* Department of Tumor Progression and Immune Defense, German Cancer Research Center, Heidelberg, Germany;
Basel Institute for Immunology, Basel Switzerland; and
Department of Applied Genetics, University of Karlsruhe, Karlsruhe, Germany
Correspondence: Margot Zöller, Department of Tumor Progression and Immune Defense, German Cancer Research Center, Im Neuenheimer Feld 280, D-69120 Heidelberg, Germany. E-mail: m.zoeller{at}dkfz.de
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Key Words: hematopoiesis • rodent • adhesion molecules • knockout • T lymphocytes
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
CD44 is well-known to play an important role in the differentiation and proliferation of myeloid and lymphoid progenitor cells at different maturation stages in the bone marrow microenvironment [23 24 25 26 27 28 ]. It facilitates stem-cell homing and immigration of pre-T cells into the thymus [15 , 16 , 27 28 29 ]. The large array of functional activities of CD44 in hematopoiesis is only partly because of the adhesive functions of CD44 [11 , 15 , 23 , 30 , 31 ]. CD44 is also known as a signal-transducing molecule, by itself or in cooperation with additional membrane molecules such as the TCR-CD3 complex or integrins [23 , 25 , 32 33 34 35 36 37 38 39 40 ]. However, an annotation of the multitude of distinct CD44 isoforms to specific functions is largely missing. This is a result of the fact that most studies on functional activity of CD44 in hematopoiesis have been performed with antibodies recognizing epitopes of the CD44 standard isoform; i.e., these antibodies bind to all CD44 isoforms and, hence, should be called panCD44-specific.
A particular involvement of CD44v6-containing isoforms in hematopoiesis became apparent when we noted that in transgenic mice, which express rat CD44v4-v7 under the control of the Thy1 promoter on thymocytes and peripheral T cells, repopulation of the thymus was accelerated significantly after lethal irradiation and reconstitution, the effect of the transgene being suppressed in the presence of a rat CD44v6-specific antibody [20 ]. Thus, expression of CD44v6 may provide a growth advantage during the intrathymic-selection processes, a feature which could well be of importance in stem-cell transplantation. Further exploration in the above-mentioned model has been hampered by the selective expression of the transgene on T cells as well as by its rat origin, which could have been a hindrance in ligand binding. With the availability of mice with a targeted deletion selectively of the CD44v6 and CD44v7 exons [41 ], we now could pursue this question. There also exist mice with a targeted deletion of the CD44v7 exon [41 ]. These mice have been characterized extensively with respect to progenitor cell homing, mobilization, and hematopoiesis [42 ]. By taking the features obtained with CD44v7-/- mice into account, it was possible to define CD44v6- or CD44v6/v7-specific functional activities, although mice with a selective deletion of CD44v6 are not yet available. The selectivity for CD44v6/CD44v6/v7 was confirmed by the use of a CD44v6-specific antibody [18 ].
Mice with a targeted deletion of CD44v6/v7, similar to mice with a targeted deletion of panCD44 [43 , 44 ], have no overt phenotype. Yet, we speculated that functional activities may become apparent under stress conditions. Indeed, we could demonstrate that CD44v6/CD44v6/v7 supports hematopoiesis and T-cell maturation within the thymus.
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Antibodies
The following hybridomas were obtained from the American Type
Culture Collection (ATCC; Manassas, VA): IM7 (anti-CD44s), RA3-3A1
(anti-CD45R/-B220), E13.161 (anti-stem cell antigen-1), M1/69
[anti-CD24/-heat stable antigen (HSA)], 33D1 [dendritic-cell
(DC)-specific], 331.12 (anti-µ), 145-2C11 (anti-CD3); or from the
European Collection of Animal Cell Cultures (Salisbury,
U.K.): YTA3.2.1 (anti-CD4), YTS169 (anti-CD8), and
YTS154.7.7.10 (anti-CD90/-Thy1). Monoclonal antibodies (mAb) specific
for H-2Dd (K9-18) and H-2Db (K7-65)
[45
] were kindly provided by G. Hämmerling (German
Cancer Research Center). The CD44v6-specific hybridoma [11A6, rat
immunoglobulin G (IgG)2a] has been described before
[18
]. mAb were purified by passage of culture
supernatants over protein G Sepharose 4B. Where indicated, purified mAb
were biotinylated or fluorescein isothiocyanate (FITC)-labeled. The
following mAb were obtained commercially: anti-CD34, anti-CD38,
anti-CD43, anti-CD117, biotinylated anticytokine antibodies, FITC- or
phycoerythrin (PE)-labeled anti-mouse, and anti-rat IgG, as well as
streptavidin-FITC and steptavidin-PE.
