Tolerance to solid organ transplants through transfer of MHC class II genes

Abstract

Donor/recipient MHC class II matching permits survival of experimental allografts without permanent immunosuppression, but is not clinically applicable due to the extensive polymorphism of this locus. As an alternative, we have tested a gene therapy approach in a preclinical animal model to determine whether expression of allogeneic class II transgenes (Tg’s) in recipient bone marrow cells would allow survival of subsequent Tg-matched renal allografts. Somatic matching between donor kidney class II and the recipient Tg’s, in combination with a short treatment of cyclosporine A, prolonged graft survival with DR and promoted tolerance with DQ. Class II Tg expression in the lymphoid lineage and the graft itself were sequentially implicated in this tolerance induction. These results demonstrate the potential of MHC class II gene transfer to permit tolerance to solid organ allografts.

Full text

Introduction
The use of potent immunosuppressive drugs in clinical
transplantation has extended the survival of vascular-
ized allografts despite genetic disparities at the MHC
class I and class II loci (1). Drug-induced nonspecific
immunosuppression is, however, associated with clini-
cal complications (2) that potentially could be avoided
by inducing graft-specific tolerance. Permanent graft-
specific unresponsiveness has previously been demon-
strated in miniature swine treated with cyclosporine A
(CsA) for only 12 days, provided that the donor and
recipient shared identity at the class II locus (3).
The beneficial impact of class II matching on the
outcome of renal allografts has also been document-
ed in clinical transplantation (4). Since matching
exclusively for class II would be difficult to achieve in
clinical settings, we tested the principle of creating a
somatic match by introducing class II genes into
hematopoietic cells of prospective allograft recipi-
ents. A pilot study in one animal has recently con-
firmed the technical feasibility of this approach and
documented that expression of an allogeneic pig
class II DRB cDNA in bone marrow–derived (BM-
derived) cells could modulate T-cell reactivity to
donor antigens (5). In this report, we present evi-
dence for induction of operational transplantation
tolerance to fully allogeneic renal allografts follow-
ing the introduction of a graft-matched DR or DQ
allele into the recipient’s immune cells.
Methods
Animals and protocols used in this study. Massachusetts
General Hospital Miniature Swine (3–5 months old)
received either allogeneic or syngeneic MHC class II
cDNA through transduction of autologous bone mar-
row cells (BMCs) (Table 1). The development of the
GS4.5 and GS7.3 recombinant retrovirus vectors con-
taining the swine DRB cDNA of the dand calleles,
respectively, as well as the development of the con-
struct NAB.V expressing the DQA and DQB cDNA
from the chaplotype have been described (5–7).
Derivation of high titer retrovirus producer clones,
conditions of virus producer cells, collection of virus-
containing supernatants, and transduction of whole
BMCs were as described (8). Cytokines were used as
follows: murine kit ligand and human IL-3/GM-CSF
fusion molecule PIXY321, both provided by Immunex
Corp. (Seattle, Washington, USA), or recombinant
porcine cytokines IL-3, GM-CSF, and stem cell factor
(SCF), kindly provided by BioTransplant Inc. (Boston,
Massachusetts, USA). Recipients were conditioned
with 5 Gy whole body irradiation given at days –1 and
The Journal of Clinical Investigation | January 2001 | Volume 107 | Number 1 65
Tolerance to solid organ transplants
through transfer of MHC class II genes
Kai-C. Sonntag,1David W. Emery,2Akihiko Yasumoto,3Gary Haller,4Sharon Germana,1
Tomasz Sablinski,5Akira Shimizu,6Kazuhiko Yamada,1Hideaki Shimada,3Scott Arn,1
David H. Sachs,1and Christian LeGuern1
1Transplantation Biology Research Center, Massachusetts General Hospital/Harvard Medical School, Boston,
Massachusetts, USA
2Division of Medical Genetics, University of Washington, Seattle, Washington, USA
3Department of Surgery (II), Chiba University School of Medicine, Chiba, Japan
4Klinikum Mannheim GmbH, Chirurgische Klinik, Mannheim, Germany
5Novartis Pharmaceuticals, Clinical Research and Development, East Hanover, New Jersey, USA
6Department of Pathology, Nippon Medical School, Tokyo, Japan
Address correspondence to: Christian LeGuern, Transplantation Biology Research Center, Massachusetts General Hospital,
MGH-East, Building 149-9019, 13th Street, Boston, Massachusetts 02129, USA.
Phone: (617) 726-4059; Fax: (617) 726-4067; E-mail: leguern@helix.mgh.harvard.edu.
Received for publication August 10, 2000, and accepted in revised form November 21, 2000.
