Fig 1.
Platelet cloaking inhibits NK cell functions.
(A,B) Quantification of platelet cloaking of ovarian and melanoma tumour cells, and the myelogenous leukaemia control cell line K562. Tumour cell lines were co-incubated with and without platelets and analysed for the surface expression of the CD42b platelet specific marker. Expression data are represented by histogram (A), as a percentage of total cells and by mean fluorescent intensity (MFI; B). (C-F) Analysis of the function consequences of platelet cloaking on NK cell functions. NK cell anti-tumour assays were performed by co-incubating PBMCs with tumour cells (cloaked and uncloaked) for 4 hours and measuring CD107a (C,E) and IFNgamma production (D,F) as markers of activation. (E,F) To dissect the respective roles of platelet contact factors (cloaked (minus the releasate)) and soluble factors (releasate (minus the platelet cloak)) platelets and releasate were isolated and used to treat NK cells in NK activation assays as previously described. (C-F) Data analysed by ANOVA—each experiment represents mean±S.E.M. of at least three independent experiments. * = p<0.05, ** = p<0.01, *** = p<0.001.
Fig 2.
Platelets modulate NKG2D receptor expression by inducing the release of NKG2DL MICA and MICB from the surface of tumour cell lines.
(A,B,C) To examine the role of platelet cloaked tumour cells, platelets and TRAP-induced platelet releasate in regulating NKG2D expression, quantification of NKG2D was performed by flow cytometry of PBMCs in the presence or absence of tumour cells, platelet cloaked tumour cells, platelets or platelet releasate for 24 hours. (C) Relative expression of MICA and MICB on tumour cell lines was quantified by MICA/MICB mAbs and flow cytometry (n = 3) (D,E) The role of the platelet cloak on the expression on known NK cell activating ligands MICA and MICB on tumour cells was investigated by measuring baseline expression (clear bars) and comparing this to platelet cloaked tumour cells (filled bars). Tumour cells were incubated for 4 (D) or 24 (E) hours in the presence or absence of platelets and ligand expression was quantified by flow cytometry. (F) Analysis of soluble MICA and MICB in the supernatants of tumour cell lines incubated for 24 hours in the presence or absence of platelets by ELISA. (G) Taqman analysis of platelet cloaked and uncloaked ovarian/melanoma tumour cells for ADAM19, ADAM10 and ADAM17 genes. Values shown are cloaked cells relative to uncloaked cells. (A,B,D) Data analysed by ANOVA—each experiment represents mean±S.E.M. of at least three independent experiments. (G) Data analysed by t-test and represented as standard deviation of at least three independent experiments, * = p<0.05, ** = p<0.01, *** = p<0.001.
Fig 3.
MICA and MICB in the releasate modulate NKG2D expression and NK cell functions.
(A) To examine the role of MICA and MICB from the platelet releasate (from platelet cloaked tumour cells) in regulating NKG2D and NK cell functions, the platelet releasate was neutralised using monoclonal antibodies against MICA and MICB as previously described. NKG2D expression (A) and NK cell functions (CD107a (B) and IFNgamma (C)) were quantified following standard 4h NK cell function assays by flow cytometry and results are expressed as the mean fluorescent intensity (MFI; A) or as a percentage of unstimulated or IgG control (B and C). (D) Schematic representation of the platelet sNKG2DL mediated decoy mechanism of tumour immune evasion (A-C) Data analysed by ANOVA—each experiment represents mean±S.E.M. of at least three independent experiments. * = p<0.05, ** = p<0.01, *** = p<0.001
Fig 4.
Platelets disrupt the CD226/CD96-CD112/CD155 NK axis for NK cell targeting of tumour cell lines.
(A) Blocking assays using mAb against CD226, CD96 or CD155/CD112 in standard anti-tumour assays (CD107a surface expression and IFNy production) to analyse the role of each molecule in NK cell targeting of tumour cell lines. (B) The role of the platelet cloak on the expression on known NK cell activating ligands CD112 and CD155 on tumour cells was investigated by measuring baseline expression (clear bars) and comparing this to platelet cloaked tumour cells (filled bars). Tumour cells were incubated for 24 hours in the presence or absence of platelets and ligand expression was quantified by flow cytometry. (C) Analysis of the effect of tumour cells with and without the platelet cloak on expression of the NK cell receptors CD226 and CD96. PBMCs (clear bars) were co-incubated with uncloaked (filled bar) and cloaked (grey bars) tumour cells for 24 hours and expression of CD226 and CD96 were quantified by flow cytometry. (D) To dissect the role of soluble factors, experiments were repeated as above with TRAP-activated degranulating platelets (+platelets group) and releasate from TRAP-activated degranulated platelets that was cleared of platelet cellular material (+releasate group). (A-D) Data analysed by ANOVA—each experiment represents mean±S.E.M. of at least three independent experiments. * = p<0.05, ** = p<0.01, *** = p<0.001.
Fig 5.
TGFbeta is an active soluble molecule in platelet derive immune evasion of circulating tumour cells.
(A) Quantification of TGFbeta (total and active) released from cloaked tumour cells (filled box) compared with uncloaked cells (clear box) using ELISA duo-set analysis of supernatants of tumours cells incubated for 4 hours in the presence or absence of platelets. (B) To dissect the role of TGFbeta, recombinant TGFbeta1 molecule was used to pre-treat NK cells for 24 hours and CD226 and CD96 receptor expression was quantified by flow cytometry. To confirm the in vitro role of platelet derived TGFbeta, neutralising antibody against TGFbeta was used to pre-clear supernatants and receptor expression was compared with untreated cells. (C) Functional assays to confirm the effect of TGFbeta on NK cell anti-tumour functions were performed by pre-treating NK cells with recombinant TGFbeta and measuring CD107a surface expression and IFNgamma production as NK cell activation markers. (D,E) Analysis of the role of TGFbeta in platelet releasate on CD107a expression (D) and IFNgamma production (E) by comparison of the effect of TRAP-induced platelet releasate versus releasate that has been cleared of TGFbeta by neutralising antibody. (C-E) Results represent the detection of the activation markers as a percentage of the average response of untreated NK cells to each individual tumour cell line. (A-E) Data analysed by ANOVA—each experiment represents mean±S.E.M. of at least three independent experiments. * = p<0.05, ** = p<0.01, *** = p<0.001.