Report

Requirement for Coronin 1 in T Lymphocyte Trafficking and Cellular Homeostasis

See allHide authors and affiliations

Science  11 Aug 2006:
Vol. 313, Issue 5788, pp. 839-842
DOI: 10.1126/science.1130563

Abstract

The evolutionarily conserved actin-related protein (Arp2/3) complex is a key component of actin filament networks that is dynamically regulated by nucleation-promoting and inhibitory factors. Although much is known about actin assembly, the physiologic functions of inhibitory proteins are unclear. We generated coronin 1–/– mice and found that coronin 1 exerted an inhibitory effect on cellular steady-state F-actin formation via an Arp2/3-dependent mechanism. Whereas coronin 1 was required for chemokine-mediated migration, it was dispensable for T cell antigen receptor functions in T cells. Moreover, actin dynamics, through a mitochondrial pathway, was linked to lymphocyte homeostasis.

The integrity of the actin cytoskeletal network is critical for a diverse range of biological processes and is dynamically regulated by a cohort of actin-associated proteins. The Wiskott-Aldrich syndrome (WAS) and the suppressor of cyclic adenosine monophosphate (cAMP) receptor (SCAR) proteins promote actin nucleation and assembly via the Arp2/3 complex (13), whereas inhibitory proteins, which include coronin, tropomysosin, and caldesmon, oppose Arp2/3 function (46). The evolutionarily conserved coronin family of actin-binding proteins has been implicated in the regulation of multiple actin-mediated cellular functions, including cell migration, cytokinesis, and cell growth of Dictyostelium discoideum and Saccharomyces cerevisiae (712). Among the seven mammalian coronin family members, coronin 1 (also known as coro1a, TACO, or p57) is preferentially expressed in cells of hematopoietic origin, where it is coexpressed with other more widely expressed coronin family members that include coronins 2, 3, and 7 (fig. S1A) (13). In mammals, coronin 1 colocalizes with F-actin surrounding phagocytic vesicles in neutrophils and macrophages and F-actin–rich membranes in activated T cells (1416).

To investigate the physiological role of coronin 1, we generated coronin 1–/– mice (fig. S1, B and C). No coronin 1 protein was detected in thymocytes, splenocytes, or bone marrow–derived cells isolated from coronin 1–/– mice, and expression of coronins 2 and 3 was not altered (fig. S1D). Analysis of lymphoid tissues revealed normal segregation of T and B cells but a paucity of T cells in spleens and lymph nodes of coronin 1–/– mice (Fig. 1A). Both CD4+ and CD8+ T cells were decreased in the blood, spleen, and lymph nodes (Fig. 1B). Naïve, but not memory/effector, splenic T cells were decreased, although both were reduced in the blood and lymph nodes of coronin 1–/– mice. Thymic cellularity and subpopulations were similar between coronin 1–/– and coronin 1+/+ mice, although a small reduction in mature CD4+ and CD8+ (CD69) coronin 1–/– thymocytes was observed (Fig. 1C and fig. S1, I to J). An analysis of coronin 1–/– mice bearing either major histocompatibility complex (MHC) class 1 restricted H-Y or class II restricted DO11.10 transgenic T cell antigen receptors (TCRs) revealed normal thymic development and decreased naïve T cells in lymph nodes (Fig. 1D).

Fig. 1.

Reduced numbers of peripheral T cells in coronin 1–/– mice. (A) Immunohistochemical staining for CD3 (red) and B220 (blue) of splenic tissue sections from coronin 1+/+ and coronin 1–/– mice. (B) Total lymphocyte, B220+ B cells, CD4+ and CD8+ T cells, and CD4+ naïve (CD44loCD62Lhi) and effector/memory (CD44hiCD62Llo) T cell subset numbers in blood (n = 4), spleen (n = 8), and inguinal lymph nodes (n = 8) were quantified from coronin 1+/+ (open bars) and coronin 1–/– (solid bars) mice (age 5 to 7 weeks). Statistical analysis was performed by two-tailed Student's t test (*P < 0.05, **P < 0.01) for all figures. Error bars indicate the standard error of the mean (SEM). Representative FACS profiles are shown in fig. S1E. The only discernible non–T cell defect was decreased circulating eosinophils (table S1 and fig. S1, F to H). (C) Thymocyte subsets were quantitated from coronin 1+/+ (open bars) and coronin 1–/– (closed bars) mice (n = 12). Representative FACS profiles are shown in fig. S1F. (D) Thymocytes and lymph node cells from female coronin 1+/+ and coronin 1–/– H-Y TCR+ rag2–/– (n = 5) and DO11.10 TCR+ (n = 9) mice were quantified.

