A new cytometry-based method reveals an accumulation of Nrf2 in dendritic cells exposed to two respiratory sensitizers

Abstract

The mechanisms underlying chemical respiratory sensitization are incompletely understood. One of the major cell types involved in this pathology are dendritic cells. In this study, the mechanisms of the NRF2–Keap1 pathway were studied using a bone marrow-derived dendritic cell model exposed to two respiratory sensitizers: ammonium hexachloroplatinate (HCP) and ammonium tetrachloroplatinate (ATCP). Expression levels for two Nrf2-regulated genes, hmox1 and srxn1, were analyzed by real time-quantitative polymerase chain reaction. A flow cytometry-based method was also developed to measure intracellular Nrf2 accumulation in dendritic cells following exposure. Exposure to HCP and ATCP increased both hmox1 and srxn1 gene expression, and was associated with accumulation of Nrf2 protein in cells. Overall, these results show that the respiratory sensitizers, in addition to skin sensitizers, can also induced markers associated with NRF2–Keap1 pathway activation in dendritic cells. This study contributes to a better understanding of the adverse outcome of respiratory sensitization.

Introduction

Respiratory allergy associated with asthma and rhinitis is an important health issue. In occupational and personal settings, allergies can be caused by exposure to chemicals that must be characterized as respiratory sensitizers [1]. Rather than waiting for clinical evidence, in vitro assays are needed to predict such properties. The mechanisms underlying chemical respiratory sensitization must be understood if we want to develop toxicological tools to detect chemical-induced allergy. Several studies have focused on skin sensitization, leading to the first described adverse outcome pathway (AOP). An AOP is a simplified organizational construct, which describes initiating molecular events and early key events (KE) at a low level of biological organization. These events subsequently lead to the adverse outcome (AO). In this context, AOPs serve as a support when developing tests to determine the sensitizing potential of a particular compound. Along the lines of the methods developed to identify skin sensitizers, we now need to develop in vitro assays to assess the respiratory sensitizing potential of chemicals [2].

The AOP for respiratory sensitization has yet to be validated [3]. However, it is generally recognized that the physiopathological mechanisms involved in respiratory sensitization share common KE with skin sensitization. Among common events, dendritic cell (DC) activation (KE 3) has been identified [4].

We have developed a bone marrow-derived dendritic cells (BMDC) assay based on this KE [5]. This in vitro assay was shown to identify some chemical respiratory sensitizers, notably the metal salt ammonium hexachloroplatinate (HCP), which has been linked to various allergic outcomes, including asthma and rhinitis [6, 7]. Indeed, HCP is considered to be a positive control for respiratory sensitization [8]. However, it must be admitted that little is known about the mechanism through which this compound and other platinum salts, such as ammonium tetrachloroplatinate (ATCP), cause sensitization. In particular, we do not know how they activate DC in in vitro assays [9]. One of the pathways that has been studied in the ‘cellular activation’ KEs is the NRF2–Keap1 pathway [9]. Nrf2 is a transcription factor that binds to the antioxidant response element (ARE) in the regulatory regions of several target genes. The repressor protein Keap1 promotes Nrf2 degradation by the proteasome [10]. Keap1 is a cysteine-rich protein, which can be modified by multiple oxidants and electrophiles, any of which may be sensitizers [10]. There are several ways to study activation of this pathway, such as by measuring Nrf2 protein accumulation or its phosphorylation, or monitoring upregulation of Nrf2-targeted genes [11]. Activation of the NRF2–Keap1 pathway is used in the KeratinoSens™ assay as an indication of the skin-sensitizing power of chemicals [12]. In contrast to skin sensitizers, and in particular in dendritic cell assays, we lack information on whether this pathway is also activated by respiratory sensitizers.

The aim of the current study was to analyze the NRF2–Keap1 pathway, by monitoring Nrf2-regulated genes and protein accumulation, in response to the respiratory sensitizers HCP and ATCP in a BMDC model. For this purpose, we adopted an approach combining commonly used quantitative real-time PCR with a more original method, analysis of intracellular Nrf2 protein accumulation by flow cytometry.

