Dermal Mast Cells Determine Susceptibility to Ultraviolet B–induced Systemic Suppression of Contact Hypersensitivity Responses in Mice

Pathogen-free, female BALB/c, DBA/2, and C57BL/6J mice were obtained (Animal Resource Center of the South Australian Department of Agriculture, Adelaide, Australia). Mice carrying the W   ^f mutation on both the C57BL/6 and DBA/2 background were obtained originally from the Walter and Eliza Hall Institute of Medical Research (Melbourne, Australia; reference 17). The homozygous mast cell–depleted mice (W   ^f/W   ^f), the heterozygous mice (W   ^f/+), and the wild-type mice (+/+) were the progeny of the mating of (C57BL/6W   ^f × DBA/2W   ^f) F1 hybrids, with the W   ^f/W   ^f, W   ^f/+, and +/+ mice identified by coat color. The parental strains for the Uvs1 mice were obtained from the Frederick Cancer Research Facility (Frederick, MD). All mice were of equivalent age (69–80 d) and were not in hair growth as determined by histology and increased skin fold thickness. All experiments were performed according to the ethical guidelines of the National Health and Medical Research Council of Australia.

Samples of dorsal skin (2–4/mouse) were rolled, fixed in 10% buffered formalin, and paraffin embedded ensuring vertical orientation of the roll, and 4-μm sections were cut and immunohistochemically stained with a 1/400 dilution of rabbit polyclonal antihistamine antibody (Chemicon, Temecula, CA). After incubation with a biotin-conjugated sheep anti–rabbit immunoglobulin (Silenus, Hawthorn, Australia) and peroxidase-conjugated streptavidin (Dako, Carpinteria, CA), the reaction product was developed with 3,3′-diaminobenzidine tetrahydrochloride (Sigma Chemical Co., St. Louis, MO). Sections were counterstained with hematoxylin solution. Control sections were incubated with diluted rabbit serum (1/400) or no primary antibody. The staining for histamine was specific as preincubation of the antibody with histamine significantly reduced the staining detected.

A video image analysis system (Video Pro 32; Leading Edge Pty Ltd., Adelaide, South Australia) was used to quantify histamine-staining cells in 15 consecutive fixed fields in each roll of skin. The epidermis was orientated horizontally and using a ×10 objective and a ×2.5 camera eyepiece; fields covering the epidermis through to the SCFT were selected. Mast cells (i.e., histamine-staining cells) and area of dermis and SCFT per fixed field of 800-μm width were measured (but expressed per mm); the coefficient of variation for measurement of mean dermal area for 15 fields was <14%. Dermal mast cell prevalence (see Figs. 1,A and 4,A) was determined by dividing the mast cell numbers per mm horizontal field length (Figs. 1,B and 4,B) by the dermal area for that field (Figs. 1,C and 4 C). In all experiments, the identity of the sections for examination by video image analysis was coded except for the positive control that was a section from skin rolls prepared from a C57BL/6J mouse. This positive control demonstrated minimal difference in staining intensity for mast cells from different staining runs. For measurement of the image background in all measurements of optical density, readings were standardized using a bright field image of the background illumination. This allowed correction of nonuniform illumination over a field of view, which would otherwise affect the discrimination of features by their gray value.

For quantification of the histamine content in each cell, the integrated optical density (representing the total amount of histamine staining in a given cell) was measured, as well as the area of histamine staining in each cell. For two rolls from each of six mice (three C57BL/6J and three BALB/c) all processed and stained within the same staining run, 200 mast cells were captured systematically in consecutive fields using a ×40 objective. The cells were categorized according to distribution in the upper or lower dermis (equal halves) or SCFT.

The UV source was a bank of FS40 sunlamps (Westinghouse Corp., Pittsburgh, PA) emitting a broad band of UV, 250–360 nm, with 65% of the output in the UVB range (280–320 nm). A polyvinylchloride plastic film was used to screen out wavelengths of <290 nm. The dose rate was monitored using a UVX radiometer with a UVX-31 sensor (Ultraviolet Products Inc., San Gabriel, CA). For irradiation of the mice, a uniform dorsal area (8 cm²) was clean shaven, the ears protected with black adhesive insulation tape, and the mice housed in individual compartments of perspex cages. The sunlamps were held 20 cm above the cages.

Histamine (0.2–200 μg/mouse; Sigma Chemical Co.) was dissolved in mouse osmolality PBS (330 mOsm/kg H₂O) and injected subcutaneously into the backs of mice 5 d before sensitization with hapten. All mice received an equal volume of diluent, with or without histamine.

