Three mice strains, 6–8 weeks old, were used: wild-type (WT) C57BL/6, OTII transgenic mice with transgenic (tg) TCRα/TCRβ receptor specific for OVA peptide (38), and double tg DO11.10 TCR/tg CARΔ (TgCAR) (Taconic), which express human coxsackie/adenovirus receptor lacking a cytosolic domain (39, 40). All the three strains were housed in a specific pathogen-free animal facility at the Institute of Molecular Genetics (Prague, Czech Republic).

For western blotting, mouse anti-RACK1, rabbit anti-pY394^Lck (Santa Cruz), anti-pY505Lck (Cell signaling), mouse anti-Lck (3A5), and phosphotyrosine-specific platinum 4G10 mAb (Millipore) were used. Mouse anti-α-actinin-1 (BM-75.2), rabbit anti-RACK1 C-end, anti-GAPDH, and cholera toxin–horseradish peroxidase (HRP) conjugate were purchased from Sigma-Aldrich, anti-α-tubulin from GeneTex, and anti-GADS and anti-LASP from Millipore. Mouse anti-rabbit and goat anti-mouse HRP light chain-specific antibodies (Jackson IRLab) were used for protein detection after immunoprecipitation (IP). For IP purposes, antibodies were coupled to Protein A or G magnetic beads (Millipore) or the bead-conjugated anti-RACK1 from Santa Cruz were used. Biotinylated anti-mouse CD4 (GK1.5) and biotinylated anti-mouse TCRβ (H57) mAbs were purchased from eBiosciences (USA). For immunofluorescence, CellTrace™ Far Red DDAO-SE and Alexa Fluor 488, 555, or 647 goat anti-mouse IgG or goat anti-rabbit IgG (H + L) were used. Actin filaments were visualized by Alexa Fluor 488 Phalloidin purchased from Life Technologies. Streptavidin, Brij58 (polyoxyethylene-20-cetyl-ether), and nocodazole were purchased from Sigma-Aldrich. Latrunculin B, lauryl maltoside (n-dodecyl-β-d-maltoside, LM), and Src-family kinase inhibitor PP2 were obtained from Calbiochem (Merck). Ultrapure grade paraformaldehyde (PFA) was obtained from Polysciences. OVA_323–339 peptide from GenScript Corporation, and staphylococcal enterotoxin E (SEE) was purchased from Toxin Technology.

The NIH3T3 cell line and Lck infectants cell lines were maintained in Iscove’s modified Dulbecco’s medium (IMDM), WT, and JCAM1.6 The Jurkat T-cell line and RAJI B-cell line were cultured in RPMI 1640 (Life Technologies). Both media were supplemented with 10% inactivated fetal calf serum (FCS) and 100 U of penicillin/10 μg streptomycin antibiotics (Sigma) per 1 ml of media. Bone marrow-derived dendritic cells (BMDCs) from OTII transgenic mice were isolated from mice femur and tibia cavities. The cells were cultured for 6 days in RPMI medium supplemented with GM-CSF-containing supernatant, which was produced by the LUTZ cell line (final concentration was adjusted to 30 ng/ml). After 3 days of cultivation, one half of the media was replenished, and on day 6, the cells were harvested and used for further experiments.

Lck constructs, described previously (21), were inserted into the murine stem cell virus (MSCV)-based internal ribosome entry site (IRES)-enhanced green fluorescent protein virus MigR1 (41), permitting the concurrent expression of a given gene and EGFP. EGFP–RACK1 and mCitrine–RACK1 was a gift from Dr. Vomastek, and CFP–Lck (clone W25) was a gift from Dr. Stockinger (42). Generation of retrovirus packaging cell lines and retrovirus stock as well as the infection of NIH3T3 or Jurkat T-cells was performed as outlined elsewhere (43, 44). Successfully infected cells were sorted to the comparable protein expression level of the desired gene.

