From www.bloodjournal.org by guest on July 12, 2017. For personal use only.
Detection of Empty HLA Class I1 Molecules on Cord Blood B Cells
By Frederic Garban, Mats Ericson, Corinne Roucard, Claire Rabian-Herzog, Helene Teisserenc, Eric Sauvanet,
Dominique Charron, and Nuala Mooney
Fetal mononuclear cells are increasingly used in transplantation of hematopoietic cells due t o a reportedly lower incidence of graft-versus-host disease. Previous studies of immune responses of fetal lymphocytes have indicated a
general hyporesponsiveness in response t o polyclonal stimulation. Fetal B lymphocytes display many features typical
of the resting state such as a low level of HLA class II expression, but a large proportion of cells also carry the activationassociated CD23 antigen. We show here that despite a low
cell surface level of all three HLA class II isotypes on fetal B
cells, their allogeneic capacity, measured as the ability t o
elicit a mixed lymphocyte reaction, is similar t o that of adult
B cells. Allogeneic stimulation is believed t o be peptide-dependent. Surprisingly, the majority of the HLA class II molecules on cord blood B cells appeared t o be devoid of stably
bound peptide as detected by the binding of a recombinant
and soluble invariant chain, as well as by the absence of
sodium dodecyl sulfate (SDS) stable a@ heterodimers in
whole cell lysates. lmmunoblot experiments showed that
HLA class II molecules of fetal B cells were predominantly
present in high molecular weight aggregates in stark contrast t o B cells of adult origin. However, a sensitive cell surface labeling technique followed by immunoprecipitation
enabled us t o detect an SDS-stable 120-kD molecule on fetal
B cells. We propose that the 120-kD molecules could correspond t o HLA class II doubledimers or superdimers. We hypothesize that the 120-kD HLA class II molecule functions as
the antigen-presenting molecule in the mixed lymphocyte
reaction of fetal B cells, as it is the major species detected
on the surface. Secondly, we suggest that a high level of
empty HLA class II molecules may be indicative of a particular stage in B-cell ontogeny.
0 1996 by The American Society of Hematology.
H
binding stabilizes the class I1 dimers that can be monitored
experimentally as an increased resistance to denaturation in
sodium dodecyl sulfate (SDS) at room temperature.'".'' All
detectable compact (52 to 56 kD) class I1 antigens are free
of 1i.l3On the other hand, it has been reported that transfected
HLA DR molecules that had not passed through a peptide
loading compartment showed a strong tendency to aggregate. l4 The aggregates could be disaggregated by exposure
to peptide suggesting an altered conformation of the empty
molecules.I4 Recent work on class I1 antigens has described
a higher molecular weight species in the mouse, which is
believed to be a double dimer or a superdimer, as it migrates
at 120 kD." The superdimers are believed to contain peptide
because an antibody preferentially recognizing the 120-kD
complexes was capable of blocking low-affinity T-cell responses.
A new probe for empty HLA class I1 molecules, a soluble
recombinant Ii chain (Ii sol) has recently been developed
and has been used to demonstrate that binding of either
the l i chain or of peptide to class I1 antigens are mutually
exclusive.'6 When empty HLA DR4 molecules expressed on
Chinese hamster ovary (CHO) cells were allowed to bind
peptide, the subsequent binding of Ii sol was inhibited. Similarly, exposure of the DR4 molecules to Ii sol blocked peptide binding indicating that they compete for the same binding site. In agreement with these results, a monoclonal
antibody (MoAb) (L243), which in many cases has proven
capable of blocking peptide binding to class I1 molecules in
vitro, could also block binding by Ii sol. The Ii sol molecule
did not bind to Epstein-Barr virus (EBV) transformed B
cell lines in which the DR4 molecules were SDS-stable (M.
Ericson, unpublished results, May 1994).
In this report, we have examined the expression and conformation of HLA class I1 molecules on fetal versus adult
B cells. Surprisingly, the majority of the class 11 molecules
in fetal B cells appear to be empty, although the fetal cells
were equally as efficient as adult B lymphocytes in eliciting
an allogeneic reaction that is thought to be peptide-dependent." Low amounts of HLA class I1 molecules that migrated
with an apparent molecular weight of 120 kD were detected
UMAN CORD blood B cells can be considered as naive
cells because they exist in an environment essentially
devoid of exogenous peptide antigen. Previous studies',*
have reported their relative incapacity to differentiate to antibody forming cells, and they are refractory to mitogenic
stimulation.' Accordingly, fewer antibody forming cells
were generated in response to various mitogens and lower
levels of soluble Igs (in particular, IgA and IgG, but also
IgM) were detected in comparison with adult cells.' This
lack of response could either be due to a failure to activate
the B cells or to an impaired interaction with T helper cells.
