Contributions of PD-1/PD-L1 pathway to interactions of myeloid DCs withT cells in atherosclerosis
Jun Lee a, Yan Zhuang b, Xin Wei b, Fujun Shang a, Jiuping Wang b, Ye Zhang b, Xiongtao Liu a, Yuhui Yang a,Li Liu a, Qiangsun Zheng a,⁎a Department of Cardiology, Tangdu Hospital, Fourth Military Medical University, Baqiao District, Xi'an 710038, PR Chinab Department of Infectious Diseases, Tangdu Hospital, Fourth Military Medical University, Baqiao District, Xi'an 710038, PR China
ls contribute to immunopathogenesis of atherosclerosis, underlying molecularmechanisms remain largely undefined. Recently, it has been demonstrated in mouse model that Programmeddeath-1 (PD-1)/PD-1 ligand (PD-L) pathway plays a critical role in proatherogenic immune responses. Herewe examined the expression of PD-1 and PD-L1 on peripheral blood mononuclear cells by flow cytometry in76 patients with coronary artery disease (CAD), and 25 healthy volunteers. The expression of PD-1 and PD-L1is significantly down-regulated on T cells and myeloid dendritic cells (mDCs) in CAD patients than in healthyindividuals, respectively. More importantly, we found that decreased PD-L1 expression on mDCs is relatedwith the increased T cell immune responses in CAD patients. In addition, stimulation of PD-L1 expression invitro could attenuate the stimulatory ability on allogeneic T cell proliferation and its cytokine production,including IFN-γ and IL-2, and also influence the production of IL-10 and IL-12 by mDCs. Taken together, wecan draw a conclusion that PD-1/PD-L1 pathway plays a key role in the regulation of proatherogenic T cellimmunity by intervening antigen presenting cell (APC)-dependent T cell activation, which associates withpro-inflammatory or anti-inflammatory cytokine production, and further studies need gain insight into thatthis pathway represents a strategy of immunotherapy for atherosclerosis.
It is well established that inflammation is an important drivingforce in every phase of atherosclerosis. Immunohistochemicalanalyses have shown that atherosclerotic plaques contain significantnumbers of T cells, including CD4+ and CD8+ subsets, at all ages oflesion development, in mouse models and humans [1,2]. However,there is no sufficient knowledge to elucidate the mechanism ofimmunomodulation in this progression. Programmed death-1 (PD-1,CD279) is an immune inhibitory receptor belonging to the B7-CD28family of costimulatory molecules, which is expressed on CD4+ T cells,CD8+ T cells, NK cells, B cells, andmonocytes upon activation. PD-1 hastwo known ligands, PD-L1 (B7–H1, CD274) and PD-L2 (B7–DC, CD273)[3,4]. PD-L1 is expressed on immunocompetent cells such as T cells, Bcells, dendritic cells (DCs), and macrophages [5]. In contrast, PD-L2expression is highly restricted to DCs and activatedmacrophages [6,7].Accumulating evidence suggests that the PD-1/PD-L pathway plays acritical role in attenuating T cell responses and promoting T celltolerance. PD-1/PD-L pathway is currently under intense investigationin that it has the potential tomodulate immune responses in a positiveor negative manner. Using mouse model, Gotsman and colleagues [8]
: +86 29 84777422.g).
l rights reserved.
have recently shown that PD-1/PD-L expression may correlate withthe extent of systemic CD4+ T cell response to hypercholesterolemiaand associated progression of arterial disease. However, it remainsunclear in human body whether PD-1/PD-L pathway influencesimmune response to plaque antigens and thereby impacts on theprocess of atherosclerosis.
Among the immune cells in atherosclerotic plaques, DCs may beparticularly important in the initiation of immune responses to plaqueantigens [9]. DCs represent the most potent inducers of T cellactivation due to their high expression levels of MHC and an arrayof accessory molecules involved in cell migration, adhesion and co-signaling. Myeloid DCs (mDCs), a subpopulation of DCs, play a pivotalrole in adaptive immunity by activating naive Tcells [10]. As amemberof co-signaling molecules, PD-L1 on mDCs may be involved in theimmune response during the process of atherosclerosis.
In this study, we examined the expression profiles of PD-1 on T cellsand PD-L1 on mDCs from patients with coronary artery disease (CAD),and evaluated the roles of PD-1/PD-L1 in the pathogenesis of athero-sclerosis. We found that PD-1/PD-L1 expressionwas down-regulated onTcells ormDCs, and this down-regulation of PD-1/PD-L1 enhanced CD4+
or CD8+ T cell proliferation and pro-inflammatory cytokine production.Additionally, reversed expression of PD-L1 onmDCs by IFN-α treatmentcould alter the inflammatory state through suppressing T cell prolifera-tion and changing the pattern of cytokine secretion.
