Toxicology Letters Comparison of Flow Cytometry And

Toxicology Letters 192 (2010) 229–237

Contents lists available at ScienceDirect

Toxicology Letters

journal homepage: www.elsevier.com/locate/toxlet

Comparison of ﬂow cytometry and immunohistochemistry in non-radioisotopic murine lymph node assay using bromodeoxyuridine

Kyoung-Mi Jung a, Il-Hong Bae a, Bae-Hwan Kim b, Wang-Ki Kim a, Jin-Ho Chung c, Young-Ho Park a, Kyung-Min Lim a,∗ a Amorepaciﬁc Corporation R&D Center, Yongin-si 446-729, Republic of Korea b Department of Public Health, Keimyung University, Republic of Korea c College of Pharmacy, Seoul National University, Seoul 151-742, Republic of Korea article info abstract

Article history: Non-radioisotopic local lymph node assay (LLNA) employing 5-bromo-2-deoxyuridine (BrdU) with ﬂow Received 25 September 2009 cytometry (FACS) or immunohistochemistry (IHC) is gaining attention due to a regulatory issue of using Received in revised form 19 October 2009 radioisotope, 3H-thymidine, in vivo in traditional LLNA. In this study, to compare the performance of Accepted 21 October 2009 these non-radioisotopic endpoints, 7 chemicals with known sensitizing potencies were examined in Available online 30 October 2009 LLNA. Mice were topically treated with chemicals or vehicle on both ears for 3 days. After intraperitoneal injection of BrdU, bilateral lymph nodes were isolated separately and undergone respectively, FACS or Keywords: IHC to determine BrdU incorporated lymph node cells (LNCs). Weight and histology of treated ears were Non-radioisotopic local lymph node assay 5-Bromo-2-deoxyuridine also examined to evaluate chemical-induced edema and irritation. Both FACS and IHC could successively Flow cytometry identify the skin sensitizers from non-sensitizers. Comparison of FACS and IHC with traditional LLNA Immunohistochemistry revealed that FACS has a higher sensitivity although both assays produced comparable sensitivity and performance to traditional LLNA. In conclusion, non-radioisotopic LLNA using FACS and IHC can successfully detect sensitizers with a good correlation to traditional LLNA. Notably, FACS showed almost equivalent sensitivity and accuracy to traditional LLNA. © 2009 Elsevier Ireland Ltd. All rights reserved.

1. Introduction regulation limiting the introduction of radioisotopes into animal in vivo, mandating a heavy ventilation, filtering and completely iso- Mouse local lymph node assay (LLNA) is a validated alternative lated animal housing facilities. Accordingly, traditional LLNA has method for testing the sensitization potential of chemicals, replac- not been generalized widely, raising an urgent need for the devel- ing the conventional guinea pig maximization tests (Farrell et al., opment and the validation of non-radioisotopic endpoints in LLNA. 2009). This method offers important advantages over guinea pig Recently, to avoid the use of 3H-thymidine in vivo and to intro- assays for the animal welfare in terms of reduction and refine- duce non-radioisotopic endpoints in LLNA, many attempts have ment without deteriorating the assay quality, leading to a wide been made. These include the examination of the phenotype of pro- acceptance of LLNA as a representative alternative method for the liferating lymphocyte subsets (Gerberick et al., 1999), the indirect identification of chemicals with a skin sensitizing potential. estimation of lymphocyte proliferation by measuring intracellu- In LLNA, the skin sensitization potential is determined by mea- lar ATP content of lymph nodes (Idehara et al., 2008) and the suring lymphocyte proliferation in the draining auricular lymph assay of ex vivo cytokines production by draining lymph node cells nodes in response to the treated chemicals (Gerberick et al., 2007). (LNCs) (Dearman et al., 1999, 1994; Ku et al., 2008; van den Berg Radiolabeling of proliferating lymphocytes using 3H-thymidine has et al., 2005). Of these approaches, the measurement of 5-bromo- been commonly employed, however many countries have a strict 2-deoxyuridine (BrdU) incorporation into DNA (Takeyoshi et al., 2001) is gathering a huge interest, owing to its similarity to the conventional LLNA method employing 3H-thymidine. Abbreviations: LLNA, local lymph node assay; FACS, flow cytometry; PPD, BrdU is a non-radioisotopic analog of thymidine, which is p-phenylenediamine; DNCB, 2,4-dinitrochlorobenzene; IS, isopropylalcohol; SLS, incorporated into DNA during the S-phase of the cell cycle. It sodium lauryl sulfate; BrdU, 5-bromo-2-deoxyuridine; IHC, immunohistochem- has substituted 3H-thymidine labeling in many biological assays istry; HCA, hexylcinnamaldehyde. ∗ measuring cellular proliferation (Porstmann et al., 1985). To mea- Corresponding author at: Amorepacific Corporation R&D Center, 314-1 Bora- sure BrdU incorporation into lymph node, many antibody-based dong, Giheung-gu, Yongin-si 446-729, Republic of Korea. Tel.: +82 31 280 5904; fax: +82 31 281 8390. assay methods are available including flow cytometric analysis E-mail address: kimlim@amorepacific.com (K.-M. Lim). (FACS) (Suda et al., 2002), immunohistochemical staining (IHC)