The essential role of HLA class I1 molecules in the presentation of antigen is clearly established, but they also function
as signal transducing molecules on B lymphocytes.' HLA
class 11 molecules are heterodimers associating a 34-kD a
chain and a 28-kD /3 chain. Intracellularly, cy0 heterodimers
are complexed with a glycoprotein, the invariant chain (Ii):,5
which inhibits peptide binding until class I1 complexes reach
a subcellular compartment where Ii is degraded and exchanged for antigenic peptide.h~x
From the peptide loading
compartment, the class I1 molecules are believed to egress to
the cell membrane.' The class TI antigens can adopt different
conformations depending on their association with peptide
and the stability of this association."' Essentially, peptide
From the Laboratoire d'lmmunogknnttique Humainr INSERM
U396, lnstitut des Cordeliers, Paris; und Service de gynecologic
Hbpital Norre Dame de 3 0 n Secours, Paris, Fmnce.
Submitted April 18, 1995; accepted December I I , 199-5.
Supported by the Institut National de la SantP et de la Recherche
Me'dicale, the Ligue contre le Cancer (comite' de Paris), and by the
Ligue Nationale contre le Cancer.
Address reprint requests to N u d a Mooney, PhD, Luboratoire
d'lmmunoge'ne'tique humaine, INSERM U396, Institut des Cordeliers, 15, rue de I'Ecole de Me'decine, 75006 Paris, France.
The publication costs of this article were defrayed in part by page
charge payment. This article must therejbre be hereby marked
"advertisement" in accordance with 18 U.S.C. section 1734 .solely to
indicate this ,fact.
0 1996 by The American Socieg i f Hemntology.
0006-49 7//9ci/8709-OO03$3.00/0
3970
Blood, Vol 87,No 9 (May 1). 1996: pp 3970-3976
From www.bloodjournal.org by guest on July 12, 2017. For personal use only.
3971
EMPTY HLA CLASS I\ ON CORD BLOOD B CELLS
on the cell surface of fetal B cells and might provide the
explanation for this phenomenon.
MATERIALS AND METHODS
Preparation of fetal and adult B lymphocytes. B lymphocytes
were prepared from the spleens of organ donors. First, mononuclear
cells were separated on a Ficoll gradient, and T cells were depleted
by rosetting with sheep red blood cells. The final preparation contains
more than 90% CD19+ cells. Resting cells were recovered after
Percoll density gradient fractionation (50% to 60% Percoll fraction)
as previously described.,
Cord blood samples from healthy donors were collected by cordocentesis, mononuclear cells were treated with anti-CD19 antibody
coupled to magnetic beads (Dynal, Oslo, Norway), and CD19’ cells
were detached according to the manufacturer’s instructions. The final
population contained more than 90% of CD19’. For immunoprecipitation, cord blood mononuclear cells were depleted by two cycles
of incubation with anti-CD2 antibody coupled to magnetic beads
(Dynal) following the manufacturer’s instructions. The final preparation contains >70% CD19* cells. Double labeling was used to demonstrate that only HLA class 11’ cells were CD19’ and vice versa.
For some experiments, purified CD4+ T cells from cord blood were
prepared with antLCD4 antibody coupled to magnetic beads as described for cord blood CD19+ cells. Cell separation by antibodies
coupled to magnetic beads was preferred to obtain a higher number
of cord blood B cells than could be obtained by rosetting with sheep
erythrocytes. In contrast, spleen is an abundant source of B cells
that permitted classical techniques of cell separation. Populations
of separated splenocytes and cord blood cells were tested by flow
cytometry for the expression of CD19, CD20, and surface immunoglobulins to ensure the purity of the B lymphocytes.
An EBV-transformed B cell line was used CA.SC.18 All the cells
were maintained in RPMI 1640 medium supplemented with 10%
fetal calf serum (FCS), 1 mmoVL sodium pyruvate, 2 mmol/L glutamine, 100 U/mL penicillin, and 0.5 mg/mL streptomycin.
Detection of HLA class II and Ii. A total of 5 X IO5 cells in
each sample was incubated with either D1.12 (anti-DR),” B7.21
(anti DP),” L2 (anti-DQ),*’ or LN2 (anti-Ii)22antibodies, washed
twice in cold phosphate-buffered saline (PBS) (1% FCS, 0.01%
NaN,) and further incubated with fluorescein isothiocyanate (FITC)
conjugated goat antimouse F(ab); (Boeringher, Mannheim, Germany) for 30 minutes on ice. The control used was FITC-conjugated
goat antimouse F(ab);. Cytofluorimetric data were acquired on a
FACscan (Becton Dickinson, Mountain View, CA).
Detection of empty HLA class II molecules by Ii-FITC fucation.
Ii-FITC fixation was performed as previously described.I6Incubation
of 5 X IO5 cells was performed with Ii-HTC (50 pg/mL final concentration) for 60 minutes on ice, washed twice in ice cold PBS (1%
FCS, 0.01% NaN,) and resuspended in PBS. FITC-labeled goat
antimouse F(ab); was used as negative control. Cytofluorimetric data
were acquired on a FACscan.
