| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
University of Cambridge School of Clinical Medicine, Department of Medicine, Cambridge, United Kingdom
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
2-microglobulin
(2m),3
is to present peptides for recognition by CD8+ T
lymphocytes, much work has concentrated on identifying B27-restricted
CTL in spondyloarthropathy patients (4, 5). However,
studies with B27+ transgenic mice have raised the
possibility that the role of B27 in joint disease may not simply be as
a restriction element for CD8+ T cells.
CD8+ T cells do not appear to be pivotal in
disease induction because B27+ mice lacking
2m
(2m-/-) with extremely
low levels of CD8+ T cells still develop
inflammatory disease. Interestingly,
2m-/- mice made
transgenic for both B27 and human 2m also
develop disease, and in these mice disease is independent of the
TAP-1 gene, which is required for efficient loading
of peptide onto MHC class I molecules (6). In addition,
adoptive transfer studies from disease-prone B27+
transgenic rats to B27+ nude rats identified that
CD4+ T cells were more efficient in transferring
inflammatory disease (7). However, in some circumstances
MHC class II molecules, which are conventionally recognized by
CD4+ T cells, are not required for disease
development because MHC class II-negative, B27+
transgenic mice still develop spontaneous disease (8),
discounting the presentation of a B27-derived peptide by class II
molecules. One possibility to explain these findings is that B27 itself
may be recognized by CD4+ T cells.
It is also possible that forms of B27 other than the conventional H
chain/2m heterodimer may be important in
disease. Support for this idea came from the finding that disease in
B27+
2m-/- transgenic mice,
which express free B27 H chains not associated with
2m, could be ameliorated by in vivo treatment
with an Ab (HC10) that recognizes free H chains (8). In
addition, a variety of unusual forms of B27 occur on the surface of
cells in vitro; empty B27 heterodimers, not containing a peptide, are
present on the surface of C1R transfected with HLA-B*2705 (C1R-B27)
(9), free B27 H chains, i.e., not associated with
2m, form on tapasin-defective cell lines
(10), and free B27 H chains have been shown to form
disulfide-bonded homodimers, dependent on the cysteine residue at
position 67 in the peptide binding groove, in TAP-deficient cell lines
(11).
To investigate the possibility that CD4+ T lymphocytes might interact with B27, we have cultured highly purified human CD4+ T cells from B27+ spondyloarthropathy patients and healthy individuals with the MHC class II-negative, TAP-negative cell line T2 transfected with B27. T2 cell line transfected with HLA-B*2705 (T2-B27) expresses a variety of forms of B27 including B27 heterodimers, free B27 H chain monomers, and homodimers. Using this coculture system we were able to isolate CD4+ T cells that recognize B27 on cell lines deficient in their Ag-processing pathways for MHC class I Ags. The form of B27 recognized appears to be dependent on mutations in genes in the MHC because the CD4+ T cells failed to respond to B27+ EBV-transformed lymphoblastoid cell lines (EBV LCLs). If unusual forms of B27 are expressed in vivo, or can be induced by factors such as intracellular infection, these B27-reactive CD4+ T cells may play a role in the pathogenesis of spondyloarthropathies.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
PBMC were isolated from the blood of six B27+ ankylosing spondylitis (AS) patients and two B27+ healthy individuals by Ficoll-Paque (Pharmacia, St. Albans, U.K.) gradient centrifugation. Monocytes were removed from PBMC by adherence. Nonadherent cells were incubated with mAbs specific for CD8, CD19, CD11b, and CD16 for 30 min at 4°C (DAKO, Ely, U.K.). Labeled cells were removed using sheep anti-mouse IgG-coated magnetic beads (Dynal Biotech, Oslo, Norway). Three rounds of negative selection resulted in the isolation of CD4+ T cell populations of 95% purity, as determined by flow cytometric analysis.
