The fate of intracellular peptides and MHC class I antigen presentation Neijssen, J.J.

(1)

The fate of intracellular peptides and MHC class I antigen presentation

Neijssen, J.J.

Citation

Neijssen, J. J. (2008, February 6). The fate of intracellular peptides and MHC class I antigen presentation. Retrieved from https://hdl.handle.net/1887/12591

Version: Corrected Publisher’s Version

License: Licence agreement concerning inclusion of doctoral thesis in the Institutional Repository of the University of Leiden

Downloaded from: https://hdl.handle.net/1887/12591

Note: To cite this publication please use the final published version (if applicable).

(2)

Direct Antigen Presenatation and Gap Junction Mediated Cross- Presentation During Apoptosis

Manuscript in preparation.

(3)

(4)

Summary

MHC class I molecules present peptides derived from intracellular proteins that are degraded in cells. During cross-presentation, peptides derived from other infected or diseased cells are presented by professional antigen presenting cells (APCs). This process is crucial for the initiation of a proper immune response to pathogens, since APCs express the required co-stimulatory molecules to activate the immune system. Apoptotic cells and apoptotic bodies have been shown to be an important source for cross-presented antigens. How these antigens enter the MHC class I route of the APC is unknown. We hypothesized that apoptotic cells are able to generate and present antigens, and that these antigens can transfer to neighboring cells via gap junctions. An inducible apoptosis system and constructs that release peptides in a caspase-dependent manner were used to test this hypothesis. We show that during apoptosis novel epitopes are generated that can be presented directly by the apoptotic cell itself or by its healthy neighbor for cross- presentation.

Introduction

Intracellular proteins are continuously degraded in the cell, yielding an excessive pool of peptides. The majority of these peptides, normally in the range of 3 to 22 amino acids in length¹, are degraded within seconds by cytosolic peptidases^2,3. Only a small part, probably less than 1%, escapes degradation by entering the endoplasmic reticulum (ER). Since peptides cannot transfer membranes by simple diffusion, transporters are required. To enter the ER, peptides have to use the transporter as- sociated with antigen presentation (TAP) for translocation. There, peptides can bind MHC class I molecules, provided that they have the correct size and sequence.

Otherwise they return to the cytosol for further trimming and degradation. We showed that peptides can have a different fate and enter neighboring cells after diffusion through gap junction channels⁴. These channels allow the transfer of information in the form of ions, metabo- lites, small molecules and peptides between cells. A gap junction is an assembly of six connexin molecules, comprising a hemi channel called connexon that forms a functional channel when it is connected to a hemi-chan-

nel present in the mebrane of an adjacent cell. Transfer of peptides between cells can result in innocent bystand- er recognition, when viral peptides from an infected cell can transfer to non-infected neighbors. We showed that intercellular peptide transfer could also facilitate cross- presentation when professional antigen presenting cells (APC) receive antigenic information through gap junctions⁵. Cross-presentation is crucial for a proper immune response since APCs express the co-stimulatory molecules that are required for activation of CD8+ cytotoxic T cells (CTL). Various routes for cross-presentation have been proposed (reviewed in ⁶), and especially apoptotic cells or apoptotic bodies have been shown to be an efﬁcient source for cross-presented antigens^7,8. It is however still unknown how extracellular apoptotic material ends up in the MHC class I presentation pathway, or simply the cytosol. There is a topological problem when apoptotic material is phagocytosed by APCs.

Apoptotic bodies are surrounded by an intact membrane and upon phagocytosis the antigen is thus surrounded by a double lipid bilayer. There is little evidence that the membrane of apoptotic bodies disintegrates in the phagosome. In order to reach the cytosol, antigens have to pass this double barrier. A partial solution for this problem could be the involvement of saposins. These lipid transfer proteins play a role in the disintegration of intralysosomal vesicles. As shown recently, these proteins are also involved in the cross-presentation of apoptotic material. Cross-presentation of antigens from apoptotic bodies was greatly enhanced after reconstitu- tion of saposins in prosaposin knock out mice⁹. Whether this is a bona ﬁde route for all extracellular material is unknown since apoptotic cells ﬁrst should be converted into apoptotic bodies for this pathway. It was shown before that apoptotic cells are coupled to healthy surrounding cells until very late stages of apoptosis¹⁰. Fur- thermore, death signals, both stimulatory and inhibitory, have been reported to be exchanged between dying and healthy cells¹¹.

