Full Terms & Conditions of access and use can be found athttp://www.tandfonline.com/action/journalInformation?journalCode=igrf20

Growth Factors

ISSN: 0897-7194 (Print) 1029-2292 (Online) Journal homepage: http://www.tandfonline.com/loi/igrf20

Reduction of CD19 autoimmunity marker on Bcells of paediatric SLE patients through repressingPU.1/TNF-α/BAFF axis pathway by miR-155

H. R. Aboelenein, M. T. Hamza, H. Marzouk, R. A. Youness, M. Rahmoon, S.Salah & A. I. Abdelaziz

To cite this article: H. R. Aboelenein, M. T. Hamza, H. Marzouk, R. A. Youness, M. Rahmoon, S.Salah & A. I. Abdelaziz (2017) Reduction of CD19 autoimmunity marker on B cells of paediatricSLE patients through repressing PU.1/TNF-α/BAFF axis pathway by miR-155, Growth Factors,35:2-3, 49-60, DOI: 10.1080/08977194.2017.1345900

To link to this article: https://doi.org/10.1080/08977194.2017.1345900

Published online: 07 Jul 2017.

Submit your article to this journal

Article views: 62

View related articles

View Crossmark data

RESEARCH PAPER

Reduction of CD19 autoimmunity marker on B cells of paediatric SLEpatients through repressing PU.1/TNF-a/BAFF axis pathway by miR-155

H. R. Aboeleneina, M. T. Hamzab, H. Marzoukc, R. A. Younessd, M. Rahmoond, S. Salahc and A. I. Abdelazize

aDepartment of Pharmacology and Toxicology, The Molecular Pathology Research Group, German University in Cairo, Cairo, Egypt;bDepartment of Clinical Pathology, Faculty of Medicine, Ain Shams University, Cairo, Egypt; cDepartment of Pediatrics, Faculty ofMedicine, Cairo University, Cairo, Egypt; dDepartment of Pharmaceutical Biology, German University in Cairo, Cairo, Egypt; eSchool ofMedicine, NewGiza University (NGU), Giza, Egypt

ABSTRACTmicroRNA-155 (miR-155) is implicated in regulating B-cell activation and survival that is import-ant in systemic lupus erythematosus (SLE) pathogenesis. PU.1, a target for miR-155, is a crucialregulator of B-cell development and enhances Tumour-Necrosis-factor-alpha (TNF-a) expression.TNF-a induces the expression of B-cell-activating-factor (BAFF). BAFF is reported to increase theexpression of the autoimmunity marker; CD19. This study aimed to investigate the regulation ofexpression of PU.1 in pediatric-systemic-lupus-erythematosus (pSLE) patients by miR-155, andhence evaluate its impact on TNF-a/BAFF/CD19 signalling pathway. Screening revealed that PU.1is upregulated in PBMCs and B-cells of pSLE patients. PU.1 expression directly correlated withsystemic-lupus-erythematosus disease-activity-index-2 K SLEDAI-2K. Ectopic expression of miR-155 and knockdown of PU.1 suppressed PU.1, TNF-a and BAFF. Finally, miR-155 decreased theproportion of BAFF-expressing-B-cells and CD19 protein expression. These findings suggest thatmiR-155 suppresses autoimmunity through transcriptional repression of PU.1 and TNF-a, whichin turn suppresses BAFF and CD19 protein expression.

ARTICLE HISTORYReceived 22 November 2016Revised 8 June 2017Accepted 12 June 2017

KEYWORDSmicroRNA 155; BAFF; PU.1;TNF-a; B cells; lupus

Introduction

Ongoing research is now focusing on the role ofmicroRNAs (miRNAs, miRs) in the regulation ofautoimmunity (Simpson & Ansel, 2015; Garo &Murugaiyan, 2016). miR-155 is among the prominentmiRNAs studied in B cell regulation (Costinean et al.,2006; Calame, 2007). The role of miR-155 in SLEremains controversial. In murine lupus models (MRL-lpr and B6-lpr mice), miR-155 was significantlyincreased in whole splenocytes, as well as in splenic Band T cells (Dai et al., 2010). Deletion of miR-155reduced autoantigen-induced GC reactions in spleensof lupus mice (Thai et al., 2013). In SLE patients,miR-155 urinary levels were reported to be signifi-cantly increased compared to healthy controls and sig-nificantly correlated with proteinuria and systemiclupus erythematosus disease activity index (SLEDAI)(Wang et al., 2012). Paradoxically, that same researchgroup had earlier reports indicating that the serumlevel of miR-155 was lower in SLE patients and corre-lated with glomerular filtration rate estimates, redblood cell count, platelet count, lymphocytic count

and the C-reactive protein (Wang et al., 2010). miR-155 expression levels also showed significant reduc-tion in the plasma of SLE patients (Wang et al.,2012). On the other hand a recent study revealed thatmiR-155 expression was reduced in CD4þT cellsfrom SLE patients versus healthy controls, and uponforcing its expression there was an increase in STAT3phosphorylation and accordingly an increase in IL-21production, therefore they hypothesized that thedecreased levels of miR-155 serves in regulating thepathologically increased IL-21 in SLE (Rasmussenet al., 2015). Similarly, our research group has demon-strated that miR-155 expression is significantly down-regulated in PBMCs from juvenile SLE patients versusage-matched healthy controls, and there was a nega-tive correlation between miR-155 expression andSLEDAI. Moreover, we have shown that forcing theexpression of miR-155 reduced PP2Ac (Protein phos-phatase 2A homologues, catalytic domain) expressionand accordingly increased IL-2 levels in PBMCs ofjuvenile SLE patients (Lashine et al., 2014).

miR-155 has several downstream targets amongwhich is PU.1 that has been previously validated as a

CONTACT Ahmed Ihab Abdelaziz ahmed.ihab.abdelaziz@gmail.com Associate Professor of Molecular Medicine, NewGiza University (NGU), Schoolof Medicine, KM 22 Cairo-Alex road, Giza, Egypt� 2017 Informa UK Limited, trading as Taylor & Francis Group

