Divergence of contact hypersensitivity in vivo compared with hapten-specific lymphocyte proliferation and interferon-γ production in vitro following ultraviolet B irradiation: the possibility that UVB does not affect the sensitizing phase of contact hypersensitivity

Abstract

A hapten-specific lymphocyte proliferation assay, which measures the in vitro stimulation of DNA synthesis (as assessed by [³H]thymidine incorporation), was used to determine systemic immunization induced by an epicutaneously applied hapten in addition to the more commonly used method which measures ear (or footpad) swelling. 2,4-Dinitrofluorobenzene (DNFB) was painted on the shaved backs of C57BL/6 mice for two consecutive days after ultraviolet B (UVB) irradiation (at 1000 J/m²), and DNFB-sensitized lymph node cells (LNC) were obtained from the regional lymph nodes 4 days later. Although the ear swelling response (ESR) was suppressed by UVB radiation, as previously reported, analysis of LNC culture supernatants showed that the production of interferon-γ, a Tc1-type cytokine, was not inhibited by the UVB irradiation. In addition, contact dermatitis was induced (at levels similar to those of non-irradiated mice) by painting DNFB on the abdomen as a secondary response. We then examined the effect of UVB exposure alone on the ESR by injecting a mast cell degranulator, compound 48/80, 7 days after irradiation. Both the ESR and the percentage of degranulated mast cells were significantly reduced in UVB-irradiated mice. These results demonstrate that UVB irradiation does not affect the sensitizing phase of contact hypersensitivity, but modulates the elicitation phase and reduces the ESR primarily by suppressing the degranulation of mast cells. Therefore, suppression of the ESR alone cannot always be considered as hapten-specific immunotolerance.

Introduction

It has long been considered that ultraviolet B (UVB) exposure induces hapten-specific immunotolerance locally as well as systemically, as contact hypersensitivity (CHS) reactions are suppressed after UVB irradiation.^1–5 However, the induction of tolerance after UVB irradiation has been proven mostly by measuring the suppression of ear or footpad swelling in mice.

Langerhans cells (LC) are largely responsible for the local immunosuppression observed after UVB irradiation in vivo, including the reduction of CHS and the induction of tolerance. It has been well documented that the number of LC in the epidermis decreases after UVB irradiation.¹ The LC that remain in the epidermis after UVB irradiation show abnormal morphology, appearing rounded and swollen, and are deficient in their dendritic processes.⁶ UVB also induces Thy-1⁺ dendritic cells, which are of bone marrow origin, and which downregulate the CHS response,³ i.e. the induction of suppressor T cells.⁷

However, allergic contact dermatitis consists of two phases: the sensitizing phase, in which immunological memory is induced by hapten application; and the elicitation phase, in which an epicutaneously painted hapten induces a hapten-specific inflammatory reaction of the skin. Those results, observed after UVB irradiation, suggest that the information about an epicutaneously applied hapten does not reach the draining lymph nodes nor does it induce memory T cells, but that it does induce suppressor T cells; therefore, the effect of UVB irradiation has been recognized to affect the sensitizing phase of CHS.

On the other hand, a clear explanation of the elicitation phase of CHS is still not well understood. Askenase & Van Loverent have reported that the activation of mast cells by antigen-specific T-cell-derived factors is necessary for the elicitation of delayed-type hypersensitivity (DTH).⁸ UVB irradiation at a low dose is not cytotoxic to mast cells because the number of mast cells and their degree of degranulation in ear skin remain unchanged.⁹ However, when mice are given an intradermal injection of a mast cell degranulator, such as compound 48/80, both the ear swelling response (ESR) and mast cell degranulation are significantly suppressed when the mice are pre-exposed to UVB (250–1000 J/m²).⁹ As DTH responses in mice are elicited preferentially at cutaneous sites enriched in mast cells, such as ears and footpads, and because defective DTH responses are elicited in two different strains of mast cell-deficient mice (W/W^V and Sl/Sl^d),⁸ the suppression of the ESR alone cannot always be considered as hapten-specific T-cell-induced immunotolerance.

To determine the effect of systemic immunization induced by an epicutaneously applied hapten, we performed a lymphocyte proliferation assay on cells obtained from the draining lymph nodes of UVB-irradiated mice.

Materials and methods

Mice

Female C57BL/6 (B6) NJcl mice were purchased from Nippon CREA Co. (Tokyo, Japan) and used at 6–8 weeks of age. Experimental procedures performed during UVB irradiation, and the measurement of ear swelling, were carried out under general anaesthesia, achieved by intraperitoneal (i.p.) injection of ketamine. Each control or experimental group consisted of four or five mice. Each experimental protocol was repeated at least three times, with similar results obtained on each occasion.