170 J. Lee et al. / Journal of Molecular and Cellular Cardiology 46 (2009) 169–176
2. Materials and methods
2.1. Patients
A total of 76 patients who had coronary atherosclerosis, defined as aN50% diameter stenosis in ≥1 coronary artery, were enrolled into ourresearch at the Department of Cardiology of Tangdu Hospital, Xi'an, PRChina.Allwere above 40years of age andofHannationality. Baseduponclinical histories, physical examination, electrocardiogram, and cor-onary angiography, the CAD patients were clinically divided into twogroups. Group I comprised 32 unrelated patients who were diagnosedwith stable angina pectoris (SA), and group II comprised 44 caseswith adiagnosis of acute coronary syndrome (ACS). SA was defined as effortangina of at least 6 month duration, without any change in symptomfrequency or severity, and angiographically documented obstructivecoronary artery disease. ACS was defined as Braunwald Class IIIBunstable angina and myocardial infarctions. Twenty-five healthyvolunteers, age and sex matched to the enrolled patients, served ascontrols. All of these controls had normal electrocardiography and hadno evidence of atherosclerosis after receiving carotid sonography.
People with evidence of infectious disease, immunological dis-orders, fever, anti-inflammatory treatment, recent major surgery orneoplastic disease were excluded from the study. All subjects gavewritten informed consent, and the study was approved by the EthicalCommittee of the Fourth Military Medical University.
Blood sample from each patient was drawn by clean venepunctureof an antecubital vein after an overnight fast. In the case of an ACS thesampling was performed immediately at admission, before anyintervention being taken. Concentrations of serum cholesterol, C-reactive protein (CRP) and white blood cell counting were measuredaccording to routine protocols. PBMCs were isolated by Ficoll–Hypaque (Sigma Aldrich, St. Louis, MO) density gradient centrifuga-tion from fresh heparinized blood samples, and were viably frozen in80% fetal calf serum (FCS; Gibco, Grand Island, NY), 10% dimethylsulfoxide (DMSO), and 10% RPMI 1640 medium (Gibco) in liquidnitrogen for subsequent analysis. The plasma was collected forinflammatory cytokine detection.
2.3. Flow cytometry
The following monoclonal antibodies were used for the flowcytometric analysis of mDCs and lymphocytes: lineage-1-FITC, HLA-DR-PerCP (isotype, mouse IgG2a, κ), CD11c-APC (isotype, mouse IgG1,κ), PD-L1-PE (isotype, mouse IgG1, κ), CD4-PerCP (isotype, mouseIgG1, κ), CD8-FITC (isotype, mouse IgG2a κ), and PD-1-APC (isotype,mouse IgG1, κ). All Abs were purchased from BD Biosciences (San Jose,NJ). PD-1 expression on T cells and PD-L1 expression on mDCs weremeasured respectively using fresh heparinized peripheral blood. Inbrief, blood samples were incubated with a mixture of Abs (lineage-1-FITC, HLA-DR-PerCP, CD11c-allophycocyanin, and PD-L1-PE for mDCs;CD4-PerCP, CD8-FITC, and PD-1-APC for T cells), as well as isotypecontrol antibodies, for 20 min at 4 °C. The cells were then lysed withFACS lysing solution (BD Pharmingen, San Diego, CA) to remove RBC,fixed, and analyzed by flow cytometry on a FACSCalibur (BDBiosciences). The results were processed with CELLQuest software(BD Biosciences) and described with the percentage of positive cells.
2.4. Preparation of cells
MDCs were purified from PBMCs by CD19-negative selection,followed by CD1c-positive selection using the CD1c (BDCA-1)Dendritic Cell Isolation Kit (Miltenyi Biotech, Bergisch-Gladbach,
Germany) with a MiniMACS separator unit. CD4+ and CD8+ T cellswere also isolated from PBMCs by positive selection using the CD4 andCD8 Positive Isolation Kit (Dynal Biotech, Brown Deer, WI). Allprocedures were performed according to the manufacturer's instruc-tions. The median purity of isolated mDCs, CD4+, or CD8+ T cells wasroutinely N90%. Unless otherwise mentioned, freshly isolated cellswere cultured in complete RPMI 1640 medium containing 10% FCS,100 U/mL penicillin, and 100 μg/mL streptomycin.
2.5. Allogeneic mixed leukocyte reactions (MLR)
Maturation of purified mDCs were induced by addition of poly (I:C)[11] (50 μg/mL; Sigma-Aldrich), performed in 96-well flat-bottomculture plates at a concentration of 1×105 cells/200 μl. To up-regulatethe expression of PD-L1 on mDCs, IFN-α (10,000 U/mL) was added inthe culture medium in some cases. Twenty-four hours later, thesupernatant was removed and mDCs were mixed at different ratios(1:5,1:10,1:20,1:40, and1:80)with freshly isolatedCD4+ or CD8+ Tcells(1×105 cells/200 μl) from a third healthy individual for 5 days. Blockingantibodies against PD-L1 (eBiosciences, San Diego, CA) or isotypecontrol antibodies (mouse IgG1, κ; eBiosciences)were added to some ofthe cultures at a concentration of 2 μg/mL. Lymphocytes alone ormDCsalone were used as negative controls. Cells were pulsed with 1 μCi [3H]thymidine (Amersham Biosciences, Uppsala, Sweden) per well for 18 hbefore harvest. Before measurement of incorporated [3H]thymidine,part of the culture supernatants was collected for cytokine detection.
