If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Cutaneous squamous cell carcinoma (SCC) is the most prevalent invasive malignancy with metastatic potential. The epidermis is exposed to a variety of environmental DNA-damaging chemicals, principal among which are polyaromatic hydrocarbons (PAHs) ubiquitous in the environment, tobacco smoke, and broiled meats. Langerhans cells (LCs) comprise a network of dendritic cells situated adjacent to basal, suprabasal, and follicular infundibular keratinocytes that when mutated can give rise to SCC, and LC-intact mice are markedly more susceptible than LC-deficient mice to chemical carcinogenesis provoked by initiation with the model PAH, 7,12-dimethylbenz[a]anthracene (DMBA). LCs rapidly internalize and accumulate DMBA as numerous membrane-independent cytoplasmic foci. Repopulation of LC-deficient mice using fetal liver LC-precursors restores DMBA-induced tumor susceptibility. LC expression of p450 enzyme CYP1B1 is required for maximal rapid induction of DNA-damage within adjacent keratinocytes and their efficient neoplastic transformation; however, effects of tumor progression also attributable to the presence of LC were revealed as CYP1B1 independent. Thus, LCs make multifaceted contributions to cutaneous carcinogenesis, including via the handling and metabolism of chemical mutagens. Such findings suggest a cooperative carcinogenesis role for myeloid-derived cells resident within cancer susceptible epithelial tissues principally by influencing early events in malignant transformation.
The skin and other epithelial tissues are continually exposed to environmental chemicals and toxins, including those with carcinogenic potential. Mutagenic polyaromatic hydrocarbons (PAHs), generated largely as a result of industrial combustion exhaust and wastewater treatment discharges, have become prevalent in the atmosphere, soil, and ground water. Automobile emissions, processed asphalt, coal burning, grilled or charred meats, and tobacco smoke also all contribute substantially to human PAH exposure. Furthermore, a variety of cosmetics and shampoos are made with coal tar and therefore may contain PAHs. As lipophilic compounds, PAHs are readily absorbed by cells; thus, PAH levels within plants and animals may be much higher than the PAH levels detected in their environments (
Skin exposure to PAHs may occur by direct contact, as well as by distribution to the skin after systemic absorption. People who smoke tobacco, for example, are exposed to high levels of the PAH benz[a]anthracene (
). Mutagenic PAHs form stable, bulky covalent DNA-PAH adducts that lead to double-stranded DNA breaks and nucleic base mismatch reads. The cellular handling and metabolism of PAHs are therefore fundamental to understanding early events in chemical carcinogenesis. Most prior studies, however, have focused on the target epithelial cells of transformation, largely ignoring the other locally resident cells that may have substantial influence on these events.
Within the epidermis, a network of dendritic Langerhans cells (LCs) extend to contact the vast majority of basal, suprabasal, and infundibular keratinocytes, in which tumor suppressor gene (e.g., p53) and oncogene (e.g., Ras) mutations may precipitate squamous cell carcinomas (SCCs). Recently, we demonstrated that LC-deficient (huLangerin-DTA transgenic; DTA (
), despite the fact that the keratinocytes of such mice contain all the requisite enzymes for metabolism of DMBA. The abrogation of the requirement of LC for keratinocyte genotoxicity and tumor induction after topical application of the 3,4-diol DMBA metabolite implicated preferential metabolism of DMBA by LC to reactive intermediates, over detoxification. Consistent with this, murine LCs were found to express cytochrome p450 enzyme CYP1B1, but not CYP1A1, and human LCs express markedly higher levels of CYP1B1 than of CYP1A1. Thus, we hypothesized that the high ratio of CYP1B1:CYP1A1 within LC biased metabolism of DMBA to mutagenic metabolites, inducing DNA damage and provoking carcinogenesis in adjacent keratinocytes that themselves expressed high levels of both p450 enzymes, but in the inverse ratio (CYP1A1>CYP1B1).
To further investigate the mechanisms by which LCs facilitate chemical carcinogenesis, we herein studied the kinetics of LC handling of DMBA and demonstrate that LCs rapidly internalize and accumulate DMBA as numerous membrane-independent cytoplasmic foci that are excluded from LC nuclei. We developed a strategy for the repopulation of LC-deficient mice using fetal liver LC-precursors, and reestablished DMBA-induced tumor susceptibility, proportionate to levels of LC reconstitution. By neonatal repopulation of LC-deficient mice with fetal liver LC-precursors isolated from CYP1B1-deficient donors, we reveal that LC expression of CYP1B1 is required for maximal rapid induction of DNA damage within adjacent keratinocytes, and as well as for their efficient neoplastic transformation. Nonetheless, LCs also show the capacity to mediate pro-carcinogenic effects independently of CYP1B1. For example, consistent with a role for LC in increasing the bioavailability of mutagen independently of metabolism, mice repopulated with LC from CYP1B1-/- donors are more susceptible to DMBA-induced DNA damage compared with LC-deficient mice. In addition, effects on tumor progression attributable to the presence of LC were revealed as CYP1B1 independent. Thus, LCs make multifaceted contributions to cutaneous carcinogenesis via their handling and metabolism of chemical mutagens, as well as their potentiation of tumor progression.