Immunoblotting of H U class 11 in cell lysates. Whole cell lysates of adult B cells, cord blood B cells, and cells from an EBVtransformed B-cell line were prepared by incubation of 3 x lo6cells
with 100 pL of NP40 containing sample buffer for 30 minutes on ice.
The lysates were centrifuged at 13,OOOg for 20 minutes to remove
membranes, nuclei, and debris. The supernatants were collected, and
30 fiL of each supernatant was diluted with an equal volume of
sample buffer containing 2.3% SDS. Samples were either incubated
at room temperature or heated at 50°C for 30 minutes. One additional
sample from each experiment was heated to 95°C for 5 minutes.
Samples were migrated on a 10% SDS polyacrylamide gel before
transfer to a membrane (Immobilon M) and immunobtotting with a
combination of MoAbs DA6.147 (anti-DRa) and SG520 (anti-DR
p). Detection was by the chemiluminescence system ECL (Amersham, UK). Bound antibodies were detected with a hydrogen peroxidase conjugated goat antimouse MoAb.
Immunoprecipitation of cell su$ace DR molecules. Surface proteins of 2 x IO6 CA cells, I O x lo6 B splenocytes, and 60 X lo6
cord blood B cells were labeled with 50 pCi (100 pCi for cord
blood) of t-buto~ycarbonyl-L-~~S
methionine, N-hydroxy succimidyl
ester (Amersham) for 30 minutes on ice. After lysis in 800 pl of
lysis buffer (6 mmol/L CHAPS, 50 mmoVL Tris-HC1 pH 6.8, 150
mmoVL NaC1, 5 mmol/L EDTA, 50 pmoVL phenylmethylsulfonyl
fluoride, 0.1 mmoVL iodoacetamide, 10 nmoVL aprotinin, 10 nmoll
L leupeptin, 10 pg/mL trypsin inhibitor, 10 nmolR. pepstatin), lysates were cleared by centrifugation for 30 minutes at 10,000g.
Supernatants were collected and precleared for 2 hours with 10 pL
of normal mouse serum and 100 pL of 50% protein G Sepharose
(Pharmacia, Uppsala, Sweden). A second preclear with 50 pL of
50% protein G Sepharose was performed before immunoprecipitation of DR molecules by D1.12 antibody at a final concentration
of 40 pg/mL overnight followed by precipitation with protein G
Sepharose. Immunoprecipitates were washed six times with lysis
buffer and resuspended in polyacrylamide gel electrophoresis (SDSPAGE) buffer (62.5 mmom Tris-HC1 pH 6.8, 10% glycerol, 5% pmercaptoethanol, and 2.3% SDS). The samples were either left at
room temperature for 30 minutes or boiled for 5 minutes before
migration on a 12.5% SDS polyacrylamide gel. After fixation in
10% acetic acid and treatment in Amplify (Amersham), the gel was
dried and autoradiographed at -80°C for 30 days.
Detection of HLA-DM B in cord blood B lymphocytes by reverse
transcriptase-polymerase chain reation (RT-PCR). A total of 3 X
lo6 cells were lysed in Trizol (GIBCO-BRL, France), and mRNA
was extracted according to the manufacturer’s instructions. Single
strand cDNAs were generated in a 50-pL mixture containing MuMlv reverse transcriptase (200 U) (GIBCO-BRL), Rnase inhibitor
(40 U) (Boehringer Mannheim), random hexaprimers pd (N) 6
(Pharmacia) and 5 pg of each RNA. PCR nucleotides used for amplification of the 3’ part of DMB (exon 3 to exon 6) were S’TGCAAGTAG-CCAAAACCACTC3’, as 5’ primer and S’GTTGTAGGA‘ITCTGCCTC3’ as 3‘ primer. PCR amplification was performed for
30 cycles of 1 minute denaturation at 94”C, 1 minute annealing at
54°C and 1 minute of extension at 72°C. Reaction mixtures included
0.2 p m o K of each primer, 200 pmol/L dNTPs, 1 U of Taq polymerase and 1OX Taq buffer. Amplified PCR fragments were analyzed
by electrophoresis on an 8% bis-acrylamide gel.
Mixed lymphocyte cultures. A total of 5 X IO4 irradiated (20
Gy) stimulating cells (cord blood cells, splenocytes, or autologous
cells) were cocultured with 5 x IO4 allogeneic mononuclear cells
from unrelated donors in triplicate in 96-well microplates. On day
5 of culture, 1 pCi of 3H thymidine (Amersham) was added to each
well. The incorporation of 3H thymidine was measured on day 6.
RESULTS
Low expression of HLA class 11 on cord blood versus
adult B lymphocytes. The expression of HLA class I1 was
compared on fetal and adult B lymphocytes. Splenocytes
were chosen as a source of adult B cells because of the
relative abundance of resting B lymphocytes in the spleen.