The requirement for coronin in cell motility in D. discoideum prompted us to examine whether coronin 1 may play a role in thymic emigration and homing to secondary lymphoid organs. CD4+ coronin 1–/– thymocytes demonstrated reduced spontaneous migration and transwell migration to CCL19, CXCL12, and CCL25 (Fig. 2A). Defects in chemotaxis were also observed in splenic CD4+ naïve and effector/memory coronin 1–/– T cells (fig. S2B). Coronin 1–/– T cells also demonstrated compromised migration in whole-organ thymic cultures and in vivo thymic egress (fig. S2, C and D). Lastly, adoptive transfer of differentially labeled coronin 1–/– and coronin 1+/+ CD4+ thymocytes revealed ∼60% decreased homing of coronin 1–/– cells to lymph nodes (Fig. 3A). Thus, coronin 1 plays important functional roles in cell motility and chemokine-mediated homing of T lymphocytes to secondary lymphocyte organs.

Fig. 2.

Defective migration and homing of coronin 1–/– T cells. (A) Transwell migration of coronin 1+/+ (open bars) and coronin 1–/– (solid bars) CD4+ thymocytes in response to the indicated chemokines. Expression of their corresponding receptors was normal (fig. S2A). Error bars indicate standard deviation (SD) from duplicate cultures. (B) Naïve CD4+ DO11.10 TCR+ coronin 1+/+ and coronin 1–/– T cells, either unstimulated or stimulated for 30 min with CCL19 (250 ng/ml), were stained for coronin 1, talin, and F-actin (phalloidin) and analyzed by deconvolution microscopy (DeltaVision, RT, Applied Precision, LLC, Issaquah, WA). Yellow arrows point to talin clusters, and white arrows indicate the position of uropod formation. Similar to the dysregulated talin staining, phosphatidylinositol 3,4,5-trisphosphate [PI(3,4,5)P3] also accumulated asymmetrically in CCL19-stimulated coronin 1–/– T cells (fig. S3C). (C) DO11.10 TCR coronin 1+/+ and coronin 1–/– T cells, stimulated with CXCL12 for the indicated times, were analyzed for Rac activation (top). Cell extracts were also subjected to immunoblotting with Rac1 and coronin 1 monoclonal antibodies (mAbs). GTP, guanosine 5′-triphosphate.

Fig. 3.

Survival defect of coronin 1–/– naïve T cells. Error bars indicate SEM. (A) Equal numbers of differentially labeled CD4+ coronin 1+/+ (open bars) and coronin 1–/– (solid bars) thymocytes were transferred into wild-type mice. Cellular constituents were analyzed at the indicated times by FACS (n = 4). (B) Percentage of annexin V+ thymocytes from coronin 1+/+ and coronin 1–/– mice (n = 5). (C) Naïve CD4+ CD62Lhi T cells from coronin 1+/+ and coronin 1–/– mice were cultured at 37°C in normal medium or kept on ice for 18 hours. Annexin V+ cells were quantified by FACS staining using duplicate cultures. (D) Naïve coronin 1+/+ and coronin 1–/– T cells were cultured in normal medium for the indicated times and then lysed, and cell extracts were analyzed by immunoblotting with mAbs against cleaved caspase-3, β-actin, or coronin 1. (E) DO11.10 TCR+ coronin 1+/+ and coronin 1–/– cells were cultured at 37°C, and activation of caspase-9 was assessed by FACS staining. The % of cells with active caspase-9 [(fluorescein isothiocyanate) FITC-LEHD-FMK+] from triplicate cultures was quantified. (F) Purified naïve coronin 1+/+ and coronin 1–/– Tcells were cultured at 37°C, and postmitochondrial cytosolic fractions were analyzed by immunoblotting with cytochrome c, β-actin, and coronin 1 mAbs (fig. S5A). Densitometric analysis from two independent experiments shows cytochrome c release in arbitrary units after normalization against β-actin.

