Actin retrograde flow controls natural killer cell response by regulating the conformation state of SHP ‐1
2018; Springer Nature; Volume: 37; Issue: 5 Linguagem: Inglês
10.15252/embj.201696264
ISSN1460-2075
AutoresOmri Matalon, Aviad Ben‐Shmuel, Jessica Kivelevitz, Batel Sabag, Sophia Fried, Noah Joseph, Elad Noy, Guy Biber, Mira Barda‐Saad,
Tópico(s)CAR-T cell therapy research
ResumoArticle15 February 2018free access Source DataTransparent process Actin retrograde flow controls natural killer cell response by regulating the conformation state of SHP-1 Omri Matalon The Mina and Everard Goodman Faculty of Life Sciences, Bar-Ilan University, Ramat-Gan, Israel Search for more papers by this author Aviad Ben-Shmuel The Mina and Everard Goodman Faculty of Life Sciences, Bar-Ilan University, Ramat-Gan, Israel Search for more papers by this author Jessica Kivelevitz The Mina and Everard Goodman Faculty of Life Sciences, Bar-Ilan University, Ramat-Gan, Israel Search for more papers by this author Batel Sabag The Mina and Everard Goodman Faculty of Life Sciences, Bar-Ilan University, Ramat-Gan, Israel Search for more papers by this author Sophia Fried The Mina and Everard Goodman Faculty of Life Sciences, Bar-Ilan University, Ramat-Gan, Israel Search for more papers by this author Noah Joseph The Mina and Everard Goodman Faculty of Life Sciences, Bar-Ilan University, Ramat-Gan, Israel Search for more papers by this author Elad Noy The Mina and Everard Goodman Faculty of Life Sciences, Bar-Ilan University, Ramat-Gan, Israel Search for more papers by this author Guy Biber The Mina and Everard Goodman Faculty of Life Sciences, Bar-Ilan University, Ramat-Gan, Israel Search for more papers by this author Mira Barda-Saad Corresponding Author [email protected] orcid.org/0000-0002-0305-7438 The Mina and Everard Goodman Faculty of Life Sciences, Bar-Ilan University, Ramat-Gan, Israel Search for more papers by this author Omri Matalon The Mina and Everard Goodman Faculty of Life Sciences, Bar-Ilan University, Ramat-Gan, Israel Search for more papers by this author Aviad Ben-Shmuel The Mina and Everard Goodman Faculty of Life Sciences, Bar-Ilan University, Ramat-Gan, Israel Search for more papers by this author Jessica Kivelevitz The Mina and Everard Goodman Faculty of Life Sciences, Bar-Ilan University, Ramat-Gan, Israel Search for more papers by this author Batel Sabag The Mina and Everard Goodman Faculty of Life Sciences, Bar-Ilan University, Ramat-Gan, Israel Search for more papers by this author Sophia Fried The Mina and Everard Goodman Faculty of Life Sciences, Bar-Ilan University, Ramat-Gan, Israel Search for more papers by this author Noah Joseph The Mina and Everard Goodman Faculty of Life Sciences, Bar-Ilan University, Ramat-Gan, Israel Search for more papers by this author Elad Noy The Mina and Everard Goodman Faculty of Life Sciences, Bar-Ilan University, Ramat-Gan, Israel Search for more papers by this author Guy Biber The Mina and Everard Goodman Faculty of Life Sciences, Bar-Ilan University, Ramat-Gan, Israel Search for more papers by this author Mira Barda-Saad Corresponding Author [email protected] orcid.org/0000-0002-0305-7438 The Mina and Everard Goodman Faculty of Life Sciences, Bar-Ilan University, Ramat-Gan, Israel Search for more papers by this author Author Information Omri Matalon1,‡, Aviad Ben-Shmuel1,‡, Jessica Kivelevitz1, Batel Sabag1, Sophia Fried1, Noah Joseph1, Elad Noy1, Guy Biber1 and Mira Barda-Saad *,1 1The Mina and Everard Goodman Faculty of Life Sciences, Bar-Ilan University, Ramat-Gan, Israel ‡These authors contributed equally to this work *Corresponding author. Tel: +972 3 5317311; Fax: +972 3 7384058; E-mail: [email protected] EMBO J (2018)37:e96264https://doi.org/10.15252/embj.201696264 PDFDownload PDF of article text and main figures. Peer ReviewDownload a summary of the editorial decision process including editorial decision letters, reviewer comments and author responses to feedback. ToolsAdd to favoritesDownload CitationsTrack CitationsPermissions ShareFacebookTwitterLinked InMendeleyWechatReddit Figures & Info Abstract Natural killer (NK) cells are a powerful weapon against viral infections and tumor growth. Although the actin–myosin (actomyosin) cytoskeleton is crucial for a variety of cellular