Toxicology Letters 209 (2012) Contents lists available at SciVerse ScienceDirect. Toxicology Letters

Transcription

1 Toxicology Letters 29 (212) Contents lists available at SciVerse ScienceDirect Toxicology Letters jou rn al h om epage: B cell increases and ex vivo IL-2 production as secondary endpoints for the detection of sensitizers in non-radioisotopic local lymph node assay using flow cytometry Kyoung-Mi Jung a, Won-Hee Jang a, Yong-Kyoung Lee b, Young Na Yum b, Soojung Sohn b, Bae-Hwan Kim c, Jin-Ho Chung d, Young-Ho Park a, Kyung-Min Lim a, a Medical Beauty Research Institute, Amorepacific CO R&D Center,Republic of Korea b National Institute of Food and Drug Safety Evaluation/Korea Food and Drug Administration,Republic of Korea c Department of Public Health, Keimyung University,Republic of Korea d College of Pharmacy, Seoul National University, Seoul ,Republic of Korea a r t i c l e i n f o Article history: Received 8 July 211 Received in revised form 22 December 211 Accepted 23 December 211 Available online 2 January 212 Keywords: Non-radioisotopic local lymph node assay 5-Bromo-2 -deoxyuridine Flow cytometry B cell Cytokine IL-2 a b s t r a c t Non-radioisotopic local lymph node assay (LLNA) using 5-bromo-2 -deoxyuridine (BrdU) with flow cytometry (FCM) is gaining attention since it is free from the regulatory issues in traditional LLNA (tllna) accompanying in vivo uses of radioisotope, 3 H-thymidine. However, there is also concern over compromised performance of non-radioisotopic LLNA, raising needs for additional endpoints to improve the accuracy. With the full 22 reference substances enlisted in OECD Test Guideline No. 429, we evaluated the performance of LLNA:BrdU-FCM along with the concomitant measurements of B/T cell ratio and ex vivo cytokine production from isolated lymph node cells (LNCs) to examine the utility of these markers as secondary endpoints. Mice (Balb/c, female) were topically treated with substances on both ears for 3 days and then, BrdU was intraperitoneally injected on day 5. After a day, lymph nodes were isolated and undergone FCM to determine BrdU incorporation and B/T cell sub-typing with B22 + and CD3e +. Ex vivo cytokine production by LNCs was measured such as IL-2, IL-4, IL-6, IL-12, IFN-, MCP-1, GM-CSF and TNF. Mice treated with sensitizers showed preferential increases in B cell population and the selective production of IL-2, which matched well with the increases in BrdU incorporation. When compared with guinea pig or human data, BrdU incorporation, B cell increase and IL-2 production ex vivo could successfully identify sensitizers with the accuracy comparable to tllna, suggesting that these markers may be useful for improving the accuracy of LLNA:BrdU-FCM or as stand-alone non-radioisotopic endpoints. 212 Elsevier Ireland Ltd. All rights reserved. 1. Introduction Skin contact with chemicals may lead to contact allergic dermatitis and other skin diseases. Therefore, an accurate evaluation of chemical-induced adverse effects on skin is important in the area of dermatotoxicology and for the development of safe pharmaceutical and industrial chemicals. The murine local lymph node assay (LLNA) is a validated model for the evaluation of contact allergic potential of chemicals, which has been endorsed by OECD (Organization for Economic Cooperation and Development) Abbreviations: LLNA, local lymph node assay; tllna, traditional local lymph node assay; BrdU, 5-bromo-2 -deoxyuridine; FCM, flow cytometry. Corresponding author at: Amorepacific Corporation R&D Center, 314-1, Boradong, Giheung-gu, Yongin , Republic of Korea. Tel.: ; fax: address: kimlim@amorepacific.com (K.-M. Lim). for an alternative to the conventionally used guinea pig tests (OECD TG46, 1992). In traditional LLNA(tLLNA), sensitization response is examined through the measurement of proliferating lymphocytes with 3 H-thymidine uptake, but many countries have a strict regulation on the in vivo use of radioisotopes (Gerberick et al., 22, 27), hindering the dissemination of tllna. Recently, to avoid the in vivo use of 3 H-thymidine, many attempts have been made to introduce non-radioisotopic endpoints in LLNA. Among them, the measurement of 5-bromo-2 -deoxyuridine (BrdU) incorporation (Takeyoshi et al., 21) is gathering a huge interest, owing to its mechanistic equivalency to tllna. For the measurement of BrdU incorporation into lymph node, several antibody-based assay methods are available including flow cytometric analysis (LLNA:BrdU-FCM) (Jung et al., 21; Suda et al., 22), immunohistochemistry (LLNA:BrdU-IHC) (Boussiquet-Leroux et al., 1995; Jung et al., 21; Lee et al., 22) and ELISA (LLNA:BrdU-ELISA) (Takeyoshi et al., 21). Previously, with 7 reference substances, we compared the performance of LLNA:BrdU-FCM and LLNA:BrdU-IHC /$ see front matter 212 Elsevier Ireland Ltd. All rights reserved. doi:1.116/j.toxlet

2 256 K.-M. Jung et al. / Toxicology Letters 29 (212) and demonstrated that LLNA:BrdU-FCM is superior in sensitivity which is comparable to tllna (Jung et al., 21). However, there is a concern over low signal to noise ratios and consequently, compromised accuracy in non-radioisotopic LLNA (Takeyoshi et al., 26), raising the needs for assessing the accuracy of LLNA:BrdU-FCM sufficiently with more reference substances. In the same context, many efforts have been made to develop additional endpoints to improve the accuracy of non-radioisotopic LLNA. Allergen-treatment induces a significant increase in lymphocyte population (Gerberick et al., 1999, 22; Inman, 1966). Both B cell and T cell populations are known to increase but a more specific and prominent expansion is observed with B cell population in response to sensitizers when compared with irritants. Flow cytometric analysis of the positivity of B22, an isoform of a transmembrane protein expressed on B cells that assists the activation of the cells, can provide a specific marker to measure B cell expansion (Gerberick et al., 22). Fortunately, flow cytometry method allows an analysis of multiple parameters at one scope (Bilsland et al., 26), which enables the simultaneous measurement of both the immunophenotyping of lymph node cells (LNCs) and BrdU incorporation. Contact allergens provoke the production of pro-inflammatory cytokines (Dearman et al., 1996a,b, 23; Ku et al., 28; Manetz et al., 21; Van Och et al., 22; Vandebriel et al., 2). The development of immune reaction by skin sensitizers is orchestrated by the activity of lymphocytes and their cytokine production. In skin sensitization, the major role is played by the cytokines secreted by T H 1 cells that include IL-2 and IFN- but significant changes are also observed with other pro-inflammatory cytokines and antiinflammatory cytokines such as IL-4, IL-6, IL-1, GM-CSF, MCP-1 and TNF (Azam et al., 25; Dearman et al., 23; He et al., 21; Ku et al., 28). Therefore, cytokine production ex vivo from the isolated LNCs from draining lymph node has been frequently examined to determine sensitizers and to discriminate from irritants or non-sensitizers. However, detailed evaluation and comparison of ex vivo cytokine production with other endpoints of LLNA using a sufficiently large number of chemicals has not been done to our best knowledge. In this study, we performed LLNA:BrdU-FCM with 22 recommended reference substances (18 minimum and 4 optional) composed of 13 sensitizers, 5 non-sensitizer, 3 false positives and 1 false negative enlisted in OECD test guideline No. 429 (OECD TG429, 21) to evaluate its performance. In addition, ex vivo cytokine assay and flow cytometric measurement of B/T cell population in LNCs were conducted which could be done with the surplus LNCs without sacrificing extra animals. We evaluated the performance of respective endpoints and compared them with tllna and real human data to provide an insight into the performance of these methods in the evaluation of sensitization potentials. 