Table 1 Tested chemicals and their reported sensitization potency.

Substance name (abbreviation) Vehicle Potency category EC3a (%) Test concentration (%)

2,4-Dinitrochlorobenzene (DNCB) AOO Extreme 0.04 0.01, 0.1, 0.25, 0.5 4-Phenylenediamine (PPD) AOO Strong 0.11 0.1, 1, 3 Isoeugenol AOO Moderate 1.5 2.5, 5, 10 Hexylcinnamaldehyde (HCA) AOO Moderate 9.9 5, 10, 25 Eugenol AOO Weak 10.1 5, 10, 25 Isopropylalcohol (IS) AOO Negative – 50 Sodium lauryl sulfate (SLS) 50% ethanol /DMF Moderate (false positive) 8.1 5, 10, 15

a Values from ICCVAM 2009, Recommended Performance Standards: Murine Local Lymph Node Assay (2009, NIH Publication No. 09-7357).

Fig. 1. Changes in auricular lymph node weights and the number of LNCs (lymph node cells). (A) Representative photograph of auricular lymph node from the treated mouse. Bar indicates 5 mm. (B) After mice were treated with 0.1, 1 and 3% PPD, 0.25% DNCB, 15% SLS or 50% IS for 3 days, auricular lymph node was collected on Day 5 and weighed. *Signiﬁcant difference from vehicle group, p < 0.05, values are mean ± S.D. (N = 5 or 6). (C) The number of LNC (×106 cells) in an auricular lymph node from treated mice. K.-M. Jung et al. / Toxicology Letters 192 (2010) 229–237 231

Fig. 2. Swelling and inflammation of ear. (A) After mice were treated with 0.1%, 1% and 3% PPD, 0.25% DNCB, 15% SLS or 50% IS for 3 days, 6 mm ear biopsy was obtained on Day 5 and weighed to evaluate ear swelling. *Significant difference from vehicle group, p < 0.05, values are mean ± S.D. (N = 5 or 6). (B) Representative microscopic photograph of H&E-stained ear tissue from treated mice. Original magnification ×100, bar indicates 200 ␮m.