GROWTH FACTORS, 2017VOL. 35, NO. 2-3, 49–60https://doi.org/10.1080/08977194.2017.1345900

potential target for miR-155, PU.1 is a member of theEts family of transcription factors, that has been iden-tified as a vital transcriptional factor for lymphoidand myeloid development (McKercher et al., 1996;Carotta et al., 2010). PU.1 overexpression has beencorrelated with SLE pathogenesis (Hikami et al.,2011), a recent study reported that PU.1 was overre-presented in the promoters of genes linked to SLEsusceptibility (Dozmorov et al., 2014). Additionally,PU.1 was demonstrated to directly activate the pro-moter of tumour necrosis factor alpha (TNF-a) (Niwaet al., 2008; Fukai et al., 2009). The contribution ofTNF-a to SLE pathogenesis has been also debatable:although initial findings conveyed TNF-a to have aprotective effect against autoimmunity, it was notuntil later when other studies reported its expressionto be directly associated with inflammation, sequentialtissue injury, and organ destruction in SLE (Jacob,1992; Aringer & Smolen, 2003), therefore, TNF-a hasbeen recently considered as a target for moleculartherapy in SLE (Xiong & Lahita, 2011). Interestingly,TNF-a was evidenced to increase the mRNA and pro-tein levels of B cell-activating factor (BAFF) (Leeet al., 2013).

BAFF overexpression has been recognized as a cru-cial factor in systemic lupus erythematosus (SLE)pathogenesis; mice over-expressing BAFF exhibitincreased autoantibody production by B cells whichleads to a progressive autoimmune disease similar toSLE (Mackay et al., 1999). BAFF was reported toinduce the expression of CD19, a major componentof the B-cell coreceptor complex (Hase et al., 2004).Overexpression of CD19 has been linked to abnormal-ities in the immune system; it disrupts peripheral tol-erance in B cells and thereby induces autoantibodyproduction (Sato et al., 2000; Inaoki et al., 1997).

Hence, the regulation of the PU.1/TNF-a/BAFFaxis seems to hold a potential role in controlling SLEdisease activity.

Therefore, this study aimed at further highlightingthe role of miR-155 in SLE pathogenesis, throughinvestigating its impact on PU.1 transcriptional factorand consequently on TNF-a/BAFF/CD19 signalling,in B cells of paediatric SLE (pSLE) patients.

Patients and methods

Patients

Thirty-four patients with SLE were recruited from theDepartment of Rheumatology, Aboelreesh PediatricHospital, Cairo University Medical School. All study-recruited patients were under the age of 18 and

fulfilled the American College of Rheumatology(ACR) 1997 revised classification criteria for SLE(Hochberg, 1997). Disease activity of each patient wasassessed at the time of sample collection using sys-temic lupus erythematosus disease activity indexSLEDAI-2K, and was used to divide patients intoactive and inactive groups where active SLE diseasewas defined as SLEDAI-2 K score of �6 and inactiveSLE disease was defined as SLEDAI-2K score of <6(Khanna et al., 2004). Absolute lymphocyte counts(ALC) were recorded; lymphopenia was defined asALC <1000/L. Additional clinical data of the pSLEpatients involved in this study is shown in Table 1.

Healthy, age-matched blood donors with no historyof autoimmune diseases or treatment with immuno-suppressive agents served as controls. Legal guardiansof all patients and controls provided a writteninformed consent before collection of peripheral bloodsamples. The study was approved by the GermanUniversity in Cairo and Cairo University’s ethicalreview committees and experiments were performedin accordance with the Helsinki Declaration of 1975.

Collection of samples and separation of PBMCs

Ten millilitres of venous peripheral blood were with-drawn from all SLE patients and healthy controls inthe presence of the anticoagulant EDTA. Peripheralblood mononuclear cells (PBMCs) were isolated fromthe whole blood using the Ficoll density gradient cen-trifugation method within 6 h of blood withdrawal.The isolated PBMCs were cryopreserved and stored at�80 �C.

B cell isolation

PBMCs of 6–7 patients were pooled and purified Bcells were subsequently isolated by negative selection

Table 1. Clinical data of SLE patients recruited in the currentstudy.Sex (male/female) 10/24Age [(years) mean ± SEM] 12.15 ± 0.5791Disease duration [(years) mean ± SEM] 3.447 ± 0.5128SLEDAI-2Ka [mean ± SEM] 5.676 ± 0.6482Activity [active/inactive] 14/20Antids-DNAb [þve/�ve] 20/14ANAc [þve/�ve] 34/0Lupus nephritis [þve/�ve] 20/14Lymphopenia [þve/�ve] 14/20Neuropsychiatric manifestations [þve/�ve] 3/31Prednisone intake [þve/�ve] 17/17Cyclophosphamide intake [þve/�ve] 4/30Azathioprine intake [þve/�ve] 19/15Hydroquine intake [þve/�ve] 31/3aSLEDAI-2 K – SLE disease activity index 2000.bAnti-dsDNA antibody – anti-double stranded DNA antibody.cANA – anti-nuclear antibody.

50 H. R. ABOELENEIN ET AL.

using the human B Cell Isolation Kit II, human(MACS, MiltenyiBiotec, San Diego, CA) according tomanufacturer’s recommendations.

Cell culture and transfection

PBMCs or B cells were cultured in RPMI 1640 media(Lonza, Visp, Switzerland) supplemented with 10%fetal bovine serum (FBS) and 1% penicillin/strepto-mycin and Mycozap (Lonza, Visp, Switzerland). Cellswere transfected with scrambled miRNAs or mimicsand inhibitors of miR-155 (Qiagen IDs: MSY0000646and MIN0000646 respectively) or PU.1 siRNA(Qiagen ID: number SI00047971) using HiPerfecttransfection reagent (all supplied by Qiagen, Hilden,Germany) according to the manufacturer’s protocol.Transfected cells were incubated for 48 h or 72 h untilanalysis.

