Evaluation of auricular lymph node cell lymphocyte proliferation and cytokine production as non-radioactive endpoints during murine contact allergy

Abstract

The murine local lymph node assay (LLNA) has been developed as a test method to assess allergic contact dermatitis. In spite of the validity of the LLNA, attention was drawn to the two disadvantages: use of radioactivity for in vivo measurement of lymph node cell proliferation ([³H]-thymidine labeling) and the possibility of false positive results caused by non-specific cell activation as a result of inflammatory processes in the skin (irritation). We aimed to investigate the following non-radioactive endpoints of LLNA: 5-bromo-2′-deoxyuridine (BrdU) incorporation ex vivo and in vivo, in vivo and ex vivo cytokine production with or without phytohemagglutinin (PHA) stimulation. Here, 8-12-week-old female BALB/c mice were treated topically with the strong sensitizer 2,4-dinitrochlorobenzene (DNCB) in acetone:olive oil (AOO, 4:1 [v/v]) at levels of 0.025, 0.05, 0.01, or 0.25% (w/v). Ear thickness was also measured to determine the differentiation index (DI) indicating the proportion of non-specific activation due to irritating properties of test compound. At the concentration of 0.05%, stimulation index (SI) value was found to be 3 for DNCB based on in vivo and ex vivo BrdU incorporation. The results of the in vivo and ex vivo non-radioactive LLNA assays were compatible both with each other and with previous radioactive LLNA data. Our results indicate that non-radioactive endpoints may be used as an alternative to the [³H]-thymidine LLNA. The levels of T_H1 cytokines (IL-2 and IFNγ) and T_H2 cytokines (IL-4 and IL-5) in lymph node cell cultures were significantly (P < 0.01) increased when DNCB was applied at the concentrations of 0.05 and 0.1%, respectively. As the DI was > 1, the applied concentrations of DNCB caused only allergic effect but not any irritant effect. This study reports that the use of these non-radioactive endpoints can assess allergic contact dermatitis caused by chemicals.

Keywords::

• Local lymph node assay
• non-radioactive endpoints
• DNCB
• cytokine production

Introduction

The murine local lymph node assay (LLNA) has been developed for identifying chemicals that have the potential to cause skin sensitization and allergic contact dermatitis (ACD), which is a common and serious health problem (Kimber and Basketter, 1992; Basketter et al., 1996, 2001). The LLNA was endorsed by the Interagency Coordinating Committee on the Validation of Alternative Methods (ICCVAM) in the United States and by the European Centre for the Validation of Alternative Methods (ECVAM) in Europe (ICCVAM, 1999; ECVAM, 2000).

In the LLNA protocol, the proliferative activity of draining auricular lymph node cells after topical application of the test chemicals is measured and compared with vehicle treated controls. In the standard LLNA, cell proliferation is measured using the incorporation of the radiolabeled thymidine into draining lymph node cells. A chemical is regarded as a skin sensitizer when the proliferation of these cells is stimulated more than three times with at least one applied concentration (Basketter et al., 2001, Kimber and Weisenberger, 1989).

Other than proliferative responses, many studies have tried to characterize the chemical sensitizers according to the cytokine profiles (Dearman et al., 1994,1996a,b, 1999, 2003; Manetz and Meade, 1999; Hayashi et al., 2001; Manetz et al., 2001; Vandebriel et al., 2000, 2003; Plitnick et al., 2002; Van Och et al., 2002, Ku et al., 2008). Cutaneous immune responses involving T-helper Type 1 (T_H1) and Type 2 (T_H2) cells, characterized by secretion of interferon (IFN)-γ and interleukin (IL)-2 and of IL-4 and IL-5, respectively, have been reported in ACD (Romagnani, 1991, Rowe and Bunker, 1998; Ulrich et al., 2001).

The LLNA has several advantages since it is quicker and cheaper, providing a quantitative endpoint. The assay also provides advantages in the context of animal welfare, i.e., compared with the traditional guinea pig tests, fewer animals are required and these animals are subjected to less trauma. However, the assay also has two disadvantages: in vivo tagging of lymph node cells with [³H]-thymidine to monitor proliferation labeling leads to the generation of radioactive biologic wastes (and, accordingly, the need for dedicated ‘hot’ housing of the hosts and special handling/disposal of all lab/animal wastes) and the possibility of false positive results caused by non-specific cell activation due to inflammatory processes in the skin (irritation) (Takeyoshi et al., 2001; Basketter et al., 2009).

