The Human Penis Is a Genuine Immunological Effector Site

The human penis is a main portal of entry for numerous pathogens, and vaccines able to control resulting infections locally are highly desirable. However, in contrast to the gastrointestinal or vaginal mucosa, the penile immune system and mechanisms inducing a penile immune response remain elusive. In this descriptive study, using multiparametric flow cytometry and immunohistochemistry, we characterized mucosal immune cells such as B, T, and natural killer (NK) cells from the urethra, fossa, and glans of human adult penile tissues. We show that memory B lymphocytes and CD138+ plasma cells are detected in all penile compartments. CD4+ and CD8+ T lymphocytes reside in the epithelium and lamina propria of the penile regions and have mostly a resting memory phenotype. All penile regions contain CD56dim NK cells surface expressing the natural cytotoxicity receptor NKp44 and the antibody-dependent cell cytotoxicity receptor CD16. These cells are also able to spontaneously secrete pro- and anti-inflammatory cytokines, such as IL-17 and IL-22. Finally, CCR10 is the main homing receptor detected in these penile cells although, together with CCR3, CCR6, and CCR9, their expression level differs between penile compartments. Unlike antigen-presenting cells which type differ between penile regions as we reported earlier, urethral, fossa, and glans content in immune B, T, and NK cells is comparable. However, median values per each analysis suggest that the glans, containing higher number and more activated NK cells together with higher number of terminally differentiate effector CD8+ T cells, is a superior effector site than the urethra and the fossa. Thus, the human penis is an immunologically active tissue containing the cellular machinery required to induce and produce a specific and effective response against mucosal pathogens. It can therefore be considered as a classic mucosal effector site, a feature that must be taken into account for the elaboration of efficient strategies, including vaccines, against sexually transmitted infections.

The human penis is a main portal of entry for numerous pathogens, and vaccines able to control resulting infections locally are highly desirable. However, in contrast to the gastrointestinal or vaginal mucosa, the penile immune system and mechanisms inducing a penile immune response remain elusive. In this descriptive study, using multiparametric flow cytometry and immunohistochemistry, we characterized mucosal immune cells such as B, T, and natural killer (NK) cells from the urethra, fossa, and glans of human adult penile tissues. We show that memory B lymphocytes and CD138 + plasma cells are detected in all penile compartments. CD4 + and CD8 + T lymphocytes reside in the epithelium and lamina propria of the penile regions and have mostly a resting memory phenotype. All penile regions contain CD56 dim NK cells surface expressing the natural cytotoxicity receptor NKp44 and the antibody-dependent cell cytotoxicity receptor CD16. These cells are also able to spontaneously secrete pro-and anti-inflammatory cytokines, such as IL-17 and IL-22. Finally, CCR10 is the main homing receptor detected in these penile cells although, together with CCR3, CCR6, and CCR9, their expression level differs between penile compartments. Unlike antigen-presenting cells which type differ between penile regions as we reported earlier, urethral, fossa, and glans content in immune B, T, and NK cells is comparable. However, median values per each analysis suggest that the glans, containing higher number and more activated NK cells together with higher number of terminally differentiate effector CD8 + T cells, is a superior effector site than the urethra and the fossa. Thus, the human penis is an immunologically active tissue containing the cellular machinery required to induce and produce a specific and effective response against mucosal pathogens. It can therefore be considered as a classic mucosal effector site, a feature that must be taken into account for the elaboration of efficient strategies, including vaccines, against sexually transmitted infections.
To reduce or prevent these STIs, vaccine strategies targeting the penis are crucially needed. Accordingly, initial HIV-1 vaccine studies were able to induce HIV-1 specific mucosal antibodies, although non-neutralizing, in the male genital mucosa (11). Furthermore, exposed seronegative (ESN) men harbor high urethral concentrations of HIV-1-specific IgA induced by nonprotected insertive sexual intercourses with seropositive female partners (12). These studies indicate that the human male genitals, as in other species (13), are effector sites. However, the lack of progress in developing vaccines to stimulate local protection in the penis is mainly due to the lack of information on the penile immune system.