The procedure was performed as previously described (14). Briefly, primary lymph node CD4⁺ T-cells (~95% purity) were isolated from mice using MACS CD4⁺ T-cell isolation kit (AutoMACS, Miltenyi Biotec). CD4⁺ T-cells were precoated with biotinylated anti-TCR (1 μg/ml, clone H57) and anti-CD4 (0.3 μg/ml, clone GK1.5) antibodies in 500 μl of PBS + 3% FCS for 30 min at 4°C, washed, and resuspended in 20 μl of PBS + 1% FCS per tube or were indicated in 20 μl of PBS + 1% FCS + 20 μM PP2. Cells were pre-warmed in 37°C, and co-aggregation-mediated activation was achieved with the addition of streptavidin to a final concentration of 50 μg/ml. Cells were vortexed and incubated at 37°C for the indicated period of time. The activation was stopped with the addition of either ice-cold lysis buffer with indicated detergent and inhibitors (see below) or immediately boiled in Laemmli sample buffer.

Cell lysis of CD4⁺ T-cells was performed in TKM (50 mM Tris, pH 7.4, 25 mM KCl, 5 mM MgCl₂, 1 mM EDTA, 0.5% Brij58) or TNE (50 mM Tris, pH 8, 12.5 mM EDTA, 1% NP-40) lysis buffer supplemented with protease and phosphatase inhibitor cocktails (Roche). Lysates were incubated on ice for 30 min, spun down at 800 × g for 2 min (removal of nuclei), and used for further applications. For RACK1 IPs, the lysates were incubated with antibody-precoated RACK1 beads overnight at 4°C on a rotational wheel. Beads were then intensively washed (six to eight times), boiled in Laemmli sample buffer for 10 min, and immunoprecipitated proteins resolved by SDS-PAGE.

Proteins resolved on polyacrylamide gels were transferred onto a PVDF membrane (Millipore) and blocked for 1 h in 5% non-fat milk (for IP) or in 1% BSA in TBS-T at room temperature. Blots were then incubated for 1 h with primary antibodies diluted in blocking buffer, washed followed by a 1-h incubation with secondary antibodies conjugated to HRP, and developed by incubation with ECL substrate (Thermo Scientific). Where indicated, densitometry quantification was performed by AIDA image analyzer software from raw image data obtained from a GS-800 Biorad densitometer scanner.

For immunofluorescence, the staining protocol was adopted from the website www.cellsignal.com with some modifications. Briefly, CD4⁺ T-cells were seeded on poly-l-lysine-coated coverslips and allowed to adhere for 15 min at 37°C, 5% CO₂ followed by cell fixation in 4% PFA for 15 min. The following procedure was applied to all cell types. After PFA fixation, the cells were permeabilized with ice-cold methanol for 10 min in −20°C (methanol step had to be omitted when phalloidin was used to stain actin cytoskeleton). The cells were blocked for 1 h with PBS containing 0.3% Triton X-100 (PBT) with the addition of 2.5% FCS and 2.5% BSA and incubated with primary and secondary antibodies for 1 h consecutively. Where indicated, cells were stained with DAPI for 10 min. Coverslips were mounted using 4% n-propyl gallate in glycerol. Samples were analyzed by sp5 confocal microscopy (Leica) or N-SIM super-resolution microscopy (Nikon). Image reconstruction was performed by Huygens Professional (SVI) or NIS elements (Nikon) software. Image post-editing and image analysis including computing of Pearson’s colocalization coefficient (PCC) were done with Fiji imaging software and its plug-in JACoB.

The preparation of conjugates for microscopic analysis of IS formation in vitro was performed as previously described (45). Briefly, BMDCs were prepared in parallel as described above. Cells were then pulsed with OVA_323–339 peptide for 2 h and TCR-transgenic CD4⁺ primary T-cells isolated from OTII mice were admixed with APCs at 3:1 ratio. The formation of APC–T-cell conjugates was achieved by short centrifugation. Conjugates were incubated in serum-free RPMI medium at 37°C. The cells were fixed with 4% PFA for 15 min, stained with anti-RACK1 and anti-Lck, and used for the microscopy analysis.