In this study, we aimed to investigate the following non-radioactive endpoints of LLNA: 5-bromo-2′-deoxyuridine (BrdU) incorporation ex vivo and in vivo, as well as the formation/release of IL-2, IFNγ, IL-4, and IL-5 by isolated lymph node cells. To differentiate between allergic and irritant potential, the Differentiation Index (DI) was also calculated with each experimental group.

Material and methods

Mice

For these studies, 8-12-week-old female Balb/c mice were used. Mice were obtained from Gulhane Military Medical Academy (Ankara, Turkey) and then housed at a temperature of 23°C and relative humidity of 55% and with a 12 h light/dark cycle. Mice were provided mouse chow (Optima, Kırklareli, Turkey) and water ad libitum. All animal procedures were conducted in an AAALAC-accredited facility under an animal protocol approved by the Ankara University Animal Experiments Ethics Board.

Chemicals

2,4-Dinitrochlorobenzene (DNCB) (Sigma-Aldrich, Steinheim, Germany) in acetone:olive oil (4:1 v/v) (AOO) (Sigma)-at concentrations of 0.025, 0.05, 0.01, or 0.25%-was applied to the animals in these studies. 5-Bromo-2′-deoxyuridine (Sigma) was dissolved in physiological saline to achieve final concentrations of 150 µg/g body weight [BW] for injection in a volume of 15 µl.

In vivo BrdU incorporation

Five groups of mice (n = 4/group) were exposed topically on the dorsum of both ears to 25 µl of the different concentration of DNCB and vehicle (AOO) alone daily for three consecutive days (i.e., Days 1–3). All mice were then injected intraperitoneally with 15 µl of BrdU-saline (150 µg/g BW) solution on Day 4. On Day 5, the mice were killed by cervical dislocation and their auricular lymph nodes were excised and weighed. Before the application of DNCB and on Day 5, ear thickness was measured using a digital micrometer (Mitutoyo Digital Caliper, Kanagawa, Japan) to calculate each host’s ‘DI’ using the formulas described by Homey et al. (1998) (see below).

After the excised right and left lymph nodes were pooled and homogenized, the cells were washed twice with phosphate-buffered solution (PBS) at 4°C and were suspended in 10 ml of PBS. The lymphocytes present were counted manually using an improved Neubauer counting chamber (Marienfeld, Lauda-Königshofen, Germany). The extent of BrdU incorporation was then measured using a BrdU cell proliferation ELISA kit (Roche, Penzberg, Germany, Cat No: 1164722901) according to manufacturer’s instructions. The absorbance at 450 nm (OD₄₅₀) was determined with the microplate reader (SpectraMAX, Molecular Devices Inc., Sunnyvale, CA); this value was defined as the BrdU labeling index.

Ex vivo BrdU incorporation

Five groups of mice (n = 4/group) were exposed topically on the dorsum of both ears to 25 µl of different concentration of DNCB and vehicle (AOO) alone daily for three consecutive days. Unlike in the ‘In vivo’ protocol above, the mice in this study were only rested on Day 4. On Day 5, all mice were killed by cervical dislocation and their auricular lymph nodes were excised and weighed. Before the application of DNCB and on Day 5, ear thickness was measured with a digital caliper to calculate each host’s DI.

Excised right and left lymph nodes were pooled and homogenized, and the released cells suspended in 15 ml physiological saline. After counting, some of the cells from the suspension were then seeded into 96 well-culture plates at 10⁵ cells/well in RPMI 1640 medium supplemented with 10% fetal bovine serum (FBS) and 1% penicillin–streptomycin (Biochrom, Berlin, Germany). BrdU (10 µl of a 10 µM BrdU labeling solution) was then added to the wells after 48 h of culture at 37°C. The cells in the wells were then recovered by aspiration and the extent of BrdU incorporation measured by ELISA (Roche), according to manufacturer’s instructions. The absorbance (OD₄₅₀) was determined with the SpectraMAX microplate reader and these values were defined as the BrdU labeling index.

Calculations of stimulation indices and differentiation indices

Stimulation Index (SI) is the ratio of the mean of (in vivo or ex vivo) BrdU incorporation measurements for each DNCB treatment group vs. that of the vehicle treatment control group. This value was calculated by dividing the mean BrdU labeling index value obtained with each treatment group by that of the control group.