2. Material and methods 2.1. Chemicals and reagents A total of 22 reference substances were selected from OECD test guideline No. 429 (OECD TG429, 21) to include 13 sensitizers, 5 non-sensitizers and 4 optional substances. The details and tested concentrations of the 22 reference substances tested are presented in Table 1. Selected 3 concentrations appropriately include the known LLNA EC3 values for sensitizers that were free from systemic toxicity and/or excessive local skin irritation. For compounds without known EC3 (negative) value, we checked the irritation potential in preliminary test with a concentration range that starts from the maximum soluble concentration to find the concentrations free from excessive irritation as advised in OECD TG429. BrdU was dissolved in phosphate-buffered saline (PBS) at a concentration of 1 mg/ml. Phycoerythrin (PE) conjugated rat monoclonal anti-mouse CD45R/B22 + and APC conjugated hamster monoclonal anti-mouse CD3e + were purchased from BD Pharmingen TM (BD Biosciences, Franklin Lakes, NJ) The following cytokine ELISA kit were purchased from Becton Dickinson ((Mississauga, Canada): Mouse IL-2, Mouse IL-4, Mouse IL-6, Mouse IL-12, Mouse IFN-, Mouse TNF, Mouse GM-CSF and Mouse MCP-1) Animal and experimental protocol Both the animal care and study protocol employed were in accordance with Institutional Animal Care and Use Committee (IACUC) of Amorepacific R&D center. They were kept under controlled conditions of temperature (23 ± 3 C) and relative humidity (5 ± 1%) with alternating 12 h light and dark. Throughout the study, animals had ad libitum access to tab water. Female Balb/c mice (7 8 weeks old, body weight g) were purchased from OrientBio (Seoul, Korea), and were used in all experiments. After they were kept for at least 1 week on laboratory solid diet (Purina Co., Korea), all animals were used at the age of 8 9 weeks. This study was performed according to the method of Jung et al. (21) with minor modifications. Groups of mice (N = 4 or 6) were treated with 25 l of the test substances in vehicle or vehicle alone on the back of both ears daily for 3 consecutive days (Days 1 3). On Day 5, mice were intraperitoneally injected with BrdU and were sacrificed after a day. After sacrifice, auricular lymph nodes were isolated, weighed and undergone lymphocyte preparation. After bilateral auricular lymph nodes were pooled on individual basis, lymph node cells (LNCs) were prepared by disaggregation through 7 m mesh (BD Biosciences, Franklin Lakes, NJ) in 1 ml PBS. The LNCs were counted using a hemacytometer after stained with trypan blue Flow cytometry analysis BrdU incorporation The LNCs ( /ml) were centrifuged (5 g) for 5 min in PBS and resuspended for fixation and permeabilization step, according to the instruction manual of BrdU Flow kits (BD Pharmingen TM, Franklin Lakes, NJ). Then LNCs were permeabilized using Cytoperm plus buffer, which contains 1% DMSO. After DNA was denatured by incubating in DNase for 1 h, LNCs were washed, and incubated with FITC conjugated anti-brdu antibody at a dilution of 1:5 for 2 min at RT in the dark. Cells were washed once more and then re-suspended in 2 l of 7-AAD solution to label DNA. Ten thousand 7-AAD expressing cells were gated, and the number of the cells expressing BrdU was analyzed using BD FACSCalibur TM system. Similar to tllna, LLNA:BrdU-FCM adopts a stimulation index (SI) value to distinguish skin sensitizers from non-sensitizers. The SI in the LLNA: BrdU-FCM is the ratio of the mean number of LNCs with incorporated BrdU from mice in each of the test substance dose groups to the mean number of LNCs with incorporated BrdU from mice in the vehicle control group Flow cytometry analysis B cell and T cell The LNCs ( /ml) were centrifuged (5 g) for 5 min in PBS and resuspended in staining buffer. For double staining, LNCs were incubated with 5 l of with hamster anti-mouse CD3e and rat anti-mouse CD45R/B22 in staining buffer for 15 min at 4 C, and washed again. Ten thousand cells were gated, and the number of the cells expressing CD45R/B22 and CD3e was analyzed using BD FACSCalibur TM system. The SI was defined as the increase in total B cells in the substance-treated group relative to that in the vehicle control group. An SI of 3 was defined as the cut-off value to identify sensitizers Measurement of cytokines (ELISA assays) Isolated lymphocytes were cultured in RPMI 164 (Sigma), 1% fetal bovine serum (Invitrogen), and 1 g/ml-1 U/ml penicillin streptomycin. Cell suspensions (5 1 6 cells/ml) were seeded into a collagen-coated 96-well tissue culture plate and cultured at 37 C in 5% CO 2 for 24 h. For cytokine assay, samples from culture supernatants (1 l) were used. The concentrations of cytokines were determined in duplicate using commercial Opt EIA Mouse set according to the manufacturer s instructions (BD Bioscience, Mississauga, Canada). OD was measured at 45 nm using a FLEX station3 (Molecular Devices, Sunnyvale, CA). Results were expressed as means ± standard deviation (SD) Statistics The EC3 value is defined as the estimated concentration that yields an SI value of 3. Among three methods of estimating EC3 values from the SI values, we used linear interpolation method (Basketter et al., 1999). The mean EC3 was estimated and classified into the appropriate chemical category (Gerberick et al., 24). Finally, the sensitivity, specificity and accuracy were calculated as the measure of relevance on the basis of the mean EC3 in order to assess the concordance of the LLNA:BrdU- FCM and LLNA:B-cell results with tllna or Guinea Pig Maximization Test/Buehler Test results (ICCVAM, 1999; OECD TG429, 21). The statistical significant of the differences between groups was determined by the one way analysis of the variance (ANOVA). In the ANOVA, when significant differences were detected, Dunnett s method as a post hoc test was used to compare treatment groups with the appropriate vehicle control group. All the statistical analyses were performed using SPSS (Statistical Package for the Social Sciences, version 13, SPSS Inc., Chicago, IL, USA) and P <.5 and P <.1 was considered significant.

