Download PDF

ARTICLES
Human and porcine early kidney precursors as a
new source for transplantation
© 2003 Nature Publishing Group http://www.nature.com/naturemedicine
BENJAMIN DEKEL1, TATYANA BURAKOVA1, FABIAN D. ARDITTI1, SHLOMIT REICH-ZELIGER1,
OREN MILSTEIN1, SARIT AVIEL-RONEN3, GIDEON RECHAVI3,4, NIR FRIEDMAN5,
NAFTALI KAMINSKI3, JUSTEN H. PASSWELL2 & YAIR REISNER1
1
Department of Immunology, Weizmann Institute of Science, Rehovot, Israel
2
Department of Pediatrics, Sheba Medical Center, Tel Hashomer, Israel
3
Functional Genomics Unit, Molecular Hemato-oncology and Respiratory Medicine, Sheba Medical Center,
Tel Hashomer, Israel
4
Department of Pediatric Hemato-Oncology, Sheba Medical Center, Tel Hashomer, Israel
5
School of Computer Science and Engineering, Hebrew University, Jerusalem, Israel
Correspondence should be addressed to Y.R.; email: [email protected]
Published online 23 December 2002; doi:10.1038/nm812
Kidney transplantation has been one of the major medical advances of the past 30 years.
However, tissue availability remains a major obstacle. This can potentially be overcome by the
use of undifferentiated or partially developed kidney precursor cells derived from early embryos
and fetal tissue. Here, transplantation in mice reveals the earliest gestational time point at which
kidney precursor cells, of both human and pig origin, differentiate into functional nephrons and
not into other, non-renal professional cell types. Moreover, successful organogenesis is achieved
when using the early kidney precursors, but not later-gestation kidneys. The formed, miniature
kidneys are functional as evidenced by the dilute urine they produce. In addition, decreased immunogenicity of the transplants of early human and pig kidney precursors compared with adult
kidney transplants is demonstrated in vivo. Our data pinpoint a window of human and pig kidney organogenesis that may be optimal for transplantation in humans.
The number of human kidney transplants has increased rapidly in
recent years, but the demand greatly exceeds organ availability.
Normal development of the mammalian kidney involves the invasion of a specialized region of intermediate mesoderm by an epithelial source (ureteric bud), which grows and branches to form a
collecting duct system, and induces disorganized metanephric
mesenchymal stem cells to group and differentiate into nephrons1.
In the developing human kidney, fresh stem cells are induced into
the nephrogenic pathway to form nephrons until 34 weeks of gestation. Thus, transplants of precursors for the adult kidney present
in early embryos and fetal tissue may be a potential source for regenerating kidney cells, and a promising solution for the current
shortage of organs for kidney transplantation. Despite their potential clinical utility for replacement therapy suggested by murine
studies2–5 performed by Hammerman et al., the fate of undifferentiated human kidney precursors after transplantation is unknown.
However, the difficulty in obtaining sufficient numbers of human
embryos, as well as the ethical problems involved with the use of
human embryonic tissue, can be circumvented by the use of
porcine embryonic donor tissue6,7. We therefore assessed growth
potential, vascularization, function and immunogenicity of kidney
precursors, derived from both human and pig embryos, after transplantation into both immunodeficient and immunocompetent
mouse hosts.
Gestational age determines growth and differentiation
An initial experiment was carried out to determine the viability
NATURE MEDICINE • VOLUME 9 • NUMBER 1 • JANUARY 2003
of transplants of adult human kidney tissue in immunodeficient
mice. At 8 weeks after transplantation all adult transplants (5/5)
were found to be sclerotic and non-viable. To assess the influence
of embryonic stage on growth and potential to differentiate, we
transplanted kidney progenitors originating from 7 to 14 weeks
of human gestation into immunodeficient mice (Table 1).
Overall, more than 80% of the human kidney grafts from all
donor ages survived, and all recovered grafts had increased in
size, with no evidence of neoplasia in any of the recovered grafts.
Nevertheless, results were distinctly different when transplants of
embryonic kidney precursors obtained from 7 and 8 weeks of
human gestation were compared with later-gestation kidneys.
Thus, at 8 weeks after transplantation, the early kidney precursors
(n = 8) had undergone remarkable growth (transplant size ratio
was 20.1 ± 2.7). Complete nephrogenesis was observed in the
grafts of early kidney precursors, which originally contain mainly
metanephric mesenchymal stem cells and ureteric buds, but no
glomeruli (5.5 ± 0.8 glomeruli and 19.3 ± 2.7 tubuli per ×40 highpower field (HPF); Figs. 1a and b). Beyond this gestational time
point, transplantation of developing whole fetal kidneys resulted
in central graft necrosis and viability was reduced. We therefore
grafted small pieces of human fetal kidney tissue into immunodeficient hosts, as previously described8,9. Under identical conditions, transplants originating from sections of 10- and 14-week
gestation kidneys (n = 14) had significantly lower transplant size
ratios (14.8 ± 2.2 and 12.3 ± 1.8, respectively, P < 0.01) as well as
glomerular and tubular counts (4.2 ± 0.8 and 15.3 ± 2.7; 3.5 ± 0.8
53
© 2003 Nature Publishing Group http://www.nature.com/naturemedicine
ARTICLES
a
b
e
f
Fig. 1 Growth and differentiation of early human and pig kidney precursors after transplantation. Shown are a macroscopic view (a) and histology
(b) (H&E; original magnification ×10) of a transplant originating from an 8week-old human embryo, 8 weeks after transplantation. Note massive
growth and the formed shape of a kidney (arrow) and appearance of layers
of glomeruli and tubuli. c and d, Macroscopic view and histology (H&E;
original magnification ×10), respectively, of a transplant originating from a
4-week-old pig embryo, 8 weeks after transplantation. Note massive
growth (arrow) and external vascular beds and numerous glomeruli and
tubuli. Transplanted early embryonic kidney cells differentiate into other
cell fates, following transplantation of E20–21 (e–g) and E24–25 (h–j) pig
kidney precursors. e, Micrograph (H&E) in low magnification (original ×4)
showing blood vessels (arrowheads), cartilage (large arrow) and bone
(small arrows) and higher magnifications (original ×40) of bone (f) and car-
c
d
g
h
i
j
tilage (g). h, Micrograph (H&E; original magnification ×10) showing myofibroblasts (arrowheads) and cartilage (large arrow), and higher magnifications (original ×40) of myofibroblasts (i) and glandular-like structure (j).