The human penis consists of four different regions: (i) the foreskin, a stratified keratinized epithelium, with a highly keratinized outer face and a less keratinized inner one facing the glans (8), (ii) the glans, a stratified keratinized epithelium; (iii) the fossa navicularis (referred to here as fossa), a stratified nonkeratinized epithelium, and (iv) the urethra, a pseudo-stratified non-keratinized epithelium (6,8). The penis susceptibility to STI depends largely on the intrinsic characteristics of the mucosal immune system of each of these regions.
Innate and adaptive immune responses contribute both to protection at mucosal surfaces (14). The mucosal innate immune system is the first line of defense against mucosal pathogens and comprises numerous components including epithelial barriers, antimicrobials peptides (15), pattern recognition receptors, such as toll-like receptors (TLRs) (16), and inflammatory immune cells, such as natural killer (NK) cells and neutrophils, which are mainly involved in apoptosis of infected cells and phagocytosis, respectively. Antigen-presenting cells that include macrophages, Langerhans cells (LCs) and dendritic cells (DCs), participate in innate immune responses, as well as the initiation of adaptive immune responses by presenting antigens to lymphocytes. Such adaptive immune responses, which take place in a second step following the innate immune responses, are pathogen specific and involve two arms, namely, the humoral response coordinated predominantly by B cells, with or without CD4 + T cells help, and the cellular response driven by cytotoxic T cells.
Penile mucosal immune cells and their interactions with STI have been little studied due to the difficulty in obtaining human tissues, whereas the foreskin immunity is better understood particularly in the context of HIV-1 infection. Hence, we showed that HIV-1 targets first LCs during sexual transmission of HIV-1 in non-circumcised men (7), providing an explanation at the cellular level to the reduction by >60% of HIV-1 acquisition in men provided by removal of the foreskin following circumcision. Circumcision also protects men efficiently against other STI including HPV and herpes simplex virus (HSV)-2 (17). In agreement with an only partial protection to STI resulting from circumcision, other penile regions are targeted by STIs. Indeed, HIV-1 also targets macrophages in the penile urethra as we reported (10). Other studies (5,6,18) have reported on the immune cell content of the penis using qualitative morphological analyses, although a detailed phenotype and the role of these mucosal immune cell populations were not assessed. To fill this gap, we provide here an in-depth characterization of B, T, and NK cells present in the different penile regions, namely, the urethra, fossa, and glans mucosae, a crucial prerequisite for the elaboration of efficient preventive and vaccinal strategies against STIs.

MaTerials anD MeThODs ethical statement
The study was performed according to local ethical regulations, following approval by the local ethical committee [Comité de Protection des Personnes (CPP) Île-de-France XI, approval no. 11016]. Written informed consent was provided by all study participants.

Tissues
Whole penile tissues were obtained from 39 individuals undergoing elective gender reassignment surgery at the Saint Louis Hospital in Paris, France (mean age 38 years, range 19-58 years). Hormonal treatment was stopped 2 months before surgery, and all patients lacked history of STIs during the 6 months before the surgery. Tissues were transported to the laboratory immediately after surgery in phosphate-buffered saline (PBS) supplemented with 20 μg/ml gentamicin (Gibco) and used within the next 2 h.

single cell suspensions
Urethra, fossa, and glans were first mechanically separated, each region cut into 8 mm × 8 mm pieces, and the underlying fat and muscle removed. To prevent denaturation or removal of cell surface molecules by dispase/trypsin enzymes that are routinely used to separate the epidermal and dermal compartments, cells were directly extracted from the tissues by collagenase type IV digestion (2 mg/ml, Sigma) in the presence of DNAse I (200 U/ml, Roche) for 2 h at 37°C (7,9). After enzyme inactivation with RPMI medium containing 10% fetal bovine serum (FBS, Gibco), tissue pieces were vortexed for 30 s. The resulting cell suspensions were filtered through 100 μm nylon cell strainers (Falcon) and centrifuged for 5 min at 450 g. , lymphocytes were gated based on their forward (FSC-A) and side (SSC-A) scatter (lymphocytes gate). After gating on CD45 + lymphocytes, penile immune cell populations were defined as CD3 − CD19 + B cells, CD3 + CD4 + T cells, CD3 + CD8 + T cells, CD3 − CD56 + natural killer cells, and CD3 + CD56 + NKT cells. Each population was then studied for their phenotypes, activation status, function as well as for the presence of potential homing receptors according to the strategy indicated in the figure. Data were acquired with an LSRII cytometer (Becton-Dickinson, Cochin CYBIO platform) and analyzed following the gating strategy shown in Figure 1 with Kaluza Software (Beckman-Coulter) as indicated before.