PBMCs or B cells referred to as mock were thosetreated with the transfecting agent without the oligo-nucleotides. After each transfection, the levels of miR-155 were assessed in cells transfected with miR-155mimics compared to mock cells to confirm transfec-tion efficiency; transfection resulted in an averageincrease in miR-155 levels of more than 300-fold inPBMCs and more than 500-fold increase in B cells.

mRNA and miRNA quantification

Total RNA was extracted using mirVana extractionkit according to the manufacturer’s protocol (AmbionApplied Biosystems, Drive Foster City, CA). Theextracted mRNA and miRNA fractions were reverse-transcribed into single stranded complementary DNA(cDNA) using the High Capacity cDNA ReverseTranscription Archive kit (Applied Biosystems, LifeTechnologies, Drive Foster City, CA). The relativeexpression of the genes of interest as well as themiRNA was then quantified using StepOneTM Real-Time PCR instrument (Applied Biosystems, DriveFoster City, CA) and StepOneTM Software (AppliedBiosystems). Expression of PU.1, TNF-a and BAFFwas normalized to 18 S housekeeping gene. miR-155expression was normalized to RNU6B housekeepingsmall RNA. All PCR reactions were carried out intriplicates.

Flow cytometric analysis

BAFF and CD19 cell-surface expression was quanti-fied using a Coulter Epics XL flow cytometer(Beckman-coulter, Miami, FL). Cell aliquots wereincubated for 30min at 4 �C with FITC conjugated

anti-BAFF (monoclonal mouse IgM) and PE-CY5conjugated anti-CD19 (monoclonal mouse IgM) ormatching isotype control (all supplied by Santa CruzBiotechnology, Dallas, TX), then washed with PBSsupplemented with 2% FCS. The mean fluorescenceintensities (MFI); which denotes the intensity ofexpression; as well as the percentages of the cellsexpressing CD19 or BAFF were analysed using Kaluzasoftware (Beckman-coulter, Miami, FL).

Statistical analysis

The relative expression of the genes is expressed inrelative quantitation (RQ¼ 2�DDCT). All values arerepresented as the mean ± standard error of the mean(SEM). Mann–Whitney test was used for statisticalanalysis. Calculations were performed using Graph-Pad Prism 5.00 software (GraphPad Software Inc., LaJolla, CA). Each experiment was independently per-formed at least three times. All statistical tests weretwo tailed and a p value <.05 was considered statistic-ally significant. For correlation analysis, Pearson’s cor-relation coefficient (r) was used; where r value showsif there is a positive or negative correlation betweenthe 2 variables and (p) indicates whether this correl-ation is significant or not.

Results

PU.1 is up-regulated in PBMCs and B cells of pSLEpatients

PBMCs were isolated from individual pSLE patients,while B cells were isolated from pooled PBMCs ofpSLE patients and age-matched healthy controls. Therelative expression of PU.1 was quantified using real-time PCR and was found to be markedly up-regulatedin PBMCs of pSLE patients (n¼ 15) compared tohealthy controls (n¼ 7) (p¼ .0289) (Figure 1(A)).Similarly, the expression of PU.1 was significantlyhigher in pooled B cells of pSLE patients (n¼ 5)compared to healthy controls (n¼ 7) (p¼ .0057)(Figure 1(A)).

To determine whether PU.1 expression wasassociated with disease activity, a correlation ana-lysis was performed between PU.1 expression inPBMCs of pSLE patients and their SLEDAI-2K,and results revealed a significant positive correlation(r¼ .5804) (p¼ .0233) (Figure 1(B)). Furthermore,patients were sub-grouped into active (SLEDAI-2K>6) and inactive (SLEDAI-2 K6) pSLE patients,and their PU.1 relative expression was then com-pared. Interestingly, active pSLE patients (n¼ 6)

GROWTH FACTORS 51

had significantly higher levels of PU.1 comparedto inactive pSLE patients (n¼ 9) (p¼ .0256)(Figure 1(C)).

Since PU.1 is an important regulator of lympho-poeisis (Iwasaki et al., 2005), it was interesting toassociate its transcript levels with lymphopenia inpSLE patients. Therefore, patients were divided intotwo groups according to the presence or absence oflymphopenia. The relative expression of PU.1 wasmarkedly elevated in lymphopenic pSLE patients (7)compared to patients without lymphopenia (n¼ 8)(p¼ .0304) (Figure 1(D)).

TNF-a and BAFF are up-regulated in PBMCs and Bcells of pSLE patients

Since PU.1 is a positive regulator of TNF-a transcrip-tion (Niwa et al., 2008) and TNF-a in turn inducesBAFF expression (Lee et al., 2013; Ittah et al., 2006)and both TNF-a and BAFF contribute to SLE patho-genesis (Jacob, 1992; Aringer & Smolen, 2003; Zhanget al., 2001; Stohl et al., 2003), hence, it was intriguingto assess TNF-a and BAFF mRNA levels in PBMCsand B cells of pSLE patients. Results showed higherexpression of TNF-a and BAFF in PBMCs of pSLE

Figure 1. Expression of PU.1 mRNA in pSLE patients and its association with disease activity and lymphopenia. (A) PU.1 was sig-nificantly up-regulated in PBMCs of pSLE patients (n¼ 15) compared to healthy controls (n¼ 7) (p¼ .0289). Likewise, the expres-sion of PU.1 was significantly higher in pooled B cells of pSLE patients (n¼ 5) compared to healthy controls (n¼ 7) (p¼ .0057).Mann–Whitney was used to assess differences between two samples. ����p< .0001, ���p< .001, ��p< .01, �p< .05. All PCRreactions were carried out in triplicate. (B) Spearman’s analysis showed a positive correlation analysis between SLEDAI-2K and PU.1expression in PBMCs of pSLE patients (r¼ .5804) (p¼ .0233). (C) Active pSLE patients (SLEDAI-2k> 6) (n¼ 6) had significantlyincreased transcripts of PU.1 compared to inactive pSLE patients (SLEDAI-2k� 6) (n¼ 9) (p¼ .0256). Mann–Whitney was used toassess differences between two samples. ����p< .0001, ���p< .001, ��p< .01, �p< .05. All PCR reactions were carried out intriplicate. (D) The relative expression of PU.1 showed a significant increase in PBMCs of pSLE patients suffering from lymphopenia(n¼ 7) compared to those without lymphopenia (n¼ 8) (p¼ .0304). Mann–Whitney was used to assess differences between twosamples. ����p< .0001, ���p< .001, ��p< .01, �p< .05. All PCR reactions were carried out in triplicate.