Preparation of hematopoietic cells
Mice were killed by cervical dislocation, and spleen, lymph
nodes, femura, tibiae, and thymus were removed under sterile
conditions. Bone-marrow cells (BMC) were obtained by flushing the bones
with 5 ml phosphate-buffered saline (PBS), containing 2% fetal calf
serum (FCS) using a 21G needle. Thymus, bone marrow, lymph nodes, and
spleen were teased through fine gauze.
Flow cytometry
BMC, lymph node cells (LNC), spleen cells (SC), and thymocytes
(TC; 5x105 cells/well) were stained according to routine
procedures. For intracellular staining of cytokines, cells were fixed
and permeabilized in advance. Negative controls were incubated with an
isotype-matched, control IgG and the secondary antibody. For flow
cytometry, purified mAb were used at a final concentration of 10
µg/ml. Programmed cell death was evaluated by annexin
V-FITC/propidium iodine (PI) staining using an apoptosis detection kit.
Analysis was performed with a FACSCalibur (Becton Dickinson,
Heidelberg, Germany).
Colony-forming assays [46
]
Colony-forming activity was determined by plating 5 x
104 BMC in 24-well plates in 0.3% semisolid agar (Difco,
Detroit, MI) in Iscove’s minimal essential medium containing
20% horse serum and keeping the cultures at 37°C, 5%
CO2 in air. For the evaluation of granulocyte-macrophage
colony-forming units (GM-CFU), 10 ng/ml GM-colony-stimulating
factor (CSF; Sigma Chemical Co., Germany) and 2 mM glutamine were
added. Colonies containing at least 50 cells were counted after 7–8
days. Macrophage (M)-CFU were cultured under the same conditions,
adding conditioned medium (20%) from L929 cells as a source of M-CSF.
For multilineage (C)-CFU, 15% conditioned medium from WEHI-3B cells
and 5% conditioned medium from L929 cells were added. Erythroid
burst-forming units (E-BFU) were determined by addition of 20% FCS
instead of HS, 1% bovine serum albumin (fraction V; Sigma Chemical
Co.), 15% conditioned medium of WEHI-3B cells, and 10 U/ml
erythropoetin (Boehringer Mannheim, Mannheim Germany). Colonies
containing at least 500 cells were counted after 12 days. Pre-B-cell
(pre-B)-CFUs were also cultured with 20% FCS, 15% supernatant of an
interleukin (IL)-7 cDNA-transfected line (kindly provided by Dr.
Rolink, Basel Institute for Immunology, Switzerland)
[47
], 2 mM L-glutamine, and 5 x
10-4 M 2-mercaptoethanol (2-ME). Colonies of at least 50
cells were counted after 7 days.
Long-term reconstitution and immunoreactivity
Long-term reconstitution was evaluated by the transfer of
limiting numbers of BMC in lethally irradiated mice. Where indicated,
the reconstituted mice received 100 µg anti-CD44v6 using rat IgG or
the rat IgG2a mAb 20.6.5 (anti-mouse-vß2) as controls. The percentage
of surviving mice, the repopulation with leukocytes, the recovery of
CFU in BMC, and the reappearance of immunocompetence have been
controlled. For the latter aspect, proliferative activity and cytokine
expression, as well as antibody production of SC and LNC in response to
an allogeneic stimulus, were evaluated by 3H-thymidine
uptake, flow cytometry, and enzyme-linked immunosorbent assay (ELISA),
respectively. Where indicated, T-cell activation was also evaluated by
incubating LNC on anti-CD3 (10 µg/ml)-coated plates and determining
3H-thymidine uptake after 48 h of incubation.
Susceptibility of TC for programmed cell death was evaluated by
incubation TC on anti-CD3 (10 µg/ml)-coated plates and staining with
annexin V-FITC and PI. In allogeneically reconstituted mice, the
percentage of donor and host cells was evaluated by flow cytometry. The
frequency of host- and donor-reactive proliferating T cells was
obtained by limiting dilution (LD) analysis. Cells (24 replicates) were
titrated from 11,200 to 100 cells/well in the presence of
104 irradiated stimulator lymphocytes.