For assay of systemic CHS responses, 5 d after UVB irradiation or histamine administration, mice (five per group) were sensitized on the shaved ventral skin with 100 μl of freshly prepared 5% (wt/vol) 2,4,6-trinitrochlorobenzene (TNCB; Tokyo Kasei Kogyo Ltd., Tokyo, Japan) in acetone. 5 d later and after coding the identities of the mouse groups, a CHS response was elicited by applying 10 μl freshly prepared 1% TNCB in acetone to each of the ventral and dorsal surfaces of both ears. 24 h after challenge, the ear thickness was measured with a micrometer (Mitutoyo Corp., Tokyo, Japan) and the extent of ear swelling for each mouse was calculated by subtracting the ear thickness before challenge. The mean swelling measured in mice that were challenged, but not sensitized, with TNCB (∼0.03 mm) was subtracted from this value. For local CHS responses, mice were sensitized 3 d after irradiation on the same dorsal site as UVB-irradiated with 100 μl freshly prepared 1% TNCB in acetone. The challenge of ears with TNCB was as described above.

4 d after UVB irradiation, W   ^f/W   ^f and W   ^f/+ mice were immunized by subcutaneous injection of 2.5 × 10⁷ allogeneic spleen cells (from BALB/c mice) in 50 μl PBS into each flank of each mouse. 7 d later, the thickness of the hind footpads of the animals was measured with a micrometer before challenge with a subcutaneous injection of 10⁷ allogeneic spleen cells in a volume of 40 μl PBS. After 24 h, the footpad thickness was measured again and the extent of swelling for each footpad was calculated by subtracting the footpad thickness before challenge.

Bone marrow cells from wild-type (+/+) mice were cultured at 10⁶/ml in Eagles's MEM, alpha modification (alpha MEM; Sigma Chemical Co.) supplemented with 10⁻⁴ M 2-mercaptoethanol, 20% horse serum (Gilles Plains Animal Resource Centre, Adelaide, Australia), and 10% pokeweed-mitogen–stimulated spleen cell–conditioned medium (PWM-SCM). PWM-SCM was prepared as previously published (18) with 2 × 10⁶ cells/ml incubated for 5 d in alpha MEM supplemented with 10% fetal calf serum, 10⁻⁵ M 2-mercaptoethanol, and 1:300 dilution of pokeweed mitogen (Sigma Chemical Co.). Culture flasks were incubated at 37°C in 5% CO₂ with half the medium replaced every 7 d. After 31 d, the cells were washed and suspended in alpha MEM (2.5 × 10⁶/ml) for injection in vivo. Each W   ^f/W   ^f mouse received 10⁶ cells in 8 × 50 μl subcutaneous injections in two rows within a 1-cm width × 2-cm length of shaved dorsal skin. From 3.00 × 10⁸ bone marrow cells harvested from 16 +/+ mice, 2.3 × 10⁷ cells (7.6% yield) were harvested at the end of the 31-d culture period. The cultured cells (>85%) stained blue after staining with alcian blue–safranin (Csaba's mast cell stain; reference 19) indicating immature mast cells.

6 wk after reconstitution, an 8-cm² dorsal area was shaved on the 18 W   ^f/W   ^f mice injected with bone marrow–derived mast cell precursors (10 male, 8 female). On the next day, 13 mice were irradiated with 12 kJ/m² UVB and 5 served as controls. 5 d later, the mice were assessed for a systemic CHS response as described above. At the end of the systemic CHS assay, the mice were killed and dorsal skin sampled for evaluation of mast cell numbers. Mast cells in the dorsal skin of the bone marrow–reconstituted W   ^f/W   ^f mice were confirmed to be mature connective tissue mast cells by staining red with the alcian blue–safranin staining procedure (data not shown).

For mast cell quantification and measurement of area of the dermis and SCFT, a mean value for each mouse was determined from 30–60 measurements. For measurement of integrated optical density and stained area per cell, the mean result from individual mice was used to calculate the mean value ± SD for three mice. A multiple comparison procedure using a one-way analysis of variance and Fisher's least significant difference test was used to determine the statistical significance (P <0.05) of differences between groups of mice from different strains. This form of analysis was also used for a test of significance between groups of mice within a CHS experiment. If the results from only two strains were compared (e.g., BALB/c versus BALB/c-Uvs1, C57BL/6J versus BALB/c for measurements of integrated optical density), an unpaired Student's t test was used. Strain-dependent effects of UVB or histamine on CHS to TNCB was examined by a multifactorial analysis of variance (MANOVA). Probabilities of <0.05 were considered significant.