Cell lines
The following cell lines were used: T2 (TAP1 and 2 negative, and MHC class II negative) (12); 721.220 (220; HLA-A and -B negative, tapasin mutant) (13, 14); C1R (hemizygous for HLA complex) (15); and the monocyte-derived cell line U-937 (16). This panel of cell lines was transfected with HLA-B*2705. In addition, 220 transfected with HLA-B8 was used. EBV LCLs were produced from PBMC as previously described (17).
Coculture of CD4+ T lymphocytes with B27-specific T cell lines
Irradiated (6 Gray) T2-B27 cells (5 x 103) were added to each 200-µl well of a 96-well plate in complete medium (RPMI 1640 (Life Technologies, Paisley, U.K.), 5% human serum, 25 mM HEPES, 2 mM glutamine, 100 U/ml penicillin, and 100 µg/ml streptomycin sulfate). Between 1 x 103 and 3 x 104 purified CD4+ T lymphocytes were added per well to the stimulator cells. The cells were incubated for 6 days at 37°C before rIL-2 (10 U/ml) was added twice weekly thereafter (Chiron, Harefield, Middlesex, U.K.). T cells lines were restimulated with irradiated stimulator cells as required. CD4+ T cell lines were further subcloned by limiting dilution in 20 µl Terasaki wells (Nunc, Roskilde, Denmark) using 106/ml irradiated allogenic PBMC as APCs, PHA, and 100 U/ml rIL-2.
T cell proliferation assays
Proliferation assays were performed at 37°C by culturing 2 x 104 T cells with 2.5 x 104 irradiated stimulator cells per well in 96-well plates. Plates were cultured for 2 days, then pulsed with [3H]thymidine (1 µCi/well; Amersham Pharmacia Biotech, Little Chalfont, U.K.). After culture for 6 h, cells were harvested, and [3H]thymidine incorporation was analyzed using a beta plate counter (Wallace, Turku, Finland). T cell proliferation was determined by subtracting background [3H]thymidine incorporation by stimulator cell lines from total [3H]thymidine incorporation.
The following mAbs were used in inhibition studies at a final concentration of 20 µg/ml; HLA-A, -B, and -C specific mAb w6/32 (18); HLA-DR specific mAb L243 (19); the free HLA-B and -C H chain specific mAb HC10 (20) (a gift from H. Ploegh, Harvard University, Boston, MA); and the CD4 specific mAb Campath-9H (a gift from J. Issacs, University of Leeds, Leeds, U.K.).
Flow cytometry
FITC-conjugated anti-CD3, anti-CD8, anti-CD16 (DAKO), and anti-CD4 were used to determine the phenotype of purified T cells. Staining with nonconjugated MHC specific mAbs L243, w6/32, ME1, and HC10 was detected using anti-mouse IgG-FITC. Flow cytometry data were collected on a FACSort (BD Biosciences) and analyzed using WinMDI.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
To investigate whether CD4+ T lymphocytes
can interact with B27 we cultured highly purified
CD4+ T cells, isolated from the PBMC of
B27+ spondyloarthropathy patients and
B27+ healthy controls, with the cell line T2-B27.
T2-B27 was initially chosen as it does not express MHC class II
molecules preventing the stimulation of MHC class II-reactive
CD4+ T cells. In addition, this TAP-negative cell
line expresses a variety of different forms of B27, including free B27
H chain monomers (shown by the weak but definite staining with HC10),
homodimers, and low levels of B27 heterodimers (Fig. 1
) (11, 21). Using this
coculture system we isolated a CD4+ T cell line
from a patient that proliferated in response to T2-B27, 220 transfected
with HLA-B*2705 (220-B27), and C1R-B27 but not to the nontransfected
parental cell lines T2, 220, and C1R (Fig. 2a). Because B27 (B*2705) is
the only HLA molecule shared among the three stimulatory cell lines
that is not expressed on the nontransfected parental cell lines, B27
must be the molecule recognized by the CD4+ T
cells (Fig. 1). In addition, the response was not cross-reactive with
other HLA-B alleles as demonstrated by the inability of the
CD4+ T cells to respond to HLA-B8 expressed on
220-B8 (see Fig. 5). At this stage the CD4+ T
cell line isolated from patient 1 contained 97%
CD4+ T cells and a small contaminant population
of CD8+ T cells (Fig. 2b). Subcloning
this line resulted in the isolation of a
CD4+CD8- T cell clone
(Fig. 2b) with the same pattern of recognition (Fig. 2a). Flow cytometry analysis showed the T cell clone
expressed an 
heterodimer TCR on the cell surface. Sequencing of
the TCR - and -chains confirmed the expression of the BV2 and AV2
TCR genes. In addition, further culture of the line resulted in the
loss of CD8+ T cells, and the line was then shown
to be clonal by TCR analysis.