The DI indicating the proportion of non-specific activation due to irritating properties of test compound was calculated using the following formulas:

Calculation of EC₃ value

The EC₃ value is the measurement of the relative skin sensitizing potency of a substance, usually expressed as the estimated concentration of chemical necessary to produce a 3-fold increase in proliferation in draining lymph nodes compared with concurrent vehicle-treated controls (Kimber et al., 2002). The method represents a simple linear interpolation of the points in the dose–response curve, which lie immediately above and below the classification threshold, i.e. an SI of 3. If the datapoints lying immediately above and below the SI value of 3 have the coordinates (a, b) and (c, d), respectively, then the EC₃ value may be calculated using the following equation (Basketter et al., 1999b, 2007): EC₃ = c + [(3 − d)/(b − d)] (a− c).

Culture of lymph node cells and cytokine determinations

Harvested lymph node cells from the in vivo and ex vivo protocols outlined above were seeded in a 24 well-culture plate at 5 × 10⁶ cells/well in 1 ml of RPMI 1640 medium supplemented with 10% FBS and 1% penicillin–streptomycin. In the case of the ex vivo method, the cells used here were obtained from the pooled cells from Day 4-rested hosts prior to removal of aliquots for use in BrdU staining/analyses (see above). Following the seeding steps, some wells were then supplemented with 5 µg/ml of phytohemagglutinin-L (PHA-L; Biochrom). PHA-L was selected as the mitogen here (as opposed to concanavalin A) in that it has been widely used for mitotic stimulation of T-lymphocytes. After 72 h of culture in a 37°C incubator containing 5% CO₂, supernatants were collected and stored at −80°C until analyzed for levels of IL-2, IFNγ, IL-4, and IL-5. The levels of each cytokine in the culture supernatants were measured using commercially available ELISA kits (BenderMed Systems, Burlingame, CA), according to the manufacturer’s instructions.

Statistical analyses

Means and standard errors were calculated for the lymph node weights and for the optical density (OD) values obtained by ELISA for each treatment group. All data were initially analyzed via one-way ANOVA; if the analyses indicated significant difference, the difference(s) between the vehicle control and each DNCB groups were then analyzed using a Dunnett t-test.

Results

Lymph node-related parameters and ear thickening outcomes

In these studies, Balb/c mice were exposed topically (on dorsum of both ears) to 0.025%, 0.05%, 0.1%, 0.25% DNCB or to AOO vehicle alone on three consecutive days (i.e., Days 1–3). In the ‘In Vivo BrdU’ studies, mice in each group were then injected intraperitoneally with 15 µl of BrdU-saline (150 µg/g BW) solution on Day 4; in the ‘Ex Vivo BrdU’ studies, mice received no BrdU injection but were instead rested on Day 4. One day later (i.e., 2 days after final application, Day 5), all mice were killed. The auricular lymph node weights, lymph node cell counts, and changes in ear thickness values associated with each mouse on Day 5 are shown in Table 1.

Table 1.  Change in ear thickness, lymph node weight, and cell count after treatment of mice with different concentrations of DNCB in AOO.

Labeling	Chemical	Vehicle	Concentration (% w/v)	Lymph node weight (g)	Cell count (10^7/ml)	Change of ear thickness (mm)	DI
In vivo	DNCB	AOO	—	0.0180 ± 0.0006	1.000 ± 0.082	0	> 1
			0.025%	0.0200 ± 0.0008	1.600 ± 0.082	0	> 1
			0.05%	0.0280 ± 0.0019*	2.400 ± 0.082*	0.010 ± 0.002*	> 1
			0.1%	0.0300 ± 0.0008*	2.800 ± 0.163*	0.030 ± 0.002*	> 1
			0.25%	0.0360 ± 0.0010*	3.000 ± 0.490*	0.030 ± 0.004*	> 1
Ex vivo	DNCB	AOO	—	0.010 ± 0.004	0.70 ± 0.24	0	> 1
			0.025%	0.012 ± 0.003	0.80 ± 0.16	0	> 1
			0.05%	0.020 ± 0.004*	1.40 ± 0.08*	0.020 ± 0.004*	> 1
			0.1%	0.030 ± 0.004*	1.60 ± 0.21*	0.030 ± 0.001*	> 1
			0.25%	0.038 ± 0.002*	1.80 ± 0.13*	0.040 ± 0.006*	> 1

Each value represents the mean ± SD.

*P < 0.01; Significant differences from vehicle control (Dunnett’s test).

