Karrie K. Wong,*,† Ismat Khatri,† Suchinta Shaha,‡ David E. Spaner,‡,§ andReginald M. Gorczynski*,†,1
*Institute of Medical Science and †Transplant Research Division, University Health Network, University of Toronto, Toronto,Ontario, Canada; ‡Division of Molecular and Cellular Biology, Research Institute, Sunnybrook Health Sciences Centre,
Toronto, Ontario, Canada; and §Toronto-Sunnybrook Regional Cancer Centre, Toronto, Ontario, Canada
RECEIVED OCTOBER 20, 2009; REVISED APRIL 1, 2010; ACCEPTED APRIL 2, 2010. DOI: 10.1189/jlb.1009686
ABSTRACTCD200 is a transmembrane protein broadly expressedon a variety of cell types, which delivers immunoregula-tory signals through binding to receptors (CD200Rs)expressed on monocytes/myeloid cells and T lympho-cytes. Signals delivered through the CD200:CD200Raxis have been shown to play an important role in theregulation of anti-tumor immunity, and overexpressionof CD200 has been reported in a number of malignan-cies, including CLL, as well as on cancer stem cells. Weinvestigated the effect of CD200 blockade in vitro on ageneration of CTL responses against a poorly immuno-genic CD200� lymphoma cell line and fresh cells ob-tained from CLL patients using anti-CD200 mAb andCD200-specific siRNAs. Suppression of functional ex-pression of CD200 augmented killing of the CD200�
cells, as well as production of the inflammatory cyto-kines IFN-� and TNF-� by effector PBMCs. Killing wasmediated by CD8� cytotoxic T cells, and CD4� T cellsplay an important role in CD200-mediated suppressionof CTL responses. Our data suggest that CD200 block-ade may represent a novel approach to clinical treat-ment of CLL. J. Leukoc. Biol. 88: 000–000; 2010.
IntroductionThe differentiation and activation of B cells involve multipleprocesses that regulate gene rearrangement, proliferation, andapoptosis. When these are disrupted, malignancies often oc-cur, including lymphomas and CLL [1]. Complete cure ofboth diseases with conventional chemotherapy remains ex-tremely rare. Although T cell-mediated anti-tumor immuneresponses have the potential to eliminate tumor cells, CLL andlymphoma cells are inherently poorly immunogenic, renderingT cell-based immunotherapies ineffective [2]. Various tech-niques have been used to try to improve immunogenicity ofCLL cells, including the use of IL-2 and TLR agonists [3]. Im-
munoregulatory molecules are known to play critical roles inregulating T cell-mediated immunotherapy, and manipulationof immunoregulatory pathways may be an important alterna-tive method to improve the efficacy of such treatments.
One immunoregulatory molecule, CD200, has beenshown to be overexpressed in a number of malignancies,including renal carcinoma, colon carcinoma, ovarian carci-noma, melanoma, AML, multiple myeloma, and CLL [4 – 8].In AML, cell surface CD200 expression on malignant cells iscorrelated with poor prognosis [8]. CD200 has also beenreported recently to be a cancer stem cell marker [9]. Theregulatory function of CD200 is delivered through bindingto a receptor, CD200R, expressed on cells of the myeloidlineage and T lymphocytes [10].
A regulatory function for CD200 in tumor immunity wassuggested following studies that showed that infusion of a solu-ble form of CD200, CD200Fc, into EL4 thymoma-bearingC57B/6 mice enhanced tumor growth [11]. Anti-CD200 mAbhave been reported recently to abrogate growth of CD200-transduced RAJI and Namalwa cells in NOD-SCID mice [12].In addition, Pallasch et al. [13] demonstrated that CD200 ex-pression on CLL cells had inhibitory effects on the prolifera-tion of autologous effector T cells, and CD200 blockade, usinga rat anti-CD200 mAb, produced a reduction in the number ofCD25�CD4�forkhead box p3� regulatory T cells in vitro. Inthe case of CLL, no correlation has been reported betweenCD200 surface expression and other CLL prognostic markers,such as CD38 expression, Ig heavy-chain variable region muta-tional status, and Binet staging system [13], and indeed, theindependent prognostic value of CD200 expression remainsunknown.
In a model system that used a poorly immunogenic lym-phoma cell line with constitutive CD200 levels or CD200� pri-mary CLL cells, we show below that blockade of CD200 bymAb or down-regulation of CD200 by specific silencers aug-mented anti-tumor CTL responses in vitro. CD4� T cells from
1. Correspondence: Institute of Medical Science and Transplant ResearchDivision, University Health Network, University of Toronto, Rm. 2-805,MaRS Tower, 101 College St., Toronto, ON, Canada, M5G1L7. E-mail:[email protected]
Epub ahead of print May 4, 2010 - doi:10.1189/jlb.1009686
Copyright 2010 by The Society for Leukocyte Biology.
splenocytes of individual CLL patients expressed CD200R, con-sistent with the hypothesis that CD200 overexpression on tu-mor cells themselves may mediate immunosuppression in CLL.
Treatments of primary CLL cells with a TLR7 agonist, aloneor in combination with phorbol esters and IL-2, have been re-ported to enhance the immunogeneicity of CLL cells and in-crease their killing by effector T cells [3, 14]. We report belowthat this treatment also reduced CD200 expression signifi-cantly on CLL cells and imply that down-regulation of CD200expression on tumor cells may improve immunogenicity ofCLL and lymphoma cells and enhance the efficacy of cell-based immunotherapies.
MATERIALS AND METHODS
CellshPBMCs were isolated from heparin-treated whole blood of healthy volun-teer donors using Ficoll-Paque PLUS gradients (GE Healthcare Bio-Sci-ences, Piscataway, NJ, USA). Five independent volunteer donors were usedon multiple occasions throughout the studies described. PBMCs were usedin CTL assays immediately after isolation. Two human cell lines propagatedfrom non-Hodgkin’s lymphomas were grown in suspension in AIM-V me-dium (Invitrogen, Carlsbad, CA, USA) supplemented with 5% FBS (Hy-clone, Logan, UT, USA) [15]. CD5�CD19� primary CLL cells were puri-fied from the fresh blood of consenting CLL patients as described previ-ously [16]. CLL spleens were obtained after splenectomy. Single cellsuspensions from CLL spleens were obtained by standard protocols. Allprotocols were approved by Institutional Review Boards. hCD200-trans-fected HEK293 cells were obtained from Genetec (Quebec, Canada). Cellswere grown in selection medium DMEM-F12, supplemented with 1 ug/mlG418 and 10% FBS.
AntibodiesThe rat anti-hCD200 mAb 1B9 and 5A9 were described previously [17].3H4, which showed no immunoreactivity against cell surface CD200 (datanot shown), was used as an isotype control in CTL assays. A polyclonal rab-bit anti-hCD200 serum was generated following immunization of rabbitswith CD200Fc and subsequent boosting �2 with HEK293-hCD200 cell ly-sate (custom immunization performed by Cedarlane Labs, Hornby, ON,Canada). Anti-Fc antibodies in the sera were removed using Fc column ab-sorption (Cedarlane Labs), and immunoreactivity of the sera was con-firmed by Western blots.
FACS analysesLymphoma cells (5�105) were washed twice with 1 ml FACS buffer (PBS,1% FBS, 0.1% NaN3, 5 mM EDTA) and incubated with 0.1 ug rat anti-hCD200 mAb (1B9) or isotype control for 45 min at 4°C. Cells werewashed �3 with PBS and incubated with goat anti-rat IgG-FITC antibody(Jackson Labs, Western Grove, PA, USA) at 1:100 dilution for 30 min at4°C. For CD200R1 staining, a mouse anti-hCD200R mAb (R&D Systems,Minneapolis, MN, USA) was used at 0.5 ug/sample, followed by incubationwith a secondary goat anti-mouse IgG-PE antibody (Jackson Labs). The fol-lowing antibodies were used at concentrations suggested by the supplier(BD Biosciences, San Jose, CA, USA): CD4-FITC, CD8-PE-Cy7, CD5-PE-Cy5,and CD19-PE. In activation experiments, CLL cells were stained withmouse anti-CD83-PE and mouse anti-CD5-FITC antibodiy (BD Biosciences),as per the manufacturer’s instruction at 24 and 48 h after stimulation. Allcells were fixed with 1% paraformaldehyde before being analyzed in aCoulter FC500 flow cytometer.