3 K.-M. Jung et al. / Toxicology Letters 29 (212) Table 1 Test chemicals. No. Substance Abb. CAS No. tllna Guinea pig Human Concentration Vehicle EC3 (%) a Category b 1 5-Chloro-2-methyl-4-isothiazolin-3-one CMI DMF.9 Extreme + +.5,.1,.5 2 2,4-Dinitrochlorobenzene DNCB AOO.49 Extreme + +.1,.1, phenylenediamine PPD AOO.11 Strong + +.1, 1, 3 4 Cobalt chloride CC DMSO.6 Strong + +.5, 1, 5 5 Isoeugenol IE AOO 1.5 Moderate + + 1, 5, Mercaptobenzothiazole MBT AOO c 1.7 Moderate + + 1, 5, 1 7 Citral Citral AOO 9.2 Moderate + + 5, 1, 25 8 Hexyl cinnamic aldehyde HA AOO 9.7 Moderate + + 5, 1, 25 9 Eugenol EUG AOO 1.1 Weak + + 5, 1, 25 1 Phenyl benzoate PB AOO 13.6 Weak + + 5, 1, Cinnamic alcohol CA AOO 21 Weak + + 1, 25, 5 12 Imidazolidinyl urea IU DMF 24 Weak + + 1, 25, 5 13 Methyl methacrylate MMA AOO 9 Weak , 9, 1 14 Chlorobenzene CB AOO NA Negative 1, 5 15 Isopropanol IS AOO NA Negative + 1, 5 16 Lactic acid LA DMSO NA Negative 1, Methyl salicylate MS AOO NA Negative 1, Salicylic acid SA AOO NA Negative 1, Sodium lauryl sulfate SLS DMF 8.1 False positive d 5, 1, 15 2 Ethylene glycol dimethacrylate EGDM MEK 28 False positive e + 1, 25, 5 21 Xylene Xylene AOO 95.8 False positive f 9, 1 22 Nickel chloride NC % EtOH c NA False negative g + + 1, 5, 1 Abb., Abbreviations; CAS No, chemical abstracts service number; DMF N,N-dimethylformamide; AOO, acetone: olive oil (4:1, v/v); DMSO, dimethyl sulfoxide; MEK, methyl ethyl ketone; 7% EtOH, 7% ethanol; NA, not applicable since stimulation index <3. Presumed to be a non-sensitizer in humans based on the fact that no clinical patch test results could be located; not included in a patch test kit allergen; and no case reports of human sensitization could be located. Data not available. a EC3 = estimated concentration needed to produce a stimulation index of 3 in traditional LLNA. b The mean EC3 was classified into the appropriate chemical category (26). c The recommended vehicle for 2-Mercaptobenzothiazole and Nickel chloride is DMF and DMSO, but we used AOO and 7% EtOH, respectively in the current study. d Sodium lauryl sulpfate was categorized as nonsensitizer based on guinea pig and human data but as sensitizer by the traditional LLNA. e Ethylene glycol dimethacrylate was categorized as nonsensitizer in guinea pig and as sensitizer in human but as sensitizer by the traditional LLNA. f Xylene was categorized as nonsensitizer in human but as sensitizer by the traditional LLNA. g Nickel chloride was categorized as sensitizer based on guinea pig and human data but as nonsensitizer by the traditional LLNA. 3. Results 3.1. Evaluation of the performance of non-radioisotopic LLNA:BrdU-FCM with 18 reference substances OECD TG 429 recommends testing 18 minimum reference substances for a newly developed or modified LLNA method. Firstly, 18 reference substances from OECD TG 429 (Table 1) were evaluated using LLNA:BrdU-FCM (Table 1). We selected 3 concentrations appropriately to include the known LLNA EC3 values for sensitizers that were free from systemic toxicity and/or excessive local skin irritation. When the cutoff SI (stimulation index) value was set at 3, LLNA:BrdU-FCM could successfully differentiate 12 out of 13 sensitizers and 5 out of 5 non-sensitizers, accomplishing an overall accuracy of 94.4% (17 out of 18, Table 2). One discordant substance, methyl methacrylate (MMA) which is classified as a sensitizer in tllna, was defined as a non-sensitizer in the LLNA:BrdU-FCM (Fig. 1) Examination of the flow cytometric measurement of B cells and T cells Immunophenotyping of LNCs as B or T cells using B22+ or CD3e+ respectively with flow cytometry, was done in combination with LLNA:BrdU-FCM to investigate if sub-typing of LNCs could improve the accuracy. While the total numbers of both B and T cells increased in LNCs after exposure to a sensitizer, 4- phenylenediamine (PPD), ratios of B cells were substantially higher than those of T cells (Fig. 2A). The percentage of B cells in LNCs of the mice treated with PPD significantly increased to 28.1% and 4.% by.1% and 1.