The three isotypes HLA-DR, DQ, and DP were consistently
detected on both cell types, although the level of expression
was considerably lower on fetal B as shown in Fig 1. The
lower level of expression (on average a lo2 shift in mean
fluorescence) was, therefore, with regard to the quantity of
class I1 expressed and not with regard to the number of cells
expressing class 11, which was greater than 85% in both
From www.bloodjournal.org by guest on July 12, 2017. For personal use only.
GARBAN ET AL
3972
SPL CTRL
CB Ctrl
0
N
0
0
0
R1 -H
FLl-H
~
loo
.... ....... .. . ... . ...........
10’
l o 2 lo3
FLl -H
lo4
0
0
FL1 -H
0
CB 87.21
SPL 8721
0
0
FL1 -H
FL1 -H
Fig 1. Cell surface expression of HLA DR DP DQ antigens and
li. Flow cytometry analysis shows a low expression of HLA class I1
molecules in cord blood 6 cells (CB) compared with 6 splenocytes
(Spl). DR, DP, and DO antigens were detected by D1.12, 67.21, and
L2 monoclonal antibodies, respectively. For each sample FITC-conjugated goat antimouse Flab); was used as a negative control (. .).
Surface expression of the invariant chain was detected by LN2 antibody. Acquisition parameters are identicalfor the cord blood and for
the splenic 6 cells in this experiment.
.
cases. Similar data were obtained using lymphocytes from
six different donors.
Lack of surface expression of li on cord blood B lvmphocyres. The invariant chain (li) is intracellularly associated
with HLA class I1 molecules,’,’ but has also been detected
The role of
on the surface of human B
surface Ii is not known, but it has been suggested that it
might serve to recycle cell surface class I1 molecules to an
endocytic ~ompartment.’~
Adult B splenocytes reproducibly
expressed low levels of surface Ii on more than 90% of the
population (Fig I). In contrast, surface expression of Ii was
never detected on fetal B cells (Fig 1). Transcripts of Ii
mRNA were, however, detected by RT-PCR, and immunoprecipitates of metabolically rather than surface labeled cells
clearly demonstrated the presence of intracellular Ii (data
not shown).
Empty HLA class I1 molecules can be detected on cord
blood B lymphocytes. The soluble Ii (Ii sol) can be used
to detect empty HLA class I1 molecules. We have previously
shown that preloading of HLA class 11 molecules with peptide blocks subsequent binding by soluble Ii, and the binding
of the Ii sol was also inhibited by the MoAb L243, which
is believed to bind at or near the peptide binding site.“ The
binding site of L243 has recently been mapped to residues
on both the (Y and the B chains.” On the contrary, bacterial
superantigens, which are believed to bind outside the peptide
binding site, did not inhibit Ii sol binding.” Fixation of fluoresceinated Ii to fetal lymphocytes was substantial, and approximately half of the cell population bound a significant
amount (Fig 2). The binding to cord blood B cells is considerably higher than to adult B splenocytes especially when
the radically different levels of surface expression of HLA
class I1 molecules are taken into account (Figs 1 and 2). The
background binding of Ii sol to class 11- T cells is shown (Fig
2). After 18 hours of culture, there was a marked increase in
Ii-FITC binding to cord blood B cells and a slight increase
in Ii FITC binding to adult B splenocytes (Fig 2). The steadystate level of cell surface class I1 expression remains unchanged after overnight culture (not shown). Because the
half-life of peptide complexes in human antigen presenting
cells is normally very close to that of the class I1 molecules
themselves,” the increased Ii-FITC binding might thus reflect either a dissociation of loosely bound peptide from cell
surface HLA molecules or, alternatively, a combination of
increased turnover of class I1 molecules combined with a
poor capacity to process antigen, thereby generating empty
molecules on the cell surface. We conclude that cord blood
B, as opposed to adult B, have empty HLA class I1 molecules
on the cell surface.
Immunoblotting of whole cell lysates shows the presence
of large HLA class I1 aggregates in cord blood B cells. The
validity of the SDS stability assay as a means to distinguish
between empty and peptide-loaded class I1 molecules has
been confirmed by various appro ache^.""^.'^.^'^^* In the
mouse, SDS-stable compact dimers are formed after peptide
binding and have an apparent molecular weight of 50 to 60
kD, while unstable dimers dissociate into free (Y and p chains
at room temperature in the presence of SDS.”.’9 Western
blotting was performed to detect different forms of HLA
class I1 molecules in fetal B cells and in an adult B-cell line
(CA). As expected, immunoblots of CA cell lysates gave
rise to a strong SDS-stable 60 kD band that corresponds to
stable (YP heterodimers. Free (Y and ,f3 chains could also be
From www.bloodjournal.org by guest on July 12, 2017. For personal use only.
