The skin is the largest organ of the body and represents the interface between the body and the external environment. Besides its numerous functions (e.g. barrier function against biological, chemical and mechanical insults; thermo regulator, sense organ etc.), it also acts as the first line of defense of the immune system. The skin is colonized by cells of the innate (Langerhans cells, dermal dendritic cells, plasmacytoid dendritic cells, natural killer cells, macrophages and mast cells) and the adaptive (T cells) immune system [1-4]. Recently, we reported that epidermal Langerhans cell precursors are already present in human embryonic skin and that their development and attraction depends on various growth and differentiation factors present in developing skin, leading to the sequential acquisition of their phenotypic repertoire [5]. Furthermore, we found that the myeloid part of the skin-resident immune system shows a higher degree of maturity already at a prenatal stage than the lymphoid part (Schuster et al. [5]; Mature and immature leukocyte subsets enter human prenatal skin at different developmental stages; manuscript under revision). Despite these recent advances in the understanding of the development of the skin immune system, the T cell composition and colonization of the human neonatal and infant skin has not been characterized yet and needs investigation.

Skin T cells predominantly express αβ T cell receptors (TCR) rather than γδ but both fulfil important roles in immunosurveillance [6-9]. Approximately one million T cells are present per cm² of normal human adult skin, resulting in 20 billion T cells per individual, which is more than twice the total number of T cells in the blood [10]. The vast majority of skin-associated T cells in non-inflamed human adult skin resides in the dermis and expresses predominantly CD4, whereas only a minimal fraction of mostly CD8⁺ T cells is found in the epidermis [11]. CD3⁺CD45RO⁺ memory T cells are the dominant T cell fraction (>95%) allowing a rapid response in case of antigenic challenge, while CD45RA+ naive T cells comprise less than 5% of all CD3⁺ resident T cells in the skin [12]. The occurrence of T cells in developing human skin has been poorly investigated. While ultrastructural studies failed to detect T cells in developing prenatal skin [13], a recent study identified naive T cells in human fetal skin using immunohistochemistry [14]. The authors describe the presence of low numbers of CD3⁺CD45RA⁺ and CD3⁺CD45RO⁺ cells in fetal skin at 18 weeks of estimated gestational age (EGA) [14]. The number of CD3⁺CD45RO⁺ cells does not appreciably rise until the 30th EGA, whereas the density of CD3⁺CD45RA⁺ cells multiplies, leading to a distribution which shows three times more naive than memory T cells within the fetal skin [14]. Total T cell numbers in fetal skin are around 4% of that from adults [14]. Our own studies have shown that CD3⁺CD45⁺ cells can already be occasionally identified in the first trimester and appear to have intracellular CD3 expression (Schuster et al. [5] Mature and immature leukocyte subsets enter human prenatal skin at different developmental stages; manuscript under revision).

Given that the predominant T cells in human fetal skin are naive T cells, while those in adult skin are memory T cells it was our goal to evaluate when memory T cells start to colonize the skin after birth to better understand the development of the skin immune system.

Skin samples

Non-inflamed human foreskin samples from routine circumcisions were obtained from neonates (1 to 27 days; n=3), infants (28 days to 24 months; n=6) and adult patients with phimosis (18 to 49 years; n=3) and prepared as described previously [15]. Because of ethical limitations it was impossible to obtain healthy adult foreskins. Thus, we have used normal adult human breast skins from mammary reduction (n=5) as a staining control (specified in the respective Figure legends) but have not included these in our quantitative analysis described in Figure 4 because a different body location may influence the results. The study was approved by the local ethics committee and conducted in accordance with the declaration of Helsinki Principles. Parents/ participants gave their written informed consent. Skin samples were stored at 4°C in PBS until use and processed no longer than 6 hours after surgery.

Skin preparation

Skin specimens were embedded in O.C.T. (Tissue Tek, The Netherlands), snap-frozen in liquid nitrogen and stored at –80°C until further processing. Five μm sections were cut, mounted on capillary gap microscope slides, fixed in ice-cold acetone for 10 min and airdried.