(Boussiquet-Leroux et al., 1995) and ELISA (Takeyoshi et al., 2003; 2. Materials and methods Yamano and Shimizu, 2009). ELISA-based immunoassay method is developed first for the measurement of BrdU incorporation 2.1. Chemicals and reagents and accordingly a validation study has been conducted earliest 2,4-Dinitrochlorobenzene (DNCB), isopropylalcohol (IS), isoeugenol, hexylcin- (ICCVAM, 2009a). However, FACS or IHC method is also enjoy- namaldehyde (HCA), eugenol, sodium lauryl sulfate (SLS), p-phenylenediamine ing a large interest since they are highly sensitive and specific to (PPD) and 5-bromo-2-deoxyuridine (BrdU) were obtained from Sigma–Aldrich (San BrdU incorporated lymphocyte population. Especially, FACS allows Diego, CA). PPD, DNCB, HCA, isoeugenol, eugenol and IS was dissolved in acetone:olive oil (AOO; 4:1) and SLS was dissolved in 50% ethanol or DMF. BrdU was a rapid analysis of many extra parameters such as surface markers, dissolved in phosphate-buffered saline (PBS) at a concentration of 20 mg/ml. cell counts or lymphocyte population identifications at one scope, providing a useful methodology and advantage for the screening 2.2. Animal housing and experimental protocol and characterization of potential sensitizers. In this study, to compare the performance of FACS and IHC, Both the animal care and the study protocol were conducted in accordance with 7 chemicals selected from recommended performance standard Institutional Animal Care and Use Committee (IACUC) of Amorepacific R&D center. ◦ chemicals of ICCVAM 2009, p-phenylenediamine (PPD), 2,4- Animals were kept under controlled conditions of temperature (23 ± 3 C) and rel- ± ◦ dinitrochlorobenzene (DNCB), isoeugenol, hexylcinnamaldehyde ative humidity (50 10 C) with alternating 12 h light and dark cycle. Throughout the study, animals had ad libitum access to tab water. Female BALB/c mice (7–8 (HCA), eugenol, sodium lauryl sulfate (SLS) and isopropylalcohol weeks old, body weight 18–22 g) were purchased from Orient Bio (Seoul, Korea), (IS) were examined in LLNA where bilateral auricular lymph nodes and were used in all experiments. Before experiments, animals were kept for at were separately isolated and undergone FACS or IHC respectively. least 1 week on laboratory solid diet (Purina Co., Korea) for acclimation. The mouse In addition, ear inflammation and lymph node changes by chemi- lymph node assay was performed according to the method of Takeyoshi et al. (2003) with minor modifications. Groups of mice (N = 5 or 6) were treated with 25 ␮lof cal treatment were evaluated by observing the morphology and the the various concentrations of test chemicals (Table 1) or vehicle on the dorsal area histology of ear and lymph node after chemical treatments. Finally, of both ears daily for 3 consecutive days (Days 0, 1 and 2). On Day 4, mice were the results from the respective assay methods were compared intraperitoneally administered with BrdU and mice were sacrificed after a day (Day with traditional LLNA data from published references to provide 5). After sacrifice, ear punch biopsies (6 mm full thickness skin) were collected and an insight into the performance of these methods in the evaluation weighed with a laboratory balance (Mettler Toledo, Columbus, OH) as a marker of ear swelling. Bilateral auricular lymph nodes were separately isolated, weighed and of the chemicals with a sensitization potential. undergone lymphocyte preparation or tissue processing for FACS assay or hema- 232 K.-M. Jung et al. / Toxicology Letters 192 (2010) 229–237

Table 2 Stimulation index obtained with FACS method.

Test article Concentration (%) Stimulation index Deﬁned potency category

IS 50 1.5 ± 0.68 Non-sensitizer

SLS (50% ethanol) 15 3.0 ± 1.08 Weak

SLS (DMF) 5 3.0 ± 0.46 Moderate 10 4.4 ± 0.36 15 6.6 ± 0.98

DNCB 0.01 1.9 ± 0.89 Extreme 0.1 9.7 ± 2.83 0.25 17.9 ± 6.41 0.5 33.2 ± 6.74

PPD 0.1 6.5 ± 1.06 Extreme 1 17.0 ± 5.02 3 29.9 ± 6.96

Isoeugenol 2.5 2.3 ± 1.41 Moderate 5 4.2 ± 0.82 10 8.4 ± 1.21

HCA 5 2.6 ± 0.25 Moderate 10 4.5 ± 0.53 25 6.8 ± 0.80

Eugenol 5 3.2 ± 0.94 Moderate 10 6.8 ± 0.46 25 10.5 ± 1.32

PPD, DNCB, HCA, isoeugenol, eugenol and IS were dissolved in acetone:olive oil (AOO; 4:1) and SLS was dissolved in 50% ethanol or DMF. Values are mean ± S.D. of 5–6 animals. Potency category was deﬁned according to Gerberick et al. (2005). toxylin and eosin staining with immunohistochemistry for BrdU incorporated lymph 1:50. Cells were washed once more and then re-suspended in 20 ␮l 7-AAD solution node cells. to label DNA. Ten thousand 7-AAD expressing cells were gated, and the number of the cells expressing BrdU was analyzed using BD FACSCaliburTM system.

