Allosensitization Does Not Increase the Risk of
Xenoreactivity to ?1,3-Galactosyltransferase
Gene-Knockout Miniature Swine in Patients on
Transplantation Waiting Lists
Banny S. Wong, Kazuhiko Yamada, Masayoshi Okumi, Joshua Weiner, Patricia E. O’Malley,
Yau-Lin Tseng, Frank J. M. F. Dor, David K. C. Cooper, Susan L. Saidman, Adam Griesemer,
and David H. Sachs
Background. The recent availability of ?1,3-galactosyltransferase knockout (GalT-KO) miniature swine has eliminated
tion. Highly-allosensitized patients, who have poor prospects of receiving a suitable cross-match negative human organ,
Methods. Using flow cytometry and complement-dependent cytotoxicity (CDC) assays, we have tested a panel of 88
human serum samples from patients awaiting cadaveric renal allotransplantation for reactivity against: 1) human; 2)
standard miniature swine; and 3) GalT-KO peripheral blood lymphocytes (PBL) and cultured endothelial cells.
Results. Anti-swine IgM and IgG antibody binding, as well as CDC, were significantly attenuated on GalT-KO versus
standard swine. No correlation was found between the degree of anti-human panel reactive antibodies (PRA) and
xenoreactivity against either standard or GalT-KO miniature swine. Treatment of sera with dithiothreitol (DTT)
showed that the majority of remaining lymphocytotoxicity against GalT-KO swine was mediated by preformed IgM
antibodies. Patients with high alloreactivity but low anti-GalT-KO xenoreactivity were readily identified.
Keywords: Panel reactive antibodies, Xenotransplantation, Sensitization, HLA antibodies.
(Transplantation 2006;82: 314–319)
ural antibodies (XNA) against the galactose-?1,3-galactose
major limiting factor in the care of patients with end-stage
enzyme, appeared an insurmountable obstacle to clinical xeno-
transplantation (3, 4). Absorption of these XNA prevented hy-
peracute xenograft rejection (HXR) (5, 6), but they always
The development of the GalT knockout (GalT-KO)
miniature swine, a herd of MHC-defined animals which lack
a functional GalT gene and therefore do not express Gal an-
tigens, has offered new promise to this field (7). Two recent
grafts from GalT-KO miniature swine transplanted into ba-
of heterotopic, cardiac xenografts (9), respectively. Further-
more, when co-transplanted with tolerance-inducing donor
thymic tissue, the GalT-KO renal xenografts showed no evi-
dence of rejection throughout the clinical course for up to 83
Selecting the most appropriate human recipients for
porcine organs will be critical in the design of future clinical
This work was supported by National Institutes of Health Program
from the Ter Meulen Fund and Prof Michael-van Vloten Fund from
Transplantation Biology Research Center, Massachusetts General Hospital,
Harvard Medical School, Boston, MA.
Address correspondence to: David H. Sachs, M.D., Transplantation Biology
Research Center, Massachusetts General Hospital, Harvard Medical
School, MGH East Building 149-9019, 13th Street, Boston, MA 02129.
Received 2 March 2006. Revision requested 3 April 2006.
Accepted 4 April 2006.
Copyright © 2006 by Lippincott Williams & Wilkins
Transplantation • Volume 82, Number 3, August 15, 2006
trials of xenotransplantation. In renal transplantation, highly-
allosensitized potential recipients with broadly-reactive, pre-
formed antibodies against antigens from most potential
donors have poor prospects for receiving a suitable cross-
match negative human donor organ (10, 11). A previous
study from this laboratory demonstrated no correlation
between the level of anti-human panel reactivity in such
sera and their overall reactivity with swine lymphocytes
(12). However, several subsequent studies found evidence
for cross-reactivity of antibodies from allosensitized pa-
tient sera against swine leukocyte antigens (SLA), suggest-
ing that allosensitivity may increase the risk of rejection
against porcine organs (13–17). These subsequent studies,
however, relied on immunoadsorption of serum anti-Gal
antibodies and the use of Gal-expressing target cells, leav-
ing open the possibility for incomplete adsorption of the
anti-Gal antibodies (see Discussion).
