lable at ScienceDirect

Pulmonary Pharmacology & Therapeutics 27 (2014) 156e163

Contents lists avai

Pulmonary Pharmacology & Therapeutics

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

A role for mitogen kinase kinase 3 in pulmonary inflammationvalidated from a proteomic approach

T. Holand a, Y. Riffo-Vasquez a, D. Spina a, B. O’Connor a, F. Woisin a, C. Sand b, M. Marber c,K.B. Bacon d,1, C. Rohlff e,2, C.P. Page a,*

a Sackler Institute of Pulmonary Pharmacology, Institute of Pharmaceutical Science, King’s College London, 5th Floor, Franklin-Wilkins Building,Waterloo Campus, London SE1 9NH, UKb Institute of Pharmaceutical Science, King’s College London, 3rd Floor, Franklin Wilkins Building, Waterloo Campus, London SE1 9NH, UKcCardiovascular Research Division, The Rayne Institute, King’s College London, St Thomas’ Hospital Campus, Lambeth Palace Rd, London SE1 7EH, UKdBayer HealthCare AG, Friedrich-Ebert-Str. 475, Wuppertal 42117, Germanye Formerly Oxford Glycosciences, 208 Bath Road, Slough, Berkshire SL1 3WE, UK

a r t i c l e i n f o

Article history:Received 15 November 2013Received in revised form15 January 2014Accepted 18 January 2014

Keywords:AsthmaMKK3ProteomicsNeutrophilsInflammation

* Corresponding author. Tel.: þ44 2078484784; faxE-mail address: clive.page@kcl.ac.uk (C.P. Page).

1 Present address: Axikin Pharmaceuticals, Inc., 1Diego, CA 92121, USA.

2 Present address: Oxford BioTherapeutics, 94A MOX14 4RY, UK.

1094-5539/$ e see front matter � 2014 Elsevier Ltd.http://dx.doi.org/10.1016/j.pupt.2014.01.006

a b s t r a c t

Proteomics is a powerful tool to ascertain which proteins are differentially expressed in the context ofdisease. We have used this approach on inflammatory cells obtained from patients with asthma toascertain whether novel drugs targets could be illuminated and to investigate the role of any such targetin a range of in vitro and in vivo models of inflammation.

A proteomic study was undertaken using peripheral blood mononuclear cells from mild asthmaticsubjects compared with healthy subjects. The analysis revealed an increased expression of the intra-cellular kinase, mitogen activated protein kinase (MKK3), and the function of this protein was investi-gated further in preclinical models of inflammation using MKK3 knockout mice.

We describe a 3.65 fold increase in the expression of MKK3 in CD8þ T lymphocytes obtained fromsubjects with asthma compared with healthy subjects using a proteomic approach which we haveconfirmed in CD8þ, but not in CD4þ T lymphocytes or human bronchial epithelial cells from asthmaticpatients using a Western blot technique. In wild type mice, bacterial lipopolysaccharide (LPS) caused asignificant increase in MKK3 expression and significantly reduced airway neutrophilia in MKK3�/� mice(median, 25, 75% percentile; wild/LPS; 5.3 (0.7e9.9) � 105 cells/mL vs MKK3�/�/LPS; 0 (0e1.9) � 105

cells/mL, P < 0.05). In contrast, eosinophilia in sensitized wild type mice challenged with allergen (0.5(0.16e0.65) � 105 cells/mL) was significantly increased in MKK3�/� mice (2.2 (0.9e3.5) � 105 cells/mL,P < 0.05).

Our results suggest that asthma is associated with MKK3 over-expression in CD8þ cells. We have alsodemonstrated that MKK3 may be critical for airway neutrophilia, but not eosinophilia, suggesting thatthis may be a target worthy of further consideration in the context of diseases associated with neutrophilactivation such as severe asthma and COPD.

� 2014 Elsevier Ltd. All rights reserved.

1. Introduction

Asthma is an inflammatory disorder of the airways that isassociated with an increase in airway hyperresponsiveness, leadingto recurrent episodes of airflow obstruction. The incidence of

: þ44 2078484788.

0835 Road to the Cure, San

ilton Park, Abingdon, Oxon

All rights reserved.

asthma continues to increase in most countries, currently affectingover 300 million people worldwide [1,2]. Despite the availability ofnumerous therapies, a significant proportion of asthma sufferersare unable to control their disease through the use of the currentlyavailable drugs. These deficiencies in current therapy have led toconsiderable research effort to identify novel drug targets, partic-ularly to help identify novel anti-inflammatory drugs to replaceglucocorticosteroids or to be used in the treatment of patients withsevere asthma who may be resistant to glucocorticosteroids [3].High throughput proteomics has become an important researchtool for the assessment of clinical samples in oncology [4,5],neurology [6], cardiovascular diseases [7], diabetes [8] and in

Abbreviations

AU absorbance unitCD cluster of designationCOPD chronic obstructive pulmonary diseaseDTT dithiothreitolHBEC human bronchial epithelial cellsIFN interferonIg immunoglobulinIL interleukinLPS lipopolysaccharideMAPK mitogen activated protein kinaseMKK3 mitogen kinase kinase 3OVA ovalbuminPBS phosphate buffered salineTc cytotoxic T cellTh T helper cellTLR toll like receptorTNF tumour necrosis factor

Table 1Demographics of the healthy and asthmatic volunteers.

Non-asthmatic Asthmatic

Number 22 18Sex (male:female) 11:11 12:6Age (yr) mean � SEM 25.8 � 1.3 28.7 � 1.7Atopic No YesPulmonary function testsFVC (litre) 4.65 � 0.16 4.53 � 0.22FVC (% predicted) 106.04 � 1.88 97.22 � 3.35FEV1 (litre) 3.92 � 0.11 2.86 � 0.13FEV1 (% predicted) 105.27 � 2.16 72.47 � 2.25PEF (litre/second) 532.25 � 22.36 443.5 � 27.97PEF (% predicted) 104.46 � 2.79 81.30 � 4.8PC20 to methacholine (mg/mL) >16 0.57 � 0.23Reversibility to salbutamol (200 mg), % Not done 14.7 � 2.76

Data expressed as mean � SD.