52 H. R. ABOELENEIN ET AL.

patients (n¼ 15) versus age-matched healthycontrols (n¼ 7) (p¼ .0145 and p¼ .024, respectively)(Figure 2(A,B)). Likewise, pooled B cells of pSLEpatients (n¼ 5) showed notably elevated TNF-a andBAFF mRNA expression compared to pooled B cellsof healthy controls (n¼ 7) (p¼ .0025 and p¼ .0025respectively) (Figure 2(A,B)).

A positive correlation between TNF-a and BAFFexpression in pSLE patients

To examine whether the expression of TNF-a andBAFF are correlated in pSLE patients, Pearson’s cor-relation test was performed. Interestingly, a direct cor-relation was found between mRNA expression of bothTNF-a and BAFF (r¼ .7631) (p¼ .0006) in PBMCs ofpSLE patients (Figure 2(C)).

miR-155 suppresses PU.1, TNF- a and BAFFexpression in PBMCs and B cells of pSLE patients

In order to confirm good transfection efficiency, thechange in expression of miRNA, relative to mock cellstreated with transfection reagent alone, was deter-mined by TaqMan real-time quantitative PCR, nor-malized to RNU6B endogenous control. PBMCs andB cells transfected with miR-155 mimics showedhigher levels of miR-155 expression compared tomock cells (p< .0001 and p¼ .007, respectively)(Figure 3(A)).

In order to assess the effect of miR-155 on PU.1,TNF-a and BAFF mRNA, the relative expression ofthe three genes was analysed after miR-155 mimickingand antagonizing in PBMCs of pSLE patients. Resultsshowed that ectopic expression of miR-155 suppressedthe expression of PU.1, TNF-a and BAFF in PBMCs

Figure 2. Expression profile of TNF-a and BAFF in pSLE patients and correlation of their expression. (A) The relative expression ofTNF-a was higher in PBMCs of pSLE patients (n¼ 15) compared to healthy controls (n¼ 7) (p¼ .0145). Consistently, pooled B cellsof pSLE patients (n¼ 5) had higher TNF-a expression compared to pooled B cells of healthy controls (n¼ 7) (p¼ .0025).Mann–Whitney was used to assess differences between two samples. ����p< .0001, ���p< .001, ��p< .01, �p< .05. All PCRreactions were carried out in triplicate. (B) Transcript levels of BAFF were higher in PBMCs of pSLE patients (n¼ 15) compared tohealthy controls (n¼ 7) (p¼ .024). Comparably, pooled B cells of pSLE patients (n¼ 5) had increased expression of BAFF mRNAcompared to pooled B cells of age-matched healthy controls (n¼ 7) (p¼ .0025). Mann–Whitney was used to assess differencesbetween two samples. ����p< .0001, ���p< .001, ��p< .01, �p< .05. All PCR reactions were carried out in triplicate. (C)Pearson’s correlation test revealed a direct correlation between mRNA expression of both TNF-a and BAFF in PBMCs of pSLEpatients (r¼ .7631) (p¼ .0006).

GROWTH FACTORS 53

of pSLE patients compared to mock cells (p¼ .0029,p¼ .0121 and p¼ .008, respectively) (Figure 3(B–D)).Interestingly knockdown of PU.1 by siRNAs remark-ably reduced transcripts of PU.1, TNF-a and BAFF inPBMCs of pSLE patients compared to mock cells(p¼ .0058, p¼ .0485 and p¼ .007, respectively)(Figure 3(B–D)). Similarly, transfection of miR-155mimics in B cells of pSLE patients significantly

down-regulated PU.1, TNF-a and BAFF expressioncompared to mock B cells (p< .0001, p¼ .0238 andp¼ .004, respectively) (Figure 3(B–D)). PU.1 siRNAsresulted in a reduced expression of PU.1, TNF-a andBAFF in B cells of pSLE patients (p¼ .0007, p¼ .0357and p¼ .0286, respectively) (Figure 3(B–D)). On theother hand, PBMCs and B cells treated with miR-155inhibitors had mRNA levels of PU.1, TNF-a and

Figure 3. Transfection efficiency of miR-155 oligonucleotides and its impact on PU.1, TNF- a and BAFF mRNA Expression in pSLEpatients. (A) Transfection of miR-155 mimics in PBMCs and B cells of pSLE patients increased miR-155 levels significantly comparedto Mock cells (p< .0001 and p¼ .007, respectively). miR-155 expression was determined by TaqMan RTqPCR and normalized toRNU6B endogenous control. All PCR reactions were carried out in triplicate. (B) Ectopic expression of miR-155 reduced PU.1 tran-scripts in PBMCs and B cells of pSLE patients compared to mock cells (p¼ .0029 and p< .0001, respectively). Moreover, knockingdown PU.1 by siRNAs suppressed PU.1 mRNA in PBMCs and B cells of pSLE patients(p¼ .0058 and p¼ .0007, respectively).Mann–Whitney was used to assess differences between two samples. ����p< .0001, ���p< .001, ��p< .01, �p< .05. All PCRreactions were carried out in triplicate. (C) miR-155 down-regulated TNF-a mRNA expression in PBMCs and B cells of pSLE patientscompared to mock cells (p¼ .0121 and p¼ .0238, respectively). PU.1 siRNAs reduced transcripts of TNF-a in PBMCs and B cells ofpSLE patients (p¼ .0485 and p¼ .0357, respectively). Mann–Whitney was used to assess differences between two samples.����p< .0001, ���p< .001, ��p< .01, �p< .05. All PCR reactions were carried out in triplicate. (D) miR-155 reduced BAFF mRNAin PBMCs and B cells of pSLE patients compared to mock cells (p¼ .008 and p¼ .004, respectively). Intriguingly, siRNAs againstPU.1 reduced BAFF mRNAin PBMCs and B cells of pSLE patients (p¼ .007 and p¼ .0286, respectively). Mann–Whitney was used toassess differences between two samples. ����p< .0001, ���p< .001, ��p< .01, �p< .05. All PCR reactions were carried out intriplicate.

