THE CELL SURFACE MOLECULE RECOGNIZED BY THE
ERYTHROCYTE RECEPTOR OF T LYMPHOCYTES
Identification and Partial Characterization Using a Monoclonal
BY THOMAS HUNIG
From the Institute for Virology and lmmunobiology of the University of W~rzburg, D-8 700
Wiirzburg, Federal Republic of Germany
Until the recent advent of monoclonal antibodies (mAb) ~ specific for human
T cells, the most widely used marker for human peripheral T cells and thymocytes
was their ability to bind sheep red blood cells (SRBC) (1-4). Early studies using
polyvalent anti-T cell sera suggested that the interaction resulting in E rosettes
is mediated by specific receptors on the human T cells (5), a notion fully
confirmed by the discovery that mAb against the T cell surface molecule carrying
T 1 1 and related markers block rosette formation (6-9). Nevertheless, E rosetting
remained a curious, if useful, phenomenon until a functional involvement of the
E receptor in T cell activation was suggested by a number of findings: First, it
was shown that the rosetting procedure rendered T cells reactive to growth
factor(s) (10). More recently, anti-T11 mAb were reported to either inhibit T
cell activation (1 1-15) and T cell-mediated cytolysis (12, 13), or, if certain
combinations of mAb were used, to lead to polyclonal T cell activation (14).
Accordingly, a "physiological ligand" of the E receptor with a role in T cell
activation was postulated (11-13, 16), which could be envisaged as either a
soluble mediator or a cell surface structure on cells with which T cells interact.
We have pursued the latter hypothesis and have developed an experimental
approach based on the following assumption: The binding of RBC to the E
receptor is not due to fortuitous crossreactivity with the unknown physiological
ligand. Since human T cells also bind autologous RBC (17), the target for T1 1
on RBC may be the physiological ligand itself; this ligand may be expressed on
other cell types as well, where it is not as easily visualized by rosetting but may
play its role in cellular interactions.
In the present report, the identification, isolation, and partial biochemical
characterization of the target structure for T 1 1 (T 1 1TS) is described. In contrast
to earlier attempts to define this molecule by conventional biochemistry (18-
This work was supported by the Deutsche Forschungsgemeinschaft through SFB 105, W/irzburg.
l Abbreviations used in this paper: BSA, bovine serum albumin; BSS, balanced salt solution; Con
A, concanavalin A; FCS, fetal calf serum; FITC, fluorescein isothiocyanate; HuRBC, human red
blood cells; IL-2, interleukin 2; mAb, monoclonal antibody(s); NMIgG, normal mouse IgG; NP-40,
Nonidet P-40; PAGE, polyacrylamide gel electrophoresis; PBL, peripheral blood lymphocytes; PBS,
phosphate-buffered saline; SDS, sodium dodecyl sulfate; SRBC, sheep red blood cells; SWBC, sheep
white blood cells; TI ITS, Tl 1 target structure.
890 J. ExP. MED. © The Rockefeller University Press • 0022-1007/85/09/0890/12 $1.00
Volume 162 September 1985 890-901
20), we have affinity purified T 11TS with an mAb to SRBC that abrogates their
binding to human and sheep T lymphocytes.
Materials and Methods
Preparation of Cells. Freshly drawn human blood from healthy individuals was centri-
fuged in balanced salt solution (BSS) with heparin. The buffy coat cells were centrifuged
over FicolI-Paque (Pharmacia, Inc., Uppsala, Sweden) to prepare peripheral blood lym-
phocytes (PBL). Human RBC (HuRBC) and PBL were washed three times in BSS before
use. Fresh SRBC were washed twice in BSS with heparin. The buffy coat was collected,
spun down, and resuspended in 90% Percoll (Pharmacia, Inc.) in BSS. 70 and 50% Percoll
suspensions and BSS were layered on top, and, after 20 min centrifugation with 180 g at
room temperature, white blood cells were collected from the interphase between the
BSS/50% Percoll and 50%/70% Percoll. Sheep white blood cells (SWBC) were stimulated
with 2 ttg concanavalin A (Con A) (Pharmacia, Inc.) per milliliter RPMI 1640 with 5%
fetal calf serum (FCS), 5 × 10 -5 M 2-mercaptoetbanol, nonessential amino acids, and
antibiotics. After 1 wk, Con A was washed out and the cells were recultured in the
presence of 5 U/ml of recombinant human interleukin 2 (IL-2), the kind gift of Drs. Fiers
and Devos, Biogent, Ghent, Belgium. Prolonged culture was performed by culturing with
IL-2 and alternating the presence and absence of Con A on a weekly schedule. Before
use, these T cell blasts were washed again in BSS with 20 mg/ml a-methyl mannoside to
remove Con A.