T. Holand et al. / Pulmonary Pharmacology & Therapeutics 27 (2014) 156e163 157

inflammation and asthma [9]. Such studies have uncovered thepotential role of numerous genes and proteins in these variousconditions, although enormous challenges remain to define thefunctional relevance of many of the proteins identified in suchstudies.

Mitogen activated protein kinase kinase 3 (MKK3) is a memberof the p38 MAPK signalling pathway, which plays a pivotal role ininflammatory processes [10]. MKK3 is expressed in various cellsand has been shown to cause a selective increase in p38MAP kinaseactivity [11]. MKK3 is known to be activated following numerousstress stimuli in a wide range of cell types [12,13] and variousstudies have demonstrated that MKK3 is an important factor in thedevelopment of non-allergic inflammatory processes and Th1 re-sponses, both in vivo and in vitro [14e16], although less is knownabout the role of MKK3 in the regulation of lung inflammationcharacteristic of diseases such as asthma and COPD.

The current study describes the identification of upregulatedMKK3 expression in peripheral blood CD8þ lymphocytes isolatedfrom subjects with asthma compared to healthy control subjectsand the characterization of this protein in the regulation of non-allergic and allergic lung inflammation.

2. Methods

2.1. Proteomic studies

2.1.1. Study subjects and sample collectionPeripheral blood (200 mL) was obtained by antecubital ven-

epuncture into lithium heparin tubes from 22 healthy non-smokingsubjects and 18 non-smoking subjects with well characterizedmildasthma (Table 1). The latter were allowed treatment with a beta2-agonist prn, but refrained from taking any medication 24 h prior tovenepuncture. This study was approved by the Ethics committee ofKing’s College London. This was conducted according to thedeclaration of Helsinki principles and informed written consentwas obtained from all participants. Lymphocytes were isolatedusing the MACS� Microbeads Cell Sorting system according to themanufacturer’s instructions.

2.1.2. Proteomic analysisProteomic analysis was performed on 50 mL of supernatant from

peripheral blood cells obtained from healthy and asthmatic sub-jects. Soluble proteins were separated by 2D gel electrophoresis: in

the first dimension isoelectric focussing was performed using pH3e10 immobilized non-linear gradient strips (GE Healthcare), fol-lowed by SDS-PAGE using 9e16% linear polyacrylamide gradientgels covalently attached to their glass support (using BindSilane, GEHealthcare). Gels were stained using Sypro Red (Molecular Probes,Inc.) and gels were scanned and processed using MELANIE7 II 2DPAGE analysis program. Features were detected using the followingparameters: smooths ¼ 2, Laplacian threshold ¼ 2, partialsthreshold ¼ 2, saturation ¼ 25, peakedness ¼ 100, minimumparameter ¼ 7. Significantly upregulated spots were excised andtryptic in-gel digests were analysed by means of peptide massfingerprinting and tandem mass spectrometry. Peptide massfingerprinting was performed using MALDI-TOF-MS (PerSeptiveBiosystems Voyager-DETM STR Matrix-Assisted Laser DesorptionIonization Time-of-Flight mass spectrometer) and resulting peptidemasses were searched against a protein database using theMOWSEsoftware [17]. Tandem MS was performed on a Quadrupole Time-of-Flight (Q-TOF) mass spectrometer (Micromass) and data anal-ysis using the SEQUEST software [18].

2.2. Expression profiling of MKK3

2.2.1. Study subjects and sample collectionHuman bronchial epithelial cells (HBECs) were cultured from

epithelial tissue explants dissected from bronchial biopsies ob-tained from healthy and asthmatic volunteers (Additional file 1:Table S1 for demographics). Peripheral venous blood was takenfrom donors to isolate different inflammatory cells. Healthy non-smoking subjects with no history of asthma, allergic rhinitis orallergic dermatitis, as well as volunteers with a history of asthmawere recruited and the study was approved by the Ethics Com-mittee of King’s College Hospital or King’s College London. Allsubjects with asthma exhibited a positive skin response to at leastone allergen, whilst none of the healthy subjects responded to anyof the allergens tested.

2.2.2. Western blottingCells were lysed in lysis buffer, (50 mM Tris, 5 M urea, 55 mM

DTT in PBS solution containing protease inhibitors) (approximately100 mL buffer for 105 cells) and sonicated for maximum lysis (SeeAdditional file 1; for further details). Tissues were lysed in lysisbuffer (approximately 500 mL buffer for 1 tissue) and homogenizedby three 30 s bursts using an Ultra Turrax T25 (IKA-Werke GmbH &Co). The protein solutionwas centrifuged at 13,000� g for 10min at4 �C and the supernatant was removed for analysis. Protein con-centration was determined using the 2-D Quant Kit Protein Assayaccording to the manufacturer’s instructions (Amersham). For

T. Holand et al. / Pulmonary Pharmacology & Therapeutics 27 (2014) 156e163158

quantification purposes, band intensities were evaluated bymeasuring relative densitometric absorption, using a BioimagingSystem, UVP Ltd., U.K. The area density of each spot was quantifiedusing Labworks software (Image Acquisition and Analysis Soft-ware). For densitometric analysis of Western blot of CD8þ T cells,the image was analysed microscopically (Zeiss Axioscope) andimages captured using a monochromatic high resolution and highsensitivity camera (Orca Flash Hamamatsu, Japan) with appropriatesoftware (HCImage).

2.2.3. ImmunofluorescenceHBECs were seeded into a 24 well plate and cultured to

confluence. Cells were washed with PBS and fixed with 4% para-formaldehyde in 0.1 M phosphate buffer for 30 min at 4 �C.Following fixation cells were washed twice with PBS and stored at4 �C for no longer than 48 h. Fluorescence was detected by a fluo-rescence microplate reader (FL600, Fisher Scientific) with excita-tion at 485 nm and emission at 530 nm. For fluorescence staining,previously rinsed coverslips were mounted on glass slides usingVectashield with DAPI (Vector Laboratories), if total cell numberwas required. The slides were kept in the dark at 4 �C until required.A Nikon Digital Camera (DXM1200) and Lucia image analysissoftware (Nikon, version 4.61) were used to obtain images of thestained cells.