Materials and Methods

Chemicals

Chemicals used were: 1-fluoro-2,4-dinitrobenzene (DNFB, CAS: 70–34-8), a skin sensitizer; ammonium hexachloroplatinate (HCP, CAS: 16919–58-7) and ammonium tetrachloroplatinate (ATCP, CAS: 13820-41-2), two respiratory sensitizers; sodium dodecyl sulfate (SDS, CAS: 151-21-3), an irritant non-sensitizer; and the Nrf2 activator tert-butylhydroquinone (tBHQ, CAS: 1948-33-0). All chemicals were purchased from Sigma-Aldrich (purity > 98%). HCP, ATCP, and SDS were prepared in Hank’s Balanced Salt Solution (HBSS). DNFB and tBHQ were prepared in dimethyl sulfoxide (DMSO). The final concentration of vehicle in the culture medium did not exceed 0.125%, at which levels it did not affect cell viability nor expression levels of the markers investigated (data not shown).

Generation of immature BMDC and treatment

A murine model of dendritic cells (BMDC) was generated as described previously [5]. After 7 days of culture, cells were trypsinized, counted, and transferred to cytometry tubes at a concentration of 2.10⁶ cells/mL. To induce an effect in cells, chemicals must be applied at a subtoxic concentration [13]. BMDC were treated with the effective concentration, which induced up to 25% mortality (EC25), as determined by trypan blue exclusion. In all experiments, cells were exposed for 3 h to one single chemical at EC25 concentration. The concentrations applied were as follows: 10 μM for DNFB, 150 μM for HCP, 200 μM for ATCP, 200 μM for SDS, and 100 μM for tBHQ.

RNA isolation and reverse transcription

After treatment, BMDC were lyzed and total RNA was isolated with the RNeasy kit (Qiagen) according to the manufacturer’s instructions. Then, 0.5 μg of RNA was reverse transcribed using the RT script kit (Bio-Rad).

Quantitative real-time PCR was performed using the SYBR GREEN kit from Bio-Rad, according to the manufacturer’s instructions. Analysis focused on expression of the ARE-dependent genes hmox1 and srxn1, and on expression of the housekeeping genes actin and tbp (tata box binding protein). The forward and reverse primers used are listed in the Supplementary data 1. Data were processed using CFX manager software (Bio-Rad), and quantification was based on the ΔΔCq method.

Flow cytometry protocol

Following exposure to chemicals, BMDC were fixed for 12 min at 37°C with Cytofix ^Tm and permeabilized for 30 min at 4°C with Perm III^Tm buffer (BD Biosciences) according to the manufacturer’s protocol. Permeabilized cells were then stained for 1 h with the following monoclonal antibodies: BlueViolet650-CD11c (Dilution 1/100; Clone B-ly6 – BD Biosciences), Phycoerythrin-Nrf2 (Dilution 1/50; Clone D1AE – Cell Signaling Technology), and their corresponding isotype controls.

Samples were analyzed on a FORTESSA x20™ cytometer (BD Biosciences). Debris were excluded from the analysis by applying a gate based on forward- and side-scattering. As CD11c + cells correspond to dendritic cells, we analyzed Nrf2 antibody fluorescence in this population. At least 20 000 cells were analyzed for each sample.

Mean fluorescence intensities (MFI) were recorded and processed with Diva and FlowJo software. Relative fluorescence intensity (RFI) values corresponded to the MFI of each marker for chemical-treated cells minus the MFI of the corresponding isotype control. Results are expressed as fold-change with respect to control cells (vehicle-treated cells).

Statistical analyses

Results were statistically analyzed using a mixed linear regression model to estimate the effect of a product on Nrf2 accumulation, or hmox1 and srxn1 expression levels. For each experiment, results are presented as the mean value ± standard error of the mean (SEM) determined from at least three independent experiments. Significance levels are indicated as follows: ^**P < 0.01 and ^***P < 0.001.

Results and Discussion

Nrf2-targeted genes are modulated in BMDC in response to sensitizers

To investigate activation of the NRF2–Keap1 pathway, expression levels for two Nrf2 target genes, hmox1 and srxn1, were studied. To do so, BMDC were treated with chemicals for 3 h, and the relative mRNA levels for these genes were measured by real time-quantitative polymerase chain reaction (RT-qPCR).