Donor/recipient MHC class II matching permits survival of experimental allografts without per-
manent immunosuppression, but is not clinically applicable due to the extensive polymorphism
of this locus. As an alternative, we have tested a gene therapy approach in a preclinical animal
model to determine whether expression of allogeneic class II transgenes (Tg’s) in recipient bone
marrow cells would allow survival of subsequent Tg-matched renal allografts. Somatic matching
between donor kidney class II and the recipient Tg’s, in combination with a short treatment of
cyclosporine A, prolonged graft survival with DR and promoted tolerance with DQ. Class II Tg
expression in the lymphoid lineage and the graft itself were sequentially implicated in this toler-
ance induction. These results demonstrate the potential of MHC class II gene transfer to permit
tolerance to solid organ allografts.
J. Clin. Invest. 107:65–71 (2001).
See related Commentary on pages 33–34

0 (10 Gy total) prior to reinfusion of transduced BMCs
(0.4–1.3 ×108cells/kg). Allogeneic kidney transplan-
tation (KTx) was performed either 150 or 273 days
after bone marrow transplantation (BMT) with CsA
treatment (10 mg/kg per day) for 12 days, starting on
the day of transplantation (9).
Detection of BM transduction rates by G418-resistant GM-
CFUs. G418-resistant GM-CFUs (G418rGM-CFUs)
were evaluated as previously described (8), and the net
percent of transduced G418rGM-CFUs was calculated
by subtracting untransduced control G418rGM-CFUs
from transduced G418rGM-CFUs at cytotoxic con-
centrations of G418.
Detection of transgene mRNA by RT-PCR. DNA and
RNA preparations from whole blood cell (WBC) and
BM samples were done as previously described (5). For
cDNA preparation, 3–5 µg of sample RNA was first
DNase-digested (Deoxyribonuclease I, Amplification
Grade; Life Technologies Inc., Gaithersburg, Mary-
land, USA) under conditions recommended by the
manufacturer, and then transcribed into cDNA using
the SuperScript Preamplification Kit (Life Technolo-
gies Inc.). DNA and cDNA were analyzed by a nested
PCR technique using oligonucleotide primers specific
for vector DRB, DQB, and long terminal repeat (LTR)
sequences as follows: DRB-1 (5′-CTCAGAGTGGAGAG-
GTCTACAG-3′), and DQB-1 (5′-AGATAGAGGAAGGCAC-
GACC-3′) with LTR-1 (5′-TACCACAGATATCCTGT TTG-
GCC-3′) in the first PCR reaction, and the primers
DRB-2 (5′-GTCACAGTGG AATGGAGGGCAC-3′), and
DQB-2 (5′-TTAGGAACGGAGACTGGACC-3′) with LTR-
2 (5′-GTTCCATCTGTTCCTGACCTTG-3′) in the second
PCR reaction. Conditions for PCR reactions were: First
PCR: in a total reaction volume of 50 µl; 0.5–1 µg DNA
or 2–15 µl of the cDNA reaction, 1.4 nM of each
primer, 100 mM dioxynucleotide triphosphates
(dNTP), 1X Assay Buffer A (Fisher Scientific, Pitts-
burgh, Pennsylvania, USA), 2.5 units of Taq I DNA
polymerase (Fisher Scientific); amplification in a PTC-
100 MJ Thermocycler (MJ-Research Inc., Waltham,
Massachusetts, USA): denaturing step at 94°C for 1
minute, annealing step at 55°C for 30 seconds, and
amplification step at 72°C for 45 seconds, for 35
cycles. Second PCR: in a total reaction volume of 50 µl;
2.5 µl of the first PCR reaction, 1.4 nM of each primer,
100 mM dNTP, 1x Opti-Prime buffer no. 6 (Strata-
gene, La Jolla, California, USA), 2.5 units of Taq I DNA
polymerase (Fisher Scientific); amplification as
described above (30 cycles). The exon 3′-specific actin
primers 5′-AACCCCAAGGCCAACCGCGAGAAGATGACC-
3′and 5′-GGTGATGACCTGGCCGTCAGGCAGCTCGTA-
3′were used in the PCR experiments to test the quali-
ty of the cDNA template. PCR products were analyzed
on 3% NuSieve GTG electrophoresis gels (BioWhit-
taker Molecular Applications, Rockland, Maine, USA)
stained with ethidium bromide.