2.3. In vivo validation studies

2.3.1. MiceMKK3�/� (C57BL/6 background; M/F) were a kind gift from Prof.

M. Marber, King’s College London. Wild type C57BL/6 (M/F) (20e25 g) mice were obtained from Charles River laboratories. Experi-ments were approved by the UK Home Office under ‘The Animals(Scientific Procedures) Act (1986)’ and local approval from theEthics Committee of King’s College London. The targeting strategyused to generate the MKK3�/� mice has been described previously[19]. A colony was re-derived from this source by crossing anMKK3�/� male to an outbred C57Bl/6 female. Animals were main-tained on standard chow and water ad libitum and 12 h day/nightlight cycle.

2.3.2. Induction of airway inflammationC57B1/6 mice (Charles River, UK) between 6 and 8 weeks old

were anaesthetized with isoflurane and then received intranasaladministration of 25 mg of Lipopolysaccharide (LPS) (1mg/mL) fromEscherichia coli (Sigma Chemical Co., United Kingdom) or 100 mg ofzymosan (1 mg/mL) (Sigma Chemical Co., United Kingdom) dis-solved in saline. Control mice received saline only. All measure-ments were performed after 24 h of exposure. In otherexperiments, mice were immunized intra-peritoneally with 10 mgof ovalbumin (OVA) (type V; Sigma Chemical Co., Gillingham, UK)absorbed to a saturated solution of aluminium hydroxide (Aldrox,100 mg per animal, Sanofi, Sao Paulo, Brazil) and immunization wasrepeated 7 days later. On days 14e16, mice were placed in a 12 Lplexiglass container and exposed once a day to a nebulized solutionof OVA (30 mg/mL, 25 min; De Vilibiss Ultraneb 90). Sham immu-nized mice were instilled with Aldrox on days 0 and 7 and alsochallenged with OVA on days 14e16. Bronchoalveolar lavage (BAL)was performed 24 h after the last aerosol challenge.

2.3.3. Bronchoalveolar lavageTwenty-four hours after the LPS or zymosan instillation (into

normal mice) or following the last antigen challenge on day 16 (ofallergic mice), animals were terminally anaesthetized with ure-thane (2 g/kg i.p.; Sigma Chemical Co.), and a cannula was insertedinto the exposed trachea. Three 0.5 mL aliquots of sterile saline

were injected into the lungs. From the BAL fluid, an aliquot of50 mL was added to 50 mL of haemolysis solution (Turk solution).The total number of cells in the lavage fluid was counted with animproved Neubauer haemocytometer. For differential cell counts,cytospin preparations were prepared from aliquots of BAL fluid(100 mL), centrifuged at 250� g for 1 min using a Shandon Cyto-spin 2 (Shandon Southern Instruments, Sewickley, PA, USA) atroom temperature and stained with Diffquick and a total of200 cells were counted to determine the proportion of neutro-phils, eosinophils and monocytes using standard morphologicalcriteria.

2.3.4. Cytokine and IgE analysisMurine cytokines in BAL fluid were assayed by ELISA using

paired Abs, according to the manufacturer’s instructions. Totalserum IgE was measured using purified IgE as standard (BD Phar-mingen). IL-5, IL-6, TNF-a, and IFNg ELISA were obtained fromeBioscience (San Diego, CA). IL-12 and the ELISA for measurementof IgE were purchased from BD Pharmingen (Devon, UK). Lowerlevels of detectionwere as follows: IL-5 and IL-6 and IgE at 8 pg/mL;IL-12 at 30 pg/mL, TNF-a at 30 pg/mL; and IFNg at 15 pg/mL.

2.3.5. Airway histology studiesAt the end of each experiment lungs were infused via the tra-

chea with 1 mL of 4% neutral formalin and immersed in the samefixative for at least 48 h. Tissues were embedded in paraffin and8 mm sections were cut and stained with H&E (SigmaeAldrich) forexamining cell infiltration microscopically (Zeiss Axioscope). ANikon Digital Camera (DXM1200) and Lucia image analysis soft-ware (Nikon, version 4.61) were used to obtain images of lungs.

2.4. Statistical analysis

The Wilcoxon Rank-Sum test was performed between the con-trol (healthy) and the asthmatic samples for each unique protein.The proteins which recorded a P-value less than or equal to 0.05were selected as statistically significant with 95% selectivity. Non-parametric tests (ManneWhitney test for two group comparison,or a KruskaleWallis test for comparisons of more than two groups,followed by a ManneWhitney with a Bonferroni correction formultiple comparisons) were used to identify differences inneutrophil or eosinophil cell recruitment and cytokine releaseamongMKK3�/� and control mice groups. Other statistical analysesincluded paired Student’s t-test and considered P < 0.05 significant(Prism, GraphPad Software).

3. Results

3.1. Comparison of healthy and asthmatic lymphoid cell proteomes

Peripheral blood samples were obtained from 22 healthy and 18asthmatics subjects. The demographics of the volunteers aresummarized in Table 1. Lysates from purified lymphocytes wereseparated by SDS-PAGE and the protein inventories of each samplewere identified using mass spectrometry. 300 proteins were iden-tified in the study and differences among the two groups wereevaluated by the presence or absence of protein detection and bydifferences in their quantity. Among the proteins identified was thesignalling molecule MKK3. Analysis of the 2D gels revealed thatMKK3 was differentially expressed in lymphocytes obtained fromhealthy controls and subjects with asthma. A 3.65 fold increase inthe expression of MKK3 was observed in CD8þ T lymphocytes fromsubjects with asthma relative to samples obtained from healthyvolunteers (Fig. 1A).