RESULTS AND DISCUSSION
LCs internalize DMBA and facilitate cutaneous carcinogenesis
Keratinocytes possess all the enzymatic requirements to metabolize DMBA to mutagenic forms; yet, application of DMBA followed by weekly promotion with the phorbol ester 12-O-tetradecanoylphorbol-13-acetate (TPA) much more readily facilitates tumor outgrowth in the presence of LC (Figure 1a). These data are consistent with mechanisms where LCs enhance DMBA mutagenesis within adjacent epidermal stem cells. Moreover, 24 hours after topical cutaneous exposure to DMBA, we observed LC mRNA expression, surface protein, and morphologic changes indicative of aryl hydrocarbon receptor–mediated activation. In addition to our previous observation of increased CYP1B1 expression (
), the LC cell line XS106 also displayed mRNA levels increased for CD80 and CD207 and decreased for TLR4 (Figure 1b). As well, 24 hours after percutaneous application of DMBA, isolated LC displayed increased CD86 surface protein (Figure 1c) and, in vivo, converted from a dendritic to a more rounded morphology (Figure 1d and e). These changes are all consistent with that previously reported for aryl hydrocarbon receptor xenobiotic response element activation in myeloid-derived cells (
). Thus, we sought to better understand the characteristics of LC internalization of DMBA, and monitored XS106, as well as isolated murine epidermal LC, by time-lapse two-photon microscopy after in vitro exposure to DMBA. The multi-ringed structures of PAHs provide predictable spectral fluorescence properties, and DMBA fluoresces in the 440 nm wavelength range upon excitation with UV light. Two-photon microscopy systems using pulsed femtosecond laser exposures provide high sensitivity with less phototoxicity, and thus may be used to monitor PAH in live cell systems. XS106 were observed to readily internalize DMBA immediately after exposure, with increasing accumulation evident as discrete intracellular foci over 1 hour, in all cells observed (Figure 2a, Supplementary Movie S1 online). LC isolated from murine skin, in contrast, showed an approximately 10 -minute delay in the appearance of intracellular DMBA foci; yet, numbers of foci per cell accumulated with rapid kinetics to reach levels comparable to that seen in XS106 cells by 20 minutes after exposure (Figure 2a, Supplementary Movie S2 online). Individual foci size increased over time (Figure 2b) indicating intracellular concentration of DMBA. One hour post DMBA exposure, high-resolution z-stack confocal images of isolated LC and XS106 cells with surface membranes labeled with antibodies to MHC II (Figure 2c and d, Supplementary Figure S1 online) revealed the internalized DMBA as numerous, variably sized foci that were excluded from the nuclei. Consistent with the foci representing cytoplasmic collections into hydrophobic micro-droplets following transmembrane diffusion, there was no evidence of endocytic vesicle formation or membrane-bound containment of the DMBA by either MHC II (Figure 2c and d, Supplementary Figure S1 online) or PKH26 red fluorescent membrane tracking dye (Figure 2e). After a 1 hour exposure to DMBA, isolated LCs that were transferred to fresh media without DMBA showed persistent depots of internalized DMBA, even after 24 hours (Figure 2f, Supplementary Movie S3 online). Thus, LCs have the capacity for rapid intracellular accumulation of DMBA. The delay in the appearance of DMBA foci within isolated LC (relative to that within XS106) suggests that normal LC may utilize an activated diffusion prior to the more passive diffusion, whereas the complete absence of MHC II-containing membranes and PKH26 around the DMBA foci indicates that this internalization is not the result of endocytosis. Furthermore, the observed nuclear exclusion is consistent with cellular processes that may protect against LC’s own genotoxicity.
Figure 1Langerhans cells (LCs) facilitate dimethylbenz[a]anthracene (DMBA)-initiated cutaneous carcinogenesis and display phenotypic and morphologic changes in response to DMBA exposure. (a) Carcinogenesis initiated in LC-intact (NLC) and LC-deficient (DTA) mice by application of DMBA (400 nmole) followed by twice weekly 12-O-tetradecanoylphorbol-13-acetate (TPA) (20 nmole, n=13–15 mice per group, ***P≤0.001). (b) XS106 was treated with 64 μM DMBA and changes in gene expression monitored by quantitative real-time PCR, *P<0.05, **P≤0.01, ***P≤0.001. (c) FVB/N dorsal bodywall was treated with vehicle (acetone) or 400 nmole DMBA, and epidermal cell suspensions were prepared 24 hours later and stained for flow cytometric analysis. Population shown is gated on CD45+ CD207+ cells. LCs from untreated (not shown) and vehicle-treated mice had similar CD86 expression (ΔMFI=20.4 vs 17.3; P=n.s.). LC CD86 expression increased following DMBA treatment (ΔMFI=54.4, P=0.0013 vs vehicle, n=3 mice/group). (d, e) FVB/N dorsal ears were treated with 35 nmoles DMBA, and epidermal sheets were prepared 24 hours later and stained with CD207 (red)+MHC II (green). LCs were identified and their shape factor (circularity ranging from 0 to 1, circle=1. Mean±SE untreated 0.1825±0.0019, DMBA treated 0.2254±0.0028, P<0.0001) measured using Volocity 6.2, n=3 mice per group. Scale bar=10 μm.