immunohistochemistry Urethra, fossa, and glans tissue pieces were fixed with PBS-4% paraformaldehyde, embedded in paraffin and serially sectioned at 4 µm. Immunohistochemistry was performed as we previously described (7, 10) using appropriate antibodies (

Memory B cells are Present in all Penile regions
B cells are an important population of adaptive immunity driving the humoral response by producing antibodies after antigenic stimulation. Characterization of their phenotypes is a fundamental prerequisite for the elaboration of efficient vaccine strategies against STIs. Therefore, the presence of B lymphocytes, defined by the expression of the pan leukocyte marker CD45 and CD19, but lack of CD3, was determined in each penile compartment by multiparametric flow cytometry (Figures 2A,B) Next, co-expression of CD21 and CD27 was used to characterize the four B cell subsets (19), namely, CD21 + CD27 − naïve (BN), CD21 + CD27 + resting memory (BRM), CD21 − CD27 + activated memory (BAM), and CD21 − CD27 − tissue-like memory (BTLM) B cells (Figures 2C,D).
The major B cell subset was BTLM in all three penile compartments. Although the frequency of BTLM was similar between Penile B cells express the iga Fc receptor-like 4 (Fcrl4) and the igg Fc receptor-like 5 (Fcrl5) The FcRL4 and FcRL5 were recently described as receptors for IgA and IgG, respectively (20). FcRL4 is expressed on BTLM cells that lack the classical CD27 memory B cell marker, and FcRL4 + B cells localize at the subepithelial and marginal zones of mucosal lymphoid tissues. FcRL4 + B cells express switched immunoglobulins that have undergone variable region somatic mutations and secrete high levels of IgG but also IgA in response to stimulation with T cell cytokines, but not following B-cell receptor (BCR) cross-linking (21). FcRL5 is also expressed specifically on CD27 negative B cells, and FcRL5 cross-linking induces B cell proliferation and surface IgA and IgG expression, suggesting that FcRL5 participates in the expansion and development of antigen-primed B cells in physiological conditions (22). Thus, we determined the surface expression of these two receptors on penile B cells by flow cytometry. The frequency of FcRL4 + B cells was higher in the urethra (20.3% [16.8-23.9]) compared with the fossa (10.6%

The Penis contains Plasma cells (Pcs)
The capacity of penile B cells to secrete antibodies was inferred from the presence of CD38 + CD138 + mucosal PCs among B cell populations in penile tissues. PCs were present in all penile regions representing less than 5% of total B cells in urethra (5% [4.5-5.6]), fossa (4.2% [3.3-5.0]), and glans (1.8% [0.9-2.6]) ( Figure 2F). The frequency of PCs was higher in the urethra and fossa compared with glans.
Staining of the tissues for CD138/syndecan-1 expression followed by immunohistochemistry ( Figure 2G) indicated that all penile regions contained CD138 + PCs in the lamina propria. In line with the flow cytometry results, the densities of PCs in the urethra and the fossa were greater than in the glans. The expression of syndecan-1 by keratinocytes precluded the specific detection of PCs in the penile epithelia using this marker.