3H-Thymidine incorporation was determined after 8 days of
culture. The frequency of proliferating cells was calculated according
to the formula F0 (fraction of nonresponding cultures) = e-u, where u = c/w (number of c cells
distributed in w wells) [48
].
Statistical analysis
Significance of differences among controls, knockout mice, and
antibody-treated groups was calculated according to the Wilcoxon
rank-sum test (in vivo assays) or the Student’s t-test (in
vitro studies). All functional assays were repeated at least three
times. Mean values and standard deviations of in vivo experiments are
derived from 20 mice/group. Mean values of in vitro studies are based
on 4–10 replicates as indicated in the individual experiments.
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
View this table: [in a new window] |
Table 1. Hematopoietic Progenitor Cells in the Bone Marrow of
CD44v6/v7-Competent and -Deficient Mice
|
Lymphocyte activation in CD44v6/v7-deficient mice
Although this aspect is not central to our question, some
information is required with respect to GvH and HvG reactivities as
described below. In brief, the distribution of leukocyte subsets
appears to be unaltered in CD44v6/v7-/- mice (Table 2
). We noted, however, a slight reduction in the response toward an
allogeneic stimulus and a strongly reduced responsiveness toward
cross-linking CD3. The latter accounted for the proliferative response
of lymph node cells as well as for induction of apoptosis in
thymocytes. Up-regulation of cytokine expression during activation was
not altered significantly.
|
View this table: [in a new window] |
Table 2. Lymphocyte Subset Distribution and Expression of Activation Markers
in Resting and Activated Lymph Node Cells of CD44v6/v7-Competent
and -Deficient Mice
|
Influence of CD44v6/v7 on long-term reconstitution
Indeed, an involvement of CD44v6 in hematopoiesis was corroborated
by long-term reconstitution experiments. BMC from
CD44v6/v7-/- mice displayed a significantly reduced
reconstitutive potential. A higher number of BMC for long-term survival
was also required when CD44v6/v7+/+ BMC were transferred,
and CD44v6 was blocked by repeated injections of anti-CD44v6
(Fig. 1 A
). Yet, reconstitution was unimpaired when
CD44v6/v7-/- mice received CD44v6/v7+/+ BMC
(unpublished results). In line with these findings was the observation
that early after reconstitution in the anti-CD44v6-treated host and the
host reconstituted with BMC from CD44v6/v7-/- mice,
reduced numbers of BMC and SC were recovered (Fig. 1B)
. In mice
reconstituted with CD44v6/v7-/- BMC, frequencies of all
CFU, in mice receiving anti-CD44v6 frequencies of C-CFU and GM-CFU,
were reduced (Fig. 1C) . Furthermore, early after the transfer, a higher
percentage of CD34+, CD90+, CD117+,
and SCA-1+ cells was recovered from the bone marrow of
reconstituted than from nontransplanted, age-matched controls (see
Table 1
). However, no increase in the percentage of CD34+
and CD117+ BMC was seen when CD44v6/v7-/- BMC
were transferred. The percentage of CD4+, CD8+,
and sIgM+ cells was similar in mice receiving
CD44v6/v7+/+ BMC with or without concomitant application of
anti-CD44v6 and in mice receiving CD44v6/v7-/- BMC. Also,
the percentage of CD44s+ and CD44v6+ cells was
unaltered in the bone marrow of transplanted as compared with
nontransplanted mice. However, expression of CD44v6 was slightly
decreased in anti-CD44v6-treated mice. With respect to CD49d
expression, a slight decrease has been seen in mice receiving
CD44v6/v7-/- BMC as well as in mice receiving anti-CD44v6
(Fig. 1D) . At 6–10 weeks after reconstitution, cellularity and the
numbers of CFU had reached control levels in most instances; i.e.,
similar numbers of BMC were recovered irrespective of the transferred
BMC and antibody treatment. No further impact of anti-CD44v6 treatment
on the number of CFU was observed, and the number of CFU in mice that
had received BMC from CD44v6/v7-/- mice was only slightly
reduced, as has been described above for the native
CD44v6/v7-/- mouse. It is important, too, that recovery
of immunocompetence as evaluated by proliferative activity,
up-regulation of cytokine expression, and antibody secretion in
response to an allogeneic stimulus was delayed significantly when the
host was treated with anti-CD44v6 or had received BMC from
CD44v6/v7-/- mice (Table 3
).