Figure 2Langerhans cells (LCs) rapidly internalize dimethylbenz[a]anthracene (DMBA) as intracellular foci. (a) DMBA (64 μM) was added to the LC cell line, XS106, or isolated LC and DMBA fluorescence followed using two-photon microscopy and time-series images collected every 2 minutes (XS106, Supplementary Movie S1 online) or 2.5 minutes (LC, Supplementary Movie S2 online) over 1 hour. DMBA foci within individual cells (n=148 XS106, n=81 LC) were identified and quantitated over the first 25 minutes using ImageJ. (b) Individual foci identified within XS106 during the first 10 minutes of DMBA exposure were measured and followed using ImageJ. (c) Compressed z-stack of isolated LC (FITC-MHC II, green) following 1 hour DMBA (blue) exposure. Single slices of z-stack (Supplementary Figure S1 online) suggest nuclear exclusion of DMBA foci. (d) XS106 cells were labeled with FITC-MHC II (green) and the nuclear dye CyTRAK Orange (red) then exposed to DMBA (blue) for 1 hour. DMBA foci were excluded from the nucleus. (e) XS106 cells were labeled with the membrane dye PKH26 (red) then exposed to DMBA (blue) for 1 hour. (f) Isolated LCs were exposed to DMBA for 1 hour, then washed and cultured for an additional 23 hours in the absence of DMBA (Supplementary Movie S3 online). At this time, 42% of LCs retained at least 1 DMBA focus, exemplified in the single slice of the z-stack shown.
Embryonic LC-precursors repopulate the skin of neonatal LC-deficient mice
Taken together, the observed characteristics of LC handling of DMBA (Figures 1 and 2) suggest LC enhancement of epidermal carcinogenesis by DMBA internalization, focal concentration, aryl hydrocarbon receptor-driven expression of metabolizing CYP1B1, and potential transfer of mutagenic metabolites to adjacent keratinocytes. Consistent with this, we previously demonstrated that DMBA applied to the skin resulted in higher rates of epidermal cell DNA damage and increased tumor formation in LC-intact mice relative to LC-deficient mice (
). To better elucidate the previously observed capacity of LC to facilitate epidermal cell genotoxicity and enhance tumor development, we developed a strategy for the efficient repopulation of LC within LC-deficient mice. Recently, it was identified that LC-progenitors first derive as an F4/80bright macrophage-precursor population within the embryonic yolk sac, emerging around embryonic age day (E) 10 to seed the primitive skin directly, or to first expand within the E10-14 liver, prior to seeding and expanding in the embryonic skin (
). Therefore, we hypothesized that E10.5 yolk sac and E12.5 liver mononuclear cells when injected intraperitoneally into LC-deficient mice would seed the skin and expand to repopulate the LC compartment by adulthood, as we and others have demonstrated previously for skin repopulation of dendritic epidermal T cells using E14-17 thymocytes (
). In fact, both E10.5 yolk sac and E12.5 liver LC precursors were capable of repopulating the epidermis when transferred into newborn LC-deficient DTA, with the E12.5 liver showing slightly greater efficiency (Figure 3a). Notably, 3 weeks following transfer, LCs were observed as circular expanding populations in the recipient epidermis (Figure 3b, and by 8 weeks eventually filled the epidermis of adult recipients (Supplementary Figure S2 online). These observations provide insight into the normal physiologic population of the epidermis by LC precursors present within embryonic yolk sac (
), and the “seed–soil” relationship of precursors that can expand over weeks after seeding the immature skin.