Haptens

2,4-Dinitrofluorobenzene (DNFB; C₆H₃FN₂O₄; MW 186·1), fluorobenzene (FB; C₆H₅F; MW 96·1) and oxazolone (Ox; 4-ethoxymethylene-2-phenyl-2-oxazolin-5-one; C₁₂H₁₁NO_3; MW 217·2) were purchased from Sigma Chemical Co. (St Louis, MO), while picryl chloride (PCl; 2,4,6-trinitrochlorobenzene; (O₂N)₃C₆H₂Cl; MW 247·6) was purchased from Tokyo Kasei Co. (Tokyo, Japan).

Culture medium

RPMI-1640 was supplemented with 5 × 10⁻⁵ m 2-mercaptoethanol, 100 µg/ml streptomycin, 100 U/ml penicillin, 1 mm l-glutamine and 10% heat-inactivated fetal calf serum (FCS) (Grand Island Biological Co., Grand Island, NY) and used as a culture medium.

Radiation

Mice were shaved on their dorsa (2 × 3 cm in size) before UVB irradiation, and both ears of each mouse were covered with black opaque tape during irradiation. Two days prior to DNFB sensitization (day 0), mice were exposed to UVB (1000 J/m²) from a bank of four FS-2000 fluorescent lamps (Dermaray^R; Torex, Tokyo, Japan). These bulbs have a broad emission spectrum (250–400 nm), and a high output primarily in the UVB range (290–320 nm, peak 305 nm). As measured using an IL 700 radiometer with a SEE240 UVB photodetector, these lamps delivered an average flux of 0·9 J/m²/second. A UV Radiometer, UVR-3036/S (Topcon, Tokyo, Japan), was used to measure UVB output each time before exposure.

Induction of CHS

Sensitization and elicitation of CHS was performed according to the method of Phanuphak et al.¹⁰ Briefly, 2 days after UVB irradiation (day 0), the dry, shaved backs of mice were painted with 25 µl of 0·4% DNFB in acetone : olive oil (at a ratio of 4 : 1) twice (on days 2 and 3).^10,11 Vehicle (acetone/olive oil)-painted mice were used as controls.

Preparation of lymph node cells (LNC)

Mice were killed 4 days after the second DNFB application (on day 7). At the same time, gross lesions and histology of the dorsal skins of the mice were examined. A single cell suspension of draining LNC was prepared under aseptic conditions by mechanical disaggregation, and was passed through sterile nylon gauze. The cells were washed three times and then resuspended in RPMI-1640 culture medium. Viable cell counts were performed using the trypan blue dye-exclusion method.

Hapten pulsed antigen-presenting cells (APC)

A single-cell suspension of normal B6 spleen cells was prepared under aseptic conditions by mechanical disaggregation and passed through sterile nylon gauze. Erythrocytes were lysed with red cell lysis solution containing 0·2% saponin (Sigma), and the remaining cells were washed three times with Hanks' balanced salt solution (HBSS; Gibco, Grand Island, NY). The cells were resuspended in culture medium and irradiated at 3000 rads. The irradiated spleen cells were then incubated for 10 min at 37° with each hapten, after which they were washed three times with HBSS to remove free haptens and then used as APC.

Lymphocyte proliferation assay

The mice were killed on day 7, and 4 × 10⁵ LNC obtained from the regional lymph nodes were co-cultured with 2 × 10⁵ various hapten-pulsed APCs or with 2 µg of concanavalin A (Con A; Sigma) in 96-well flat-bottomed microculture plates (Falcon, Xxxxxxxxxx, Xxxxxxxx). Cultures were maintained in 95% air and 5% CO₂ at 37° for 4 days. Twenty-four hours before harvesting, 1 µCi (37 kBq) of [³H]thymidine ([³H]TdR) was added to each well. Incorporation of [³H]TdR in each well was determined the next day using a liquid scintillation counter. All cultures were established in triplicate and the results are expressed as the mean number of counts per minute [c.p.m.; standard error of the mean (SEM) <10%].

Cytokine analysis by enzyme-linked immunosorbent assay (ELISA)

At the same time as the proliferation assay, culture supernatants of regional LNC and hapten-pulsed APC were harvested in 96-well flat-bottomed microculture plates (day 10), and stored at −80° until use. ELISAs for interleukin (IL)-2, IL-4, IL-10 and interferon-γ (IFN-γ) were performed using sandwich ELISA kits purchased from Endogen (Woburn, MA). All protocols were performed according to the manufacturer's instructions. Samples were assayed in duplicate and the results are expressed as the mean concentrations (SEM < 10%). The sensitivities for these assays are as follows: IL-2, <3 pg/ml; IL-4, <5 pg/ml; IL-10, <12 pg/ml; and IFN-γ, <10 pg/ml.