With the use of a video image analysis system, the prevalence of histamine-staining dermal mast cells (mast cells per mm²) was found to be significantly greater in C57BL/6J mice than in DBA/2 mice (P <0.05), which in turn had a significantly greater dermal mast cell prevalence than measured in BALB/c mice (P <0.05; Fig. 1,A). When mast cell numbers and tissue areas were considered separately in the different mouse strains, similar significant differences (P <0.05) were detected in dermal mast cell numbers per mm horizontal field length (Fig. 1,B). In the SCFT, there was no significant difference in the number of mast cells per mm horizontal field length of DBA/2 and BALB/c mice (Fig. 1,B). However, there was a significant decrease in mast cell prevalence in the SCFT of C57BL/6J mice (P <0.05; Fig. 1,B). The area of the dermis or SCFT per unit horizontal length of tissue for the three strains of mice examined did not vary significantly (Fig. 1,C). There was a linear relationship between the prevalence of histamine-staining dermal mast cells and the log of the dose of UVB demonstrated by others (5) to cause 50% suppression of CHS responses in the different mouse strains (Fig. 1,D). Representative histologic sections from BALB/c and C57BL/6J mice are shown in Fig. 2, A and B, respectively.

Differences in the histamine content of mast cells from C57BL/6J and BALB/c mice were investigated by measurement of the integrated optical density for histamine staining in 400 skin mast cells/mouse. The relative distribution of the 400 mast cells assessed varied for the different mouse strains (Fig. 3,A), as well as the tissue area required to be scanned in order to measure 400 mast cells. The mean area for histamine staining in mast cells from all mice was similar, the variation in stained areas approximating a normal distribution. The mean integrated optical density for histamine staining per cell was very similar for C57BL/6J and BALB/c mice (Fig. 3 B).

Mast cell numbers in the dermis and SCFT of BALB/c-Uvs1 mice and BALB/c mice of the same parental origin were investigated (Fig. 4). Within each of the three experiments (five BALB/c-Uvs1 and five BALB/c in experiment 1, three BALB/c-Uvs1 and two BALB/c in experiment 2, and three BALB/c-Uvs1 and two BALB/c in experiment 3), all BALB/c-Uvs1 animals were littermates. Dermal mast cell numbers were significantly higher in the BALB/c-Uvs1 mice (P = 0.0001; Fig. 4,A), although there were no differences in the numbers of mast cells in the SCFT (Fig. 4,B). It should be noted that variation in dermal mast cell numbers within the BALB/c- Uvs1 group occurred in all three experiments (three high and two low in experiment 1, two high and one low in experiment 2, and two high and one low in experiment 3). Dermal areas for the BALB/c-Uvs1 and BALB/c mice were not significantly different (Fig. 4 C).

To confirm that histamine was a mast cell product that could limit quantitatively UVB-induced immunosuppression in BALB/c (low susceptibility) versus C57BL/6J (high susceptibility) mice, the different mouse strains were UVB irradiated or injected subcutaneously with increasing doses of histamine 5 d before sensitization with the hapten, TNCB, and assay of a CHS response to that hapten (Fig. 5). The strain-dependent effects of UV on CHS to TNCB were significant by multifactorial analysis of variance, with 50% suppression at UVB doses as published (reference 5; Fig. 5,A). However, there was no difference between the strains at a dose of histamine that was immunosuppressive (20 μg/mouse; Fig. 5 B).

Mast cell–depleted mice (W  ^f/W  ^f; reference 17) were then tested for susceptibility to UVB-induced systemic immunosuppression. The prevalence of dorsal dermal mast cells in the homozygous (W  ^f/W  ^f) mice was ∼12% of that measured in wild-type mice of the same genotype (Fig. 6,A). Systemic CHS responses by W  ^f/W  ^f mice, in contrast to heterozygous (W  ^f/+) and wild-type (+/+) mice, were not suppressed by UVB at the dose of irradiation tested (Fig. 6,B). However, when the mast cell product histamine was exogenously introduced 5 d before hapten sensitization, CHS responses in these animals were reduced (Fig. 6 C).

In a manner similar to regulation of systemic CHS responses, UVB irradiation was able to suppress a DTH response to allogeneic spleen cells in heterozygous (W  ^f/+), but not in homozygous (W  ^f/W  ^f), mice (representative of two experiments in Table 1). In contrast, mast cell depletion in mice did not affect downregulation of local CHS responses by UVB irradiation, i.e., CHS responses after sensitization with hapten applied to the irradiated site (representative of two experiments in Table 1).