Figure 3Embryonically derived Langerhans cells (LC)-precursors repopulate LC-deficient huLangerin-DTA transgenic Langerhans cell deficient mouse (DTA) mice. (a) Cell suspensions prepared from w.t. E10.5 yolk sac or E12.5 liver were injected i.p. into newborn LC-deficient DTA recipients. Three or eight weeks later, flow cytometric analysis of epidermal cell suspensions prepared from recipients as well as LC-intact (NLC) and DTA controls were used to determine the extent of LC repopulation. (b) Immunofluorescent staining of an epidermal sheet prepared from a DTA recipient 3 weeks following E12.5 liver transfer shows expanding groups of LC (CD207, green). Scale bar=500 μm. Inset scale bar=100 μm. (c) E12.5 liver suspensions prepared from w.t. or CYP1B1-/- donors were injected i.p. into recipients of indicated ages. Epidermal cell suspensions prepared 8 weeks later and analyzed by flow cytometry show decreasing LC repopulation with increasing recipient age at the time of transfer. No repopulation was detectable when recipients were 5 weeks old at transfer. (d) Flow cytometric analysis of epidermal cell suspensions prepared from bodywall skin of 8-week-old NLC, DTA, and DTA recipients of E12.5 liver. Upper panels are gated on total epidermal cells; lower panels are gated on CD45+ cells. (e) Epidermal sheets prepared from bodywall skin 8 weeks following E12.5 liver transfer show typical LC (CD207, green) and dendritic epidermal T cell (TCRγδ, red) density and morphology when examined by confocal microscopy. Scale bar=10 μm.
To examine the effect of recipient age at the time of transfer on the ability of precursors to enter the epidermis and expand, E12.5 liver-derived LC-precursors were injected intraperitoneally into LC-deficient recipients ranging in age from <36 hours to 5 weeks, and the resulting epidermal LC populations analyzed and compared with age-matched normal controls. LCs were reestablished most efficiently when transferred into newborn recipients (Figure 3c) and displayed phenotypic (Figure 3d) and morphologic (Figure 3e) characteristics of normal LC. The requirement for neonatal skin, with marked decreasing efficiency of LC repopulation in recipients >36-hour old, is in contrast to our experience with repopulation of dendritic epidermal T cell where E16 thymocyte precursors will repopulate adult epidermis provided the skin is in active hair growth at the time of transfer (Supplementary Table S1 online).
CYP1B1 expression by LC is required for maximal genotoxicity of adjacent keratinocytes and tumor development
Despite the fact that keratinocytes express the enzymes (CYP1A1, CYP1B1, and epoxide hydrolase) necessary to metabolize DMBA to its ultimate mutagenic form, DMBADE, we previously proposed that the higher ratio of CYP1A1:CYP1B1 within keratinocytes favors detoxification (
). Given the capacity of LC to internalize and accumulate DMBA (Figures 1 and 2), and our previous demonstration of the capacity of LC to metabolize DMBA to its penultimate mutagenic form, DMBA-3,4-diol, we hypothesized that CYP1B1 expression within LC would be required for maximal genotoxicity of adjacent keratinocytes after epidermal exposure to DMBA. Thus, we repopulated LC-deficient mice with E12.5 liver LC-precursors from CYP1B1-/- mice (courtesy of F. Gonzalez;
), which we first backcrossed to the FVB background by speed congenic breeding, and compared them to LC-deficient mice repopulated with precursors from wild-type (w.t., CYP1B1-competent) FVB mice. LC repopulation levels and LC surface phenotype and morphology were equivalent in DTA recipients of w.t. or CYP1B1-/- LC precursors (Figure 3c–e).
Phosphorylated (γ)H2Ax is recruited to double-stranded DNA breaks and DNA-PAH adducts and is therefore a nonspecific but sensitive indicator of DNA damage, including that as a result of DMBA exposure (
). Twenty-four hours after a single application of DMBA, LC-deficient mice repopulated with LC-precursors from wild-type mice showed levels of γH2Ax comparable to that of wild-type mice (Figure 4a and b). Thus, repopulation of LC by intraperitoneal injection of embryonic liver isolated LC-precursors into neonatal recipients results by adulthood in functionally competent LC, capable of facilitating keratinocyte genotoxicity after DMBA exposure. In contrast, LC-deficient mice repopulated with LC-precursors from syngeneic CYP1B1-/- mice showed significantly reduced levels of γH2Ax relative to LC-intact controls but which were statistically distinguishable from that of mice devoid of all LCs (Figure 4a and b). Therefore, CYP1B1 expression specifically within LC contributes substantially to neighboring keratinocyte DNA damage following DMBA exposure, even though this DMBA metabolizing enzyme is present in the surrounding cells. In addition, the observation that DTA mice repopulated with CYP1B1-/- LC had increased DNA damage compared with mice devoid of all LCs indicates that LCs also participate in CYP1B1-independent events that facilitate damage of surrounding cells. Within each of the DMBA-exposed groups, only ∼1% of LCs showed γH2AX staining indicative of DNA damage, whereas LC density 24 hours post DMBA exposure was not distinguishable from that of untreated mice (Supplementary Table S2 online). This, together with the observed lack of nuclear DMBA fluorescence within LC (Figure 2d), suggests that LC may possess mechanisms to guard against their own genotoxicity while contributing to adjacent keratinocyte damage.