(g) Immunohistochemistry analysis of CD138 + PCs (brown), by comparison with isotype control (inset), in urethra, fossa, and glans [representative of n = 5 different donors for each tissue; mean age 41 years ]. Cells were stained with an anti-CD138 antibody or a rabbit IgG isotype control and revealed using diaminobenzidine peroxidase substrate (brown staining). Scale bar = 50 μm. All box-and-whisker plots represent minimum-to-maximum values, and each point corresponds to one donor. Statistical analyses were performed first by the Kruskal-Wallis test; pairwise comparisons were performed by the Mann-Whitney U-test. *p < 0.05, **p < 0.01, and ***p < 0.001.  3B,D). Tissue distribution of CD4 + and CD8 + T cells was also evaluated by immunohistochemistry following co-labeling with anti-CD3 (blue staining) and anti-CD8 (brown staining) antibodies. Hence, CD8 + T cells co-express CD3 and CD8 (blue/ brown staining), whereas CD4 + T cells expressed CD3 but not CD8 (blue staining only) ( Figure 3E). Matched isotypes served as negative control (data not shown). In the urethra, fossa, and glans, both CD4 + and CD8 + T cells localized in the lamina propria, close to the mucosal epithelium, whereas the majority of cells within the epithelial compartment were CD8 + T cells. However, it cannot be excluded that some of the cells detected by this technique also express T cell markers on their surface such as NKT cells. Next, T cell maturation was assessed using CD45RA and CCR7 to distinguish four T cell subsets (23), namely, CD45RA + CCR7 + naive (TN), CD45RA − CCR7 + central memory (TCM), CD45RA − CCR7 − effector memory (TEM), and CD45RA + CCR7 − effector (TE) T cells. CD4 + T cells (Figures 3F,G Intermediate memory T cells (TIM) were next quantified based on CD27 expression in the CCR7 negative memory T cells population ( Figure 3H). TIM represented around 5% of the memory population of both CD4 + and CD8 + T cells ( Figure 3I) Last, T cell activation, defined by co-expression of HLA-DR and CD38, was assessed ( Figure 4A). In the urethra and fossa, CD4 + T cells exhibited a resting phenotype, as activated cells coexpressing CD38 and HLA-DR represented only 8.2%

resident Memory T cells Form a Minor T cell subset in all Penile regions
Memory T cells, particularly CD8 + cells that can remain in tissues without recirculation through the blood (24), are referred to as resident memory cells, and can be characterized by the expression of CD69, L-selectin (CD62L) and CD103 (αEβ7 integrin). Both CD4 + and CD8 + T cells poorly expressed CD69 (Figures 4C,D) and CD62L (Figures 4E,F) 6-6.8] in the urethra, fossa, and glans, respectively) ( Figures 4G,H), although their distribution was very heterogeneous between individuals.
Overall, CD4 + and CD8 + T cells are equally present in all penile tissue regions and harbor a resting phenotype. Whereas CD4 + T cells are mostly (>90%) EM cells, CD8 + T cell subsets include mainly EM cells but also effector cells, especially in the glans. Furthermore, the higher expression of CD103 by CD8 + relative to CD4 + T cells translates into a higher epithelial CD8 + T cells distribution than that of CD4 + T cells, although both subsets are equally present in the lamina propria.    (Figures 5A,B, left panel). However, TNF-α (Figures 5B,C middle panels) and IL-2 ( Figures 5B,C, right panels) secreting cells were low in these tissues (less than 1% of CD4 + and CD8 + cell populations, representative dot plots for CD4 + cells shown on Figure 5A).
The secretion of Th2-like cytokines by CD4 + cells was also evaluated (Figures 5D,E). Th17 cells correspond to a pro-inflammatory subset of CD4 + T cells that secrete IL-17 after stimulation. Th17 cells provide a protective inflammatory response toward pathogens in mucosal tissues, such as skin, gut, and lung. IL-17 mediates the recruitment of neutrophils and macrophages to infected tissues, thereby acting in host defense against extracellular pathogens (26). Although CD3 + CD4 + cells from all penile tissues secreted IL-17 (Figures 5F,G, left panel), IL-17 production in the glans appeared to be slightly higher (2.0% [1.7-2.3]), compared with the urethra and fossa (1% [0.6-1.3] and 0.9% [0.7-1.2], respectively). Finally, we evaluated the secretion by penile CD4 + cells of IL-22, a cytokine that modulates the expression of many genes encoding proteins involved in tissue protection, survival, differentiation, and remodeling, possibly exerting pro-inflammatory functions (26) (Figures 5F,G,

Penile nK cells have an activated Profile and the Machinery to Mediate antibody-Dependent cell cytotoxicity (aDcc)
Natural killer cells, phenotypically defined as CD3 − CD56 + lymphocytes, are part of the innate immune system acting as first line of defense of the organism due to their capacity of destroying infected cells without prior activation/stimulation. The intensity of CD56 expression distinguishes two subpopulations of NK cells (27), namely, the CD56 dim cytotoxic and the CD56 bright immunoregulatory NK cells.