T. Holand et al. / Pulmonary Pharmacology & Therapeutics 27 (2014) 156e163 159

3.2. MKK3 protein expression during inflammation

Western blot analysis revealed MKK3 expression in humanCD4þ lymphocytes, CD8þ lymphocytes, and human epithelial cells(HBECs) (Fig. 1). Using densitometric analysis, no differences inMKK3 expression were observed in CD4þ lymphocytes (Fig. 1B)obtained from healthy or asthmatic subjects, or in HBECs (Fig. 1C).Similarly, no difference in MKK3 expression in HBECs taken fromhealthy controls and subjects with asthma was detected usingimmunofluorescence staining (Fig. 1D,E). We could not detect anyMKK3 in human neutrophils or eosinophils from healthy andasthmatic subjects (data not shown). In contrast, we detectedsignificantly greater MKK3 expression in human CD8þ T lympho-cytes from subjects with asthma compared with healthy subjects(0.38 � 0.11 vs 0.04 � 0.02 arbitrary units of intensity, P ¼ 0.039,n ¼ 3 per group) (Fig. 1F), confirming the results of the proteomicstudy.

3.3. The role of MKK3 in airway inflammation in vivo

3.3.1. Non-allergic inflammationDensitometric analysis of Western blot data showed that LPS

treated mice developed a marked increase in total lung (control;49.6� 1.9 AU vs LPS; 70.1�0.7 AU, P< 0.0001, n¼ 4 per group) and

Fig. 1. MKK3 expression in haemopoietic and airway cells from healthy and asthmatic subjeperipheral blood taken from 18 asthmatic subjects compared to 22 normal control subjectmeasured relative to the control samples. (B) CD4þ lymphocytes and (C) HBECs between hexperiment, P > 0.05). In other experiments, constitutive MKK3 expression in HBECs fromquantified (n ¼ 4 subjects per group, data from one experiment). Data are expressed as me(Lanes 1e3), and atopic asthmatic (Lanes 4e6). Lane 7 represents purified recombinant MK

spleen (control; 70.6 � 1.3 AU vs LPS; 77.4 � 0.76 AU, P ¼ 0.0036,n ¼ 4 per group) MKK3 protein expression compared to saline-treated mice (Additional file 1: Fig. S1 for an example WesternBlot). There was a significant increase in total cell and neutrophilnumbers in the BAL fluid 24 h after LPS treatment (Fig. 2A P¼ 0.005,Fig. 2B P ¼ 0.003, respectively) or zymosan treatment in wild typemice (Fig. 3A P¼ 0.0003, Fig. 3B P¼ 0.0003, respectively) comparedwith saline controls. In contrast, this response was significantlyattenuated in MKK3�/� mice.

We further measured different cytokine levels in BAL fluid ob-tained from LPS and zymosan treated wild type and MKK3�/� miceusing conventional ELISA (Figs. 2 and 3). LPS induced a significantincrease in the levels of IL-6, IL-12, and IFNg released into BAL fluidin wild type mice compared to saline treated mice (Fig. 2CeF).There was some evidence of the presence of IL-6, IL-12 and TNFa inthe BAL fromMKK3�/�mice challenged with LPS, although this wasnot statistically different from cytokine levels found in salinetreated animals. In the case of IFNg, there was a small, but statis-tically significant, increase (P ¼ 0.024) in levels found in BAL fluidobtained from MKK3�/� mice compared with the saline treatedgroup. In the case of zymosan treatment, there was a significantincrease in IL-6 in BAL fluid of MKK3�/� mice (P ¼ 0.0117; Fig. 3C),but this was not significantly different fromwild type mice treatedwith zymosan. In contrast, therewas a significant reduction in IL-12

cts. (A) MKK3 expression is significantly increased in CD8þ lymphocytes isolated froms (*P < 0.05). In asthma samples, a 3.65-fold increase in the expression of MKK3 wasealthy and asthmatic subjects, respectively (n ¼ 4 subjects per group, data from onehealthy and asthmatic subjects was measured by (D) immunofluorescence and (E)

an � S.E.M. P > 0.05. (F) MKK3 expression in CD8þ T lymphocytes in healthy subjectsK3 as a positive control.

Fig. 2. The effect of genetic ablation of MKK3 on various indices of airway inflammation in response to LPS. (A) Total cell number and (B) neutrophil number was significantlyincreased in BAL fluid following LPS instillation in wild type mice but not MKK3�/� mice (n ¼ 9e13 mice per group, combined data from two individual experiments) and expressedas individual values. Horizontal line represents median. Saline vs LPS. *P ¼ 0.005 (total), P ¼ 0.003 (neutrophils) and LPS wild type vs LPS MKK3�/� mice; yP ¼ 0.027 (total), P ¼ 0.034(neutrophils). The concentration of various cytokines in BAL fluid including (C) IL-6, (D) IL-12, (E) TNF-a and (F) IFNg was significantly increased following instillation of LPS(**P < 0.05 vs wild type saline treated animals; yP < 0.05 vs MKK3�/�). There was no significant increase in these cytokine levels in MKK3�/� mice compared with saline treatedanimals (n ¼ 9e10 mice per group, combined data from two individual experiments). Data are expressed as individual data points. Horizontal line represents median. (G) His-tological examination of lung tissue neutrophilia in wild type (saline control i, LPS treated ii) and MKK3�/� (saline control iii, LPS treated iv). Lung tissues were fixed, sectioned at8 mm thickness, and stained with H&E to identify tissue neutrophils. Bars represent 20 mm.