Immunofluorescence

Staining was performed overnight at 4°C with conjugated anti- CD45RA mAb (clone T6D11; Miltenyi Biotech, Germany) and anti- CD3 mAb (clone UCHT1; ADG, Austria). Anti-CD45RO mAb (clone UCHL1; AbD Serotec, UK) staining was visualized with AlexFluor-546 F(ab’)₂ goat anti-mouse IgG mAb (Invitrogen, CA, USA). Slides were mounted with fluorescence mounting medium (Vector Laboratories Inc., CA, USA). Control samples were stained with appropriate isotype-matched Ab.

Imaging and quantification of cells in skin sections

Images were recorded with a Nikon eclipse 80i optical microscope. For the enumeration of naive and memory T cells, at least 10 randomly chosen areas per donor (n=3) were analyzed. To that end, the surface area was defined with Lucia general software (Nikon Corporation, Tokyo Japan) and the numbers of single and double positive cells/ cm2 were determined and placed in relation (given as percentage of all CD3⁺ cells).

Statistical analysis

The results are expressed as the mean ± S.E.M. Differences among the groups were assessed with one way analysis of variance followed by the Tukey test. All statistical analyses were conducted using the GraphPad Prism 5 software (GraphPad, San Diego, CA, USA).

Non-inflamed skin samples, were stained by doubleimmunofluorescence with the T cell specific marker CD3 and markers for naive (CD45RA), and memory (CD45RO) T cells. The distribution pattern and the density of naive and memory T cells in the epidermis and dermis were determined in neonatal, infant, and, for comparison, adult skin specimens. We found that the neonatal epidermis was devoid of CD3⁺CD45RA⁺ naive T cells which is in line with observations in the fetal epidermis [14] (Schuster et al; Mature and immature leukocyte subsets enter human prenatal skin at different developmental stages; manuscript under revision). Naive T cells were identified in the infant epidermis for the first time and were located within the basal and suprabasal keratinocyte layer (Figure 1). CD3⁺CD45RA⁺ naive T cells were present in neonatal, infant and adult dermis. They were distributed irregularly in the papillary and reticular dermis and were found often in close apposition to CD3⁺CD45RA^- cells, presumably memory T cells (Figure 2).

Figure 1: Naive T cells are present in infant and adult but not in neonatal epidermis. Cryosections from neonatal (3 days), infant (5 months) and adult (31 years) foreskin were labeled with anti-CD3-FITC and anti-CD45RA-PE mAbs. Arrows indicate CD3⁺CD45RA⁺ cells in the epidermis. Shown is one representative experiment out of 10. Scale bars: 5 μm. E = Epidermis, D = Dermis.

Figure 2: Naive T cells are present in newborn, infant and adult dermis. Cryosections from neonatal (5 days) and infant (5 months) foreskin and, as positive control, adult male breast skin (19 years) were included and stained with an anti-CD45RA-PE mAb and counterstained with an anti-CD3-FITC mAb. CD3⁺CD45RA⁺ cells (arrows) were found in the dermis of the indicated age groups. Inset shows colocalization of a CD3⁺CD45RA⁺ T cell with a CD3+CD4RAT cell. Shown is one representative experiment out of 10. Scale bars: 20 μm. E = Epidermis, D = Dermis.

Some rare CD3⁺CD45RO⁺ memory T cells were located in the neonatal epidermis and their number increased with age (Figure 3). In the dermis, memory T cells were found more frequently already in neonatal skin compared to CD45RA⁺ naive T cells (Figure 3, arrows). Furthermore, a similar T cell distribution pattern was observed between infant and adult skin but not neonatal skin. Consequently, we further determined the frequency of naive and memory T cells in neonatal and infant skin. CD3⁺CD45RA⁺ naive T cells were absent in the neonatal epidermis as described in Figure 1, but a small population of naive T cells (12.26%±7.67 of all epidermal T cells) was determined in infant epidermis (Figure 4A). While all T cells in neonatal epidermis displayed a memory T cell expression profile, not all T cells in infant epidermis (95.3%±8.14) expressed CD45RO (Figure 4A). In contrast to the T cell composition of the neonatal epidermis, the T cell population of the neonatal dermis was similar to infants. The frequencies of naive T cells (5.77%±4.2 in neonates vs. 5.95%±1.69 in infants) and memory T cells (97.67%±2.72 in neonates compared to 90.22%±4.89 in infants) in the dermis were already consistent to those described for adult skin (Figure 4B) [16].