Assay for E Rosettes. SRBC and HuRBC were treated with neuraminidase (Testneura-
minidase; Behringwerke AG, Marburg, FRG) diluted 1:50 in BSS for 30 min at 37°C,
washed, and adjusted to 10S/ml and 2.5 × 107/ml, respectively in BSS with 5% FCS (BSS/
FCS). Human PBL were either left untreated (for sheep E rosettes), or neuraminidase
treated (for autologous rosettes), washed, and adjusted to 8 x 106/ml in BSS/FCS. Sheep
T cell blasts were also treated with neuraminidase and adjusted to 8 × 106 BSS/FCS. 50
ttl of lymphocytes were mixed with 50 IA of RBC and 50 tA of BSS/FCS. For rosettes of
human PBL with SRBC, the mixture was left for 10 min at room temperature, centrifuged
for 2 min at 200 g, and left for another 30 min at room temperature. Autologous human
and sheep rosettes were prepared by incubating the cell mixture for 20 min at 37°C,
centrifuging for 5 min at 200 g, and leaving the pelleted cells for 60 min on ice. All
rosettes were microscopically enumerated after gently resuspending the cells and staining
with crystal violet. 400 nucleated cells were examined per test, and nucleated cells with
three or more bound RBC were scored as rosettes.
E Rosette Inhibition Assay. mAb to SRBC were assayed by preincubating 50 #1 of SRBC
with 50 ~1 of antibody preparations for 30 min at room temperature, followed by the
addition of lymphocytes and the rosetting procedure as described above. Anti-T11 mAb
(Coulter Immunology, Hialeah, FL) or partially purified T1 ITS were assayed by prein-
cubating 50 tLl of these preparations with 50 ttl lymphocytes for 30 min at room
temperature before proceeding with the assay. To determine whether lymphocytes or red
cells are the target of inhibition, each partner was preincubated with the relevant
preparation for 30 min at room temperature, washed twice with BSS/FCS followed by
the addition of the untreated partner, and the assay performed as described above.
Isolation ofmAb L180/1. Supernatants from a total of 86 SRBC-specific hybridomas,
produced in two separate fusions from SRBC primed and boosted BALB/c mice with the
nonsecreting myeloma X63-Ag 8.653, were screened for inhibition of rosettes between
human PBL and SRBC. These hybridomas were kindly provided by M. Lohoff at this
institute. One culture, termed L180, scored positive. It was subcloned, and antibody was
purified from culture supernatant and ascites by ammonium sulphate precipitation and
salt-gradient elution from DEAE-Affigel Blue (Bio-Rad Laboratories, Richmond, CA).
The antibody L180/1 is of the IgG class and binds to protein A. Fab fragments were
prepared by digestion with insolubilized papain (Sigma GmbH, Taufkirchen, FRG) and
examined for purity by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-
PAGE). For controls, normal mouse IgG (NMIgG) was purified from serum. Protein
content was determined using a protein assay (Bio-Rad Laboratories).
TARGET STRUCTURE FOR THE E RECEPTOR
Preparation of Surface-labeled Cell Extracts.
times in phosphate-buffered saline (PBS) containing 5 X 10 -6 M KI and resuspended in
0.9 ml of this buffer. 2.5 U of lactoperoxidase (Sigma GmbH), 1 mCi of ~5I as NaI, and
10 #l of 0.3% H~O~ were added. 1 U lactoperoxidase and 10 #1 of 0.3% H~O~ were added
again, twice, at 4-min intervals. The reaction was stopped after another 8 rain by washing
the cells three times in PBS containing 5 x 10 -3 M KI. For labeling of cell surface
carbohydrates, 5 x 108 SRBC were treated with 0.25 U neuraminidase and 25 U galactose
oxidase (Sigma GmbH) in 5 ml BSS for 45 rain at 37°C, washed once in cold PBS, and
resuspended in 2 ml cold PBS containing 30 mCi tritiated BH4- (NEN, Dreieich, FRG).