The results indicate that at concentrations ≥ 0.05%, the application of DNCB in AOO caused significant increases in lymph node weight, cell counts, and in change of ear thickness as compared to the values noted with the mice that received vehicle treatments only (P < 0.01). Regardless of the mice received BrdU on Day 4, the effect on lymph node weight reflected an apparent dose-trend response; the trends were not as apparent for the cell count or change in ear thickness endpoints. Interestingly, the mice that had received the BrdU injection on Day 4 appeared to consistently have the greater cell counts when compared to the ex vivo groups receiving the same dose of DNCB were made. Unfortunately, again, this trend was not apparent for the other two endpoints outlined in the Table 1.

A DI was defined describing the relation between skin-draining lymph node cell activation (lymph node cell count index) and skin inflammation (ear swelling). A DI > 1 indicates an allergic reaction pattern, whereas a DI < 1 demonstrates an irritant potential of a chemical (Homey et al., 1998). As the differentiation indices were found to be > 1 in our study, the applied concentrations of DNCB were deemed to have caused only an allergic effect, and not any irritant effect.

The results indicate that the SI was found to attain a level of 3 at the concentration of 0.05% DNCB. Also, when the EC₃ values derived from the in vivo and ex vivo BrdU labeling study results were calculated, it was seen that there were insignificant differences between the two protocols utilized (i.e., EC₃ values of 0.059% and 0.054%, respectively). The data for all OD₄₅₀, SI, and EC₃ values from both the in vivo and ex vivo BrdU labeling studies are presented in Table 2. In no case did the difference in BrdU treatment (or not, on Day 4) influence these outcomes.

Table 2.  OD₄₅₀, SI, and EC₃ values from the in vivo and ex vivo BrdU labeling studies.

Labeling	Chemical	Vehicle	Concentration (%w/v)	OD450	SI	EC3(%)
In vivo	DNCB	AOO	—	0.230 ± 0.012	1.00	0.059%
			0.025%	0.310 ± 0.018	1.34
			0.05%	0.690 ± 0.022	3.00
			0.1%	1.154 ± 0.025	5.00
			0.25%	1.580 ± 0.020	6.80
Ex vivo	DNCB	AOO	—	0.054 ± 0.010	1.00	0.054%
			0.025%	0.104 ± 0.016	2.00
			0.05%	0.152 ± 0.018	3.00
			0.1%	0.236 ± 0.012	4.60
			0.25%	0.329 ± 0.036	6.40

The OD₄₅₀ values shown are in terms of mean ± SD.

Cytokine determinations

As noted above, on Day 5, auricular lymph nodes were excised from each mouse in the various treatment groups and lymph node cell suspensions were prepared. These cells were then cultured in medium alone or in the presence of PHA mitogen to assess background and inducible formation of select cytokines (e.g., IL-2, IFNγ, IL-4, and IL-5) associated with T_H1 and T_H2-type responses. The levels of IL-2, IFNγ, IL-4, and IL-5 present in the lymph node cell culture supernatants and the SI values are provided in Tables 3–6, respectively.

Table 3.  IL-2 levels in cultures of auricular lymph node cells from mice treated daily (3 days) with DNCB.

DNCB (in AOO) concentration (% w/v)	In vivo (+) PHA	SI in vivo (+) PHA	In vivo ( − ) PHA	Ex vivo (+) PHA	SI ex vivo (+) PHA	Ex vivo (−) PHA
0	80.65 ± 11.07^a	1.00	ND	7.89 ± 1.52	1.00	ND
0.025%	112.31 ± 13.36	1.39	ND	13.73 ± 1.53	1.74	ND
0.05%	228.99 ± 50.22*	2.84	ND	67.04 ± 15.10*	8.50	19.46 ± 3.75
0.1%	329.64 ± 65.89*	4.09	ND	97.54 ± 1.35*	12.36	79.13 ± 29.54
0.25%	593.15 ± 30.20*	7.35	ND	235.87 ± 66.40*	29.89	122.25 ± 16.62

^aIL-2 levels in terms of pg/ml (mean ± SE). *P < 0.01; Significant differences from vehicle control (Dunnett’s test).

ND: Not detectable.

Table 4.  IFNγ levels in cultures of auricular lymph node cells from mice treated daily (3 days) with DNCB.