Elicitation of secondary response

Five days after the primary immunization on the dorsal skin, the right ear of each mouse was challenged with 20 µl of 0·2% DNFB (10 µl on each side of the ear). Twenty-four hours later, the ear thickness of each mouse was measured using a micrometer (Mitutoyo, Kawasaki, Japan). In addition, to determine the secondary response, the shaved abdomens of DNFB-sensitized mice were painted with 25 µl of 0·2% DNFB for two consecutive days (days 7 and 8) and, 48 hr later, the mice were killed and skin specimens from their abdomens were fixed in buffered formalin, embedded in paraffin and stained with haematoxylin and eosin.

Effect of compound 48/80 on ESR and mast cell degranulation

Seven days after exposure to UVB, at the same time as the ear challenge with 0·2% DNFB described above, the mice were given an intradermal injection of 20 µl of 2·5 mg/ml compound 48/80, dissolved in phosphate-buffered saline (PBS), pH 7·4 (Gibco), into the dorsal side of the ear. Ten mice were used in each group. Ear thickness was measured before and 45 min after the compound injection, as previously reported.⁹ The interval of 45 min was chosen because the mechanical swelling induced by a saline injection had returned nearly to the normal level by that time. The ESR represents the mean data of an increase in ear thickness above the baseline value. Ears from UVB-irradiated mice, from non-irradiated normal controls, from compound 48/80-challenged mice and from 0·2% DNFB-challenged mice, were excised, fixed in buffered formalin, embedded in paraffin and stained with toluidine blue. Each group consisted of five mice. Mast cells were counted at ×400 magnification by two individuals, each blinded to the experimental conditions, and were classified as degranulated when eight or more granules were found outside the cell membrane, according to the method of Maurer et al.¹² The percentage degranulation was calculated as follows:

Statistical analysis

The statistical significance of differences between the means of each experimental group was determined using analysis of variance. Mean differences were considered to be significant when the P-value was <0·01. Statistical analyses were performed using StatView 4·5.

Results

UVB irradiation suppresses local inflammatory changes at the irradiated site

Many studies have demonstrated that UVB irradiation not only inhibits contact dermatitis when a hapten is applied at the irradiated site but also induces hapten-specific tolerance in UVB-susceptible strains of mice (such as B6), measured as the ESR.^1–7

In preliminary experiments, we exposed the shaved backs of normal mice to different doses of UVB in order to establish the maximum dose of UVB irradiation that does not influence the skin macroscopically or histopathologically. Several doses of UVB (range 400–1400 J/m²) were used to irradiate the dry, shaved backs of those mice. There were no significant changes in gross lesions at UVB doses up to 1000 J/m². However, when ≥ 1200 J/m² UVB was used, moderate-to-severe skin ulcers were observed on the backs of the mice (data not shown). Therefore, we used an irradiation dose of 1000 J/m² throughout this study.

The minimum concentration of DNFB solution which induced contact dermatitis when applied for two consecutive days, was also established. Two DNFB applications on the dorsa at concentrations of ≥ 0·4% caused moderate-to-severe contact dermatitis (data not shown). Therefore, we chose to use 0·4% DNFB for this study. Each control or experimental group consisted of four or five mice. Each experimental protocol was repeated at least three times, with similar results obtained on each occasion. Representative data are shown in each figure.

Figure 1 demonstrates the gross skin appearance and a haematoxylin and eosin-stained section of each group. There were no significant changes between the UVB-irradiated mice or vehicle only-treated mice (Fig. 1b), compared with untreated mice (Fig. 1a). Therefore, irradiation with UVB alone does not have a discernible influence. In contrast, severe contact dermatitis with crust formation was induced by painting a 0·4% DNFB solution on the dorsa of non-irradiated mice. A thickened epidermis and a severe lymphoid cell infiltration in the upper dermis was observed (Fig. 1c). However, the UVB-irradiated mice did not develop those changes, even when 0·4% DNFB was applied for two consecutive days (Fig. 1d). No significant dermatitis was observed in those mice, either macroscopically or histopathologically, as in the normal controls. Therefore, as previously reported, UVB irradiation suppressed local inflammatory changes at the irradiated site.

Figure 1.

Macroscopic and histopathological findings of the shaved dorsa of mice (haematoxylin & eosin ×40). Each control or experimental group consisted of four or five mice. Each experimental protocol was repeated at least three times (with similar results obtained on each occasion), and representative data for each experimental group are described in the figure. (a) The shaved back of a non-irradiated mouse, as a normal control. (b) The shaved back of a vehicle (acetone : olive oil; 4 : 1)-treated mouse 2 days after exposure to ultraviolet B (UVB) at 1000 J/m². No significant changes were found compared with the unirradiated control. (c) A non-irradiated mouse painted with 25 µl of 0·4% 2,4-dinitrofluorobenzene (DNFB) solution for 2 days developed severe contact dermatitis with crust formation on the back. A thickened epidermis and a severe lymphoid cell infiltration in the upper dermis were observed. (d) A UVB-irradiated mouse did not develop significant dermatitis, even when 0·4% DNFB was painted for two consecutive days.