After 30 min on ice the cells were washed three times in PBS. Surface-labeled cells were
lysed in 2 ml PBS containing 0.5% Nonidet P-40 (NP-40) (Roth, Karlsruhe, FRG), 0.02%
deoxycholate (Sigma GmbH), and 70 #g/ml aprotinin (Trasylol; Bayer, Leverkusen, FRG).
After 15 rain on ice the lysate was spun for 30 rain at 20,000 rpm, and the pellet was
Purified mAb L180/1 or NMIgG were covalently
coupled to agarose beads according to the manufacturer's suggested procedure (Affigel
10; Bio-Rad Laboratories). ~3.5 mg IgG bound per milliliter of gel. Before radioimmu-
noprecipitation (RIP), the beads were washed three times in lysis buffer. 100-200 #! of
radiolabeled lysate were kept with 50 #i of NMIgG-coupled beads for 30 min on ice. The
supernatant was collected by centrifugation, and 50 #1 L18011 or NMIgG-coupled beads
were added. After 60 min on ice, the beads were washed three times in lysis buffer
containing 0.5 M KCI and three times in lysis buffer.
Large-Scale Enrichment of TIlTS.
100 ml of PBS-washed, packed SRBC were lysed
with an equal volume of PBS containing 1% NP-40, 0.04% deoxycholate, and 70 #g
aprotinin. After 30 min on ice, the lysate was centrifuged for 30 min at 20,000 rpm in
the cold. The supernatant was passed dropwise over 1 ml of mAb L 180/1-coupled Affigel
10. The immunosorbent column was washed with 50 ml lysis buffer containing 0.5 M
KCI and 50 ml lysis buffer without KCI. The gel was removed from the column, residual
buffer was aspirated, and the bound material was eluted overnight with 2 ml 3.5 M
NH4SCN, pH 8.0 containing 0.2% NP-40. The eluate was passed over a Sephadex G-25
column equilibrated with 0.1% NP-40 in PBS. Two or three 1-ml fractions that were
positive in the Bio-Rad protein assay were collected. For studies on the interaction of
affinity-enriched material with viable cells, the gel was washed with PBS before salt elution
and detergent was omitted during the following steps. Protein content was determined
using the Bio-Rad protein assay.
Crude SRBC Membranes.
50 #1 of packed SRBC were lysed with 1 ml H20. Membranes
were pelleted and washed three times in 0.2% NaCi by centrifugation at 20,000 rpm for
SDS-PAGE and Autoradiography.
Samples were boiled in Laemmli sample buffer (21)
with or without 2-mercaptoethanol and run on SDS slab gels containing 10 or 12%
acrylamide. ~4C-labeled markers were: cytochrome c (12 kilodaltons [kD]), carbonic
anhydrase (30 kD), ovalbumin (45 kD), bovine serum albumin (BSA) (66 kD), phospho-
rylase b (92 kD) (all from Amersham Buchler, Braunschweig, FRG). Unlabeled markers
were: lysozyme (14.3 kD), trypsinogen (24 kD), ovalbumin (45 kD), BSA (66 kD) (Sigma
GmbH). Gels were fixed, stained with Coomassie Blue, and dried. Gels containing SH-
labeled material were fluorographed using dimethyl sulfoxide/PPO. Autoradiography
was performed with Cronex 2 x-ray film (DuPont Co., Wilmington, DE) preexposed to
light, using enhancing screens.
Indirect Immunofluorescence and Flow Cytometry. 2.5 x l0 s human PBL were preincu-
bated for 30 min at room temperature with either 100 ~1 PBS or PBS containing 1 #g
protein eluted from the L 180/1 affinity column. The cells were washed once and treated
with 4/zg NMIgG, anti-T11 mAb diluted 1:100, or anti-T3 mAb (OKT3; Ortho Diag-
nostic Systems, Inc., Raritan, N J) diluted 1:50, in 0.1 ml PBS containing 0.1% BSA and
0.02% sodium azide (PBS/BSA/azide). After 30 rain on ice, the PBL were washed in this
buffer, stained with goat anti-mouse, fluorescein isothiocyanate (FITC)-coupled antibodies
(Coulter Diagnostics, Hialeah, FL) diluted 1:50 for 30 min on ice, and washed three times
100 ~1 of packed SRBC were washed three
in PBS/BSA/azide. Analysis was performed on an Epics V flow cytometer (Coulter
Electronics, Inc., Hialeah, FL), using a forward angle light scatter gate for lymphocytes
and collecting integrated green fluorescence signals from a logarithmic amplifier.