Detection of transgene expression in different hematopoietic
cell lineages. T-cell and macrophage purification proce-
dures included either flow cytometry (FCM) cell sort-
ing (FACScan II; BD Immunocytometry Systems, San
Jose, California, USA) or positive magnetic bead selec-
tion using the MACS system (Miltenyi Biotec, Auburn,
California, USA). T and natural killer (NK) cells were
sorted by using the anti-pig CD2 (MSA4) (10) and
monocytes/granulocytes (M/G) by an anti-pig M/G
(74-22-15A) (11) murine mAb. FCM-sorted cells were
stained at saturating Ab concentrations for 30 minutes
at 4°C. MACS cell purification involved cell incubation
with Ab for 15 minutes at 4°C followed by interaction
with rat anti-mouse IgG1 or IgG2a mAb coupled to
superparamagnetic MACS microbeads. Cell-bead com-
plexes were then passed through a separation column
(BS) placed in a magnetic separator. Cells collected as
the magnetic fraction were analyzed in FCM assays to
evaluate the T-cell as well as monocyte phenotype and
respective contamination with other cell types. Purity
of cell fractions ranged from 97 to 99%.
Histopathology of kidney biopsies. Sequential wedge kid-
ney biopsies were performed on postoperative days
(PODs) 30, 60, 100, and >100 through a flank incision.
Tissues were stained using hematoxylin and eosin, and
coded slides were examined according to standard
66 The Journal of Clinical Investigation | January 2001 | Volume 107 | Number 1
Table 1
Animals used in this study
Animal no. Transduction KTx donor KTx Outcome of first Outcome of second
(SLA haplotype) (Gene→haplotype) (Haplotype) (days after BMT) kidney graft kidney graft
Allogeneic gene transfer
10736 (cc) IId(DRB)→Ic/IIcdd (Id/IId) 150 sacrifice day 1120
11585 (gg) IIc(DRB)→Ic/IIdjj (Ia/IIc) 150 graft-loss day 322 graft-loss day 29
11782 (cc) IId(DRB)→Ic/IIcdd (Id/IId) 273 graft-loss day 184 graft-loss day 108
12307 (dd) IIc(DQA+B)→Id/IIdcc (Ic/IIc) 150 sacrifice day 500
12426 (dd) IIc(DQA+B)→Id/IIdcc (Ic/IIc) 150 tolerant day 658 tolerant day 139
Syngeneic gene transfer
10807 (cc) IIc(DRB)→Ic/IIcdd (Id/IId) 150 rejection day 120
11077 (cc) IIc(DRB)→Ic/IIcdd (Id/IId) 150 rejection day 5
11625 (cc) IIc(DRB)→Ic/IIcdd (Id/IId) 150 rejection day 82
13286 (cc) IIc(DQA+B)→Ic/IIcdd (Id/IId) 150 rejection day 39
Animals as well as BMC transductions, BMT, and KTx are described in Methods. The SLA haplotypes and alleles at the class I (I) and class II (II)
loci are indicated.

pathologic criteria (12) based on the “Banff Schema” of
the International Society of Nephrology as applied to
clinical KTx (13). All animals that died underwent com-
plete postmortem examinations.
Mixed lymphocyte reaction assays. Mixed lymphocyte
reaction (MLR) culture conditions for porcine T-cell
proliferation have been described previously (14).
Briefly, 4 ×105peripheral blood lymphocyte (PBL)
responders from experimental (Table 1) or naive con-
trol (swine leukocyte antigen [SLA] dhaplotype) ani-
mals were mixed with an equal number of irradiated
(25 Gy) stimulator PBLs (SLA chaplotype and Yucatan
pig [YUC] as third-party control) and were incubated
in 200 µl of standard MLR media using flat-bottomed
96-well plates (Costar, Cambridge, Massachusetts,
USA). After 5 days of incubation, 1 µCi of 3H-thymi-
dine was added to each well, followed by an additional
5-hour incubation. 3H incorporation into newly syn-
thesized polynucleotides was determined in triplicate
samples by liquid scintillation. Stimulation indices (SI)
were derived by dividing cpm of individual allo-
responses with the corresponding cpm of self-response.
The percentages of anti-donor versus third party
response were calculated as follows: SI experimental /
SI third party YUC ×100. Responses against third-
party YUC ranged between 7,500 and 241,000 cpm.
Statistical analyses. Group comparisons were made by
Student’s ttest. Values of BM transduction rates were
evaluated by G418rGM-CFUs as previously described
(8). Graft survival data were based on blood creatinine
levels and histopathology of rejected grafts. Pvalues less
than 0.05 were considered statistically significant.