Fig. 3. The effect of genetic ablation of MKK3 on various indices of airway inflammation in response to zymosan. (A) Total cell number and (B) neutrophil number in BAL fluid insaline and zymosan treated wild type or MKK3�/� mice (n ¼ 8e10, combined data from two individual experiments). Data are expressed as individual data points. The horizontalline represents the median data. Saline vs zymosan in wild type. *P ¼ 0.0003 (total), P ¼ 0.0003 (neutrophils) and MKK3�/� yP ¼ 0.33 (total), P ¼ 0.0015 (neutrophils). This wasconfirmed by a significant difference in cell number between genotype (wild type vs MKK3�/� mice treated with zymosan P ¼ 0.0006 (total), P ¼ 0.0012 (neutrophils)). Theconcentration of various cytokines in BAL fluid including (C) IL-6 and (D) IL-12 was significantly increased following instillation of zymosan in MKK3�/� and wild type micecompared with saline groups, respectively **P ¼ 0.0117 (IL-6), P ¼ 0.02 (IL-12). Data are expressed as individual data points. The horizontal line represents the median data (n ¼ 5e12, combined data from two individual experiments). *P < 0.05 vs saline. (E) Histological examination of lung tissue obtained from wild type (saline control i, zymosan treated ii)and MKK3�/� (saline control iii, zymosan treated iv mice). Lung tissues were fixed, sectioned at 8 mm thickness, and stained with H&E to identify tissue neutrophils. Bars represent20 mm.

T. Holand et al. / Pulmonary Pharmacology & Therapeutics 27 (2014) 156e163160

T. Holand et al. / Pulmonary Pharmacology & Therapeutics 27 (2014) 156e163 161

levels in lavage fluid in MKK3�/� mice compared with wild typemice treated with zymosan (P ¼ 0.02, Fig. 3D). Histological analysisof lung tissue showed an intense accumulation of neutrophils in thelungs obtained fromwild type animals treated with LPS (Fig. 2G) orzymosan (Fig. 3E) compared to saline control mice. In contrast, nomajor inflammatory cell accumulationwas observed in the lungs ofLPS or zymosan treated MKK3�/� mice (Figs. 2G and 3E,respectively).

3.3.2. Allergic inflammationSensitization and challenge of the airways of allergic mice with

ovalbumin did not result in an increase in MKK3 protein expres-sion in the lung or spleen in comparison with sham-immunizedmice (data not shown), suggesting that the expression of MKK3protein is differentially activated in non-allergic and allergic in-flammatory responses. The total number of cells (Fig. 4A) andeosinophils (Fig. 4B) in BAL fluid was significantly increased insensitized mice challenged with aerosols of OVA in wild type mice.Similarly, MKK3�/� mice still retained the capacity to recruit in-flammatory cells to the airways as reflected in the total number ofcells and eosinophils (Fig. 4A,B, respectively). Indeed, we observeda 4.4-fold increase in median eosinophil number in OVA sensitizedMKK3�/� mice compared with OVA sensitized wild type mice(P ¼ 0.0096).

The concentration of IL-5 in BAL fluid (Fig. 4C) and total IgElevels in blood serum (Fig. 4D) was measured in allergic mice usingconventional ELISA. Allergic MKK3�/� mice challenged with OVAdemonstrated a significant increase in the concentration of IL-5 inmice compared to sham-immunized mice (Fig. 4D, P ¼ 0.034).MKK3�/� mice (sham) also revealed constitutively increased serumlevels of IgE compared with shamwild typemice, although this wasonly statistically significant with OVA treatment (Fig. 4D). Histo-logical analysis of the lung tissue demonstrated an increasedeosinophil accumulation in OVA sensitized wild type micecompared to sham-immunized mice (Fig. 4E). However, anincreased inflammatory response was observed in the lungs ofMKK3�/� mice irrespective of allergic status (Fig. 4E).

Fig. 4. The effect of genetic ablation of MKK3 on various indices of airway inflammation in alnumber in BAL fluid are increased following OVA challenge of sensitized wild type or MKKmedian data (n ¼ 10e13 mice per group, combined data from two individual experiments).sham, yP ¼ 0.0096 vs wild type ova group. The concentration of (C) IL-5 in BAL fluid was signand (D) serum IgE was significantly increased following sensitization to OVA in MKK3�/� micdata points. The horizontal line represents median data (n ¼ 10e13 per group, combined datafrom wild type (sham control i, ovalbumin treated ii) or MKK3�/� (sham control iii, ovalbumwith H&E to identify tissue eosinophils. Bars represent 20 mm.

4. Discussion

We have identified the upregulation of MKK3 in CD8þ lym-phocytes isolated from patients with well characterized atopicasthma compared to healthy subjects. However, our in vivo study inMKK3 knockout mice has revealed the central importance thisprotein plays in the development of neutrophilic, but not eosino-philic lung inflammation.

As anticipated, LPS-challenged mice showed a significant neu-trophilia in the lung that was associated with an increase in MKK3expression in lung and spleen tissue. Furthermore, we have shownthat neutrophil recruitment to the lung following LPS stimulationwas significantly attenuated in MKK3�/� mice in comparison towild type mice, which likely results from the reduction in cytokinessuch as IL-6, IL-12, IFNg and TNF-a observed in these mice, cyto-kines that have been implicated in the recruitment of neutrophils[15,19,20]. These data extend previous studies showing a stronglink between p38 MAPK signalling and LPS in a number of cells[19,21e23], as well as in LPS-induced pulmonary inflammation inthe rat [24] and mouse [25]. We have further demonstrateddecreased neutrophilia to the lung, as well as suppression of levelsof IL-12, in MKK3�/� mice following zymosan stimulation whichextends previous findings that zymosan signals via a TLR2/6 [26]and MKK3 dependent process [27]. The role of downstream tar-gets of MKK3 was beyond the scope of this study, although it isapparent that both MKK3 independent (e.g. NFkappaB) anddependent (p38 MAPK, NFkappaB) pathways are stimulated by LPSvia TLR4 resulting in the secretion of cytokines (e.g. TNFa) andrecruitment of neutrophils to sites of inflammation [28e30].