In line with this CD56 dim NK cell profile, CD56 + cells were detected at low, but specific, frequency in the lamina propria of the urethra after comparative immunochemical detection with an anti-CD56 and isotype control-matched antibody (Figure 6D, right and left panels, respectively). However, CD56 + cells appeared absent from the fossa and the glans using this technique (data not shown), most likely because CD56 expression was too low for visualization by immunohistochemistry.
The main activating receptors of penile NK cells, namely, the members of the natural cytotoxicity receptors (NCR) family NKp30, NKp44, and NKp46 (27) In addition to T cells, innate immune cells including NK cells also secrete the inflammatory cytokines IL-17 and IL-22 (28). We thus evaluated whether IL-17 was spontaneously produced by penile NK cells. Only urethral NK cells secreted IL-17 (4.5% Altogether, all penile tissue regions are equally populated by NKp44 + -activated NK cells residing in the lamina propria that are equipped to trigger ADCC via the CD16 receptor, and are able to secrete spontaneously pro-inflammatory cytokines such as IL-17 and IL-22.

homing receptors expression by Penile immune cells
We first evaluated CCR5 and CXCR4 expression on T cells (Figures 7A,B), two chemokine receptors that additionally permit HIV-1 infection. CCR5 + CD8 + T cells were more numerous than      Next, we assessed the surface expression of homing receptors, namely, CCR3, CCR6, CCR9, and CCR10. CCR6 and CCR10 were expressed at a significantly higher level than CCR3 and CCR9 on B, CD4 + , and CD8 + T cells in all penile regions (Figures 7C-E). Furthermore, expression of CCR6 and CCR10 on NK cells (Figure 7F) in the different penile regions was higher compared with B, CD4 + , and CD8 + T cells. Finally, whereas less than 5% of B and T cells expressed CCR3 and CCR9 in all penile regions (Figures 7C-E), NK cells expressed higher levels of both CCR3 in the urethra, and CCR9 in all penile compartments ( Figure 7F).
Thus, CCR10 and CCR6, and to a lesser extend CCR3 and CCR9, might participate in the differential homing of these mucosal immune cells into penile tissues.

Urethra, Fossa, and glans have Different cytokine environments
Cell homing is driven by chemokine receptor expression, and their capacity to bind to the appropriate chemokines/cytokines expressed specifically in each target mucosal tissue. We therefore evaluated which chemokines/cytokines were constitutively expressed in the various penile mucosal regions. Tissue pieces derived from the three penile regions were lysed, and the supernatants were analyzed for the presence of MCP-1/CCL2, MCP-4/ CCL13, Eotaxin/CCL11 (ligands of CCR3), MIP-3α/CCL20 (one ligand of CCR6), CCL25 (the ligand of CCR9), CTACK/ CCL27 and CCL28 (the ligands of CCR10), RANTES/CCL5 (one ligand of CCR5), TRAIL/TNFS10 (a member of the tumor necrosis factor family of ligands capable of initiating apoptosis through engagement of cell death receptors), and the inflammatory cytokines IL-4 and IL-13 using the luminex and quantikine technologies (Figure 8).
Altogether, the tissue cytokine milieu varies with the penile tissue regions, with the urethra appearing to be the richest one.