2.3. Lymph node cell counts in auricular lymph nodes and ﬂow cytometry analysis for BrdU positive LNCs 2.4. Immunohistochemical staining for BrdU positive LNCs

Lymph node cells (LNCs) were prepared from lymph node by disintegration For immunohistochemical staining (IHC) of incorporated BrdU, representative through 70 ␮m mesh (BD Biosciences, Franklin Lakes, NJ) in 1 ml PBS. The LNCs tissue samples were fixed in 10% neutral phosphate-buffered formalin, processed in were counted using a hemacytometer after staining with trypan blue. The LNCs a routine manner, embedded in paraffin, and sectioned at 5 ␮m thickness. Sections (2 × 106/ml) were washed once by centrifugation (300 × g) for 5 min with PBS and were incubated with monoclonal mouse anti-BrdU antibody (M0744, DakoCytoma- re-suspended for fixation and permeabilization step, according to the instruction tion, Glostrup, Denmark) as a primary antibody at a dilution of 1:150 overnight at manual of BrdU Flow kits (BD PharmingenTM, Franklin Lakes, NJ). Then LNCs were 4 ◦C. After rinsed three times, sections were incubated with polymer-HRP secondary permeabilized using Cytoperm plus buffer, which contains 10% DMSO. After DNA antibody (Klear Mouse DAB Kit, Golden Bridge International, Mukilteo, WA) for 1 h was denatured by incubating for 1 h with DNase, LNCs were washed, and incubated at RT. Immunoreactivity was detected by the standard avidin–biotin immunoper- with FITC conjugated anti-BrdU antibody for 20 min at RT in the dark at a dilution of oxidase method. Counterstaining with Meyer’s hematoxylin was then performed

Table 3 Stimulation index obtained with IHC method.

Test article Concentration (%) Stimulation index Deﬁned potency category

IS 50 1.3 ± 0.16 Non-sensitizer

SLS (50% ethanol) 15 1.6 ± 0.28 Non-sensitizer

SLS (DMF) 5 3.1 ± 0.90 Moderate 10 3.5 ± 0.48 15 5.0 ± 2.00

DNCB 0.01 1.5 ± 0.23 Extreme 0.1 3.5 ± 0.38 0.25 4.6 ± 0.54 0.5 9.0 ± 1.76

PPD 0.1 3.4 ± 0.49 Extreme 1 5.4 ± 0.44 3 6.2 ± 0.49

Isoeugenol 2.5 1.7 ± 0.63 Moderate 10 5.3 ± 3.29

HCA 5 1.3 ± 0.17 Non-sensitizer 10 1.7 ± 1.22 25 2.6 ± 0.97

Eugenol 5 2.8 ± 0.97 Weak 10 4.0 ± 0.79 25 2.9 ± 1.40

Values are mean ± S.D. of 5–6 animals. Potency category was deﬁned according to Gerberick et al. (2005). K.-M. Jung et al. / Toxicology Letters 192 (2010) 229–237 233

Fig. 3. Flow cytometric analysis of LNCs for the BrdU incorporation. After mice were treated with 0.1%, 1% and 3% PPD, 0.25% DNCB, 15% SLS or 50% IS for 3 days, auricular lymph node was collected on Day 5. (A) LNCs from auricular lymph nodes were stained with FITC conjugated anti-BrdU antibody and analyzed with FACS. Nuclei were counterstained using 7-AAD. (B) Representative FACS histogram for LNCs from AOO and 3% PPD treated mice. (C) Percentage of positive BrdU incorporated LNCs. *Signiﬁcant difference from vehicle group, p < 0.05, values are mean ± S.D. (N = 5 or 6).

for 5 min. Thereafter they were evaluated under light microscope (Olympus, Japan) Correlation between the tests was evaluated by Pearson’s correlation analysis. Sta- and analyzed with Image-Pro® Plus (Media Cybernetics, Bethesda, MD). To demon- tistical analysis was performed using SPSS software (Chicago, IL). strate the difference of the BrdU-labeled lymphocytes of each group quantitatively, images of 9 sites were randomly selected (1300 × 1030 pixels, magniﬁcation ×400) and analyzed. The results were expressed as mean number of BrdU incorporated 3. Results LNCs per group. 3.1. Changes in lymph node weight and lymph node cell count 2.5. Statistics To compare the performance of FACS and IHC LLNA method, Values are shown as mean ± S.D. for all groups. The data were subjected to one- way analysis of variance followed by Duncan’s multiple range test or Student t-test 7 chemicals with known potencies (Table 1) were examined in to determine which means were signiﬁcantly different from the control (p < 0.05). LLNA where bilateral auricular lymph nodes were separately iso- 234 K.-M. Jung et al. / Toxicology Letters 192 (2010) 229–237