% PPD, respectively (17.1% in vehicle control). When we measured the increases of B cell and T cell population for 18 reference substances, the stimulation index (SI) estimated with B cell increase was better-correlated with the proliferation index (BrdU-FCM, r =.91), than those of B/T cell ratio (r =.738) or T cell (r =.882) (Fig. 2B). When an increase of more than 3 fold (SI value of 3) was set as cutoff, 12 out of 13 sensitizers and 5 out of 5 non-sensitizers were successfully defined, achieving an overall accuracy of 94.4% (17 out of 18). Like LLNA:BrdU-FCM, MMA was evaluated as a non-sensitizer(a false negative) in the LLNA:B-cell (Fig. 2C) Measurement of ex vivo cytokine release from LNCs We measured the production of pro-inflammatory cytokines from LNCs of PPD-treated mice to investigate if ex vivo cytokine production could be employed as a secondary endpoint for LLNA:BrdU-FCM. Among the measured pro-inflammatory cytokines, IL-2, IL-6, IFN-, MCP-1, GM-CSF and TNF- were increased by PPD in a dose-dependent manner with statistical significance (P <.1, Fig. 3A). To further examine if the ex vivo production of these cytokines could be employed for the differentiation of sensitizers from non-sensitizers or irritants, extreme to weak sensitizers (DNCB, isoeugenol, HCA and eugenol) and a nonsensitizer, isopropanol and an irritant, SLS were treated and the cytokine production was measured. As shown in Fig. 3B, IL-2, IL-6 and IFN- were significantly induced by sensitizers while they were not affected by isopropanol or SLS. Evaluation of ex vivo production

4 258 K.-M. Jung et al. / Toxicology Letters 29 (212) Table 2 Performance of the LLNA:BrdU-FCM and LLNA:B-cell compared with the tllna recommended reference substances. Substance tllna LLNA:BrdU-FCM LLNA:B-cell EC3 Category EC3 Category EC3 Category 1 5-Chloro-2-methyl-4-isothiazolin-3-one.9 Extreme <.1 Extreme <.1 Extreme 2 2,4-Dinitrochlorobenzene.4 Extreme.12 Extreme <.1 Extreme 3 4-phenylenediamine.1 Strong <.1 Strong <.1 Strong 4 Cobalt chloride.6 Strong <.5 Strong <.5 Strong 5 Isoeugenol 1.5 Moderate 2.5 Moderate 1. Moderate 6 2-Mercaptobenzothiazole 1.7 Moderate 2.6 Moderate 4.5 Moderate 7 Citral 9.2 Moderate 14.1 Weak 19.1 Weak 8 Hexyl cinnamic aldehyde 9.9 Moderate 5.4 Moderate 9.1 Moderate 9 Eugenol 1.1 Weak 2. Moderate 3.8 Moderate 1 Phenyl benzoate 13.6 Weak 11. Weak 18.6 Weak 11 Connamic alcohol 21 Weak 4.1 Weak 42.7 Weak 12 Imidazolidinyl urea 24 Weak 45.9 Weak 44.7 Weak 13 Methyl methacrylate 95.8 Weak Negative Negative 14 Chloronenzene Negative Negative Negative 15 Isopropanol Negative Negative Negative 16 Lactic acid Negative Negative Negative 17 Methyl salicylate Negative Negative Negative 18 Salicylic acid Negative Negative Negative Predictive capacity Sensitivity 92.3% (12/13) 92.3% (12/13) Specificity 1% (5/5) 1% (5/5) Accuracy 94.4% (17/18) 94.4% (17/18) Predictive capacity, sensitivity is the proportion of all positive substances that are classified as positive; specificity is the proportion of all negative substances that are classified as negative; accuracy is the proportion of correct outcomes (positive and negative) of a test method when compared to the results of tllna. of these cytokines for 18 reference substances demonstrated a high correlation between the fold increase of cytokine production and SI values of LLNA:BrdU-FCM (vs IL-2, r =.83; vs IL-6, r =.563; vs IFN-, r =.742, Fig. 4A C). When a statistically-significant increase of cytokines was set as the criteria for defining of sensitizers, IL-2, IL-6 and IFN- could differentiate sensitizers with the accuracies of 88.9%, 77.8% and 72.2%, respectively (Fig. 4D), reflecting that IL-2 was the most selective marker for the identification of sensitizers Comparison of the performances of LLNA:BrdU-FCM, LLNA:B-cells and LLNA:cytokine for 4 optional reference substances In addition to 18 minimum reference substances, OECD recommends testing 4 optional reference substances for a new or modified LLNA, that are defined as either false positives or false negative in tllna (Table 3). We further evaluated the performances of BrdU incorporation, B cell and ex vivo cytokine production with A 4 Low 2 Extreme/Strong Middle High B Moderate Stimulation index CMI DNCB PPD CC Stimulation index IE MBT Citral HCA C 2 9 D Stimulation index Weak Stimulation index 6 3 Non-sensitizer EUG PB CA IU MMA CB IS LA MS SA Fig. 1. Results of LLNA:BrdU-FCM for 18 minimum reference substances. (A) Extreme and strong sensitizer CMI (.5%,.1%,.5% in DMF), DNCB (.1%,.1%,.5% in AOO), PPD (.1%, 1.%, 3.% in AOO) and CC (.5%, 1.%, 5.% in DMSO); (B) moderate sensitizer IE (1%, 5%, 1% in AOO), MBT (1%, 5%, 1% in AOO), Citral (5%, 1%, 25% in AOO) and HCA (5%, 1%, 25% in AOO); (C) weak sensitizer EUG (5%, 1%, 25% in AOO), PB (5%, 1%, 25% in AOO), CA (1%, 25%, 5% in AOO), IU (1%, 25%, 5% in DMF) and MMA (75%, 9%, 1% in AOO); (D) negative CB (1%, 25% in AOO), IS (1%, 25% in AOO), LA (1%, 25% in DMSO), MS (1%, 25% in AOO) and SA (1%, 25% in AOO).