54 H. R. ABOELENEIN ET AL.

BAFF comparable to mock cells. To assess the effectof miR-155 on BAFF protein expression in B cells,PBMCs were stained with FITC-conjugated anti-BAFFand PE-CY5 conjugated anti-CD19 antibodies. TheMFIs of CD19 and BAFF were analyzed by flowcytometry. BAFF surface expression was significantlyreduced on CD19þB cells treated with miR-155mimics or PU.1 siRNA (p¼ .0095 and p¼ .0190,respectively) (Figure 4(A,B)).

CD19 expression is increased in pSLE patients

CD19 is an important regulator of B cell function andautoimmunity (Hasegawa et al., 2001). Flow cytomet-ric analysis showed that CD19 expression is signifi-cantly increased in pSLE patients (n¼ 10) comparedto healthy controls (n¼ 4) (p¼ .033) (Figure 4(C,D)).

miR-155 reduces CD19 expression in pSLE patients

In order to investigate the impact of miR-155 andPU.1 on CD19 protein expression in pSLE patients,PBMCs were transfected with miR-155 mimics/inhibi-tors or PU.1 siRNAs. CD19 was markedly suppressedon miR-155 mimicked cells compared to mock cells(p¼ .0186). Interestingly, cells treated with PU.1siRNAs also had reduced CD19 protein expressioncompared to mock cells (p¼ .0167), while miR-155antagonized PBMCs showed levels of CD19 similar tomock cells (Figure 4(E)).

miR-155 reduces BAFFþB cells in pSLE patients

The proportion of BAFF-positive cells were analyzedin mock, miR-155 mimicked/antagonized and PU.1siRNA-treated CD19þB cells. Results revealed thatmiR-155 as well as PU.1 siRNA dramaticallydecreased the proportion of BAFFþB cells (p¼ .0095and p¼ .0095, respectively) (Figure 5(A,B)).

Discussion

The controversial role of miR-155 in SLE triggeredour interest to evaluate the impact of this miRNA onPU.1 transcription factor, one of its novel downstreamtargets that is highly involved in B cell developmentand regulation (Vigorito et al., 2007; McKercher et al.,1996). We have recently revealed miR-155 to beunderexpressed in pSLE patients (Lashine et al.,2014).

Therefore, the current study initially analysed PU.1expression profile in pSLE patients. Our resultsshowed that PU.1 expression is remarkably increased

in both PBMCs and B cells from pSLE patients versusage-matched healthy controls. This may be supportedby previous data reporting PU.1 gene to be hypome-thylated in SLE (Javierre et al., 2010), it also stands inline with data stating that PU.1 transcription factor isinterlinked with the promoters of genes related to SLEpathogenesis (Dozmorov et al., 2014).

We further aimed at examining the associationbetween PU.1 expression and clinical features in SLEpatients. Interestingly, PU.1 expression demonstrateda positive correlation with SLEDAI-2K. Similarly,active SLE patients and those suffering from lympho-penia showed notably higher expression of PU.1 thantheir inactive or non-lymphopenic counterparts. Thesefindings provide further support for a previous studythat suggested PU.1 up-regulation to play a role inthe pathogenesis of SLE (Hikami et al., 2011).

This study investigated the effect of manipulatingmiR-155 on PU.1 expression in PBMCs and B cells ofpSLE patients and its consequent impact on B cellpathology. Our results revealed that forcing theexpression of miR-155 has markedly reduced PU.1mRNA expression in PBMCs and B cells of pSLEpatients compared to mock cells. Indeed, the suppres-sive effect of miR-155 on PU.1 was comparable to theeffect of knocking down PU.1 by siRNAs.

PU.1 overexpression has been suggested to enhanceTNF-a transcription (Niwa et al., 2008). This notionwas further validated by a study demonstrating thatknocking down PU.1 noticeably reduced TNF-a pro-tein in dendritic cells, while increasing PU.1 led to anincreased production of TNF-a. Therefore, it was pro-posed that TNF-a promoter is directly activated byPU.1(Fukai et al., 2009). For these reasons, it was ofour interest to analyse the expression profile of TNF-a in pSLE patients and to explore the effect of manip-ulating PU.1 expression by miR-155 on the expressionof TNF-a. Our Results showed considerable augmen-tation of TNF-a baseline expression in PBMCs and Bcells of pSLE patients. This finding is consistent withstudies reporting that TNF-a is markedly increased inSLE (Gabay et al., 1997; Sabry et al., 2006; Studnicka-Benke et al., 1996) and that anti-TNF therapyachieved a good outcome in arthritis and nephritisSLE patients (Aringer & Smolen, 2008). Nevertheless,it may contradict previous data that have reportedTNF-a levels to be higher in SLE patients withinactive disease compared to patients with active dis-ease (Gomez et al., 2004).