UVB exposure does not affect the production of hapten-specific LNC in vitro, as demonstrated by the lymphocyte proliferation assay

To determine whether systemic immunization might be induced by the epicutaneously applied hapten, we measured lymphocyte proliferation of LNC obtained from the draining lymph nodes of UVB-irradiated mice, in addition to measuring the ESR. UVB-irradiated (day 0) and non-irradiated mice were sensitized by painting 0·4% DNFB on their shaved dorsa twice after UVB irradiation (days 2 and 3). Mice were killed on day 7, and 4 × 10⁵ LNC from their regional lymph nodes were cultured with 2 × 10⁵ APC pulsed with various haptens, and proliferative responses were measured 4 days later (Fig. 2).

Figure 2.

No effect of ultraviolet B (UVB) radiation on the proliferative response of lymph node cells (LNC). 2,4-Dinitrofluorobenzene (DNFB) (0·4%) was painted for two consecutive days on the shaved dorsa of B6 mice after UVB irradiation. LNC (4 × 10⁵) from regional lymph nodes were cultured with 2 × 10⁵ antigen-presenting cells (APC) pulsed with various haptens, and proliferative responses were measured 4 days later. The data reported represent the mean counts per minute (c.p.m.) of triplicate cultures [standard error of the mean (SEM) < 10%]. LNC from UVB-irradiated or from non-irradiated mice showed strong proliferative responses only when co-cultured with DNFB and were statistically significant. *P < 0·0001 when using a one-way factorial analysis of variance (anova). FB, fluorobenzene; Ox, 4-ethoxymethylene-2-phenyl-2-oxazolin-5-one; PCl, picryl chloride (2,4,6-trinitrochlorobenzene).

Surprisingly, LNC from the UVB-irradiated mice showed strong proliferative responses when co-cultured with DNFB-pulsed APC, with the responses equal to or even higher than the levels observed in non-irradiated mice. In contrast, APC pulsed with other haptens, such as FB, Ox or PCl, did not stimulate LNC proliferation. Therefore, a strong hapten-specific proliferation was produced, which was statistically significant (P < 0·0001). Hapten-specific LNC were observed at levels similar to or higher than those obtained in non-irradiated mice.

However, the possibility remained that DNFB was merely stimulatory and induced the strong lymphocyte proliferation non-specifically. To rule out that possibility, normal LNC were cultured with 2 × 10⁵ DNFB-pulsed APC. No proliferative responses were observed, although these normal LNC showed strong proliferation when cultured with 2 µg/ml Con A (data not shown). These results demonstrate that epicutaneously applied DNFB, at sites irrespective of UVB irradiation, induces hapten-specific lymphocytes in the regional lymph nodes.

No suppressive effect of UVB irradiation on IFN-γ production

The most reasonable explanation for the hapten-specific tolerance after UVB irradiation was the induction of suppressor T cells or antigen-specific tolerance-inducing cells. It has also been reported that IFN-γ is essential for eliciting the CHS.^13,14 To determine whether LNC obtained after hapten application at a UVB-irradiated site can produce IFN-γ, we analysed cytokine production in the culture supernatant.

A total of 4 × 10⁵ LNC from regional lymph nodes were cultured with various hapten-pulsed APCs and, 72 hr later, the culture supernatants were harvested and the concentrations of IFN-γ and other cytokines (such as IL-2, IL-4 and IL-10) analysed by ELISA. Surprisingly, the production of IFN-γ, reported to be essential for eliciting the CHS,^13,14 was not inhibited by UVB irradiation. In fact, the level of IFN-γ was similar to or even higher in UVB-irradiated mice compared with non-irradiated mice (Table 1), and the level of IL-10, reported to be essential for the induction of tolerance, was not elevated. Levels of IL-2 and IL-4 were similar to those of normal controls. These results demonstrate that the production of IFN-γ, type 1 cytotoxic cells (Tc1), in regional lymph nodes was not suppressed by UVB irradiation. If the above condition occurs in vivo, the contact dermatitis should be induced as a secondary response, independent of the UVB irradiation.

Table 1.

No suppressive effect of ultraviolet B (UVB) irradiation on interferon-γ (IFN-γ) production

		UVB(+)			UVB(−)
	non	DNFB	Ox	non	DNFB	Ox
IFN-γ	140	4746	35	0	2740	80
IL-2	80	82	60	75	90	85
IL-4	0	0	0	0	0	0
IL-10	0	0	0	0	0	0

The regional lymph node cells (LNC) were cultured with hapten-pulsed antigen-presenting cells (APC). After 3 days, culture supernatants were harvested and the concentrations of interleukin (IL)-2, -4 and -10, and IFN-γ, were measured by enzyme-linked immunosorbent assay (ELISA). The lowest detection levels of these cytokines are as follows: IL-2, <3pg/ml; IL-4, <5 pg/ml; IL-10, <12 pg/ml; and IFN-γ, <10 pg/ml.