We therefore hypothesize that APCs might be connected to (early) apoptotic cells via gap junctions, or make even new connections with dying cells. This would solve the topological problem because the cytosol of the dying cell and the APC are coupled, thereby facilitating transfer of peptide. This would explain how an intracellular antigenic peptide from apoptotic source can enter

Direct Antigen Presentation and Gap Junction Mediated Cross-Presentation During Apoptosis.

Joost Neijssen, Baoxu Pang, Lennert Janssen, Christoph Lippuner and Jacques Neefjes.

Division of Tumor Biology, The Netherlands Cancer Institute, Amsterdam, The Netherlands.

93

(5)

94

Intercellular Peptide Transfer through Gap Junctions

the MHC class I presentation route of the APC. To test this hypothesis we used an inducible apoptosis system¹² and constructs that release peptides in a caspase dependent manner. Here we show that during apoptosis novel epitopes can be generated that can be presented directly by the apoptotic cell or by its healthy neighbor for cross- presentation.

Results

Requirements for presentation of antigen from apop- totic cells

Apoptotic cells can only present peptides in the MHC class I route if some basic criteria are met. The proteasome and cytosolic peptidases must be still active in apoptotic cells in order to degrade antigens in to peptides that can be presented by MHC class I molecules. In ad- dition, ATP is required for degradation, peptide translocation and MHC class I transport and should be present during the initiation and onset of apoptosis. To test these requirements, we used cells in which apoptosis could be induced in a synchronized manner by adding a small drug to the medium. Specific induction of apoptosis was achieved by expressing FKBD-Casp9 IRES GFP (F- Casp9) and adding the dimeriser AP20187(dimeriser) to the culture medium, as described before by Straathof and colleagues¹². Apoptotic cell blebbing could be observed already after 3 hours of exposure to the dimeriser, but only in cells expressing F-Casp9 (Fig 1A). The drug did not affect growth or viability of cells lacking the F-Casp9 construct that were present in the same culture. Cells with the caspase construct were co-cultured with the parental cell line. All cells were stained with MitoTracker Red as a marker for healthy cells and the cells with the F-Casp9 could be distinguished easily by the expression of GFP. Specific induction of apoptosis was observed only in the cells expressing GFP after exposure to the dimeriser (Figure 1A). Apoptosis was con- firmed by staining cells with AnnexinV at different time points after induction of apoptosis and analyzed by flow cytometry. The percentage of AnnexinV positive cells increased in time showing that apoptosis was effectively induced. This preceded cell leakage since propidium iodide was not able to enter the cells, even when PARP was completely digested. Apoptosis was shown biochemically by generating lysates at different time points after drug exposure. The lysates were analyzed by SDS- PAGE and Western blotting with antibodies against PARP. PARP is cleaved by activated caspase3 during the apoptosis process. Cleavage of PARP was observed 2 hours after induction of apoptosis. Apoptosis could be effectively inhibited by adding Z-VAD-FMK (Z-VAD), a pan-caspase inhibitor (Figure 1C). Protein synthesis was measured by pulsing cells with ³⁵S-methionine at

different time points after induction of apoptosis. Incor- poration of the label was monitored by analyzing the samples by SDS-PAGE and autoradiographs. Radioac- tive amino acids were incorporated in proteins for at least 3 hours after induction of apoptosis, indicating that protein synthesis continues after induction of apoptosis (Figure 1D). The peptidase activity in apoptotic cells was measured in in vitro peptidase assays. Lysates from cells exposed to drugs for different time points were added to internally quenched fluorescent peptides and appearance of fluorescence was recorded. Upon degradation of the peptides, the quencher and the fluoro- phore are spatially separated and fluorescence can be observed. No differences in peptidase activity were observed, for both 9-mer (figure 1E) and 15-mer peptides (data not shown). To test whether ATP is still present in apoptotic cells, the A431 F-Casp9 cells were transiently transfected with constructs for luciferase and renilla.