The Nrf2-activator tBHQ-induced upregulation of both hmox1 (7.6-fold) and srxn1 (8.6-fold) (Fig. 1). Transcription of these two ARE-dependent genes was even more markedly induced following exposure of BMDC to the skin sensitizer DNFB (fold-changes of 20.6 and 14.5, respectively). In contrast, the irritant SDS had no effect on expression of these genes. As mentioned above, the KeratinoSens™ assay uses activation of the NRF2–Keap1 pathway in a keratinocyte cell line as a marker of the skin sensitizing power of chemicals. In our BMDC model, the skin sensitizer DNFB also activated this pathway. These results are in accordance with those of a previous study that demonstrated that the DC-like cell line THP-1 could upregulate Nrf2-regulated genes, such as hmox1 or nqo1, in response to exposure to skin sensitizers [14].

Figure 1.

Nrf2-regulated gene expression in BMDC is significantly enhanced by both skin and respiratory sensitizers. BMDC were exposed to HCP, ATCP, DNFB, SDS, and tBHQ for 3 h. The mRNA levels of hmox1 (A) and srxn1 (B) were determined by RT-qPCR. Data are represented as fold-change for expression levels normalized relative to vehicle-exposed cells. Data correspond to the mean ± SEM from three independent experiments. Statistical analysis: linear mixed regression model. ^***P < 0.001; ^**P < 0.01 (comparison between treated and untreated cells). tBHQ: tert-butylhydroquinone, SDS: sodium dodecyl sulfate, DNFB: 1-fluoro-2,4-dinitrobenzene, HCP: ammonium hexachloroplatinate, ATCP: ammonium tetrachloroplatinate.

The respiratory sensitizers tested, HCP and ATCP, also induced upregulation of hmox1 (10.3- and 7.1-fold) and srxn1 (3.3 and 2.8-fold) expression in BMDC. These results suggest that respiratory sensitizers can also upregulate ARE genes in this BMDC model, although the extent of upregulation was weaker than that observed with the skin sensitizer or with the Nrf2 activator. Previous reports on human DC-like ‘THP-1’ cells indicated that the kinetics of the redox imbalance triggered by sensitizers depends on the nature of the sensitizer, with respiratory sensitizers taking longer than skin sensitizers to induce measurable effects [15]. Therefore, in other experiments, we examined gene expression after 6 h exposure to HCP and ATCP. In these conditions, stronger induction was observed (data not shown).

According to the literature, the two classes of sensitizers also showed non-equivalent potencies in the local lymph node assay (LLNA), with extreme sensitization reported for DNFB [11], and strong sensitization for HCP and ATCP [16, 17]. These distinctions could at least partly explain the differences in expression induced here.

The RT-qPCR data showed that in the BMDC model, within a few hours, both skin and respiratory sensitizers induced overexpression of ARE-dependent genes in dendritic cells, whereas an irritant did not. Thus, with the products tested, expression levels of Nrf2-regulated genes could be used to distinguish between sensitizers and irritants.

Intracellular Nrf2 accumulates in murine dendritic cells in response to sensitizers

Since NRF2–Keap1 is a complex pathway with multistep-regulation, it was necessary to confirm the RT-qPCR results by a method exploring intracellular Nrf2 protein levels. We therefore performed a cytometry-based assay to measure the accumulation of Nrf2 protein in chemical- and vehicle-treated cells.

To validate the flow cytometry method, we used the Nrf2 activator tBHQ as a positive control. As expected, tBHQ increased Nrf2 accumulation by 4.6-fold compared to vehicle (Fig. 2).

Figure 2.

Nrf2 protein accumulates to a significant extent in murine BMDC following exposure to sensitizers. Cells were exposed to chemicals for 3 h. Intracellular Nrf2 accumulation was analyzed by flow cytometry using a specific antibody on permeabilized cells. Representative flow cytometry data from three independent experiments are shown and compare vehicle condition (light grey) to treated condition (dark grey) (A). MFI was measured by flow cytometry on a FORTESSA x20, and data were analyzed using FlowJo. Results are presented as fold-change, determined as the ratio of the RFI for chemical-treated cells over the RFI for vehicle-treated cells (B). Statistical analysis: linear mixed regression model. ^***P < 0.001 (comparison between treated and untreated cells). tBHQ: tert-butylhydroquinone, SDS: sodium dodecyl sulfate, DNFB: 1-fluoro-2,4-dinitrobenzene, HCP: ammonium hexachloroplatinate, ATCP: ammonium tetrachloroplatinate.

