| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Departments of
* Medicine and
Pathology and Immunology, Washington University School of Medicine, St. Louis, MO 63110; and
Program in Molecular Pathogenesis, Skirball Institute of Biomolecular Medicine, New York, NY 10016
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResults and DiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResults and DiscussionReferences |
|---|
During locomotion, T cells form a distinct cellular structure at the trailing edge of the cell termed the uropod. Many proteins, including CD43, localize to the uropod. The leading edge of the cell has significantly increased sensitivity to Ag, despite similar densities of TCR (14). Relocalization of CD43 to the uropod is, therefore, assumed to support both T cell adhesion and activation by minimizing steric hindrance at the leading edge. However, this model of CD43 function, as well as the requirement for exclusion of CD43 from the T cell-APC contact zone, has not been formally tested.
The localization of CD43 to the cellular uropod in migrating lymphocytes has been thought to be dependent on the interaction of cytoplasmic domain of CD43 with the adapter protein ezrin, a member of the ezrin, radixin, and moesin (ERM) family (15, 16). ERM proteins link transmembrane proteins to actin by either direct or indirect interactions (17, 18). Thus, the anchoring of actin filaments to CD43 may be important in regulating cellular processes such as the establishment of cell polarity or cell motility, or in forming an initiating nucleus for the assembly of signal-transducing complexes (19).
We formally evaluated the requirement for CD43 redistribution for T cell activation, proliferation, and motility. Reconstitution of CD43-deficient primary T cells by retroviral gene transfer with specific mutants of CD43 demonstrates that the failure of CD43 to redistribute from the T cell-APC contact zone does not inhibit formation of an immunological synapse or proliferation in response to Ag. In addition, the redistribution of CD43 to the cellular uropod is not required for T cell locomotion. Taken together, our findings suggest a function of CD43 other than steric hindrance in inhibition of T cell function.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResults and DiscussionReferences |
|---|
CD43-deficient mice were backcrossed to the DO11.10 TCR-transgenic background (provided by K. Murphy, Washington University, St. Louis, MO (20)) for seven generations. BALB/c mice were purchased from The Jackson Laboratory (Bar Harbor, ME). Animals were maintained in specific pathogen-free housing at Washington University School of Medicine. All protocols have been approved by the Animal Studies Committee at Washington University School of Medicine.
Antibodies
Anti-CD43 hybridoma (S7, rat IgG2a) was provided by J. G.
Frelinger (Rochester University, Rochester, NY). Anti-TCR
hybridoma
(H57-597, Armenian hamster IgG) was provided by P. M. Allen
(Washington University). All other mAb were purchased from BD
PharMingen (San Diego, CA). F(ab')2 were
generated using Pierce Immunopure F(ab')2 kit
(Pierce, Rockford, IL). Abs and F(ab')2 were
conjugated to fluorochromes (CyChrome 3, CyChrome 5 (Cy5), and
Alexa 488; Molecular Probes, Eugene, OR).
Constructs
Full-length murine CD43 cDNA was cloned into the retroviral vector GFPRV (21) (provided by W. Sha, University of California, San Francisco, CA) or tailless human CD4RV (hCD4RV, provided by K. Murphy). All mutations were introduced by oligonucleotide site-directed mutagenesis using the Quickchange kit (Stratagene, La Jolla, CA). The CD43/CD102 chimera was made by overlap extension PCR (CD102 cDNA provided by T. Springer, Center for Blood Research, Boston, MA). Sequences were confirmed by automated fluorescent sequence analysis (Big Dye; Applied Biosystems, Lincoln, CA).
Retroviral infections
Retroviral vectors containing CD43 were transiently transfected into the Phoenix Eco packaging cell line (provided by G. Nolan, Stanford University, Palo Alto, CA) and T cells infected with retroviral supernatant as previously described (22). Expression of CD43 was confirmed in all experiments by flow cytometry. When indicated, T cells retrovirally infected with the hCD4RV constructs were enriched using immunomagnetic cell sorting and anti-human CD4 beads with an AutoMACS sorter (Miltenyi Biotec, Auburn, CA).