Figure 4Langerhans cell (LC) expression of CYP1B1 is necessary for optimal dimethylbenz[a]anthracene (DMBA)-induced DNA damage. (a) Ears of LC-intact (NLC) and huLangerin-DTA transgenic Langerhans cell deficient mouse (DTA) controls as well as DTA repopulated with either w.t. or CYP1B1-/- LC were treated with 35 nmole DMBA. Twenty-four hours post treatment epidermal sheets were prepared for γH2AX+ cell enumeration. Each dot represents one mouse. *P<0.05, **P≤0.01, ***P≤0.001. (b) Representative confocal images from (a) show LC (CD207, green) and nuclei containing DNA damage (γH2AX, red). Scale bar=10 μm.
When groups of LCs repopulated and control mice were followed for tumor development using the standard DMBA-TPA model, LC-deficient mice repopulated with LC-precursors from wild-type embryos were again similar to normal FVB (LC-intact) mice (Figure 5a–c), and the level of LC reconstitution measured at 8 weeks, i.e., immediately prior to initiation with DMBA, was directly correlated with the extent of tumor development (R=0.7528; P=0.0155, data not shown). In contrast, LC-deficient mice repopulated with CYP1B1-deficient LC precursors were indistinguishable by tumor area per mouse (Figure 5a) or tumor area/tumor-bearing mouse (Supplementary Table S3 online) from LC-deficient control DTA mice through 12 weeks of tumor promotion, and showed markedly reduced tumor susceptibility relative to LC-deficient mice repopulated with w.t. LC precursors and NLC (w.t.) controls (Figure 5a–c).
Figure 5Langerhans cell (LC) expression of CYP1B1 is required for optimal cutaneous chemical carcinogenesis. (a) Chemical carcinogenesis was initiated in LC-intact (NLC) and huLangerin-DTA transgenic Langerhans cell deficient mouse (DTA) controls as well as DTA repopulated with either w.t. or CYP1B1-/- LC using 400 nmoles dimethylbenz[a]anthracene (DMBA) followed by twice weekly 12-O-tetradecanoylphorbol-13-acetate (TPA) promotion (20 nmole); ***P<0.001 DTA+CYP1B1-/- LC vs DTA+w.t. LC and vs NLC; and P<0.001 for DTA vs DTA+w.t. LC and vs NLC. N=5–15 mice/group for panels a–c. (b) Representative images from (a) taken 12 weeks following initiation. (c) The percentage of tumor-free mice over time depicted in Kaplan–Meier plot. P<0.05 for DTA vs DTA+CYP1B1-/- LC, P≤0.01 for DTA+CYP1B1-/- LC vs DTA+w.t. LC and vs NLC, P<0.0001 for DTA vs NLC and vs DTA+w.t. LC. (d, e) Fourteen weeks post initiation, DTA repopulated with CYP1B1-/- LC have more total tumors (d) and carcinomas (e) compared with LC-deficient DTA; n=7 DTA, 5 DTA+CYP1B1-/-LC. (f) Epidermal hyperplasia, as measured by minimal epidermal thickness, in response to repeated TPA application (4 weeks) is greater in the presence of LC; n=5 DTA, 5 NLC. (g) Ear thickness 48 hours after a single TPA application is greater in the presence of LC; n=6 DTA, 8 NLC. Baseline ear thickness (g) and minimal epidermal thickness (f) are not different in DTA vs NLC. *P<0.05, **P≤0.01, ***P≤0.001. Each dot represents one mouse in panels d–g.
However, tumors in DTA mice repopulated with CYP1B1-/- LC did arise earlier than in DTA controls (Figure 5c). This suggested that LC exerted CYP1B1-independent effects on tumor promotion. Therefore, we continued to follow co-caged cohorts of DTA and DTA repopulated with CYP1B1-/- LC. Notably, tumor growth was observed to statistically increase in CYP1B1-deficient LC-repopulated mice, relative to LC-deficient controls at 14 weeks, further suggesting a CYP1B1-independent tumor-promoting role for LC (Figure 5d and e). Consistent with this, independent studies in DMBA-naive mice showed increased epidermal hyperplasia in LC-intact skin relative to LC-deficient skin after repeated exposure to TPA alone (Figure 5f) and increased ear thickness 48 hours following a single application of TPA (Figure 5g). Each of these results suggests that LCs influence the response to epidermal perturbation, directly or indirectly promoting eventual keratinocyte proliferation that is fundamental to tumor promotion.
Collectively, our data reveal multiple distinct roles for LC in chemical carcinogenesis: (1) morphologic and structural features that support efficient PAH internalization and collection within LC, (2) metabolism to mutagenic metabolites that are specifically dependent on LC expression of CYP1B1, and (3) CYP1B1-independent contributions to DNA damage within adjacent keratinocytes and to tumor promotion. The contributions of immune cells to tumor growth have been previously characterized to include the activities of tumor-associated macrophages, tumor-promoting T cells (
), and T-regulatory cells that may increase oxidative stress, provide growth factors, and suppress anti-tumor immunity to facilitate tumor immune evasion and growth. We have extended these untoward effects of immune cells under the concept of cooperative carcinogenesis to emphasize that immune cells may be fundamental to the full spectrum of initiation and promotion.