3973
EMPTY HLA CLASS II ON CORD BLOOD B CELLS
(lane 3) in five independent experiments, although the intensity of the bands corresponding to free a and p chains increases with heating (lanes 1 and 2). This increase is probably due to the disaggregation of large, nonspecific aggregates
of HLA class I1 molecules present in lysates of fetal cells
that were visible after transfer of the stacking gel (Fig 3, see
arrow). The formation of large aggregates may, or may not,
be an experimental artifact, but they were not detected in
lysates from an EBV-transformed B-cell line. What is important to the present issue, however, is that only empty
class I1 molecules, or class I1 molecules loosely associated
with peptide, tend to aggregate.'4.30
120-kDSDS stable HLA class 11 molecules are preferentiall>)detected on the su$ace of fetal B lymphocytes. Because low levels of SDS stable class I1 polypeptides could
have gone undetected in the immunoblot assays, we per-
UJO
0
FLI -H
t
N
N
0
0
FL1-H
FLI -H
- 60 kD
-u
Fig 2. Detection of empty HLA class II molecules on cord blood B
cells by fixation of Ii-FITC. Flow cytometry detection of empty HLA
class II molecules on EBV-transformed B-cell line (CAI, resting B
splenocytes (Spl), and cord blood B cells (CB). Cord blood T cells
were used as a negative control. For each sample FITC-conjugated
goat antimouse F(ab); was used as a negative control (. . I . li-FITC
did not fix CA t o any greater extent than the class 11- cord blood T
cells. There was minimal fixation t o the splenocytes, while there was
considerable fixation t o the cord blood B lymphocytes. Overnight
incubation (18 hours1 in medium (CB 18H, Spl 18HI considerably increased binding of li t o cord blood B cells.
-P
.
seen (Fig 3, lanes 4 to 6 ) . The increase of free a and /3
chains on heating the samples (lanes 4 & 5 ) correlates with
a decrease in the intensity of the 60-kD band, demonstrating
that the dissociated a and p chains derive from the latter.
Because the anti-/3 chain MoAb was less efficient than the
anti-cu chain MoAb in Western blotting, detection of the /3
chain was always poor. On the contrary, SDS stable 60 kD
heterodimers were never observed in fetal B-cell lysates
1
2
3
4
5
6
Fig 3. Western blotting detection of HLA DR molecules in whole
cell lysates. SDS containing supernatants of lysates from cord blood
B cells (lanes 1 t o 31 and from an EBV-transformed B-cell line [lanes
4 t o 6) were either incubated at room temperature for 30 minutes
(lanes 3 and 61, incubated at 50°C for 30 minutes (lanes 2 and 51, or
heated t o 95°C for 5 minutes (lanes 1 and 6). lmmunoblots of the
stacking gels are at the top of the figure. HIA class II molecules were
clearly visible in nonheated cord blood B-cell lysates (see arrow).
lmmunoblots of the separating gels detected SDS stable HLA class
II heterodimers in all but the boiled lysates of an EBV-transformed B
cell line [lanes 5 and 61 and only dissociated chains in the boiled
lysate (lane 4). On the contrary, SDS stable 60-kD heterodimam were
not detected in cord blood B-cell lysates, although dissociated chains
became visible on heating. Cord blood cell lysates from four different
donors were examined and the results shown are typical.
From www.bloodjournal.org by guest on July 12, 2017. For personal use only.
GARBAN ET AL
3974
1
2
3
4
5
6
7
8
120-
60-
a-
P-
120-
60-
6). In immunoprecipitates from cord blood B cells. the SDS
stable material instead migrated with an apparent molecular
weight of 120 kD (lanes 3 and 7). Dissociated chains were
visible after heating (lanes 4 and 8) corresponding with the
disappearance of the material migrating with an apparent
molecular weight of 120 kD. As this labeling method depends on the presence of amino groups, the CY chain is more
strongly labeled than the p chain. The 60-kD band was never
observed in fetal B-cell immunoprecipitates, not even after
a prolonged (3 months) exposure of the gel. The 120-kD
protein could correspond to the previously reported double
dimers or superdimers of class I1 molecules described in the
mouse.''
Detection of mRNA rfthe HLA DMB genes in cord blood
B cells. The DM molecule is an CY@
heterodimer that dissociates proteolytic Ii fragments from HLA class I1 molecules
in a peptide-loading compartment to promote loading of antigenic
Messenger RNA transcripts of the HLA
DMB gene were assayed by RT-PCR (Fig 5) in cord blood
B cells (lane 2), in adult B cells (lane I), and in an EBVtransformed B-cell line (lane 3) that was included as a positive control. Although these studies were not quantitative,
the HLA DMB transcript is clearly present in cord blood B
lymphocytes.
Equivalent stimulation of allogeneic responses bs cord
blood B and by adult B lymphocytes. To explore a functional aspect of the HLA class I1 molecules in cord blood B
lymphocytes, we performed mixed lymphocyte reactions
with irradiated B splenocytes or irradiated cord blood B cells
as stimulators. Table 1 shows that cord blood B cells, despite
the lower level of HLA class I1 expression, can induce an
allogeneic response equally as well as adult B.