Apoptosis was induced and at different time points cell lysates were taken and a luciferase/renilla assay was performed. Luciferase signal could be observed in the samples, indicating that ATP was still present for more than 3 hours after induction of apoptosis (Figure 1F).

(6)

Gap junctional coupling of apoptotic and healthy cells To test whether apoptotic cells are still coupled via gap junctions we co-cultured A431/Cx43 F-Casp9 with A431/Cx43 transiently transfected with H2B-mRFP for detection. After induction of apoptosis the cells were ﬁxed and stained with antibodies against Cx43, and analyzed using CLSM. Cx43 was detected between healthy and dying cells, indicating that connexin molecules are still present at the interphase of healthy and dying cells (Figure 2A).

To show that apoptotic cells are still coupled via gap junctions, apoptosis was induced in A431/Cx43 F- Casp9 cells after labeling with calcein AM and a FRAP experiment was performed. One cell of a small group of cells was photobleached to destroy the fluorescence in that cell completely and fluorescence recovery as a result of diffusion of dye from surrounding cells was monitored in time. Redistribution of the fluorescent signal over the group of apoptotic cells was observed, showing that apoptotic cells are still coupled via functional gap junctions (Figure 2B). Transfer of fluorescent dye between healthy and apoptotic cells was monitored in an analogue experiment. A431/Cx43 F-Casp9 were co-cultured with A431/Cx43 H2B-mRFP. Upon induction of apoptosis by adding the homo-dimeriser, apoptotic cells could be distinguished morphologically but also due to exclusion of H2B-mRFP. Dye transfer was observed between photobleached apoptotic cells and healthy cells indicating that healthy and apoptotic cells Figure 1. Requirements for presentation of antigen from apoptotic cells.

A. The inducible apoptosis system was tested by expressing F-Casp9 IRES GFP in A431 cells. These cells were co-cultured with normal A431 cells and stained with MitoTracker Red. The cells were analyzed by CLSM at t=0h and 4h of exposure to the dimeriser.

Only the transfected (GFP+) cells undergo apoptosis after exposure to the dimeriser. B. Apoptosis was monitored by ﬂow cytometry.

Cells were stained with AnnexinV and propidium iodide (PI) at different time points after induction of apoptosis. The percentage of cells positive for apoptosis (AnnexinV) increased in time whereas the percentage of necrotic cells (PI) did not change. C. Apoptosis was shown biochemically by generating lysates at different time points of drug exposure. The lysates were analyzed by SDS-PAGE and Western blotting with antibodies against PARP. Cleavage of PARP, a substrate of active caspase 3, can be observed 2 hours after induction of apoptosis and apoptosis could be effectively inhibited by adding Z-VAD. The membranes were stained with antibodies against GFP as loading control. D. To test whether translation continues during apoptosis cells were pulsed with 35S-Methionine for 30 minutes and the lysates of both floating and adherent cells were analyzed by SDS-PAGE and autoradiographs. Incorporation of the radioactive amino acids continues to more than 3 hours after induction of apoptosis. E. The peptidase activity in apoptotic cells was tested in an in vitro peptidase assay. Cell lysates were added to quenched fluorescent peptides and fluorescence was recorded.

F. The presence of ATP in apoptotic cells was monitored by expressing luciferase and renilla in A431 F-Casp9 cells. The cells were lysed at different time points after apoptosis induction and substrate conversion was measured by measuring luminescence.

Figure 2. Gap junctional coupling of apoptotic and healthy cells.

A. To visualize gap junctions on the interphase between apoptotic cells and healthy cells, A431 F-Casp9 cells were co-cultured with A431 H2B-mRFP cells and apoptosis was induced. After ﬁxation the slides were stained with antibodies against Cx43 and imaged by CLSM. B. Gap junctional coupling of apop- totic cells was shown by FRAP experiments. A431 cells treated with Edelfosin (ALP) were loaded with calcein AM and all ﬂuorescence was destroyed by photobleaching in the left cell. The recovery of the dye in the left cell was monitored (red trace). C. Gap junctional coupling of apoptotic and healthy cells was determined by FRAP. A431 F-Casp9 cells were co-cultured with A431 H2B-mRFP cells and apoptosis was induced. FRAP analysis revealed that dye transfer to the bleached apoptotic cell was still possible.