Western blotsCells were lysed in 0.025% SDS and run on SDS-PAGE gels. After transfer,the blots were blocked with 5% milk in TBST overnight at 4°C. Blots were
then probed with the rabbit anti-hC200 sera at a 1:6000 dilution. In siRNAexperiments, MAPK was used as a housekeeping protein for reference. Gelswere divided at �40 kd, and the top part of the gel blot (�40 kd) wasprobed for CD200, and the bottom part (�40 kd) was probed with amouse anti-MAPK mAb, Tag-100 (Qiagen, Valencia, CA, USA), at 1:1000dilution. After thorough washing, blots were probed with goat anti-rabbitIgG-HRP (1:6000) or goat anti-mouse IgG-HRP (1:2000; Jackson Labs) for1 h at room temperature. Blots were developed using an ECL Western blotdetection kit (GE Healthcare Bio-Sciences).
CD200 column absorptionThe rat anti-hCD200 mAb 1B9 was conjugated to cyanogen bromide-acti-vated Sepharose beads (Cedarlane Labs). For CD200 absorption, 1 ml neatserum samples from healthy controls and CLL patients was incubated with250 �l 1B9-conjugated beads on a shaker overnight at 4°C. Preabsorbedserum and CD200-absorbed serum samples were subsequently used in MLCassays at designated dilutions.
RT-PCR and real-time PCRRNA was extracted from cells using TRIZOL reagent, and cDNA was ob-tained using OliogDT primers (Invitrogen). To detect CD200 mRNA level,the following primer pair was designed to detect �100 bp amplicons: for-ward, AATACCTTTGGTTTTGGGAAGATCT; reverse, GGTGGTCTTCA-GAGAATTTGTAGTGA. Primer mixes for GAPDH and TATA box-bindingprotein were purchased from Qiagen and used as housekeeping genes fornormalization of CD200 gene expression level. All primers were used inregular PCR and real-time PCR. For real-time PCR experiments, 50 ngcDNA was used per reaction.
siRNA transfectionThree commercial CD200 siRNA, designated CD200 siRNA #1, CD200siRNA #4, and CD200 siRNA #6, were obtained from Qiagen. Two controlsiRNAs, a positive control GAPDH silencer and a negative control, werepurchased from Ambion (Austin, TX, USA) for use in silencing experi-ments. Lymphoma cells (7.5�105) were transfected with 2 ug siRNA usinglipofectamine2000 (Invitrogen) as a transfection reagent at a 1:6 ratio.Transfection was performed in triplicate in 12-well plates according to themanufacturer’s instructions. Cells were harvested for RNA at 48 h and forprotein at 72 h after transfection. In some experiments, cells were used inCTL assays as stimulator cells 72 h after transfection.
CTL assaysPBLs (1.2�106) were stimulated with mitomycin C-treated Ly5 or Ly2 cellsat a 15:1 responder:stimulator ratio in 96-well plates. In some wells, 8 ug1B9 rat anti-hCD200 mAb was added for functional neutralization ofCD200 expression. Supernatants were harvested from each well 18 and42 h after stimulation to assay for cytokines. After 6 days, fresh lymphomacells were labeled overnight with [3H]TdR at 37°C and washed three timesin PBS, and 1 � 104 cells were added to each well in the 96-well plate. Theplate was harvested for [3H]TdR analysis at 18 h. All assays were performedin triplicate, and geometric means were used in quantitation of CTL activ-ity. Cytotoxic killing of lymphoma cells was calculated from the [3H]TdRremaining in cells with reference to unstimulated controls and the totalcounts added in the targets. All results shown were obtained from a mini-mum of three independent experiments. Where CD5�CD19� primary CLLcells were used as stimulators, 51Cr release assays were performed to assesskilling. At 7 days after stimulation, with unstimulated PBL cells set up asnegative controls, 51Cr-labeled CLL cells were added into each well askilling targets, and 51Cr release was assessed in supernatant at 6 h afteraddition of 51Cr-labeled CLL targets. CLL cells from three different pa-tients were used as targets for the same PBL effectors in three indepen-dent experiments. In experiments where CD4� or CD8� T cells weredepleted, depletion was performed using EasySep immunomagnetic cell
2 Journal of Leukocyte Biology Volume 88, August 2010 www.jleukbio.org
ELISASupernatant samples harvested from CTL assays were assayed for TNF-�,IFN-�, TGF-�, IL-4, IL-6, IL-10, and IL-12 using ELISA kits purchased fromeBioscience (San Diego, CA, USA), as per the manufacturer’s instruction.A standard curve was obtained in each assay to quantify cytokine present inthe supernatant.
Activation of CLL cellsPurified CLL cells (2�106) were cultured in serum-free AIM-V mediumplus 2-ME (Sigma-Aldrich, St. Louis, MO, USA) in 24-well plates at 37°C in5% CO2 in the presence or absence of the following immunomodulators:TLR-7 agonist Imiquimod (LKT Laboratories, St. Paul, MN, USA), PMA(Sigma-Aldrich), and hrIL-2. Imiquimod and PMA powders were reconsti-tuted in DMSO as 1 mg/ml and 10 mg/ml stock solutions. For activationof CLL cells, Imiquimod, PMA, and IL-2 were used at a final concentrationof 3 ug/ml, 30 ng/ml, and 500 U/ml, respectively. At 24 and 48 h after
stimulation, cells were harvested, and cell surface expressions of CD200,CD83, and CD5 were determined by FACS. Up-regulation of CD83 expres-sion was used as an indicator for response to stimulation.
Statistical analysesP values for all experimental data were obtained using the Student’s t-testto determine the significance between sample means.
RESULTS
Comparison of CD200 expression on primary CLLcells and two independent human lymphoma celllinesTo explore the effect of CD200 expression on induction ofanti-tumor immunity in vitro, we first characterized expressionof CD200 using a number of independent isolates of primaryCLL cells and established cell lines using PCR, Western blots,
TABLE 1. CLL Patient Characteristics of Individuals Used in Study
Patient # Sex AgeYears afterdiagnosis
RaiStagea WBCb Treatmentc % CD38 Cytogeneticsd
3 M 72 4 IV 175 S, CHOP, CVP 2 13q–/17p–4 F 84 10 III 200 PC, SR 46 13q–5 M 62 12 IV 121 CP, CVP, S, FC 2 13q–8 M 74 7 IV 220 FR, splenectomy, S na 17p–
12 M 55 6 IV 23 CHOP, P 81 13q-, 17p-14e M 71 6 II 186 none 72 11q–, 13q–16 M 69 6 IV 189 CP 4 normal22 M 54 6 IV 8 CP, FC, FCR, S 1 normal27 F 82 11 III 125 none 8 T1236 M 59 7 IV 10 CP 7 Na41 F 59 6 III 35 CP, splenectomy 40 normal46e M 55 5 III 121 none 4 T1257 F 77 10 IV 120 CP na 13q–60 M 88 6 IV 62 none 1 13q–, 17p–61f M 81 6 III 89 P 1 normal65 F 61 26 IV 140 splenectomy 18 13q–, 11q–70 F 78 5 III 33 none 13 13q–71 M 57 2 II 21 none 1 T12,13q–72 F 48 2 0 13 none 2 na73e F 53 13 IV 65 none 2 13q–74 M 77 10 III 250 none 1 13q–75f F 63 3 II 22 none 71 na78f M 72 5 IV 80 CP,FCR 13 normal79f M 51 10 IV 140 C�2 1 13q–80f M 51 1 II 88 None 2 13q–Ig M 72 5 IV 96 C�2, splenectomy 60 17p–
IIg M 58 15 IV 77 C, F, FC, S, Splen 65 13q–, 11q–
a Rai stage: 0, Lymphocytosis; I, with adenopathy; II, with hepatosplenomegaly; III, with anemia; IV, with thrombocytopenia.b WBC, White blood cell count (�106 cells/ml) in the peripherial blood.c CVP, Cyclophosphamide/Vincristine/Prednisone; CHOP, Cyclophosphamide/Oncovin/Prednisone/Doxorubicin; C, Cyclophosphamide;
P, Prednisone; F, fludarabine; R, rituximab; S, solumedrol; PC, Prednisone followed by Cyclophosphamide; SR, Solumedrol followed by Rituximab;CX2, 2 Courses of Cyclophosphamide treatment.
d T12, Trisomy 12; na, not available.e Cells from indicated patients were used as stimulators in CTL assays (see Fig 6A).f Cells from indicated patients were used in activation experiments (see Fig 7). Cells from all patients in this table were stained for CD200
(Fig. A), with the exception of Patients I and II (g).g Spleens were obtained from indicated patients after splenectomy.Corresponding splenocytes were analyzed for CD200R expression (see Fig 6C).