APC-T cell interaction
DO11.10 CD43-/- T cells were infected with retrovirus containing CD43 with the indicated mutations. L cells transfected with I-Ad and ICAM-1 (provided by A. S. Shaw, Washington University) were cultured in complete medium with 10 µm OVA323–339 peptide on coverslips for 24 h. Heated (37°C) FCS2 units (Bioptechs, Butler, PA) were assembled with coverslips to which L cells were adhered, and retrovirally infected T cells stained with conjugated Abs and/or F(ab')2 were injected into the units. Digitized three-color fluorescence images were obtained using a Zeiss Axiovert LSM 510 confocal microscope (Zeiss, Thornwood, NY) as described previously (23). Z-stacks were performed by acquiring 30 images at optimal confocal planes. False color image analysis was performed using LSM 510 software (Zeiss). Quantification of the fluorescence in T cell:APC contacts was done using National Institutes of Health ImageJ software (National Institutes of Health, Bethesda, MD). Nine independent readings of fluorescence intensity were taken over both the area of the T cell:L cell contact and over noncontact areas of the T cell. Three separate cells expressing each contact were analyzed. Presented is the mean fluorescence of the contact and noncontact areas for each cell ± SEM.
Proliferation assays
DO11.10 CD43-/- T cells were infected with retrovirus encoding for either full-length CD43 or mutant CD43 proteins and 2.5 x 105 T cells cocultured with 1.25 x 106 irradiated BALB/c splenocytes in complete medium. Graded doses of OVA323–339 peptide were added and proliferation was determined by tritiated thymidine incorporation for the final 8 h of a 42-h culture. All conditions were plated in quadruplicate and the mean ± SD of the quadruplicate wells was presented.
T cell motility assay
Supported planar lipid bilayers containing Cy5-conjugated GPI-ICAM-1 were prepared as previously described (1). Ab- and F(ab')2-stained retrovirally infected T cells were injected into a heated FCS2 chamber (Bioptechs) and allowed to adhere, and data were acquired for 360 s using the Zeiss LSM 510 confocal microscope. Data analysis of trajectory and velocity of T cells was performed using National Institutes of Health ImageJ analysis software. Two-tailed t tests were performed to determine the statistical significance of any differences.
Statistical analysis
Statistical significance of the data obtained from motility studies and fluorescence intensity was determined by an unpaired two-tailed t test using Excel software (Microsoft, Seattle, WA).
|
Results and Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResults and DiscussionReferences |
|---|
To determine the domain within CD43 that regulates its
redistribution upon T cell polarization, we reconstituted primary T
cells from CD43-deficient mice with mutant CD43 protein using
retroviral gene transduction. Wild-type or mutant CD43 constructs (Fig. 1
a) were cloned into the
retroviral vector GFPRV, which encodes for a bicistronic message
allowing for expression of green fluorescent protein (GFP) and CD43.
Flow cytometric analysis demonstrated that the expression of the
retrovirally encoded CD43 protein was at similar levels to the
endogenous protein, with the exception of the CD43
cyto (Fig. 1b). The CD43cyto construct consistently had lower levels
of expression, suggesting that the stability of the expressed protein
may be decreased due to deletion of the cytoplasmic domain. The
retrovirally expressed CD43 was recognized normally by a variety of
anti-CD43 mAb (Fig. 1 and data not shown), some of which recognize
carbohydrate-dependent epitopes, suggesting normal glycosylation of the
retrovirally encoded CD43. In addition, as we used primary T cells
derived from mice in which the CD43 gene has been deleted by homologous
recombination, it is unlikely that there are defects in the glycosyl
transferase enzymes that mediate posttranslational modification of
CD43. Therefore, it is likely that the retrovirally expressed CD43 is
glycosylated and sialylated in a manner similar to endogenous
CD43.
|
a). In these cells, the
entire amount of surface CD43 is concentrated in the uropod, leading to
a higher local density of CD43. Within the membrane proximal region,
the amino acid sequence KRR, at position 276–278, has been
demonstrated to mediate interaction of CD43 with the cytoskeletal
adapter protein ezrin (24). Consistently, mutation of the
KRR motif to NGG within the full-length cytoplasmic tail of CD43
abrogated the redistribution of CD43 to the uropod. False color image
analysis of the fluorescence intensity demonstrated high expression of
CD43 throughout the cell in the nonredistributing mutants (Fig. 2a). In both the CD43NGG- and CD43cyto-expressing cells,
there is some suggestion of nonuniform clustering of CD43; however, the
significance of this is unclear. Expression of a chimeric protein
consisting of CD43 extracellular and transmembrane domain fused with
the cytoplasmic portion of ICAM-2 (CD102), which also contains a
membrane proximal ERM-binding site, restored the polar localization of
the chimeric receptor. Analysis of the percentage of cells that form
uropods revealed no difference in the ability of the cells to form a
uropod based on the expression of mutant CD43. However, virtually all
of the cells that did form uropods redistributed CD43, with the
exception of the CD43cyto and CD43NGG constructs, in which
essentially none of the cells redistributed CD43 (Table I). The number of cells forming uropods
in cells expressing the CD43 STOP1–4 constructs were similar to wild
type (data not shown). Thus, the redistribution of CD43 on the cell
surface is an active process requiring a specific interaction of the
cytoplasmic domain of CD43 with intracellular proteins.