95

(7)

can still be coupled via gap junctions (Fig 2C).

Direct antigen presentation of apoptotic antigens When apoptotic cells still contain the machinery for antigen generation, translocation and presentation, in principle antigen should be presented following apoptosis. To test this, we required an antigen selectively generated after apoptosis. A variant of the GFP-Ub-peptide construct as used before⁴ was generated. The peptide is removed from the chimeric protein by ubiquitin hydrolase activity that cleaves after the ultimate glycine (G₇₆) of the ubiquitin moiety. By replacing G₇₆ for a DEVD sequence the fusion protein is stable under normal conditions. Since DEVD is the cleavage sequence of activated caspase 3 and 7, the peptide should be released during apoptosis (Figure 3A). To test this, we transfected Hela-Kb cells with OVA_257-264 (GFP-Ub- DEVD-SL8). Apoptosis was induced by adding 10 μM Edelfosin (ALP) to the culture medium¹³. Biochemical analysis using Western blot showed that this treatment induced cleavage of PARP, a bona fide marker for apoptosis. To show that the fusion protein was expressed in these cells an antibody against GFP was used. To reveal the 1 kD difference between the intact chimera and the cleaved form, the lysates were separated by 15% SDS PAGE. A lower shadow band just under the normal GFP band appeared only in the apoptotic sample, indicating partial cleavage upon induction of apoptosis (Figure 3B). To test whether this released epitope could be presented by in the MHC class I presentation route, a CTL assay was performed. The HeLa cells expressed the proper mouse MHC class I molecules H2-Kb (restrictive element), and B3Z cells were added to the cultures to measure peptide presentation. B3Z cells will express LacZ upon activation by binding to specific MHC class I-peptide complexes. The expression of LacZ (i.e. T-cell activation) was increased significantly (p=0.012) after induction of apoptosis, as measured in an enzymatic assay (Figure 3C), suggesting that induction of apoptosis resulted in the release and presentation of the peptide.

To show presentation of apoptosis derived peptides in cells in which apoptosis could be induced, A431 F- Casp9 cells were transfected with a red mRFP variant of the apoptosis-speciﬁc construct (mRFP-Ub-DEVD- SL8) and the restrictive H2-Kb element (Figure 3D).

Western blot analysis showed release of the epitope. The SIINFEKL signal disappeared upon induction of apoptosis, in contrast to the mRFP signal that did not change in time (Figure 3E). This process could be blocked by adding Z-VAD during the experiment. Pre sentation of SL8 was measured in a B3Z assay. Presentation of the released peptide could be observed only when apoptosis was induced by adding the dimeriser and this could be blocked by adding ZVAD (Figure 3F).

Intercellular transfer of apoptosis speciﬁc peptide To test whether the released peptide was able to diffuse into and presented by neighboring cells, a co-culture experiment was performed. A431/Cx43 F-Casp9 cells were transiently transfected with mRFP-Ub-DEVD- SL8 and co-cultured with B16-Cx43 cells that express H2-Kb molecules (Figure 4A). Since the A431 cells lack the correct MHC class I molecules, presentation of the epitope could only occur after transfer to B16 cells.

The dependence on gap junctional transfer was determined using A431 with or without the expression of Cx43 as the peptide donor to apoptotic cells. Transfer of the released peptide was assayed in a B3Z assay. T cell activation could be observed only when antigen is expressed, apoptosis induced and when gap junctions are present (Figure 4B). This shows that epitopes, speciﬁ- cally generated in apoptotic cells, can be transferred to healthy neighboring cells via gap junctions.

Discussion

In this study we show that peptides produced in apoptotic cells can be presented by the apoptotic cell and be transferred to healthy surrounding cells via gap junctions for cross-presentation. This novel route of antigen presentation solves a topological problem for peptides derived from an apoptotic source. Apoptotic cells or material are possibly important sources for cross-presentation as shown in multiple reports^7,8. How these antigens enter the MHC class I route of the APC is unknown.