processes, the role of mechanotransduction, the conversion of actomyosin mechanical forces into signaling cascades, was never explored in NK cells. Here, we demonstrate that actomyosin retrograde flow (ARF) controls the immune response of primary human NK cells through a novel interaction between β-actin and the SH2-domain-containing protein tyrosine phosphatase-1 (SHP-1), converting its conformation state, and thereby regulating NK cell cytotoxicity. Our results identify ARF as a master regulator of the NK cell immune response. Since actin dynamics occur in multiple cellular processes, this mechanism might also regulate the activity of SHP-1 in additional cellular systems. Synopsis SHP-1 is crucial for suppressing natural killer (NK) cell activation. Actin retrograde flow regulates SHP-1 conformation and enzymatic activity upon inhibitory receptor engagement, thereby controlling NK cell mediated cytotoxicity towards cancer cells. The velocity of the actin retrograde flow is slower at the inhibitory versus the activating NK immunological synapse. Slower actin retrograde flow enables the interaction of SHP-1 with β-actin. Interaction with β-actin regulates SHP-1 conformation and induces its activation. Blocking actin dynamics during the NK inhibitory response results in "closed" SHP-1 conformation and reduced SHP-1 activity. The reduced enzymatic activity of SHP-1 following actin retrograde flow arrest results in SHP-1 substrate activation and NK cell cytotoxicity. Introduction Natural killer (NK) cells are lymphocytes of the innate immune system, crucial for killing transformed and virally infected cells. The balance between the signals initiated upon the engagement of a variety of activating and inhibitory receptors determines whether an NK cell will kill its target or remain tolerant (Vivier et al, 2004). Although the mechanisms regulating NK cell activation are crucial for controlling multiple processes including tumor growth, viral infections, autoimmunity, and graft-versus-host disease, the global forces dictating NK cell signaling and regulating their functional outcome are poorly understood. The actin cytoskeleton is crucial for multiple cellular processes essential for an optimal immune response. The diversity and flexibility of lymphocyte functions are facilitated by the formation of various actin structures that provide the mechanical forces necessary for lytic granule secretion, motility, adhesion, and tissue invasion. These processes depend on the rapid assembly of filamentous actin (F-actin; Pollard & Borisy, 2003; Chhabra & Higgs, 2007; Matalon et al, 2013). The actin network also provides the structural basis for the formation of the immunological synapse (IS), namely the lymphocyte–target cell interaction site, and for the integration of molecular complexes and signaling effectors (Orange, 2008; Barda-Saad et al, 2010; Reicher & Barda-Saad, 2010, 2011; Pauker et al, 2011; Reicher et al, 2012; Matalon et al, 2013; Ritter et al, 2013; Fried et al, 2014; Joseph et al, 2014). In addition, physical forces generated by the actomyosin network are responsible for mechanotransduction, the conversion of mechanical forces into biochemical signals. In this process, the "pushing" force generated by actin polymerization and the "pulling" force of myosin are translated into signaling cascades (Vogel & Sheetz, 2006; Jaalouk & Lammerding, 2009). Increasing evidence strongly indicates that a dynamic actin network, rather than a static one, is crucial for regulating cellular responses (Babich et al, 2012; Yi et al, 2012; Comrie et al, 2015). Nevertheless, the actin cytoskeleton in NK cells was regarded merely as a static platform/scaffold, and studies mainly focused on the activating and inhibitory mechanisms that control actin rearrangement; however, little is known regarding whether and how the actin machinery controls NK cell signaling and function. Moreover, the role of actin movement and its spatial–temporal dynamics in eliciting NK cell effector function was never explored. Upon interaction of NK cells with a potential target, engagement of NK activation receptors with ligands on the target cell results in the activation of protein tyrosine kinase (PTK)-dependent signaling pathways. Inhibition of NK cells is controlled by