 View larger version (29K): [in a new window] |
Figure 1. Survival and recovery after lethal irradiation and reconstitution with
syngeneic BMC from CD44v6/v7-competent and -deficient mice. (A)
Lethally irradiated CD44v6/v7+/+ F1 mice were reconstituted
with titrated numbers of BMC from syngeneic CD44v6/v7-/-
F1 mice or from syngeneic CD44v6/v7+/+ F1 mice. Where
indicated, mice were treated with anti-CD44v6 [100 µg/mouse,
intravenously (i.v.), twice per week], starting at the day of
reconstitution. Survival of animals (20/group) was monitored for 16
weeks. The majority of nonsurviving mice became moribund between 15 and
25 days after reconstitution. (B–D) Lethally irradiated F1 mice,
reconstituted with 3 x 104 BMC as described above,
were sacrificed at 2, 3, 4, 5, and 6 weeks after reconstitution to
evaluate (B) the number of BMC and SC, (C) the number of CFU, and (D)
at 2 weeks after reconstitution, the composition of BMC, as revealed by
flow cytometry. Mean values ± SD of five experiments
are shown. *, Significance of differences.
|
|
View this table: [in a new window] |
Table 3. Influence of CD44v6 on the Recovery of Immunocompetence after Lethal
Irradiation and Syngeneic Reconstitution
|
When titrated numbers of CD44v6/v7+/+ BMC were transferred into the lethally irradiated allogeneic host, a higher number of cells were required upon transfer in the syngeneic host to guarantee survival. Instead, when CD44v6/v7-/- BMC were transferred into lethally irradiated, allogeneic BALB/c mice, a comparable number of BMC, as in the syngeneic host, sufficed for survival, although still a higher number were required as compared with BMC from a CD44v6/v7+/+ donor. When the allogeneic host received anti-CD44v6, the number of transferred BMC required for survival was not increased as compared with the syngeneic host (Fig. 2 A ). Finally, a lower number of CD44v6/v7+/+ BMC sufficed for survival of the allogeneic CD44v6/v7-/- host than of the CD44v6/v7+/+ host (Fig. 2C) .
 View larger version (27K): [in a new window] |
Figure 2. Survival after lethal or nonlethal irradiation and allogeneic
reconstitution with BMC from CD44v6/v7-competent or -deficient mice.
Lethally (A) and nonlethally (B) irradiated BALB/c mice were
reconstituted with titrated numbers of BMC from
CD44v6/v7-/- or CD44v6/v7+/+ F1 mice. Where
indicated, mice were treated with anti-CD44v6 (100 µg/mouse, i.v.,
twice per week), starting at the day of reconstitution. Lethally (C)
and nonlethally (D) irradiated CD44v6/v7-/- and
CD44v6/v7+/+ F1 mice were reconstituted with titrated
numbers of allogeneic BALB/c BMC. Survival of animals (20/group) was
monitored for 16 weeks. (E) The percentage of mice surviving after the
transfer of 1–10 x 104 syngeneic or allogeneic BMC
in dependence on anti-CD44v6 treatment and CD44v6/v7 expression is
shown.
|
Thus, as in the lethally irradiated allogeneic host, a deficit in CD44v6/v7 had only a minor negative impact on reconstitution as compared with the significant negative impact on syngeneic reconstitution. Antibody treatment appeared to be advantageous. To reveal the underlying mechanism, we evaluated repopulation, recovery of donor versus host hematopoiesis, and GvH as well as HvG reactivities.