DNCB (in AOO) concentration (% w/v)	In vivo (+) PHA	SI in vivo (+) PHA	In vivo ( − )PHA	Ex vivo (+) PHA	SI ex vivo (+) PHA	Ex vivo (−) PHA
0	266.06 ± 57.89^a	1.00	ND	88.28 ± 8.05	1.00	ND
0.025%	585.66 ± 191.48^†	2.20	ND	117.57 ± 15.21^†	1.33	45.5 ± 9.49
0.05%	912.54 ± 198.70*	3.43	6.23 ± 1.54	382.42 ± 25.13*	4.33	81.77 ± 8.34
0.1%	1358.8 ± 132.38*	5.11	27.12 ± 13.84	516.28 ± 32.33*	5.85	277.26 ± 25.60
0.25%	1539.07 ± 20.16*	5.78	272.18 ± 15.69	1378.37 ± 121.17*	15.61	682.47 ± 12.94

^aIFNγ levels in terms of pg/ml (mean ± SE).

^†P < 0.05. *P < 0.01; Significant differences from vehicle control (Dunnett’s test).

ND: Not detectable.

Table 5.  IL-4 levels in cultures of auricular lymph node cells from mice treated daily (3 days) with DNCB.

DNCB (in AOO) concentration (% w/v)	In vivo (+) PHA	SI In vivo (+) PHA	In vivo ( − ) PHA	Ex vivo (+) PHA	SI Ex vivo (+) PHA	Ex vivo (−) PHA
0	24.76 ± 4.65^a	1.00	ND	9.36 ± 2.29	1.00	ND
0.025%	29.00 ± 1.48	1.17	1.52 ± 0.61	13.98 ± 1.19	1.49	6.06 ± 3.35
0.05%	32.37 ± 5.14	1.31	4.65 ± 7.64	15.55 ± 0.69	1.66	8.33 ± 5.43
0.1%	62.62 ± 2.26*	2.53	7.64 ± 1.20	24.72 ± 2.79*	2.64	23.7 ± 1.22
0.25%	79.03 ± 8.24*	3.19	15.32 ± 1.16	44.16 ± 4.09*	4.72	27.92 ± 1.93

^aIL-4 levels in terms of pg/ml (mean ± SE).

*P < 0.01; Significant differences from vehicle control (Dunnett’s test).

ND: Not detectable.

Table 6.  IL-5 levels in cultures of auricular lymph node cells from mice treated daily (3 days) with DNCB.

DNCB (in AOO) concentration (% w/v)	In vivo (+) PHA	SI In vivo (+) PHA	In vivo ( − )PHA	Ex vivo (+) PHA	SI Ex vivo (+) PHA	Ex vivo (−) PHA
0	90.65 ± 43.41^a	1.00	ND	67.11 ± 2.34	1.00	ND
0.025%	280.19 ± 21.50	3.09	8.21 ± 4.52	150.26 ± 27.17	2.24	33.5 ± 14.14
0.05%	305.44 ± 37.36	3.37	25.38 ± 10.43	186.94 ± 27.17	2.78	57.76 ± 9.01
0.1%	510.43 ± 44.14*	5.63	55.6 ± 4.52	340.35 ± 27.46*	5.07	199.25 ± 9.36
0.25%	806.14 ± 61.28*	8.89	223.28 ± 25.65	560.19 ± 77.73*	8.35	354.23 ± 32.33

^aIL-5 levels in terms of pg/ml (mean ± SE).

*P < 0.01; Significant differences from vehicle control (Dunnett’s test).

ND: Not detectable.

In the absence of DNCB treatment and PHA stimulation in vitro, detectable levels of each of these cytokines could not be measured in the medium of the lymph node cell suspensions. With regard to IL-2, in the case of mice that had received a BrdU injection on Day 4, none of the DNCB doses led to increased formation/release of the cytokine unless PHA was present. When PHA was presented to cells from these mice, the formation of IL-2 was strongly elevated in a manner reflecting a clear (DNCB) dose-related trend. This strong elevation also was reflected in the calculated SI values. However, the relationship to dose was not directly linear, i.e., as the dose of DNCB increased by 10 (from 0.025 to 0.25%), the level of PHA-induced IL-2 was only nearly tripled. These dose-trend patterns (as well as lack of linear nature) were also evident with the cells from mice that had not received BrdU on Day 4. Also, as was noted with the lymph node physical parameter results (see above), the cells from mice that had received the BrdU injection on Day 4 appeared to consistently have the greater ability to produce IL-2 in response to PHA, when comparisons between groups receiving the same dose of DNCB were made. Interestingly, the opposite was the case when background levels (spontaneous formation/release, no PHA) of the cytokine were analyzed.