ligation of the inhibitory receptors, including the killer cell immunoglobulin-like receptor (KIR) and the CD94-NKG2A receptor, to self-major histocompatibility complex (MHC) class I molecules (Vivier et al, 2004). This engagement antagonizes activating pathways by recruiting the protein tyrosine phosphatase (PTP), Src homology region 2 domain-containing phosphatase-1 (SHP-1) to the NK/target interaction site, the NK immunological synapse (NKIS; Lanier, 2008; Long, 2008). SHP-1 dephosphorylates key signaling molecules, such as VAV1, thereby blocking NK cell activation (Stebbins et al, 2003). We recently revealed that the linker for activation of T cells (LAT), phospholipase Cγ1 (PLCγ1), and PLCγ2 serve as additional SHP-1 substrates during NK cell inhibition. We showed that SHP-1 dephosphorylation of these proteins inhibits NK cell activation and cancer cell killing (Matalon et al, 2016). Here, we demonstrate that the velocity of the actomyosin retrograde flow (ARF) is significantly slower at the inhibitory versus the activating NKIS, enabling the interaction of SHP-1 with β-actin. Following this interaction, actin dynamics govern the conformational structure of SHP-1, dictating its catalytic activity. Indeed, blocking actin dynamics during the inhibitory response results in reduced SHP-1 activity, by confining SHP-1 to its inactivated "closed" conformation. This reduced enzymatic activity of SHP-1 leads to increased phosphorylation of SHP-1 substrates, an elevation of intracellular calcium flux, and NK cell cytotoxicity. Our data suggest that SHP-1 plays a major role as a sensor of ARF-generated forces in the process of mechanotransduction, and reveal a novel mechanism by which regulation of SHP-1 by ARF dictates NK cell killing decisions. Results SHP-1 binds to β-actin during the NK cell inhibitory response To characterize the role of the actomyosin network in regulating NK cell signaling, we examined the interaction of the PTP SHP-1 with signaling molecules during inhibitory vs. activating interactions. YTS NK cells expressing the inhibitory receptor KIR2DL1 (YTS-2DL1 cells, Fig EV1A) were incubated with 721.221 target cells expressing the inhibitory HLA-Cw4 haplotype (221-Cw4 cells), or the irrelevant HLA-Cw7 haplotype (221-Cw7 cells), resulting in NK cell activation. The association of SHP-1 with tyrosine phosphorylated signaling proteins was determined by immunoprecipitation (IP) of SHP-1, revealing a prominent 42-kDa band upon inhibitory interaction (Fig 1A, indicated by arrow). Mass-spectrometric analysis identified this band as β-actin (Fig EV1B), suggesting that SHP-1 interacts with the actin machinery during the inhibitory NK cell response. The mass spectrometry data were confirmed by blotting the membrane with an antibody specific to β-actin (Fig 1A) and were further validated by identifying an SHP-1:β-actin complex in KIR2DL1-expressing isolated primary NK cells incubated with inhibitory 221-Cw4 target cells vs. 721.221 activating target cells (pNK-2DL1 cells, Fig 1B; for NK cell purification, Fig EV1C). To eliminate the possibility of unspecific binding of β-actin to A/G agarose beads, we either incubated cell lysates with the A/G agarose beads without antibody (Fig 1A and B; No Ab), or with beads attached to an irrelevant IgG isotype control antibody (Fig EV1D), demonstrating the lack of co-immunoprecipitation in the control samples. Moreover, SHP-1 was immunoprecipitated from cell lysates of YTS cells alone or target cells alone, demonstrating the lack of binding of SHP-1 to β-actin in the control samples (Fig EV1E). These data suggest that the SHP-1:β-actin complex is formed following NK cell inhibition, rather than disassociating following activation. Click here to expand this figure. Figure EV1. Analysis of SHP-1:β-actin complex formation following NK cell inhibition YTS-2DL1 cells were stained with anti-KIR2DL1 antibody and a secondary Alexa-488 antibody. The expression level of the KIR2DL1 receptor was determined by FACS. YTS-2DL1 cells were incubated with 221-Cw4 target cells for 5 min at 37°C. The cells were lysed, and immunoprecipitates (IP) of SHP-1 were resolved by SDS–PAGE and stained with Coomassie blue. A band of ˜42 kDa was subjected to analysis by mass spectrometry, as described in the Appendix