Major lines of epidemiologic and experimental study have linked mutagenic PAHs to cancer within a variety of tissues (e.g., skin, lung, gastrointestinal, bladder, and breast) (
). In addition, a role for mutagenic PAHs in cutaneous carcinogenesis, including their potentiation of UVR, has been implicated by several lines of laboratory investigation. Purified particles of air pollution applied to the skin of mice can serve as an initiator of cutaneous carcinogenesis, and this is dependent on the expression of the intracellular aryl hydrocarbon receptor for PAHs (
). Initiation by topical application of mutagenic benzo[a]pyrene followed by exposure to UV radiation synergistically provokes skin tumor formation in mice (
). Furthermore, chronic exposure of mice to both inhalation of PAH-containing tobacco smoke and UV radiation induces the outgrowth of cutaneous SCC at a much greater rate compared with either agent alone (
). A single application of the PAH 7,12-DMBA prior to UV radiation of human skin xenografts markedly increases the development of actinic keratoses and SCC, relative to the same dose of UV alone (
). Thus, although UV radiation is well established as a genotoxic agent associated with human skin cancer risk, these studies collectively reveal the potential of chemical mutagenesis by PAHs as an underappreciated cofactor in the induction of human cutaneous malignancy.
SCSs constitute the most common human solid tumor malignancy including those of the skin, gastrointestinal, genitourinary, breast, and lung origins. They derive, by definition, from continuously regenerating epithelial cells that encompass large surface areas at the interface with the environment, where they may be exposed to myriad environmental compounds, including mutagenic PAHs. In addition to direct contact with exogenous agents of DNA-damaging potential, systemic absorption of such mutagens through the epithelium (e.g., via smoking, industrial exposures, diets high in grilled meats) can lead to more global tissue exposures. As CYP1B1 has also been shown to metabolize the carcinogenic PAH benzo[a]pyrene to its mutagenic intermediate (
Metabolism of benzo[a]pyrene to trans-7,8-dihydroxy-7,8-dihydrobenzo[a]pyrene by recombinant human cytochrome P450 1B1 and purified liver epoxide hydrolase.
), it is conceivable that a spectrum of benz[a]athracene and benzo[a]pyrene-related compounds present in the environment might be predicted to undergo similar handling and activation within the skin and other tissues.
Moreover, it is also possible that differential cellular expression of metabolizing enzymes other than CYP1B1 may also contribute to carcinogenesis. For example, LC also preferentially expresses CYP2E1 (
) and azoxymethane (Sohn et al., 2001). Thus, although we believe our findings definitively demonstrate that LC expression of CYP1B1 is fundamental to DMBA mutagenesis of neighboring keratinocytes, further investigation is warranted to fully appreciate the potential generalizability to other dendritic cell populations, other metabolizing enzymes, and other epithelial and non-epithelial tissues.
The roles (beyond antigen presentation) of Langerhans cells in cutaneous biology have been expanded to also include their potential for immune regulation (
), it is unlikely that T cell–mediated regulation has a significant role in this stage of chemical carcinogenesis. In contrast, the previously described tumor-promoting αβ T cells and anti-tumor γδ T cells (
The distinct contributions of murine T cell receptor (TCR) gammadelta+ and TCRalphabeta+ T cells to different stages of chemically induced skin cancer.
) raise the possibility that these components of adaptive immunity may interact with LC to, in aggregate, contribute to tumor promotion. In this regard, application of higher doses of DMBA (or UV light) to mouse skin results in depletion of LC largely due to migration to draining lymph nodes, and percutaneous immunization in such mice shows blunted antigen-specific responses that might include anti-tumor effects (
). Thus, although we have herein provided additional insight into mechanisms by which LCs influence chemical carcinogenesis, further studies are necessary to more clearly delineate the complexity of interactions between LC and other immune cells in regulating tumor promotion.
MATERIALS AND METHODS
Animals and housing
Mice were maintained under specific pathogen-free conditions in a facility accredited by the Association for Assessment of Laboratory Animal Care with food and water provided ad libitum. The Yale Animal Care and Use Committee approved all in vivo studies. FVB/N mice were purchased from The Jackson Laboratory. FVB/N LC–deficient (DTA) mice were generated with a bacterial artificial chromosome–based transgene containing human Langerin modified to drive expression of diphtheria toxin subunit A from an internal ribosome entry site inserted in the 3′ untranslated region of the langerin gene (
). Cyp1b1-/- mice were kindly provided by F. Gonzalez (NCI, Bethesda, MD) and backcrossed onto the FVB background via speed congenic methodology (testing provided by IDEXX RADIL, Columbia, MO) for six generations.