Induction of contact dermatitis on the abdomen as a secondary response

To determine the secondary response using a method other than the ESR, the shaved abdomens of DNFB-sensitized mice were challenged with 25 µl 0·2% DNFB for two consecutive days (days 7 and 8). Forty-eight hours later, those mice were killed and the skin of their abdomens excised. In preliminary experiments, we determined that this amount of DNFB is the maximum dose which induces contact dermatitis as a secondary response, but does not induce contact dermatitis in unsensitized mice. The shaved abdomens of vehicle-painted mice served as normal controls.

Compared with normal controls, non-sensitized mice did not develop dermatitis (as determined both macroscopically and histopathologically) when 0·2% DNFB was applied on their shaved abdomens (Fig. 3a). Neither UVB-irradiated nor vehicle-painted mice developed dermatitis when challenged with 0·2% DNFB (Fig. 3b) Therefore, irradiation with UVB alone does not have a discernible influence on this protocol. However, non-irradiated 0·4% DNFB-sensitized mice developed moderate contact dermatitis on their abdomens 48 hr later, when challenged with 0·2% DNFB. Moderate lymphoid cell infiltrations were observed throughout the dermis to the subcutaneous fat (Fig. 3c). Surprisingly, UVB-irradiated 0·4% DNFB-sensitized mice also developed moderate contact dermatitis compared with non-irradiated mice, following challenge with 0·2% DNFB on their abdomens (Fig. 3d). In contrast to the ESR results, allergic contact dermatitis was induced equally on the abdomens of mice as a secondary response, whether or not they were UVB irradiated.

Figure 3.

Induction of contact dermatitis on the abdomen as a secondary response (haematoxylin & eosin ×40). Mice were immunized with 25 µl of 0·4% 2,4-dinitrofluorobenzene (DNFB) for two consecutive days on their shaved backs after ultraviolet B (UVB) irradiation. Seven days after UVB irradiation, the shaved abdomens of mice were challenged with 25 µl of 0·2% DNFB for two consecutive days (days 7 and 8). Forty-eight hours later, mice were killed and the skin of their abdomens excised. The shaved abdomens of vehicle-painted mice served as normal controls. Each control or experimental group consisted of four or five mice. Each experimental protocol was repeated at least three times, with similar results obtained on each occasion. (a) The shaved abdomen of a non-immunized and 0·2% DNFB-applied mouse. Compared with a normal control, no significant changes (as evidenced either macroscopically or histopathologically) were induced by application of 0·2% DNFB. (b) The shaved abdomen of a UVB-irradiated, vehicle-sensitized and 0·2% DNFB-challenged mouse. There were no significant changes observed compared with non-irradiated mice. (c) A non-irradiated 0·4% DNFB-sensitized mouse developed moderate contact dermatitis on the abdomen when challenged with 0·2% DNFB. A moderate lymphoid cell infiltration was observed throughout the dermis to the subcutaneous fat. (d) Similarly to the non-irradiated mouse, a UVB-irradiated 0·4% DNFB-sensitized mouse also developed moderate contact dermatitis after painting of 0·2% DNFB on its abdomen.

These results demonstrate that, at least in our system, UVB irradiation suppresses local inflammatory changes at the irradiated site, but does not inhibit the production of hapten-specific LNC, which cause allergic contact dermatitis on the abdomen. In other words, UVB does not affect the sensitization phase of CHS. It was therefore of interest to determine whether UVB irradiation itself can suppress the ESR.

Suppression of the DNFB-induced ESR by UVB irradiation

There are many reports^1–7 that UVB irradiation inhibits systemic immunization, and such inhibition was usually demonstrated by a diminished ESR after hapten application. Hence, we next examined the ESR to assess the state of systemic immunization.

Five days after sensitization with DNFB, the right ear of each mouse was challenged with 20 µl of 0·2% DNFB, and the ESR was measured 24 hr later. The negative control refers to mice that were challenged but had not been sensitized to DNFB (Fig. 4). The ESR of UVB-irradiated mice was significantly suppressed compared with that of non-irradiated mice, as previously reported,^1,4,5 but the fact that the allergic contact dermatitis was induced on the abdomen to a similar extent in both groups, suggests that UVB does not affect the sensitizing phase of CHS. Therefore, we focused on characterizing the elicitation phase, especially differences between the skin of the abdomen and the ears.

Figure 4.