Supplementary Materials and Methods. Coverage of the identified β-actin protein following trypsin digestion is shown. The covered area is indicated in red (70% coverage). Isolated primary KIR2DL1+ NK cells were stained with anti-KIR2DL1 antibody, and a secondary Alexa-488 antibody. The expression level of the KIR2DL1 receptor was determined by FACS. Primary NK-KIR2DL1 cells were incubated with 221-Cw4 or 721.221 cells for 5 min at 37°C, and SHP-1 IP was subjected to IB using anti-β-actin or anti-SHP-1. As a negative control, IgG isotype antibody was used. YTS-2DL1 cells were incubated with 221-Cw4 or Cw7 cells for 5 min at 37°C. Cells were lysed, and the SHP-1 IP was immunoblotted using anti-β-actin. SHP-1 was also immunoprecipitated from lysates of YTS-2DL1, 221-Cw4, or 221-Cw7 target cells alone as negative controls. YTS CFP-actin cells transiently expressing YFP-SHP-1 were incubated on slides pre-seeded with mCherry-expressing 221-Cw4 or 221-Cw7 target cells for 5 min at 37°C. The cells were fixed, and FRET analysis was then performed. Graph summarizing the mean percentage FRET efficiencies ± SEM (n = 43 cells for Cw4, n = 37 cells for Cw7). Scale bars indicate 5 μm. White arrows indicate the NKIS site formed between YTS and target cells. Statistical significance was calculated with Student's t-tests used for unpaired, two-tailed samples. MST analysis for the binding of YFP-SHP-1 and actin-derived peptides was performed as described in Appendix Supplementary Materials and Methods. Lysates of 293T cells expressing YFP-SHP-1 were incubated with serially diluted (400 μM–48 nM) peptides, including WT actin peptide (KEKLCYVALDF), mutant actin peptide (KEKLCFVALDF), and irrelevant control peptide (EYQKASGVSG). Binding curves were generated by the NanoTemper analysis software (MO.Affinity Analysis v2.1.3). Graph summarizing the changes of the fluorescent thermophoresis signals as a function of peptide concentration from at least three independent experiments. Data information: Data are means ± SEM. Statistical significances were calculated with Student's t-tests used for unpaired, two-tailed samples. Source data are available online for this figure. Download figure Download PowerPoint Figure 1. SHP-1 binds β-actin at the inhibitory NKIS YTS-2DL1 cells were incubated with 721.221-Cw4 or Cw7 (221-Cw4 or Cw7) cells for 5 min at 37°C. The cells were lysed, and immunoprecipitates (IP) of SHP-1 were immunoblotted (IB) with anti-pTy (top panel), anti-β-actin (middle panel), or anti-SHP-1 (bottom panel) antibodies. Primary NK-2DL1 cells were incubated with 221-Cw4 or 721.221 cells for 5 min at 37°C, and SHP-1 IPs were subjected to IB using anti-β-actin or anti-SHP-1. YTS-2DL1 cells were incubated at 37°C with 221-Cw4 or Cw7 cells for 5 min, 1 μM of JAS was added, and the cells were incubated for an additional 5 min. The cells were lysed, and IPs of SHP-1 were subjected to IB using anti-β-actin or anti-SHP-1. Data information: Results shown are representative of at least three independent experiments. Source data are available online for this figure. Source Data for Figure 1 [embj201696264-sup-0014-SDataFig1.pdf] Download figure Download PowerPoint To further support our findings of SHP-1–β-actin interaction following NK cell inhibition, we utilized two additional experimental approaches. The first was analysis of fluorescence resonance energy transfer (FRET) efficiency between SHP-1 and β-actin to measure their complex formation on the nanometer scale. YTS-2DL1 cells stably expressing CFP-actin (YTS CFP-actin cells) and transfected with YFP-SHP-1 were examined following interaction with inhibitory 221-Cw4 or activating 221-Cw7 target cells expressing mCherry. FRET analysis demonstrated a significantly higher binding of SHP-1 to β-actin at the inhibitory vs. activating NKIS (18.9 ± 3.6% vs. 8.5 ± 2.7%; P = 0.02; Fig EV1F), supporting the biochemical data. The SHP-1–β-actin complex formation was also examined using microscale thermophoresis (MST) technology. We measured the binding of SHP-1 to a WT β-actin-derived peptide that contains an ITIM (KEKLCYVALDF). The binding of SHP-1 to a single aa mutant form of the β-actin peptide, containing a tyrosine to phenylalanine substitution (Y-F mutant), or irrelevant control peptide was also