and 11.2 ± 2.2 per HPF, respectively, P < 0.05). Thereby, in contrast to transplantation of more mature human fetal renal fragments, grafting of the undifferentiated kidney precursors led to
better growth and nephrogenesis.
We used the same approach described above to assess the
growth and potential to differentiate of porcine kidney precursor
cells (Table 1). Here, transplants of 6- and 8-week pig gestation
kidneys exhibited central fibrosis and necrosis and graft deterioration, whereas sectioned grafts had a transplant size ratio of 10.5 ±
2.2 and 8.2 ± 1.2, respectively, at 8 weeks after transplantation. For
characterization of early pig kidney precursors, transplants originating from 20–21 (E20–21), 24–25 (E24–25) and 27–28 (E27–28)
embryonic days (combined data of 30 transplants) were analyzed. The transplants of kidney
Table 1 Transplantation of human and pig kidney precursors in immunodeficient
precursors originating from the E27–28 pig
mice
donors all exhibited significant growth with a
Gestation No. of
Method of
Grafta
Graft differentiationb
transplant size ratio at 8 weeks post-transplant of
age
transplants transplantation
growth
Renal
Non-renal
Necrosis
28.3 ± 4.1 and full differentiation into mature
Human
glomeruli and tubuli (7.0 ± 1.0 glomeruli and
14w
3
whole
3/3
none
none
3/3
35.5 ± 5.1 tubuli per HPF; Figs. 1c and d). In con14w
8
fragments
7/8
7/7
none
none
trast, six of nine transplants from the E20–21 pig
10w
2
whole
2/2
none
none
2/2
donors failed to develop or had evolved into
10w
6
fragments
6/6
6/6
none
none
growths containing few glomeruli and tubuli, but
8w
5
whole
5/5
5/5
none
none
containing other differentiated derivatives, such
7w
3
whole
3/3
3/3
none
none
as blood vessels, cartilage and bone (Figs. 1e–g).
Pig
Furthermore, pig kidney progenitors from
8w
7
whole
5/7
none
none
5/5
E24–25 donors were inferior to E27–28 trans8w
6
fragments
6/6
6/6
none
none
plants for nephrogenesis, as non-renal cell types
6w
5
whole
4/5
none
none
4/4
and disorganized cell clusters were found in three
6w
6
fragments
6/6
6/6
none
none
of nine transplants (Figs. 1h–j). Our findings
E27-E28
12
whole
12 / 12
12 / 12
none
none
complement recent in vitro data10, which both inE24-E25
9
whole
8/9
5/8
3/8
none
E20-E21
9
whole
6/9
3/6
3/6
none
dicate that early in gestation the embryonic kida
b
ney contains progenitor cells with the ability to
Transplant growth and differentiation were assessed at 8 weeks after transplantation. Differentiation was
categorized to renal (only nephrons), non-renal (differentiated derivatives other than renal) and necrosis (in
generate many cell types (that is, multipotent
addition to nephrons, appearance of necrotic areas mostly in center of transplant).
progenitors or embryonic renal stem cells).
54
NATURE MEDICINE • VOLUME 9 • NUMBER 1 • JANUARY 2003
© 2003 Nature Publishing Group http://www.nature.com/naturemedicine
ARTICLES
a
b
e
f
i
j
Vascularization of kidney precursors by host vessels
The ability of cellular or tissue transplants to grow in a foreign
microenvironment depends first on the ability to sustain angiogenesis11. In addition, the immunological barrier to xenotransplantation is conditioned to a large extent by the manner in
which the transplant derives its blood supply12. To determine the
ability of recipient mice to support angiogenesis of the early avascular human and pig kidney precursors by ingrowth of recipient
vessels, we identified mouse PECAM (CD31), a marker of sprouting endothelial cells, on the developing transplants with immunostaining. Counts of immunoreactive vessels reflecting the
combined total number of capillary and larger vessels of host origin were performed per HPF (ref. 13). Control sections of vascularized human and pig fetal kidneys showed no staining using
antibody against mouse CD31 (Figs. 2a and b). At 4 weeks after
transplantation, we found 23.5 ± 4.0 and 21.3 ± 3.2 vessels of
host origin per HPF supplying the developing human and pig
transplants, respectively. Among these, all external vessels
stained positive for mouse CD31 (Figs. 2c and d). In addition, we
detected medium and small size capillaries of host origin in both
parenchyme (Figs. 2e and f) and glomeruli (Figs. 2g and h) of the
developing human and pig transplants. In transplants originating from mature, vascularized human and pig fetal kidneys, we
found a significantly reduced mouse CD31-positive vessel count
(10.2 ± 1.8 and 11.5 ± 2.2, respectively, P < 0.001) composed of
c
g
d
h
Fig. 2 Vascularization of developing kidney transplants by recipient vessels.