5 K.-M. Jung et al. / Toxicology Letters 29 (212) Fig. 2. Immuno-phenotyping of B cell and T cell in isolated LNCs using flow cytometry. (A) Flow cytometric analysis of the draining lymph nodes of mice treated with PPD. The left three plots, vehicle (AOO),.1% and 1.% PPD, respectively are CD45R/B22 versus CD3e plots. Single cell suspensions were prepared from the draining lymph nodes for flow cytometry 72 h following the final treatment. The histogram in the right is the comparison between vehicle and PPD 1.% for CD3e and CD45R/B22; (B) correlation of stimulation index of LLNA:BrdU-FCM versus B/T ratio (r value =.738), stimulation index of T cell (r value =.882) and B cell (r value =.91); (C) stimulation index of LLNA:B-cell for 18 recommended reference substances. The values shown are the means ± S.D (n = 4 6). these 4 optional reference substances. In LLNA:BrdU-FCM, false positives (SLS, ethylene glycol dimethacrylate (EGDM) and xylene) and false negative, nickel chloride (NC) in tllna were all determined to be positives. On the other hand, B cell increase (SI > 3.) determined EGDM as non-sensitizer while SLS, xylene and NC, as sensitizers. Ex vivo IL-2 assay (statistically significant increase, P <.5) defined EGDM, NC and xylene as sensitizers while SLS was defined as a non-sensitizer Comparison of the endpoints of non-radioisotopic LLNA and tllna with guinea pig and human data When the results for the LLNA:BrdU-FCM, LLNA:B-cell and IL-2 were compared with tllna, LLNA:BrdU-FCM had a higher concordance to tllna (9.9%) than LLNA:B-cell (86.4%) and ex vivo IL-2 (81.8%). When the results from the three endpoints of nonradioisotopic LLNA and tllna were compared with the known Table 3 Results of the 4 optional materials. Substance tllna BrdU-FCM B-cell ex vivo cytokine IL-2 IL-6 IFN- Sodium lauryl sulfate Moderate a (8.1%) b Moderate (5.4%) Moderate (5.2%) c Ethylene glycol dimethacrylate Weak (28.%) Weak (44.2%) Negative d Xylene Weak (95.8%) Weak (<9.%) Weak (1.%) e Nickel chloride Negative Moderate (3.4%) Moderate (4.6%) a The mean EC3 was classified into the appropriate chemical category (26). b Mean EC3. c No significant change. d Statistically significant increase, p <.1. e Statistically significant increase, p <.5.

6 26 K.-M. Jung et al. / Toxicology Letters 29 (212) A Cytokine release (% of control) IL-2 IL-4 IL-6 IL-12 IFN-γ MCP-1 GM-CSF AOO PPD.1% PPD 1% PPD 3% TNFα B Cytokine release (% of control) 5 DNCB IE HCA 4 EUG IS SLS IL-2 IL-6 IFN-r MCP-1 GM-CSF TNF-a Fig. 3. Ex vivo cytokine production from isolated LNCs. (A) The ex vivo production of cytokines from isolated LNCs from the draining lymph nodes of mice treated daily with vehicle and PPD (.1%, 1.% and 3.%) over 3 consecutive days. The auricular lymph nodes were excised approximately 72 h after the final treatment; (B) each cytokine levels in lymph nodes of mice treated with DNCB.5%, IE 1%, HCA 25%, EUG 25% as sensitizer, IS 25% as a non-sensitizer and SLS 15% as an irritant. The values shown are the means ± S.D. (n = 5). P <.5 and P <.1 compared to vehicle. results from guinea pig, LLNA:BrdU-FCM and ex vivo IL-2 had accuracies comparable to tllna (85.7% [18/21]) while LLNA:B-cell had a higher accuracy rate (9.5% [19/21]) (Table 4). When compared with known human data, LLNA:BrdU-FCM, ex vivo IL-2 and tllna exhibited comparable accuracies (8% [16/2]) while B cell increase showed a lower accuracy. In addition, despite of similar accuracies, LLNA:BrdU-FCM and tllna had a higher sensitivity than ex vivo IL-2 (87.5% vs 81.3%) while ex vivo IL-2 had a higher specificity than LLNA:BrdU-FCM or tllna (75% vs 5%). 4. Discussion Previously, we have developed and pre-validated LLNA:BrdU- FCM method with 7 reference standards through comparison with LLNA:BrdU-IHC (Jung et al., 21). In the present study, we further examined the performance of LLNA:BrdU-FCM with all 22 recommended reference substances enlisted in OECD TG429 (21). The performance standard in the guideline states that 18 minimum and 4 optional reference substances should be tested for evaluation of newly proposed or modified LLNA methods. Of 18 minimum reference substances, LLNA:BrdU-FCM could successfully differentiate 12 out of 13 sensitizers and 5 out of 5 non-sensitizers, producing overall concordance of 94.4% (17 out of 18, Table 2) to tllna. Of 4 optional reference substances, SLS, EGDM and xylene are false positives in tllna and NC, a false negative. These 4 optional reference substances including the false negative, NC, were all determined to be sensitizers in LLNA:BrdU-FCM, alluding a certain difference of LLNA:BrdU-FCM from tllna. Although other possibilities such as difference of strain (Balb/C in the current study vs CBA/J or CBA/JN in tllna) or biological variance between the individual assays (Basketter et al., 211) may not be excluded, we could find that the discrepancy may stem from vehicle selection. For NC, OECD guideline recommends DMSO as vehicle but in our preliminary solubility test, it was insoluble in DMSO, indicating that the sensitizing potential of NC could not be properly evaluated with DMSO. Therefore, we screened other applicable vehicles for NC and found that Table 4 Predictive capacity of the non-radioactive LLNA compared with tllna, guinea pig and human data. Predictive capacity tllna a Guinea pig a Human a S = 16, N = 6 S = 14, N = 7 S = 16, N = 4 BrdU B cell Cytokine tllna BrdU B cell Cytokine tllna BrdU B cell Cytokine Sensitivity No. 15/16 14/16 13/16 13/14 13/14 13/14 12/14 14/16 14/16 13/16 13/16 % Specificity No. 5/6 5/6 5/6 5/7 5/7 6/7 6/7 2/4 2/4 2/4 3/4 % Accuracy No. 2/22 19/22 18/22 18/21 18/21 19/21 18/21 16/2 16/2 15/2 16/2 % a The category data based on OECD TG 429. S, sensitizer; N, non-sensitizer. Data for xylene is not given for guinea pig, and those for CB and LA are not provided for human by OECD TG 429.