Wong et al. CD200 controls immunity to CLL
www.jleukbio.org Volume 88, August 2010 Journal of Leukocyte Biology 3
and FACS. Although essentially all primary tumor cells (n�25;Table 1) analyzed in this study stained positive for CD200, ascompared with the respective isotype controls, the MFI of thestaining varied widely (Fig. 1A). Heterogeneity in the CD200cell surface expression level was not correlated independentlywith any of the various clinical parameters of CLL analyzed,including Rai disease stage, CD38 expression, and cytogeneticstatus. To date, there are no data, to our knowledge, directlyaddressing the issue of altered CD200 cell surface expressionwith disease progression. In the two lymphoma cell lines stud-ied, Ly5 cells showed CD200 expression levels comparable withthat of primary CLL cells, and Ly2 cells failed to stain forCD200 (Fig. 1B). Results from FACS studies were confirmedusing Western blot analyses, in which CD200 was detected as aband at approximately 48 kd in lysates of hCD200-transfectedHEK293 cells and Ly5 cells, but not Ly2 cells (Fig. 1C). CD200expression on Ly5 cells, on the other hand, remained consis-tently high after prolonged in vitro passage (�20; data not
shown). Ly5 and Ly2 cells were used in the functional studiesdescribed below, designed to investigate the functional conse-quences of the presence of CD200 expression on tumor cell-induced immunity.
The effect of CD200 blockade in the killing ofCD200� Ly5 and CD200– Ly2 cellsDifferent epitopes of hCD200 are recognized by independentlyderived rat anti-hCD200 mAb [17]. Among this panel of ratanti-human mAb (all IgG2a), 1B9 and 5A9 recognized the ex-tracellular domain of CD200 (Fig. 1B), and another mAb,3H4, failed to stain any of the CD200� cells identified by 1B9(data not shown). We have used 3H4 as an isotype control inthe experiments discussed below.
We explored the effect of addition of 1B9, 5A9, or 3H4 invitro on MLCs using hPBL from healthy blood donors as effec-tor cells and mitomycin C-treated CD200� Ly5 and CD200–
Ly2 cells as stimulators. [3H]TdR-based cytotoxicity assays were
A250.0
Mean Fluorescent Intensity (MFI) of CD200 cell surface expression
Figure 1. Expression of CD200 on primary CLL cellsand two CLL cell lines. (A) MFI of CD200 expressionon a panel of primary CLL cells (n�25), two lym-phoma cell lines (Ly5 and Ly2 cells), and normal Bcells (Nor). The dotted, horizontal line designatesthe level of CD200 expression on normal B cells. (B)Constitutive cell surface expression of CD200 on Ly5cells and absence of expression on Ly2. The mAb1B9 and 5A9 stained CD200 equally well on Ly5 cells,and Ly2 cells showed no CD200 staining. Stainingwas performed using 0.1 ug 1B9, 5A9, or rat IgG iso-type control (shaded histogram). FL1, Fluorescence1. (C) Detectable CD200 levels in Ly5 cell lysate butnot in Ly2 lysate using a rabbit anti-hCD200 sera at1:6000 dilution.
4 Journal of Leukocyte Biology Volume 88, August 2010 www.jleukbio.org
performed 7 days after stimulation at three different E:T ratiosto assess the effect of CD200 blockade on the killing of Ly5and Ly2 cells by activated effectors. Data shown in the figuresbelow are for E:T ratios of 10:1. Figure 2A shows pooled re-sults from nine independent experiments using PBL stimula-tors from five different donors. Ly5 and Ly2 cells were poorlyimmunogenic when used alone, and optimal lysis was only�6%. Addition of the 1B9 anti-CD200 mAb, but not of 5A9, pro-duced an approximate fivefold increase in the killing of CD200�
Ly5 cells and no significant change in the killing of CD200– Ly2cells (data not shown). The isotype control antibody 3H4 failedto augment killing of Ly5 or Ly2 cells (Fig. 2B).
The enhanced lysis seen using 1B9 was relatively independentof the PBMC effector source and occurred even after pretreat-ment of tumor cells (but not PBMC) with mAb (data not shown),consistent with the primary target, CD200, expressed on the Ly5tumor cells themselves. Moreover, the killing of Ly5 cells was ab-rogated following depletion of CD8� cells, suggesting that CD8�
cytotoxic T cells are likely involved in tumor killing in this system(Fig. 2C). Interestingly, when CD4� T cells were depleted, lysis ofLy5 cells increased threefold, even without CD200 blockade (Fig.2C), which may be taken to reflect an intrinsic, autoregulatoryrole for CD4� cell subsets.
Functional inhibition of expression of CD200 in Ly5lymphoma cells by siRNAAs an alternate approach to modifying functional CD200 ex-pression on tumor cells, we used synthetic siRNAs to down-regulate CD200 expression. Three independent, commercialsiRNAs were examined for their ability to modify CD200 ex-
pression at the mRNA (Fig. 3A) and protein level (Fig. 3B).Optimal silencing was seen using siRNAs #4 and #6 (Fig. 3).Western blotting and FACS analysis were used to monitorknockdown of CD200 at the protein level following siRNAtranfection. By Western blots, incomplete silencing was ob-served. Similarly, cell surface level CD200 expression on Ly5cells was reduced by �50% at 24 h after transfection of siRNA#4 and #6 by FACS (data not shown).
We compared the relative increase in induction of CTL byLy5 cells in vitro using siRNAs or anti-CD200 mAb to decreasefunctional CD200 expression on tumor stimulator cells. Asshown in Figure 4, 1B9 and anti-CD200 siRNAs #4 and #6 aug-mented induction of CTL for lymphoma cells in vitro (Fig. 4;one of three such studies). Consistent with data in Figure 2using CD200 blockade by mAb, neither of the two siRNAsmodulated the killing of Ly2 cells. These data confirm thatfunctional inhibition of CD200 expression on tumor cells, bymAb or siRNA silencing, augments generation of anti-tumorimmunity in vitro.