2.6. ELISA for cytokines
Plasma of studied population and culture supernatants werefrozen and kept at −80 °C before analysis. IFN-γ, IL-2, IL-10, and IL-12production were measured using ELISA kits (eBiosciences), inaccordance with the manufacturer's instructions. The sensitivities ofthese ELISA kits were 4 pg/mL for IFN-γ, IL-12 p70, and IL-10, and 2 pg/mL for IL-2. Absorbance was measured on an automatic plate reader.
2.7. Statistical analysis
All experimentswere repeated three times. Values are expressed asmean±SD, unless otherwise stated. Statistical analysis was performedusing GraphPad Prism version 5.00 for Windows (GraphPad Software,San Diego, CA). Continuous clinical variables between groups werecompared using the two-sample Student's t test. The percentages ofPD-1-positive or PD-L1-positive cells between groups were comparedusing the Mann–Whitney test. The distribution of plasma cytokinelevels was skewed, and comparisons were performed by the nonpara-metric Wilcoxon Mann–Whitney U test. All p values were two-tailedand values b0.05 were considered statistically significant.
3. Results
3.1. Characteristics of the study individuals
The characteristics of individuals in the present study are shown inTable 1. There were no significant differences in age, gender, bodymass index, lipid, diabetes, hypertension, or cigarette smoking inpatient groups (all pN0.05). Although people who underwent anti-inflammatory treatment were excluded from our study, some othermedicines, such as lipid-lowing agent statins [12] and anti-plateletagent aspirin [13], have anti-inflammatory properties that go beyondtheir major effects. There is no significant difference in medicationincluding statins and aspirin between patient groups. As expected,CRP levels and white blood cell counts were higher in patientssuffering from ACS comparedwith thosewith SA and healthy controls,respectively (pb0.01). Patients with ACS had no greater extent of CADor higher degree of diameter stenosis than those with SA.
Table 1Baseline characteristics of the studied population
Data are presented as mean±standard deviation, median (interquartile range) or number. ⁎pb0.01 ACS group vs SA group and control.LDL= low-density lipoprotein; CRP=C-reactive protein; P=probability of significance (difference among three groups).
171J. Lee et al. / Journal of Molecular and Cellular Cardiology 46 (2009) 169–176
3.2. Reduced expression of PD-1 on T lymphocytes and PD-L1 on mDCsfrom CAD patients
To explore whether PD-1 and PD-L1 were involved in T–mDCsinteractions in the development of atherosclerosis, we first examinedthe frequency of PD-1-expressing T-cell subsets and that of PD-L1-expressing mDCs from CAD individuals (n=76), and compared these
Fig. 1. Expression of PD-1 on T cells and PD-L1 onmDCs from CAD patients. Fresh heparinized(A, C and E) Representative dot plots of PD-1 staining or PD-L1 staining for studied individuexpression was determined by FACS gated on Lin1−HLA-DR+CD11c+ cells. (B, D and F) Percenshown, respectively. PD-1 expression on CD4+ or CD8+ T cells and PD-L1 expression on mDCshealthy control; ▵, SA patient; ◊, ACS patient. Horizontal bars, mean values. Values of p are
with cells from healthy controls (n=25). To this end, whole bloodsamples from CAD patients and healthy controls were incubated witha mixture of Abs and the expression of PD-1 and PD-L1 weredetermined by flow cytometric analysis. Representative dot blots ofPD-1+ or PD-L1+ distinct subpopulations were shown in Figs. 1A, C andE. As shown in Fig. 1B, PD-1+ distinct subpopulations of CD4+ T cellswere significantly decreased in peripheral blood cells from patients
peripheral blood were stained with a mixture of mAbs and analyzed by flow cytometry.als. PD-1 expression was determined by FACS gated on CD4+ or CD8+ cells, and PD-L1tages of PD-1+CD4+ T cells, PD-1+CD8+ T cells, and PD-L1+Lin1−HLA-DR+CD11c+ cells areare significantly down-regulated in CAD patients, as compared with healthy controls.○,shown.
172 J. Lee et al. / Journal of Molecular and Cellular Cardiology 46 (2009) 169–176
with SA (7.62±2.31) and ACS(7.90±2.44) compared with healthyindividuals (15.91±4.00; all pb0.001). CD8+-gated PD-1+ T cells werealso decreased in both SA (12.63±3.95) and ACS (12.53±3.46)compared with controls (Fig. 1D; 25.72±5.70; all pb0.001). Thefrequency of PD-L1+ mDCs (Lin-1−HLA-DR+CD11c+-gated PD-L1+ cells)from CAD individuals were significantly decreased in both SA and ACScompared with healthy controls (Fig. 1F; SA, 0.31±0.14; ACS, 0.32±0.13; healthy controls, 0.76±0.28; all pb0.001). The decreasedexpression of both PD-1 and PD-L1 on peripheral blood cells fromCAD patients suggested that the PD-1/PD-L1 interaction mightparticipate in the pathogenesis or regulation of atherosclerosis.