Results
Gene transfer of MHC class II in BMCs of recipients. As in
humans, the pig DRαpolypeptide sequence is invari-
ant, whereas the DRβ, DQα, and DQβsequences are
polymorphic (15). To ensure the transfer of a fully
allogeneic class II, retroviral vectors were constructed
to express either the DRβ(DRB cDNA) chain along
with the gene for neomycin resistance (Neor) (6, 8) or
the DQαand β(DQA+B) chains (7). Animals includ-
ed in this study received either allogeneic or control
syngeneic DRB or DQA+B cDNA through transduc-
tion of autologous BMCs (Table 1) (5, 8). Levels of
DRB vector expression were evaluated by the percent-
age of G418rGM-CFUs in serial rib biopsies (8). We
have previously shown that initial BMC transduction
rates of 8.9 ± 2.1% were not sufficient to permit the
induction of specific tolerance to transgene-matched
(Tg-matched) renal allografts (5). The addition of
murine c-kit ligand and human IL-3/GM-CSF fusion
molecule PIXY321 (8) or the recombinant porcine
cytokines IL-3, GM-CSF, and SCF substantially
increased the initial transduction levels to 21.3 ± 3.5%
G418rGM-CFUs (P< 0.02 as compared with trans-
duction without cytokines). Drug-resistant GM-CFUs
were still detected in BMCs of all transplanted recipi-
ents at the time of immunological reconstitution at
approximately 120 days after BMT, as well as at the
time of KTx (data not shown). The percentage of
G418rGM-CFUs declined uniformly in all animals, a
decay that may correlate with the turnover rate of
transduced cells. Due to the absence of a selectable
marker in the DQ vector, transduction rates were esti-
mated by PCR analysis of randomly picked unselect-
ed GM-CFUs. DQ proviral sequences were detected in
approximately 10% of single colonies, suggesting
lower transduction rates as compared with those
observed with the DR gene transfer (data not shown).
Long-term expression of the class II Tg. DRB and DQB Tg
presence and transcription in vivo was monitored in
BM and WBC samples by a nested PCR designed to
amplify proviral DNA and cDNA, respectively. Using
these assays, transgenic DRB mRNA was detected up to
413 days after BMT in samples from long-term
graft–accepting animals (Figure 1). Transcription of
the DRB Tg was confirmed at time of KTx as shown
previously (5), but ceased around 171 days before graft
loss (no. 11585, Figure 1b). In contrast, DQ Tg presence
and expression were detected only up to 84 days after
BMT for animal no. 12307 and were, consequently, no
longer present at time of KTx (Figure 1b, arrows).
DRB Tg expression in hematopoietic lineages. Lineage
distribution of DRB Tg expression was evaluated by
RT-PCR of RNA isolated from antibody-sorted CD2+
cells or M/G subpopulations from peripheral blood
(Table 2). Transcription of the Tg was observed at var-
ious time points in the M/G lineage of the five animals
studied, including those recipients whose BMCs were
transduced in the absence of cytokines with resultant
lower transduction efficiencies (5). In contrast, the
long-term graft–accepting animals, which received
BMCs transduced in the presence of cytokines, exhib-
ited additional Tg expression in the CD2+lym-
phoid/NK lineage (Table 2). The short expression
course of Tg in DQ-engineered pigs (Figure 1) did not
allow similar lineage analyses.
Graft survival in genetically engineered pigs. The impact
of DR or DQ Tg matching on the immunity of the
class II–engineered animals was evaluated by chal-
The Journal of Clinical Investigation | January 2001 | Volume 107 | Number 1 67
Table 2
Detection of Tg expression in different hematopoietic cell lineage by
RT-PCR
Last day positive
Animal for Tg RT-PCR Graft survival
no. Cytokines Myeloid Lymphoid (days)
10736 + 364 364 1120
11782 + 385 385 184
10660A– 154 ∅10
10680A– 154 ∅22
10697A– 154 ∅40
Purification of T-cell and M/G fractions, as well as detection of Tg transcrip-
tion, was as described in Methods. + and – indicate the presence or absence
of cytokines during BMC transduction, respectively. ∅, no detection of Tg
transcripts. AAnimals no. 10660, 10680, and 10697 received autologous BMC
transduced with an allogeneic MHC class II DRB cDNA in the absence of
cytokines as described (5).