Augmented cytokine production in MLCs followingdecreased CD200 expressionIn addition to exploring whether anti-CD200 or CD200 siRNAscould alter induction of CTL in vitro in MLCs with tumorcells, we asked whether these same reagents would alter cyto-kine production in vitro. Supernatants from PBMCs stimulatedwith Ly5 or Ly2 cells were collected at 18 and 42 h after stimu-lation, with typical data (one of four such studies) shown inFigure 5 (Fig. 5A shows TNF-� production; Fig. 5B, IFN-� pro-duction). Consistent with the data in Figure 2, minimal induc-
A
30
35
% specific lysis of Ly5 cells with or without anti-CD200 mAb
10
15
20
25
% s
peci
fic ly
sis
0
5
0
PBL+Ly5 PBL+Ly5+3H4 PBL+Ly5+1B9 PBL+Ly5+5A9
40% specific lysis of Ly2 cells with or without anti-CD200 mAb
B
20
25
30
35
40
cific
lysi
s
0
5
10
15
% s
pec
Ly2 only Ly2+1B9
40 UntreatedCD4 depletion
% Specific lysis of Ly5 cells by PBL following depletion of distinct cell subsets
C
5
10
30 CD8-depletionCD4-depletion
spec
ific
lysi
s
9B1+5yLylno 5yLylno 9B1
-5
0
% s
∗∗
Figure 2. Effect of anti-CD200 mAb 1B9 on induction of CTL by lym-phoma cells. hPBLs (1.2�106) were stimulated with 8 � 104 mitomy-cin C-treated Ly5 or Ly2 cells for 7 days in the presence or absence ofthe rat anti-hCD200 antibodies 1B9 and 5A9 using 3H4 as an isotypecontrol antibody. (A) 1B9 (**, P�0.0001) and 5A9 (P�0.05) en-hanced CTL responses to Ly5 cells, and 3H4 showed no significanteffect. (B) Failure of anti-CD200 mAb to augment CTL responses in-duced by Ly2 cells. (C) Depletion of CD8� T cells abrogated augmen-tation of Ly5 lysis by 1B9, and depletion of CD4� T cells enhancedLy5 lysis significantly, even in the absence of 1B9 (**, P�0.05). All Pvalues were calculated using percent killing obtained from effectorPBMCs with stimulation by Ly5 cells as a reference.
Wong et al. CD200 controls immunity to CLL
www.jleukbio.org Volume 88, August 2010 Journal of Leukocyte Biology 5
tion of cytokine production occurred using Ly2 cells as stimu-lator, and there was no further augmentation using anti-CD200 mAb. In contrast, although Ly5 cells induced minimalcytokine production in the absence of additional manipula-tion, inclusion of anti-CD200 mAb or pretreatment of tumorcells with siRNAs augmented induction of TNF-� and IFN-�.Again, these effects were not seen using control mAb (3H4) orsiRNAs (Fig. 5). CD200 blockade did not affect the productionlevel of a number of other cytokines, including IL-4, IL-6,IL10, IL-12, and TGF-�. Moreover, the changes in TNF-� andIFN-� levels were not observed in the absence of stimulationby Ly5 cells (data not shown).
Augmented killing of primary CLL cells by CD200blockade
Immunodeficiency is one of the clinical hallmarks of CLL. Tcells from CLL patients generally show Th2 polarization andexpress low levels of CD80, CD86, and CD154 [18]. As CLLcells express high levels of CD200, the CD200:CD200R axismay be an important pathway involved in suppression of T cellactivity by CLL cells. We thus examined the effect of CD200blockade on the killing of primary CLL cells using 1B9 (themAb with the most profound effect in the studies describedabove). Effector PBL from two different donors was stimulated
A
0.001
0.0015
NA
leve
l
Relative CD200 RNA levels in Ly5 cells after siRNA transfection
0
0.0005
Rel
ativ
e m
RN
Reagent only CD200 siRNA#1 CD200 siRNA#4 CD200 siRNA#6
B1 2 3 4 5 6
CD200
MAPK (House-keeping) protein)
Figure 3. Silencing of CD200 expression by specific oli-godeoxynucleotides. Ly5 cells were transfected with 2ug ofthe CD200 siRNAs #1, #4, or #6. All silencers were designedby Qiagen. A control using lipofectamine treatment but nosiRNA was included. Cells were assayed ay 48 h (RNA) or72 h (protein) following transfection. (A) CD200 mRNA levelas shown by real-time PCR. CD200 silencer #1 showed noeffect. (B) Expression of CD200 protein was decreased aftertransfection with silencers #4 and #6, but not #1.
20
25
30
lysi
s
** **
**
% Specific lysis of Ly5 cells transfected with control or CD200 specific siRNA
0
5
10
15
% s
peci
fic
PBL+ly5
PBL+ly5+1
B9
PBL+ly5+ -
ve siR
NA
PBL+ly5+s
iRNA#4
PBL+ly5+s
iRNA#6
Figure 4. siRNAs for CD200 enhanced cytotoxic killing ofLy5 cells to a similar degree as that achieved using anti-CD200 mAb 1B9 (**, P<0.05, with both CD200 silencers).The negative control silencer (#1) had no effect.
6 Journal of Leukocyte Biology Volume 88, August 2010 www.jleukbio.org
with mitomycin C-treated primary CLL cells from three differ-ent CLL patients (see Table 1), and killing in this case wasassessed using a 51Cr release assay. Interestingly, effector cellsderived from Donor 1 showed only low levels of killing of allthree CLL targets (Fig. 6A; data are shown as mean�sd forkilling of all three CLL targets), whereas effector cells derivedfrom Donor 2 killed all targets to a greater degree (data notshown). However, regardless of the quantitative level of killing,CD200 blockade (but not control 3H4 antibody) increasedkilling of all CLL targets for both effector populations.
As was observed for killing of Ly5 cells, depletion of CD8� Tcells prior to stimulation resulted in minimal killing of CLLtargets, indicating involvement of CD8� effector cells in CLLkilling (Fig. 6B). Interestingly, depletion of CD4� T cells alonewas sufficient to augment killing of Ly5 targets in the absence
of CD200 blockade (Fig. 2C), and augmented killing of CLLtargets was seen only using CD200 blockade and depletion ofCD4� T cells (Fig. 6B). We speculate that this may reflect theinvolvement of different effector populations for the two tar-get populations studied.
As CD200 induces immunoregulation following binding to areceptor CD200R, receptor expression was examined on cellsharvested from the spleens of two patients who had under-gone splenectomy for clinical reasons associated with diseasetreatment (Patients I and II; see Table 1). Greater than 90%of cells were CD19�CD5� CLL cells in the spleen of Patient I,whereas T cells constituted �50% of all cells from the spleenof Patient II (Fig. 6, D and E). Despite these differences incellular constitution, �90% of CD4� T cells in both spleenpopulations stained positive for CD200R, and only a minorpopulation of CD8� T cells (�1%) expressed CD200R (Fig.6D). Splenic CD5� CLL cells, on the other hand, did notshow detectable levels of CD200R. A direct comparison ofCD200R expression on splenic and circulating CD4� T cellsand CLL cells in the same patients could not be made, as pe-ripheral blood from the two splenectomized patients was notavailable at the time of study. However, unlike CD200 (Fig.6E), CD200R was never detected on CD5� CLL cells in spleen(Fig. 6D) or peripheral blood (data not shown).
As an adjunctive approach, we also investigated whether CLLserum, which we have found in independent studies to be capa-ble of suppressing human allogeneic CTL immune responses invitro (unpublished), lost this suppressive capacity after passageover a CD200 immunoadsorbent column. Data in Figure 6C showthat CTL activity (measured in human allogeneic MLCs at Day 6)was inhibited by CLL serum but that this inhibition was attenu-ated following absorption of CD200 from the serum, againconsistent with an important role for CD200 (in this case,in soluble form in CLL serum) in suppressing T cell-medi-ated immunity.
Association of down-regulated CD200 expression withincreased immunogenicity of CLL cellsTreatments of primary CLL cells with immunomodulators,such as TLR7 agonists, IL-2, and PKC agonists, have beenshown to improve immunogenicity of CLL cells in vitro, poten-tially by increasing expression of costimulatory molecules andrendering them more effective targets for lymphokine-acti-vated killer cells [3, 14, 16]. As cell surface expression ofCD200 provides immunosuppressive signals that counter theeffect of costimulatory molecules, we asked whether treatmentsdesigned to modulate immunogenicity of CLL cells wouldhave a concomitant effect on CD200 expression. Primary CLLcells from five patients (see Table 1) at different stages of dis-ease (Rai Stages II–IV) were treated with a TLR7 agonist ofthe imidazoquinoline family, Imiquimod, alone or in combina-tion with hrIL-2 and PMA for 24 h, and then assessed for cellsurface CD200 and CD5 expression by FACS. Data from stimu-lation of CLL cells from Patient 61 are shown as representedbelow (Fig. 7).