DISCUSSION
a-
PFig 4. SDS stabilii of HLA class II molecules. Surface SDS stable
60kD HLA-DR complexes were detected on CA cells (lane 1).B splenocytes (lane 51, but not on cord blood B cells (lanes 3 and 7). Dissociated chains are visible in the heated samples (lanes 2, 4, 6, 8). The
only SDS stable material on the surface of cord blood B cells migrated
with an apparent molecular weight of 120 kD. Cell surface proteins
were labeled with "S. DR molecules were immunoprecipitatedwith
D1.12 antibody. Migration was performed in a 12.5% polyacrylamide
gel. For each cell type samples were boiled (lanes 2,4, 6, 81 and not
boiled (lanes 1, 3, 5, 71 to distinguish free a,p chains, and compact
DR complexes. The positions of SDS stable 120 kD forms (120). SDS
stable 60 kD forms 1601, and the cr and p chains are shown.
formed a sensitive cell surface labeling followed by immunoprecipitation of HLA DR molecules. Adult B cells displayed
stable 60 kD heterodimers that dissociated on boiling, thus
confirming the above results (Fig 4,lanes 1 to 2 and 5 to
A low level of major histocompatibility complex (MHC)
class I1 molecules has been previously reported as a major
feature in B-cell ontogeny in the mouse (31). We suggest
that a low level of HLA class I1 expression can similarly be
used to distinguish fetal B cells from adult B cells in the
human. Furthermore, we argue that the results shown in this
report, using a recombinant and soluble Ii chain as a molecular probe, demonstrate the existance of an important population of HLA class I1 molecules, which are not saturated with
peptide, on fetal B cells. The antigenic world seen by B cells
in utero is much impoverished compared with that of adult
splenocytes. After birth, the antigenic universe becomes exponentially larger, and the number of exogenous antigens
that must be internalized and proteolytically processed increases. However, it is at present not clear whether the high
proportion of empty class I1 molecules in fetal B cells is
actually due to the relative deficiency of exogenous antigen
in fetal blood. In theory, it could also result from the absence
of some accessory molecule needed for an efficient intracellular association between the class I1 molecule and peptide.
Two such molecules are the Ii chain and HLA-DM. The Ii
chain is required for the transport of HLA class I1 molecules
to an intracellular compartment where peptide-loading occurs.6-8.32The HLA DM polypeptide induces the dissociation
of Ii-derived peptides from HLA class 11 dimers to facilitate
From www.bloodjournal.org by guest on July 12, 2017. For personal use only.
EMPTY HLA CLASS I1 ON CORD BLOOD B CELLS
3975
1
2
.\I
3
Fig 5. Detection of HLA-DMB by RT-PCR. Detection of mRNA transcripts of the HLA DM B gene transcript was detected by RT-PCR in adult B cells (lane
1). cord blood B cells (lane 21, and an EBV-transformed B-cell line (lane 31.
peptide
Because messenger RNA transcripts of
both the Ii and HLA DMB genes could be detected by RTPCR, the fetal B cells seem to dispose of these two essential
functions for peptide presentation. Nonetheless. we have not
quantified either li or HLA-DMB, and a recent study describes the predominance of free cy and /3 chains over cy/3
dimers in mice lacking the Ii.'?
Somewhat surprising, the fetal B cells were equally efficient as adult splenocytes in generating a mixed lymphocyte
reaction. Such an example of an allogeneic response is believed to be peptide-dependent." We were unable to detect
the typical SDS stable (and, therefore, peptide loaded) 60kD class 11 heterodimer on fetal B cells, but after surface-
Table 1. Mixed Lymphocyte Reaction
Thymidine Incorporation x 1,000 cpm
Responding Cells
Stimulating cells
Experiment 1
CB1 CD19'
CB1 MNC
CB2 MNC
Spll
Donor 1
Donor 2
Donor 3
Experiment 2
CB3 CD19'
CB3 MNC
Spl 2
Donor 1
Donor 2
Donor 3
Donor 3
Donor 1
Donor 2
93
114
94
82
0.9
92
113
84
108
66
69
56
0.2
90
39
45
56
15
30
33
1.1
33
36
119
1.1
71
7.4
72
83
127
92
1.1
39
53
39
159
70
103
0.2
Irradiated stimulating cells: purified CD19' cells (CD19) or MNC
from cord blood (CB1, CB2, and CB31 and B splenocytes (Spll and
Sp12) were cocultured with unrelated control MNC (Donor 1, Donor
2, and Donor 3). 'H-thymidine incorporation was measured at day 6.
3H-thymidine incorporation in control cultures is shown. The incorporation of 'H-thymidine was indicative of a successful allogeneic stimulation whether B and MNC were of either fetal or adult origin.
Abbreviation: MNC, mononuclear cells.
labeling and immunoprecipitation. an SDS stable 120-kD
species was visible. This form might represent a double dimer or superdimer as suggested from both murine studiesI5
and from crystallographic studies in the human." Schafer
and Piercen5have recently argued that these superdimers are
important for the generation of low-affinity T-cell responses.