lenging the BM-reconstituted recipients with kidney
grafts that were class II identical to the Tg but fully dis-
parate to the recipient MHC. All pigs received a graft
either at 150 or 273 days after BMT along with a 12-
day course of CsA to prevent T-cell reactivity to the
class I disparity (3) (Table 1). As shown in Figure 2c,
control animals engineered to express the syngeneic
class II Tg rejected their transplants between 5 and 120
days (mean survival time = 60 ± 50.1 days), a time
frame similar to that reported for the survival of fully
mismatched kidneys under the same immunosup-
pressive regimen (3). These animals never exhibited
normal renal function and eventually lost their grafts
as a consequence of acute cellular rejection (ACR)
(Figure 3b). Recipient animals engineered with the
allogeneic class II DR demonstrated either full accept-
ance (no. 10736) or significantly prolonged survival
(P= 0.0246) of DR Tg-identical grafts, with stable renal
function for at least 170 days (Figure 2a). Two of the
latter animals, nos. 11782 and 11585, eventually lost
their grafts at PODs 184 and 322, respectively, due to
a chronic allograft glomerulopathy (CAG) and/or
chronic rejection (Figure 3a). In contrast to the
DR-treated recipients, the animals that received the
allogeneic class II DQA+B genes developed full unre-
sponsiveness to the DQ Tg-identical transplants, with-
out exhibiting clinical or histopathological signs of
graft rejection (Figures 2b and 3c).
Possible mechanisms involved in tolerance induction. In an
attempt to understand the mechanisms which may
control hyporesponsiveness to kidney allografts, a sec-
ond set of kidney transplants, MHC-matched to the
first grafts, were implanted into two of the DR- and
one of the DQ-engineered animals without further
CsA treatment. The two nontolerant DR animals, nos.
11585 and 11782, accepted the second graft for 29
and 108 days, respectively, although blood creatinine
never reached basal levels (Figure 2a). Loss of the
grafts was attributed to ACR for no. 11585 and to
CAG for no. 11782. Survival of the second grafts in
these animals was markedly increased compared with
the 7-day survival of a first graft implanted in a
Tg-engineered pig without CsA treatment (data not
shown). The tolerant DQ animal, however, which was
68 The Journal of Clinical Investigation | January 2001 | Volume 107 | Number 1
Figure 1
PCR analysis of MHC class II DRB and DQB transgenes. (a) Results from a representative long-term graft–accepting animal, no. 11585
(DRB, top), and a tolerant animal, no. 12307 (DQA+B, bottom). WBC and BM samples, taken at various time points after BMT, were ana-
lyzed for actin and retroviral DRB or DQB cDNA. Specific PCR products for actin (414 bp), DRB (358 bp), and DQB (461 bp) are shown.
GS4.5-16 and STP29.20 are DRB- and DQA+B-virus producing cell clones, respectively. (b) RT-PCR and DNA-PCR analyses performed on
WBC and BM samples from long-term graft–accepting pigs. Filled, gray, and open bars indicate the presence of DRB or DQB Tg DNA, DRB
or DQB Tg transcription, and graft survival, respectively. KTx was performed at times indicated by arrows. Numbers over brackets show graft
persistence (in days) following cessation of Tg transcription.
Figure 2
Outcome of allogeneic kidney grafts in the
genetically engineered animals. Blood creati-
nine levels (mg/dl) of allogeneic DRB-engi-
neered (a) and DQA+B-engineered animals
(b) are displayed. Inset (c) shows results from
syngeneic DRB-engineered and DQA+B-engi-
neered control animals. Arrows indicate time
of second transplantation of Tg-matched kid-
ney grafts without CsA treatment.

already unresponsive to the first graft, accepted the
second SLA-matched kidney for the entire period of
the study without immunosuppression (139 days,
Figure 2b). The second transplant displayed normal
renal function as well as no histological indication of
cellular or humoral rejection. Since the BMT and the
short-term immunosuppressive treatment may have
had cumulative effects upon the induction of non-
specific hyporesponsiveness, the immune status of
this DQ recipient was tested with a third-party allo-
graft. When the second transplant was replaced with
an outbred pig allograft at POD 139, rejection
occurred within 5 days, demonstrating the specificity
of tolerance as well as the immune competence of the
recipient (data not shown).
To further investigate the role of the graft in this spe-
cific unresponsiveness, peripheral T-cell activation was
examined in three DQ-engineered (including one
rejector animal) and three DR-engineered animals by
MLR assays. Results in Figure 4 indicated that, prior
to kidney grafting, all pigs tested had full in vitro
peripheral T-cell immunocompetence to the subse-
quently grafted tissues. A progressive decrease of anti-
donor reactivity was, however, observed following KTx
in the two tolerant DQ animals. A comparable
decrease in anti-donor T-cell proliferation was detect-
ed in three DR-transferred animals 100 days after KTx
(Figure 4). These findings were indicative of a con-
comitant role of the class II Tg and the graft itself in
induction of tolerance. In addition, persistence of allo-
grafts in the engineered animals coincided with the
disappearance of peripheral T-cell cytotoxicity to
donor-specific antigens (results not shown).