Suppression of the ear swelling response (ESR) by ultraviolet B (UVB) irradiation. Groups of B6 mice were exposed to UVB (at 1000 J/m²) on their shaved dorsal skin. The ears were covered with black tape during the irradiation. Two days later, 0·4% 2,4-dinitrofluorobenzene (DNFB) was painted on the irradiated skin for two consecutive days. The right ears of the mice were challenged on day 7 with 0·2% DNFB and the ESR was measured 24 hr later. The negative control refers to mice that were challenged but not immunized. n = 15 mice/group; *P < 0·01 (unpaired Student's t-test) compared with the ESR of non-irradiated, DNFB-sensitized mice.

Suppression of compound 48/80-evoked ESR by UVB irradiation

DTH responses in mice are elicited preferentially at cutaneous sites enriched in mast cells, such as ears and footpads. Mast cell-deficient mice are reported to be unable to elicit DTH responses. Haptens evoke the ESR in systemically immunized mice and this response is achieved primarily by the degranulation of mast cells. Compound 48/80 is well known as a mast cell degranulator and it evokes cutaneous swelling at sites rich in mast cells, such as the ears and footpads. UVB irradiation at a low dose is not cytotoxic to mast cells because the number of mast cells and their degree of degranulation in the ear skin remain unchanged.⁹ Therefore, it is possible that the suppressive effect of UVB irradiation on the ESR is a result of the inhibition of mast cell degranulation. It is surprising that UVB irradiation suppressed the ESR, even though the ears were protected by black tape during irradiation.

To determine the effect of UVB exposure alone on the ESR, mice were given an intradermal injection of 20 µl of 2·5 mg/ml compound 48/80 into the ears, 7 days after exposure to UVB, at the same time as the ear challenge with 0·2% DNFB. Ear thickness was measured before and 45 min after injection of compound 48/80, and the ESR represents the mean increase in ear thickness above the baseline value. The ESR was significantly suppressed in UVB-irradiated mice compared with non-irradiated mice, at a statistically significant level (Fig. 5). This fact demonstrates that UVB exposure alone suppresses the mast cell-related ESR of B6 mice spontaneously, at least on day 7.

Figure 5.

Suppression of the compound 48/80-evoked ear swelling response (ESR). Seven days after ultraviolet B (UVB) exposure, mice were given an intradermal injection of 20 µl of compound 48/80 (a mast cell degranulator) into their ears. Ear thickness was measured before and 45 min after the injection. The ESR represents the mean increase in ear thickness above the baseline value. The ESR was significantly suppressed in UVB-irradiated mice compared with non-irradiated mice. UVB exposure itself suppresses the ESR evoked by injection of compound 48/80 into B6 mice, at least until day 7 after irradiation. n = 10 mice/group; *P < 0·01 (unpaired Student's t-test) compared with the ESR of non-irradiated mice.

Suppression of compound 48/80-evoked mast cell degranulation after UVB irradiation

To determine the direct effect of UVB on mast cells, the ears of UVB irradiated–compound 48/80-challenged mice were excised and stained with toluidine blue (Fig. 6). The ears of mice were covered with opaque tape during UVB exposure (at 1000 J/m²) of the shaved dorsa. Seven days after the irradiation, mice were given an intradermal injection of 20 µl of 2·5 mg/ml compound 48/80 into one side of their ears. The ears of non-irradiated mice served as normal controls (Fig. 6a). UVB irradiation itself did not show significant changes compared with normal controls (Fig. 6b). Non-irradiated mice showed a strong ESR when compound 48/80 was injected (Fig. 6c). UVB-irradiated–compound 48/80-injected mice developed less ear swelling than non-irradiated mice (Fig. 6d). The ears of 0·2% DNFB-challenged mice were also removed surgically and were stained with toluidine blue at the same time. The percentage of degranulated mast cells was counted at ×400 magnification by two individuals, each blinded to the experimental conditions (Fig. 7). UVB irradiation itself did not influence the degree of mast cell degranulation. However, compound 48/80-evoked mast cell degranulation was significantly suppressed in UVB-irradiated mice compared with non-irradiated mice at a statistically significant level (Fig. 7). Mice challenged with 0·2% DNFB also tended to show similar results. This fact demonstrates the possibility that UVB exposure itself suppresses the mast cell degranulation of B6 mice spontaneously, at least on day 7, even though their ears were covered with black tape.

Figure 6.