Interestingly, our study demonstrated a notableinhibitory effect of miR-155 on TNF-a expression inPBMCs and B cells of pSLE patients suggesting that

GROWTH FACTORS 55

Figure 4. Flow cytometric analysis of BAFF and CD19 cell surface expression in pSLE patients. (A) Flow cytometric histogram show-ing BAFF and CD19 expression (MFI; x-axis) on PBMCs of pSLE patients after different manipulations. PBMCs were stained withanti-CD19 antibody conjugated to PE-CY5 and anti-BAFF antibody conjugated to FITC. CD19 and BAFF levels were decreased oncells treated with miR-155 mimics or those treated with PU.1 siRNA compared to Mock cells. While miR-155 antagonized cells hadsimilar expression of CD19 as Mock cells. (B) Percentages of MFI, as measured by flow cytometry, of BAFF protein expressionassessed on CD19þ B cells of pSLE patients. miR-155 and PU.1 siRNA notably reduced the expression of BAFF on CD19þ B cellsof pSLE patients (p¼ .0095 and p¼ .0190, respectively). Mann–Whitney was used to assess differences between two samples.����p< .0001, ���p< .001, ��p< .01, �p< .05. The experiment was done in triplicate. (C) Flow cytometric histogram showingCD19 expression on PBMCs of pSLE patients and healthy controls. Background fluorescence was determined with an irrelevant iso-type-matched antibody. SLE patients had higher CD19 protein expression compared to healthy controls. (D) Percentages of MFI ofCD19 expression were increased on PBMCs of pSLE patients (n¼ 10) compared to healthy controls (n¼ 4) (p¼ .033).Mann–Whitney was used to assess differences between two samples. ����p< .0001, ���p< .001, ��p< .01, �p< .05. The experi-ment was done in triplicate. (E) miR-155 mimicked and PU.1 knocked-down PBMCs had decreased CD19 expression (p¼ .0186 andp¼ .0167, respectively) compared to Mock cells. miR-155 inhibitors did not affect CD19 expression. Mann–Whitney was used toassess differences between two samples. ����p< .0001, ���p< .001, ��p< .01, �p< .05. The experiment was done in triplicate.

56 H. R. ABOELENEIN ET AL.

this is the impact of repressing PU.1 which was con-firmed by bioinformatics analysis that did not predictmiR-155 to have any binding sites on the 3’UTR ofTNF-a. Therefore, for further confirmation thatTNF-a reduction is due to PU.1 repression, PU.1 wasknocked down using siRNAs and a suppressive effecton TNF-a expression similar to that obtained by miR-

155 mimics was detected in both PBMCs and B cellsof pSLE patients. Hence, miR-155 may be an indirectregulator of TNF-a mediating its action throughPU.1.

In SLE patients, serum levels of BAFF were foundto be aberrantly high and correlated with diseaseactivity (Zhang et al., 2001; Stohl et al., 2003).

Figure 5. Flow cytometric analysis of BAFF positive CD19þ B cells. (A) Representative dot plots of flow cytometry data showingisotype controls, CD19-PE and BAFF-FITC expression on mock PBMCs, PBMCs treated with miR-155 mimics/inhibitors or PU.1 siRNA.Numbers represent the percentages of BAFFþ CD19þ cells, data shown is representative of four experiments (n¼ 4). The percent-age of BAFFþ B cells was remarkably decreased in cells treated with miR-155 mimics or PU.1 siRNA. (B) The proportion ofBAFFþ B cells (n¼ 4) was significantly reduced upon mimicking with miR-155 (n¼ 4) or upon knocking down PU.1 (n¼ 4).Mann–Whitney was used to assess differences between two samples. ����p< .0001, ���p< .001, ��p< .01, �p< .05. The experi-ment was done in triplicate.

GROWTH FACTORS 57

Moreover, Belimumab, a monoclonal antibody againstBAFF, has been recently approved for SLE therapy(Navarra et al., 2011).

Our study further quantified BAFF gene expressionin PBMCs and B cells of pSLE patients and revealed itto be evidently enhanced in pSLE patients. We havealso shown that BAFF mRNA expression is positivelycorrelated with TNF-a mRNA expression in PBMCsof pSLE patients and since TNF-a shown in severalstudies to enhance BAFF expression (Lee et al., 2013;Ittah et al., 2006; Assi et al., 2007), it was intriguingto detect the impact of miR-155 and PU.1 on BAFFexpression. Interestingly, both miR-155 mimics andPU.1 siRNAs considerably suppressed BAFF mRNAand protein levels in B cells and PBMCs of pSLEpatients, and also markedly decreased the proportionof BAFF-positive B cells as quantified by flowcytome-ter. These findings appear to partially oppose previousliterature which reported miR-155 inhibitors to sig-nificantly reduce BAFF-R expression in B cells; how-ever the latter study was performed on B cells frommyasthenia gravis patients (Wang et al., 2014), anexplanation for this might be that miRNAs act in acell-specific or disease-specific manner where Linet al. found that miR-206 can act in an opposite man-ner on the expression of the same target (KLF4) innormal versus cancer cells (Lin et al., 2011). Thisshows that miRNAs action might depend on thepathological status of the cell. In similarity with theTNF-a bioinformatics results, we found no predictedbinding regions for miR-155 on the BAFF 3’UTR.Hence, it is plausible that the effect of miR-155 onBAFF is through transcriptional repression of PU.1 aswell as TNF-a.

Dysregulated CD19 expression has been associatedwith abnormalities of the immune system, whereupregulation of CD19 was linked to the developmentof autoantibodies (Sato et al., 2000). In SLE, CD19was somehow controversial; while a study reportedCD19 levels to be highly expressed on B cells fromSLE patients versus healthy controls (Mei et al., 2012).Other studies reported about 20% reduction in CD19expression levels by naïve and memory B cells fromSLE patients (Sato et al., 2000; Sato et al., 2004).Accordingly, and in a trial to identify the status ofCD19 in SLE patients, CD19 expression was assessedusing FACS analysis; our results showed that CD19protein expression is inherently upregulated in pSLEpatients compared to healthy controls. This goes inline with the fact that the expression of CD19 on all Bcell subsets; memory cells, plasmablasts and naive Bcells correlates with each other in SLE patients.

However, CD19 expressed by plasmablasts of SLEpatients was found to be less than that expressed bymemory cells and naive B cells (Mei et al., 2012).Therefore, further investigation is needed to assessCD19 expression in different B cell subsets of SLEpatients. CD19 has been reported to be indirectlyenhanced by BAFF-R activation. Such activation mod-ulates the transcription factor Pax5 and consequentlyaugments CD19 expression, resulting in an enhancedB cell Receptor signalling (Hase et al., 2004).Therefore, the impact of miR-155 on CD19 expressionwas analysed. Results showed that forcing the expres-sion of miR-155 as well as knocking down PU.1 bysiRNAs significantly reduced CD19 proteinexpression.