Fig. 4. Immunohistochemical analysis of LNCs for the BrdU incorporation. (A) Representative low-power microscopic photographs of IHC for BrdU incorporation in lymph nodes. The number of BrdU corporated lymphocytes (brown) increased in proportion with PPD concentration. Original magnification ×12.5, bar indicates 1000 ␮m. (B) High-power microscopic photograph of lymph node sections of the mice treated with AOO, IS, SLS, DNCB and PPD. Arrow denotes BrdU positive cell of the treated mice. Original magnification ×400, bar indicates 50 ␮m. (C) Number of LNCs immunopositive to BrdU antibody. *Significant difference from vehicle group, p < 0.05, values are mean ± S.D. (N = 5 or 6). (For interpretation of the references to color in this figure legend, the reader is referred to the web version of the article.)

lated and undergone FACS or IHC. Firstly, the changes in auricular 3.2. Changes in mouse ears lymph node weights and the number of LNCs (lymph node cells) were measured to evaluate the effects of chemicals on lymph Biopsies of ear pad were collected after sacrifice and examined nodes. Fig. 1A–C shows the representative data from PPD, DNCB, to evaluate the irritation. Fig. 2A and B shows the representative SLS and IS. The weights of the lymph nodes of the mice treated data from PPD, DNCB, SLS and IS. The weights of the ears treated with sensitizers, PPD (0.1%, 1% and 3%), DNCB (0.25%) and irri- with 15% SLS, 1% and 3% PPD, were increased significantly compared tant, SLS (15% in 50% ethanol), were increased significantly when to respective vehicle control group (Fig. 2A), reflecting that SLS compared to the respective vehicle control groups (Fig. 1A and C). and PPD induced ear edema. These ear edema data well matched Meanwhile, the numbers of LNCs were significantly increased by the inflammatory infiltration and epidermal hyperplasia by PPD sensitizers, PPD and DNCB while IS or SLS did not affect LNC num- (1% and 3%) and SLS (15%) as observed by increased inflammatory bers (Fig. 1B). cells in the histological examination (Fig. 2B), confirming that these K.-M. Jung et al. / Toxicology Letters 192 (2010) 229–237 235

Fig. 5. Comparison of stimulation index (SI) values obtained by flow cytometry, immunohistochemistry and traditional LLNA. (A) Comparison of SI values between traditional LLNA (LLNA) and FACS measurement (FC). (B) Comparison of SI values between LLNA and IHC measurement (IHC). (C) Comparison of SI values between FC and IHC. Correlation between the tests was evaluated by Pearson’s correlation analysis and regression analysis. Correlation coefficient (CC) and p values were obtained by Pearson’s correlation analysis and slope, by regression analysis. Values of traditional LLNA are from the previously reported results (Betts et al., 2006; Gerberick et al., 2005; Warbrick et al., 1999). S is SLS 15%(in 50% ethanol) and D is 0.25% DNCB. chemicals had skin irritation potential at the tested doses. In con- PPD, DNCB, SLS and IS. In the analysis of 10,000 LNCs with FACS, trast, topical applications of vehicle (AOO or 50% ethanol), IS 50%, all the tested doses of PPD and DNCB increased BrdU positive LNC PPD 0.1% or DNCB 0.25% did not alter the ear weight or histological population significantly while IS or SLS treatment induced no sig- appearance. nificant increases in BrdU incorporated LNCs (Fig. 3A–C). The stimulation index (SI) was calculated from the fold increase 3.3. Flow cytometry analysis for BrdU incorporated LNCs of the total number of positive BrdU LNCs in a lymph node compared to that of the respective vehicle control groups. Table 2 shows To evaluate the effects of test article treatment on auricular the SI values obtained with FACS method for the 7 test chemicals. LNC proliferation, DNA synthesis was assessed by measuring the While FACS method produced comparable results to traditional incorporation of BrdU into LNCs with FACS using FITC conjugated LLNA for most of the chemicals, PPD and eugenol were determined anti-BrdU antibody. Fig. 3A–C shows the representative data from to be stronger sensitizers than traditional LLNA. SI values for 15% 236 K.-M. Jung et al. / Toxicology Letters 192 (2010) 229–237

Table 4 Estimated EC3 values and performance of FACS and IHC in identifying sensitizers based on the reported results from T-LLNA.