Effect of ultraviolet B (UVB) irradiation on the compound 48/80-evoked ear swelling response (ESR) (toluidine blue staining ×40). To determine the effect of UVB irradiation on mast cells, the ears of UVB-irradiated, compound 48/80-challenged mice were excised and stained with toluidine blue. The ears of mice were covered with opaque tape during UVB exposure at 1000 J/m² on the shaved dorsa. Seven days after the irradiation, mice were given an intradermal injection of 20 µl of 2·5 mg/ml compound 48/80 into one side of their ears. Each control or experimental group consisted of four or five mice. Each experimental protocol was repeated at least three times (with similar results obtained on each occasion), and representative data for each experimental group are described in the figure. (a) The ear of a non-irradiated mouse, as a normal control. (b) Seven days after the irradiation. No significant changes were found compared with a normal mouse. (c) A non-irradiated mouse developed a strong ESR when compound 48/80 was injected into the dorsal side of the ear. Note the degranulated mast cell in the high-power view. (d) A UVB-irradiated, compound 48/80-injected mouse developed less ESR than a non-irradiated mouse that received compound 48/80 injection. Note the non-activated mast cell in the high-power view.

Figure 7.

Suppressive effect of ultraviolet B (UVB) irradiation on mast cell degranulation. The ears of UVB-irradiated mice, non-irradiated normal controls, compound 48/80-challenged mice, and 0·2% 2,4-dinitrofluorobenzene (DNFB)-challenged mice were surgically removed and stained with toluidine blue. Mast cells were counted at ×400 magnification and were classified as degranulated when eight or more granules were found outside the cell membrane, according to the method of Maurer et al.¹² The percentage degranulation was calculated as follows: % degranulation = [(number of degranulated mast cells) ÷ (number of total mast cells) × 100]. UVB irradiation itself did not influence the degree of mast cell degranulation. However, compound 48/80-evoked mast cell degranulation was significantly suppressed in UVB-irradiated mice compared with non-irradiated mice at a statistically significant level. n = 5 mice/group; *P < 0·01, **P < 0·05 (unpaired Student's t-test) compared with non-irradiated mice.

Discussion

It is well documented that ultraviolet radiation, especially UVB, has a suppressive effect on the immune system of the skin.¹ Several experimental models have clearly demonstrated that exposure to UVB in vivo inhibits CHS reactions and induces hapten-specific tolerance.^1,4,5 UVB also induces the expression of Thy-1⁺ dendritic cells, which are of bone marrow origin, and downregulates the CHS response,³ i.e. it induces suppressor T cells.⁷

These results suggest that information of an epicutaneously applied hapten does not reach the draining lymph nodes nor does it induce memory T cells, but it does induce suppressor T cells, and thus the effect of UVB irradiation is primarily in the sensitizing phase of CHS. However, the induction of tolerance after UVB irradiation has been proven by the suppression of the ESR or of foodpad swelling, and not by the complete inhibition of those reactions. Measurements of the ESR are sometimes irregular, and can differ with each measurement.

We performed a lymphocyte proliferation assay to measure the systemic immunization induced by an epicutaneously applied hapten, because we suspected that the suppression of the ESR alone cannot be considered as hapten-specific immunotolerance. In our results described above, DNFB-sensitized LNC were obtained from the regional lymph nodes of UVB-irradiated mice at levels equal to or higher than those obtained from the regional lymph nodes of non-irradiated mice. This fact suggests that UVB exposure does not affect the production of hapten-specific LNC in vivo, i.e. it does not affect the sensitizing phase of CHS.

Therefore, the role of various cytokines in the elicitation phase of the CHS remains unclear. Among the T-cell-derived cytokines, both type 1 and type 2 are found at sites of CHS responses, whereas type 1 and Tc1 cytokines, such as IFN-γ and IL-12, are predominant in the course of a CHS response.^13–15 Furthermore, the type of immune response appears to depend upon the specific immunogen used. For instance, DNFB induces a Tc1 cytokine response, whereas fluorescein isothiocyanate induces type 2 cytokines.¹⁶ Cutaneous reactions at DNFB-painted sites of B6 mice induce CD4⁺ T helper (Th)1 and CD8⁺ Tc1 cells, which are essential for the full development of CHS or regulatory CD4⁺ Th2 cells.¹³ As injection of IL-10 inhibits elicitation of the CHS, a negative regulatory effect of IL-10 on the CHS appears to be well established.^7,17

In this study, the concentrations of IL-2, IL-4, IL-10 and IFN-γ in LNC culture supernatants were measured by ELISA. Only IFN-γ showed a strong increase when co-cultured with DNFB-pulsed APC, resulting in similar or higher levels in UVB-irradiated mice. Furthermore, the level of IL-10, reported to be essential for the induction of tolerance, was not increased in the regional lymph nodes of UVB-irradiated mice. This fact demonstrates that the production of IFN-γ, a Tc1-type cytokine reported to be the cytokine most essential for the production of CHS in the regional lymph nodes, was not inhibited or suppressed by UVB irradiation in this study.