Discussion
The present study demonstrates, for the first time to
our knowledge, the potential of somatic transgenesis
of MHC class II genes for the induction of transplan-
tation tolerance to vascularized organs. We have shown
that the transfer and expression of a single allogeneic
MHC class II molecule in BMCs of reconstituted ani-
mals allows prolonged survival of subsequent Tg-
matched kidney allografts without permanent phar-
macological immunosuppression. The aftermath of
somatic class II matching on graft survival is depend-
ent upon the nature of the class II genes transferred as
well as the timing of their expression. While early
expression of the allogeneic DR prolonged kidney graft
survival, DQ promoted the induction of tolerance.
Although transcription of proviral DR mRNA could
reach substantial levels as detected in WBC by North-
ern blotting up to 28 weeks after BMT (5), we believe
that overall the quantity of retroviral material tran-
scribed in peripheral cells was very low, since RT-PCR
was required for its detection (Figure 1). Alternatively,
a limited number of peripheral clones with higher lev-
els of Tg expression may have been present but were
not detected by our method of analysis. The results
from the lineage studies in the DR-engineered recipi-
ents (Table 2) supported this view by showing that Tg
expression in the lymphoid lineage coincided with
graft survival. This correlation also suggests that the
use of cytokines during BM transduction resulted in
the transduction of a subpopulation of tolerance-
inducing cells of lymphoid origin, possibly thymic
The Journal of Clinical Investigation | January 2001 | Volume 107 | Number 1 69
Figure 3
Histology of kidney biopsies from DRB- and DQA+B-treated ani-
mals, with hematoxylin and eosin staining. Arterioles (a) and
glomeruli (g) are indicated. (a) Long-term graft–accepting pig no.
11585 at POD 322. There are typical signs of chronic rejection
including chronic vasculopathy with intimal hyperplasia in small
arteries, CAG, focal infiltration (arrow), and diffuse interstitial
fibrosis. (b) Control animal no. 11625 at POD 82. Severe ACR with
acute and severe endothelialitis in small vessels, diffuse mononu-
clear cell infiltration, and acute allograft glomerulopathy. (c) Tol-
erant animal no. 12307 at POD 100. No evidence of ACR, CAG,
endothelialitis, or cellular infiltration.

dendritic cell precursors, which have also been impli-
cated as modulators of the anti-graft response (16–18).
All of the DRB-engineered pigs, with the exception
of animal no. 10736, eventually lost their grafts due
to CAG and/or chronic rejection (Figures 2a and 3a).
Graft loss was not mediated through sensitization to
donor-type antigens since these animals did not
develop detectable levels of donor-specific antibod-
ies. The fact that a DQ disparity between graft donor
and host persisted in the DR-treated recipients could
have accounted for the onset of delayed glomeru-
lopathy. The animals receiving the DQA+B Tg did
not, however, develop such a pathology (Figures 2b
and 3c). Alternatively, the differential effects of DR
and DQ on tolerance induction might be related to
a distinctive expression pattern, i.e., the surface
assembly of the DRβTg product with the DRαchain
relies on the endogenous production of the latter
chain in the targeted transduced cells, whereas any
DQA+B transduced cell type should be able to
assemble the DQ heterodimer regardless of its
intrinsic ability to express endogenous class II mol-
ecules. A final possibility may invoke posttranscrip-
tional mechanisms, which have been implicated in
the mutually exclusive expression of DR or DQ in
human cell lines (19). Interestingly, human progen-
itors of hematopoietic cells were shown to express
DR and DP before DQ (20–22).
Early expression of allogeneic class II had no effect
on the recipient’s peripheral anti-donor T-cell
response prior to KTx, as evaluated by MLR (Figure
4). In addition, transcription of the Tg was not
detectable in BMCs or WBC from DQ-engineered
animals long before KTx (Figure 1). This result
implies that induction of sustained hyporesponsive-
ness is not mediated solely by the class II Tg products
but rather by a combination of events, which may be
sequential, involving the transient expression of allo-
geneic class II (as dimers or peptides) and the pres-
ence of graft-associated antigens. The observation
that the donor-specific T-cell response diminished
only after kidney allograft implantation (Figure 4)
directly implicates graft antigens, rather than class II
microchimerism alone, in shaping graft-specific
unresponsiveness as has been observed in other mod-
els (23, 24). The fact that the concomitant presence
of the class II Tg and the graft itself was not required
for induction of specific unresponsiveness implies
that early, transient Tg expression may generate reg-
ulatory mechanisms, which persist after the cessation
of class II Tg transcription. Given that these regula-
tory mechanisms are likely to be class II–dependent,
and that our previous studies demonstrated selective
involvement of CD4 Th1 cells in kidney rejection (25)
as well as increased transcription of the IL-10 gene in
accepted grafts (26), we hypothesize that CD4+Th2
regulatory cells are involved in the early phase of tol-
erance induction. Such immunoregulatory control of
peripheral tolerance has been described in other
models (27, 28) and may operate by altering the acti-
vation pathway of antigen-presenting cells (18), local
inhibition of T-cell functions with cytokines such as
IL-4 and IL-10 (29–34), and/or direct interaction
with T effector cells (35).