|
|
b, blue staining) at the uropod indicates that
the failure to redistribute CD43 was not the result of a global defect
in surface protein reorganization. Presence of CD43 in the T cell-APC contact zones does not affect T cell proliferation
We evaluated whether expression of nonredistributing CD43 mutants
would prevent the formation of an immunological synapse and/or inhibit
the T cell proliferative response to Ag. CD43 constructs were expressed
in DO11.10 TCR-transgenic CD43-deficient T cells by retroviral gene
transfer. Infected cells were sorted by immunomagnetic selection and
CD43 expression was confirmed by flow cytometry (data not shown).
Sorted T cells were cocultured with ICAM-1- and
I-Ad-transfected L cells previously loaded with
OVA323–339 peptide. T cell-APC contact and CD43
exclusion were monitored by confocal microscopy. T cells infected with
virus encoding full-length CD43 (CD43FL) excluded CD43 from cell-cell
contacts, as did CD43+/+ controls (Fig. 3a and data not shown). Cells
expressing CD43/CD102 also redistributed the chimeric protein to areas
of no cellular contact, consistent with ezrin redistribution (Fig. 3a and data not shown). The CD43NGG and CD43cyto mutants
failed to exclude CD43 from the site of contact (Fig. 3a)
without affecting the ability of the T cell to generate sustained
cSMACs. All cells expressing CD43FL or CD43/CD102 that formed a stable
contact with the APC redistributed CD43, whereas all CD43NGG and
CD43cyto cells failed to redistribute. Quantification of the
intensity of CD43 staining at both the T cell:L cell contact and the
region of the T cell not in contact with the L cell confirmed a failure
to redistribute CD43 in the CD43cyto-expressing cells (Fig. 3b). In addition, there was no difference in the intensity
of CD43 staining in the cSMAC as compared with the pSMAC.
Interestingly, there was partial exclusion of CD43 from the contact
region in cells expressing CD43NGG. Previous studies have demonstrated
that there is a second ERM protein binding site in the tail of CD43
(24). This site is preserved in the CD43NGG mutant and may
mediate the movement of this protein from the contact. Nonetheless,
neither the CD43cyto nor the CD43NGG constructs were excluded to the
same degree as the full-length CD43 or CD43/CD102 constructs, which
were virtually completely redistributed away from the T cell:L cell
contact. Both the CD43cyto- and CD43NGG-expressing cells formed a
stable immunologic synapse, suggesting that, despite the large size and
negative charge of the protein, redistribution away from the region of
cell:cell contact is not required for formation of this structure.
Interestingly, these observations are in contrast to data obtained by
Wild et al. (26) which determined that extension of
another protein, CD48, can inhibit T cell recognition of Ag, presumably
by disrupting close apposition of cell membranes, thus disrupting
long-term protein-ligand interaction, which would otherwise lead to
cellular activation.
|
c). Thus, the presence of
CD43 at the site of cell-cell contact does not inhibit functional TCR
engagement. Consistent with the finding that large molecules can be
resident among smaller-sized proteins without necessarily disrupting
formation of an immunological synapse, CD45, another large molecule,
was found to actively migrate into cSMACs during immunological synapse
formation (23). The ability of TCR engagement to occur in the absence of CD43 redistribution is supported by a recent series of papers examining the role of ERM proteins in the segregation of T cell surface proteins during T cell activation (27, 28, 29). Delon et al. (28) demonstrated that CD43 interacted with moesin and that dephosphorylation of moesin resulted in the release of CD43. The dephosphorylation occurred in response to TCR signaling and led to the release of CD43 from the cytoskeleton, which permitted an initial passive exclusion of the molecule from the contact. Re-anchoring of CD43 to the cytoskeleton was then required for complete exclusion. However, the initial release required engagement of the TCR, suggesting that effective TCR engagement occurs before redistribution of CD43. Furthermore, TCR-induced Ca2+ flux was normal in cells expressing a nonredistributing mutant of CD43. Similarly, Roumier et al. (27) found redistribution of ezrin to occur in response to TCR signaling and to depend upon Lck activity. Allenspach et al. (29) found the predominant ERM protein interacting with CD43 to be ezrin. Expression of a GPI-linked CD43 ectodomain was not excluded from the T cell:APC contact, yet CD69 was normally up-regulated, indicating T cell activation. Interestingly, both groups demonstrate a reduction in IL-2 secretion by cells expressing nonredistributing mutants of CD43, whereas we find that proliferation is normal. Thus, it may be that while initial engagement is normal in the absence of CD43 redistribution, prolonged signaling is not maintained. However, our data suggest that this defect is not sufficient to lead to a decrease in proliferative responses.