PMA and Imiquimod treatments reduced CD200 expressionon CLL cells significantly in all patients tested, and expressionof the CLL surface marker CD5 remained relatively un-
A TNFαα production 18 and 42hr after stimulation with Ly5
240
290
340
TNFα production 18 and 42hr after stimulation with Ly5
40
90
140
190
pg/m
lpg
/ml
18h42h
-10
PBL only
PBL+ly5 N
o reag
ent
PBL+ly5+
1B9
PBL+ly5 -
ve si
RNA
PBL+ly5 s
iRNA#4
PBL+ly5 s
iRNA#6
Ly5 only
P
B
800
900
IFNγ production 18 hr after stimulation with Ly5
300
400
500
600
700
800
-100
0
100
200
PBL only
No reag
ent
BL+ly5+1
B9
5 -ve si
RNA
ly5 siR
NA#4
ly5 siR
NA#6
Ly5 only
PBL+ly5 N PBL
PBL+ly5
PBL+ly5
PBL+ly5
Figure 5. Modulation of cytokine production in MLCs using anti-CD200 mAb or CD200-specific siRNAs. Supernatants were harvested18 and 42 h after stimulation and assayed at 1:5 and 1:10 dilution, re-spectively, for (A) TNF-� (18 and 42 h supernatant) and (B) IFN-�(18 h supernatant). Suppression of functional CD200 expression bymAb 1B9 or the CD200-specific silencers #4 and #6 augmented pro-duction of TNF-� and IFN-� by responder PBLs. Neither mAb norsiRNAs affected production of TNF-� or IFN-� when PBLs were stimu-lated with Ly2 cells.
Wong et al. CD200 controls immunity to CLL
www.jleukbio.org Volume 88, August 2010 Journal of Leukocyte Biology 7
changed (Fig. 7). Expression of CD83, a costimulatory mole-cule and an activation marker, was also increased in responseto PMA and Imiquimod, showing that the CLL cells were in“activated” states following treatment. PMA-induced CD200down-regulation was observed in all CLL cells, and Imiquimod-induced CD200 down-regulation was observed in cells fromfour out of five patients (data not shown). IL-2, on the otherhand, had no effect on CD200 expression and produced mini-
mal increase in CD83 expression (Fig. 7). In agreement withprevious reports, in which CLL cells were shown to exhibit aheterogeneous response to PMA and Imiquimod in the up-regulation of CD83, CD80, and CD86 expressions [3], the ef-fect of these two stimuli on CD200 expression also variedamong patients (data not shown). Reduction in CD200 expres-sion was most pronounced by concomitant treatment of PMA,Imiquimod, and IL-2, which as reported previously [3], also
25
30
is
A % Specific lysis of CLL cells in the presence or absence of anti-CD200 mAb
5
10
15
20
% s
peci
fic ly
s
0
9B1GgI taRB
25
30 PBL+CLLPBL+CLL+control IgG
10
15
20
% L
ysis
PBL+CLL+1B9
0
5
No depletion CD4 depletion CD8 depletion
C
CD20
0R1
CD5 CD8 CD4
CLL spleen 1
CLL spleen 2
CD5 CD8 CD4
CD20
0R1
0.1%
96.6%
0.6%
99.3% 0.2%
99.0%
0.5%
0.6%
99.0%
0.0%
90.3%
99.5%
CD
19C
D19
CD5
CD5
CD5 and CD19 gated CLL splenocytes:Whole CLL splenocytes
CD200
CD200
CLL spleen-1
CLL spleen-2
E
Whole CLL splenocytesCD5 and CD19 gated
CLL splenocytes:
D
Figure 6. CD200 blockade augments killing of primary CLL cells byallogenic effector PBLs. Effector PBLs were stimulated by mitomycinC-treated primary CLL cells in the presence of rat IgG or 1B9. Killingwas measured 7 days after stimulation by 51Cr release assay. (A) Killing
was shown as an average of three independent experiments using three different CLL targets and one PBL effector. Unstimulated PBLs wereused as negative controls. (B) Depletion of CD4� T cell from MLCs further augmented killing of CLL targets using CD200 blockade, andCD4� T cell depletion alone had no effect on CLL killing. Depletion of CD8� T cells reduced the killing of CLL cells to levels akin to thoseseen with unstimulated PBLs, even in the presence of 1B9. (C) Effect of CD200 absorption on immunosuppression in hMLCs using CLL pa-tient serum: CD200 was absorbed from CLL serum by overnight incubation of pooled CLL serum (obtained from 15 donors) with 1B9-conju-gated Sepharose beads. CLL serum before (CLL) or after (absorbed) CD200 absorption was added to hMLC at the indicated dilutions; resultsshow percent lysis of 51Cr-labeled target cells at a 30:1 E:T ratio in 6 h. CLL serum suppressed MLC reactivity in a dose-dependent mannercompared with controls (P�0.05), but this inhibition was lost after absorption. (D) Expression of CD200R on cells gated on CD5, CD8, andCD4. Of the three populations, over 90% of CD4� T cells stained positive for CD200R using cells from CLL spleen of both CLL patients, andneither CD8� T cells nor CLL (CD5�) cells expressed detectable levels of CD200R. (E) Expression of CD200 on splenic CLL cells: CLL popula-tions, as determined by the cell surface markers CD19 and CD5. CD5�CD19� CLL cells from both CLL spleen populations express high levelsof CD200.
8 Journal of Leukocyte Biology Volume 88, August 2010 www.jleukbio.org
resulted in the greatest increase in CD83 expression. The pres-ence of IL-2 in combination with PMA and Imiquimod, al-though causing further augmentation of CD83 expression, hadno effect on CD200 down-regulation.
DISCUSSION
Immunomodulatory molecules contributing to negative sig-naling of T cells are thought to play a pivotal role in regu-lating anti-tumor responses and tumor progression in hu-man malignancies. As examples, altered expression of im-munomodulatory molecules of the B7 family, B7-H1, B7-H3,
and B7-H4, have been detected in lung, prostate, ovarian,and kidney carcinomas and neuroblastoma [19]. In prostatecancer and clear cell renal carcinoma, B7-H3 overexpres-sion on tumor cells is associated with poor prognosis [20,21]. In ovarian cancer, the serum B7-H4 level has beenidentified as another marker that predicts poor prognosis[22]. Functional blockade of these immunomodulatory mol-ecules might thus provide a novel therapy for such malig-nancies. Indeed, blockade of CTLA4, a negative regulator ofT cells, using a fully humanized antibody, is currently underdevelopment in Phase III clinical trials in patients with ad-vanced melanoma and other malignancies [23].
Figure 7. Expression of CD200 on CLLcells in response to stimulation by PMA,Imiquimod, and IL-2. Fresh CLL cellswere treated with 30 ng/ml PMA, 3ug/ml Imiquimod, and 500 U/ml hrIL-2,alone or in combination, for 24 h. CD200expression (y-axis) on the surface of CLLcells was reduced in response to PMAand Imiquimod but not IL-2. CD200 ex-pression is down-regulated further whenCLL cells were treated with all three stim-ulants, whereas change in expression ofthe cell surface marker CD5 (x-axis) wasminimal. CD83 expression was also up-regulated in response to stimulation byPMA and Imiquimod (CD83� cells repre-sented by dark spots). The experimentwas repeated with fresh CLL cells fromfive patients, and typical data from onesuch study are shown.
Wong et al. CD200 controls immunity to CLL
www.jleukbio.org Volume 88, August 2010 Journal of Leukocyte Biology 9
In B cell malignancies, including lymphomas and CLL, thetumor cells themselves are known to be poorly immunogenic,despite the expression of high levels of MHC molecules andtumor antigens [24, 25]. A number of strategies have beeninvestigated to develop clinically applicable methodologies toenhance the immunogenicity of CLL cells [26, 27]. For exam-ple, transduction of CLL cells with the CD40 ligand has beenshown to enhance antigen-specific recognition of tumor cellsby autologous T cells in vitro [27]. Various attempts have alsobeen made to improve efficacy of vaccines targeting CLL-spe-cific antigens [28, 29].