Because the actin cytoskeleton is required for cellular polarization and lymphocyte migration, we analyzed the morphologic changes induced by CCL19. Whereas stimulated coronin 1+/+ T cells acquired a polarized phenotype with unipolar accumulation of talin beneath the cell membrane opposite of the uropod, coronin 1–/– T cells failed to develop a uropod and formed multiple patch-like talin-rich clusters that were distributed irregularly around the cell cortex (Fig. 2B). Notably, coronin 1–/– T cells had increased amounts of F-actin. Despite enhanced basal F-actin formation, coronin 1–/– naïve T cells demonstrated lower degrees of CCL19-induced F-actin formation (fig. S3A). These cytoskeletal defects were associated with a selective defect in SDF1α-mediated Rac1 activation but not activation of Erk, Akt, or Rsk signaling pathways (Fig. 2C and fig. S3B). Although cytoskeletal reorganization is also requisite for TCR function, coronin 1–/– DO11.10 TCR+ T cells demonstrated normal antigen-induced proliferation and accumulated F-actin at the T cell–antigen-presenting cell (APC) interface (fig. S4, A to C). Thus, coronin 1 plays a critical role in chemokine-regulated uropod formation, talin polarization, and migration but is dispensable for TCR-mediated functions.

Although adoptively transferred coronin 1–/– T cells were compromised in their ability to migrate to lymph nodes, no compensatory increase of coronin 1–/– T cells was observed in the circulation (Fig. 3A), which suggested that loss of coronin 1 may also compromise cell survival. Correspondingly, coronin 1–/– mice had a higher percentage of annexin V+ CD4+ and CD8+ thymocytes, particularly in mature CD69lo cells (Fig. 3B). A similar enhancement in annexin V+ cells was observed in peripheral naïve T cells. This increased rate of in vivo apoptosis was cell intrinsic, because in vitro incubation of coronin 1–/– naïve T cells exhibited greater spontaneous cell death compared with coronin 1+/+ T cells (Fig. 3C). Increased apoptosis was associated with cleavage of caspases 3 and 9 and reversed by caspase inhibitors (Fig. 3, D and E). In addition, coronin 1–/– T cells demonstrated increased spontaneous release of cytochrome c to the cytoplasm (Fig. 3F). In contrast, effector/memory T and splenic B cells did not demonstrate any increase in spontaneous apoptosis (fig. S5B). Thus, coronin 1 is required for naïve T cell survival.

The linkage of actin dynamics to mitochondrial membrane potential (MMP), caspase activation, and cellular viability in yeast (17) and the interaction of coronins with the Arp2/3 complex (4, 18) prompted us to examine the effects of coronin 1 deficiency on actin polymerization and apoptosis. Fluorescence microscopy and fluorescence-activated cell sorting (FACS) analysis revealed enhanced basal F-actin and, conversely, decreased G-actin in coronin 1–/– T cells (Fig. 4, A and B); these effects were reversed by reexpression of wild-type coronin 1 (fig. S6S). Additionally, coronin 1–/– naïve T cells demonstrated a loss in MMP (Fig. 4C). Lastly, incubation with the actin depolymerizing agent latrunculin A partially reversed the loss in MMP and the increased apoptosis observed in coronin 1–/– T cells (Fig. 4, D and E). Conversely, consistent with studies in T cell lines (19), primary T cells treated with jasplakinolide, which decreases actin turnover, demonstrated increased MMP and increased spontaneous apoptosis (Fig. 4, D and E). Thus, actin dynamics in mammalian cells also controls cellular viability through a mitochondrial-dependent pathway.

Fig. 4.