|
|
|
). These lines showed a lower but statistically significant response
to C1R-B27; in later experiments it was found that increasing the
numbers of C1R-B27 stimulator cells enhanced responses of B27-reactive
T cell lines due to improved T cell-stimulator cell contact.
Nevertheless, it is noteworthy that the responses were inversely
correlated with B27 expression as detected by ME1, i.e., T2-B27 was
most stimulatory and C1R-B27 least, suggesting that a novel form of B27
might be being recognized. Abs specific for class I MHC Ags inhibit the response by CD4+ T cells
To identify the surface molecules recognized by the cross-reactive
CD4+ T cells from patient 1, the ability of mAbs
specific for MHC molecules to inhibit the proliferative responses of
the CD4+ T cell clone to T2-B27 was tested (Fig. 3). The mAb w6/32, which recognizes MHC
class I heterodimers associated with 2m and
free B27 H chain homodimers, inhibited the proliferative response of
CD4+ T cells to T2-B27. In addition, the response
to T2-B27 was inhibited by the HLA-B7, -B27 specific mAb ME1. Because
HLA-B7 is not expressed on T2-B27, B27 is the only molecule available
for ME1 binding. No inhibition was observed using the HLA-DR specific
mAb L243 or the free HLA-B and -C H chain specific mAb HC10. It is
difficult to interpret the lack of inhibition by HC10, as the ability
of this mAb to inhibit T cell responses in an in vitro system has not
previously been determined. The response to 220-B27 and C1R-B27 was
also inhibited by w6/32 and ME1 (data not shown). Because T2-B27 does
not express MHC class II molecules and L243 did not inhibit responses
of the CD4+ T cells to the MHC class II-positive
cell lines 220-B27 and C1R-B27 (data not shown), MHC class II molecules
do not appear to be involved in the recognition of these three cell
lines, excluding the possibility of recognition of a B27-derived
peptide on MHC class II.
|
To determine whether the CD4+
T cell response to B27 was dependent on the CD4 coreceptor, as occurs
with a conventional CD4+ T cell response to MHC
class II molecules, the effects of a CD4-specific mAb on responses to
T2-B27 were tested. The CD4+ T cell response to
T2-B27 was inhibited by 64% by the CD4 mAb (Fig. 4a), similar to the inhibition
demonstrated using a conventional MHC class II-restricted
CD4+ T cell clone (Fig. 4b). The
inability of the CD4 Ab to inhibit a response to PHA (Fig. 4) suggests
that CD4-independent responses are not affected by this mAb. A similar
pattern of inhibition was observed with 220-B27 and C1R-B27 as
stimulator cells (data not shown).
|
To determine whether the form of B27 recognized by the
CD4+ T cells was expressed in vivo and on cell
lines with intact Ag-processing pathways, the ability of the
CD4+ T cells to respond
B27+ EBV LCL and U937 transfected with HLA-B27
was tested. The B27-reactive CD4+ T cells failed
to respond to EBV LCL derived from B27+ patients,
B27+ healthy individuals, or to the
monocytic-derived cell line U937-B27 (Fig. 5), suggesting that the form of B27
recognized by the CD4+ T cells is not present on
EBV LCL and cell lines with intact Ag-processing pathways.