determined. Lysates of HEK 293T cells transiently expressing YFP-SHP-1 were incubated with decreasing concentrations of the different peptides followed by MST analysis. Strikingly, the WT actin peptide bound SHP-1 with a dissociation constant (KD) of 39.5 ± 6.2 μM, whereas no binding was detected with the mutant form or irrelevant peptides (Fig EV1G). All together, these data indicate a direct interaction of SHP-1 with β-actin, specifically following NK cell inhibition, suggesting a possible role of the actin network in inhibitory signaling cascades. To elucidate the role of actin dynamics in SHP-1 signaling, a pharmacological inhibitor was used. Jasplakinolide (JAS) inhibits actin turnover by blocking F-actin depolymerization and depleting the pool of cellular G-actin, thereby stabilizing actin filaments (Cramer, 1999). JAS was reported to inhibit actin dynamics, and specifically, actin flow in several systems (Babich et al, 2012; Yi et al, 2012). The effect of JAS on the SHP-1:β-actin complex was determined following NK cell interaction with inhibitory or activating target cells. Strikingly, JAS treatment substantially increased SHP-1 binding to β-actin following inhibitory interactions, whereas no binding was detected following activation, regardless of JAS treatment (Fig 1C). Thus, we used JAS to determine the effect of actin dynamics on SHP-1 signaling and activity. Actin dynamics control SHP-1 movement at the inhibitory vs. activating NKIS The roles of actin movement and its spatial–temporal dynamics in governing the NK inhibitory versus activating signaling and response were never explored. To address these issues, we first assessed the profile of F-actin and myosin IIA distribution at the inhibitory versus activating contact site, to distinguish between the well-defined actin cytoskeletal regions, the lamellipodium (LP), the lamellum (LM), and the cell body (CB; Babich et al, 2012; Yi et al, 2012). To this end, the F-actin probe, F-tractin GFP (Fig EV2A and B), was utilized. F-tractin was shown to be the ideal reporter for visualizing F-actin organization and dynamics, as it neither affects the depolymerization rate of actin filaments nor interferes with the formation of the different F-actin structures (Johnson & Schell, 2009; Yi et al, 2012). As previously described (Orange et al, 2003), F-actin accumulated at the periphery of the activating NKIS (the LP), which is a preferred site for actin dynamics and retrograde flow (Bunnell et al, 2001; Kaizuka et al, 2007; Babich et al, 2012), whereas F-actin demonstrated a disperse distribution at the inhibitory NKIS (Fig EV2C and D; P ≤ 0.00001). Myosin IIA, however, was depleted from the LP and accumulated at the LM and the CB, which are located behind the LP, in both the inhibitory and activating systems (Fig EV2C). These results define the regions of actin and myosin accumulation, demonstrating differential organization of F-actin in the inhibitory versus activating settings. Click here to expand this figure. Figure EV2. Differential F-actin and SHP-1 dynamics and distribution at the activating versus inhibitory NKIS YTS-2DL1 cells were transfected with plasmids encoding F-tractin-GFP or GFP alone, and cell lysates were prepared after 24 h. F-tractin GFP protein level and size were analyzed by IB with anti-GFP. YTS-2DL1 cells were transfected with F-tractin-GFP, and cells stably expressing the protein were generated (YTS F-tractin GFP cells). F-tractin GFP expression level was determined by FACS. YTS F-tractin GFP cells transiently expressing mCherry-Myosin IIA were seeded over coverslips pre-coated with activating anti-CD28 (top panel) or inhibitory anti-KIR2DL1 antibody (bottom panel), and allowed to spread for 4 min at 37°C before fixation. F-actin and myosin IIA distributions from multiple cells were profiled by ImageJ, as described in the Materials and Methods, and the average normalized fold intensities of F-actin and myosin IIA along the diameter of the activating (F-actin: n = 68; myosin IIA: n = 66; top graph) versus inhibitory (F-actin: n = 68; myosin IIA: n = 66; bottom graph) site are shown. Scale bar indicates 5 μm. YTS F-tractin GFP cells were incubated on slides with mCherry-expressing 221-Cw7 or 221-Cw4 target cells for 5 min at 37°C, and then fixed. Z stack