The main problem for an extracellular antigen is that it cannot traverse membranes without the help of a transporter. As apoptotic cells and bodies retain membrane integrity, the antigen is surrounded by an intact lipid bilayer. In contrast to necrotic cells the antigens are therefore not released in the extracellular space. When an apoptotic cell or body is phagocytosed by an APC, an additional phagocytic membrane is surrounding these structures. Whether the membrane of apoptotic bodies in the phagosome remains intact is not known, but there is no evidence for lipase activity in phagosomes. The antigen has to traverse two lipid bilayers to reach the cytosol, where it can enter the MHC class I antigen presentation route. It has been shown that cross-presented antigens enter the cytosol where it must be degraded by the proteasome (proteins) or bind TAP (peptides). Here we propose a solution for this topological problem.

Apoptotic cells are coupled until very late stages via gap junctions¹⁰. As shown in ﬁgure 1 apoptotic cells still have Cx43 at the contact surface with healthy cells, indicating that gap junctions are indeed present. These channels are functional, as shown in

96

(8)

Figure 3. Direct antigen presentation of apoptotic antigens.

A. In order to have an antigen selectively generated after apoptosis, GFP-Ub-DEVD-SL8 was made. The last glycine of ubiquitin was replaced by DEVD, the cleavage sequence of activated caspase 3 and 7. The peptide should be released in the cytosol during apoptosis. B. Release of the peptide after induction of apoptosis was tested biochemically. Cells were treated with ALP and lysates of different time points after induction were taken and analyzed by SDS-PAGE and Western Blot. The cleaved product comprising of GFP-Ub could be observed in the GFP panel as a lower shadow band. C. To test whether the released peptide can also be presented in the MHC class I route, HeLa cells expressing H2-Kb were transfected with GFP-Ub-DEVD-SL8. Apoptosis was induced and the SL8 speciﬁc hybridoma cell line B3Z was added as a read out for presentation. D. Release of peptide in a caspase dependent man- ner was also tested in A41 cells transfected with an inducible apoptosis construct. The GFP in the DEVD construct was replaced by mRFP, because the apoptosis construct already expressed GFP. E. Release of the peptide after induction of apoptosis was tested biochemically. Cell lysates were taken at different time points of exposure to the dimeriser and analyzed by SDS-PAGE and Western blotting. The membranes were stained with antibodies against SIINFEKL and this signal disappeared in time after exposure to the dimeriser, indicating that the peptide was released. F. Direct presentation of the released peptide was shown by expressing mRFP- Ub-DEVD-SL8 and H2-Kb in the A431 F-Casp9 cells. Presentation was monitored by measuring the activation of peptide speciﬁc

Figure 4. Intercellular transfer of apoptosis speciﬁc peptide.

A. Peptide transfer from an apoptotic cell was tested by co-cul- turing A431+/-Cx43 stably transfected with FKBD-Casp9, and transiently transfected with mRFP-Ub-DEVD-SIINFEKL with B16-Cx43 cells that express H2-Kb molecules. B. Presentation of the released peptide was shown by adding B3Z to the de- scribed above. T cell activation was observed only during apop- tosis when mRFP-Ub-DEVD-SL8 and Cx43 was present in the A431 F-Casp9 cells.

97

(9)

ﬁgure 1B, where dye from neighboring cells can still be transferred to apoptotic cells after photobleaching.

To facilitate in vivo cross-presentation it is not enough to have gap junctional communication between healthy and apoptotic cells. First of all, proteins still need to be degraded by the proteasome. For degradation to occur proteins ﬁrst need to be unfolded by the 19S cap of the proteasome and for this process energy in the form of ATP is required. It has been previously reported that ATP levels drop somewhat during apoptosis but no to zero. We also show that upon induction of apoptosis luciferase activity is still observed in cells transfected with a luciferase construct (Figure 1F). It has been reported that proteasomal activity is severely compromised in apoptotic cells as a result of cleavage of 3 subunits of the 19S cap by caspase 3¹⁴. Menendez-Benito and colleagues report however, that proteasomal activity is not affected in cells treated with apoptosis inducing drugs by using a Ub-R-YFP reporter construct¹⁵. Secondly, protein synthesis needs to be continued during apoptosis to produce MHC class I molecules and DriPs. That the machinery for antigen generation, transfer and presentation is active at the early phase of apoptosis is best il- lustrated by our ﬁnding that apoptotic cells still directly present antigens.