Influence of CD44v6/v7 on engraftment and graft acceptance
The overall recovery of hematopoietic cells during the starting 6
weeks after allogeneic reconstitution of the nonlethally irradiated
host supported our assumption that death in the nonlethally irradiated
allogenically reconstituted host might have been a result of GvH and/or
HvG reactions (Fig. 3
). Although the number of BMC increased, albeit much more slowly
than in the syngeneically reconstituted host, less TC and SC were
recovered 3 and 4 weeks as compared with 2 weeks after reconstitution
of the CD44v6/v7+/+ allogeneic host with
CD44v6/v7+/+ BMC. This did not hold true when the donor was
CD44v6/v7-/-, in which case fewer cells were recovered
from the bone marrow, the thymus, and the spleen early after
reconstitution, yet the numbers increased steadily. A steadily
increasing number of cells were also recovered from allogeneically
reconstituted, anti-CD44v6-treated mice. Finally, it should be noted
that the recovery of BMC in the CD44v6/v7-/- host was
extremely low, and an increased number of cells were recovered from the
spleen. This finding, however, may not be a result of the absence of
CD44v6. Instead, it is likely a consequence of the absence of CD44v7,
because similar findings were observed with CD44v7-/-
mice [42
]. Evaluating the proliferative activity of BMC,
SC, and TC 4 weeks after reconstitution indirectly supported our
interpretation that the stagnation in the expansion of BMC and the
reduction in the number of SC and TC have been because of GvH and HvG
reactions; i.e., freshly harvested cells proliferated vigorously, and
proliferation of BMC exceeded the one of SC and, more pronounced, of
TC. It should be noted that there has been only one exception; i.e., a
higher proliferation rate of SC than of BMC has been observed when
CD44v6/v7-/- mice received CD44v6/v7+/+ BMC
from BALB/c mice. It is interesting, too, that despite the higher
recovery of cells from anti-CD44v6-treated mice, BMC and SC of those
mice displayed reduced proliferative activity.
 View larger version (23K): [in a new window] |
Figure 3. Recovery of hematopoietic cells after nonlethal irradiation and
allogeneic reconstitution with BMC from CD44v6/v7-competent or
-deficient mice. Nonlethally irradiated BALB/c mice (A and C) and
CD44v6/v7+/+ as well as CD44v6/v7-/- F1 mice
(B and D) were reconstituted with 2.5 x 104
allogeneic BMC as described in Figure 2
. (A and B) The number of BMC,
SC, and TC was monitored at 2, 3, 4, and 6 weeks after reconstitution.
(C and D) The proliferative activity (mean values±SD of
four experiments) of freshly harvested BMC, SC, and TC at 4 weeks after
reconstitution is shown. *, Significance of differences.
|
 View larger version (24K): [in a new window] |
Figure 4. Distribution of donor and host cells in nonlethally irradiated,
allogeneically reconstituted mice in dependence on CD44v6 blockade or
CD44v6/v7 deficiency of donor or host. Nonlethally irradiated BALB/c
mice and CD44v6/v7+/+ as well as CD44v6/v7-/-
F1 mice were reconstituted with 2.5 x 104 allogeneic
BMC as described in Figure 2
. The percentage of donor- and host-derived
cells was evaluated at 2, 3, 4, and 6 weeks after reconstitution in
bone marrow (A), thymus (B), and spleen (C) by staining with
FITC-labeled H-2Dd- or H-2Db-specific mAb. The
total number of donor- and host-derived cells (mean of three
independently performed experiments) is presented.
|
 View larger version (27K): [in a new window] |
Figure 5. Distribution of donor and host T and B cells in the spleen of
nonlethally irradiated, allogeneically reconstituted mice in dependence
on CD44v6 blockade or CD44v6/v7 deficiency of donor or host. SC were
double-stained with biotinylated anti-CD4/anti-CD8 (counterstaining
with Strep-PE) or anti-mIgM-PE and FITC-labeled anti-H-2Dd
or H2Db at 2, 3, 4, and 6 weeks after reconstitution. The
total number of donor- and host-derived CD4+,
CD8+, and sIgM+ cells (mean of three
independently performed experiments) is shown.
|
Finally, it should be mentioned that the defect in T-cell maturation of CD44v6/v7-/- BMC had long-term consequences on the reconstitution of the allogeneic, nonmyeloablatively, pretreated host. In control and anti-CD44v6-treated mice, over 90% of hematopoietic cells were donor-derived 12 weeks after reconstitution (unpublished results). Different patterns were seen in mice receiving allogeneic CD44v6/v7-/- BMC. Roughly 50% of the mice displayed a mixed donor-host hematopoietic chimerism. In the remaining animals, donor or host cells were dominating with up to over 90% of hematopoietic cells displaying the H-2d or the H-2b haplotype (unpublished results).