The WT or Lck-deficient Jurkat T-cell line JCAM1.6 was retrovirally transfected with either an EGFP–RACK1 or CFP–Lck construct, respectively. In addition, the mCitrine–RACK1 construct was electroporated (BTX ECM 830, 300 V, 10 ms, 4 mm cuvette) into CFP–Lck stable infectants. Jurkat T-cells that were positive for both CFP and mCitrine were sorted (BD, Influx cell sorter) and rested for 1 day in 37°C/CO₂. The formation of RAJI–Jurkat conjugates using live cell imaging was performed as previously described (46). Briefly, RAJI cells, serving as APCs, were labeled with DDAO-SE (Life Technologies), then loaded with 1 μg/ml SEE, and transferred into cover glass chamber (Ibidi). CFP–Lck/mCitrine–RACK1 Jurkat T-cells were subsequently added at a 1:1 ratio. Cells were observed in a 37°C/CO₂ climate chamber using a DeltaVision Core/Olympus IX71 microscope under CFP/YFP/mCherry filter cubes. Images were acquired every 15 s. Image post-editing and time-lapse movies were done with Fiji imaging software.

To analyze the concentricity and apposition of RACK1 and Lck (or GADS), the distance from the centroid of the cell to the cell edge was measured using a fluorescent intensity profile (see details in Figure S1 in Data Sheet 1 in the Supplementary Material). All microscopic quantitative analyses were performed by ImageJ program. Statistical analyses were performed with GraphPad Prism 5 using a paired two-tailed t-test.

Quantitation of the recruitment of RACK1 to the IS was calculated as described previously (47). Briefly, using the selection brush tool in ImageJ program, the T cell area adjacent to the synapse region, outside synapse region and the background area outside of the cell, was demarcated. The relative recruitment index (RRI) was calculated as the [mean fluorescence intensity (MFI) at synapse region minus the background]/[MFI of the outside synapse region minus background]. Quantitative measurements of MFI were performed with the program ImageJ. Statistical analysis was performed with GraphPad Prism 5 using a paired two-tailed t-test.

This size-exclusion chromatography procedure is based on a previously described protocol (48) with slight modifications (14). Briefly, a 5-ml pipette tip plugged with a small piece of glass wool and filled with Sepharose 4B beads (Sigma-Aldrich) was used as the column. Cells were lysed in TKM-Brij58 lysis buffer for 30 min and spun down for 2 min at 800 × g. Supernatant at a total volume of 1/10 of the stationary bead volume was loaded onto the top of the column and eluted with cell lysis buffer. All steps were performed at 4°C. In this setting, fraction #4 contains complexes of >10⁷ Da; most of the pentameric IgM and IgG standards eluted in fractions #7 and #9, respectively (48).

The isolation of DRMs was performed as described previously (14, 16). For the assessment of protein distribution in fractions obtained by gel filtration, pooled fractions were mixed 1:1 with 80% sucrose and subjected to the same protocol. Light DRMs, corresponding to classical LRs, are enriched in top fractions (#1–3), while the bottom fractions (#8–10) concentrate heavy DRMs together with soluble proteins.

For assessing the involvement of cytoskeletal components in the redistribution of Lck after TCR/CD4 aggregation, CD4⁺ T-cells were resuspended in PBS + 3% FCS and treated with either 2 μg/ml of latrunculin B or 10 μM of nocodazole for 30 min in 37°C and 5% CO₂. Subsequent precoating of the cells and TCR/CD4 co-aggregation were also performed in the presence of these inhibitors on ice. Before activation, small aliquots of latrunculin B- and nocodazole-treated samples, as well as untreated controls sample, were pre-warmed and fixed at 37°C with PFA for 15 min, stained as described above, and analyzed by microscopy.