Category Material EC3a (%)

T-LLNA FACS IHC

DNCB 0.04 0.017 0.11 PPD 0.11 <0.10 <0.10 Sensitizer Isoeugenol 1.5 3.4 5.2 HCA 9.9 5.5 >25 Eugenol 10.1 2.8 5.8

Non-Sensitizer IS – – –

False positive SLS (in DMF) 8.1 5.4 5.9

Sensitivity (%) – 100 (6/6) 83 (5/6)

Speciﬁcity (%) – 100 (1/1) 100 (1/1)

a EC3 values are obtained with least square method with the linear ranges.

SLS, an irritant was estimated to be 3.0 ± 1.08 with 50% ethanol as a 3H-thymidine. To our best knowledge, it is the first study to com- vehicle but it showed 6.6 ± 0.98 with DMF as a vehicle, suggesting pare the performance of non-radioisotopic endpoints of LLNA in that the choice of vehicle is important for LLNA. the same animals. Non-radioisotopic LLNA methods can be more generally used in 3.4. Immunohistochemical staining for BrdU incorporation in many countries since the disposition of the radioactive waste, con- lymph nodes taminated animal corpses and exhaled air are strictly regulated, limiting the widening of traditional LLNA employing radioactive Auricular lymph nodes from treated mice were undergone IHC 3H-thymidine in vivo. Recently, the Interagency Coordinating Com- staining with anti-BrdU antibody and polymer-HRP secondary anti- mittee on the Validation of Alternative Methods (ICCVAM) has body. Fig. 4A–C shows the representative data from PPD, DNCB, presented evaluation reviews on the validation status of the LLNA: SLS and IS. PPD and DNCB treatment increased both lymph node BrdU-ELISA and LLNA: BrdU-FC as a non-radioisotopic modification size and BrdU positive LNCs (0.25% DNCB, 90.3 ± 10.6 and 3% of the traditional LLNA and ECVAM published a draft guideline for PPD, 123.3 ± 9.71, respectively vs 19.8 ± 7.18 for AOO). The num- BrdU-FC (ICCVAM, 2009a,b). In these reviews, it was concluded that ber of BrdU incorporated LNCs of the mice were also significantly these non-radioisotopic endpoints might be acceptable but require increased by the treatment of SLS (34.0 ± 5.20 vs 20.8 ± 8.96 in further validation studies. While the traditional LLNA assesses LNC 50% EtOH), whereas IS-treated mice showed no significant increase proliferation by measuring the incorporation of radioactive 3H- (26.0 ± 3.08, Fig. 4C). Table 3 shows the SI values obtained with IHC thymidine using liquid scintillation count, the LLNA:BrdU-FC and method for the 7 test chemicals. In contrast to FACS, HCA and SLS (in LLNA:BrdU-ELISA assess the incorporation of BrdU, an analog of 50% ethanol) were determined to be non-sensitizer in IHC method. thymidine (Takeyoshi et al., 2004), depending on the affinity of anti-BrdU antibody. In the current study, we demonstrated that 3.5. Comparison of non-radioisotopic endpoints and traditional LLNA:BrdU-FC and LLNA:BrdU-IHC method give well-correlated LLNA data to traditional LLNA, supporting that these methods can replace traditional LLNA in the future. We compared the SI values obtained from FACS, IHC and the In the present study, the strong irritant, SLS in 50% ethanol was reported results from traditional LLNA with 3H-thymidine uptake determined to be a weak sensitizer in FACS while IHC defined it assay (Betts et al., 2006; Gerberick et al., 2005; Warbrick et al., 1999) as non-sensitizer. Although most non-sensitizers were concluded with Pearson’s correlation and regression analysis. As shown in as negative in the traditional LLNA, false positive results could be Fig. 5A and B, both FACS and IHC methods showed well-correlated often observed with a non-sensitizer and a strong irritant like SLS SI data to traditional LLNA as determined by strong statistical sig- (Basketter et al., 1998). Indeed, Jacobs et al. (2006) have reported nificances (p < 0.01) and linear relationship (correlation coefficients that toxic dose of non-sensitizing skin irritants can cause the death close to +1) in Pearson’s correlation analysis. However, the com- of epidermal keratinocytes and dying keratinocytes can serve as an parison between FACS and IHC indicated that FACS method was adjuvant for stimulating dendritic cell migration and maturation in more sensitive than IHC as determined by a high slope in regression a non-antigen-dependent manner. We could confirm indeed that analysis (slope = 