Sections were immunostained with antibody against mouse CD31 (PECAM).
a and b, Lack of staining in vascularized human and pig fetal kidneys, respectively (original magnification ×20). Immunostaining of early embryonic
human (8-week origin) (c, e, g) and pig (4-week origin) (d, f, h) kidney precursors 4 weeks after transplantation. Positive staining (arrowheads) in larger
vessels (c, d), medium and small-size capillaries (e, f), and developing
glomeruli (g, h). Lack of staining in glomeruli and small-size capillaries of 4week-old transplants originating from mature 16-week human (i) and 8-week
pig (j) fetal kidneys. All magnifications are original ×40.
mainly external larger vessels, but not endothelial cells in
glomeruli and small capillaries (Figs. 2i and j). Thus, recipient
mice have a significantly larger contribution to vasculogenesis of
the transplants of early human and pig kidney precursors including the formation of the microcirculation.
Early human and pig kidney precursors produce dilute urine
To determine kidney functionality, we measured levels of urea nitrogen and creatinine in cyst fluid collected from cysts arising
from transplants of early kidney precurors. Large cysts were
mostly found in transplants established in the abdomen and
therefore were not limited by the renal subcapsular space (Figs. 3a
and b). We evaluated kidney transplants that originated from 8week human (n = 2) and 4-week pig (n = 4) embryos, 6–8 weeks
after transplantation. As the transplants cannot use the host’s excretory system for urine drainage, fluid was derived by insertion
of a permanent microcatheter into the developing renal grafts.
a
b
Fig. 3 Transplants of early embryonic human and pig kidney precursors produce urine. Macroscopic view of human, 8-week origin (a) and
pig, 4-week origin (b), intra-abdominal grafts containing large cysts (indicated by arrows) 8 weeks after transplantation. Analysis of cyst fluid revealed dilute urine.
NATURE MEDICINE • VOLUME 9 • NUMBER 1 • JANUARY 2003
55
ARTICLES
a
b
© 2003 Nature Publishing Group http://www.nature.com/naturemedicine
c
d
e
f
g
h
Fig. 4 Differential effect of human PBMC on developing kidney transplants.
a, Growth curves of transplanted human kidney precursors obtained at different stages of kidney organogenesis in the presence () or absence () of alloreactive human PBMC. In transplants originating from 14- or 10-week-old
human fetuses, 8 weeks after transplantation, P < 0.01 and P < 0.05 compared with controls, respectively. b, Immunostaining with antibodies against
human CD3 (original magnification ×40) demonstrating destruction of
tubule (right) and glomerulus (left) by human T cells in a transplant originating from a 14-week-old fetus. c, Lack of staining for infiltrating human CD3
and preserved glomeruli and tubuli (original magnification ×40) in a transplant originating from an 8-week-old embryo. d, Growth curves of transplanted pig kidney precursors obtained at different stages of kidney
organogenesis in the presence () or absence () of xenoreactive human
PBMC. In transplants originating from 8- or 6-week-old pig fetuses, 8 weeks
after transplantation, P < 0.01 and P < 0.05 compared with controls, respectively. e, Immunostaining with antibodies against human CD3 (original magnification ×40) shows destruction of transplant tissue, originating from a
8-week-old pig fetus, by invading human T cells. f, Transplant originating
from a 4-week-old embryo demonstrates preserved glomeruli and tubuli with
no CD3+ infiltrate at a higher magnification (original, ×40). g, The growth
curve (left) of early human kidney precursors (8-week embryonic origin) re56
i
ceiving two independent infusions of alloreactive human PBMC, at the time
of transplantation and 6 weeks post-transplant () is similar to that of control
early human precursors transplanted in absence of alloreactive human PBMC
(). The growth curve (right) of transplants originating from 14-week-old
human fetuses demonstrates halted growth () when the latter are transplanted in conjunction with the second donor human PBMC, and compared
with those not subjected to PBMC () (P < 0.05, 8 weeks after transplantation). h, The growth curve (left) of early pig kidney precursors (4-week embryonic origin) receiving 2 independent infusions of xenoreactive human
PBMC, at the time of transplantation and 4 weeks post-transplant (), is similar to that of control early pig precursors transplanted in absence of xenoreactive human PBMC (). The growth curve (right) of transplants originating
from 8-week-old pig fetuses demonstrates arrested growth () when the latter are transplanted in conjunction with the second donor human PBMC and
compared with those not subjected to PBMC () (P < 0.05, 8 weeks after
transplantation). i, Analysis of co-stimulatory mRNA expression in normal
human developing kidneys (pre-transplant), in transplanted developing
human kidneys immediately after transplantation but before the administration of allogeneic human PBMC (post-transplant), and at 2, 4 and 6 weeks
after mice were reconstituted with human immune cells. Transplants originated from 8-week, 14-week and 22-week human gestation kidneys.
NATURE MEDICINE • VOLUME 9 • NUMBER 1 • JANUARY 2003
© 2003 Nature Publishing Group http://www.nature.com/naturemedicine
ARTICLES
Average levels (mg/dl) of urea nitrogen and creatinine were
higher in cyst fluid compared with those found in the sera of
transplanted mice (518 ± 169 versus 45 ± 8 and 7.2 ± 1.9 versus
0.46 ± 0.048, respectively; P < 0.001), indicating that the human
and pig transplants had produced urine. These results are in line
with the demonstration that murine precursor kidneys can develop into functional nephrons2–5. Levels of urea nitrogen and
creatinine in the cyst fluid were significantly lower compared
with native bladder urine (518 ± 169 versus 4279 ± 402 and 7.2 ±
1.9 versus 54 ± 6, respectively; P < 0.001). The dilute urine in the
cyst fluid is compatible with a reduced capacity of the fetal kidney to concentrate urine.