The form of HLA-B27 recognized on T2-B27 is stabilized at 26°C
Incubation of TAP-negative cells at 26°C has been shown
to stabilize expression of MHC class I on the cell surface, including
expression of empty heterodimers associated with
2m. Incubation of T2-B27 at 26°C for 24
h resulted in an increase of ME1 staining and w6/32 staining compared
with incubation of T2-B27 at 37°C (Fig. 6, a and b). This
reduction of temperature had no stabilizing effect on MHC expression on
nontransfected T2, indicating that the MHC class I molecule stabilized
on T2-B27 was B27. When CD4+ T cells responses to
incubation of T2-B27 at 37°C and 26°C were compared, a 2-fold
increase in proliferation was observed using T2-B27 incubated at 26°C
(Fig. 6c) suggesting that the form of B27 recognized is
stabilized on T2-B27 by incubation at 26°C.
|
To investigate whether the B27 free H chain disulfide-bonded
homodimers were recognized, the ability of the
CD4+ T cell to proliferate in response to C1R
transfected with B27 containing a substitution of cysteine at position
67 to serine (C1R-C67S-B27) was tested. Although the chosen C67S
transfectant expressed B27 at a lower level than C1R-B27, B27
expression was equivalent to levels on 220-B27 and was higher than
levels found on T2-B27 (Fig. 7a). The inability of the
CD4+ T cells to proliferate in response to
C1R-C67S-B27 (Fig. 7a) implies that the form of B27
recognized is dependent on cysteine at position 67.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Although it is well described that CD4+ T cells recognize peptide presented by MHC class II molecules, we have provided clear evidence that B27 is the molecule recognized by the CD4+ T cells. This is, to our knowledge, the first time that B27-reactive CD4+ T cells have been identified. In agreement to the identification of this unusual interaction between an MHC class I molecule and CD4+ T cells, several studies have previously isolated HLA-A2-restricted CD4+ T cells from humans (23, 24). Ab inhibition with w6/32 confirmed that these CD4+ T cells were MHC class I restricted. In one case the CD4 coreceptor appeared to play a role in this interaction (23). MHC class I-restricted CD4+ T cells have also been identified in mice; one subset uses an invariant TCR and recognizes a glycolipid presented by CD1d, but others appear to be CD1 independent, including cells capable of inducing gut inflammation, a feature of some B27-associated spondyloarthropathies (25, 26).
There are a number of forms of B27 that can be expressed on the cell surface of the stimulator cells which were recognized that could interact with the CD4+ T cells; these include B27 heterodimers or free B27 H chains expressed as monomer or homodimer forms. B27 heterodimers, which are detected by the mAbs ME1 and w6/32, are expressed on all three cell lines. The pattern of responses seen could be explained if the B27 heterodimers from transfected T2, 220, and C1R cell lines contain a peptide that is absent from B27+ EBV LCL or U937-B27. Alternatively, B27 heterodimers devoid of peptide, expressed on all the stimulatory APC lines (9, 10, 21, 27), may be recognized by the CD4+ T cells. These "empty" B27 molecules are also reactive with w6/32 and ME1, consistent with the strong inhibition of proliferative responses shown by these Abs. Incubation of T2 cell lines at 26°C has been shown previously to stabilize B27 molecules that do not contain an antigenic peptide (28, 29, 30). Thus, increased responses by the CD4+ T cell clone to T2-B27 incubated at this temperature may indicate that this form of B27 is being recognized.
Free B27 H chains may also be a potential target of the CD4+ T cells because they are expressed in TAP- and tapasin-negative cell lines. Because ME-1 Abs do not recognize free H chain homodimers of B27 (when prepared as a recombinant protein and tested for Ab binding by ELISA) (11), the inhibition of the response is not likely to occur by direct blocking of the CD4+ T cell-B27 homodimer interaction. However, studies have shown that free H chain forms of B27 are generated from w6/32-reactive (i.e., heterodimeric) B27 molecules on the cell surface (31). Therefore, an alternative explanation for the ability of ME1 to block the CD4+ T cell proliferative responses may be that B27 heterodimers are stabilized by the presence of ME1 so that the dissociation of the B27 heterodimer to free H chain forms of B27 is inhibited.