In contrast, when investigating the role of MKK3 in allergicinflammation, our findings demonstrated no significant changes inMKK3 protein levels following antigen challenge of ovalbuminsensitized mice. Furthermore, mice deficient in MKK3 still retainedthe capacity to recruit eosinophils and produce IL-5 (with a trendfor increased levels of these parameters), suggesting that MKK3 isnot central to Th2 regulated inflammation in these mice. This wasfurther verified by measuring circulating total IgE levels in these

lergic mice in response to ovalbumin challenge. (A) Total cell number and (B) eosinophil3�/� mice. Data are expressed as individual data points. The horizontal line representsTotal cells; *P ¼ 0.005; **P ¼ 0.0003 vs sham. Eosinophils *P ¼ 0.0117, **P ¼ 0.0015 vsificantly increased in OVA sensitized MKK3�/� mice compared with sham (*P ¼ 0.034)e compared with wild type sham mice (**P ¼ 0.0015). Data are expressed as individualfrom two individual experiments). (E) Histological examination of lung tissue obtainedin treated iv mice). Lung tissues were fixed, sectioned at 8 mm thickness, and stained

T. Holand et al. / Pulmonary Pharmacology & Therapeutics 27 (2014) 156e163162

animals. Indeed, significantly higher levels of total IgE were foundin MKK3�/� mice compared with their wild type counterpartsirrespective of immunization. IL-4 levels were not measured inthese experiments, as previous data has shown IL-4 to peak atearlier time points in this murinemodel thanwe investigated in thepresent study [31].

In accordance with previous studies our MKK3�/� mice possessa deficient Th1 cytokine milieu [15,19], which might result in askewing towards a Th2 phenotype. Indeed, Th1 and Th2 responseshave been reported to reciprocally regulate each other by changingthe local cytokine environment. IFNg has been shown to inhibit Th2responses, whereas IL-4 and IL-10 were reported to counteract Th1mediated actions [32,33]. Previous data have suggested CD8þ Tlymphocytes to be the primary cells responsible for producingprotective immunity (Tc1), as well as a local IFNg dominant cyto-kine milieu, which is capable of suppressing the development ofallergic airway inflammation [34]. However other Tc subsets haveemerged (e.g. Tc2, Tc9, Tc17 and Tcreg), although their role inallergic inflammation remains to be established [35,36]. A recentstudy has shown that IFNg secreting CD8þ T lymphocytes suppressan established Th2 allergic inflammatory response and in theabsence of IFNg both neutrophil and eosinophil recruitment to theairways was enhanced [37]. The implication of these findings is thatIFNg from CD8þ T lymphocytes counterregulate Th2 and Th17 re-sponses, and in the absence of this cytokine, lead to an increase inneutrophil and eosinophil recruitment to the lung.

Using a proteomic approach, we have demonstrated anincreased expression ofMKK3 in CD8þ T lymphocytes frompatientswith mild asthma that was also confirmed in CD8þ T lymphocytesisolated from mild asthmatic patients using a Western blot tech-nique. However, we were unable to document MKK3 expression inneutrophils or eosinophils from healthy or asthmatic subjects,which contrasts with previous studies that have reported expres-sion of phosphorylated MKK3 in these cells from healthy subjects[21,38]. Nonetheless, this increase in MKK3 expression in CD8þ Tcells is of interest in view of clinical studies demonstratingincreased spontaneous production of IFNg in the sputum from thiscell type that has been shown to be related to disease severity [39].Furthermore, CD8þ T lymphocyte numbers in bronchial tissue iscorrelated with the annual decline in lung function in patients withbronchial asthma [40] and there is evidence of increased tran-scriptional activity, predominantly in circulating CD8þ comparedwith CD4þ T lymphocytes, in patients with severe asthma [41]. Incontrast, our data suggests that the upregulation ofMKK3 in CD8þ Tlymphocytes from patients with mild asthma is likely not con-trolling the eosinophilia seen in such patients, although it remainsto be established whether this is also a consequence of increasedTc1 and Tc2 CD8þ T lymphocyte activity. Why MKK3 is consistentlyupregulated in CD8þ T lymphocytes from patients with mildasthma is not known, but experimental data showing thatMKK3�/�

mice lose the ability to recruit neutrophils to the lung in response tonon-allergic stimuli may suggest that MKK3 in CD8þ T lymphocytesis important in regulating non-allergic responses in such patients[42].

Collectively, our findings demonstrate that MKK3 abrogationin vivo effectively reduced non-allergic neutrophilic inflammationin the lung due to reduction of Th1-like cytokines. This may be ofrelevance to our understanding of asthma phenotypes such as se-vere asthma and asthma exacerbations that are known to be linkedto neutrophil recruitment to the lung [43e46]. Therefore, it is alsoof interest that the transcriptional activity of CD8þ T lymphocytes isincreased in patients with severe asthma [41]. Given the recentfailure of several novel therapeutic approaches specifically directedtowards inhibiting eosinophil function in subjects with asthma[47,48], our current results provide evidence that perhaps targeting

MKK3 regulated signalling pathways in CD8þ T lymphocytes mayprovide a more fruitful approach to targeting airway diseasesassociated with neutrophilia such as severe asthma with aneutrophilic phenotype [3] and our findings may well have impli-cations for other airway diseases where neutrophils have beenimplicated such as COPD and in patients with so called “overlapsyndrome” (i.e. patients exhibiting characteristics of asthma andCOPD) [49,50].

In conclusion our findings provide evidence for a direct role ofMKK3 in the regulation of non-allergic recruitment of neutrophilsinto the airways.

Disclosure statement

The authors have no competing interests to declare.

Authors’ contribution

CP, KBB and CR conceived and designed the study. TH, YRV, DSdesigned and performed the human in vitro and animal in vivostudies. CS performed Western blot experiments presented inFig.1f. MMprovided theMKK3 breeding colony. BC, FW responsiblefor recruiting patients for proteomic study. CR and KBB performedthe proteomic analysis. TH, YRV, DS and CP drafted the manuscript.All authors contributed to the final draft of the manuscript.

Acknowledgements

We thank Professor Richard Flavell, Section of Immunobiology,Howard Hughes Medical Institute, Yale University School of Medi-cine, New Haven, CT, USA and Roger Davies Howard Hughes Med-ical Institute, University of Massachusetts, Worcester, MA, USA, forthe kind gift of the MKK3�/�mouse line to Professor Mike Marber,King’s College London. We also thank Bayer for funding part of thisresearch.