This novel pathway of cross-presentation may explain how apoptotic antigens are being cross-presented al- though it obviously does not exclude other mechanisms.

We show that a protein is cleaved by active caspase3, which is only active in apoptotic cells. This might yield unique peptides and therefore a mechanism against pathogens that induce apoptosis of cells. These pathogens are recognized by direct presentation or cross- presentation during early phases of apoptosis. Of note, some tumor cells express basal levels of caspase activi- ties¹⁶ and these cells thus have cytosolic proteases other than the proteasome for the generation of antigenic fragments. Unique tumor-specific epitopes could then be generated, not because the antigens are tumor-specific but because of the degradation pattern specific for these tumors.

Materials and methods

Cells, antibodies, peptides and reagents

A431 cells were stably transfected with human Cx43 cDNA in pcDNA3 (kindly provided by Dr. B. Giep- mans) as described before4. HeLa cells stably transfected with mouse H2-Kb MHC class I molecules (HeLa- Kb) were kindly provided by Dr. K. Rock. Mouse B16 cells were stably transfected with human Cx43 cDNA.

The GFP/mRFP-Ub-DEVD-OVA_257-264 construct was made by PCR on ubiquitin using a reverse primer with a ﬂanking region containing the sequence encoding the

caspase3 cleavage site DEVD and the OVA peptide, thereby deleting the ubiquitin-splicing site following Gly76 of ubiquitin. The OVA_257-264 peptide sequence:

SIINFEKL. This product was cloned in a normal or mod- iﬁed pEGFP-C1 vector (Clontech, USA) as described in 17. A431 and A431/Cx43 were transduced with the inducible apoptosis construct FKBD-Casp9 IRES GFP as described¹². B3Z cells expressing the LacZ gene under the control of the interleukin-2 promoter were used to determine presentation of SIINFEKL in the context of H2-Kb. The following antibodies were used: rabbit anti- GFP¹⁸, mAb anti-SIINFEKL¹⁹, rabbit anti-PARP (Cell Signaling Technology, USA), rabbit anti-Cx43 (Sigma Aldrich, The Netherlands) and mAb anti-tubulin²⁰. HRP coupled secondary antibodies from Daco (Copenhagen, Denmark).

Requirements for presentation of antigen from apop- totic cells

A431/Cx43 F-Casp9 cells were co-cultured with A431/

Cx43 on glass coverslips. Staining of mitochondria was performed by incubation with 10 nM MitoTracker Red (Molecular Probes, Leiden, The Netherlands).

Apoptosis was induced by adding 0.1 nM of AP20187 (Ariad, USA) to the culture medium. The cells were imaged using confocal microscopy (Leica AOBS). To determine the percentage and speed of apoptosis A431/

Cx43 F-Casp9 were exposed to AP20187 for different periods and times. The cells were then harvested by trypsinisation and stained wit AnnexinV-APC and propidium iodide in 1x binding buffer as supplied in the AnnexinV-APC apoptosis detection kit I (BD, Bel- gium) for 15 minutes at room temperature. The cells were then analyzed by ﬂow cytometry. Apoptosis was checked biochemically by SDS-PAGE and Western blot analysis. Apoptosis was inhibited by adding 50μM Z-VAD-FMK to the culture medium during exposure to the dimeriser. Lysates were generated by adding reduc- ing sample buffer at different time points of exposure to the dimerising drug. The membranes were stained with anti-PARP and anti-GFP. Protein synthesis during apoptosis was measured by pulsing cells for 30 minutes with ³⁵S-methionine in cystein / methionine free medium at different time points after induction of apoptosis.

Then the adherent cells and centrifuged (10 minutes at 14000 rpm) supernatant were lysed with 1% NP40.

TCA precipitation was used to remove free radioactive nucleotides. The samples were analyzed by SDS-PAGE and autoradiographs. For in vitro peptidase assays, A431/Cx43 F-Casp9 cells were lysed in cold PBS at different timepoints after induction of apoptosis by using a Cell Cracker (EMBL, Germany). The cell lysates were added to 1 μl peptide (1 mg/ml) and appearance of ﬂuorescence was recorded using a FLUOstar OP- TIMA (BMG Labtechnologies, Germany) ﬂuorescent

98

(10)

plate reader. The presence of ATP in apoptotic was assayed by transient transfection of A431/Cx43 F-Casp9 with ﬁreﬂy luciferase and renilla luciferase constructs.