With IFNγ, in the case of mice that had received a BrdU injection on Day 4, again there were no significant increases in formation/release of the cytokine unless PHA was present; this now occurred when DNCB levels ≥ 0.025% had been applied. As was the case for IL-2, the changes in expression of IFNγ seemed to reflect a DNCB dose-related trend overall. Once again, the cells from mice that had received the BrdU injection on Day 4 appeared to consistently have the greater ability to produce the cytokine in response to PHA and the lesser capacity in the case of background spontaneous formation/release (no PHA) of the cytokine. Overall, there was a noteworthy enhancement of SI values among the PHA-treated cells, especially for the cells generated in the ex vivo method as compared to in the case of the in vivo protocol.

In regard to the results for IL-4 and IL-5, in the case of mice that had received a BrdU injection on Day 4, none of the DNCB doses led to any significant increase in formation/release of the cytokine (relative to the vehicle control values) unless PHA was present. Furthermore, in the latter cases, this only now occurred (unlike with IFNγ or IL-2) when the levels of DNCB that had been used were ≥ 0.1%. Thus, while changes in expression of the IL-4 or IL-5 T_H2-type cytokines did reflect DNCB dose-related trends, the dose range over which this occurred appeared to be somewhat narrower (i.e., only first starting at 0.05%) than that for the two T_H1-type cytokines assessed here. Nevertheless, even with these underwhelming outcomes, the cells from mice that had received the BrdU injection on Day 4 again appeared to repeatedly have the greater ability to produce these proteins in response to PHA and a lesser capacity with regard to their background spontaneous formation/release. In general, there was an enhancement in the SI values of these two cytokines that was related to the doses of DNCB used (for cells generated in either protocol), but the increases were not as dramatic as those noted in the cases of IL-2 and IFNγ.

In summary, the levels of IL-2 and the two T_H2 cytokines (i.e., IL-4 and IL-5) in the culture supernatants were significantly increased when DNCB was applied at the concentrations of 0.05% and 0.1%, respectively (P < 0.01). The significant induction of IFNγ was detected at the 0.025% concentration (P < 0.05).

Discussion

DNCB has been previously shown to induce positive responses in the standard LLNA and has been categorized as an extremely effective contact allergen (Kimber et al., 2003) based on EC₃ values of 0.05% (Betts et al., 2006). EC₃ values demonstrate the estimated concentration of chemical required to induce an SI of 3 relative to concurrent vehicle-treated controls (Basketter et al., 1999a; Gerberick et al., 2007). In the study reported here, we sought to investigate the utility of select non-radioactive endpoints (e.g., ex vivo/in vivo BrdU incorporation and IL-2, IFNγ, IL-4, IL-5 release) for the LLNA following topical application of DNCB.

BrdU, a non-radioactive alternative to [³H]-thymidine, has been used in various studies. Flow cytometric analysis and immunohistochemical assays have been developed to count BrdU incorporation into lymph node cells (Boussiquet-Leroux et al., 1995; Lee et al., 2002, Suda et al., 2002). Takeyoshi et al. (2001) investigated BrdU incorporation into lymph node cells by using a cell proliferation ELISA kit but their results for a number of allergens suggested an insufficient sensitivity because the increase in lymph node cell counts was not taken into account when calculating SI values (Takeyoshi et al., 2006). Using this method, DNCB yielded positive results at doses of ≥ 0.1%, although an EC₃ value was found to be 0.05% for DNCB in the standard LLNA method. However, recently, ICCVAM has approved the LLNA:BrdU-ELISA method that was based on the protocol developed by Takeyoshi et al. (2001) (ICCVAM, 2010).

In our study, we found significant increases at doses of 0.05% for DNCB. Using the in vivo BrdU incorporation method, an EC₃ value of 0.059% was calculated using the earlier-noted formula. The protocol for the non-radioactive LLNA developed by Takeyoshi et al. (2001), after the auricular lymph nodes were excised and weighed, they were stored at −20°C until the day of analysis. In the in vivo BrdU study here, BrdU incorporation into the lymph node cells was measured on the same day as excision of these nodes. Therefore, this study reports an improvement in the sensitivity of the non-radioactive LLNA, and the value is found to be equivalent to that obtained using a standard LLNA. This study indicates the importance of preparing the suspension of lymph node cells on the same day as the excision of the auricular lymph nodes.