DMBA fluorescence
The LC-like cell line XS106 was kindly provided by A. Takashima (U of Toledo, OH) and cultured as previously described (
). Epidermal cell suspensions were prepared as previously described (Strid et al., 2008) and LC enriched using either MHC II- or CD45-magnetic beads (Miltenyi Biotec, San Diego, CA) followed by staining with PE or FITC-MHC II (eBioscience, San Diego, CA) and MoFlo (DAKO, Carpinteria, CA) cell sorting to obtain 98% purity. XS106 or purified LCs were cultured overnight in 35 mm glass bottom culture dishes (MatTek Corp, Ashland, MA) in phenol red-free RPMI 1640 (Invitrogen, Grand Island, NY) containing 10% FBS, 10 mM Hepes, 1% nonessential amino acids, 2 mM L-glutamine, 1 mM sodium pyruvate, 0.05 mM β-mercaptoethanol, and antibiotics before use in DMBA fluorescence monitoring. DMBA (Sigma, St Louis, MO; 64 μM) was added after images of untreated cells were collected. Imaging was performed using a Zeiss (Oberkochen, Germany) 710 Duo NLO/Multiphoton microscope running Zen 2010 software and equipped with a Spectra-Physics MaiTai titanium:sapphire laser for two-photon excitation, and an enclosed CO2 incubator (37 °C). DMBA fluorescence was collected from 400 to 490 nm (peak 440 nm). In some experiments FITC or PE-MHC II (eBioscience) antibody was added (0.25 μg ml-1) to permit LC cell membrane visualization. XS106 was labeled with CyTRAK Orange (5 μM, eBioscience) or membrane dye PKH26 (Sigma) according to the manufacturer’s instructions and exposed to DMBA (16 μM) for 1 hour before z-stacked images were obtained. Live time-series images were taken every 2 or 2.5 minutes over 1 hour. Individual cells and contained DMBA foci were identified and quantitated using ImageJ v1.47 (NIH).
Confocal microscopy
Epidermal sheets were prepared following a 2 hour 37 °C incubation in 20 mM EDTA (for LC morphology post DMBA application and LC and dendritic epidermal T cell visualization in repopulated mice) or a 20 minute 37 °C incubation in 0.5 M ammonium thiocyanate (for γH2AX analysis), fixed 20 minutes in ice-cold acetone, rehydrated in PBS, blocked with 2% BSA, and stained using the following antibodies: anti-CD207 (1.5 μg ml-1, RMUL.2, eBioscience)+Alexa488- or Cy3-Donkey-anti-rat IgG (Jackson Immunoresearch, West Grove, PA), anti-TCRγδ (1.5 μg ml-1, GL3, Becton Dickinson, San Jose, CA)+Cy3-goat anti-hamster IgG (Jackson Immunoresearch), anti-γH2AX (1 μg ml-1, JBW30, Millipore, Billerica, MA)+Alexa568-Goat-anti-mouse IgG (Invitrogen), and FITC-MHC II (1.6 μg ml-1, M5/114.15.2, Biolegend, San Diego, CA). Samples were examined using a Zeiss 510Meta confocal microscope. Analysis of γH2AX+ cells was performed as previously described (
). Briefly, 35 nmole DMBA was applied to the dorsal surface of each ear, and 24 hours later epidermal sheets were prepared, stained, and z-stacked images of 10 fields per sheet (2 sheets per mouse) collected in a set pattern in 2 mm intervals. ImageJ (NIH) particle analyzer software was used to identify and quantitate γH2AX+ cells. For analysis of LC morphology 24 hours post DMBA application, images collected as described above were analyzed using Volocity 6.2 (Perkin Elmer, Waltham, MA) for LC identification and measurement of shape factor (circularity on a scale of 0–1 with 1=circle). The mean and distribution of shape factor measurements were calculated using Prism 6.0 (GraphPad, La Jolla, CA).
Gene expression analysis
RNA was isolated from XS106 cells using the RNeasy Micro kit (Qiagen, Valencia, CA) and transcribed using the High-Capacity cDNA Reverse Transcription kit (ABI, Carlsbad, CA). Real-time quantitative RT-PCR (ABI 7500, SDS 2.0 software) was performed using Taqman assays and Taqman Gene Expression Mastermix (ABI). Obtained Ct values were normalized to β-actin and expression differences to untreated samples calculated using RQ=2-ΔΔCt.