In conclusion, we attempted to further evaluatemiR-155 in SLE, through investigating its net impacton autoimmune stimulatory proteins BAFF andCD19. Our results revealed that miR-155 acts as asuppressor of autoimmunity through transcriptionalrepression of PU.1 and TNF-a, which in turn sup-presses BAFF and CD19 protein expression.

Disclosure statement

The authors declare no commercial or financial conflict ofinterest.

References

Aringer M, Smolen JS. 2003. SLE – Complex cytokineeffects in a complex autoimmune disease: Tumor necrosisfactor in systemic lupus erythematosus. Arthritis ResTher 5:172–177.

Aringer M, Smolen JS. 2008. Efficacy and safety of TNF-blocker therapy in systemic lupus erythematosus. ExpertOpin Drug Saf 7:411–419.

Assi LK, Wong SH, Ludwig A, Raza K, Gordon C, Salmon M,Lord JM, Scheel-Toellner D. 2007. Tumor necrosis factoralpha activates release of B lymphocyte stimulator by neu-trophils infiltrating the rheumatoid joint. Arthritis Rheum56:1776–1786.

Calame K. 2007. MicroRNA-155 function in B Cells.Immunity 27:825–827.

Carotta S, Dakic A, D’Amico A, Pang SH, Greig KT, NuttSL, Wu L. 2010. The transcription factor PU.1 controlsdendritic cell development and Flt3 cytokine receptorexpression in a dose-dependent manner. Immunity32:628–641.

Costinean S, Zanesi N, Pekarsky Y, Tili E, Volinia S,Heerema N, Croce CM. 2006. Pre-B cell proliferationand lymphoblastic leukemia/high-grade lymphoma inE(mu)-miR155 transgenic mice. Proc Natl Acad Sci USA103:7024–7029.

Dai R, Zhang Y, Khan D, Heid B, Caudell D, Crasta O,Ahmed SA. 2010. Identification of a common lupus

58 H. R. ABOELENEIN ET AL.

disease-associated microRNA expression pattern in threedifferent murine models of lupus. PLoS One 5:e14302.

Dozmorov MG, Wren JD, Alarcon-Riquelme ME. 2014.Epigenomic elements enriched in the promoters of auto-immunity susceptibility genes. Epigenetics 9:276–285.

Fukai T, Nishiyama C, Kanada S, Nakano N, Hara M,Tokura T, Ikeda S, et al. 2009. Involvement of PU.1 inthe transcriptional regulation of TNF-alpha. BiochemBiophys Res Commun 388:102–106.

Gabay C, Cakir N, Moral F, Roux-Lombard P, Meyer O,Dayer JM, Vischer T, et al. 1997. Circulating levels oftumor necrosis factor soluble receptors in systemic lupuserythematosus are significantly higher than in otherrheumatic diseases and correlate with disease activity.J Rheumatol 24:303–308.

Garo LP, Murugaiyan G. 2016. Contribution of MicroRNAsto autoimmune diseases. Cell Mol Life Sci 73:2041–2051.

Gomez D, Correa PA, G�omez LM, Cadena J, Molina JF,Anaya JM. 2004. Th1/Th2 cytokines in patients with sys-temic lupus erythematosus: Is tumor necrosis factor alphaprotective? Semin Arthritis Rheum 33:404–413.

Hase H, Kanno Y, Kojima M, Hasegawa K, Sakurai D,Kojima H, Tsuchiya N, et al. 2004. BAFF/BLyS canpotentiate B-cell selection with the B-cell coreceptorcomplex. Blood 103:2257–2265.

Hasegawa M, Fujimoto M, Poe JC, Steeber DA, Lowell CA,Tedder TF. 2001. A CD19-dependent signaling pathwayregulates autoimmunity in Lyn-deficient mice.J Immunol 167:2469–2478.

Hikami K, Kawasaki A, Ito I, Koga M, Ito S, Hayashi T,Matsumoto I, et al. 2011. Association of a functionalpolymorphism in the 3'-untranslated region of SPI1 withsystemic lupus erythematosus. Arthritis Rheum63:755–763.

Hochberg MC. 1997. Updating the American College ofRheumatology revised criteria for the classification of sys-temic lupus erythematosus. Arthritis Rheum 40:1725.

Inaoki M, Sato S, Weintraub BC, Goodnow CC, Tedder TF.1997. CD19-regulated signaling thresholds control per-ipheral tolerance and autoantibody production in B lym-phocytes. J Exp Med 186:1923–1931.

Ittah M, Miceli-Richard C, Eric Gottenberg J, Lavie F,Lazure T, Ba N, Sellam J, et al. 2006. B cell-activating fac-tor of the tumor necrosis factor family (BAFF) isexpressed under stimulation by interferon in salivarygland epithelial cells in primary Sjogren's syndrome.Arthritis Res Ther 8:R51.

Iwasaki H, Somoza C, Shigematsu H, Duprez EA, Iwasaki-Arai J, Mizuno S, Arinobu Y, et al. 2005. Distinctive andindispensable roles of PU.1 in maintenance of hemato-poietic stem cells and their differentiation. Blood106:1590–1600.

Jacob CO. 1992. Tumor necrosis factor alpha in auto-immunity: Pretty girl or old witch? Immunol Today13:122–125.

Javierre BM, Fernandez AF, Richter J, Al-Shahrour F,Martin-Subero JI, Rodriguez-Ubreva J, Berdasco M, et al.2010. Changes in the pattern of DNA methylation associ-ate with twin discordance in systemic lupus erythemato-sus. Genome Res 20:170–179.

Khanna S, Pal H, Pandey RM, Handa R. 2004. The relation-ship between disease activity and quality of life in

systemic lupus erythematosus. Rheumatology (Oxford)43:1536–1540.