Another goal of this study was aimed at improving animal welfare by having isolated lymphocytes labeled with BrdU ex vivo (by adding BrdU at the 48-h point in the lymphocyte cell culture) instead of having this done in vivo (in situ) (i.e., labeling by intraperitoneal BrdU injection). Using the ex vivo BrdU labeling method, an EC₃ value was calculated to be 0.054%, a result comparable with the in vivo EC₃ value.

Contact sensitivity is a typical example of the delayed-type (Type IV) hypersensitivity, i.e. mediated by effector T-lymphocytes. Upon primary contact sensitization with antigen, these effector T-lymphocytes are generated in lymphoid tissues within 3–4 days and are recruited into the local tissue following local skin challenge with the same antigen. At 24-h postchallenge, the inflammation associated with contact sensitivity reaches its maximum level, with massive infiltration of antigen-non-specific mononuclear leukocytes and granulocytes (i.e., neutrophils and eosinophils) (Roupe and Ridell 1979; Itakura et al., 2006). It has long been known that effector T-lymphocytes of contact sensitivity are recruited into the antigen-challenged local tissue. Against the common belief that these effectors are recruited by antigen-non-specific mechanisms, it has been shown that antigen-specific immunoglobulin M (IgM) antibodies are required for effector T-lymphocyte recruitment, leading to elicitation of contact sensitivity responses. These IgM antibodies are produced by B-1 cells, a subpopulation of B-lymphocytes (Tsuji et al., 2002, Hardy and Hayakawa,1994).

An increase in cytokine levels is related to T-helper cell proliferation; each of these factors plays an important role in the induction and elicitation of contact sensitivity. Many studies have tried to characterize the chemical sensitizers according to the cytokine profiles they provoke (Dearman et al., 1996a,b, 1999, 2003; Manetz and Meade, 1999; Hayashi et al., 2001; Manetz et al., 2001; Plitnick et al., 2002; Vandebriel et al., 2000, 2003; Van Och et al., 2002). The regulation of contact sensitivity responses by T_H1 and T_H2 cytokines has been studied intensively. The T_H1 cytokine IFNγ is widely known as one of the most important effector cytokines for the onset/maintenance of contact sensitivity (Fong and Mosmann, 1989). Azam et al. (2005) suggested that an IL-2-based LLNA yielded similar outcomes (at both qualitative and quantitative levels) when compared to those obtained via the traditional LLNA. In our study here, we also detected significant increases in IL-2 and IFNγ levels at 0.05% and 0.025% DNCB, respectively.

In contrast, the T_H2 cytokines (primarily IL-4 and IL-5) have long been considered suppressive against contact sensitivity (Berg et al., 1995; Xu et al., 1996; Biedermann et al., 2001) and IL-4 is considered as a key regulator of humoral immunity. Also, it has been postulated that T_H2 cytokines can be used to distinguish respiratory sensitizers from skin sensitizers (Dearman and Kimber, 2000; De Jong et al., 2009). However, Ulrich et al. (2001) suggested that the use of T_H2 cytokine ‘profiles’ to distinguish respiratory sensitizers from skin sensitizers did not appear to be an appropriate tool. We found significant increases in both of these T_H2 cytokines at a 0.1% DNCB concentration. Similar to our results, Ku et al. (2008) reported that both IFNγ and IL-4 were significantly up-regulated in draining lymph node (LN) cells by the sensitizers but not by the irritants. It was also found out that IL-4 plays a positive role in elicitation of contact sensitivity. Thus, IL-4 appears to stimulate B-1 cells to produce antigen-specific IgM antibodies required for local recruitment of effector T-lymphocytes (Campos et al., 2003). Traidl et al. (1999) investigated the role of IL-4 in contact hypersensitivity reactions in IL-4-deficient mice. These investigators found that IL-4-deficient mice sensitized to DNCB showed a significant reduction in magnitude and duration of their contact hypersensitivity responses in comparison with those observed with their wild-type counterparts. The results of our observations here appear to be compatible with these latter studies.

IL-5 is one of the most important cytokines for eosinophils and controls their differentiation, survival, and function (Lopez et al., 1988; Yamaguchi et al., 1988; Hitoshi et al., 1991). It was reported that IL-5 plays an essential role in elicitation of contact sensitivity, with an effect on both eosinophils and B-1 cells (Itakura et al., 2006). These researchers showed that contact sensitization responses were impaired in IL-5-deficient mice. We found significantly induced IL-5 levels produced ex situ by murine auricular lymph node cells in hosts that had been treated with 0.1% DNCB. Consequently, these findings demonstrate that not only T_H1 cytokines such as IL-2 and IFNγ but also T_H2 cytokines such as IL-4 and IL-5 can play important roles in elicitation of contact sensitivity.