3.3. Effect of PD-L1 deficiency on mDCs activation of T cells
After confirming the decreased expression of PD-1 on T cells andPD-L1 on mDCs, we studied whether decreased PD-L1 expression onmDCs from patients with atherosclerosis can induce overreaction of Tcells. To do so, in vitro allogeneic MLR was performed. Mature mDCsfrom healthy subjects and CAD patients were cocultured at differentratios with T cells from a third healthy individual. We found that theT cell-stimulating capacity of mDCs was enhanced in CAD patients.Fig. 2 showed that mDCs isolated from CAD patients were more
Fig. 2. Effects of PD-1/PD-L1 pathway on T cell proliferation costimulated by mDCs. mDCs isoCD8+ T cells from a normal individual (as the third party), with or without treatment of IFN-cultures with [3H]thymidine for the last 18 h. Results are expressed as mean cpm±SD of trtreated mDCs from patient groups;■, addition of anti-PD-L1 mAbs; and□, addition of isotypenhanced the proliferation of CD4+ or CD8+ T cells, at all ratios compared with healthy individallostimulatory capacity upon T cells in comparison with healthy controls. The pretreatmsignificantly reversed their capacity to stimulate allogeneic CD4+ or CD8+ T cells, and the lev(all pN0.05). The blockade of PD-L1 signals significantly raised the stimulatory capacity of IFcontrol antibodies (all pb0.05).
efficient in inducing both CD4+ T cell (Figs. 2A and B) and CD8+ T cell(Figs. 2C and D) proliferation at all ratios, compared with mDCsisolated from controls (pb0.05 for both).
3.4. In vitro PD-L1 up-regulation by IFN-α attenuates the allostimulatorycapacity of mDCs from CAD patients
It has been demonstrated that both IFN-α and IFN-γ can stimulatemDCs to express PD-L1 [14]. Therefore, to explore the role of PD-1/PD-L1 pathway in T cell immune response during atherosclerosis, wecompared the effects of mDC-associated PD-L1 regulation onallogeneic T cell proliferation with or without the stimulation ofIFN-α. As expected, after mDCs being treated with IFN-α for 24 h, theexpression of PD-L1 on mDCs from CAD patients was significantly up-regulated (data not shown). The augmented stimulatory capacity ofmDCs frompatients can beweakened, leading to results similar to thatof normal mDCs (Fig. 2, all pN0.05). We further examined the effectsof IFN-α-associated mDCs' PD-L1 up-regulation on allogeneic MLR inthe presence of anti-PD-L1. It was found that weakened capacity ofmDCs to stimulate Tcell proliferationwas reversed by addition of anti-PD-L1mAbs, similar to that of untreatedmDCs, but higher than that oftreatedmDCswith isotype control antibodies. These data indicate that
lated from patients or healthy controls were cocultured at different ratios with CD4+ orα. After 5 days at 37 °C, proliferation of alloreactive T cells was assessed by pulsing theiplicate wells. ○ indicates healthy control; ▵, SA patient; ◊, ACS patient; ▴ or O, IFN-α-e control IgG1. (A and C) Peripheral mDCs isolated from SA patients (n=10) significantlyuals (n=6). (B and D) mDCs of ACS patients (n=10) also showed a significantly increasedent of IFN-α up-regulated the expression of PD-L1 on mDCs (data not shown) andel of T cell proliferation was not statistically significant compared with those of controlsN-α-treated mDCs on T cell proliferation at all ratios compared with addition of isotype
173J. Lee et al. / Journal of Molecular and Cellular Cardiology 46 (2009) 169–176
PD-L1 down-regulation on mDCs appears to be responsible, at least inpart, for the increased proliferation of CD4+ and CD8+ T cells in CADpatients, and the up-regulation of PD-L1 can reduce the allostimula-tory capacity of matured mDCs.
3.5. Plasma levels of cytokines
Cytokine production is an important criterion for evaluating T cellfunction. Here we detected the plasma levels of pro- and anti-inflammatory cytokines in CAD patients to find out evidence ofsystemic immune activation (Fig. 3). The concentration of IFN-γ in SApatients and ACS patients was increased (57.22±19.33 pg/mL,pb0.001; 120.4±45.05 pg/mL, pb0.001), compared to those of con-trols (32.75±12.38 pg/mL). The level of IL-2 was significantly higher inpatient groups (SA, 47.06±14.25 pg/mL, p=0.009; ACS, 78.76±17.00 pg/mL, pb0.001) than those of controls (32.64±10.98 pg/mL).Differences in the level of IFN-γ and IL-2 between two patient groupswere also significant (all pb0.001). The level of IL-10 had significantdifference among the groups (43.13±11.01 pg/mL in control group,35.34±9.85 pg/mL in SA group, and 13.55±5.17 pg/mL in ACS group;all pb0.05). Mean level of IL-12 was 106.5±37.43 pg/mL in SA groupand 136.7±31.51 pg/mL in ACS group, higher than that of controls(75.26±30.41 pg/mL, all pb0.01).