images of NK-target conjugates were collected, and 3D projections of the NKIS planes were assembled. F-actin distribution from multiple cells were profiled by ImageJ, as described in the Materials and Methods, and the average normalized fold intensities of F-actin along the diameter of the activating vs. inhibitory NKIS are shown (n = 44 for Cw7, n = 46 for Cw4). Scale bars indicate 5 μm. NKIS sites are indicated by white arrowheads. Primary NK cells were transfected with F-tractin GFP, and the expression level of F-tractin GFP was determined by FACS. YTS F-tractin GFP cells were dropped over coverslips coated with either anti-CD28 or anti-KIR2DL1, and live cell imaging was performed as in Fig 2A. Kymographs of F-actin dynamics were compiled along the contact site radius. Kymographic traces of F-actin were compiled into a graph to show the average traces for each cell type at the LP. The y-axis shows distance travelled (μm), whereas the x-axis shows time (s) (anti-CD28: total traces = 632 from 7 movies; anti-KIR2DL1: total traces = 296 from 7 movies). Primary NK F-tractin GFP cells were dropped over coverslips coated with anti-NKG2D or anti-NKG2A, and quantitative analysis of F-actin traces from the LP of activating vs. inhibitory contact sites was performed as in (F) (anti-NKG2D: total traces = 532 from 5 movies; anti-NKG2A: total traces = 312 from 5 movies). YTS F-tractin GFP cells were dropped over coverslips coated with anti-CD28 or anti-KIR2DL1, and quantitative analysis of F-actin traces from the LM and CB of activating vs. inhibitory contact sites was performed as in (F) (anti-CD28: total traces = 231 from 7 movies; anti-KIR2DL1: total traces = 395 from 7 movies). Primary NK F-tractin GFP cells were dropped over coverslips coated with anti-NKG2D or anti-NKG2A, and quantitative analysis of F-actin traces from the LM and CB of activating vs. inhibitory contact sites was performed as in (F) (anti-NKG2D: total traces = 157 from 5 movies; anti-NKG2A: total traces = 98 from 5 movies). YTS F-tractin GFP cells were dropped over isotype IgG coated coverslips, and imaged as in Fig 2A. F-actin kymographs were compiled along the cell radius. Quantitative analysis of F-actin traces was performed as in Fig 2C (IgG: total traces = 92 from 6 movies). YTS-2DL1 cells expressing mCherry-SHP-1 were dropped over anti-CD28- or anti-KIR2DL1-coated coverslips and imaged as in Fig 2E. SHP-1 kymographs were compiled along the cell radius. Quantitative analysis of SHP-1 traces was performed as in (F) (anti-CD28: total traces = 194 from 4 movies; anti-KIR2DL1: total traces = 121 from 5 movies). Data information: Data are means ± SEM. Data are representative of three independent experiments. Source data are available online for this figure. Download figure Download PowerPoint Next, we followed the actin centripetal retrograde flow following engagement of NK cell inhibitory versus activating receptors in live cells. Imaging of F-actin dynamics was performed at the contact site of fully spread NK cells. YTS-2DL1 or primary NK cells expressing F-tractin GFP (pNK F-tractin GFP cells, Fig EV2E) were seeded over surfaces pre-coated with the stimulatory anti-CD28 or NKG2D, or the inhibitory antibodies, anti-KIR2DL1 or NKG2A. Monitoring of ARF revealed that the F-actin network at the LP of activating NKIS demonstrated fast and continuous retrograde flow, while at the inhibitory contact site, this continuous flow was slow and barely detectable; instead, random and inconsistent F-actin movements were observed (Movies EV1, EV2 and EV3). To further characterize actin dynamics, ARF velocity was quantified using kymograph analysis. Kymographs were obtained along the radius of F-tractin-expressing NK cells. F-actin features were monitored either at the outer margin of the kymographs, representing the LP (Fig 2A and B; blue arrows), or at the intermediate and inner regions, representing the LM and CB, respectively (Fig 2A; red arrows). In this analysis, the angle of the F-actin trace designates its velocity; a vertical orientation indicates slow to negligible velocity, whereas tendency to horizontal orientation indicates a faster velocity. Strikingly, F-actin average traces of the LP demonstrated a shallower slope under the inhibitory conditions (Fig EV2F and