Interaction of coronin 1 with Arp2/3 is required for normal actin dynamics. (A) Immunofluorescence microscopy of naïve coronin 1+/+ and coronin 1–/– T cells stained with phalloidin or coronin 1 mAb and analyzed by deconvolution microscopy (DeltaVision). (B) Cellular F- and G-actin content of naïve coronin 1+/+ (black) and coronin 1–/– (red) T cells was assessed by phalloidin and deoxyribonuclease I FACS staining, respectively. (C) Naïve CD4+ DO11.10 TCR+ coronin 1+/+ and coronin 1–/– T cells were cultured at 37°C in normal medium for the indicated times, and changes in the MMP were assessed. (D and E) Naïve coronin 1+/+ and coronin 1–/– T cells were preincubated on ice with 5 μg/ml latrunculin A, 1 μM jasplakinolide, or dimethyl sulfoxide (DMSO) carrier. Cells were cultured at 37°C for 2 hours, and loss of MMP (D) or annexin V+ (E) was quantified by FACS. Error bars indicate SD from duplicate cultures. (F) Wild-type (WT) and mutant coronin 1 were expressed in A20 cells. Lysates were immunoprecipitated with Flag mAbs and immunoblotted with antibodies against Arp2 and Flag. (G) Basal F-actin of DO11.10 TCR+ coronin 1–/– cell lines, transfected with the indicated coronin 1 internal ribosomal entry site (IRES)–green fluorescent protein (GFP) expression constructs, was assessed by staining with phalloidin and quantified by FACS. Results represent the change in mean fluorescent intensity (ΔMFI) between GFP and GFP+ cells. Protein expression was assessed in fig. S7. The ΔMFI between coronin 1–/– and coronin 1+/+ T cells was –122. Data are representative of two independent experiments.

To determine whether the interaction between coronin 1 and the Arp2/3 complex was required for controlling actin dynamics in T cells, we used a set of coronin 1 mutants, S2D and ΔCC, that were compromised in their ability to bind Arp2/3 (Fig. 4F) (18). Whereas expression of wild-type or Flag-tagged wild-type coronin 1 in coronin 1–/– T cells resulted in decreased phalloidin staining to levels nearing those of coronin 1+/+ T cells, expression of coronin 1 mutants (S2D and ΔCC) did not reverse the enhanced phalloidin staining in coronin 1–/– T cells (Fig. 4G and fig. S7). Thus, coronin 1 plays an inhibitory role in the steady-state F-actin equilibrium via an Arp2/3-dependent mechanism.

Consistent with an actin-regulatory role of coronin proteins, we have found a selective requirement for coronin 1 in chemokine-induced but not TCR-mediated functions. Our data also indicate an Arp2/3-dependent inhibitory function of coronin 1 on steady-state F-actin equilibrium. Two lines of in vitro experimentation provide the mechanistic basis for this disturbance in actin dynamics. Purified yeast coronin inhibits Arp2/3-mediated actin polymerization (4), and electron microscopy (EM) images indicate that coronin binds near p35 (ARPC2) to skew Arp2/3 to a more open and “inactive” conformation (20). In coronin 1–/– T cells, increased F-actin was associated with enhanced apoptosis and shortened survival that was partially reversed with the addition of actin depolarizing agents. This link between actin dynamics and cellular homeostasis may account for the decreased circulating T cell numbers observed in dock2–/– and wasp–/– mice and in WAS patients with dysregulated cytoskeletal rearrangements (2123).

A link between actin dynamics and cellular longevity has been demonstrated recently in yeast (17). Yeast with reduced actin dynamics exhibit accumulation of F-actin, loss of MMP, release of reactive oxygen species (ROS), and increased cell death. Conversely, yeast with increased actin dynamics exhibit decreased ROS, increased cellular viability, and prolonged yeast life span. In contrast to our observations on mammalian coronin 1, coronin-null mutants (crn1Δ) in yeast have no discernible phenotype. However, crn1Δ act1-159 and crn1Δ cof1-22 mutations accumulate a large filamentous actin mass and exhibit reduced cell growth (11). Together, these data indicate that the link between actin dynamics and cellular homeostasis exists in both yeast and mammalian cells. Perturbations in basal F-actin content and decrements in naïve T lymphocyte homeostasis have also been reported in aged mice and humans (24, 25). Our studies reinforce and expand our appreciation of the diversity of functions controlled by the actin cytoskeleton that range from the well-recognized functions of lymphocyte chemotaxis and function to also include lymphocyte survival.

Supporting Online Material

www.sciencemag.org/cgi/content/full/313/5788/839/DC1

Materials and Methods

Figs. S1 to S7

Table S1

References and Notes

View Abstract

Navigate This Article