Given the dominant nature of immunomodulatory signals, thestimulation of costimulatory molecules on tumor cells alone maynot be sufficient to overcome the poor immunogenicity of thetumor. Expression of CD200, a known immunoregulatory mole-cule, has been reported on CLL and lymphoma cells (ref. [10];see also Fig. 1A). Although the CD200 cell surface expressionlevel does not seem to correlate with other CLL clinical markers,it remains unknown whether CD200 expression levels on CLLcells vary in response to treatment or during disease progression.Our results described herein and data from other groups supportthe hypothesis that CD200 expression on tumor cells might beone of the contributors for the poor immunogenicity of leuke-mic/lymphoma cells. Blockade of functional CD200 expressionwould thus provide a promising approach to enhance immuno-genicity of such tumor cells. In support of this hypothesis, Kretz-Rommel et al. [12] demonstrated recently that blockade ofCD200 using specific humanized mAb enhanced anti-tumor re-sponses using hPBMCs and tumor cells artificially transfectedwith a CD200 lentiviral vector.
A drawback to the study reported by Kretz-Rommel et al. [12]is that lentiviral transfection often produces protein expressionlevels not reflective of those seen physiologically. Accordingly, wehave been interested in induction of tumor responses directedagainst primary CLL cells isolated from peripheral blood of pa-tients, as well as a B-lymphoma cell line, Ly5, which constitutivelyexpresses CD200 at levels paralleling those expressed by primaryCLL cells. A CD200– cell line, Ly2, was used as a control. CTLassays using these two cell lines as targets showed that althoughboth cell lines were poorly immunogenic, killing of CD200� Ly5cells and indeed, of primary CLL cells, but not CD200– Ly2 cells,was augmented greater than fivefold by the presence of a rat anti-hCD200 mAb 1B9 when compared with an isotype control anti-body, 3H4.
Although CD200 is expressed on normal B cells, and its ex-pression is increased on T cells upon activation, the effect of theCD200 blockade was PBMC donor-independent (unpublishedobservations; see also ref. [30]) and appears to represent target-ing of CD200 expressed on tumor cells. Antibody-mediatedCD200 blockade, as a means of enhancing CTL responses, wasaffected by the CD200 epitopes targeted, as another CD200-spe-cific mAb 5A9 produced much less augmentation of CTL induc-tion than 1B9, despite equivalent staining of Ly5 cells in FACS by5A9 and 1B9 (Fig. 1B). This is consistent with previous data indi-cating heterogeneity in the activity of different anti-CD200 mAbin different functional assays [17]. Thus, biochemical and func-tional characterization of the epitopes recognized by anti-CD200
mAb is of crucial importance in the design of CD200-specificmAb therapies.
CD200 blockade by 1B9- and CD200-specific siRNAs enhancedproduction of TNF-� and IFN-� in vitro from effector cells, sug-gesting that CD200 blockade may affect anti-tumor immunitythrough other (cytokine-mediated) mechanisms [31]. The datadescribed used two independent CD200 siRNAs (#4 and #6),both of which showed specific knockdown of CD200 at RNA andprotein levels. Interestingly, CD200 knockdown by siRNA #6 re-sulted in higher production of TNF-� and IFN-�, despite similaraugmentation of the CTL response to Ly5 cells after transfectionwith both silencers (#4 and #6). This may simply reflect the dif-ference in the effector populations responsible for activity inthese two assays.
The in vitro killing of tumor cells in our CTL assays was me-diated by CD8� cytotoxic effector cells, as demonstrated bythe minimal CTL response to Ly5 and primary CLL cells whenCD8� T cells were depleted from responder populations. NKcells, which have been shown to express CD200R [32], couldalso potentially play a role in the killing of tumor targets, par-ticularly in assays with CLL targets, where killing was not abro-gated completely after CD8 depletion (Fig. 6B). In our hands,�30% of the PBMCs stained with anti-CD56 mAb in FACSanalysis before cultures. We were unable to detect statisticallysignificant changes in the percent of cells stained with anti-CD56 mAb following culture (immediately before assaying lyticactivity), with levels 2–3% in all groups, and levels of CD4�/CD8� cells in nondepleted (by anti-CD4/CD8) cultures were�10%. We presume the equivalent and low survival of CD56�
cells following various cell depletion strategies reflects the ab-sence of production of mediators (cytokines) from CD4�
and/or CD8� cells in tumor cell-stimulated cultures, whichcould contribute to NK survival/growth in vitro. In addition,no significant changes in CD56� cells were seen in cultures incu-bated with anti-CD200 mAb (data not shown). We conclude thatthe differential killing activity seen following the manipulationsshown in Figure 6 is best explained by our hypothesis that CD8�
cells are the primary effector population assayed.Interestingly, the killing of Ly5 and CLL cells was affected sig-
nificantly by the absence of CD4� effectors. Although depletionof CD4� T cells was sufficient to enhance killing of Ly5 cells (Fig.2C), CD4 depletion augmented the killing of CLL cells, only inthe presence of CD200 blockade (Fig. 6B). We interpret thesedata as suggestive of the involvement of CD4� T cells in regula-tion of killing, directly (as a regulatory cell population itself; note,we have not independently assessed the effect of depletion ofCD25� cells in these assays) or indirectly, acting to affect the ac-tivity of other regulatory cells. The exact mechanism(s) involvedremain unexplored.
CLL cell-mediated T cell defects have been well documented.Recently, the formation of immunological synapses between CLLcells and autologous T cells was shown to be impaired [33]. Thisimpairment appears to be CLL-dependent, as incubation of CLLcells with allogenic T cells also led to failure in formation of nor-mal immunologic synapse between CLL cells and normal T cells.The expression of activation markers on T cells was also impairedafter incubation with CLL cells in a mechanism involving directcell-cell contact as well as soluble factors secreted by CLL cells
10 Journal of Leukocyte Biology Volume 88, August 2010 www.jleukbio.org
[34]. The results of our CTL studies showed that CD200 may beone of the cell-surface factors contributing to CLL-mediated Tcell suppression, as CD200 blockade resulted in enhanced killingof CLL cells.
Further evidence for an important role of the CD200:CD200Raxis in CLL was supported by the high frequency of CD200R�
CD4� T cells in the spleen of CLL patients as detected by FACSanalyses (Fig. 6D). CD8� T cells and CLL cells, on the otherhand, showed no detectable level of CD200R expression, consis-tent with the hypothesis that the primary target for CLL-derived,CD200-mediated, immunosuppressive signals represents CD4�
and/or other CD200R� (but non-CD8�) cells. It remains to bedetermined whether CD200R�CD4� T cells and CD200� CLLcells exist in close proximity in vivo in CLL microenvironments.However, the observation that CD200R-expressing CD4� T cellsand CD200-expressing CLL cells are present in the same micro-environment (spleen, in this case) supports a model, in whichCD200-mediated suppression of CD200R� CD4� T cells is, inpart, at least responsible for the Th2 cytokine polarization anddiminished CD8� cytotoxic T cell function observed in CLL pa-tients. The ability of anti-CD200 to augment lysis of fresh CLLcells following CD4 depletion may suggest a role for such (anti-CD200) therapy in CLL, alongside treatment with, e.g., fludara-bine and alemtuzumab, both of which have a significant ability tokill CD4� cells [35, 36]. A potential concern for immunothera-pies targeting the CD200:CD200R pathway is autoimmunity, asthis pathway has been shown to play important regulatory roles ina number of autoimmunity models in rodents, including collag-en-induced arthritis and experimental allergic encephalomyelitis[37–39]. In vivo models of CLL will be needed to address suchsafety:efficacy questions. It also remains open to speculationwhether CD200 blockade may even enhance treatments such asallogenic bone marrow transplantation, in which killing of tumorcells is mediated by allogenic T cells.