Early precursors are less susceptible to human leukocytes
We determined whether transplants of early, undifferentiated
kidney progenitors have an immunological advantage over latergestation and adult kidneys. Four weeks after transplantation of
adult human kidney fragments in immunodeficient recipients
together with 100 × 106 alloreactive human peripheral blood
mononuclear cells (PBMC), massive infiltration, tissue destruction and rejection were observed, as previously described14,15. At
4 weeks after transplantation, grafts originating from 14-week
gestation kidneys had an average of 39.8 ± 7.8 donor human lymphocytes per HPF. Despite the presence of T cells in these grafts,
early rejection similar to the adult transplants did not occur, and
growth of all transplants took place during the first month8,9 (Fig.
4a). Nevertheless, analysis of the grafts at later time points (6–8
a
Fig. 5 Global gene expression patterns of immunerelated genes in normal adult and fetal kidneys.
a, Hierarchical clustering dendogram of the experimental groups on the basis of the similarity of their expression profiles demonstrates that adult and fetal
kidneys cluster separately22. b, Gene expression patterns in the 231 immune-related genes showed that
122 of them had a TNoM = 0 or 1 (ref. 37). Gene expression values were divided by a geometric mean of
all samples, log transformed and then plotted using
the Plottopgene program37. Yellow represents maximal expression and purple, minimal. Note that most of
the immune-related genes were lower in fetal kidney
compared with adult kidney. c, Gene expression of 68
genes that had a TNoM = 0 (P < 0.05). Plots are the
mean expression values of all genes in the group. To
eliminate outlier effect, genes were normalized to a
range of [0,1], meaning that the maximum value for
every gene was set to be 1, the minimum value to be
0, and the rest of the values were linearly fitted to this
range. Note again that most statistically significant
genes (57/68) were lower in fetal kidney compared
with adult kidney. d, Examples of significantly upregulated immune-related genes in the adult kidneys were
classified according to functional categories.
NATURE MEDICINE • VOLUME 9 • NUMBER 1 • JANUARY 2003
b
weeks post-transplant) revealed the harmful effects of the infiltrating cells as graft deterioration became apparent and transplant growth was significantly halted compared with transplants
from animals that were not subjected to infusion of human
leukocytes (Figs. 4a and b). Similar results were obtained for grafts
obtained from 10-week human gestation kidneys (Fig. 4a). In
contrast, upon infusion of 100 × 106 human cells into the host’s
peritoneum, kidney transplants originating from 7- and 8-week
embryos exhibited no donor human T cells and grew similarly to
transplants of control mice with no apparent signs of destruction
or rejection (Figs. 4a and c). Moreover, when the experimental
protocol was altered so that two inocula of 100 × 106 alloreactive
human PBMC from different donors were infused 6 weeks apart,
immune cells did not reject the 8-week human gestation kidney
precursors, but 14-week human gestation kidneys transplanted in
conjunction with PBMC of the second donor were rejected (Fig.
4g). Thus, the immunogenicity of the differentiated tissue, which
developed for 6 weeks following implantation of 8-week embryonic kidney precursors, was still greatly reduced compared with
14-week gestation kidneys.
We also assessed subcapsular transplants of pig kidney tissue
after intraperitoneal infusion of xenogeneic human leukocytes
(100 × 106 cells). Within 3 weeks, five of six adult pig kidney transplants were infiltrated and histologic damage and destruction
were apparent (data not shown). In kidney transplants obtained
from 8-week pig fetuses, human T cells were readily detectable in
all six grafts with an average of 40.5 ± 6.75 lymphocytes per HPF at
c
d
57
© 2003 Nature Publishing Group http://www.nature.com/naturemedicine
ARTICLES
4 weeks after transplantation. Analysis at later time points indicated that five of six transplants had signs of rejection (Figs. 4d
and e). These results and similar findings from other experiments
with 6-week pig fetal kidney donors (Fig. 4e) showed that at these
stages of organogenesis the xenogeneic human PBMC induced rejection of the developing pig kidney transplants. In contrast, rejection by the human immune cells was prevented when early pig
kidney precursors, originating from E21 and E28 pig donors, were
transplanted (Figs. 4d and f). Furthermore, transplants of E28
donors subjected to a second infusion of 100 × 106 xenogeneic
human PBMC, 4 weeks after transplantation, were not rejected,
whereas kidney precursors of an 8-week pig fetus transplanted directly with these cells were eventually rejected (Fig. 4h).
Co-stimulatory molecules on donor antigen-presenting cells
(APC) are crucial in the alloimmune response16–19. We therefore
analyzed the mRNA expression of B7-1, B7-2, CD40 and CD40L
before and after transplantation of developing human kidneys in
the absence and presence of allogeneic human PBMC. Our results
demonstrate differential expression of co-stimulatory molecules
in both normal and transplanted developing human kidneys of
different gestation age, with a distinct deficiency (especially CD40
and B7-1) in the human kidney precursors originating from the
8-week embryo and a progressive increase in expression in latergestation kidneys (Fig. 4h).
Advantage of early precursors in immunocompetent mice
We further determined the immunogenicity of the kidney precursors by transplanting adult pig kidney tissue and early pig kidney
precursors (E27–28) into immunocompetent BALB/c mice.