DiscUssiOn
Numerous pathogens including viruses and bacteria use the human penis as a portal of entry into the body and a vaccine locally active against such pathogens is key to control the spread Penile Mucosae Are Immunologically Active Frontiers in Immunology | www.frontiersin.org December 2017 | Volume 8 | Article 1732 of STIs. Establishment of immune memory against pathogens occurs naturally after infection and permits to establish a faster specific immune response following a second contact with the pathogen. Unlike that of the other penile regions, the mucosal immune system of the foreskin has been intensively studied and characterized. Thus, the lamina propria and the epithelium of the human foreskin selectively contain macrophages and CD1a + HLA-DR + LCs, respectively (6,7,9). Furthermore, the foreskin also harbors dendritic cells and numerous EM T cells, including CD4 + Th17 cells (29). Finally, the foreskin epithelium is also protect by Mucin proteins, such as MUC1 and MUC4, and expresses TLRs, particularly TLR5 (16), TLR4 and TLR3 (30), thus endowing this tissue with an important anti-infectious potential.
An effective vaccine must ideally induce such pathogenspecific immune memory by inducing local specific humoral and cellular responses, in turn preventing mucosal pathogen entry and establishment of infection. Therefore, the first requirement for elaborating a protective vaccine in the penis is to evaluate whether the penile tissue possesses the appropriate set of mucosal immune cells able to target each penile compartment, and to develop a protective local immune response.
Mucosal B cells, and particularly the most abundant memory B cells we found in all penile regions, are fundamental for the development of a long-lived vaccine humoral response. Longlived humoral memory comprises two major components, namely, memory B cells that search for their specific antigen in the periphery allowing for later affinity maturation, proliferation, and differentiation (31), and the PCs that locally produce specific immunoglobulins.
Penile memory B cells comprise CD21 − CD27 − FCRL4 + BTLM, most abundant in all penile regions and reaching even 75% of glans B cells. BTLM that lack the memory marker CD27 are an exhausted B cells subset (32), which also express numerous inhi bitory receptors containing immunoreceptor tyrosine-based inhibitory motifs as well as FcRL4 (32). The high expression of FcRL4 on urethral B cells we report may control BCR-dependent proliferation (19,31,33). Hence in the context of HIV-1 infection, FcRL4 in BTLM cells negatively controls the production of cytokines by BTLM cells, but also limits the secretion rate of HIV-1 specific antibodies.
Penile memory B cells include also the CD21 − CD27 + BAM that are long-lived, express the BCR, and have already undergone the processes of class switching and antibody affinity maturation. Upon pathogen infection, these BAM transform into CD27 + CD38 + CD138 + PCs that will produce large amounts of antibodies, particularly IgG or IgA. BAM represent 30% of total B lymphocytes in the urethra and fossa but only 15% in the glans. CD38 + CD138 + PCs were found in all penile compartments although in limited quantity, residing in the tissue lamina propria. Although we did not determine the immunoglobulin isotype secreted by penile PCs, these cells in the urethra and fossa are known to predominantly secrete polymeric IgA and IgM that translocate through urethral epithelium in a polymeric immunoglobulin receptor-dependent manner, thus releasing secretory SIgA or SIgM in the urethral canal (5,6). In addition, mucosal PCs secreting IgG locally are also present in various mucosa (34) including the foreskin (35) where IgGs translocate across the epithelium in a FcRn-dependent manner (36). Thus BAM cells and in turn PCs will most likely respond locally to a mucosal vaccine, allowing a rapid and specific response against the targeted pathogen.
How B cells home specifically into the various penile compartments remains unclear. Among the chemokine receptors expressed on penile B cells in the urethra, fossa, and glans, we now show that CCR6 and CCR10 likely drive penile homing, although CCR6 usually mediates mucosal homing to inflamed tissues including lung and gut (37). Especially in the urethra, CCR6 might promote homing as CCR6 chemotaxes toward CCL20/ MIP-3α, a chemokine we found enriched in the urethral compartment. Furthermore CCL28, one ligand of CCR10 detected in all penile tissues, might also participate in penile homing of B cells, as in the female genital tract (38).