3.6. PD-L1 expressed on mDCs is associated with cytokine production byT cells and mDCs themselves
To determinewhether the expression of PD-L1 influences cytokines'production, we further measured the level of cytokines in MLR culturesupernatants. Using ELISA detection, we found that the mDCs of CADpatients were more efficient in inducing both CD4+ and CD8+ T cells tosecrete IFN-γ (Fig. 4) and IL-2 (Fig. 5) than those of healthy controls(pb0.05). Moreover, we found that following the up-regulation of PD-L1, the production of both IFN-γ and IL-2 by T cells was reduced. Nosignificant differences were found between the two patient groups. Asone of the most important anti-inflammatory cytokines, IL-10 secreted
Fig. 3. Plasma concentrations of pro-inflammatory and anti-inflammatory cytokines in patien3 groups were presented. The concentrations of cytokines were shown as a box with the 25percentiles. In SA group and ACS group, there was an enhanced level of pro-inflammatory cysuch as IL-10, compared with those in the control group. Values of p are shown.
by mDCs of CAD patients was significantly lower than that of healthycontrols (pb0.05; Fig. 6). On the contrary, the level of IL-12 in MLRculture supernatants from patient groups was significantly higher thanthat of normal donors (pb0.05; Fig. 7). IL-10 and IL-12 production canalso be reversed after IFN-α treatment on mDCs. There were nosignificant differences between SA group and ACS group (pb0.05 forall). We also examined the levels of these cytokines in the presence ofanti-PD-L1 mAbs during MLR. We found that changes of thesecytokines' secretion after IFN-α treatment can be intercepted byaddition of anti-PD-L1mAbs. Basing upon these results, wemay drawaconclusion that inhibitory signals linked with PD-L1 down-regulationon mDCs seems to be enhanced in the process of atherosclerosis,resulting in an increase of T cell proliferation and cytokine production,while these inhibitory signalsmay bemostly restored aftermDCs beingtreatedwith IFN-α, further demonstrating that down-regulation of PD-L1 may be responsible for enhanced T cell responses.
4. Discussion
Atherosclerosis is a major global cause of morbidity and mortality.Systemic inflammation plays an important role in its development.Found in arterial lesions, inflammatory cells are believed to be pivotalin a variety of processes, such as atherosclerotic plaque progression,plaque disruption and thrombosis [15]. Understanding the mechan-isms of inflammation in the process of atherosclerosis will have aprofound influence on the establishment of immune therapeuticregimens to inhibit activation of immune system and thereby improveor terminate atherosclerosis.
It has been well established that in addition to the first signal (thebinding between MHC/peptide complex and T cell receptor), costi-mulatory or coinhibitory molecules provide the second signal which isessential to influence the magnitude of T cell activation or T celltolerance [4,16]. Recently, several novel molecules have beendiscovered, including new B7-CD28 family members PD-1 and itsligands. The activation of naive T cells is considered as a sum ofpositive and negative signals conducted by the relative expression of
ts with coronary artery disease. Levels of IFN-γ (A), IL-2 (B), IL-10 (C) and IL-12 (D) in theth to 75th percentiles containing the median line and the lines showing the 1st to 99thtokines such as IFN-γ, IL-2 and IL-12, and reduced level of anti-inflammatory cytokines
Fig. 4. Effect of IFN-α-induced PD-L1 up-regulation on IFN-γ production by T cells inMLR. IFN-α-treated mDCs from all 3 groups were cocultured at different ratios with Tcells from a third healthy control. IFN-γ concentrations in supernatants obtained after5 days of culture were evaluated by ELISA. Data are expressed in pg/mL. There was asignificantly increased capacity of mDCs from patients (SA group, n=10; ACS group,n=10) to stimulate allogeneic CD4+ (A) and CD8+ (B) T cells to produce IFN-γ at theratio of 1:10, compared with controls (n=6). Up-regulation of PD-L1 by IFN-α couldreverse the IFN-γ production by T cells in MLR. Blocking PD-L1 by monoclonalantibodies could result in increased secretion of IFN-γ compared with IFN-α-treatedgroup and addition of isotype Ig group, with no statistically difference in comparisonwith original patient groups, respectively. α-mDC, IFN-α-treated mDCs. ○, healthycontrol;▵, SA patient; ◊, ACS patient;▴ or O, IFN-α-treatedmDCs from SA or ACS group;■, addition of anti-PD-L1 mAbs; and □, addition of isotype control IgG1. Horizontalbars, median values. Values of p are shown.
Fig. 5. Effect of IFN-α-induced PD-L1 up-regulation on IL-2 production by Tcells inMLR.IFN-α-treated mDCs from all 3 groups were cocultured at different ratios with T cellsfrom a third healthy control. IL-2 concentrations in supernatants obtained after 5 daysof culture were evaluated by ELISA. Data are expressed in pg/mL. There was asignificantly increased capacity of mDCs from patients (SA group, n=10; ACS group,n=10) to stimulate allogeneic CD4+ (A) and CD8+ (B) T cells to produce IL-2 at theratio of 1:10, compared with controls (n=6). Up-regulation of PD-L1 by IFN-α canreverse the IL-2 production by T cells in MLR. Blockade of PD-L1 signals by monoclonalantibodies could result in increased secretion of IL-2 compared with IFN-α-treatedgroup and addition of isotype Ig group, with no statistically difference in comparisonwith original patient groups, respectively. α-mDC, IFN-α-treated mDCs. ○, healthycontrol;▵, SA patient; ◊, ACS patient;▴ or O, IFN-α-treatedmDCs from SA or ACS group;■, addition of anti-PD-L1 mAbs; and □, addition of isotype control IgG1. Horizontalbars, median values. Values of p are shown.