4.11 in FACS vs IHC) (Fig. 5C). In addition, compar- when SLS was applied in DMF, another accepted vehicle for LLNA, ison between FACS and traditional LLNA showed that FACS method stronger and persistent irritation was induced and higher SI values could detect sensitizers with equivalent sensitivities to, or higher (6.6 ± 0.98 in FACS and 5.0 ± 2.00 in IHC) were obtained, leading to than traditional LLNA (slope = 0.429 in traditional LLNA vs FACS) a conclusion that SLS is a sensitizer with a moderate potency. This (Fig. 5A). In support of this, comparison of EC3 values and definition result indicates the importance of the vehicle selection in LLNA test of sensitizers obtained from these methods (Table 4) demonstrated and the necessity of the multiple vehicle tests for the chemicals that FACS method shows an equivalent performance to traditional with ambiguous results. Additionally, the LLNA results of strong LLNA. irritants should be reviewed carefully to discern possible compli- cations from strong irritation and keratinocyte cytotoxicity. 4. Discussion While many reports suggested the utility of FACS or IHC as the new non-radioisotopic endpoints for LLNA, there has been no In this study, we demonstrated that the skin sensitization poten- research comparing these two endpoints in the same animals to tial could be successfully identified in LLNA by non-radioisotopic our best knowledge. Both FACS and IHC could successively iden- endpoints, FACS and IHC for BrdU incorporated lymph node cells tify the sensitizers with a good correlation to traditional LLNA. (LNCs). Both FACS and IHC could successively differentiate sensi- Although FACS or IHC is more time-consuming and expensive, tizers with comparable accuracies to traditional LLNA employing they can be conducted non-radioisotopically, indicating that these K.-M. Jung et al. / Toxicology Letters 192 (2010) 229–237 237 methods could be more generally used in the countries having strict Dearman, R.J., Scholes, E.W., Ramdin, L.S., Basketter, D.A., Kimber, I., 1994. The local regulations for the in vivo use of radioisotopes. Incidentally, these lymph node assay: an interlaboratory evaluation of interleukin 6 (IL-6) production by draining lymph node cells. J. Appl. Toxicol. 14, 287–291. countries should pay a large resource in facility, ventilation and Dearman, R.J., Hilton, J., Basketter, D.A., Kimber, I., 1999. Cytokine endpoints for the 3 the disposal of radioactive waste to use H-thymidine, making the local lymph node assay: consideration of interferon-gamma and interleukin 12. traditional LLNA to be more expensive assay method than FACS J. Appl. Toxicol. 19, 149–155. Farrell, J., Jenkinson, C., Lavergne, S.N., Maggs, J.L., Kevin Park, B., Naisbitt, D.J., 2009. and IHC. Moreover, the signal to noise level, that is, the stimulation Investigation of the immunogenicity of p-phenylenediamine and Bandrowski’s index data of FACS method were comparable or superior to tradi- base in the mouse. Toxicol. Lett. 185, 153–159. tional LLNA (Fig. 5A) and IHC method (Fig. 5C), suggesting that FACS Gerberick, G.F., Cruse, L.W., Ryan, C.A., 1999. Local lymph node assay: differentiating method could be a good alternative for traditional LLNA. allergic and irritant responses using flow cytometry. Methods 19, 48–55. Gerberick, G.F., Ryan, C.A., Kern, P.S., Schlatter, H., Dearman, R.J., Kimber, I., Patlewicz, FACS, along with cell counting, can measure various markers G.Y., Basketter, D.A., 2005. Compilation of historical local lymph node data in one process such as cellularity, BrdU incorporation, cell viabil- for evaluation of skin sensitization alternative methods. Dermatitis 16, 157– ity and surface markers with high specificity, reducing further the 202. Gerberick, G.F., Ryan, C.A., Dearman, R.J., Kimber, I., 2007. Local lymph node noise from non-lymphocytes or non-specific binding. Meanwhile, assay (LLNA) for detection of sensitization capacity of chemicals. Methods 41, IHC determines the BrdU incorporated cells depending on anti- 54–60. body binding, and it could not differentiate non-lymphocytic cells, ICCVAM, 2009a. Revised Draft Background Review Document—Non-radioactive Murine Local Lymph Node Assay: BrdU-ELISA Test Method