Treatment with imidazoquinolines, a family of TLR7 agonists,along with IL-2 and PKC agonists, has also been considered as ameans to improve immunogenicity of CLL cells [2]. In vitrotreatment of CLL cells with these immunmodulators is effectivein transforming CLL cells to a dendritic cell-like phenotype withhigh expression of costimulatory molecules, production of inflam-matory cytokines, and the ability to stimulate T cell proliferation,at least in vitro [3, 14, 16]. We found that expression of CD200on the surface of CLL cells was down-regulated in response toImiquimod or PMA, with an optimal decrease observed followingcombined use of Imiquimod and PMA. IL-2 treatment did notaffect CD200 cell surface expression on CLL cells. Given that re-versal of CD200-mediated suppression does not seem to requirecomplete abrogation of CD200 cell surface expression (see Figs. 3and 4 using siRNAs), this reduction of CD200 expression on tu-mor cells achieved by Imiquimod and PMA may represent a keyfeature of their immunostimulatory activities. As PMA and Imi-quimod are global activators of multiple pathways, further investi-gations are required to evaluate the contribution of alteredCD200 expression to the biological effects produced by theseagents. However, our data provide evidence for the potential oftherapies targeting the CD200:CD200R axis, in combination withtreatments to enhance immunogenicity of tumor cells, as a mech-anism to augment anti-tumor responses.
Whether the down-regulation of CD200 expression in re-sponse to PMA and Imiquimod is mediated through tran-scriptional control or mechanisms involving ectodomaincleavage by proteases is currently unknown. PMA is a knownactivator of a disintegrin and metalloproteinase domainfamily of proteases and is responsible for the inducibleshedding of a number of cytokines and chemokines, includ-ing TNF, TNFRI, IL-6R, and CX3CL1 [40 – 42]. It is thuspossible that CD200 down-regulation following PMA stimu-lation involves proteolytic cleavage of cell surface CD200,and preliminary observations using in vitro studies of CLLcells support this hypothesis [43]. Such a shedding eventmight also contribute to the existent of a soluble, immuno-suppressive form of CD200 in CLL serum (see Fig. 6C).
In sum, we have shown that CD200-mediated immunosuppres-sion is an important mechanism used by CLL cells as a means toinhibit anti-tumor immune responses. We predict that inhibitionof CD200 expression on tumor cells in general may have impor-tant clinical implications in developing novel immunotherapies.
AUTHORSHIP
Karrie K. Wong contributed to all research work, the design ofstudies, and writing of the manuscript. Ismat Khatri contrib-uted to biochemistry and FACS analysis. Suchinta Shaha con-tributed to CLL cell preparation. David E. Spaner contributedto the design of all studies and writing of the manuscript.Reginald M. Gorczynski contributed to research, the design ofall studies, and writing of the manuscript.
ACKNOWLEDGMENTS
This research was supported by grants from CIHR and Heartand Stroke Association (to R. M. G.) and CIHR and OntarioInstitute for Cancer Research (to D. E. S.). K. K. W. was sup-ported by a CIHR-Training Program in Regenerative Medicinegraduate student fellowship.
REFERENCES
1. Shaffer, A. L., Rosenwald, A., Staudt, L. M. (2002) Lymphoid malignan-cies: the dark side of B-cell differentiation. Nat. Rev. Immunol. 2, 920–932.
2. Spaner, D. E., Masellis, A. (2007) Toll-like receptor agonists in the treat-ment of chronic lymphocytic leukemia. Leukemia 21, 53–60.
3. Tomic, J., White, D., Shi, Y., Mena, J., Hammond, C., He, L., Miller,R. L., Spaner, D. E. (2006) Sensitization of IL-2 signaling through TLR-7enhances B lymphoma cell immunogenicity. J. Immunol. 176, 3830–3839.
4. McWhirter, J. R., Kretz-Rommel, A., Saven, A., Maruyama, T., Potter,K. N., Mockridge, C. I., Ravey, E. P., Qin, F., Bowdish, K. S. (2006) Anti-bodies selected from combinatorial libraries block a tumor antigen thatplays a key role in immunomodulation. Proc. Natl. Acad. Sci. USA 103,1041–1046.
5. Moreaux, J., Veyrune, J. L., Reme, T., De Vos, J., Klein, B. (2008) CD200:a putative therapeutic target in cancer. Biochem. Biophys. Res. Commun.366, 117–122.
6. Petermann, K. B., Rozenberg, G. I., Zedek, D., Groben, P., McKinnon, K.,Buehler, C., Kim, W. Y., Shields, J. M., Penland, S., Bear, J. E., Thomas,N. E., Serody, J. S., Sharpless, N. E. (2007) CD200 is induced by ERK andis a potential therapeutic target in melanoma. J. Clin. Invest. 117, 3922–3929.
7. Siva, A., Xin, H., Qin, F., Oltean, D., Bowdish, K. S., Kretz-Rommel, A.(2008) Immune modulation by melanoma and ovarian tumor cellsthrough expression of the immunosuppressive molecule CD200. CancerImmunol. Immunother. 57, 987–996.
Wong et al. CD200 controls immunity to CLL
www.jleukbio.org Volume 88, August 2010 Journal of Leukocyte Biology 11
8. Tonks, A., Hills, R., White, P., Rosie, B., Mills, K. I., Burnett, A. K., Dar-ley, R. L. (2007) CD200 as a prognostic factor in acute myeloid leukemia.Leukemia 21, 566–568.
9. Kawasaki, B. T., Mistree, T., Hurt, E. M., Kalathur, M., Farrar, W. L.(2007) Co-expression of the toleragenic glycoprotein, CD200, with mark-ers for cancer stem cells. Biochem. Biophys. Res. Commun. 364, 778–782.
10. Gorczynski, R. M. (2005) CD200 and its receptors as targets for immuno-regulation. Curr. Opin. Investig. Drugs 6, 483–488.
11. Gorczynski, R. M., Chen, Z., Hu, J., Kai, Y., Lei, J. (2001) Evidence of arole for CD200 in regulation of immune rejection of leukemic tumorcells in C57BL/6 mice. Clin. Exp. Immunol. 126, 220–229.
12. Kretz-Rommel, A., Qin, F., Dakappagari, N., Ravey, E. P., McWhirter, J.,Oltean, D., Frederickson, S., Maruyama, T., Wild, M. A., Nolan, M. J.,Wu, D., Springhorn, J., Bowdish, K. S. (2007) CD200 expression on tu-mor cells suppresses antitumor immunity: new approaches to cancer im-munotherapy. J. Immunol. 178, 5595–5605.
13. Pallasch, C. P., Ulbrich, S., Brinker, R., Hallek, M., Uger, R. A., Wendt-ner, C. M. (2009) Disruption of T cell suppression in chronic lympho-cytic leukemia by CD200 blockade. Leuk. Res. 33, 460–464.
14. Shi, Y., White, D., He, L., Miller, R. L., Spaner, D. E. (2007) Toll-like re-ceptor-7 tolerizes malignant B cells and enhances killing by cytotoxicagents. Cancer Res. 67, 1823–1831.
15. Tweeddale, M. E., Lim, B., Jamal, N., Robinson, J., Zalcberg, J., Lock-wood, G., Minden, M. D., Messner, H. A. (1987) The presence of clono-genic cells in high-grade malignant lymphoma: a prognostic factor. Blood69, 1307–1314.
16. Hammond, C., Shi, Y., Mena, J., Tomic, J., Cervi, D., He, L., Millar, A. E.,Debenedette, M., Schuh, A. C., Baryza, J. L., Wender, P. A., Radvanyi, L.,Spaner, D. E. (2005) Effect of serum and antioxidants on the immunoge-nicity of protein kinase C-activated chronic lymphocytic leukemia cells.J. Immunother. 28, 28–39.
17. Chen, D. X., Gorczynski, R. M. (2005) Discrete monoclonal antibodiesdefine functionally important epitopes in the CD200 molecule responsi-ble for immunosuppression function. Transplantation 79, 282–288.
18. Gorgun, G., Holderried, T. A., Zahrieh, D., Neuberg, D., Gribben, J. G.(2005) Chronic lymphocytic leukemia cells induce changes in gene ex-pression of CD4 and CD8 T cells. J. Clin. Invest. 115, 1797–1805.