Inhibition of Heterologous and Autologous E Rosette Formation by Anti-T11 and
The culture supernatants from a total of 86 SRBC-specific hy-
bridomas from two independent fusions were screened for their ability to inhibit
rosette formation between human T cells and SRBC. One positive culture (L 180)
was identified, subcloned, and the antibody purified from culture supernatant
and ascitic fluid. Fab fragments were prepared to eliminate hemagglutination
and tested again for rosette inhibition.
As shown in Fig. 1 A, nanogram amounts of L 180/1 Fab fragments completely
abrogated the formation of rosettes between human PBL and SRBC. Since mAb
L180/1 does not bind to HuRBC (data not shown), it was expected that autolo-
gous human rosettes were not affected by L180/1 Fab (Fig. 1B). On the other
hand, anti-T 11, which is known to block binding of SRBC to human T cells (6-
9) (Fig. 1 A), also blocked autologous human rosette formation (Fig. 1 B), sug-
gesting that the same E receptor binds SRBC and HuRBC. The absence of the
epitope recognized by L180/1 from HuRBC could either mean that the target
structure recognized by the human E receptor on SRBC and HuRBC are
unrelated, or that there is antigenic variation of the same molecule between
species, as is commonly found.
If, as hypothesized, the putative target structure for the E receptor is a cell
HUMAN PSL - SRBC
IIA-- IW,~ - -
HUMAN PB1. - HuRBC SHEEP ConA BLASTS- SRBC
Oil. of J, T11
II #.T 11
• FEb L 180/1
FIGURE 1. Inhibition of autologous and xenogeneic E rosette formation by anti-T cell and
anti-erythrocyte mAb. NMIgG (O), mAb to TI 1 (m), and Fab ofmAb L180/1 (4,) were tested
for blocking of E rosettes as described in Materials and Methods. In the absence of antibody,
the fraction of rosette-forming cells was: 62% (A), 24% (B), 9.6% (C).
TARGET STRUCTURE FOR THE E RECEPTOR
FrcuR~ 2. Autologous rosette formed between sheep T cell blast and SRBC.
interaction molecule, it must be functioning in an autoiogous situation such as
in the binding of human T cells to human erythrocytes. Since, as discussed
above, mAb L180/! detects the putative ligand of T11 on sheep but not on
human RBC, we have devised a protocol for the formation of sheep-sheep
rosettes. Although freshly isolated SWBC do not form stable autologous E
rosettes, ~ 10% of SWBC activated by Con A and propagated in human recom-
binant IL-2 bound numerous SRBC to form stable rosettes (Fig. 2). This rosette
formation was completely inhibited by L180/1 Fab (Fig. 1C), suggesting that
sheep T cell blasts have a receptor for autologous RBC that binds the same
target structure also recognized by the human T11 molecule. The mAb to T11
that was used neither bound to sheep T cell blasts (data not shown) nor inhibited
sheep autologous rosettes (Fig. 2C). In summary, it appears that analogous
receptor-ligand systems exist in man and sheep, and that the mAb to T11 and
mAb L180/! recognize species-specific determinants on the E receptor and its
Isolatin of TI 1TSfrom SRBC.
SRBC were either surface iodinated or surface
labeled by treatment with neuraminidase plus galactose oxidase followed by
reduction with tritiated borohydride. Immunoprecipitates from detergent lysates
were analyzed by SDS-PAGE and autoradiography. As can be seen in Fig. 3A, a
band with an apparent molecular weight of ~42,000 was precipitated from ~z5I-
labeled extracts by mAb L180/1 coupled to agarose beads under both reducing
and nonreducing conditions. The additional band of higher molecular weight
was, in all experiments where it was found in L180/1 immunoprecipates, also
present in control precipitates. It is therefore regarded as nonspecific. Immu-
noprecipitation of 3H-labeled glycosylated membrane molecules revealed a single
band of slightly smaller apparent molecular weight than observed with iodinated
extracts. Note that this band does not comigrate with the bulk of the labeled
FIGURE 3. SDS-PAGEofsurface-labeledSRBCextractsradioimmunoprecipitatedwithmAb
L180/1. (A) 2~sI-labeled, 10% acrylamide; (B) 3H-labeled, 10% acrylamide reduced; (C) ~sI-
labeled, 12% acrylamide reduced. Neuraminidase (N'ase) treatment of the lysate (C) was
performed for 30 min at 37 °C with 1:50 dilution of Tesmeuraminidase.