Adenoviral vectors and virus particles containing shRNA hairpins, as well as control empty vector/viruses, were prepared using the Knockout RNAi system and the Adeno-X Expression system 1 (Clontech) and used according to the manufacturer’s protocol. Target sequences of shRNA for RACK1 are as follows: RAO#2-gcaagatcattgtagatgaat, RAO#4-ctcccacttcgttagtgat, and RAO#5-ggatgagagtcattcagaatg. Transduction of T-cells was performed as described elsewhere (49). Briefly, isolated naïve CD4⁺ T-cells were resuspended in DMEM medium containing 2% FCS and desired MOI of adenoviral particles and incubated at 37°C/CO₂ for 1 h. Then, the cell/virus mixture was transferred to a culture dish and incubated for the indicated time at 37°C/CO₂ in complete RPMI media supplemented with 2 ng/ml of rmIL-7 (PeproTech). Then, live T-cells were sorted, rested for 4 h, and used for further experiments.

The stained protein bands were processed according to the standard protocol generally used for mass spectrometry (MS) protein identification (50) with minor modifications. The gel slices containing the proteins of interest were washed, proteins reduced with dithiothreitol, alkylated with iodoacetamide, and digested with trypsin. The extracted peptides were separated using a home-made microgradient device (51) with C18 reversed phase capillary column (i.d. 200 μm, length 70 mm) for LC MALDI-TOF/TOF MS and MS/MS analysis using 4800 Proteomics Analyzer (Applied Biosystems, Framingham, MA, USA) with α-cyano-4-hydroxycinnamic acid as MALDI matrix. Protein database identification was carried out with Protein Pilot 2.0 software using the SwissProt part of the UniProt database server.

First, we determined the expression of RACK1 in mouse primary lymph node CD4⁺ T-cells. RT-PCR analysis performed on total mRNA isolated from FACS-sorted cells (99.3% purity) demonstrated detectable levels in non-activated CD4⁺ T-cells (data not shown). Confocal microscopy confirmed the presence of RACK1 on the protein level and showed that Lck is localized almost exclusively to the plasma membrane. We observed that RACK1 is positioned just beneath Lck, in a constrained cytoplasmic niche between the plasma membrane and nucleus, the latter fulfilling the vast majority of intracellular space (Figure 1A). To assess RACK1 subcellular distribution more accurately, we performed super-resolution microscopy. It confirmed that in resting CD4⁺ T-cells, Lck and RACK1 are juxtaposed concentrically (Figures 1B–D) and exhibit a mild overlap (Figure 1E) in accordance with the PCC >0.6 (Figure 1F).

Next, we assessed the cellular co-distribution of RACK1 and Lck in activated primary lymph node CD4⁺ T-cells during early phases of IS formation. Microscopic examination of cell conjugates of OVA peptide-pulsed BMDCs with transgenic T-cells specific for OVA peptide showed that Lck and RACK1 concomitantly translocated to and enriched in IS at early phases of its formation (2–5 min) (Figures 2A,B). Lck, the accumulation of which in IS has been previously demonstrated (20, 45), was used as an internal control. Interestingly, we observed that RRI for RACK1 is even higher than that for Lck, strongly suggesting the physiological importance of RACK1 enrichment in the forming IS (Figure 2B).

To analyze the co-redistribution of RACK1 and Lck during early phases of IS formation in more detail, we performed experiments that would visualize the kinetics of this process. Toward this end, Jurkat T-cells, which are able to form cell conjugates with RAJI cells were infected with a RACK1–EGFP retroviral construct, and the kinetics of RACK1–EGFP translocation into IS was examined in a time-dependent manner using live fluorescent microscopy (Figure 3A; Video S1 in Supplementary Material). RACK1–EGFP protein was found moderately enriched in the forming IS (Figure 3B) with the total time of its transient residency from 6 min to more than 18 min (Figure 3C).