Evaluation of adult (n = 10) and early kidney (n = 10) precursors
after 2 weeks showed rejection of both tissues. Following shortterm administration of CTLA4–Ig, an immunoglobulin fusion
protein that directly affects T-cell recognition of B7+ on APC (ref.
20), at 2–4 weeks post-transplant, all adult grafts (8/8) had a disturbed morphology, necrotic tissue and a high degree of lymphocyte infiltration. In contrast, at the same time point, infusion of
CTLA4–Ig resulted in growth and differentiation of 6 of 10 of the
early kidney precursor transplants, which was not seen in the untreated animals, indicating the immune advantage of the developing precursor transplants over developed adult kidney transplants
in fully immunocompetent hosts.
Multiple immune genes are decreased developing kidneys
To investigate the inherent immunogenic properties of the developing kidney, that might account for its decreased immunogenicity, we determined global gene expression by microarray
analysis during human kidney development and in mature adult
human kidneys. We further analyzed 231 genes that have a direct
role in the immune response (the complete list of genes can
be found at http://www.weizmann.ac.il/immunology/reisner/
immunogenicity.xls). These included genes encoding HLA molecules, cytokines, chemokines, chemokine receptors, apoptosisrelated molecules, adhesion molecules, metalloproteinases, molecules of innate immunity and other immunomodulators.
Hierarchical clustering21 of all genes on the basis of similarity in
gene expression among the experimental groups revealed two
main clusters, separating the adult from fetal tissues. Moreover,
the immune-related genes were grouped according to gestational
age with a cluster of genes within the least-mature fetal kidneys
and a cluster of genes within the adult kidneys on opposing sides
(Fig. 5a). The patterns of “immune” gene expression are presented using Plottopgene program22 (Fig. 5b). We found that 68
58
genes were significantly changed between adult and fetal tissues
(P < 0.05, total number of misclassifications (TNoM) = 0).
Expression profiles of these genes demonstrated those increased
in the adult tissues (n = 57 genes; Fig. 5c, top) and those decreased
(n = 11 genes; Fig. 5c, bottom). Examples of the most significantly
changed immune-related genes include those encoding molecules participating in both the acquired and the innate immune
response (Fig. 5d).
Discussion
Our results show that when human and pig kidney precursors are
obtained from 7- to 8-week human or 3.5- to 4-week pig gestation
and transplanted into immunodeficient mice, they survive, grow
and undergo complete nephrogenesis, forming a functional organ
able to produce urine. Embryonic renal cells of earlier origin fail to
mature into the desired professional cell fate and form nonrenal differentiated derivatives and disorganized cell clusters.
Furthermore, the successful organogenesis of the kidney precursors is achieved when early progenitors, but not later-gestation
kidneys (whole or fragmented), are completely isolated and transplanted into mice. At these stages both human and pig kidney
precursors contain renal mesenchymal stem cells and ureteric bud
branches, but no glomeruli, emphasizing their remarkable potential to differentiate after transplantation. Their growth and development are facilitated by the vascular contribution of host origin.
It has been known for over four decades that embryonic tissues
are less immunogenic compared with their adult counterparts23.
Thus, our definition of the earliest time point in human or pig
renal gestation at which normal differentiation and subsequent
kidney function are possible may also pinpoint the ideal time for
harvesting the tissue least prone to immune rejection.
Accordingly, graft acceptance may reflect the progressive development of a complex array of cell surface molecules and soluble
factors that determine immune recognition in the fetal organ.
We established by microarray analysis that the development of
immunological maturity in the human kidney is a rather late
event in gestation. Altogether, the developing kidneys (representing gestational time points through which the developing transplants progress) are restricted in multiple factors that determine
immune recognition. Thus, 13 of the 57 genes that were significantly upregulated in adult versus fetal kidney tissues belong to
the HLA class I and class II systems. In addition, molecules that
mediate trafficking of leukocytes into the graft, such as the
chemokines RANTES and MCP-1 (ref. 24), the adhesion molecule
E-selectin25, pro-inflammatory cytokines such as osteopontin26–28
and complement genes known to be associated with innate immunity29, may all be responsible for the reduced immunogenicity
of the developing kidneys.
The immunogenicity of the kidney precursors was also evaluated in two different mouse models. The first, a “humanized” system, renders host mice immunodeficient and uses isolated human
mononuclear cells transferred into mice to test imunogenicity.
The level of immunity afforded in this model following infusion
of human PBMC has been well documented30,31. Here, both first
and second rounds of transplanted human T cells, obtained from
two separate donors, were not capable of mounting cellular rejection of developing human and pig kidneys when the latter were
isolated during early kidney organogenesis. Whereas the global
gene analysis indicates that the acceptance of these early grafts is
likely associated with multiple immune pathways, the reduction
in CD40 and B7-1 expression implies a possible absence or immaturity of donor hematopoietic APCs. In addition, the reduced imNATURE MEDICINE • VOLUME 9 • NUMBER 1 • JANUARY 2003
© 2003 Nature Publishing Group http://www.nature.com/naturemedicine
ARTICLES
munogenicity could also be associated with the observed depletion of donor endothelial cells, shown recently to perform as APC
and/or as targets for T-cell–mediated cytotoxicity in direct allorecognition32. Allogeneic tissue engineered human skin, devoid
of donor endothelial cells and limited in its antigen-presenting capabilities, has been shown to perform similarly to the early kidney
precursors in a humanized model of skin rejection33.