material (Fig. 3B). The difference in molecular weight between iodinated and
"~H-labeled material is accounted for by the removal of sialic acid during the
labeling procedure used for the latter, since the same shift in molecular weight
was observed if 12~I-labeled extracts were neuraminidase treated before immu-
noprecipitation (Fig. 3 C). The apparent glycoprotein nature of T 11TS has been
further substantiated by trypsin and endoglycosidase degradation (data not
The ready availability of the starting material also permitted large-scale puri-
fication of T 11TS. As an initial step, a detergent lysate prepared from 100 ml
packed SRBC was passed over insolubilized mAb L180/1 and eluted with high
salt. As can be seen in Fig. 4C, T11TS is a major, although not the only band
visible upon Coomassie Blue staining of the eluate in SDS-PAGE. Subsequent
radiolabeling of this affinity-enriched material and immunoprecipitation (data
not shown) confirmed that the band indicated by arrow in Fig. 4 is the material
precipitated by mAb L180/1 as described above. To determine whether mAb
L180/1 precipitates a major protein of the red ceil membrane, a crude SRBC
membrane preparation (Fig. 4, A and B) and total detergent lysate (Fig. 4, D and
E) were run in parallel. It appears that T11TS is only a minor component of the
red cell membrane as detected by this method. T11TS can be further purified
by adsorption to Con A-Sepharose (not shown), and our current efforts are
directed towards obtaining enough homogeneous material for microsequencing.
Binding of PartiaUy Purified Tl lTS to the E Receptor of Human T Cells.
experiments showed that after elution of detergent-solubilized T11 TS from the
L180/1 affinity column with high salt and transfer into PBS via salt-exchange
chromatography, the yield of T11TS was greatly enhanced if the presence of
detergent was maintained. However, enough material for detection by protein
stains in SDS-PAGE was recovered if the high salt eluate was transferred to
detergent-free PBS, permitting investigation of the interaction between the
affinity-enriched material and the E receptor on viable cells. Two types of
experiments were performed with such preparations. First, we investigated
whether partially purified T11TS blocks formation of rosettes between human
TARGET STRUCTURE FOR THE E RECEPTOR
crude SRBC membranes; (C) eluate from L180/1 affinity column; (D, E) 1:I00 and 1:10
dilutions of detergent lysate applied to affinity column. SDS-PAGE (12% acrylamide) was run
under reducing conditions and stained with Coomassie Blue.
Large-scale affinity enrichment of T1 ITS. (A, B) 1:10 and 1:100 dilutions of
• ! I i i
ng partially purified TllTS added
500 125 31
FICURE 5. Inhibition of E rosette formation by affinity-enriched T11TS. Human PBL were
preincubated with the eluate from the affinity column before addition of neuraminidase-
treated SRBC. In the absence of TI ITS, 58% of PBL formed rosettes.
PBL and SRBC (Fig. 5). Complete inhibition was indeed observed when 500 ng
of protein was added to the assay system. To assess whether the observed effect
was caused by the putative ligand of the E receptor or by antibody leaking from
Effect of Preincubation of PBL or SRBC with mAb L 180/1, anti-T11,
or Partially Purified T11TS on E Rosette Formation
Percent inhibition of E rosettes after
PBL or SRBC were preincubated for 30 min with mAb to T11 (h300
final), 2 ug Fab of mAb L180/1, or 2 ug of partially purified T11TS,
washed, and tested for rosette formation. 66% of nucleated cells formed
rosettes in untreated controls.
anti-T1 I anti-T3
Log Integral Fluorescence
FIGURE 6. Interference of partially purified T11TS with the binding of anti-T11 to human
PBL. PBL were preincubated with affinity-enriched T1 ITS, washed, and stained with anti-
T 11 and anti-T3, followed by FITC-coupled goat anti-mouse Ig as described in Materials and
Methods. NMIgG was included as a negative control but, for reasons of clarity, is not shown.
the immunosorbent column, PBL or SRBC were separately preincubated with
either anti-T11 mAb, Fab L180/1, or partially purified T11TS, then washed,
mixed with the untreated partner, and examined for rosettes (Table I). As
expected, T1 1 mAB and partially purified T1 ITS blocked rosette formation if
the PBL were preincubated, whereas mAB L180/1 was only effective after
pretreatment of red cells. It appears, therefore, that a protein was present in the
column eluate that was capable of competing with the binding site for red cells
on human T cells.