Our results support the view that transduction of
recipient BMCs with allogeneic class II vectors for both
the αand βpolypeptide chains affects the regulatory
T-cell compartment of the recipient immune system to
generate local and specific unresponsiveness to subse-
quent class II–matched renal allografts. These data also
suggest that such an immunomodulatory therapy has
potential to downregulate T-dependent responses to
grafted tissues and autoimmune antigens as well.
Acknowledgments
We thank A. Foley, M. Murphy, A. Hansen, K. Allison,
V. Ferrara, and I. Lubenec for expert technical assis-
tance; I. McMorrow, J. Iacomini, J. Madsen, and S. Pil-
lai for critical reading of the manuscript; and L.A.
Bernardo for assistance in manuscript preparation.
K.-C. Sonntag and G. Haller were supported in part
by a grant from Deutsche Forschungsgemeinschaft
(DFG); D.W. Emery was supported in part by a fel-
lowship from the Medical Foundation Inc./Charles
A. King Trust (Boston, Massachusetts, USA). This
work was supported by NIH grants 2RO1 AI31046
and 5RO1 HL46532, and by the Yates Foundation.
Novartis Pharmaceuticals is also acknowledged for
generously providing CsA.
1. Vanrenterghem, Y.F. 1997. Impact of new immunosuppressive agents on
late graft outcome. Kidney Int. Suppl. 63:S81–S83.
2. Shaw, L.M., Kaplan, B., and Kaufman, D. 1996. Toxic effects of immuno-
suppressive drugs: mechanisms and strategies for controlling them. Clin.
Chem. 42:1316–1321.
3. Rosengard, B.R., et al. 1992. Induction of specific tolerance to class I dis-
parate renal allografts in miniature swine with cyclosporine. Transplanta-
70 The Journal of Clinical Investigation | January 2001 | Volume 107 | Number 1
Figure 4
Peripheral T-cell reactivity of class II–treated animals. MLRs were per-
formed as described in Methods. Anti-donor response is given as per-
cent of the third-party response used as individual control in each
experiment (anti-donor/third-party ×100). MLR assays were per-
formed the day of the first KTx (day 0) and at indicated days there-
after for DQ and DR animals. The second KTx in animal no. 12426
(POD 658 after first transplant) is indicated by an arrow. T-cell pro-
liferation of control animal no. 13286 was tested at PODs 0 and 35,
four days prior to graft rejection.

tion. 54:490–497.
4. Nojima, M., et al. 1996. The significant effect of HLA-DRB1 matching on
acute rejection in kidney transplants. Transpl. Int. 9(Suppl. 1):S11–S15.
5. Emery, D.W., et al. 1997. Expression of an allogeneic MHC DRB trans-
gene, through retroviral transduction of bone marrow, induces specific
reduction of alloreactivity. Transplantation. 64:1414–1423.
6. Shafer, G.E., et al. 1991. Expression of a swine class II gene in murine bone
marrow hematopoietic cells by retroviral-mediated gene transfer. Proc.
Natl. Acad. Sci. USA. 88:9760–9764.
7. Shimada, H., et al. 1999. MHC class II alpha/beta heterodimeric cell sur-
face molecules expressed from a single proviral genome. Hum. Gene Ther.
10:2397–2405.
8. Emery, D.W., Shafer, G.E., Karson, E.M., Sachs, D.H., and LeGuern, C.
1993. Retrovirus-mediated transfer and expression of an allogeneic major
histocompatibility complex class II DRB cDNA in swine bone marrow
cultures. Blood. 81:2460–2465.
9. Smith, C.V., et al. 1992. Successful induction of long-term specific toler-
ance to fully allogeneic renal allografts in miniature swine. Transplanta-
tion. 53:438–444.
10. Hammerberg, C., and Schurig, G.G. 1986. Characterization of mono-
clonal antibodies directed against swine leukocytes. Vet. Immunol.
Immunopathol. 11:107–121.
11. Pescovitz, M.D., Lunney, J.K., and Sachs, D.H. 1984. Preparation and char-
acterization of monoclonal antibodies reactive with porcine PBL. J.
Immunol. 133:368–375.
12. Colvin, R.B. 1996. The renal allograft biopsy. Kidney Int. 50:1069–1082.
13. Solez, K. 1994. International standardization of criteria for histologic
diagnosis of chronic rejection in renal allografts. Clin. Transplant.