Appendix A. Supplementary data

Supplementary data related to this article can be found at http://dx.doi.org/10.1016/j.pupt.2014.01.006.

References

[1] Masoli M, Fabian D, Holt S, Beasley R. The global burden of asthma: exec-utive summary of the GINA Dissemination Committee report. Allergy2004;59:469e78.

[2] Rees J. Asthma control in adults. BMJ 2006;332:767e71.[3] Wenzel SE. Asthma: defining of the persistent adult phenotypes. Lancet

2006;368:804e13.[4] Cho WC, Cheng CH. Oncoproteomics: current trends and future perspectives.

Expert Rev Proteomics 2007;4:401e10.[5] Lin Y, Dynan WS, Lee JR, Zhu ZH, Schade RR. The current state of proteomics in

GI oncology. Dig Dis Sci 2009;54:431e57.[6] Maurer MH. Proteomics of brain extracellular fluid (ECF) and cerebrospinal

fluid (CSF). Mass Spectrom Rev 2010;29:17e28.[7] Macri J, Rapundalo ST. Application of proteomics to the study of cardiovas-

cular biology. Trends Cardiovasc 2001;11:66e75.[8] Suss C, Solimena M. Proteomic profiling of beta-cells using a classical

approach e two-dimensional gel electrophoresis. Exp Clin Endocrinol Dia-betes 2008;116(Suppl. 1):S13e20.

[9] Benson LM, Mason CJ, Friedman O, Kita H, Bergen III HR, Plager DA. Extensivefractionation and identification of proteins within nasal lavage fluids fromallergic rhinitis and asthmatic chronic rhinosinusitis patients. J Sep Sci2009;32:44e56.

[10] Pettus LH, Wurz RP. Small molecule p38 MAP kinase inhibitors for thetreatment of inflammatory diseases: novel structures and developmentsduring 2006-2008. Curr Top Med Chem 2008;8:1452e67.

[11] Raingeaud J, Whitmarsh AJ, Barrett T, Derijard B, Davis RJ. MKK3- and MKK6-regulated gene expression is mediated by the p38 mitogen-activated proteinkinase signal transduction pathway. Mol Cell Biol 1996;16:1247e55.

T. Holand et al. / Pulmonary Pharmacology & Therapeutics 27 (2014) 156e163 163

[12] Cuenda A, Alonso G, Morrice N, Jones M, Meier R, Cohen P, et al. Puri-fication and cDNA cloning of SAPKK3, the major activator of RK/p38 instress- and cytokine-stimulated monocytes and epithelial cells. EMBO J1996;15:4156e64.

[13] Cuenda A, Cohen P, Buee-Scherrer V, Goedert M. Activation of stress-activatedprotein kinase-3 (SAPK3) by cytokines and cellular stresses is mediated viaSAPKK3 (MKK6); comparison of the specificities of SAPK3 and SAPK2 (RK/p38). EMBO J 1997;16:295e305.

[14] Inoue T, Hammaker D, Boyle DL, Firestein GS. Regulation of p38 MAPK byMAPK kinases 3 and 6 in fibroblast-like synoviocytes. J Immunol 2005;174:4301e6.

[15] Inoue T, Boyle DL, Corr M, Hammaker D, Davis RJ, Flavell RA, et al. Mitogen-activated protein kinase kinase 3 is a pivotal pathway regulating p38 acti-vation in inflammatory arthritis. Proc Natl Acad Sci U S A 2006;103:5484e9.

[16] Wang L, Ma R, Flavell RA, Choi ME. Requirement of mitogen-activated proteinkinase kinase 3 (MKK3) for activation of p38alpha and p38deltaMAPK isoformsby TGF-beta 1 in murine mesangial cells. J Biol Chem 2002;277:47257e62.

[17] Pappin DJ, Hojrup P, Bleasby AJ. Rapid identification of proteins by peptide-mass fingerprinting. Curr Biol 1993;3:327e32.

[18] Eng JK, McCormack AL, Yates JR. Interaction of angiotensin peptides and zincmetal ions probed by electrospray ionization mass spectrometry. J Am SocMass Spectrom 1994;5:976e89.

[19] Lu HT, Yang DD, Wysk M, Gatti E, Mellman I, Davis RJ, et al. Defective IL-12production in mitogen-activated protein (MAP) kinase kinase 3 (Mkk3)-deficient mice. EMBO J 1999;18:1845e57.

[20] Son YH, Jeong YT, Lee KA, Choi KH, Kim SM, Rhim BY, et al. Roles of MAPK andNF-kappaB in interleukin-6 induction by lipopolysaccharide in vascularsmooth muscle cells. J Cardiovasc Pharmacol 2008;51:71e7.

[21] Nick JA, Avdi NJ, Young SK, Lehman LA, McDonald PP, Frasch SC, et al. Selectiveactivation and functional significance of p38alpha mitogen-activated proteinkinase in lipopolysaccharide-stimulated neutrophils. J Clin Invest 1999;103:851e8.

[22] Han J, Lee JD, Bibbs L, Ulevitch RJ. A MAP kinase targeted by endotoxin andhyperosmolarity in mammalian cells. Science 1994;265:808e11.

[23] Lee JC, Laydon JT, McDonnell PC, Gallagher TF, Kumar S, Green D, et al.A protein kinase involved in the regulation of inflammatory cytokinebiosynthesis. Nature 1994;372:739e46.

[24] Haddad EB, Birrell M, McCluskie K, Ling A, Webber SE, Foster ML, et al. Role ofp38 MAP kinase in LPS-induced airway inflammation in the rat. Br J Phar-macol 2001;132:1715e24.

[25] Schnyder-Candrian S, Quesniaux VF, Di PF, Maillet I, Noulin N, Couillin I, et al.Dual effects of p38 MAPK on TNF-dependent bronchoconstriction and TNF-independent neutrophil recruitment in lipopolysaccharide-induced acuterespiratory distress syndrome. J Immunol 2005;175:262e9.