7 K.-M. Jung et al. / Toxicology Letters 29 (212) Fig. 4. Evaluation of ex vivo cytokine production as an endpoint for the determination of sensitizers. (A C) Correlation of stimulation index of LLNA:BrdU-FCM versus IL-2(A, r value =.83), IL-6(B, r value =.563) and IFN- (C, r =.742); (D) evaluation and predictive capacity of ex vivo cytokine production. The accuracy, sensitivity and specificity of IL-2 vs tllna were found to be 88.9%, 84.6% and 1%, respectively. The IL-6 and IFN- had an accuracy of 77.8% and 72.2%, a sensitivity of 84.6% and 61.5%, specificity of 6% and 1%, respectively. Statistically significant difference with P <.1 between substance-treated mice and vehicle-treated mice. Statistically significant difference with P <.5 is no significant. 7% ethanol could dissolve NC to a complete solution, which has been adopted for the assay. In 7% ethanol, NC was determined to be a moderate sensitizer with SI values of 7.91 at 1%. In this regard, we suggest that NC should be tested in tllna again with a soluble vehicle to clarify the sensitization potential of NC. Issue of vehicle selection was also encountered with MBT. The recommended vehicle for MBT in OECD TG429 is N,Ndimethylformamide (DMF). When DMF was used as recommended, we could not get a positive result up to the highest soluble concentration of MBT. However, when we tested MBT with AOO, MBT was determined to be a sensitizer (SI of 4.5 at 1%, a moderate sensitizer). Indeed, this result was in line with the study of van Och et al. (2). These results highlight the importance of vehicle selection in LLNA again, which has been continuously pointed out in the variants of LLNA (Basketter et al., 21). Other difference from tllna can be found with MMA which was determined to be a non-sensitizer in LLNA:BrdU-FCM. MMA is a well-known sensitizer in human (Borak et al., 211), and in tllna performance standard (OECD TG429, 21), it was defined as a weak sensitizer. However, a recent report of Basketter et al. (211) also showed that MMA can be defined as a non-sensitizer in tllna, suggesting that MMA is a marginally weak sensitizer and it may yield equivocal results depending on the condition of respective experiments. In our study, MMA was consistently defined as a non-sensitizer by B cell increase and IL-2 production indices. Interestingly, LLNA:DA, another non-radioisotopic LLNA endorsed by OECD TG442A (21), also defines MMA as a false negative, suggesting that the issue of MMA might be shared by other nonradioisotopic LLNA methods although further comparison should be necessary to confirm it. Generally, allergic contact dermatitis is considered as T cell mediated immune response but some studies have found a significant increase in the proliferation of B cells in draining lymph nodes as well after exposure to sensitizers (Gerberick et al., 22; Manetz et al., 21). Gerberick et al. reported a significant and sensitizerspecific increase in the ratio of B22 + B cells in LNCs. Manetz and Meade reported that activated B cell population increases in both IgE-inducing and T cell-mediated hypersensitivity responses whereas it is not affected by irritants. B cells increase in draining lymph nodes after chemical sensitization, most probably owing to the proinflammatory cytokines produced by activated T cells (Neumann et al., 1992). Consequence of B cell increase in sensitization, however, is still controversial. B cells may execute antibody responses to hapten by producing anti-haptenic antibody (Thomas et al., 1976). On the other hand, B cells may contribute to the local tolerance to sensitizers by inducing suppressive regulatory

8 262 K.-M. Jung et al. / Toxicology Letters 29 (212) T cells (Singh et al., 28). In our study, the percentage of B22 + cells increased in a dose-dependent manner by PPD which was in a good agreement with the previous reports. With the 22 reference substances, we could find that B cell increase was bettercorrelated with proliferation index (BrdU-FCM, r =.889) than B/T cell ratio (r =.735) or T cell (r =.86). With the SI 3. criteria, LLNA:BrdU-FCM and LLNA:B-cell exhibited same performances for 18 minimum reference substances, i.e., sensitivity of 92.3%, specificity of 1% and accuracy of 94.4% with 17 out of 18 substances in agreement with tllna. For 4 optional substances, SLS, xylene and NC were determined to be positive while EGDM was negative in LLNA:B-cell, which was different from the result of LLNA:BrdU- FCM where all 4 substances including EGDM were determined to be positive. Actually, EGDM is a known sensitizer in human (Basketter et al., 2; Isaksson et al., 25) but it was negative in the guinea pig test (ICCVAM, 1999). It is yet unclear why this discrepancy occurred between these endpoints but we believe that further future studies will be enlightening for the role of B cell proliferation in the manifestation of contact allergy. Pro-inflammatory, T H 1- and T H 2-type cytokines play key roles in the initiation and regulation of the immune response in the skin sensitization (Ku et al., 28; Naisbitt et al., 2; Ravel et al., 24; Van Och et al., 22; Vandebriel et al., 2, 23). In our study, we evaluated the production of pro-inflammatory (IL-6, IL-12, MCP-1, TNF-, GM-CSF), T H 2 (IL-4) and T H 1 (IL-2, IFN- ) cytokines. In the PPD-treated mice, all of the 8 cytokine significantly increased, but 6 cytokines including IL-2, IL-6, IFN-, MCP-1, TNF- and GM-CSF were the most pronounced. Further experiments with more reference substances confirmed that 3 cytokines, IL-2, IL-6 and IFN- were selective and sensitive to sensitizers. This finding is well in line with the previous reports that cytokines including IL-2, IL-6 and IFN- are profusely produced during the elicitation phase in contact hypersensitivity (Hope et al., 1994; Thomson et al., 1993). Of these 3 cytokines, IL-2 exhibited the highest accuracy for the identification of sensitizers. Previously, Azam et al. (25) and Ku et al. (28) found that IL-2 mrna and protein may be used for a marker for the detection of sensitizers. However, a small number of substances were tested in those studies. Here, we could demonstrate that even