In our study, CD3 + T cells represent the major immune cell population of all the penile regions. Although CD8 + T cellmediated immunity is crucial in antiviral defense, CD4 + T cells remain the determinant T cell subpopulation mediating a wide variety of actions on other immune actor cells, either directly by cell-cell contact or indirectly via chemokine/cytokine secretion. Hence, CD4 + T cells represent half of all T cells in all penile regions, thus contrasting with the female genital tract (39). Of note, using morphological studies, other reported that CD8 + predominate over CD4 + T cells in the male genital tract (6), although this apparent contradiction most likely resides in the differences in techniques used, namely, immunohistochemistry (6) vs the more quantitative flow cytometry approach we used here.
Similar to other genital tissues, such as the foreskin (29), cervix and endometrium (40), and other mucosa (41), we show here that resting EM cells are the dominant T cell subset in all penile compartments. Within the CD4 + T population, EM cells are long-lived cells derived from effector cells, which migrate into infected tissues to eliminate. Upon reinfection, CD4 + TEM secrete different effector cytokines and chemokines resulting in a faster and more efficient secondary response. All penile tissues contained cells able to secrete pro-and antiinflammatory cytokines that regulate the immune response and the recruitment of immune cells to sites of pathogen infection (25). Their proportions in the different penile mucosa correspond to that described in the foreskin (29) or the sigmoid colon (42,43), but differ from that of the female genital tract mucosa such as the endometrium (44). As penile T cells are resting cells, one might suggest that upon activation, as can occur during infection, for example, by HIV, these T cell subsets are induced to secrete their characteristic cytokines.
Of note, the amount of effector T cells among the CD8 + T cell population is higher than that of CD4 + T cell population, especially in the glans. Unlike their CD4 + counterparts, CD8 + effector cells correspond to a terminally differentiated cell population and are derived from TEM (45). Thus, the increased presence of these effector cells in the lamina propria of the different penile tissues reflects an intense immune response, but also a continuous immunological monitoring at these sites.  Penile CD8 + T cells are essentially memory cells that express the mucosal αE/β7 integrin (CD103) which could explain their epithelial localization throughout the penile regions. Accordingly, this homing receptor retains T cells in genital epithelia but also in other mucosa, such as the gut, by interacting with E-cadherin expressed by epithelial cells (46). Such CD103 + T cells are most likely the non-recirculating resident memory T cell subset known to remain in tissues and prone to initiate a rapid immune response after resensitization (24).
Conversely, the recirculating T cell subset looses expression of CD103 and also the early activation marker CD69, thus cells devoid of these two markers might represent penile CD4 + T cells. Indeed, CD3 + CD4 + CD103 − CD69 − cells are present in all penile regions studied here, similar to most non-lymphoid tissues (40). This particular CD103 − CD69 − double negative phenotype associated with the expression of the HIV-1 co-receptor CCR5 on penile CD4 + T cells might explain the dissemination of HIV-1 in the host following penile infection of resident urethral macrophages (10), and potential subsequent transfer to CD4 + T cells. Although CD103 participates in the mucosal homing and residence of T cells in the female genital tract (46), the role of other receptors implicated in T cells homing into male or female genital tissues remains elusive, although T cells homing in other mucosa is better documented. For example, CCR9 and the α4β7 integrin participate in homing to the gut (47), whereas CCR3 and CCR10 are involved in T cell-mediated skin inflammation (48) and in CD4 + T-cell infiltration in the nasal mucosa (49). We suggest here that in addition to CCR10 and CCR6, CCR5 that is also expressed by penile T cells might participate in penile tissue homing due to the high concentration, especially in the urethra and fossa, of penile RANTES/CCL5, the natural ligand of CCR5.