In the second model, adult pig tissue and early pig kidney precursors were transplanted in immunocompetent hosts. Rejection
in such hosts can be triggered by donor APC transferred in the implant or alternatively by cross priming against host APC loaded
with donor antigens in a fashion similar to the normal process for
the presentation of bacterial or viral antigens18,34. Because the early
kidney precursors possibly lack mature APC, in addition to a relative reduction in homing receptors and specific cytokines or
chemokines, blockade of cross priming might be sufficient to alleviate the observed rejection of these implants. Indeed, the relatively reduced immunogenicity of the early kidney precursors was
revealed only when the indirect presentation was neutralized, at
least in part, by a short-term, co-stimulatory blockade with
CTLA4–Ig, a protocol that failed to prevent rejection of the transplants of developed adult kidneys. This set of experiments has direct implications for designing immunosuppressive regimens if
transplantation of both allogeneic and xenogeneic early kidney
precursors in human subjects is to be considered.
Finally, similar to previous transplantation studies of renal
precursors obtained from murine embryos2–5,35, we were not able
to demonstrate a connection of donor human and pig nephrons
to the collecting system of hosts. However, our transplants did
integrate into the host’s microenvironment and use its blood
vessels, and urine was produced separately of the native kidneys.
Further experimentation needs to be developed to produce adequate urinary anastomosis and diversion of blood supply to the
kidney grafts sufficient to correct biochemical aberrations in a
uremic individual. Increasing the number of transplants and/or
administering specific human growth factors might support
functional replacement.
Our data pinpoint a window of human and pig embryogenesis
that may be optimal for transplantation in humans. Considering
the limited availability of human fetal tissue, pig kidney precursors could afford an unlimited source for renal transplantation.
When tested in large animal models or in patients, our data predict that these early stage human and pig kidney progenitors
should require less immunosuppression compared with that currently used in renal transplantation.
Methods
Animals. We used 3-month-old BALB/c (from Harlan Olac, Shaw’s Farm,
Blackthorn, Bicester, Oxon., UK) as hosts for the transplantation studies. For
transplantation in immunodeficient mice, lethally irradiated BALB/c recipients
were radioprotected with bone marrow from NOD/SCID (severe combined
immunodeficiency) mice and subsequently implanted with human tissues as
described30,31. NOD/SCID mice were obtained from the Weizmann Institute
Animal Breeding Center, Rehovot, Israel. Animal experiments were carried
out according to the National Institutes of Health Guide for Care and Use of
Experimental Animals and approved by the Weizmann Institute of Science
Animal Care Committee.
Tissues. Embryonic human kidney precursors were obtained following curettage. Adult kidney specimens were obtained from the normal kidneys removed for stage I clear cell carcinoma. Pig kidney precursors and adult pig
kidney tissue were obtained with the assistance of the Lahav Institute for
Animal Research, Kibbutz Lahav. All specimens for gene expression analysis
were snap frozen in liquid nitrogen. Tissues for transplantation were kept in
NATURE MEDICINE • VOLUME 9 • NUMBER 1 • JANUARY 2003
sterile conditions at 4 °C (for approximately 2 h) in either RPMI 1640 or
Dulbecco’s modified Eagle’s medium containing 10% fetal calf serum
(Biological Industries, Beit HaEmek, Israel).
Transplantion procedure. Transplantation of human and pig kidney precursors as well as adult kidney tissue under the kidney capsule was done as described14. Early kidney precursors (human ≤ 8-week gestation; pig ≤ 4-week
gestation) were transplanted in whole. Later-gestation kidneys were transplanted in whole or in fragments of 1–2 mm in diameter. In some experiments an intra-abdominal incision was made and kidney precursors were
implanted and sutured (5-0 suture) onto the testicular fat of mice, in conjunction with a left nephrectomy.
Isolation and transfer of human PBMC. Generation, transfer into mice and
analysis for engraftment of human PBMC were carried out as described30.
Transplant growth and differentiation. The animals receiving implants
were sacrificed at 4–10 weeks after transplantation. Kidneys and their capsules were then removed and fixed in 10% paraffin. Transplants were sectioned and mounted on slides coated with poly-L-lysine and sections were
stained with hematoxylin and eosin (H&E). We measured the long (L) and
short (W) axes of the grafts and calculated the post- and pre-transplant size
ratio by multiplying L × W, both for the original (pretransplant) and the graft
at the time of sacrifice. Assessment of graft development was performed by
counting the number of mature glomeruli and tubuli in 10 consecutive fields
(×40 high-power field (HPF))/transplant in 3 transplants/group. For the determination of human T cell infiltration, we immunostained sections with rabbit
antibody against human CD3 (Zymed, San Francisco, California; pan T-cell)
as described15 and counted the number of human CD3+ cells in 10 consecutive fields (×100 HPF)/transplant in 3 transplants/group for the determination
of human T-cell infiltration.
Analysis of host vessel vascularization. Paraffin sections (5 µm) were immunostained with a rat antibody against mouse PECAM-1 (Pharmingen, San
Diego, California) using a histostain plus kit (Zymed, San Francisco,
California), according to the manufacturer’s instructions. Vessel counts were
performed in similar regions within kidney grafts per high-power field (5 consecutive fields/transplant in 5 transplants/group).
Urine collection and analysis. Urine was collected following exposure of developing human and pig transplants in anesthetized mice (midline or left
flank incision) and insertion of a plastic catheter to an identifiable area of fluid
concentration. At the site of insertion, graft walls were sutured around the
catheter using a 5-0 nylon suture, which then traversed the skin and brought
into small plastic PCR tubes sutured onto the skin of the mice. Drained fluid
was analyzed for urea nitrogen and creatinine concentrations.