To compare the potential utility of the in vivo vs. the ex vivo methods used here as they relate to cytokine release, absolute levels of each cytokine released (with and without presence of mitogen) were analyzed. It is clear that in the absence of PHA, with increasing host DNCB treatment dose, spontaneous release of each of the four cytokines was higher by cells generated in the ex vivo protocol than by those in the in vivo method. Unexpectedly, completely opposite outcomes were seen when the mitogen was present. Furthermore, even in the absence of any DNCB treatments, the levels of cytokine release were consistently far greater by cells of BrdU-injected mice than by cells from non-injected counterparts. In light of this conundrum, SI values were then generated to provide an additional endpoint for our comparison of the two methods. In this case, it was clear (except for the case of IL-5) that the ex vivo approach yielded SI values for cells from DNCB-treated hosts that trended higher than those associated with cells of mice that underwent the in vivo method. Based on these SI results, we are able to suggest that measuring ex vivo BrdU incorporation into lymph node cells could be used as a sensitive and useful method to assess ACD caused by chemicals-without adversely affecting on additional measures of cell function, such as cytokine formation/release.

The obvious question engendered by these above outcomes remains why there were such stark differences in SI values/levels (magnitude) of cytokine release between the ex vivo and in vivo data. One potential basis for these differences might be that the mice were ‘dosed’ on Day 4 (with the BrdU solution) in the in vivo protocol and not ‘rested’ as in the ex vivo method. It is clear that the stress incurred during the Day 4 injection process, brief that it was for each animal, could have implications for the functionality (such as mitogen responsiveness) of their immune system cells (Plytycz and Seljelid, 2002). In retrospect, our study should have ‘sham-injected’ the hosts whose cells were used for the ex vivo protocols so that if there was any confounding effect from handling/injection on Day 4, it would have been uniform across the two protocols.

A second factor could be that the cells in the ex vivo study for cytokine analyses were not exposed to BrdU at any point, i.e., these cells were harvested from the Day 4-rested hosts on Day 5 and then removed from among all the pooled cells for use in the cytokine studies (while other cells in this pool were used for BrdU treatment). It has been demonstrated by Brent et al. (1977) that there is a clear degree of toxicity from BrdU upon lymphocytes in culture. Whether there is any similar direct toxicity from BrdU in situ remains to be determined. Nevertheless, the fact that the in vivo method cytokine studies used cells that had ‘seen’ BrdU while those in the ex vivo method never ‘saw’ the agent (and therefore did not become functionally modified it) might be an additional critical factor that works in the favor of the use of the ex vivo BrdU method to assess ‘truer’ measures of function of lymphocytes from hosts being used to assess a potential for ACD caused by a test chemical.

Homey et al. (1998) and Vohr et al. (2000) presented an integrated model for differentiating skin (IMDS) reaction that assessed ear swelling and cell proliferation in lymph nodes to detect and differentiate chemical-induced irritant and allergic skin responses. In our study, since DI values were > 1 for all concentrations, DNCB is an allergic compound. These results are compatible with those of Suda et al. (2002).

Conclusion

In conclusion, this study suggests that the use of these non-radioactive endpoints can be used to evaluate ACD caused by chemicals in any laboratory where it might be difficult to handle/employ radioisotopes. The data also suggest that measuring ex vivo BrdU incorporation into lymph node cells could be preferable as a sensitive method with respect to improving overall animal welfare in that it allows for one less set of injections to be imposed on a treated host. The data also clearly show that the ex vivo approach does not adversely affect the measures of formation/release of key T_H1/T_H2 cytokines that can play important roles in the elicitation/propagation of contact sensitivity. Nevertheless, further studies with different chemicals (including known irritants and allergens) are still needed in order to be able make more general comments about the true utilizability of this novel protocol. In fact, the study presented here is a part of a project supported by the Scientific and Technological Research on Council of Turkey (TUBITAK) wherein the impact of some moderate and weak allergens are also to be tested using this method. That particular project is still ongoing; as such, that data will be presented in a follow-on paper.

Declaration of interest

This study has been supported by the Scientific and Technological Research on Council of Turkey (TUBITAK) (Project number: 107S365). The authors report no conflicts of interest. The authors are alone responsible for the content and writing of the paper.