The possibility that the homodimeric form of B27 is recognized by the CD4+ T cell clone is suggested by the lack of responses to C1R transfected with B27 in which amino acid 67 has been mutated from cysteine to serine. This substitution in the B pocket allows the normal formation of HLA-B27 heterodimers (32) and is reported not to effect its ability to be recognized by some peptide-specific CD8+ CTLs (11). However, it should also be considered that subtle changes to the peptide binding groove caused by this mutation may result in loss of the epitope recognized by these B27-reactive CD4+ T cells.
In any case, we do not propose that only one form of B27 can be recognized by CD4+ T cells because we have isolated CD4+ T cell lines from both B27+ AS and B27+ healthy individuals that exhibit different patterns of specificity for B27-transfected cell lines. For example, we have isolated CD4+ T cells lines from B27+ AS patients that recognize T2-B27 but no other B27-transfected cell line (L. H. Boyle, manuscript in preparation).
The CD4 coreceptor appears to play a role in the response of these CD4+ T cells in a manner similar to that shown for class II-restricted CD4+ T cells. We suggest that CD4 might interact with some form of B27, possibly the homodimeric form, or that CD4-associated lck may be prevented by mAb binding from associating with the TCR (33).
Although we have provided strong evidence that CD4+ T cells are capable of interacting with B27, this situation is somewhat artificial involving cell lines defective in Ag-processing pathways or containing extensive deletions in the MHC. However, in vivo it is possible that normal APCs acquire properties similar to these mutant cell lines. This has already been shown to occur following infection with CMV (34). In view of the association between spondyloarthropathy and bacterial infection it is worth considering the possibility that intracellular bacteria can subvert MHC expression in the same way as viruses, leading to expression of altered forms of B27. Indeed, normal intestinal flora are essential in the induction of arthritis in B27 transgenic rodents (3, 35).
It cannot be determined at this stage whether the B27-specific
CD4+ T cells we have described have any
pathogenic role in humans. In support of the role of MHC class
I-restricted CD4+ T cells in pathogenesis,
apparently MHC class I-restricted CD4+ T cells
derived from MHC class II-deficient mice can induce aggressive
inflammatory bowel disease in congenic, immunodeficient hosts
(26). Future studies will determine the frequency of these
B27-reactive CD4+ cells in healthy individuals
and in patients with spondyloarthropathy. It may also be worthwhile to
determine whether B27-reactive CD4+ T cells can
be isolated from B27 transgenic rodents, and whether these interact
with novel forms of B27 such as free H chains, as suggested by the
occurrence of disease in B27,
2m-/- mice. Further
investigations will determine whether this interaction is similar to
the TCR-MHC class II interaction of conventional class II-restricted
CD4+ T cells.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
2 Address correspondence and reprint requests to Dr. J. S. Hill Gaston, Department of Medicine, University of Cambridge, Box 157, Level 5, Addenbrooke’s Hospital, Hills Road, Cambridge, CB2 2QQ, U.K. E-mail address: jshg2{at}medschl.cam.ac.uk
3 Abbreviations used in this paper: 2m, 2-microglobulin; 220, 721.220 cell line; 220-B27, 220 transfected with HLA-B*2705; T2-B27, T2 cell line transfected with HLA-B*2705; C1R-B27, C1R transfected with HLA-B*2705; EBV LCL, EBV-transformed lymphoblastoid cell line; AS, ankylosing spondylitis.
Received for publication April 23, 2001. Accepted for publication July 12, 2001.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
2-microglobulin: a model of human spondyloarthropathies. J. Exp. Med. 182:1153.2-microglobulin-free H chain homodimer structure. J. Immunol. 162:5045.