Flow cytometric analysis
To examine LC phenotype following DMBA exposure, epidermal cell suspensions were prepared as previously described (Strid et al., 2008) from dorsal bodywall of FVB/N mice that had been clipped and treated with depilatory cream 5 days earlier and treated with vehicle (acetone) or DMBA (400 nmole) 24 hours before harvest. After blocking FcR (CD16/32), cells were stained with PerCPCy5.5-CD45, PE-MHC II, APC-CD86 (all from eBioscience), and Alexa488-CD207 (Imgenex, San Diego, CA) and data collected using FACSCalibur (Becton Dickinson) and analyzed with FlowJo (TreeStar, Ashland, OR). To assess LC repopulation levels, epidermal cell suspensions were prepared from either ear punch biopsy (prior to DMBA application in carcinogenesis studies) or from untreated bodywall skin. Epidermal cells from LC-intact (NLC) and LC-deficient (DTA) as well as those from each repopulated mouse were blocked (CD16/32) and stained with APC-CD45+PE-MHC II+Alexa488-CD207. In each experiment the average percentage of CD45+MHC II+CD207+ epidermal cells in NLC controls was calculated and the percent LC repopulation of each individual yolk sac or liver recipient determined by comparison.
LC repopulation
E10.5 yolk sac or E12.5 liver was isolated from donor embryos. Yolk sac was digested as described (
) by incubation in 0.25% collagenase in PBS containing 20% FBS for 40–60 minutes at 37 °C. Liver was dissociated by gentle pipetting. Cells were washed once in PBS+10% FBS and resuspended in PBS at 4 donor equivalents per 30 μl. Newborn LC-deficient recipients received 30 μl donor cells by intraperitoneal injection. The level of LC repopulation was assessed by flow cytometric analysis.
) at age 7 weeks, dorsal bodywall hair was removed by shaving with electric clippers followed by application of depilatory cream. One week later carcinogenesis was initiated by application of DMBA (400 nmole, Sigma) in acetone. One week following initiation, twice weekly applications of TPA (20 nmole, Sigma) in ethanol began. Tumors were counted, measured, and scored weekly as clinically apparent papillomas (i.e., well demarcated, symmetrical, pedunculated, or dome-shaped papules, without erosion or ulceration) or clinically apparent carcinomas (poorly demarcated, asymmetrical, non-pedunculated, or dome-shaped papules with erosion or ulceration) by an observer blinded to the experimental groups, as previously described (
The distinct contributions of murine T cell receptor (TCR) gammadelta+ and TCRalphabeta+ T cells to different stages of chemically induced skin cancer.
). After 12 weeks of promotion, tumor burden was a humane concern for NLC and DTA+w.t. LC groups, therefore, only co-caged cohorts of DTA and DTA+CYP1B1-/- LC groups were carried through 14 weeks. At the conclusion of experiments, representative tumors were excised, fixed, and embedded for hematoxylin and eosin staining and were examined by a certified dermatopathologist (C.K.) for histologic confirmation. The terms “tumor area” and “tumor number” are used to refer to all tumors present (papillomas+carcinomas).
Irritant responses
For analysis of epidermal hyperplasia, TPA (20 nmole) in ethanol was applied to dorsal bodywall skin twice weekly for 4 weeks, then formalin fixed paraffin-embedded samples prepared, stained with hematoxylin and eosin, and examined using an Olympus BH10 microscope equipped with an optical micrometer. Minimal epidermal thickness was measured from the basal to the granular layer of the epidermis in 12 fields per section in 4 samples per mouse. Ear thickness following a single application of TPA (0.5 nmole to each side of each ear) was measured using an engineer’s micrometer (Mitutoyo 7301). Measurements were made by an observer blind to the experimental groups.
Statistical analysis
One-tailed Student’s t-test for unpaired data was used to determine the statistical significance of difference between experimental groups in all cases, except Figure 1a, b, and d and Figure 5a and c, with significance at P<0.05. Curves in Kaplan–Meier plots (Figure 5c) were compared using the log-rank test. Changes in tumor area or gene expression over time (Figure 1a and b and Figure 5a) were compared using repeated-measures ANOVA with appropriate correction for multiple time point and multiple group comparisons (Tukey’s correction for Figure 5a, Sidak’s correction for Figure 1a and b). Frequency distributions (Figure 1d) were compared with the nonparametric Kolmogorov–Smirnov test. All statistical analyses were performed with GraphPad Prism 6.0 software.
ACKNOWLEDGMENTS
Funding for this work was provided by the NIH grant R01CA102703 (to MG) and the Novartis Foundation for Medical-Biological Research (to CB). We thank F Geissman (KCL) and F Ginhoux (A*STAR) for protocol advice, A Takashima (U Toledo) for XS106 cells, and F Gonzalez (NCI) for Cyp1b1-/- mice.
The distinct contributions of murine T cell receptor (TCR) gammadelta+ and TCRalphabeta+ T cells to different stages of chemically induced skin cancer.
Metabolism of benzo[a]pyrene to trans-7,8-dihydroxy-7,8-dihydrobenzo[a]pyrene by recombinant human cytochrome P450 1B1 and purified liver epoxide hydrolase.
37256-0&id=gr1.jpg){kind=link}
37256-0&id=gr2.jpg){kind=link}
37256-0&id=gr3.jpg){kind=link}
37256-0&id=gr4.jpg){kind=link}
37256-0&id=gr5.jpg){kind=link}