It thus seems feasible that the mixed lymphocyte reaction is
stimulated by the SDS stable 120-kD molecule in fetal B
cells.
On the other hand, when whole cell lysates were analyzed
by immunoblotting. a major fraction of the class I1 antigens
in fetal B cells were not detectable in nonboiled samples. On
boiling however, free cy- and &chains were clearly visible.
Subsequent experiments with nonboiled samples showed
that these molecules were aggregated and unable to enter
the SDS-polyacrylamide gel. In fact. it is known that empty
class I1 molecules have a strong tendency to form aggreg a t e ~ " , ~and
' consequently, the presence of a large nonspecific aggregates of class II molecules in cord blood B cells
is likely to reflect the absence of bound peptide. We cannot
at present formally exclude the possibility of a loose association between peptides and class II molecules on fetal B cells.
as loosely bound peptides do not necessarily confer SDS
stability and because li sol could theoretically displace such
peptides under the experimental conditions used.
Cord blood mononuclear cells are increasingly used in
transplantation due to the lower incidence of graft-versushost reactions.'5 and it has been proposed that cord blood
transplantations could be useful in a wide variety of diseases
ranging from autoimmunity to human immunodeficiency virus (HIV) infection.'' Another interesting possibility that
now seems feasible would be to artificially charge empty
HLA class 11 molecules on fetal B cells with synthetic peptides for vaccination purposes.
REFERENCES
I . Splawsky JR. Jelinek DF. Lipsky PE: Delineation of the functional capacity of human neonatal lymphocytes. J Clin Invest 87545.
1991
2. Tucci A. Mouzaki A. James H, Bonnefoy JY. Zubler RH: Are
cord blood B cells functionally mature? Clin Exp lmmunol 84:389.
1991
3. Mooney NA. Grillot-Courvnlin C. Hivroz C. J u LY. Charron
From www.bloodjournal.org by guest on July 12, 2017. For personal use only.
3976
D: Early biochemical events after MHC class I1 mehated signaling
on human B lymphocytes. J Immunol 145:2070, 1990
4. Kvist S, Wiman K, Claesson L, Peterson PA, Dobberstein B:
Membrane insertion and oligomeric assembly of HLA-DR histocompatibility antigens. Cell 29:61, 1982
5. Lamb CA, Cresswell P: Assembly and transport properties of
invariant chain trimers and HLA-DR invariant chain complexes. J
Immunol 148:3478, 1992
6. Roche PA, Cresswell P: Invariant chain association with HLA
DR molecules inhibits immunogenic peptide binding. Nature
345:615, 1990
7. Amigorena S, Drake JR, Webster P, Mellman I: Transient
accumulation of new class I1 MHC molecules in a novel endocytic
compartment in B lymphocytes. Nature 369: 113, 1994
8. Qiu Y, Xu X, Wandinger-Ness A, Dalke DP, Pierce SK: Separation of subcellular compartments containing distinct functional
forms of MHC class 11. J Cell Biol 125:595, 1994
9. Cresswell P: Assembly, transport and function of MHC class
I1 molecules. Annu Rev Immunol 12:259, 1994
10. Sadegh-Nasseri S, Germain RN: A role for peptide in determining MHC class I1 structure. Nature 35:167, 1991
11. Germain RN, Rinker AG: Peptide binding inhibits protein
aggregation of invariant free class I1 dimers and promotes surface
expression of occupied molecules. Nature 363:725, 1993
12. Hedley ML, Urban RG, Strominger JL: Assembly and peptide
binding of major histocompatibility complex class II heterodimers
in an in vitro translation system. Proc Natl Acad Sci USA 91:10479,
I994
13. Neefjes JJ, Ploegh HL: Inhibition of proteolytic activity by
leupeptin blocks surface expression of MHC class I1 molecules and
their conversion to SDS resistant ap heterodimers in endosomes.
EMBO J 11:411, 1992
14. Stem LJ, Wiley DC: The human class I1 MHC protein HLA
DR-1 assembles as empty ap heterodimers in the absence of antigenic peptide. Cell 68:465, 1992
15. Schafer P, Pierce S: Evidence for dimers of MHC molecules
in B lymphocytes and their role in low affinity T cell response.