As in the RT-qPCR assay, exposure to the irritant SDS had no effect on the cellular response. In contrast, exposure to DNFB, HCP, and ATCP significantly enhanced Nrf2 accumulation in BMDC, inducing an approximately 4-fold increase in accumulation compared to vehicle. In parallel, cells were co-exposed to sensitizers plus brusatol, a well-known Nrf2 inhibitor. This co-exposure reduced the signal measured for both sensitizers (data not shown), confirming the specificity of the response.

In complement, other chemicals such as naphtol, a skin sensitizer and methyl salicylate, an irritant non-sensitizer have been tested. They showed similar response to chemicals of their respective category (Supplementary data 2, Figs 1 and 2). Taken together, these results indicate that sensitizers induce upregulation of Nrf2-regulated genes, and lead to Nrf2 protein accumulation inside the cell. From a methodological point of view, the cytometry-based method developed here is an alternative to western blots (WB), which are extensively used in Nrf2-activation studies. The availability of an alternative method is a significant advance, as both we and others had difficulty unequivocally detecting Nrf2 in WB in murine in vitro models. These difficulties have been suggested to be the result of questionable specificity of commercial antibodies in WB applications [18, 19]. This method has many advantages such as it multiplexing capability – the possibility to analyze several proteins in the same sample for a defined population – and the quantifiable response it provides. Flow cytometry is also little time-consuming and easy to process with a basic cytometer.

The results obtained with the two methods tested showed that the respiratory sensitizers, HCP and ATPC, can activate the NRF2–Keap1 pathway in dendritic cells. Thus, NRF2–Keap1 pathway activation is not specific to skin sensitizers. Lalko [20] hypothesized that respiratory sensitizers are lysine-reactive compounds, which have been suggested to be unable to activate the NRF2–Keap1 pathway [21]. Although HCP and ATCP may not be exclusive respiratory sensitizers, the results presented here do not support Lalko’s theory. In accordance with our findings, HCP and ATCP were recently demonstrated to show a cysteine-depleting profile [16].

NRF2–Keap1 pathway activation in cutaneous epithelial cells is a marker of sensitizing capacity in the validated KeratinoSens™ method. In another model, the BEAS-2B epithelial pulmonary cell line was used to assess respiratory sensitizers [22], and HCP also activated NRF2–Keap1 pathway in this epithelial model. However, in this previous study, SDS – a non-sensitizing irritant that triggers a well-known false-positive response in the predictive LLNA assay – also activated the NRF2–Keap1 pathway. Similarly, in assays using the DC-surrogate THP-1 cell line, SDS induced overexpression of Nrf2-regulated genes [14]. The irritant property of SDS could explain these false-positive results. However, this type of cross-reaction between irritants and sensitizers is not universal. Consequently, in the BMDC model presented here, SDS induced no response in either the RT-qPCR or flow cytometry assays. Therefore, activation of the NRF2–Keap1 pathway in this model might serve to distinguish between sensitizing and irritant capacities for both skin and respiratory trigger compounds.

Conclusion

In complement to measuring the expression of Nrf2-regulated genes by RT-qPCR, a cytometric method has been developed to study NRF2–Keap1 activation. This method allows the analysis of intracellular Nrf2 protein levels. After validation with a well-known Nrf2 activator, tBHQ, this method was applied to BMDC exposed to sensitizers. Both skin and respiratory sensitizers induced protein accumulation, and confirmed in accordance with RT-qPCR results the activation of the NRF2–Keap1 pathway in dendritic cells. These results contribute to a better characterization of the mechanisms induced by respiratory sensitizers in the key event ‘activation of dendritic cells’. More respiratory sensitizers will now need to be tested to confirm these observations.

Markers associated with the activation of the NRF2–Keap1 pathway in this model could serve to distinguish between sensitizers and non-sensitizers for both skin and respiratory trigger compounds.

Authors’ Contributions

Isabelle Sponne conceived the study. Adrien Audry performed data collection, data analysis, and produced the figured and scripts, with overall guidance from Isabelle Sponne. All authors wrote the manuscript. Isabelle Sponne and Adrien Audry deposited the data.

Conflicts of Interest

There are no conflicts of interest to declare.