[26] Sato M, Sano H, Iwaki D, Kudo K, Konishi M, Takahashi H, et al. Direct bindingof Toll-like receptor 2 to zymosan, and zymosan-induced NF-kappa B acti-vation and TNF-alpha secretion are down-regulated by lung collectin surfac-tant protein A. J Immunol 2003;171:417e25.

[27] Lee HY, Takeshita T, Shimada J, Akopyan A, Woo JI, Pan H, et al. Induction ofbeta defensin 2 by NTHi requires TLR2 mediated MyD88 and IRAK-TRAF6-p38MAPK signaling pathway in human middle ear epithelial cells. BMCInfect Dis 2008;8:87.

[28] Kang Y, Wang F, Lu Z, Ying H, Zhang H, Ding W, et al. MAPK kinase 3 po-tentiates Chlamydia HSP60-induced inflammatory response through distinctactivation of NF-kappaB. J Immunol 2013;191:386e94.

[29] De Stefano D, Coletta C, Bianca R, Falcone L, d’Angelo I, Ungaro F, et al. A decoyoligonucleotide to NF-kappaB delivered through inhalable particles preventsLPS-inducedratairway inflammation.AmJRespirCellMolBiol2013;49:288e95.

[30] Nick JA, Young SK, Brown KK, Avdi NJ, Arndt PG, Suratt BT, et al. Role of p38mitogen-activated protein kinase in a murine model of pulmonary inflam-mation. J Immunol 2000;164:2151e9.

[31] Riffo-Vasquez Y, Spina D, Page C, Tormay P, Singh M, Henderson B, et al. Effectof Mycobacterium tuberculosis chaperonins on bronchial eosinophilia andhyper-responsiveness in a murine model of allergic inflammation. Clin ExpAllergy 2004;34:712e9.

[32] Zhang XR, Zhang LY, Devadas S, Li L, Keegan AD, Shi YF. Reciprocal expressionof TRAIL and CD95L in Th1 and Th2 cells: role of apoptosis in T helper subsetdifferentiation. Cell Death Differ 2003;10:203e10.

[33] Glimcher LH, Murphy KM. Lineage commitment in the immune system: the Thelper lymphocyte grows up. Genes Dev 2000;14:1693e711.

[34] Marsland BJ, Kopf M, Le GG. Viral infection and allergy. Nat Immunol 2004;5:865e6.

[35] Shrikant PA, Rao R, Li Q, Kesterson J, Eppolito C, Mischo A, et al. Regulatingfunctional cell fates in CD8 T cells. Immunol Res 2010;46:12e22.

[36] Visekruna A, Ritter J, Scholz T, Campos L, Guralnik A, Poncette L, et al. Tc9 cells,a new subset of CD8(þ) T cells, support Th2-mediated airway inflammation.Eur J Immunol 2013;43:606e18.

[37] Tang Y, Guan SP, Chua BY, Zhou Q, Ho AW, Wong KH, et al. Antigen-specificeffector CD8 T cells regulate allergic responses via IFN-gamma and dendriticcell function. J Allergy Clin Immunol 2012;129:1611e1620.e4.

[38] ten HoveW, Houben LA, Raaijmakers JA, Bracke M, Koenderman L. Differentialregulation of TNFalpha and GM-CSF induced activation of P38 MAPK inneutrophils and eosinophils. Mol Immunol 2007;44:2492e6.

[39] Cho SH, Stanciu LA, Holgate ST, Johnston SL. Increased interleukin-4, inter-leukin-5, and interferon-gamma in airway CD4þ and CD8þ T cells in atopicasthma. Am J Respir Crit Care Med 2005;171:224e30.

[40] van Rensen EL, Sont JK, Evertse CE, Willems LN, Mauad T, Hiemstra PS, et al.Bronchial CD8 cell infiltrate and lung function decline in asthma. Am J RespirCrit Care Med 2005;172:837e41.

[41] Tsitsiou E, Williams AE, Moschos SA, Patel K, Rossios C, Jiang X, et al. Tran-scriptome analysis shows activation of circulating CD8þ T cells in patientswith severe asthma. J Allergy Clin Immunol 2012;129:95e103.

[42] Loh LC, Vyas B, Kanabar V, Kemeny DM, O’Connor BJ. Inhaled endotoxin inhealthy human subjects: a dose-related study on systemic effects and pe-ripheral CD4þ and CD8þ T cells. Respir Med 2006;100:519e28.

[43] Singh M. The burden of asthma in children: an Asian perspective. PaediatrRespir Rev 2005;6:14e9.

[44] Wark PA, Johnston SL, Moric I, Simpson JL, Hensley MJ, Gibson PG. Neutrophildegranulation and cell lysis is associated with clinical severity in virus-induced asthma. Eur Respir J 2002;19:68e75.

[45] Barnes PJ. Medicine. Neutrophils find smoke attractive. Science 2010;330:40e1.

[46] Qiu Y, Zhu J, Bandi V, Guntupalli KK, Jeffery PK. Bronchial mucosal inflam-mation and upregulation of CXC chemoattractants and receptors in severeexacerbations of asthma. Thorax 2007;62:475e82.

[47] Walsh GM. Mepolizumab and eosinophil-mediated disease. Curr Med Chem2009;16:4774e8.

[48] Walsh GM. Reslizumab, a humanized anti-IL-5 mAb for the treatment ofeosinophil-mediated inflammatory conditions. Curr Opin Mol Ther 2009;11:329e36.

[49] Shirtcliffe P, Weatherall M, Travers J, Beasley R. The multiple dimensions ofairways disease: targeting treatment to clinical phenotypes. Curr Opin PulmMed 2011;17:72e8.

[50] Pleasants RA, Ohar JA, Croft JB, Liu Y, Kraft M, Mannino DM, et al. Chronicobstructive pulmonary disease and asthma e patient characteristics and healthimpairment. COPD; 2013. http://dx.doi.org/10.3109/15412555.2013.840571.