More direct evidence for the binding of partially purified T11TS to the E
receptor is presented in Fig. 6. Human PBL were preincubated with the eluate
from the affinity column or with control buffer, washed, treated with mAb to
T11 or to T3, stained with FITC-conjugated second antibody, and analyzed by
flow cytometry. Pretreatment with the preparation believed to contain T1 1TS
resulted in a marked reduction of staining with the anti-T11 mAb, but had no
effect on staining with anti-T3 mAb. This indicates a specific blocking of the
binding site recognized by anti-T 11 and thus confirms the presence of E receptor
binding material in the affinity-enriched preparation.
TARGET STRUCTURE FOR THE E RECEPTOR
The results presented here show that an mAb to SRBC, termed L180/1,
completely inhibits rosette formation between human or sheep T lymphocytes
and SRBC. In addition, the molecule recognized by this mAb, a minor glycopro-
tein of the SRBC membrane with an apparent molecular weight of ~42,000 (42
K), also inhibited rosette formation and the binding of anti-T11 mAb to the E
receptor on human T cells. Since these findings appear to indicate that mAb
L180/1 detects the cell surface molecule recognized by the E receptor, we have
given this glycoprotein the preliminary name T11 target structure, or T11TS.
This is not the first report claiming the identification of the SRBC membrane
molecule involved in rosette formation, although blocking of the E receptor as
detected by reduced binding of mAb to T11 (Fig. 6) has, to our knowledge, not
been described. Kitao et al. (18) described a sialoglycopeptide of 10 K mol wt
released from SRBC by trypsin, which blocked the formation of E rosettes;
Fletcher et al. (19) demonstrated rosette inhibition with glycoproteins of 37,
26.5, and 9 K mol wt that were isolated by biochemical means from SRBC
membranes. In contrast, Gfirtler (20) detected rosette-inhibiting activity in the
glycolipid fraction of SRBC membranes. The relationship between these prepa-
rations and the molecule described here is unclear at present, but it should be
noted that in these reports the amount of glycoprotein or glycolipid required to
block rosette formation was several hundred-fold higher than is reported here
for the partially affinity-purified material. Thus, the active component may have
comprised only a small minority of the glycoprotein or glycolipid preparations
isolated biochemically. Indeed, the results in Figs. 3 and 4 show that the molecule
recognized by mAb L180/1 is not one of the major glycoproteins of the well-
studied red cell membrane (22), a finding corroborated by the expression of
T11TS on white blood cells also (H/inig, manuscript in preparation). We are
presently attempting to purify enough material for microsequencing, so that the
identity of T11TS will hopefully be soon resolved.
The nature of the interaction between the E receptor and its target structure
remains unclear, although several laboratories have presented evidence for the
recognition of sugar moieties on the red cell membrane. Boldt and Armstrong
(23) demonstrated rosette inhibition by several glycoproteins rich in sialic acid,
galactose, N-acetyl glucosamine, and mannose, although at concentrations several
orders of magnitude higher than were required for blocking in our experiments.
Moreover, Fletcher et al. (19) reported that the glycoproteins which they isolated
lost their activity in the rosette inhibition assay after enzymatic desialylation, an
apparent contradiction to the enhanced rosette-forming capacity of neuramini-
dase-treated SRBC. We have confirmed this finding to some extent with the
glycoprotein studied here (data not shown), but the poor solubility of T11TS in
the absence of detergents suggests that the reduction in inhibitory activity after
desialylation may not be due to destruction of the site recognized by the E
receptor but rather to a further loss of solubility in aqueous buffers. It should
be added that neuraminidase treatment of SRBC or iodinated lysates did not
impair the binding of mAb L180/1 as determined by radioimmunoprecipitation
(Fig. 3 C) and quantitative flow cytometry (data not shown), indicating that sialic
acid residues are not directly involved in the antigenic determinant recognized
by this mAb. Nevertheless, it cannot be distinguished at present whether mAb
L180/1, much less the E receptor, recognize the protein or the sugar moiety of
TllTS. Experiments addressing this question are under way. Quite clearly,
however, the complete inhibition of rosette formation by nanogram amounts of
Fab fragments of an mAb specific for only one minor glycoprotein of the red
cell membrane demonstrates that the interaction of the E receptor with its
complementary structure is highly selective.