Protocol (LLNA: compromising the specificity and sensitivity to a certain degree. In BrdU-ELISA). ICCVAM. support of this, HCA, a moderate sensitizer was determined to be ICCVAM, 2009b. Draft ICCVAM Test Method Recommendations—Non-Radioactive negative in IHC method (Table 4). In this regard, we suggest that Local Lymph Node Assay: BrdU- FC. ICCVAM. Idehara, K., Yamagishi, G., Yamashita, K., Ito, M., 2008. Characterization and eval- FACS assay appears to be a better method than IHC in the evalua- uation of a modified local lymph node assay using ATP content as a non-radio tion of sensitization potential of chemicals with non-radioisotopic isotopic endpoint. J. Pharmacol. Toxicol. Methods 58, 1–10. LLNA. Jacobs, J.J., Lehe, C.L., Hasegawa, H., Elliott, G.R., Das, P.K., 2006. Skin irritants and contact sensitizers induce Langerhans cell migration and maturation at irritant In conclusion, we have confirmed that the evaluation of prolifer- concentration. Exp. Dermatol. 15, 432–440. ation in the auricular lymph node with FACS and IHC in LLNA could 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, provide a good alternative screening method for contact allergens, D.Y., 2008. Intracellular expression of cytokines and granzyme B in auricular without the use of radioisotope. And we could demonstrate that lymph nodes draining skin exposed to irritants and sensitizers. Toxicology 250, 116–123. FACS could show a higher sensitivity than IHC in the measurement Porstmann, T., Ternynck, T., Avrameas, S., 1985. Quantitation of 5-bromo-2- of BrdU incorporation suggesting that this method could be a good deoxyuridine incorporation into DNA: an enzyme immunoassay for the substitute for traditional LLNA using radioactive 3H-thymidine. assessment of the lymphoid cell proliferative response. J. Immunol. Methods 82, 169–179. Suda, A., Yamashita, M., Tabei, M., Taguchi, K., Vohr, H.W., Tsutsui, N., Suzuki, R., Conflict of interest Kikuchi, K., Sakaguchi, K., Mochizuki, K., Nakamura, K., 2002. Local lymph node assay with non-radioisotope alternative endpoints. J. Toxicol. Sci. 27, 205–218. Takeyoshi, M., Yamasaki, K., Yakabe, Y., Takatsuki, M., Kimber, I., 2001. Development The authors have declared no conflicts. of non-radio isotopic endpoint of murine local lymph node assay based on 5- bromo-2-deoxyuridine (BrdU) incorporation. Toxicol. Lett. 119, 203–208. Acknowledgement Takeyoshi, M., Sawaki, M., Yamasaki, K., Kimber, I., 2003. Assessment of statistic analysis in non-radioisotopic local lymph node assay (non-RI-LLNA) with alpha- hexylcinnamic aldehyde as an example. Toxicology 191, 259–263. This research was supported by a grant (09162KFDA557) from Takeyoshi, M., Noda, S., Yamazaki, S., Kakishima, H., Yamasaki, K., Kimber, I., 2004. Korean Food & Drug Administration in 2009. Assessment of the skin sensitization potency of eugenol and its dimers using a non-radioisotopic modification of the local lymph node assay. J. Appl. Toxicol. 24, 77–81. References van den Berg, F.A., Baken, K.A., Vermeulen, J.P., Gremmer, E.R., van Steeg, H., van Loveren, H., 2005. Use of the local lymph node assay in assessment of immune Basketter, D.A., Miettinen, J., Lahti, A., 1998. Acute irritant reactivity to sodium lauryl function. Toxicology 211, 107–114. sulfate in atopics and non-atopics. Contact Dermat. 38, 253–257. Warbrick, E.V., Dearman, R.J., Lea, L.J., Basketter, D.A., Kimber, I., 1999. Local lymph Betts, C.J., Dearman, R.J., Heylings, J.R., Kimber, I., Basketter, D.A., 2006. Skin sensitiza- node assay responses to paraphenylenediamine: intra- and inter-laboratory tion potency of methyl methacrylate in the local lymph node assay: comparisons evaluations. J. Appl. Toxicol. 19, 255–260. with guinea-pig data and human experience. Contact Dermat. 55, 140–147. Yamano, T., Shimizu, M., 2009. Skin sensitization potency and cross-reactivity of Boussiquet-Leroux, C., Durand-Cavagna, G., Herlin, K., Holder, D., 1995. Evaluation p-phenylenediamine and its derivatives evaluated by non-radioactive murine of lymphocyte proliferation by immunohistochemistry in the local lymph node local lymph node assay and guinea-pig maximization test. Contact Dermat. 60, assay. J. Appl. Toxicol. 15, 465–475. 193–198.