These data suggest that cell-cell interactions should not be regarded purely as flat and rigid membrane interactions but rather as highly active undulating processes in which differently sized molecules can be in relative proximity without sterically hindering one another. Alternatively, molecules like CD43 may display considerable flexibility at physiological ion concentrations, such that they are accommodated in 15-nm contact regions formed at the site of TCR:MHC interactions. The electron microscopy studies that show CD43 as a rigid rod were performed at low ionic strength, where the low dielectric constant reduces the shielding between sialic acids. The sialic acids then strongly repel each other, resulting in a fully extended conformation where the distance between sialic acids is maximal (10). At physiological ionic strengths the dielectric constant is higher and the repulsion between sialic acids is lower. Under these conditions CD43 should be much more compressible and may coexist with the TCR contacts where the relatively slow (seconds to minutes) contact formation process allows time for conformational changes in CD43 leading to interdigitation of CD43 chains in the 15-nm contact area (30). The negative role of CD43 in adhesive interactions with endothelial cells in flow conditions (31) may be more related to the very rapid kinetics of these interactions that do not allow time for the CD43 to move out of the way of interacting receptors.
These studies demonstrate that during a physiologically relevant interaction, i.e., T cell with an APC, the presence of CD43 in the cSMAC does not inhibit T cell proliferation. However, it remains a formal possibility that the kinetics of cell activation could be altered by the presence of CD43 in the cSMAC. Changes in the time course of effective TCR engagement would not be evident by measurement of proliferation. However, these data do establish that failure to exclude CD43 from the T cell-APC contact does not preclude effective T cell activation.
Presence of CD43 in the adhesion contact zone does not inhibit T cell motility but affects de-adhesion
Polarization of T cells occurs during cell migration, resulting in
the redistribution of CD43 to the uropod. The leading edge, which is
depleted in CD43, is the site at which initial adhesion receptor
engagement occurs. The activity of both adhesion receptors and the TCR
is increased in this area of the cell (14, 32). This has
been thought to be due in part to the removal of large proteins such as
CD43, which may sterically inhibit the interaction of these receptors
with ligand. To test this hypothesis, primary T cells expressing either
wild-type or mutant CD43 were placed on a supported planar lipid
bilayer containing GPI-anchored ICAM-1. Cross-sectional reconstituted
images (Z-stack) obtained by confocal microscopy confirmed the presence
of CD43 at the cell-bilayer interface (Fig. 4). Time lapse analysis of migrating
cells demonstrated that expression of nonredistributing mutant of CD43
did not impair T cell trajectory (Fig. 4a) and/or velocity
(Fig. 4b) compared with redistributing constructs. Thus, the
redistribution of wild-type CD43 to the uropod, while a consequence of
the movement of ezrin to the uropod (16), which itself may
or may not have functional consequences, is not required for effective
engagement of adhesion receptors by their ligands.
|
). Motile cells expressing CD43NGG had a
striking phenotype of long residual strands of membrane trailing the
cell that remained adherent to the bilayer which was not observed in
CD43cyto-expressing cells. These stained heavily for CD43, and the
persistent adherence of the strands did not inhibit migration velocity
over the 6-min observation period. IRM images obtained in conjunction
with Z-stack reconstitution of retrovirally infected cells revealed
another striking observation. The uropods of cells expressing CD43NGG
and CD43cyto were in full contact with the bilayer compared with
endogenous CD43-, CD43FL-, and CD43/CD102-expressing cells, which
clearly elevated uropods from the bilayer (Fig. 5, arrowheads in
Z-stack and IRM panels).
