khan et al., 2018) and inflammation (Voss et al., 2018). Higher CALML5 levels were implicated in terminal differentiation (Sun et al., 2015) and barrier repair (atopic dermatitis) (Donovan et al., 2013). CTSC prompted serine protease generation in immune cells (Meyer-Hoffert, 2009). High ALB was associated with reduced skin hydration, consistent with our observation of lower hydration/skin dryness, particularly in LPT and FT shortly after birth. TF was associated with inflammation (Mehul et al., 2017).

Over the time from birth until 2–3 months later, a greater number of biomarkers were differentially expressed for infants versus adults, in addition to those involving filaggrin processing. S100A7, S100A8, S100A9, and MPO were significantly higher for all infants versus adults (Figures 4A–C). Protease inhibitors/enzyme regulators, PI3, SERPINB3, and SERPINB4 remained higher for all infant groups versus adults, and SERPINB1, SERPINB9, SERPINB12 and CSTA became significantly higher over time (Figures 4A–C). SERPINB1, located in macrophages and the cytoplasm and granules of neutrophils (Majewski et al., 2016), functions as an antimicrobial in infections and can guard against apoptosis. SERBINB9 has been described to react with enzymes in bacteria, yeasts, and fungi (Meyer-Hoffert, 2009) and serves in host-defense against bacteria and viruses in the lung, another epithelial tissue (Askew and Silverman, 2008). SERPINB12 is ubiquitous in human tissue, including the epidermis and eccrine duct, and is thought to guard macrophages from their internal protease inhibitors as well as from exogenous sources (Niehaus et al., 2015).

Collectively, these unique protein expression profiles suggest that the processes and pathways regulated by these proteins continue to be important for the provision of epidermal immunity well after birth. Neonates respond to multiple system transitions at birth by 1) producing NMF and lowering skin pH, 2) mitigating desquamation by inhibiting specific protease activity, and 3) increasing the antimicrobial features of the epidermal barrier. Neonatal skin must adapt over time to provide a sufficient level of epidermal immunity and maturation.