To visualize the Lck and RACK1 co-redistribution event, we co-transfected Jurkat T-cells with two constructs: RACK1–mCitrine and Lck–CFP. mCitrine and CFP double-positive cells were FACS-sorted, and time-lapse images were recorded (Video S2 in Supplementary Material). As presented in Figure 3D, shortly after conjugate formation (1–2 min, see the arrow in the 2 min time frame), RACK1 and Lck co-redistributed to the forming IS where they both persisted for several minutes. Between 6 and 8 min, RACK1–mCitrine slowly moved distally from IS. At the 9-min mark until 11 min (Figure 3D, bottom two rows of images), suddenly and rapidly RACK1–mCitrine re-translocated back to the bulk volume of cytoplasmic space of the T-cell, leaving Lck behind in the IS where it remained for at least 7–8 min (Figure 3E). It is of note that the Jurkat T-cell which appears in the upper right hand corner of the 2-min image of the time-lapsed video and image and in which the amount of RACK1–mCitrine markedly exceeded that of Lck–CFP (red color is below the visible range) displays similar kinetics of RACK1 translocation to the IS and reverse re-translocation back to the cytoplasm as described above (Figure 3D; Video S2 in Supplementary Material).

Taken together, these data suggest that T-cell activation induces a rapid, cooperative, and IS-directed movement of Lck and RACK1. This co-redistribution pattern provided the first evidence that RACK1 could participate in the mechanism guiding the redistribution of membrane Lck upon TCR engagement.

To determine if RACK1 is involved in proximal T-cell signaling via its interaction with Lck and if it is constitutive or activation inducible, we utilized the model of antibody-mediated T-cell activation enabling to study the kinetics of interaction between two interacting proteins during the first seconds after TCR/CD4 co-aggregation (14, 16) (Figure 4A). On average, a threefold to sixfold enrichment of Lck in complexes with RACK1 reached its maximum level 10 s after TCR–CD4 engagement (Figure 4B). Then, in the following 30–90 s, these levels diminished to those observed in non-activated cells (Figures 4A and 5A,C). Consistent with the colocalization analysis (Figures 1C–F) only low, background levels of RACK1–Lck complexes were detected in the non-precoated control sample (Figures 4A and 5A, “non-precoated” lane).

We recently demonstrated that a small preactivated pool of pY394^Lck, which plays an important role in TCR triggering in resting primary CD4⁺ T-cells, associates with an LM-sensitive type of membrane microdomains called heavy DRMs (14). This high molecular weight (HMW) fraction can be obtained from Brij58 detergent-solubilized cells by gel filtration (53, 54). Interestingly, it is the heavy DRM-associated pool of Lck which, upon TCR/CD4 engagement, not only significantly increases its kinase activity but also exclusively translocates to distinct types of microdomains called light DRMs (14, 55). These considerations led us to assess the subcellular distribution of RACK1 in CD4⁺ T-cells. As illustrated in Figure 4C, RACK1, similar to the pool of pY394^Lck, is enriched in the complexes that are associated with high HMW fractions #4–6. Next, we assessed the physical association of Lck and RACK1 in the fractions that were prepared from resting and activated T-cells. As shown in Figure 4D, the immunopreciptation of RACK1 from pooled HMW fractions #4 + 5 and #6 + 7 confirmed the activation-induced formation of RACK1–Lck complexes that were detected only in fractions #4 + 5 that were prepared from activated T-cells. These results suggest that in resting T-cells, the pool of preactivated pY394^Lck and RACK1 are in physical proximity by co-distributing to heavy DRMs, which spatially restricts their transient interaction upon TCR activation. These data thus support the prediction that upon TCR/CD4 triggering, RACK1 can bind a spatially restricted pool of kinase active Lck and functions as a transportation vehicle that assists the redistribution of Lck to light DRMs, as documented previously (14, 16).