174 J. Lee et al. / Journal of Molecular and Cellular Cardiology 46 (2009) 169–176
receptors and ligands on T cells and antigen presenting cells (APCs). Inthis setting, those possible causes of atherosclerosis may exploit thecosignaling balance to induce activated DC to provide sufficientpositive signals and down-regulate negative signals, resulting in alarge scale of T cell activation. It has been demonstrated that PD-1 andPD-Ls have a profound influence on T cell responses and athero-sclerosis in hypercholesterolemic mice [8], indicating that PD-1/PD-Lspathway may regulate the systemic immune and the resulting arterialpathology. Our study further demonstrated that PD-1/PD-L1 mole-cules expressed on PBMCs from CAD patients were down-regulatedand the attenuation of PD-1/PD-L1 was related to the hyperrespon-siveness of T cells, indicating that PD-1/PD-L1 pathway participated inthe immunopathogenesis of atherosclerosis.
PD-1 has been shown to be involved in the negative regulation ofsome immune responses and to take great efforts in the regulation ofperipheral tolerance [17]. Its two ligands, PD-L1 and PD-L2, might
bidirectionally regulate DC-T cell interactions [6]. PD-L1 is found onmany cell types, while the expression of PD-L2 is limited [18].Therefore, we focus on the effects of PD-L1. PD-1 engagement by PD-L1 has been shown to have potent inhibitory effects on immunestimulation [6,19], and this pathway is widely involved in immunor-egulatory processes, e.g. tumor escape mechanisms, inflammatoryresponses, organ-specific autoimmunity, and transplantation [4,20–22]. These results suggest that PD-1/PD-L1 pathway acts as animportant negative regulator in immune system. The present studyprovided evidence that down-regulation of PD-1 expression onfunctional T cells existed in CAD patients. Decreased PD-L1 expressionwas also found on mDCs. The reduction of PD-1/PD-L1 expressionwasrelatedwith enhanced Tcell proliferation. BecausemDCs play a pivotalrole in Ag presentation and the regulation of effector T cells, functionaldeficiencies in the mDC-T cell interaction may be responsible for theprogression of atherosclerosis. Our results showed that mDCs from
Fig. 6. Effect of IFN-α-induced PD-L1 up-regulation on IL-10 production by mDCs inMLR. IFN-α-treated mDCs from all 3 groups were cocultured at different ratios withCD4+ (A) or CD8+ (B) T cells from a third healthy control. IL-10 concentrations insupernatants obtained after 5 days of culture were evaluated by ELISA. Data areexpressed in pg/mL. There was a significantly decreased capacity of mDCs from patients(SA group, n=10; ACS group, n=10) to produce IL-10 compared with controls (n=6). Up-regulation of PD-L1 by IFN-α can reverse the IL-10 production in MLR. Blockade of PD-L1 signals by monoclonal antibodies could result in decreased secretion of IL-10compared with IFN-α-treated group and addition of isotype Ig group, with no statis-tically difference in comparison with original patient groups, respectively. α-mDC, IFN-α-treated mDCs. ○, healthy control;▵, SA patient; ◊, ACS patient;▴ or O, IFN-α-treatedmDCs from SA or ACS group;■, addition of anti-PD-L1 mAbs; and□, addition of isotypecontrol IgG1. Horizontal bars, median values. Values of p are shown.
Fig. 7. Effect of IFN-α-induced PD-L1 up-regulation on IL-12 production by mDCs inMLR. IFN-α-treated mDCs from all 3 groups were cocultured at different ratios withCD4+ (A) or CD8+ (B) T cells from a third healthy control. IL-12 concentrations insupernatants obtained after 5 days of culture were evaluated by ELISA. Data areexpressed in pg/mL. There was a significantly increased capacity of mDCs from patients(SA group, n=10; ACS group, n=10) to produce IL-12 compared with controls (n=6). Up-regulation of PD-L1 by IFN-α can reverse the IL-12 production in MLR. And blocking PD-L1 signals by monoclonal antibodies could result in increased secretion of IL-12compared with IFN-α-treated group and addition of isotype Ig group, with nostatistically difference in comparisonwith original patient groups, respectively.α-mDC,IFN-α-treated mDCs. ○, healthy control; ▵ SA patient; ◊, ACS patient; ▴ or O, IFN-α-treated mDCs from SA or ACS group;■, addition of anti-PD-L1 mAbs; and□, addition ofisotype control IgG1. Horizontal bars, median values. Values of p are shown.
175J. Lee et al. / Journal of Molecular and Cellular Cardiology 46 (2009) 169–176
CAD patients were more powerful in stimulating allogeneic T cellproliferation than mDCs from healthy individuals, suggesting that anintrinsic defect of these mDCs may be responsible for enhancedstimulatory capacity.