T-cell recognition of Yersinia enterocolitica. Immunology 91:503.[Medline]
This article has been cited by other articles:
|
|
 |
|
  C Duftner, C Dejaco, W Kullich, A Klauser, C Goldberger, A Falkenbach, and M Schirmer Preferential type 1 chemokine receptors and cytokine production of CD28- T cells in ankylosing spondylitis Ann Rheum Dis, May 1, 2006; 65(5): 647 - 653. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. J. Turner, D. P. Sowders, M. L. DeLay, R. Mohapatra, S. Bai, J. A. Smith, J. R. Brandewie, J. D. Taurog, and R. A. Colbert HLA-B27 Misfolding in Transgenic Rats Is Associated with Activation of the Unfolded Protein Response J. Immunol., August 15, 2005; 175(4): 2438 - 2448. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T K Kvien, J S H Gaston, T Bardin, I Butrimiene, B A C Dijkmans, M Leirisalo-Repo, P Solakov, M Altwegg, P Mowinckel, P-A Plan, et al. Three month treatment of reactive arthritis with azithromycin: a EULAR double blind, placebo controlled study Ann Rheum Dis, September 1, 2004; 63(9): 1113 - 1119. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. M. Tran, N. Satumtira, M. L. Dorris, E. May, A. Wang, E. Furuta, and J. D. Taurog HLA-B27 in Transgenic Rats Forms Disulfide-Linked Heavy Chain Oligomers and Multimers That Bind to the Chaperone BiP J. Immunol., April 15, 2004; 172(8): 5110 - 5119. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. N. Antoniou, S. Ford, J. D. Taurog, G. W. Butcher, and S. J. Powis Formation of HLA-B27 Homodimers and Their Relationship to Assembly Kinetics J. Biol. Chem., March 5, 2004; 279(10): 8895 - 8902. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Roddis, R. W. Carter, M.-Y. Sun, T. Weissensteiner, A. J. McMichael, P. Bowness, and H. C. Bodmer Fully Functional HLA B27-Restricted CD4+ as well as CD8+ T Cell Responses in TCR Transgenic Mice J. Immunol., January 1, 2004; 172(1): 155 - 161. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. H. Boyle and J. S. Hill Gaston Breaking the rules: the unconventional recognition of HLA-B27 by CD4+ T lymphocytes as an insight into the pathogenesis of the spondyloarthropathies Rheumatology, March 1, 2003; 42(3): 404 - 412. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
E. May, M. L. Dorris, N. Satumtira, I. Iqbal, M. I. Rehman, E. Lightfoot, and J. D. Taurog CD8{alpha}{beta} T Cells Are Not Essential to the Pathogenesis of Arthritis or Colitis in HLA-B27 Transgenic Rats J. Immunol., January 15, 2003; 170(2): 1099 - 1105. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Hulsmeyer, R. C. Hillig, A. Volz, M. Ruhl, W. Schroder, W. Saenger, A. Ziegler, and B. Uchanska-Ziegler HLA-B27 Subtypes Differentially Associated with Disease Exhibit Subtle Structural Alterations J. Biol. Chem., November 27, 2002; 277(49): 47844 - 47853. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
N. S. Dangoria, M. L. DeLay, D. J. Kingsbury, J. P. Mear, B. Uchanska-Ziegler, A. Ziegler, and R. A. Colbert HLA-B27 Misfolding Is Associated with Aberrant Intermolecular Disulfide Bond Formation (Dimerization) in the Endoplasmic Reticulum J. Biol. Chem., June 21, 2002; 277(26): 23459 - 23468. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
L. Sesma, V. Montserrat, J. R. Lamas, A. Marina, J. Vazquez, and J. A. L. de Castro The Peptide Repertoires of HLA-B27 Subtypes Differentially Associated to Spondyloarthropathy (B*2704 and B*2706) Differ by Specific Changes at Three Anchor Positions J. Biol. Chem., May 3, 2002; 277(19): 16744 - 16749. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
cpm, calculated as described in
Materials and Methods. Data shown are the means of
duplicate wells ± SDs. These data are representative of at least
three similar experiments. b, Flow cytometric analysis
of the phenotype of the CD4+ lines and the CD4+
clone.