The expression of T11TS on white blood cells (Hiinig, manuscript in prepa-
ration) points to the possible biological significance of this interaction; the present
demonstration of autologous rosettes in the sheep system that appear to use the
same molecule on SRBC as that recognized by human T lymphocytes (Fig. 1), is
intriguing in this respect. The E receptor as defined by anti-T! 1 and related
mAb is widely distributed among primates (24), and the present results suggest
that an analogous system also exists in more distantly related mammals. In this
context, it is of interest that anti-T11 mAb blocked both heterologous rosettes
between human T cells and SRBC, and autologous human rosettes (Fig. 1). This
suggests that the E receptor target system is highly conserved and that SRBC
rosetteing is more than an immunological curiosity. One may speculate that red
cell rosetting involves the interaction of complementary cell interaction mole-
cules, a notion fully consistent with published experiments on the regulation of
T cell activation via the E receptor (10-16).
A monoclonai antibody (mAb) to sheep red blood cells (SRBC), termed L180/
1, is described that completely blocks rosette formation between SRBC and
human or sheep T lymphocytes. L180/1 precipitated a minor glycoprotein of
about ~42,000 mol wt from surface-labeled SRBC. This glycoprotein was par-
tially affinity purified and found to block E rosette formation and to compete
with anti-T11 mAb for the E receptor. The molecule detected by mAb L180/1
thus appears to be recognized by the E receptor and was given the preliminary
name, T11 target structure (T11TS). Since the mAb to sheep T11TS blocks the
binding of SRBC to both human and sheep T cells, and mAb to T11 blocks the
binding of red cells from human and sheep to the human E receptor, we
concluded that analogous receptor-ligand (T 11-T 11TS) systems exist in man and
sheep that are crossreactive over the species barrier. The possibility is discussed
that the E receptor, which is known to be involved in T cell activation, and
T 11TS function as complementary cell interaction molecules in T cell responses.
I thank Rita Mitnacht for excellent technical assistance, M. Lohoff for his collection of
SRBC-specific hybridomas, Drs. Fiefs and Devos for their gift of recombinant human IL-
2, Drs. A. Schimpl and E. Wecker for helpful discussions, and Ch. Stoppe and B. Jordan
for typing this manuscript.
Received for publication 30 April 1985 and in revised form 5June 1985.
1. Bach, J. F., J. Dormont, M. Dardenne, and H. Balner. 1969. In vitro rosette inhibition
TARGET STRUCTURE FOR THE E RECEPTOR
by antihuman antilymphocyte serum. Correlation with skin graft prolongation. Trans-
plantation (Baltimore). 8:265.
2. Brain, P., J. Gordon, and W. A. Willets. 1970. Rosette formation by peripheral
lymphocytes. Clin. Exp. Immunol. 6:681.
3. Coombs, R. R. A., B. W. Gurner, A. B. Wilson, G. Holm, and B. Lindgren. 1970.
Rosette formation between human lymphocytes and sheep red cells not involving
immunoglobulin receptors. Int. Arch. Allergy Appl. Immunol. 39:658.
4. Lay, W. H., N. F. Mendes, C. Bianco, and V. Nussenzweig. 1971. Binding of sheep
red blood cells to a large population of human lymphocytes. Nature (Lond.). 230:531.
5. Wortis, H. H., A. G. Cooper, and M. C. Brown. 1973. Inhibition of human lympho-
cyte rosetting by anti-T sera. Nature (Lond.). 243:109.
6. Howard, F. D., J. A. Ledbetter, J. Wong, C. P. Bieber, E. B. Stinson, and L. A.
Herzenberg. 1981. A human T-lymphocyte differentiation marker defined by mon-
oclonal antibodies that block E-rosette formation. J. Immunol. 126:2117.