Lashine YA, Salah S, Aboelenein HR, Abdelaziz AI. 2014.Correcting the expression of miRNA 155 repressesPP2Ac and enhances the release of IL-2 in PBMCs ofjuvenile SLE patients. Lupus 24:240–247.

Lee GH, Lee J, Lee JW, Choi WS, Moon EY. 2013. B cellactivating factor-dependent expression of vascular endo-thelial growth factor in MH7A human synoviocytesstimulated with tumor necrosis factor-alpha. IntImmunopharmacol 17:142–147.

Lin CC, Liu LZ, Addison JB, Wonderlin WF, Ivanov AV,Ruppert JM. 2011. A KLF4-miRNA-206 autoregulatoryfeedback loop can promote or inhibit protein translationdepending upon cell context. Mol Cell Biol31:2513–2527.

Mackay F, Woodcock SA, Lawton P, Ambrose C, Baetscher M,Schneider P, Tschopp J, Browning JL. 1999. Mice transgenicfor BAFF develop lymphocytic disorders along with auto-immune manifestations. J Exp Med 190:1697–1710.

McKercher SR, Torbett BE, Anderson KL, Henkel GW,Vestal DJ, Baribault H, Klemsz M, et al. 1996. Targeteddisruption of the PU.1 gene results in multiple hemato-poietic abnormalities. EMBO J 15:5647–5658.

Mei HE, Schmidt S, Dorner T. 2012. Rationale of anti-CD19 immunotherapy: An option to target autoreactiveplasma cells in autoimmunity. Arthritis Res Ther14(Suppl 5):S1.

Navarra SV, Guzm�an RM, Gallacher AE, Hall S, Levy RA,Jimenez RE, Li EK, et al. 2011. Efficacy and safety of beli-mumab in patients with active systemic lupus erythema-tosus: A randomised, placebo-controlled, phase 3 trial.Lancet 377:721–731.

Niwa Y, Nishiyama C, Nakano N, Kamei A, Kato H,Kanada S, Ikeda S, et al. 2008. Opposite effects of PU.1on mast cell stimulation. Biochem Biophys Res Commun375:95–100.

Rasmussen TK, Andersen T, Bak RO, Yiu G, Sørensen CM,Pedersen KS, Mikkelsen JG, et al. 2015. Overexpressionof microRNA-155 increases IL-21 mediated STAT3 sig-naling and IL-21 production in systemic lupus erythema-tosus. Arthritis Res Ther 17:154.

Sabry A, Sheashaa H, El-Husseini A, Mahmoud K,Eldahshan KF, George SK, Abdel-Khalek E, et al. 2006.Proinflammatory cytokines (TNF-alpha and IL-6) inEgyptian patients with SLE: Its correlation with diseaseactivity. Cytokine 35:148–153.

Sato S, Fujimoto M, Hasegawa M, Takehara K. 2004.Altered blood B lymphocyte homeostasis in systemicsclerosis: Expanded naive B cells and diminished but acti-vated memory B cells. Arthritis Rheum 50:1918–1927.

Sato S, Hasegawa M, Fujimoto M, Tedder TF, Takehara K.2000. Quantitative genetic variation in CD19 expressioncorrelates with autoimmunity. J Immunol 165:6635–6643.

Simpson LJ, Ansel KM. 2015. MicroRNA regulation oflymphocyte tolerance and autoimmunity. J Clin Invest125:2242–2249.

Stohl W, Metyas S, Tan SM, Cheema GS, Oamar B, Xu D,Roschke V, et al. 2003. B lymphocyte stimulatoroverexpression in patients with systemic lupus erythema-tosus: Longitudinal observations. Arthritis Rheum48:3475–3486.

GROWTH FACTORS 59

Studnicka-Benke A, Steiner G, Petera P, Smolen JS. 1996.Tumour necrosis factor alpha and its soluble receptorsparallel clinical disease and autoimmune activity in sys-temic lupus erythematosus. Br J Rheumatol 35:1067–1074.

Thai TH, Patterson HC, Pham DH, Kis-Toth K, KaminskiDA, Tsokos GC. 2013. Deletion of microRNA-155reduces autoantibody responses and alleviates lupus-likedisease in the Fas(lpr) mouse. Proc Natl Acad Sci USA110:20194–20199.

Vigorito E, Perks KL, Abreu-Goodger C, Bunting S, Xiang Z,Kohlhaas S, Das PP, et al. 2007. microRNA-155 regulatesthe generation of immunoglobulin class-switched plasmacells. Immunity 27:847–859.

Wang G, Tam LS, Kwan BC, Li EK, Chow KM, Luk CC, LiPK, Szeto CC. 2012. Expression of miR-146a and miR-155 in the urinary sediment of systemic lupus erythema-tosus. Clin Rheumatol 31:435–440.

Wang G, Tam LS, Li EK, Kwan BC, Chow KM, Luk CC,Li PK, Szeto CC. 2010. Serum and urinary cell-free

MiR-146a and MiR-155 in patients with systemic lupuserythematosus. J Rheumatol 37:2516–2522.

Wang H, Peng W, Ouyang X, Li W, Dai Y. 2012.Circulating microRNAs as candidate biomarkers inpatients with systemic lupus erythematosus. Transl Res160:198–206.

Wang YZ, Tian FF, Yan M, Zhang JM, Liu Q, Lu JY, ZhouWB, et al. 2014. Delivery of an miR155 inhibitor by anti-CD20 single-chain antibody into B cells reduces theacetylcholine receptor-specific autoantibodies and amelio-rates experimental autoimmune myasthenia gravis. ClinExp Immunol 176:207–221.

Xiong W, Lahita RG. 2011. Novel treatments for systemiclupus erythematosus. Ther Adv Musculoskelet Dis3:255–266.

Zhang J, Roschke V, Baker KP, Wang Z, Alarc�on GS,Fessler BJ, Bastian H, et al. 2001. Cutting edge: A rolefor B lymphocyte stimulator in systemic lupus erythema-tosus. J Immunol 166:6–10.

60 H. R. ABOELENEIN ET AL.