Furthermore, even though the ESR was suppressed after UVB irradiation, allergic contact dermatitis was induced to a similar extent on the abdomens of mice after challenge with 0·2% DNFB as a secondary response, regardless of whether or not they were UVB irradiated. This suggests that DNFB-specific LNC obtained from regional lymph nodes of UVB-irradiated mice are not all tolerance-inducing cells or suppressor T cells. It is safe to assume that they contain some effector or memory T cells, as the allergic contact dermatitis was induced, even in UVB-irradiated mice. If UVB irradiation, especially with respect to local immunosuppression, affects the sensitization phase of CHS, as previously reported,^1,4,5 the contact dermatitis should not be induced as a secondary response in vivo.

However, some reports state that LC are not necessarily required for T-cell sensitization,^4,6 and therefore other APCs, such as dermal dendritic cells, may also contribute to the priming of naïve T cells after epicutaneous contact with a hapten. Therefore, information about an epicutaneously applied hapten is delivered somehow to the draining lymph nodes where it induces hapten-specific T cells, even after UVB irradiation, i.e. the sensitizing phase of the CHS is not affected by UVB exposure.

Then why did those previous reports obtain results different from ours? Most studies on the pathophysiology of CHS have focused on the sensitizing phase, and little is known about the elicitation phase. However, the ESR shows a peak level at 24 hr against allergic contact dermatitis, such as patch testing, which is mediated by allergen-specific lymphocytes, usually induced 48 hr after elicitation.

We then considered differences between the skin of the ear and of the abdomen. The key parameter was the population of mast cells in those tissues, and the ears and footpads of mice are known to be enriched in mast cells.⁸ We initially focused on the report that activation of mast cells by antigen-specific T-cell-derived factors is necessary to elicit DTH.⁸ The earliest histopathological findings during a CHS response are mast cell degranulation, vasodilation and the influx of neutrophils, followed by infiltration of mononuclear and T cells.¹⁵

Activation of mast cells by antigen-specific T-cell factors initiates a cascade of cellular interactions.⁸ Three experimental approaches have been performed that focus on mast cells:⁸

1. DTH responses in mice are elicited preferentially at cutaneous sites enriched in mast cells, such as the ears and footpads.⁸

2. Defective DTH responses are elicited in two different strains of mast-cell-deficient mice (W/W^V and Sl/Sl^d) that have independent genetic defects which lead to their mast cell deficiency.⁸

3. DTH responses are inhibited by treatment with the cromolyn-like drug, proxicromil, which prevents the release of mediators by mast cells.⁸

In other words, UVB exposure influences the ESR, a tissue where mast cells are numerous.

Therefore, we hypothesized that UVB irradiation affects mainly the elicitation phase of CHS by modulating mast cell activation. UVB irradiation at a low dose was, alone, not cytotoxic to mast cells because the numbers of mast cells and their degrees of degranulation in the ear skin were unchanged.⁹

The ESR induced by intradermal injection of the mast cell degranulator, compound 48/80, was significantly suppressed in UVB-irradiated mice compared with non-irradiated mice. The percentage of mast cell degranulation was also suppressed in UVB-irradiated mice. This fact demonstrates the possibility that UVB exposure suppresses the ESR of B6 mice spontaneously, at least on day 7. In other words, the hapten-specific ESR might also be influenced by suppression of mast cells induced by antigen-specific T-cell derived factors after UVB irradiation, even though the ears were protected by black tape during irradiation.

Therefore, in this study, we demonstrated that UVB exposure suppresses the ESR and local inflammation at the irradiated site, but that it does not affect the production of hapten-specific LNC and IFN-γ, a Tc1-type cytokine, in the regional lymph nodes. In addition, contact dermatitis was induced by painting DNFB on the abdomen as a secondary response, at levels equal to those observed in non-irradiated mice. Furthermore, compound 48/80 evoked the ESR, and the number of degranulated mast cells was significantly reduced in UVB-irradiated B6 mice, even 7 days after irradiation. These results demonstrate that UVB irradiation does not affect the sensitizing phase of CHS, but modulates the elicitation phase primarily by suppressing the degranulation of mast cells. Hence, the suppression of ESR alone cannot always be considered as hapten-specific immunotolerance.

Abbreviations

APC

antigen-presenting cells

B6

C57BL/6

CHS

contact hypersensitivity

Con A

concanavalin A

DNFB

2,4-dinitrofluorobenzene

DTH

delayed-type hypersensitivity

ELISA

enzyme-linked immunosorbent assay

ESR

ear swelling response

FB

fluorobenzene

FCS

fetal calf serum

HBSS

Hank's balanced salt solution

LC

Langerhans cells

LNC

lymph node cells

Ox

4-ethoxymethylene-2-phenyl-2-oxazolin-5-one

PCl

picryl chloride (2,4,6-trinitrochlorobenzene)

UVB

ultraviolet B