with more diverse substances, IL-2 may be a useful marker for detection of sensitizers, indeed. IL-2 is secreted by activated T cells and is essential for the expansion of antigen-selective T-cell clones, contributing to both T H 1 and T H 2 cell responses (Ku et al., 28). Furthermore, IL-2 promotes the proliferation of antigen-responsive cytotoxic T-cells (Stern and Smith, 1986) and regulatory T-cells (Sakaguchi et al., 1995), playing an important role for development of T cell immunologic memory. IFN- represents T H 1-type cytokine along with IL-2. IFN- production by LNCs following sensitization is considered important for the development of effecter T cells, which is critical for the accomplishment of contact dermatitis (He et al., 21). IL-6 is released principally from sensitized dendritic cells (Hope et al., 1994, 2) and plays an important role in the acute and later phases of inflammation. It induces the host-defense responses against pathogen, stimulates the differentiation of B cells into immunoglobulin-producing cells (Hirano et al., 1986) and antagonizes the reconstitution of regulatory T cells (Chen et al., 29), driving pro-inflammatory reaction to sensitizers. In contrast, the change in IL-4 level, a representative T H 2 type cytokine was only minimal in our study. Contact dermatitis represents Type IV hypersensitivity reaction which is mainly orchestrated by T H 1 cells (Aeby et al., 24), explaining these cytokine profiles well. In conclusion, by providing a comprehensive assay result for the full 22 reference substances enlisted in OECD TG429, this study demonstrated that the evaluation of proliferation in the draining lymph node using BrdU incorporation with flow cytometry can provide a good non-radioisotopic endpoint replacing the traditional radioisotopic tllna. Additionally, we showed that the stimulation index using B cell or ex vivo cytokine releases may be useful as secondary or independent endpoints for non-radioisotopic LLNA that can help discriminate sensitizers more accurately. It is a great advantage over the conventional guinea-pig method to be able to obtain multiple, quantitative and non-radioisotopic secondary endpoints in the same animals without sacrificing extra animals. Thus, we propose that our non-radioisotopic LLNA:BrdU-FCM using BrdU incorporation, B cell increase and ex vivo cytokine production may be useful for assessing the allergic potential of chemicals. Conflict of interest The authors have declared no conflict. Acknowledgements This research was supported by grants ( KFDA58 and KFDA575) from Korean Food & Drug Administration. References Aeby, P., Wyss, C., Beck, H., Griem, P., Scheffler, H., Goebel, C., 24. Characterization of the sensitizing potential of chemicals by in vitro analysis of dendritic cell activation and skin penetration. J. Invest. Dermatol. 122, Azam, P., Peiffer, J.L., Ourlin, J.C., Bonnet, P.A., Tissier, M.H., Vian, L., Fabre, I., 25. Qualitative and quantitative evaluation of a local lymph node assay based on ex vivo interleukin-2 production. Toxicology 26, Basketter, D., Kolle, S.N., Schrage, A., Honarvar, N., Gamer, A.O., Ravenzwaay, B., Landsiedel, R., 211. Experience with local lymph node assay performance standards using standard radioactivity and nonradioactive cell count measurements. J Appl Toxicol. Basketter, D.A., Balikie, L., Dearman, R.J., Kimber, I., Ryan, C.A., Gerberick, G.F., Harvey, P., Evans, P., White, I.R., Rycroft, R.J., 2. Use of the local lymph node assay for the estimation of relative contact allergenic potency. Contact Dermatitis 42, Basketter, D.A., Gerberick, G.F., Kimber, I., 21. Skin sensitisation, vehicle effects and the local lymph node assay. Food Chem. Toxicol. 39, Basketter, D.A., Lea, L.J., Dickens, A., Briggs, D., Pate, I., Dearman, R.J., Kimber, I., A comparison of statistical approaches to the derivation of EC3 values from local lymph node assay dose responses. J. Appl. Toxicol. 19, Bilsland, J.G., Haldon, C., Goddard, J., Oliver, K., Murray, F., Wheeldon, A., Cumberbatch, J., McAllister, G., Munoz-Sanjuan, I., 26. A rapid method for the quantification of mouse hippocampal neurogenesis in vivo by flow cytometry validation with conventional and enhanced immunohistochemical methods. J. Neurosci. Methods 157, Borak, J., Fields, C., Andrews, L.S., Pemberton, M.A., 211. Methyl methacrylate and respiratory sensitization: a critical review. Crit. Rev. Toxicol. 41, Boussiquet-Leroux, C., Durand-Cavagna, G., Herlin, K., Holder, D., Evaluation of lymphocyte proliferation by immunohistochemistry in the local lymph node assay. J. Appl. Toxicol. 15, Chen, X., Das, R., Komorowski, R., Beres, A., Hessner, M.J., Mihara, M., Drobyski, W.R., 29. Blockade of interleukin-6 signaling augments regulatory T-cell reconstitution and attenuates the severity of graft-versus-host disease. Blood 114, Dearman, R.J., Basketter, D.A., Kimber, I., 1996a. Characterization of chemical allergens as a function of divergent cytokine secretion profiles induced in mice. Toxicol. Appl. Pharmacol. 138, Dearman, R.J., Betts, C.J., Humphreys, N., Flanagan, B.F., Gilmour, N.J., Basketter, D.A., Kimber, I., 23. Chemical allergy: considerations for the practical application of cytokine profiling. Toxicol. Sci. 71, Dearman, R.J., Moussavi, A., Kemeny, D.M., Kimber, I., 1996b. Contribution of CD4+ and CD8+ T lymphocyte subsets to the cytokine secretion patterns induced in mice during sensitization to contact and respiratory chemical allergens. Immunology 89, Gerberick, G.F., Cruse, L.W., Ryan, C.A., Local lymph node assay: differentiating allergic and irritant responses using flow cytometry. Methods 19, Gerberick, G.F., Cruse, L.W., Ryan, C.A., Hulette, B.C., Chaney, J.G., Skinner, R.A., Dearman, R.J., Kimber, I., 22. Use of a B cell marker (B22) to discriminate between allergens and irritants in the local lymph node assay. Toxicol. Sci. 68, Gerberick, G.F., Ryan, C.A., Dearman, R.J., Kimber, I., 27. Local lymph node assay (LLNA) for detection of sensitization capacity of chemicals. Methods 41, Gerberick, G.F., Ryan, C.A., Kern, P.S., Dearman, R.J., Kimber, I., Patlewicz, G.Y., Basketter, D.A., 24. A chemical dataset for evaluation of alternative approaches to skin-sensitization testing. Contact Dermatitis 5, He, B., Munson, A.E., Meade, B.J., 21. Analysis of gene expression induced by irritant and sensitizing chemicals using oligonucleotide arrays. Int. Immunopharmacol. 1,