Immunity 1:699, 1994
16. Ericson M, Sundstrom M, Sansom DM, Charron DJ: Mutually
exclusive binding of peptide and invariant chain to major histocompatibility class I1 antigens. J Biol Chem 269:26531, 1994
17. Weber DA, Terrell NK, Zhang Y, Strindberg G, Martin J,
Rudensky A, Braunstein NS: Requirement for peptide in alloreactive
CD4' T cell recognition of class Il molecules. J Immunol 154:5153,
1995
18. Charron DJ, MacDevitt HO: Analysis of HLA-D region associated molecules with monoclonal antibody. hoc Natl Acad Sci
USA 76:6567, 1979
19. Carrel S, Tosi R, Gross N, Tanigaki R, Carmagnola AL,
Accola RS: Subset of human Ia like molecules defined by monoclonal antibodies. Mol Immunol 18:403, 1981
20. Rubinstein D, Roska AK, Lipsky PE: Liver sinusoidal lining
cells express class I1 major histocompatibility antigens but are poor
GARBAN ET AL
stimulators of fresh allogeneic T lymphocytes. J Immunol 137:1803,
1986
21. Fermand JP, Schmitt C, Brouet C: Distribution of class I1 DQ
antigens normal and leukemic lymphoid cells. Eur J lmmunol
15:1183, 1985
22. Epstein AL, Marder RJ,Winter JN, Fox RI: Two new monoclonal antibodies (LNI LN2) reactive in B5 formalin fixed parffin
embedded tissues with follicular and mantle zone human lymphocytes and derived tumors. J Immunol 133:1028, 1984
23. Koch N, Moldenhauer G, Hofmann WJ, Moller P: Rapid
intracellular pathway gives rise to cell surface expression of the
MHC class I1 associated invariant chain (CD74). J Immunol
147:2643, 1991
24. Roche PA, Teletski CL, Stang E, Bakke 0, Long EO: Cell
surface HLA DR-Invariant chain complexes are targeted to endosomes by rapid internalization. Proc Natl Acad Sci USA 90:8581,
I993
25. Fu X, Karr RW: HLA-DRa chain residues located on the
outer loops are involved in non-polymorphic and polymorphic antibody binding epitopes. Hum Immunol 39:253, 1994
26. Lanzavecchia A, Reid PA, Watts C: Irreversible association
of peptides with class I1 MHC molecules in living cells. Nature
357:249, 1992
27. Benaroch P, Yilla M, Raposo G, It0 K, Miwa K, Geuze
HJ, Ploegh HL: How MHC class I1 molecules reach the endocytic
pathway. EMBO J 14:37, 1995
28. Nelson CA, Petzol SJ, Unanue ER: Identification of two distinct properties of class I1 major histocompatibility complex associated peptides. Proc Natl Acad Sci USA 90:1227, 1993
29. Dezin LK, Cresswell P HLA DM induces CLIP dissociation
from MHC class I1 ab dimers and facilitates peptide loading. Cell
82:155, 1995
30. Takaosu NT, Lower JA, Robertson EJ, Bikoff EK: Major
histocompatibility class I1 peptide occupancy, antigen presentation
and CD4' T cell function in mice lacking the p41 isoform of invariant chain. Immunity 3:385, 1995
3 1. Hayakawa K, Tarlinton D, Hardy R: Absence of MHC class
I1 expression distinguishes fetal from adult B lymphopoiesis in mice.
J Immunol 152:4801, 1994
32. Tulp A, Verwoed D, Dobberstein B, Ploegh HL, Pieters J :
Isolation of the intracellular MHC class I1 compartment. Nature
369:120, 1994
33. Ceman S, Rudersdorf RA, Petersen JM, DeMars R: DMA
and DMB are the only genes in the class I1 region oh the human
MHC needed for class I1 associated antigen processing. I Immunol
154:2545, 1995
34. Arimilli S, Cardoso C, Mukku P, Baichwal V, Nag B: Refolding and reconstitution of functionally active complexes of human
leukocyte antigen DR2 and myelin basic protein peptide from recombinant a and p polypeptide chains. J Biol Chem 270:971, 1995
35. Wagner JE, Kernan NA, Broxmeyer HE, Gluckman E: Allogeneic umbilical cord blood transplantation report of results in 26
patients. Blood 10:86, 1993 suppl 1
36. Thompson C: Umbilical cords: Turning garbage into clinical
gold. Science 268:805, 1995
From www.bloodjournal.org by guest on July 12, 2017. For personal use only.
1996 87: 3970-3976
Detection of empty HLA class II molecules on cord blood B cells
F Garban, M Ericson, C Roucard, C Rabian-Herzog, H Teisserenc, E Sauvanet, D Charron and N
Mooney
Updated information and services can be found at:
http://www.bloodjournal.org/content/87/9/3970.full.html
Articles on similar topics can be found in the following Blood collections
Information about reproducing this article in parts or in its entirety may be found online at:
http://www.bloodjournal.org/site/misc/rights.xhtml#repub_requests
Information about ordering reprints may be found online at:
http://www.bloodjournal.org/site/misc/rights.xhtml#reprints
Information about subscriptions and ASH membership may be found online at:
http://www.bloodjournal.org/site/subscriptions/index.xhtml
Blood (print ISSN 0006-4971, online ISSN 1528-0020), is published weekly by the American
Society of Hematology, 2021 L St, NW, Suite 900, Washington DC 20036.
Copyright 2011 by The American Society of Hematology; all rights reserved.