RT–PCR. Grafts were dissected carefully from the subcapsular site and RNA
was isolated as described15. Total RNA (1 µg) was reversed transcribed into
cDNA and amplified using Access RT-PCR kit (Promega Corp., Madison,
Wisconsin) and specific primers for human B7-1, B7-2, CD40, CD40L and a
housekeeping gene, β-actin29. We used the following primers: 5′-GACCAAGGAAGTGAAGTGGC-3′ (sense) and 5′-AGGAGAGGTGAGGCTCTGGAAAAC-3′
(antisense) for human B7-1 (410 bp); β-actin (764 bp); 5′-CACTATGGGACTGAGTAACATTC-3′ (sense) and 5′-GCACTGACAGTTCAGAATTCATC-3′ (antisense) for human B7-2 (383 bp); 5′-CTCTGCAGTGCGTCCTCTGGGG-3′
(sense) and 5′-GATGGTATCAGAAACCCCTGTAGC-3′ (antisense) for human
CD40 (410 bp); 5′-TATCACCCAGATGATTGGGTCAGC-3′ (sense) and 5′CCAGGGTTACCAAGTTGTTGCTCA-3′ (antisense) for human CD40L (349
bp); 5′-ATGAAGGTCTCCGCGGCAGCCC-3′ (sense) and 5′-CTAGCTCATCTCCAAAGAGTTG-3′ (antisense) for human HLA-DR (215 bp); and 5′-ACCATCAAGCTCTGCGTGACTG-3′ (sense) and 5′-GCAGGTCAGTTCAGTTCCAG
GTC-3′ (antisense) for β-actin (310 bp). After amplification, the sample was
separated on an agarose gel containing ethidium bromide and bands visualized and photographed using an ultraviolet transilluminator.
Co-stimulatory molecule blockade. Following transplantation, BALB/c mice
were treated for a short time with 250 µg of CTLA4–Ig (a fusion protein made
from the extracellular portion of the mouse gene encoding CTLA4 and the
constant region of human IgG1, kindly provided by Steffen Jung, Hadassa
59
ARTICLES
Medical School, Jerusalem, Israel), given i.p. six times over a 2-week period.
Control mice were injected with phosphate buffered saline or control immunoglobulin.
© 2003 Nature Publishing Group http://www.nature.com/naturemedicine
Statistical analysis. Comparisons between groups were evaluated by the
Student’s t-test. Data were expressed as mean ± s.e.m., and were considered
statistically significant if P values were 0.05 or less.
Microarray analysis. Labeled cRNA preparation and hybridization to a
Genechip Human Genome HU95A array were performed as recommended by the manufacturer of the microarrays. Analysis of Genechip
data was performed as described22,36. For cluster analysis we used the
Cluster, Gene Cluster, Treeview programs21 and Scoregene Package
(http://FGUSheba.cs.huji.ac.il/). Fold ratios were calculated for each sample against the geometric mean of all the samples. A detailed description
of the scoring methods and our approach to analysis of microarray data
have been published37.
Acknowledgments
Supported in part by a grant from Mrs. E. Drake and the Gabriella Rich Center
for Transplantation Biology Research and the Edward H. Kass Award from the
American Physicians Fellowship (B.D.). Y.R. holds the Henry H. Drake Professorial
Chair in Immunology.
Competing interests statement
The authors declare competing financial interests: see the website
http://www.nature.com/naturemedicine for details.
RECEIVED 4 OCTOBER; ACCEPTED 4 DECEMBER 2002
1. Woolf, A.S. in Pediatric Nephrology 4th edn. (eds. Barratt, T.M., Avner, A. & Harmon,
W.) 1–19 (Williams & Wilkins, Baltimore, Maryland, 1999).
2. Rogers, S.A., Lowell, J.A., Hammerman, N.A. & Hammerman, M.R. Transplantation of
developing metanephroi into adult rats. Kidney Int. 54, 27–37 (1998).
3. Rogers, S.A., Liapis, H. & Hammerman, M.R. Transplantation of metanephroi across
the major histocompatibility complex in rats. Am. J. Physiol. Regul. Integr. Comp.
Physiol. 280, R132–136 (2001).
4. Rogers, S.A. & Hammerman, M.R. Transplantation of rat metanephroi into mice. Am.
J. Physiol. Regul. Integr. Comp. Physiol. 280, R1865–1869 (2001).
5. Rogers, S.A. & Hammerman, M.R. Transplantation of metanephroi after preservation
in vitro. Am. J. Physiol. Regul. Integr. Comp. Physiol. 281, R661–665 (2001).
6. Auchincloss, H. & Sachs, D.H. Xenogeneic transplantation. Annu. Rev. Immunol. 16,
433–470 (1998).
7. Hammerman, M.R. Xenotransplantation of renal primordia. Curr. Opin. Nephrol.
Hypertens. 11, 11–16 (2002).
8. Dekel, B. et al. Engraftment of human kidney tissue in rat radiation chimera: II.
Human fetal kidneys display reduced immunogenicity to adoptively transferred
human PBMC and exhibit rapid growth and development. Transplantation 64,
1550–1558 (1997).
9. Dekel, B. et al. In vivo modulation of the allogeneic immune response by human fetal
kidneys: the role of cytokines, chemokines, and cytolytic effector molecules.
Transplantation 69, 1470–1478 (2000).