Cytokine is a possible link between systemic inflammation andlocal inflammatory processes. Ligation of PD-1 receptor by PD-L1 cannot only inhibit T cell proliferation, but also influence cytokineproduction by activated T cells or mDCs [23]. Here we examined thecytokine levels by ELISA in plasma of CAD patients and culturesupernatants during allogeneic MLR. In comparison with healthysubjects, plasma levels of IFN-γ and IL-2 were significantly higher inpatients with SA and ACS, and increased IL-12 and decreased IL-10secretion were also detected in two patient groups. Similar resultsabout these cytokines were found in the coculture supernatantsduring allogeneic MLR. In coronary atherosclerosis, these moleculesare of crucial importance in promoting or suppressing cellular andhumoral responses to endothelial toxins [24,25]. The toxins may
activate monocytes and a T helper type 1 (Th1) response follows,characterized by IL-12 production, down-regulation of IL-10, T cellrecruitment, and subsequently up-regulation of IL-2 and IFN-γ [25].IL-10 is one of the predominated anti-inflammatory cytokines andplays a protective role in CAD [26]. On the contrary, IL-12 is asignificant inducer of the Th1 cell response in animal models ofatherosclerosis [25,26]. Our data suggest that PD-1/PD-L1 pathwaymay modulate the secretion of pro-inflammatory or anti-inflamma-tory cytokines and then influence the process of atherosclerosis.
It has been demonstrated that activation of Th1 responses isinvolved in lesion rupture [27,28]. The difference of CRP levels, whiteblood cell counts, and cytokine levels between SA group and ACSgroup could be explained by the acute inflammatory state of ACScharacterized by Th1 cell activation. However, the difference of PD-1or PD-L1 expression on T cells or mDCs between patient groups is notsignificant, indicating that PD-1/PD-L1 pathwaymight not be involvedin the development of plaques instability.
176 J. Lee et al. / Journal of Molecular and Cellular Cardiology 46 (2009) 169–176
Little is known about the stimuli for PD-1 expression [6]. Theexpression of PD-L1 protein on hepatocytes, endothelial cells, andmDCs can be strongly up-regulated by the stimulation of IFN-α andIFN-γ [14,29,30]. In this study, we use IFN-α as stimulator to reversethe expression of PD-L1 on mDCs isolated from CAD patients. Thesuppressed responsewas also evident when analyzing type 1 cytokinesecretion by T cells cocultured with IFN-α-treated mDCs. This PD-L1recovery was also accompanied with IL-12 down-regulation and IL-10up-regulation in coculture supernatants. These findings furtherconfirmed the impact of PD-1/PD-L1 up-regulation on T-cell pro-liferation and APC function. IFN-γ displays similar characteristics withrespect to PD-L1 up-regulation as IFN-α, but the overall immuneeffects of these cytokines differ considerably. Although this cytokinedisplays anti-inflammatory properties under certain conditions [31],IFN-γ is preferably considered as a pro-inflammatory molecule.Considering the high level of IFN-γ in plasma of patients withatherosclerosis, we could assume that pro-inflammatory activity of Tcells might be modulated through a negative feedback loop. In thisway, activated Th1 cells producing high level of IFN-γ are forced toreduce their cytokine levels via IFN-γ-induced PD-L1 expression.Nevertheless, interferons represent components of a complex net-work and their effects depending on the disease state and the time ofapplication [32]. Under the circumstance of atherosclerosis, thisnegative feedback loop might not function any more, and inflamma-tion exerts a dominant effect in the process of this disease.
Medication like statins, aspirin, etc, has become routinely treat-ments under the circumstance of atherosclerosis. Besides their majoreffects like lowing lipid or anti-platelet, they can affect circulating celltranscriptomes and play an anti-inflammatory role in the course ofthis disease [12,13]. But the mechanism of inflammation is verycomplicated and the inflammatory state of patients in our study wasnot obviously interfered, as well as the results of our experiments.
In summary, our data demonstrated the down-regulation of PD-1on T cells and PD-L1 onmDCs in CAD patients. The down-regulation ofPD-1/PD-L1 pathway can improve, at least in part, T cell activation inthe development, progression, and complications of atherosclerosis.Furthermore, it also influences the secretion of pro- and anti-inflammatory cytokines, which contribute to pathogenic and regula-tory immunity in the context of atherosclerosis. Therefore, manip-ulating the expression of PD-1/PD-L1 molecules may represent a newtherapeutic option for atherosclerosis.
Acknowledgments
We thank all CAD individuals and healthy participants in this study.
Appendix A. Supplementary data
Supplementary data associated with this article can be found, inthe online version, at doi:10.1016/j.yjmcc.2008.10.028.
[4] Greenwald RJ, Freeman GJ, Sharpe AH. The B7 family revisited. Annu Rev Immunol2005;23:515–48.
[5] Dong H, Zhu G, Tamada K, Chen L. B7–H1, a third member of the B7 family,co-stimulates T-cell proliferation and interleukin-10 secretion. NatMed 1999Dec;5(12):1365–9.
[6] Keir ME, Francisco LM, Sharpe AH. PD-1 and its ligands in T-cell immunity. CurrOpin Immunol 2007 Jun;19(3):309–14.
[7] Latchman Y, Wood CR, Chernova T, Chaudhary D, Borde M, Chernova I, et al. PD-L2is a second ligand for PD-1 and inhibits T cell activation. Nat Immunol 2001 Mar;2(3):261–8.