The recent development of GalT-KO miniature
swine eliminates the potential effects of anti-Gal antibod-
ies and provides a unique opportunity to study the role
that anti-non-Gal antibodies may play in presensitization
to xenografts. In this study, we have utilized GalT-KO tar-
get cells to test whether patients with high PRA are at in-
creased risk for presensitization against inbred GalT-KO
Serum samples from 88 patients on the waiting list for
cadaveric renal transplantation at the Massachusetts General
Hospital (Boston, MA) were selected for this study. Allosen-
sitization was determined as percent of panel reactive anti-
microlymphocytotoxicity assay using lymphocytes from a
allosensitized patients with PRA greater than 80%, 21 were
from moderately-allosensitized patients with PRA from 20%
to 80%, and 38 were from mildly or non-allosensitized pa-
tients with PRA less than 20%.
Distinction between IgG- and IgM-mediated reactivity
was determined by treatment of serum with dithiothreitol
(DTT, Sigma Chemical, St. Louis, MO) to inactivate IgM ac-
tivity. Equal volumes of undiluted serum and a 0.01 M DTT
assaying for serum reactivity. Where possible, the HLA spec-
ificities of the PRA were determined based on the defined
specificities of the HLA-typed lymphocyte panel.
Peripheral blood mononuclear cells (PBMC) were
obtained from a random human volunteer blood donor, a
standard SLAdd, GalT?/? miniature swine (std), and two
Freshly heparinized whole blood was diluted 1:2 with HBSS
(GIBCO BRL, Gaithersburg, MD), and mononuclear cells
were obtained by density gradient centrifugation using lym-
phocyte separation medium (Organon Teknika, Durham,
NC). Mononuclear cells were washed once with HBSS, and
contaminating red blood cells were lysed with ammonium
chloride potassium buffer (Cambrex BioScience, Walkers-
ville, MD). Cells were washed once more with HBSS and re-
suspended in tissue culture medium. All cell suspensions
were kept at 4°C until used.
Confirmation of GalT-KO Genotype and
Genomic DNA was purified from the PBMC of putative
gestions, and Southern Blot hybridization with probe Bip419
were performed as previously described (18). Detection of sur-
that in place of human serum, 10 ul of either purified baboon
anti-Gal serum or FITC-conjugated IB4 lectin (Sigma) were
and cultured aortic endothelial cells (EC) were suspended at
sodium azide). Volumes of 10 ul of untreated or DTT-treated
human sera were added to 100 ?l of cells at the above concen-
trations and incubated for 30 min at 4°C. Cells were washed
twice in FACS buffer before addition of 10 ?l of a secondary
antibody mix containing a 1:50 dilution of FITC-conjugated
goat Fabanti-human IgM Fcantibodies (Jackson Immuno-
Research, West Grove, PA) and 1:50 dilution of PE-conjugated
goat Fabanti-human IgG Fcantibodies (Jackson Immuno-
Research). Cells were incubated for 30 min in the dark at 4°C
and then washed three times in FACS buffer before two-color
flow cytometric analysis on a FACScan (Becton Dickinson,
Franklin Lake, NJ). Data for FITC and PE fluorescence were
collected for 5?103cells per sample using a 256-channel loga-
rithmic scale. FACScan was calibrated regularly using Cali-
BRITE beads (Becton Dickinson) to ensure normalization of
readings over the course of the study. Flow cytometry data
were analyzed using Cell Quest (Becton Dickinson) and
WinListXP (Verity Software, Topsham, ME) software.
Complement-Dependent Cytotoxicity (CDC)
CDC of untreated or DTT-treated human sera, in two-
fold dilutions from 1:2 to 1:16 in M199 with 2% FBS, against
lymphocytotoxicity technique of the National Institutes of
Health (20). CDC of human sera (untreated or DTT-treated
measured using a modification of the same technique in
which detection of lysed cells was performed using the fluo-
rescent viability stain 7-actinoaminomycin D (7-AAD, Sigma)
Flow cytometry results were expressed as relative mean
fluorescence (RMF), defined as the mean channel fluores-
cence (MCF) of an experimental sample divided by the MCF
of the negative control sample (secondary antibodies only).
CDC results were expressed as the % of lysed cells within the
total cell population as reflected by positive staining with ei-
ther trypan blue or 7-AAD. Calculation of 95% confidence
intervals and p-values were performed using Microsoft Excel
© 2006 Lippincott Williams & Wilkins
Wong et al.
(Microsoft Corp., Redmond, Washington, USA), and corre-
lation analysis was performed with the Spearman non-
parametric method using Prism 4 (GraphPad Software, San
Diego, California, USA).