19. Flies, D. B., Chen, L. (2007) The new B7s: playing a pivotal role in tumorimmunity. J. Immunother. 30, 251–260.
20. Crispen, P. L., Sheinin, Y., Roth, T. J., Lohse, C. M., Kuntz, S. M.,Frigola, X., Thompson, R. H., Boorjian, S. A., Dong, H., Leibovich, B. C.,Blute, M. L., Kwon, E. D. (2008) Tumor cell and tumor vasculature ex-pression of B7–H3 predict survival in clear cell renal cell carcinoma. Clin.Cancer Res. 14, 5150–5157.
21. Roth, T. J., Sheinin, Y., Lohse, C. M., Kuntz, S. M., Frigola, X., Inman,B. A., Krambeck, A. E., McKenney, M. E., Karnes, R. J., Blute, M. L.,Cheville, J. C., Sebo, T. J., Kwon, E. D. (2007) B7–H3 ligand expressionby prostate cancer: a novel marker of prognosis and potential target fortherapy. Cancer Res. 67, 7893–7900.
22. Oikonomopoulou, K., Li, L., Zheng, Y., Simon, I., Wolfert, R. L., Valik,D., Nekulova, M., Simickova, M., Frgala, T., Diamandis, E. P. (2008) Pre-diction of ovarian cancer prognosis and response to chemotherapy by aserum-based multiparametric biomarker panel. Br. J. Cancer 99, 1103–1113.
23. Comin-Anduix, B., Lee, Y., Jalil, J., Algazi, A., de la Rocha, P., Camacho,L. H., Bozon, V. A., Bulanhagui, C. A., Seja, E., Villanueva, A., Straatsma,B. R., Gualberto, A., Economou, J. S., Glaspy, J. A., Gomez-Navarro, J., Ribas,A. (2008) Detailed analysis of immunologic effects of the cytotoxic T lym-phocyte-associated antigen 4-blocking monoclonal antibody tremelimumabin peripheral blood of patients with melanoma. J. Transl. Med. 6, 22.
24. Rezvany, M. R., Jeddi-Tehrani, M., Rabbani, H., Lewin, N., Avila-Carino,J., Osterborg, A., Wigzell, H., Mellstedt, H. (2000) Autologous T lympho-cytes may specifically recognize leukemic B cells in patients with chroniclymphocytic leukemia. Br. J. Haematol. 111, 608–617.
25. Rezvany, M. R., Jeddi-Tehrani, M., Wigzell, H., Osterborg, A., Mellstedt, H.(2003) Leukemia-associated monoclonal and oligoclonal TCR-BV use in pa-tients with B-cell chronic lymphocytic leukemia. Blood 101, 1063–1070.
26. Kofler, D. M., Mayr, C., Wendtner, C. M. (2006) Current status of immu-notherapy in B cell malignancies. Curr. Drug Targets 7, 1371–1374.
27. Mayr, C., Kofler, D. M., Buning, H., Bund, D., Hallek, M., Wendtner,C. M. (2005) Transduction of CLL cells by CD40 ligand enhances an an-
tigen-specific immune recognition by autologous T cells. Blood 106, 3223–3226.
28. Hus, I., Schmitt, M., Tabarkiewicz, J., Radej, S., Wojas, K., Bojarska-Junak,A., Schmitt, A., Giannopoulos, K., Dmoszynska, A., Rolinski, J. (2008)Vaccination of B-CLL patients with autologous dendritic cells can changethe frequency of leukemia antigen-specific CD8� T cells as well asCD4�CD25�FoxP3� regulatory T cells toward an antileukemia re-sponse. Leukemia 22, 1007–1017.
29. Palma, M., Adamson, L., Hansson, L., Kokhaei, P., Rezvany, R., Mellstedt,H., Osterborg, A., Choudhury, A. (2008) Development of a dendritic cell-based vaccine for chronic lymphocytic leukemia. Cancer Immunol. Immu-nother. 57, 1705–1710.
30. Kretz-Rommel, A., Qin, F., Dakappagari, N., Cofiell, R., Faas, S. J., Bowd-ish, K. S. (2008) Blockade of CD200 in the presence or absence of anti-body effector function: implications for anti-CD200 therapy. J. Immunol.180, 699–705.
31. Dunn, G. P., Koebel, C. M., Schreiber, R. D. (2006) Interferons, immu-nity and cancer immunoediting. Nat. Rev. Immunol. 6, 836–848.
32. Clark, D. A., Wong, K., Banwatt, D., Chen, Z., Liu, J., Lee, L., Gorczynski,R. M., Blajchman, M. A. (2008) CD200-dependent and nonCD200-depen-dent pathways of NK cell suppression by human IVIG. J. Assist. Reprod.Genet. 25, 67–72.
33. Ramsay, A. G., Johnson, A. J., Lee, A. M., Gorgun, G., Le Dieu, R., Blum,W., Byrd, J. C., Gribben, J. G. (2008) Chronic lymphocytic leukemia Tcells show impaired immunological synapse formation that can be re-versed with an immunomodulating drug. J. Clin. Invest. 118, 2427–2437.
34. Ramsay, A. G., Gribben, J. G. (2009) Immune dysfunction in chronic lym-phocytic leukemia T cells and lenalidomide as an immunomodulatorydrug. Haematologica 94, 1198–1202.
35. Hegde, U., Chhabra, A., Chattopadhyay, S., Das, R., Ray, S., Chakraborty,N. G. (2008) Presence of low dose of fludarabine in cultures blocks regu-latory T cell expansion and maintains tumor-specific cytotoxic T lympho-cyte activity generated with peripheral blood lymphocytes. Pathobiology 75,200–208.
36. Mone, A. P., Cheney, C., Banks, A. L., Tridandapani, S., Mehter, N.,Guster, S., Lin, T., Eisenbeis, C. F., Young, D. C., Byrd, J. C. (2006) Ale-mtuzumab induces caspase-independent cell death in human chroniclymphocytic leukemia cells through a lipid raft-dependent mechanism.Leukemia 20, 272–279.
37. Gorczynski, R. M., Chen, Z., Lee, L., Yu, K., Hu, J. (2002) Anti-CD200Rameliorates collagen-induced arthritis in mice. Clin. Immunol. 104, 256–264.
38. Gorczynski, R. M., Chen, Z., Yu, K., Hu, J. (2001) CD200 immunoadhesinsuppresses collagen-induced arthritis in mice. Clin. Immunol. 101, 328–334.
39. Chitnis, T., Imitola, J., Wang, Y., Elyaman, W., Chawla, P., Sharuk, M.,Raddassi, K., Bronson, R. T., Khoury, S. J. (2007) Elevated neuronal ex-pression of CD200 protects Wlds mice from inflammation-mediated neu-rodegeneration. Am. J. Pathol. 170, 1695–1712.
40. Bell, J. H., Herrera, A. H., Li, Y., Walcheck, B. (2007) Role of ADAM17in the ectodomain shedding of TNF-� and its receptors by neutrophilsand macrophages. J. Leukoc. Biol. 82, 173–176.
41. Matthews, V., Schuster, B., Schutze, S., Bussmeyer, I., Ludwig, A., Hun-dhausen, C., Sadowski, T., Saftig, P., Hartmann, D., Kallen, K. J., Rose-John, S. (2003) Cellular cholesterol depletion triggers shedding of thehuman interleukin-6 receptor by ADAM10 and ADAM17 (TACE). J. Biol.Chem. 278, 38829–38839.
42. Sanderson, M. P., Erickson, S. N., Gough, P. J., Garton, K. J., Wille, P. T.,Raines, E. W., Dunbar, A. J., Dempsey, P. J. (2005) ADAM10 mediatesectodomain shedding of the � cellulin precursor activated by p-amino-phenylmercuric acetate and extracellular calcium influx. J. Biol. Chem.280, 1826–1837.
43. Wong, K., Shaha, S., Spaner, D.S., Gorczynski, R. (2009) Potential rolefor serum soluble CD200 in human chronic lymphocytic leukemia. J. Im-munol. 182, 88.15 (abs.).
KEY WORDS:cancer � immune evasion � cancer immunology
12 Journal of Leukocyte Biology Volume 88, August 2010 www.jleukbio.org