[8] Gotsman I, Grabie N, Dacosta R, Sukhova G, Sharpe A, Lichtman AH. Proatherogenicimmune responses are regulated by the PD-1/PD-L pathway in mice. J Clin Invest2007 Oct;117(10):2974–82.
[10] Liu YJ. Dendritic cell subsets and lineages, and their functions in innate andadaptive immunity. Cell 2001 Aug 10;106(3):259–62.
[11] Ponsaerts P, Van den Bosch G, Cools N, Van Driessche A, Nijs G, Lenjou M, et al.Messenger RNA electroporation of human monocytes, followed by rapid in vitrodifferentiation, leads to highly stimulatory antigen-loaded mature dendritic cells.J Immunol 2002 Aug 15;169(4):1669–75.
[12] Patel TN, Shishehbor MH, Bhatt DL. A review of high-dose statin therapy: targetingcholesterol and inflammation in atherosclerosis. Eur Heart J 2007 Mar;28(6):664–72.
[13] Khan Q, Mehta JL. Relevance of platelet-independent effects of aspirin to itssalutary effect in atherosclerosis-related events. J Atheroscler Thromb 2005;12(4):185–90.
[14] Chen L, Zhang Z, Chen W, Zhang Z, Li Y, Shi M, et al. B7–H1 up-regulation onmyeloid dendritic cells significantly suppresses T cell immune function in patientswith chronic hepatitis B. J Immunol 2007 May 15;178(10):6634–41.
[15] Ross R. Atherosclerosis—an inflammatory disease. N Engl J Med 1999 Jan 14;340(2):115–26.
[16] Watts TH. TNF/TNFR family members in costimulation of T cell responses. AnnuRev Immunol 2005;23:23–68.
[17] Nishimura H, Honjo T. PD-1: an inhibitory immunoreceptor involved in peripheraltolerance. Trends Immunol 2001 May;22(5):265–8.
[18] Okazaki T, Iwai Y, Honjo T. New regulatory co-receptors: inducible co-stimulatorand PD-1. Curr Opin Immunol 2002 Dec;14(6):779–82.
[19] Grakoui A, John Wherry E, Hanson HL, Walker C, Ahmed R. Turning on the offswitch: regulation of anti-viral T cell responses in the liver by the PD-1/PD-L1pathway. J Hepatol 2006 Oct;45(4):468–72.
[20] Rini BI. Current status and future directions of molecular markers in renal cellcarcinoma. Curr Opin Urol 2006 Sep;16(5):332–6.
[21] Dong H, Chen X. Immunoregulatory role of B7–H1 in chronicity of inflammatoryresponses. Cell Mol Immunol 2006 Jun;3(3):179–87.
[22] Freeman GJ, Wherry EJ, Ahmed R, Sharpe AH. Reinvigorating exhausted HIV-specific T cells via PD-1-PD-1 ligand blockade. J Exp Med 2006 Oct 2;203(10):2223–7.
[23] Freeman GJ, Long AJ, Iwai Y, Bourque K, Chernova T, Nishimura H, et al.Engagement of the PD-1 immunoinhibitory receptor by a novel B7 family memberleads to negative regulation of lymphocyte activation. J Exp Med 2000 Oct 2;192(7):1027–34.
[24] Szodoray P, Timar O, Veres K, Der H, Szomjak E, Lakos G, et al. TH1/TH2 im-balance, measured by circulating and intracytoplasmic inflammatory cytokines–immunological alterations in acute coronary syndrome and stable coronary arterydisease. Scand J Immunol 2006 Sep;64(3):336–44.
[25] Fernandes JL, Mamoni RL, Orford JL, Garcia C, Selwyn AP, Coelho OR, et al.Increased Th1 activity in patients with coronary artery disease. Cytokine 2004May7;26(3):131–7.
[26] Tedgui A, Mallat Z. Cytokines in atherosclerosis: pathogenic and regulatorypathways. Physiol Rev 2006 Apr;86(2):515–81.
[27] Binder CJ, Chang MK, Shaw PX, Miller YI, Hartvigsen K, Dewan A, et al. Innate andacquired immunity in atherogenesis. Nat Med 2002 Nov;8(11):1218–26.
[28] Veillard NR, Steffens S, Burger F, Pelli G, Mach F. Differential expression patterns ofproinflammatory and antiinflammatory mediators during atherogenesis in mice.Arterioscler Thromb Vasc Biol 2004 Dec;24(12):2339–44.
[29] Muhlbauer M, FleckM, Schutz C,Weiss T, FrohM, Blank C, et al. PD-L1 is induced inhepatocytes by viral infection and by interferon-alpha and -gamma and mediatesT cell apoptosis. J Hepatol 2006 Oct;45(4):520–8.
[30] Mazanet MM, Hughes CC. B7–H1 is expressed by human endothelial cells andsuppresses T cell cytokine synthesis. J Immunol 2002 Oct 1;169(7):3581–8.
[31] Krakowski M, Owens T. Interferon-gamma confers resistance to experimentalallergic encephalomyelitis. Eur J Immunol 1996 Jul;26(7):1641–6.
[32] TownsendMJ, McKenzie AN. Unravelling the net? Cytokines and diseases. J Cell Sci2000 Oct;113(Pt 20):3549–50.