7. Kamoun, M., P.J. Martin, J. A. Hansen, M. A. Brown, A. W. Siadeck, and R. C.
Nowinski. 1981. Identification of a human T lymphocyte surface protein associated
with the E-rosette receptor.J. Exp. Med. 153:207.
8. Bernard, A., C. Gelin, B. Raynal, D. Pham, C. Grosse, and L. Boumsell. 1982.
Phenomenon of human T cells rosetting with sheep erythrocytes analyzed with
monoclonal antibodies.J. Exp. Med. 155:1317.
9. Verbi, W., M. Greaves, C. Schneider, K. Konbek, G. Janossy, H. Stein, P. Kung, and
G. Goldstein. 1982. Monoclonal antibodies OKTll and OKTllA have pan°T
reactivity and block sheep erythrocyte "receptors". Eur.J. Immunol. 12:81.
10. Larsson, E. L., J. Andersson, and A. Coutinho. 1978. Functional consequences of
sheep red blood cell rosetting for human T-cells: gain of reactivity to mitogenic
factors. Eur. J. Immunol. 8:693.
11. Palacios, R., and O. Martinez-Maza. 1982. Is the E-receptor on human T-|ymphocytes
a "negative signal receptor"?J. Immunol. 129:2479.
12. Martin, P. J., G. Longton, J. A. Ledbetter, W. Neumann, M. P. Braun, P. G. Beatty,
and J. A. Hansen. 1983. Identification and functional characterization of two distinct
epitopes on the human T-cell surface protein tp 50. J. Immunol. 131 : 180.
13. Krensky, A. M., F. Sanchez-Madrid, E. Robbins, J. A. Nagy, T. A. Springer, and S.
J. Burakoff. 1983. The functional significance, distribution, and structure of LFA-1,
LFA-2, and LFA-3: cell surface antigens associated with CTL-target interactions. J.
14. Wilkinson, M., and A. Morris. 1984. The E-receptor regulates interferon-gamma
production: four-receptor model for human activation. Eur. J. Immunol. 14:708.
15. Reed, J. c., w. Tadmori, M. Kamoun, G. Koretzky, and P. C. Nowell. 1985.
Suppression of interleukin 2 receptor acquisition by monoclonal antibodies recogniz-
ing the 50 kd protein associated with the sheep erythrocyte receptor on human T-
lymphocytes. J. Immunol. 134:163 I.
16. Meuer, S. C., R. E. Hussey, M. Fabbi, D. Fox, O. Acuto, K. A. Fitzgerald, J. C.
Hodgdon, J. P. Protentis, S. F. Schlossman, and E. L. Reinherz. 1984. An alternative
pathway of T-cell activation: a functional role for the 50 KD TI 1 sheep erythrocyte
receptor protein. Cell. 36:897.
17. Baxley, G., G. B. Bishop, and H. H. Wortis. 1973. Rosetting of human red blood
cells to thymocytes and thymus-derived cells. Clin. Exp. Immunol. 15:385.
18. Kitao, T., M. Takeshita, and K. Hattori. 1976. Studies on glycopeptide released by
trypsin from sheep erythrocytes. J. lmmunol, t 17:310.
19. Fletcher, M. A., K. E. Caldwell, Z. A. Latif, M. Cayer, and A. Claflin. 1982.
Immunochemical studies of infectious mononucleosis. VIII. A glycoprotein from
sheep erythrocytes with sialic acid-dependent receptor properties. J. Immunol.
90. Gfirtler, L. G. 1981. Glycolipid and p40 are the binding sites in the sheep erythrocyte
and T-lymphocyte membrane responsible for rosette formation. Immunobiology.
21. Laemmli, U. K. 1970. Cleavage of structural proteins during the assembly of the
head of bacteriophage T4. Nature (Lond.). 227:680.
22. Steck, T. L. 1974. The organization of proteins in the human red blood cell
membrane. J. Cell Biol. 62:1.
23. Boldt, D. H., andJ. P. Armstrong. 1976. Rosette formation between human lympho-
cytes and sheep erythrocytes. Inhibition of rosette formation by specific glycopeptides.
J. Clin. Invest. 57:1068.
24. Haynes, B. F., B. L. Dowell, L. L. Hensley, F. Gore, and R. S. Metzgar. 1982. Human
T-cell antigen expression by primate cells. Science (Wash. DC). 215:298.