10. Oliver, J.A., Barasch, J., Yang, J., Herzlinger, D. & Al-Awqati, Q. Metanephric mesenchyme contains embryonic renal stem cells. Am. J. Physiol. Renal Physiol. 283,
F799–809 (2002).
11. Gritsch, H.A. et al. The importance of nonimmune factors in reconstitution by discor-
60
dant xenogeneic hematopoietic cells. Transplantation 57, 906–917 (1994).
12. Cascalho, M. & Platt, J.L. The immunological barrier to xenotransplantation.
Immunity 14, 437–446 (2001).
13. Vermeulen, P.B. et al. Quantification of angiogenesis in solid human tumours: An international consensus on the methodology and criteria of evaluation. Eur. J. Cancer
32A, 2474–2484 (1996).
14. Dekel, B. et al. Engraftment of human kidney tissue in rat radiation chimera: I. A new
model of human kidney allograft rejection. Transplantation 64, 1541–1550 (1997).
15. Dekel, B. et al. Acute cellular rejection of human renal tissue by adoptive transfer of allogeneic human peripheral blood mononuclear cells into chimeric rats: sequential
gene expression of cytokines, chemokines and cytolytic effector molecules, and their
regulation by CTLA-4-Ig. Int. Immunol. 11, 1673–1683 (1999).
16. Boussiotis, V.A. et al. Differential association of protein tyrosine kinases with the T cell
receptor is linked to the induction of anergy and its prevention by B7 family-mediated
costimulation. J. Exp. Med. 184, 365–376 (1996).
17. Schwartz, R.H. Models of T cell anergy: Is there a common molecular mechanism? J.
Exp. Med. 184, 1–8 (1996).
18. Sayegh, M.H. & Turka, L.A. The role of T-cell costimulatory activation pathways in
transplant rejection. N. Engl. J. Med. 338, 1813–1821 (1998).
19. Li, Y. et al. Blocking both signal 1 and signal 2 of T-cell activation prevents apoptosis
of alloreactive T cells and induction of peripheral allograft tolerance. Nature Med. 5,
1298–1302 (1999).
20. Linsley, P.S. et al. CTLA-4 is a second receptor for the B cell activation antigen B7. J.
Exp. Med. 174, 561–569 (1991).
21. Eisen, M.B., Spellman, P.T., Brown, P.O. & Botstein, D. Cluster analysis and display of
genome-wide expression patterns. Proc. Natl. Acad. Sci. USA 95, 14863–14868
(1998).
22. Zuo, F. et al. Gene expression analysis reveals matrilysin as a key regulator of pulmonary fibrosis in mice and humans. Proc. Natl. Acad. Sci. USA 99, 6292–6297
(2002).
23. Medawar, P.B. Some immunological and endocrinological problems raised by the
evolution of viviparity in vertebrates. Symp. Soc. Exp. Biol. 7, 320–323 (1953).
24. Nelson, P.J. & Krensky, A.M. Chemokines, chemokine receptors, and allograft rejection. Immunity 14, 377–386 (2001).
25. Tedder, T.F., Steeber, D.A., Chen, A. & Engel, P. The selectins: vascular adhesion molecules. FASEB J. 9, 866–873 (1995).
26. O’Regan, A.W., Nau, G.J., Chupp, G.L. & Berman, J.S. Osteopontin (Eta-1) in cell-mediated immunity: Teaching an old dog new tricks. Immunol. Today 21,
475–478 (2000).
27. Ashkar, S. et al. Eta-1 (osteopontin): An early component of type-1 (cell-mediated)
immunity. Science 287, 860–863 (2000).
28. Xie, Y. et al. Expression, roles, receptors, and regulation of osteopontin in the kidney.
Kidney Int. 60, 1645–1657 (2001).
29. Pratt, J.R., Basheer, S.A. & Sacks, S.H. Local synthesis of complement component C3
regulates acute renal transplant rejection. Nature Med. 8, 582–587 (2002).
30. Segall, H., Lubin, I., Marcus, H., Canaan, A. & Reisner, Y. Generation of primary antigen-specific human cytotoxic T lymphocytes in human/mouse radiation chimera.
Blood 88, 721–730 (1996).
31. Reisner, Y. & Dagan, S. The Trimera mouse: generating human monoclonal antibodies and an animal model for human diseases. Trends Biotechnol. 16, 242–246 (1998).
32. Kreisel, D. et al. Non-hematopoietic allograft cells directly activate CD8+ T cells and
trigger acute rejection: An alternative mechanism of allorecognition. Nature Med. 8,
233–239 (2002).
33. Briscoe, D.M. et al. The allogeneic response to cultured human skin equivalent in the
hu-PBL-SCID mouse model of skin rejection. Transplantation 67, 1590–1599 (1999).
34. Benichou, G., Valujskikh, A. & Heeger, P.S. Contributions of direct and indirect T cell
alloreactivity during allograft rejection in mice. J. Immunol. 162, 352–358 (1999).
35. Woolf, A.S., Palmer, S.J., Snow, M.L. & Fine, L.G. Creation of a functioning chimeric
mammalian kidney. Kidney Int. 38, 991–997 (1990).
36. Kaminski, N. et al. Global analysis of gene expression in pulmonary fibrosis reveals distinct programs regulating lung inflammation and fibrosis. Proc. Natl. Acad. Sci. USA
97, 1778–1783 (2000).
37. Kaminski, N. & Friedman, N. Practical approaches to analyzing results of microarray
experiments. Am. J. Respir. Cell Mol. Biol. 27, 125–132 (2002).
NATURE MEDICINE • VOLUME 9 • NUMBER 1 • JANUARY 2003