After 48 hours apoptosis was induced by adding dimeriser and the cells were lysed by adding 1x passive ly- sis buffer (Promega) and shake for 20 minutes at room temperature at 950 rpm on an Eppendorf Thermomixer Comfort (Eppendorf, Germany). A luciferase assay was performed using a Dual Luciferase Reporter Kit (Pro- mega, USA) and luminescence was measured using a FLUOstar OPTIMA (BMG Labtechnologies, Germany) ﬂuorescent plate reader.

Gap junctional coupling of apoptotic and healthy cells A431/Cx43 F-Casp9 cells were co-cultured with A431/

Cx43 transiently transfected with H2B-mRFP on glass coverslips. Apoptosis was induced by adding 0.1 nM of AP20187 (Ariad, USA) to the culture medium. Cells were fixed with paraformaldehyde and stained with antibodies against Cx43 and mounted on coverglass with Vectashield. A431/Cx43 F-Casp9 cells were loaded with 1 μg/ml calcein-AM (Molecular Probes, Leiden, The Netherlands) for 30’ at 37ôC. Apoptosis was induced by adding 0.1 nM of AP20187 to the culture medium. Gap junctional coupling was assayed by photobleaching one cell of a small group and recording fluorescence in the bleached cell (FRAP) using confocal microscopy (Leica AOBS and SPII). To show transfer between healthy and dying cells A431/Cx43 F-Casp9 cells were co-cultured with A431/Cx43 cells expressing H2B-mRFP and loaded with calcein-AM as described above. Upon induction of apoptosis, apoptotic cells with adjacent healthy H2B-mRFP expressing cells were photobleached and recovery of fluorescence was recorded.

Direct antigen presentation of apoptotic antigens HeLa-Kb cells were transiently transfected with GFP- Ub-DEVD-OVA257-264 by electroporation using a Biorad GenePulser Xcell. Apoptosis was induced by adding 10 μM Edelfosin (ALP) to the culture medium.

After washing away the ALP, cells were co-cultured with B3Z cells. Presentation of SIINFEKL was measured by B3Z activation that was assayed by enzymatic conversion of the β-D-galactopyranoside analog chlo- rophenol red-β-D-galactopyranoside (CPRG). Target cells were seeded 5*10e4 per 0.6 cm2 well and 1*10e5 B3Z cells were added. After o/n incubation at 37°C, 5%

CO2, CPRG (in 10mM phosphate buffer (pH 7.4) containing 1mM MgCl₂ and 0.125% NP-40) was added and OD was measured at 595 nm. Cleavage of the vector was shown biochemically by analyzing lysates of the cells by SDS-PAGE (15% gel) and subsequent West- ern blotting by probing the membrane with antibodies against GFP, and apoptosis was shown by probing with

the PARP antibody.

Similar experiments were performed in cells with an inducible apoptosis construct. A431/Cx43 F-Casp9 cells were stably transfected with H2-Kb. The cells were transfected transiently with mRFP-Ub-DEVD- SL8 as described above. Western blotting was performed as above and the membranes were stained with anti-tubulin as loading control, anti-SIINFEKL and anti-mRFP antibodies. To measure peptide presentation B3Z assays were performed as described above.

Intercellular peptide transfer and cross-presentation assays

A431 or A431/Cx43 F-Casp9 cells were transiently transfected with mRFP-Ub-DEVD-OVA_257-264 by electroporation using a Biorad Gene Pulser Xcell. Apoptosis was induced by adding 0.1 nM of AP20187 to the culture medium. Release of the epitope was checked biochemically. Cell lysates were obtained by using a Cellcracker (EMBL, Germany) and analysed by SDS-PAGE. After transfer of the proteins to nitrocellulose membrane, the integrity of the construct was determined by staining with antibodies against SIINFEKL. The pan caspase inhibitor F-ZVAD-FMK (ZVAD) (Sigma Aldrich, The Netherlands) was added to the medium to show that release of the epitope was dependent on caspase activation. Intercellular transfer of released peptide was assayed by transient expression of mRFP-Ub-DEVD- OVA_257-264. After over night incubation the cells were co- cultured with B16-Cx43 cells and the cells were allowed to adhere and grow to 70% conﬂuency. Apoptosis was induced by adding 0.1 nM of AP20187 to the culture medium. Subsequently, B3Z cells were added to assay transfer to and presentation by the B16 cells.