Of the 18 W  ^f/W  ^f mice administered bone marrow–derived mast cell precursors, only one was judged to be poorly reconstituted after 8 wk (6 wk + 2 wk CHS assay; Fig. 7,A). In the reconstituted mice, the mast cells were found nonuniformly distributed in the dorsal skin and the data summarized in Fig. 7 A comprise an average for 15 consecutive fields from the skin of each mouse, some with numerous mast cells and others with very few. There was no significant difference in the dorsal mast cell numbers of reconstituted mice that had been either UVB irradiated or nonirradiated 12 d before skin sampling.

Susceptibility to UVB (12 kJ/m²) for systemic suppression of CHS responses was evaluated in 13 mast cell–reconstituted W  ^f/W  ^f mice. The CHS response by all mice except one was significantly suppressed by UVB irradiation; this mouse was the one shown in Fig. 7,A to be poorly reconstituted and the results for this mouse have been excluded from the results of Fig. 7 B.

The evidence supporting the hypothesis that mast cells determine susceptibility to UVB-induced systemic immunosuppression is of two types. (a) In different strains of mice, susceptibility to UVB immunosuppression correlated with dermal mast cell prevalence and the prevalence in W  ^f/W  ^f mice was insufficient to mediate susceptibility. This established a correlation. (b) The ability of reconstitution of the dorsal skin of W  ^f/W  ^f mice with mast cells to confer susceptibility to UVB-induced systemic immunosuppression confirmed the functional role of mast cells in this response, and eliminated the concern that the lack of responsiveness of unreconstituted W  ^f/W  ^f mice to UVB might be the result of other abnormalities (anemia, albino nature) of these mice. The results with the albino BALB/c- Uvs1 mice also argued against the possibility that pigment was contributing in this study to the extent of susceptibility of mice to systemic immunomodulation by UVB. In a previous study, it was confirmed that Uvs1 was not genetically linked to 12 chromosomal markers including the pigment genes b (brown) and c (albino) (8).

By use of a combination of histamine types 1 and 2 receptor antagonists, we have previously shown that histamine is an important regulatory molecule in the pathway(s) by which UVB regulates systemic CHS responses in mice (14). The source of the histamine had not been confirmed. These studies implicate mast cell–derived histamine as an important contributor to a pathway that determines degree of susceptibility of mice to UVB-induced systemic, but not local, suppression of CHS responses. The histamine content of dermal mast cells as estimated using quantitative video image analysis did not vary significantly for different mouse strains. However, there was greater than a twofold increase in mast cell prevalence in the dermis of the C57BL/6J mice, i.e., UV highly susceptible mice, than in the dermis of the poorly susceptible BALB/c mice. In fact, there was a direct correlation between dermal mast cell prevalence and susceptibility to UVB-induced systemic suppression of CHS responses. Dermal mast cells in the BALB/c-Uvs1 mice were also increased to a level similar to that of DBA/2 mice (Fig. 1 A). The level of mast cell prevalence is consistent with the previously reported extent of suppression of CHS responses in BALB/c-Uvs1 mice with 4.5 kJ/m² UVB (8). Overall, these objective measures of histamine in stained sections suggest a correlation between total histamine that may be released from dermal mast cells after degranulation and the extent of systemic immunomodulation after UVB exposure.

That histamine was a component of a pathway defining degree of susceptibility of mice to UVB-induced systemic suppression of CHS responses was supported by further in vivo experiments. C57BL/6J and BALB/c mice differed in their susceptibilities to UVB for decreased CHS responses, but were similar in their susceptibilities to histamine-induced suppression of CHS responses. In addition, systemic CHS responses by mast cell–depleted mice (W  ^f/W  ^f) were not suppressed after UVB irradiation, but were susceptible to suppression by histamine. If keratinocytes were producing histamine after UV irradiation as recently reported in vitro (20), it was not sufficient to suppress the CHS response in W  ^f/W  ^f mice. We believe that we have administered physiological concentrations of histamine to the mice; we estimated that the dermis of 8 cm² dorsal skin (∼270 μm deep, ∼40 mast cells/mm² horizontal section of dermis, and 3–5 pg histamine/mast cell; reference 20) contains ∼8 μg histamine. This must be considered a crude estimation of available histamine, as not all histamine would be released from degranulating mast cells. Whether histamine was the only important mediator from mast cells was not established. A low, but significant, level of UVB-induced immunosuppression was still detected in the mice administered histamine receptor antagonists (14). The potential contribution of mast cell–derived TNF-α and IL-4 to UVB-associated immunomodulation is currently under investigation. However, it may be that there was insufficient histamine receptor antagonist available at histamine's site(s) of action to ensure total inhibition.