9 K.-M. Jung et al. / Toxicology Letters 29 (212) Hirano, T., Yasukawa, K., Harada, H., Taga, T., Watanabe, Y., Matsuda, T., Kashiwamura, S., Nakajima, K., Koyama, K., Iwamatsu, A., et al., Complementary DNA for a novel human interleukin (BSF-2) that induces B lymphocytes to produce immunoglobulin. Nature 324, Hope, J.C., Campbell, F., Hopkins, S.J., 2. Deficiency of IL-2 or IL-6 reduces lymphocyte proliferation, but only IL-6 deficiency decreases the contact hypersensitivity response. Eur. J. Immunol. 3, Hope, J.C., Dearman, R.J., Kimber, I., Hopkins, S.J., The kinetics of cytokine production by draining lymph node cells following primary exposure of mice to chemical allergens. Immunology 83, ICCVAM, The Murine Local Lymph Node Assay: A Test Method for Assessing the Allergic Contact Dermatitis Potential of Chemicals/Compounds. Inman, M.T., A monaural paediatric chest stethoscope for three pence threefarthings. Anaesthesia 21, Isaksson, M., Lindberg, M., Sundberg, K., Hallander, A., Bruze, M., 25. The development and course of patch-test reactions to 2-hydroxyethyl methacrylate and ethyleneglycol dimethacrylate. Contact Dermatitis 53, Jung, K.M., Bae, I.H., Kim, B.H., Kim, W.K., Chung, J.H., Park, Y.H., Lim, K.M., 21. Comparison of flow cytometry and immunohistochemistry in nonradioisotopic murine lymph node assay using bromodeoxyuridine. Toxicol. Lett. 192, Ku, H.O., Jeong, S.H., Kang, H.G., Pyo, H.M., Cho, J.H., Son, S.W., Kim, H.R., Lee, K.J., Ryu, D.Y., 28. Intracellular expression of cytokines and granzyme B in auricular lymph nodes draining skin exposed to irritants and sensitizers. Toxicology 25, Lee, J.K., Park, J.H., Park, S.H., Kim, H.S., Oh, H.Y., 22. A nonradioisotopic endpoint for measurement of lymph node cell proliferation in a murine allergic contact dermatitis model, using bromodeoxyuridine immunohistochemistry. J. Pharmacol. Toxicol. Methods 48, Manetz, T.S., Pettit, D.A., Meade, B.J., 21. The determination of draining lymph node cell cytokine mrna levels in BALB/c mice following dermal sodium lauryl sulfate, dinitrofluorobenzene, and toluene diisocyanate exposure. Toxicol. Appl. Pharmacol. 171, Naisbitt, D.J., Gordon, S.F., Pirmohamed, M., Park, B.K., 2. Immunological principles of adverse drug reactions: the initiation and propagation of immune responses elicited by drug treatment. Drug Saf. 23, Neumann, C.M., Oughton, J.A., Kerkvliet, N.I., Anti-CD3-induced T-cell activation in vivo I. Flow cytometric analysis of dose-responsive, time-dependent, and cyclosporin A-sensitive parameters of CD4+ and CD8+ cells from the draining lymph nodes of C57Bl/6 mice. Int. J. Immunopharmacol. 14, OECD TG46, OECD guideline for the testing of chemicals: skin sensitization (Guinea Pig Maximization Test; Buehler Test). OECD TG429, 21. OECD guideline for the testing of chemicals-skin sensitization: local lymph. OECD TG442A, 21. Skin sensitization local lymph node assay: DA. Ravel, G., Christ, M., Horand, F., Descotes, J., 24. Cytokine release does not improve the sensitivity and specificity of the direct popliteal lymph node assay. Toxicology 2, Sakaguchi, S., Sakaguchi, N., Asano, M., Itoh, M., Toda, M., Immunologic selftolerance maintained by activated T cells expressing IL-2 receptor alpha-chains (CD25). Breakdown of a single mechanism of self-tolerance causes various autoimmune diseases. J. Immunol. 155, Singh, A., Carson W.F.t., Secor Jr., E.R., Guernsey, L.A., Flavell, R.A., Clark, R.B., Thrall, R.S., Schramm, C.M., 28. Regulatory role of B cells in a murine model of allergic airway disease. J. Immunol. 18, Stern, J.B., Smith, K.A., Interleukin-2 induction of T-cell G1 progression and c-myb expression. Science 233, Suda, A., Yamashita, M., Tabei, M., Taguchi, K., Vohr, H.W., Tsutsui, N., Suzuki, R., Kikuchi, K., Sakaguchi, K., Mochizuki, K., Nakamura, K., 22. Local lymph node assay with non-radioisotope alternative endpoints. J. Toxicol. Sci. 27, Takeyoshi, M., Noda, S., Yamasaki, K., Kimber, I., 26. Advantage of using CBA/N strain mice in a non-radioisotopic modification of the local lymph node assay. J. Appl. Toxicol. 26, 5 9. Takeyoshi, M., Yamasaki, K., Yakabe, Y., Takatsuki, M., Kimber, I., 21. Development of non-radio isotopic endpoint of murine local lymph node assay based on 5- bromo-2 -deoxyuridine (BrdU) incorporation. Toxicol. Lett. 119, Thomas, W.R., Asherson, G.L., Watkins, M.C., Reaginic antibody produced in mice with contact sensitivity. J. Exp. Med. 144, Thomson, J.A., Troutt, A.B., Kelso, A., Contact sensitization to oxazolone: involvement of both interferon-gamma and interleukin-4 in oxazolone-specific Ig and T-cell responses. Immunology 78, van Och, F.M., Slob, W., de Jong, W.H., Vandebriel, R.J., van Loveren, H., 2. A quantitative method for assessing the sensitizing potency of low molecular weight chemicals using a local lymph node assay: employment of a regression method that includes determination of the uncertainty margins. Toxicology 146, Van Och, F.M., Van Loveren, H., De Jong, W.H., Vandebriel, R.J., 22. Cytokine production induced by low-molecular-weight chemicals as a function of the stimulation index in a modified local lymph node assay: an approach to discriminate contact sensitizers from respiratory sensitizers. Toxicol. Appl. Pharmacol. 184, Vandebriel, R.J., De Jong, W.H., Hendriks, J.J., Van Loveren, H., 23. Impact of exposure duration by low molecular weight compounds on interferon-gamma and interleukin-4 mrna expression and production in the draining lymph nodes of mice. Toxicology 188, Vandebriel, R.J., De Jong, W.H., Spiekstra, S.W., Van Dijk, M., Fluitman, A., Garssen, J., Van Loveren, H., 2. Assessment of preferential T-helper 1 or T-helper 2 induction by low molecular weight compounds using the local lymph node assay in conjunction with RT-PCR and ELISA for interferon-gamma and interleukin-4. Toxicol. Appl. Pharmacol. 162,