Figure 3: Memory T cells are present in newborn, infant and adult skin. Frozen human foreskin samples from a neonate (3 days) an infant (6 months) and male breast skin (19 years) were labeled with an anti-CD45RO mAb, visualized with an AlexaFluor-546 goat anti-mouse IgG and counterstained with an anti- CD3-FITC mAb. CD3+CD45RO+ cells were extremely rare in newborn epidermis (not shown here), but their numbers increased with age. CD3⁺CD45RO⁺ cells are indicated with arrows (dermis) and arrowheads (epidermis). Shown is one representative experiment out of 10. Scale bars: 20 μm. E = Epidermis, D = Dermis.

Figure 4: Frequency of naive and memory T cells in neonatal and infant skin. Frozen human foreskin samples from neonates (n=3) and infants (n=3) were analyzed. Diagrams show the relation of naive (CD3⁺CD45RA⁺, white bars) and memory (CD3⁺CD45RO⁺, black bars) T cells in human skin. T cell numbers in the epidermis (A) and dermis (B) were evaluated, by counting naive and memory T cells in different donors on cryosections. ns = not significant.

In this study, we have shown that the composition of T cells in human neonatal skin is comparable in some aspects to adult as well as to fetal skin. Studies in human adult skin have highlighted that the great majority of T cells have a memory phenotype [7], while the majority of T cells in human fetal skin display a naive phenotype [14]. We found that the neonatal skin contains a large pool of memory T cells and that the epidermis is devoid of naive T cells, thus, reflecting the composition of the T cell population in adult and fetal skin, respectively.

CD3⁺ T cells are detectable in the human fetal circulation at about 15-16 weeks of gestation [17]. In line with this, naive T cells but almost no memory T cells (due to the scarcity of exogenous antigen and/or a lack of cytokines critical for the attraction of T cells) are present in fetal skin [5,14,18] (Schuster et al. [5] Mature and immature leukocyte subsets enter human prenatal skin at different developmental stages; manuscript under revision). It has to be unravelled whether fetal skin T cells represent passenger lymphocytes. It is also unclear whether skin homing fetal T cells are functional and may be of importance for the creation of tolerance [19]. Most recent observations on the occurrence of a fetal dermatitis with inflammatory infiltrates including CD3⁺ T cells upon microbial invasion of the amniotic fluid implies that circulating T cells can be attracted under certain conditions [20].

While in fetal skin only a few memory T cells can be identified [14], we observed a progressive colonization of memory T cells in the skin after birth when skin is challenged with a huge variety of pathogens. It is conceivable that memory T cells are recruited to the skin because antigen-presenting cells within the skin mount vigorous immune responses after birth once the immunosuppressive fetal environment is overcome as previously shown in mice [21] and suggested in humans [5,22]. Alternatively, the few already existing memory T cells expand rapidly but this remains to be investigated.

Our observation that naive T cells can only be observed in infant but not yet in neonatal epidermis is in favour with the idea of a change of the microenvironment now providing the gradient for an influx of naive T cells from the dermis into the epidermis.

We believe that these findings lead to a better understanding of the development of skin immune responses and will form the basis for future work.

This project was funded by an Austrian Science Foundation grant (P19474-B13 to A. E-B). We thank Drs. Gero Kramer (Department of Urology, Medical University of Vienna, Austria) and Anton Stift (Department of Surgery, Medical University of Vienna, Austria) for the organization of human skin samples. This work was performed in Vienna, Austria