UVB was also unable to suppress DTH responses in the W  ^f/W  ^f mice and implicates mast cells and their products in the regulation of this other systemic response. In contrast, there was no evidence for mast cell involvement in the modulation by UVB of CHS responses when the sensitizing hapten was applied to the irradiated site. These results suggest that antigen-presenting cells at a distant, nonirradiated site, but not at the irradiated site, are susceptible to immunomodulation, directly or indirectly, by mast cell products.

It is uncertain whether UV triggers dermal mast cell degranulation directly or indirectly. Signals from sensory nerve fibres originating in the epidermis may be involved (21). Trans-urocanic acid (UCA) is a histidine-derived product in the stratum corneum of the skin (22) and has been implicated as a photoreceptor for UV-induced immunosuppression (23). There has been a report that cis-UCA, an immunosuppressive molecule and the product of UV-induced isomerization of trans-UCA, can stimulate mast cell degranulation (24). Mast cell degranulation induced by cis-UCA would support observations by us assaying systemic CHS responses (our unpublished data) and by others (25) of differences in response to cis-UCA by BALB/c and C57BL/6J mice, the latter being better responders to cis-UCA.

The mechanisms by which histamine alone, or together with other mast cell or keratinocyte products, signals systemic immunosuppression are not known. Histamine stimulates prostaglandin E₂ production by cultured human keratinocytes (12, 13); prostaglandin E₂ may act locally on epidermal cells (26), in lymph nodes (27), or systemically (28). Further, by indomethacin administration to mice, prostanoids have been shown to be responsible for at least 50% of the systemic suppression of CHS responses by UVB irradiation (13). There is also evidence that DNA is a UVB photoreceptor that initiates immunosuppression (29, 30). Damage to DNA can regulate the transcription in epidermal cells of regulatory cytokines (e.g., IL-10 [31], TNF-α [7, 32]) and adhesion molecules that enable cell–cell interaction (e.g., intercellular adhesion molecule 1 [ICAM-1]; reference 33). However, whether there is a relationship between histamine and other mast cell products, and cyclobutyl pyrimidine dimer formation in DNA of UVB-irradiated cells (30), and the enzymes responsible for dimer repair, is not known. With respect to UV-induced IL-10 production, it has been reported that UVB can suppress systemic CHS responses, but not DTH responses, in IL-10–deficient mice (34) and suggests that IL-10 may not be an important molecule in determining responses to UVB irradiation in W  ^f/ W  ^f mice. Histamine may directly influence the migration of immune cells into the UV-irradiated site, or to nearby lymph nodes, and regulate their proliferation and cytokine profiles (35–38).

In conclusion, our results show that mast cells contribute critically to the manifestation of UVB-induced systemic suppression of CHS and DTH responses. Further, they add to our previous suggestion that histamine is one important component of the pathway to immunosuppression (14) and indicates mast cells as the source of the histamine. Downregulation of systemic immune responses in mice after UVB exposure correlated directly with the prevalence of histamine-containing dermal mast cells. In addition, the strains of mice investigated were equivalently susceptible to the immunomodulatory effects of exogenous histamine, further emphasizing the relationship of dermal mast cell prevalence with susceptibility to UV systemic immunomodulation. The mechanism(s) by which histamine, and potentially other mast cell products, can modulate immune reactivity remains an unanswered question. The answers will be very important in our quest to understand and prevent or treat the harmful effects of UVB that are becoming of greater concern due to greater ozone depletion in the earth's stratosphere.

We are grateful to Mr. J. Brennan and Ms. K. Clark for performing the Csaba's mast cell stain, and to our colleagues for critical reading of the manuscript.

This work was supported by the Anti-Cancer Foundation of South Australia, the National Health and Medical Research Council of Australia, the Flinders Medical Centre Foundation (to P.H. Hart and J.J. Finlay-Jones) and in part by National Institutes of Health grant RO1 CA-53765 (to F.P. Noonan).

CHS

contact hypersensitivity

DTH

delayed type hypersensitivity

SCFT

subcutaneous fatty tissue

TNCB

2,4,6-trinitrochlorobenzene

UCA

urocanic acid

UVB

UV radiation of wavelength 280–320 nm