8:345–350.
14. Pescovitz, M.D., Lunney, J.K., and Sachs, D.H. 1985. Murine anti-swine
T4 and T8 monoclonal antibodies: distribution and effects on prolifera-
tive and cytotoxic T cells. J. Immunol. 134:37–44.
15. LeGuern, C., et al. 1995. Retrovirus-mediated transfer of MHC class II
cDNA into swine bone marrow cells. J. Mol. Med. 73:269–278.
16. Sykes, M. 1996. Chimerism and central tolerance. Curr. Opin. Immunol.
8:694–703.
17. Matzinger, P. 1994. Tolerance, danger, and the extended family. Annu. Rev.
Immunol. 12:991–1045.
18. Field, E.H., Gao, Q., Chen, N.X., and Rouse, T.M. 1997. Balancing the
immune system for tolerance: a case for regulatory CD4 cells. Transplan-
tation. 64:1–7.
19. De Lerma, B., Sartoris, S., Tosi, G., Nicolis, M., and Accolla, R.S. 1994. Evi-
dence for a specific post-transcriptional mechanism controlling the
expression of HLA-DQ, but not -DR and -DP, molecules. J. Immunol.
153:4530–4538.
20. Robinson, J., Sieff, C., Delia, D., Edwards, P.A., and Greaves, M. 1981.
Expression of cell-surface HLA-DR, HLA-ABC and glycophorin during
erythroid differentiation. Nature. 289:68–71.
21. Falkenburg, J.H., et al. 1984. Polymorphic and monomorphic HLA-DR
determinants on human hematopoietic progenitor cells. Blood.
63:1125–1132.
22. Katz, F.E., Davis, L., Myers, C.D., and Greaves, M.F. 1985. Selective expres-
sion of class-II MHC antigens during hemopoietic differentiation. Exp.
Hematol. 13:1182–1187.
23. Hamano, K., Rawsthorne, M.A., Bushell, A.R., Morris, P.J., and Wood, K.J.
1996. Evidence that the continued presence of the organ graft and not
peripheral donor microchimerism is essential for maintenance of toler-
ance to alloantigen in vivo in anti-CD4 treated recipients. Transplantation.
62:856–860.
24. Ramsdell, F., and Fowlkes, B.J. 1992. Maintenance of in vivo tolerance by
persistence of antigen. Science. 257:1130–1134.
25. Giangrande, I., et al. 1997. Selective increase in CD4-positive graft-infil-
trating mononuclear cells among the infiltrates in class I disparate kid-
ney grafts undergoing rejection. Transplantation. 63:722–728.
26. Blancho, G., et al. 1995. Molecular identification of porcine interleukin-
10: regulation of expression in a kidney allograft model. Proc. Natl. Acad.
Sci. USA. 92:2800–2804.
27.Qin, S., et al. 1993. “Infectious” transplantation tolerance. Science.
259:974–977.
28. Van de Keere, F., and Tonegawa, S. 1998. CD4(+) T cells prevent sponta-
neous experimental autoimmune encephalomyelitis in anti-myelin basic
protein T cell receptor transgenic mice. J. Exp. Med. 188:1875–1882.
29. Buer, J., et al. 1998. Interleukin 10 secretion and impaired effector func-
tion of major histocompatibility complex class II-restricted T cells aner-
gized in vivo. J. Exp. Med. 187:177–183.
30. Groux, H., et al. 1997. A CD4+ T-cell subset inhibits antigen-specific T-
cell responses and prevents colitis. Nature. 389:737–742.
31. Powell, T.J.J., and Streilein, J.W. 1990. Neonatal tolerance induction by
class II alloantigens activates IL-4-secreting, tolerogen-responsive T cells.
J. Immunol. 144:854–859.
32. Maeda, H., Takata, M., Takahashi, S., Ogoshi, S., and Fujimoto, S. 1994.
Adoptive transfer of a Th2-like cell line prolongs MHC class II antigen
disparate skin allograft survival in the mouse. Int. Immunol. 6:855–862.
33. Onodera, K., et al. 1997. Type 2 helper T cell-type cytokines and the devel-
opment of “infectious” tolerance in rat cardiac allograft recipients. J.
Immunol. 158:1572–1581.
34. Scully, R., et al. 1997. A role for Th2 cytokines in the suppression of CD8+
T cell-mediated graft rejection. Eur. J. Immunol. 27:1663–1670.
35. Chai, J.G., et al. 1999. Anergic T cells act as suppressor cells in vitro and
in vivo. Eur. J. Immunol. 29:686–692.
The Journal of Clinical Investigation | January 2001 | Volume 107 | Number 1 71