References

1. Kisselev, A. F. et al. The sizes of peptides generated from protein by mammalian 26 and 20 S proteasomes. Implications for understanding the degradative mechanism and antigen presentation. J Biol Chem 274, 3363-71 (1999).

2. Reits, E. et al. A major role for TPPII in trimming proteasomal degradation products for MHC class I antigen presentation. Im- munity 20, 495-506 (2004).

3. Reits, E. et al. Peptide diffusion, protection, and degradation in nuclear and cytoplasmic compartments before antigen presentation by MHC class I. Immunity 18, 97-108 (2003).

4. Neijssen, J. et al. Cross-presentation by intercellular peptide transfer through gap junctions. Nature 434, 83-8 (2005).

5. Neijssen, J. et al. Gap junction-mediated intercellular communication in the immune system. Prog Biophys Mol Biol 94, 207-18 (2007).

6. Groothuis, T. A. et al. MHC class I alleles and their explo- ration of the antigen-processing machinery. Immunol Rev 207, 60-76 (2005).

7. Blachere, N. E. et al. Apoptotic cells deliver processed anti-

99

(11)

gen to dendritic cells for cross-presentation. PLoS Biol 3, e185 (2005).

8. Albert, M. L. et al. Immature dendritic cells phagocytose apoptotic cells via alphavbeta5 and CD36, and cross-present antigens to cytotoxic T lymphocytes. J Exp Med 188, 1359-68 (1998).

9. Winau, F. et al. Apoptotic vesicles crossprime CD8 T cells and protect against tuberculosis. Immunity 24, 105-17 (2006).

10. Wilson, M. R. et al. Cell population dynamics (apoptosis, mitosis, and cell-cell communication) during disruption of ho- meostasis. Exp Cell Res 254, 257-68 (2000).

11. Cusato, K. et al. Gap junctions remain open during cyto- chrome c-induced cell death: relationship of conductance to ‘by- stander’ cell killing. Cell Death Differ 13, 1707-14 (2006).

12. Straathof, K. C. et al. An inducible caspase 9 safety switch for T-cell therapy. Blood 105, 4247-54 (2005).

13. Ruiter, G. A. et al. Anti-cancer alkyl-lysophospholipids in- hibit the phosphatidylinositol 3-kinase-Akt/PKB survival pathway. Anticancer Drugs 14, 167-73 (2003).

14. Sun, X. M. et al. Caspase activation inhibits proteasome function during apoptosis. Mol Cell 14, 81-93 (2004).

15. Menendez-Benito, V. et al. Endoplasmic reticulum stress compromises the ubiquitin-proteasome system. Hum Mol Genet 14, 2787-99 (2005).

16. Yang, L. et al. Coexistence of high levels of apoptotic signaling and inhibitor of apoptosis proteins in human tumor cells:

implication for cancer speciﬁc therapy. Cancer Res 63, 6815-24 (2003).

17. Johansson, M. et al. Activation of endosomal dynein motors by stepwise assembly of Rab7-RILP-p150Glued, ORP1L, and the receptor betalll spectrin. J Cell Biol 176, 459-71 (2007).

18. van Ham, S. M. et al. HLA-DO is a negative modulator of HLA-DM-mediated MHC class II peptide loading. Curr Biol 7, 950-7 (1997).

19. Hilton, C. J. et al. Anti-peptide antibody blocks peptide binding to MHC class I molecules in the endoplasmic reticulum.

J Immunol 166, 3952-6 (2001).

20. Kilmartin, J. V. et al. Rat monoclonal antitubulin antibodies derived by using a new nonsecreting rat cell line. J Cell Biol 93, 576-82 (1982).

100