Absence of Functional GalT Gene Locus and Gal
Expression in GalT-KO Miniature Swine
To confirm the absence of functional GalT genes, PCR
analysis was performed on genomic DNA purified from
PBMC from each putative GalT-KO pig used in this study.
predicted size corresponding to the targeted deletion at the
GalT locus. Southern blot of the digested DNA was also per-
firmed for PBMC from the putative GalT-KO pigs by FACS
using FITC-conjugated IB4, and separately using purified
anti-Gal primary antibodies and FITC- and PE-conjugated
secondary antibodies. Both the IB4 FACS and anti-Gal anti-
body FACS confirmed the lack of Gal antigen expression on
A High Degree of Xenoreactivity Against
Standard Swine Is Significantly Attenuated
Against the GalT-KO Swine
Using standard (std) and GalT-KO PBMC from swine
to detect the degree of human serum antibody binding to the
PBMC targets. Figure 1A shows that the amount of IgM and
IgG antibodies were greatly reduced against GalT-KO targets
compared to standard targets, the reduction in IgM binding
ing 10-fold. Figure 1B shows that this decrease in antibody
binding was also reflected in significantly reduced levels of
cytotoxicity against GalT-KO targets. Because of the nearly
identical genetic background for the PBMC targets, the dif-
ference in levels of xenoreactivity against GalT-KO vs. stan-
dard cells can be attributed entirely to anti-Gal antibodies.
PRA Has No Correlation with Degree of
Reactivity Against Either Standard or GalT-KO
Patient sera were stratified into groups with high PRA
(??80%), moderate PRA (between 20 and 80%), and low
PRA (??20%) according to AHG-augmented CDC against
a panel of human target cells. Statistical analysis was per-
formed to compare the levels of serum IgM binding, IgG
binding, and CDC against GalT-KO target cells across the
three groups. Figure 2A shows that there was no statistically
significant difference in either serum IgM or IgG binding
across the three PRA groups when tested against GalT-KO
targets. Figure 2B shows a similar lack of correlation between
the anti-swine complement-mediated cytotoxic activity and
allosensitization as stratified by PRA.
Anti-Gal Cytotoxicity Is Mediated by IgM and IgG
Antibodies, Whereas Anti-non-Gal Cytotoxicity Is
Mediated Predominantly by IgM Antibodies
To distinguish between cytotoxicity mediated by anti-
bodies of either the IgM or the IgG class, DTT treatment was
performed on a subset of 29 patient serum samples chosen
from the panel of 88 used in previous assays, based on avail-
ability of sera. The selective inactivation of serum IgM func-
tion by DTT reduction was confirmed by flow cytometry,
showing complete elimination of serum IgM binding (Fig.
3A) but maintenance of full IgG binding activity (Fig. 3B) to
GalT-KO PBMC in DTT-treated serum when compared to
the same serum without DTT treatment. The remaining cy-
totoxicity after DTT treatment was therefore representative
showed IgG-mediated cytotoxicity to standard pig PBMC
shows that against GalT-KO PBMC, only two out of the 29
serum samples showed IgG-mediated cytotoxicity greater
than 20%, under the same conditions. Thus, while anti-Gal
activity was potent and resided in predominantly in the IgM
PBMC is significantly attenuated compared to reactivity
against standard (std) pig PBMC. A panel of human sera
were assayed for (A) degree of IgM and IgG binding by
FACS analysis and (B) complement-mediated cytotoxicity.
Human serum reactivity against GalT-KO
Transplantation • Volume 82, Number 3, August 15, 2006
antibody fraction, there was also considerable IgG anti-Gal
activity. In contrast, anti-non-Gal activity was low and re-
sided almost exclusively among IgM antibodies.
Patients with PRA Demonstrate High
Cytotoxicity to Human Target Cells Mediated
Primarily by IgG
To assure that high PRA sera as determined by AHG-
CDC were also highly cytotoxic when measured by the CDC
assay used for testing against the swine targets, PBMC from a
human donor were used as targets in a CDC assay with and
without DTT, testing sera from a subset of nine patients
with high PRA and two patients with low PRA. As shown in
Figure 4, both patients with low PRA (#36 and #87) showed
low cytotoxicity (?10%) against the human PBMC, while
those with high PRA showed uniformly high cytotoxicity
ous findings implicating IgG as the predominant antibody
class mediating alloreactivity, DTT treatment had little effect
on anti-human cytotoxicity.
when elimination of the anti-Gal response is complete. The re-
sults of this study show that PRA has no correlation with the
degree of serum antibody binding or cytotoxicity against
GalT-KO swine. Furthermore, IgG antibodies contribute only
significant amounts of cytotoxic IgG against the GalT-KO
swine, one patient (Pt #50) was highly-allosensitized, with a
PRA of 100%, while the other (Pt #87) was a non-allosensi-
tized, patient with a PRA of 0% and no previous history of
failed transplants or blood transfusions. While we did not
determine the antibody specificities of either serum sample,
it is highly unlikely that Pt #87 has anti-SLA antibodies
cross-reactive with HLA, since Pt #87 showed no evidence of
sensitization to HLA in our panel. Given the extensive HLA
representation of our lymphocyte panel, especially for the
major cross-reactive groups, any possible anti-HLA antibod-
HLA epitope specific to a particular HLA allele, rather
vation does not support the hypothesis that interspecies
MHC cross-reactivities occur predominantly against con-
individual animals from widely disparate species (13, 14).