|
cyto and the CD43NGG mutations resulted in contact of
the uropod with the bilayer. However, the trailing strands were
observed only in the CD43NGG mutation. Thus, the specific loss of the
ezrin binding site in CD43, while keeping the remainder of the
cytoplasmic interactions intact, resulted in this phenotype. This
suggests that the binding of a distinct cytoplasmic protein to CD43 in
the absence of ezrin binding to the proximal tail is responsible or,
alternatively, that the introduction of the NGG mutation resulted in
new protein-protein interaction with CD43. The precise function of CD43 remains controversial (9). The most consistent observation has been that CD43-deficient cells have increased adhesion. This has been demonstrated both in vitro and in vivo (12, 13, 31, 33). The inhibitory function of CD43 on cell adhesion has been presumed to be mediated by the structural features of the extracellular domain, resulting in the steric hindrance of receptor ligand interactions. The data presented in this work suggest that this mechanism is not a tenable explanation. Disruptions in cytoskeletal associations or other signaling pathways could result in the alterations in the regulation of cell adhesion molecule activity. These and other possibilities remain to be explored as we move beyond the steric barrier model of CD43 function.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
2 Address correspondence and reprint requests to Dr. Jonathan M. Green, Washington University School of Medicine, 660 South Euclid Avenue, Box 8052, St. Louis, MO 63110. E-mail address: greenj{at}msnotes.wustl.edu
3 Abbreviations used in this paper: cSMAC, central supramolecular activation complex; pSMAC, peripheral SMAC; ERM, ezrin, radixin, and moesin; Cy5, CyChrome 5; GFP, green fluorescent protein.
Received for publication December 11, 2001. Accepted for publication February 11, 2002.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResults and DiscussionReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  M. L. Dustin Modular Design of Immunological Synapses and Kinapses Cold Spring Harb Perspect Biol, July 1, 2009; 1(1): a002873 - a002873. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. L. Cannon, A. Collins, P. D. Mody, D. Balachandran, K. J. Henriksen, C. E. Smith, J. Tong, B. S. Clay, S. D. Miller, and A. I. Sperling CD43 Regulates Th2 Differentiation and Inflammation J. Immunol., June 1, 2008; 180(11): 7385 - 7393. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. G. Garcia, A. A. S. Akha, and R. A. Miller Age-Related Defects in Moesin/Ezrin Cytoskeletal Signals in Mouse CD4 T Cells J. Immunol., November 15, 2007; 179(10): 6403 - 6409. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Y. Huang and J. K. Burkhardt T-cell-receptor-dependent actin regulatory mechanisms J. Cell Sci., March 1, 2007; 120(5): 723 - 730. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 B. J. Evans, A. McDowall, P. C. Taylor, N. Hogg, D. O. Haskard, and R. C. Landis Shedding of lymphocyte function-associated antigen-1 (LFA-1) in a human inflammatory response Blood, May 1, 2006; 107(9): 3593 - 3599. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. K. Randhawa, H. J. Ziltener, J. S. Merzaban, and R. W. Stokes CD43 Is Required for Optimal Growth Inhibition of Mycobacterium tuberculosis in Macrophages and in Mice J. Immunol., August 1, 2005; 175(3): 1805 - 1812. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. J. Fernandez-Cabezudo, C. Vijayasarathy, D. L. Pflugh, A. L. M. Bothwell, and B. K. al-Ramadi Evidence for a dual pathway of activation in CD43-stimulated Th2 cells: differential requirement for the Lck tyrosine kinase Int. Immunol., August 1, 2004; 16(8): 1215 - 1223. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Kyoizumi, T. Ohara, Y. Kusunoki, T. Hayashi, K. Koyama, and N. Tsuyama Expression Characteristics and Stimulatory Functions of CD43 in Human CD4+ Memory T Cells: Analysis Using a Monoclonal Antibody to CD43 That Has a Novel Lineage Specificity J. Immunol., June 15, 2004; 172(12): 7246 - 7253. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. Tong, E. J. Allenspach, S. M. Takahashi, P. D. Mody, C. Park, J. K. Burkhardt, and A. I. Sperling CD43 Regulation of T Cell Activation Is Not through Steric Inhibition of T Cell-APC Interactions but through an Intracellular Mechanism J. Exp. Med., May 3, 2004; 199(9): 1277 - 1283. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
F. E. McCann, B. Vanherberghen, K. Eleme, L. M. Carlin, R. J. Newsam, D. Goulding, and D. M. Davis The Size of the Synaptic Cleft and Distinct Distributions of Filamentous Actin, Ezrin, CD43, and CD45 at Activating and Inhibitory Human NK Cell Immune Synapses J. Immunol., March 15, 2003; 170(6): 2862 - 2870. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. Barat and M. J. Tremblay Engagement of CD43 Enhances Human Immunodeficiency Virus Type 1 Transcriptional Activity and Virus Production That Is Induced upon TCR/CD3 Stimulation J. Biol. Chem., August 2, 2002; 277(32): 28714 - 28724. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