Innate immune cells are also present in penile tissues including macrophages, as we have shown earlier (10), and NK cells  (50). Throughout the penis, NK cells are essentially CD56 dim and harbor the Fc-γ Receptor CD16, associated with the "cytotoxic" NK cell subset described in the blood. Such CD56 dim NK cells proliferate little and produce low level of cytokines, at least in vitro (27). Accordingly, CD16 is the principal Fc-γ receptor of glans NK cells (51). Furthermore, penile NK cells strongly express all the members of the NCR family (27), namely, NKp30, NKp44, and NKp46 as we show herein. NKp30 and NKp46 are constitutively expressed on all NK cells of healthy individuals, whereas NKp44 is expressed only after activation (23,27). Penile NK cells are phenotypically similar to lung NK cells that are also CD56 dim CD16 + and mostly NKp46 + (52), thus corresponding to blood NK cells. By contrast, penile NK cells differ from CD56brightCD16+/-NK cells gut (53) and female reproductive tract (FRT) (54) that are comparable to immature NK cells found in secondary lymphoid tissues and known to produce higher amounts of inflammatory cytokines and harbor low cytotoxicity potential. NKp44 + and NKp46 + expression indicates that penile NK contains the two subsets of NK cells identified at the mucosal level, namely, the NKp44 + NK cells able to secrete IL-17 and IL-22, and the NKp46 + NK cells producing IFN-γ (28,55). IL-17 is a pro-inflammatory cytokine allowing for the accumulation of immune cells on the mucous surface, but also for the secretion of antimicrobial peptides by epithelial cells (26), whereas IL-22 confers protection to mucosal surfaces against bacterial and fungal infection, and promotes inflammation and the recruitment of immune cells at these sites (43). In addition, one of the other roles of IL-22 is to stimulate the secretion of the anti-inflammatory cytokine IL-10, by the epithelial cell barrier, allowing the preservation of its integrity (56). The secretion of these cytokines depends on IL-23 that is mainly produced by DCs, macrophages, and NK cells themselves following stimulation by lipopolysaccharides and the activation of the TLR4 as part of the innate immunity.
Thereby, in addition to their immune surveillance and putative cytotoxic functions, penile NK cells are therefore involved in the inflammatory reactions, thus allowing the establishment of the immune response orchestrated by B and T cells previously described.
Overall, as schematized in Figure 9, we show here that the human penis resembles other mucosal immunological tissues. Concerning T cell subsets, their relative proportions ressemble that of the foreskin (29) and ectocervix (57) but also non-genital tissues (41). B cells relative proportion is also similar to that of other immunologically active mucosa such as the FRT (57), but clearly different from that of the GIT (58) where B cells represent one quarter of the total immune cells. Furthermore, if penile NK cells are phenotypically similar to those found in the lungs, they approximate in proportion to those found in the cervix but are much less abundant than in the endometrium (57). Finally, taken the cytokine/chemokine environment together with chemokine receptors we studied here, CCR10 the best expressed one together with CCR6, CCR3, and CCR9 might participate in the homing of penile immune cells. Differential chemokine receptor expression in each penile compartment suggests that each specific combination of these receptors contributes to specific homing to each penile region. Of note, the frequency of CCR6 and CCR10 expression is always greater than that of the two other receptors, in particular for NK cells.
In conclusion, the human male urethra and the fossa exhibit an immunological composition close to those of other mucosal effector sites, as the foreskin (29) or the FRT (57). Strikingly, the glans shows effector features superior to that of the urethra and the fossa, due to the presence of more activated NK cells and the presence of more terminally differentiate effector CD8 + T cells. Thus, one might suggest that the glans is a site more sensitive to an infection such as HIV, whereas the two other more inner penile compartments, namely, fossa and urethra, allow for protection due of their specific immune composition. Nevertheless, urethra, fossa, and glans contain all the components necessary to host an innate immune response with the presence of activated NK cells, as well as an effector adaptive immune response due to the presence of PCs, memory B, and EM T cells. However, our results suggest that the induction of B and T cell responses most likely does not occur in these tissues but rather in draining lymph node, since lymphoid follicles are absent in the penis (6).
Altogether from the present study, the human penis emerges as an efficient mucosal effector site, which contains cell subsets required to induce and generate specific and effective immune responses against mucosal pathogens. This study provides an invaluable source of information on the penile immune system that might contribute to elaborating vaccine strategies against STIs targeting the male genital tract, such as HIV-1.

eThics sTaTeMenT
The study was performed according to local ethical regulations, following approval by the local ethical committee [Comité de Protection des Personnes (CPP) Île-de-France XI, approval no. 11016]. All subjects gave written informed consent in accordance with the Declaration of Helsinki.