G), indicating slower movement. Indeed, F-actin flow at the LP of both YTS and pNK cells was significantly faster at the activating contact site (YTS 0.13 ± 0.0037 μm/s, pNK 0.2 ± 0.006 μm/s) relative to the inhibitory contact site (YTS 0.024 ± 0.0014 μm/s, pNK 0.017 ± 0.0009 μm/s; P ≤ 0.00001; Fig 2C and D). Furthermore, we found that the velocity of F-actin depends on its location along the NKIS radius. Kymograph analysis of the vertical traces in the LM and CB indicated an immobile F-actin network (Fig 2A, red arrow). Indeed, quantitative analysis demonstrated slower or negligible ARF movement at these regions (Figs 2C and D, and EV2H and I). As a negative control, YTS F-tractin GFP cells were seeded over slides coated with IgG isotype antibody, followed by analysis of ARF velocity. Live cell imaging and kymograph analysis demonstrated negligible actin flow velocity, regardless of the location across the NKIS radius, which was significantly slower than ARF velocity at the LP of the inhibitory NKIS (IgG: 0.006 ± 0.0004 μm/s, KIR2DL1: 0.024 ± 0.0014 μm/s; P ≤ 0.00001; Fig EV2J). Figure 2. The inhibitory vs. activating NKISs are characterized by different F-actin dynamics A. YTS F-tractin GFP cells were dropped over coverslips coated with either anti-CD28 or anti-KIR2DL1 and imaged at a single focal plane at one frame per second. Representative images from movies are shown, and kymographs of F-actin dynamics were compiled along the contact site radius (represented as a dashed line). Blue arrowheads indicate traces at the LP, and red arrowheads indicate traces at LM and CB. B. Primary NK cells expressing F-tractin GFP were dropped over anti-NKG2D- or anti-NKG2A-coated coverslips, imaged as in (A), and kymographs were compiled along the cell radius. Blue arrowheads indicate traces at the LP. C, D. Kymographic analysis of F-actin traces was compiled into a graph to show the distribution of F-actin velocity (μm/s) along the radius of the contact site (YTS/anti-CD28: total traces = 863 from 7 movies; YTS/anti-KIR2DL1: total traces = 691 from 7 movies; pNK/anti-NKG2D: total traces = 748 from 5 movies; pNK/anti-NKG2A: total traces = 537 from 5 movies). 0, cell center; 1, cell periphery. E. YTS-2DL1 cells expressing mCherry-SHP-1 were dropped over anti-CD28- or anti-KIR2DL1-coated coverslips and imaged as in (A). SHP-1 kymographs were compiled along the cell radius. Blue arrowheads indicate traces at the LP. F. Quantitative analysis of SHP-1 traces was performed as in (2C) (anti-CD28: total traces = 247 from 4 movies; anti-KIR2DL1: total traces = 315 from 5 movies). Data information: Scale bars indicate 5 μm. Data are means ± SEM. Statistical significances were calculated with Student's t-tests used for unpaired, two-tailed samples. Data are representative of three independent experiments. Download figure Download PowerPoint These data suggest differential distribution and F-actin dynamics at the activating versus inhibitory NKIS. We demonstrate here the increased F-actin accumulation at the periphery of activating NKIS, with rapid ARF at the LP site relative to the inhibitory settings. Since slower ARF was observed at the inhibitory versus activating settings, with increasing formation of the SHP-1:β-actin molecular complex, we asked whether these molecular events are related. Thus, the role of ARF in SHP-1 dynamics was examined. YTS-2DL1 cells transiently expressing mCherry-SHP-1 were seeded over activating or inhibitory surfaces, and live cell imaging of SHP-1 dynamics was performed. Interestingly, quantitative kymograph analysis indicated faster SHP-1 retrograde flow at the LP of activating (0.15 ± 0.0076 μm/s) vs. inhibitory (0.021 ± 0.0028 μm/s; P ≤ 0.00001) contact sites (Figs 2E and F, and EV2K). These findings demonstrate similar velocities of ARF and SHP-1, suggesting that the translocation of F-actin and SHP-1 might be related. F-actin turnover and myosin IIA contractile force are required for F-actin flow Actomyosin retrograde flow is potentially driven by F-actin polymerization, resulting in "pushing" forces toward the membrane edge of a spreading cell, and/or myosin contractile forces that "pull" the F-actin network away from the cell membrane (Babich et al, 2012; Yi et al, 2012; Hammer & Burkhardt, 2013
Referência(s)