Previous studies presented evidence that anti-HLA an-
cross-react against SLA (13–16, 21). However, the degrees of
cross-reactivity between HLA and SLA implicated by these
studies differed. Of note, these studies relied on immunoad-
sorption of anti-Gal antibodies from human sera to selec-
tively eliminate the anti-Gal response before testing of sera
adsorption may have confounded the interpretation of these
studies, since there are several reports of difficulty in achiev-
ing complete adsorption of anti-Gal antibodies both in vivo
and in vitro (22–24). Given the polyclonal nature of anti-Gal
antibodies, a subset of these antibodies could be specific for a
Gal epitope that is dependent on the molecule to which it is
conjugated (25). Therefore, adsorption using a particular
type of Gal polysaccharide or Gal-conjugated protein may
was often performed with the same or similar types of mole-
bind and detect the anti-Gal antibodies that escaped the ini-
tial adsorption. The present study shows that IgG anti-Gal
Because the SLA glycoprotein may also contain Gal epitopes,
incomplete adsorption of IgG anti-Gal antibodies may have
been mistaken for IgG anti-HLA cross-reactive against SLA.
cytotoxicity remained in the sera of several patients when
tested against GalT-KO cells. Cytotoxicity mediated by anti-
non-Gal IgG antibodies, however, was found in only two out
of 29 human sera tested. While the potential role of anti-
gree of reactivity against GalT-KO pig PBMC. Degree of
human serum IgM and IgG binding as stratified by PRA
levels, against PBMC a GalT-KO pig is shown as measured
by (A) FACS analysis of relative mean fluorescence (RMF)
and (B) complement-mediated cytotoxicity (CDC).
No correlation is found between PRA and de-
© 2006 Lippincott Williams & Wilkins
Wong et al.
donor IgM antibodies in xenotransplantation remains to be
determined, the presence of pretransplant, donor-specific
IgM antibodies has no correlation with short-term and long-
term renal graft outcome in allotransplantation (26). In fact,
renal allotransplantation in spite of high levels of pretrans-
plant IgM has been performed successfully in the context of
accommodation (27, 28), although accommodation has yet to
rience using GalT-KO donors in a pig-to-baboon model of
tibody response should be avoidable by strategies aimed at
inducing T cell tolerance, as no humoral rejection was ob-
mic xenografts (8). Despite the low incidence of significant
IgG mediated cytotoxicity in this study, it would still be impor-
tant and easy to perform a prospective crossmatch to check for
the presence of preformed anti-pig IgG, since the donor and
In summary, in considering potential candidates for
potential human donors should be at no increased risk of
xenoreactivity against GalT-KO swine compared to patients
who are not allosensitized. Since highly-allosensitized patients
have poor prospects of receiving an appropriate cross-match
negative, human organ, they may be appropriate candidates to
benefit from xenotransplantation using GalT-KO donors in a
well-constructed protocol with a consent form that clearly de-
of human serum IgM and IgG antibodies: DTT selectively
inactivates (A) IgM binding, but not (B) IgG binding, as
DTT treatment is used to distinguish the roles EffectsofDTTtreatmentdistinguishtherelativerolesofIgM
totoxicity of a panel of human sera including both high and
low PRA, when tested against PBMC from a normal volun-
teer human donor.
Lack of an effect of DTT treatment on the cy-
Transplantation • Volume 82, Number 3, August 15, 2006
ACKNOWLEDGMENTS Download full-text
The authors would like to thank Drs. Isabel McMorrow
Maria Doherty for editorial assistance.
1. Cooper DK, Gollackner B, Sachs DH. Will the pig solve the transplan-
tation backlog? Annu Rev Med 2002; 53: 133.
Sachs DH, Sykes M, Robson SC, Cooper DK. Xenotransplantation Adv
Immunol 2001; 79: 129.
and other carbohydrate epitopes that are bound by human anti-pig
antibodies: relevance to discordant xenografting in man. Transpl Im-
munol 1993; 1: 198.