Next, we sought to determine which domain or tyrosine residue, which regulates Lck kinase activity, mediates its interaction with RACK1. Toward this end, we prepared a NIH3T3 fibroblast cell line expressing wild type Lck (WT), constitutively active Lck (Y505F) and Y505F Lck backbone with additionally inactivated either SH3 (W97K) or SH2 (R154K) domain (Figures S2A–C in Data Sheet 1 in the Supplementary Material). Endogenous RACK1 was immunoprecipitated with anti-RACK1 antibody, and its aliquots were blotted for Lck and RACK1. We found that in NIH3T3 cells, the WT, Y505F, and even kinase compromised Y505F/Y394F Lck interacted comparably with endogenous RACK1. However, this complex formation was nearly abolished in the variant expressing non-functional SH3 domain and severely compromised in the SH2 mutant of Y505F Lck (Figures 4E,F). This suggests that the presence of both functional modular SH3 and SH2 domains of Lck is a prerequisite for RACK1–Lck complex formation.

The above results revealed that RACK1–Lck complex formation, at least in fibroblast cells, proceeds irrespective of the activation status of Lck (Figure 4E, Y505F versus Y394F/Y505F Lck). We next assessed if such a mode of interaction would also occur in primary CD4⁺ T-cells. As shown in Figure 5A, the kinetics of TCR/CD4 co-aggregation allowed the IP of RACK1–Lck complexes, which contain detectable levels of pY394^Lck. Their highest levels were observed at 5–10 s after activation (Figure 5A, pY394^Lck panel, and Figure 5B), whereby the latter time point correlated with the peak of enrichment of Lck in these complexes (Figure 5A, Lck panel, and Figure 5C). This would suggest that a TCR/CD4-induced increase in pY394^Lck precedes and thus predicates the formation of RACK1–Lck complexes. However, the activation-induced RACK1–Lck complex formation was not ablated in T-cells that were pretreated with the SFK kinase inhibitor, PP2, which effectively inhibited the activation-induced global tyrosine phosphorylation of TCR downstream substrates (Figure 5D, 4G10 bottom panel, right lane, PP2+) and blocked the activation-induced enhancement of pY394^Lck levels (Figure 6D, pY394^Lck panel, PP2+). Together, these biochemical results strongly suggest that the physical interaction between Lck and RACK1 is independent of the activation status of Lck. However, under normal circumstances during TCR/CD4 engagement, upon which the preactivated pool of Lck is significantly increased (56), these complexes contain a sizeable fraction of pY394^Lck (Figure 5A).

Our observation that the formation of RACK1–Lck complexes was dependent on both the SH2 and SH3 domain of Lck (Figure 4E) suggested that this interaction could be indirect and involved additional structural components. Consistent with this notion are results from our pull-down experiments using in vitro translation that failed to show direct RACK1–Lck binding (data not shown). To gain added insight into the complexity of RACK1–Lck interaction, we first assessed the presence of additional components in RACK1 immunoprecipitates from activated primary CD4⁺ T-cells. To recognize their presence in immunoprecipitated RACK1 complexes, we took advantage of the fact that TCR signaling is a tyrosine phosphorylation (pY)-driven event, and thus, at least those components of RACK1 complex which change their pY status could be readily detectable.

As illustrated in Figure 6A, several tyrosine phosphorylated proteins co-immunoprecipitated with RACK1. Specifically, we found four major phosphoproteins in activated T-cells, three of which possessed the molecular weights (MWs) of ~100, ~70, and ~38 kDa, and their pY status was associated with RACK1 in an activation-inducible manner. The fourth phosphoprotein which possessed an MW of ~56 kDa matched the MW of Lck. To reveal the identity of the former three phosphoproteins, the corresponding areas from a coomassie-stained gel were excised and their protein content subjected to MS analysis (Table S1 in the Supplementary Material). Due to the fact that RACK1 is a scaffold protein that plays an essential role in transcription, epigenetics, and translation as well as binds diverse signaling and structural proteins, we zoomed in on those proteins that potentially act outside these activities. Among them, we focused on proteins that were known to be involved in proximal TCR signaling and cytoskeleton regulation of the forming IS (Figure 6B). The capacity of these candidate proteins to bind RACK1 before and after T-cell activation was assessed. GADS, LASP1, and α-actinin-1 exhibited sizeable changes in the kinetics of interaction with RACK1 upon TCR/CD4 co-aggregation. Notably, while α-actinin-1 binds to RACK1 with increasing intensity (from 0 to 90 s, Figure 6C, upper panel), the level of RACK1 interaction with GADS and LASP1 was diminished with similar kinetics over the time tested (Figure 6C, GADS and LASP1 panels). RACK1–GADS complex formation in resting primary T-cells was microscopically corroborated by their colocalization in resting T-cells (Figure S3 in Data Sheet 1 in the Supplementary Material).

While RACK1 interactome involves unusually high number of partners (BioGRID database; http://thebiogrid.org), the binding of α-actinin-1 to RACK1 complexes is an original finding. Actinins are primarily considered to be actin-cross-linking proteins but can also link transmembrane proteins to the cytoskeleton and membrane trafficking events (57). Our data showed that in resting CD4⁺ T-cells, α-actinin-1 inducibly associates with RACK1 upon TCR/CD4 co-aggregation (Figure 6D, TCRβ/CD4, 0 versus 10 s, α-actinin-1 panel, and Figure 6E). However, as described above, while RACK1–Lck complex formation proceeds even in the presence of SFK inhibitor PP2 when Lck kinase activity is severely compromised (Figure 6D, TCRβ/CD4, 10 s/PP2+, pY394^Lck panel), binding of this complex to α-actinin-1 is blocked (Figure 6D, last lane, α-actinin-1 panel, and Figure 6E). This advocates for a scenario in which TCR/CD4 co-ligation induces the formation of complexes which contain RACK1 and kinase active Lck (Figure 5A), the latter required for linking these complexes to α-actinin-1 component within the cytoskeleton. If Lck activity is blocked, the formation of RACK1–Lck complexes still proceeds, but α-actinin-1 is not engaged.

The timely and spatially coordinated complex formation between Lck and RACK1 in heavy DRM, their co-redistribution to the forming IS and linkage to α-actinin-1 suggest that the cytoskeletal network is actively involved. We have also reported that the activation-induced translocation of Lck to light DRMs is blocked by nocodazole, an inhibitor of microtubular assembly (14). To further investigate the potential involvement of RACK1 in the microtubular network-assisted translocation of Lck, we assessed the effect of nocodazole treatment on RACK1–Lck complex formation. The results showed that nocodazole, but not latrunculin-mediated inhibition of the actin cytoskeleton, effectively blocked the formation of RACK1–Lck complexes (Figure 7A, Lck panel, and Figure 7B). Of note, while lantrunculin and nocodazole treatment effectively disrupted the actin and microtubular network, respectively (Figure 7C), the activation-induced global tyrosine phosphorylation in the presence of these inhibitors was comparable to the untreated control sample (Figure 7A, bottom panel, 4G10). These data potentially provide a mechanistic explanation for the activation-dependent redistribution of Lck to light DRMs by virtue of linking the TCR/CD4-Lck complex to the microtubular cytoskeletal network via RACK1. This also suggests that at the very early stages of T-cell engagement, actin cytoskeleton does not affect Lck mobility even though α-actinin, which can potentially crosslink actin microfilaments, is already part of the Lck redistribution machinery.

We next attempted to evaluate the potential functional relevance of RACK1–Lck interaction for the redistribution of Lck to light DRMs (16). CD4⁺ T-cells from TgCAR transgenic mice expressing the receptor for adenovirus on T-cells were infected with adenovirus containing shRNAs to downregulate RACK1 (Figure 8A). After 96 h of adenovirus infection, the cells were harvested and activated by TCR/CD4 co-aggregation. The distribution of Lck to light DRMs was then compared to cells that were infected with a mock construct. Regardless of the technical caveats that are associated with our limited ability to consistently generate viable RACK1 knockdown (KD) immune cells (see commentary in Figure S4 in Data Sheet 1 in the Supplementary Material), our data showed that in activated T-cells, the reduced levels of RACK1 correlated with diminished redistribution rate of Lck to light DRMs (Figure 8B). These results points to the potential involvement of RACK1-based multiprotein signaling network in Lck redistribution during proximal T-cell signaling.