Sandrin MS, Vaughan HA, Dabkowski PL, McKenzie IF. Anti-pig IgM
antibodies in human serum react predominantly with Gal(alpha
1–3)Gal epitopes. Proc Natl Acad Sci U S A 1993; 90: 11391.
Xu Y, Lorf T, Sablinski T, et al. Removal of anti-porcine natural anti-
plantation 1998; 65: 172.
Kozlowski T, Shimizu A, Lambrigts D, et al. Porcine kidney and heart
transplantation in baboons undergoing a tolerance induction regimen
and antibody adsorption. Transplantation 1999; 67: 18.
Kolber-Simonds D, Lai L, Watt SR, et al. Production of alpha-1,3-
galactosyltransferase null pigs by means of nuclear transfer with fibro-
blasts bearing loss of heterozygosity mutations. Proc Natl Acad Sci U S A
2004; 101: 7335.
Yamada K, Yazawa K, Shimizu A, et al. Marked prolongation of
porcine renal xenograft survival in baboons through the use of al-
pha1,3-galactosyltransferase gene-knockout donors and the co-
transplantation of vascularized thymic tissue. Nat Med 2005; 11: 32.
Kuwaki K, Tseng YL, Dor FJ, et al. Heart transplantation in baboons
using alpha1,3-galactosyltransferase gene-knockout pigs as donors:
initial experience. Nat Med 2005; 11: 29.
Gjertson DW. A multi-factor analysis of kidney regraft outcomes. Clin
Transpl 2002: 335.
panel reactive antibodies and their impact on kidney transplantation
outcome. Transpl Int 1992: 5 Suppl 1: S116.
Bartholomew A, Latinne D, Sachs DH, et al. Utility of xenografts: Lack
of correlation between PRA and natural antibodies to swine. Xeno-
transplantation 1997; 4: 34.
Diaz VI, Sanchez MP, Centeno CA, et al. Cross-reactivity between swine
leukocyte antigen and human anti-HLA-specific antibodies in sensitized
Oostingh GJ, Davies HF, Tang KC, et al. Sensitisation to swine leuko-
Am J Transplant 2002; 2: 267.
sera of sensitized patients awaiting renal transplant are also reactive to
swine leukocyte antigens. Transplantation 1998; 66: 1074.
Barreau N, Godfrin Y, Bouhours JF, et al. Interaction of anti-HLA
antibodies with pig xenoantigens. Transplantation 2000; 69: 148.
Cooper DK, Tseng YL, Saidman SL. Alloantibody and xenoantibody
cross-reactivity in transplantation. Transplantation 2004; 77: 1.
Lai L, Kolber-Simonds D, Park KW, et al. Production of alpha-1,3-
2002; 295: 1089.
Dor FJ, Tseng YL, Cheng J, et al. alpha1,3-Galactosyltransferase gene-
Patel R, Terasaki PI. Significance of the positive crossmatch test in
kidney transplantation. N Engl J Med 1969; 280: 735.
Taylor CJ, Tang KG, Smith SI, et al. HLA-specific antibodies in highly
cytes. Transplantation 1998; 65: 1634.
Teranishi K, Alwayn IP, Buhler L, et al. Depletion of anti-Gal antibod-
ies by the intravenous infusion of Gal type 2 and 6 glycoconjugates in
baboons. Xenotransplantation 2003; 10: 357.
Teranishi K, Gollackner B, Buhler L, et al. Depletion of anti-gal anti-
bodies in baboons by intravenous therapy with bovine serum albumin
conjugated to gal oligosaccharides. Transplantation 2002; 73: 129.
Kuwaki K, Knosalla C, Moran K, et al. Reduction of anti-Galalpha1,3Gal
antibodies by infusion of types 2 and 6 gal trisaccharides conjugated to
Parker W, Stitzenberg KB, Yu PB, et al. Biophysical characteristics of
anti-Gal(alpha)1–3Gal IgM binding to cell surfaces: implications for
xenotransplantation. Transplantation 2001; 71: 440.
Bryan CF, Martinez J, Muruve N, et al. IgM antibodies identified by a
DR phenotype. Clin Transplant 2001; 15 Suppl 6: 28–35.
Tanabe K, Takahashi K, Sonda K, et al. Long-term results of ABO-
Transplantation 1998; 65: 224.
Shishido S, Asanuma H, Tajima E, et al. ABO-incompatible living-donor
© 2006 Lippincott Williams & Wilkins
Wong et al.