So far, our results supported the interpretation that the unimpaired reconstitution of the anti-CD44v6-treated, nonlethally irradiated, allogeneic host may have been a result of a reduction in GvH and HvG reactivities, which compensated for the impaired hematopoiesis. The mild impairment seen after the transfer of CD44v6/v7-/- BMC could well be because of a reduction in GvH reactivity. Accordingly, the unimpaired reconstitution of the CD44v6/v7-/- host could be the result of reduced HvG reactions. This interpretation was controlled by evaluating the frequencies of donor- and host-reactive T cells in the thymus and the spleen of the nonlethally irradiated, allogeneically reconstituted host.
Influence of CD44v6/v7 on tolerance induction
Antidonor and antigraft reactivity were evaluated starting 3 weeks
after the transfer, because most of the animals succumbing to runt
disease became moribund between 15 and 25 days after BMC
transplantation. Frequencies of host- and donor-reactive TC and SC were
evaluated by thymidine incorporation under LD conditions. The values in
Table 4
represent the actual frequencies of donor- and host-reactive T
cells; i.e., overall frequencies have been corrected for the percentage
of host- and donor-derived T cells as revealed by flow cytometry.
|
View this table: [in a new window] |
Table 4. Frequencies of Donor- and Host-Reactive T Cells in Thymus and Spleen of
Allogeneically Reconstituted Mice in Dependence on CD44v6 Blockade
or CD44v6/v7 Deficiency of Donor or Host
|
Considering the influence of CD44v6/v7 on T-cell maturation, 3 weeks after reconstitution, frequencies of donor-reactive TC and SC were strongly reduced when the host was devoid of CD44v6/v7. Yet, a deficiency of CD44v6/v7 on host cells had no impact on the frequencies of donor antihost-reactive T cells. Correspondingly, when the donor was CD44v6/v7-/-, host antidonor reactivity was unaltered (TC) or slightly reduced (SC), but the frequencies of donor-derived, host-reactive TC and SC were strongly diminished. When mice received anti-CD44v6 donor-reactive frequencies (but not host-reactive), TC and SC were strongly reduced in the BALB/c host. When F1 mice received BALB/c BMC, anti-CD44v6 influenced antidonor and antihost reactivity, the frequencies of antihost-reactive T cells being more severely injured.
In all settings, donor cells apparently were fully tolerant at 6–8 weeks after reconstitution; i.e., very low frequencies of host-reactive T cells were detected. There remained a low level of host antidonor reactivity, which was independent of CD44v6/v7 expression of the host as well as of anti-CD44v6 treatment (unpublished results).
Thus, frequencies of GvH- and HvG-reactive T cells were strongly reduced in the absence of CD44v6/v7 on the respective T (progenitor) cells. Hence, CD44v6/v7 is involved in T-cell maturation and T-cell activation. Instead, tolerance induction/negative selection appeared to be independent of CD44v6/v7 expression on T cells as well as on dendritic cells/monocytes. This is suggested by the finding that 6 weeks after reconstitution, GvH- as well as HvG-reactive T cells were hardly detected, irrespective of CD44v6/v7 expression/deficiency on donor T cells and, importantly, on host-derived, antigen-presenting cells in the thymus.
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
CD44v6 is expressed at a low density [10 , 16 , 49 ] on subpopulations of BMC and TC. The majority of CD44v6+ BMC expresses CD90, CD117, and SCA-1. Except for a slight reduction in CD49d+ cells, we did not observe gross changes in marker-defined subpopulations of BMC from CD44v6/v7-/- versus CD44v6/v7+/+ BMC. This included expression of CD44s. Yet, the number of CFU, particularly of multilineage progenitors, was decreased. This weakness of CD44v6/v7-/- mice, although not resulting in a pathological phenotype under nonstress conditions, became important upon lethal irradiation and syngeneic reconstitution; i.e., a significantly higher percentage of mice reconstituted with CD44v6/v7-/- BMC died, the period of recovery was prolonged, and regain of immunocompetence was delayed. Similar features were observed in syngeneically reconstituted mice receiving anti-CD44v6. It should be mentioned that reduced numbers of CD34+ and CD117+ BMC were recovered early after the transfer of CD44v6/v7-/- BMC, and expression of CD49d was reduced slightly in BMC of mice reconstituted with CD44v6/v7-/- BMC or receiving anti-CD44v6. We are currently exploring whether expression of CD44v6 or ligand-binding of CD44v6 may directly influence CD34, CD49d, and CD117 expression and whether it is CD44v6/v7/CD44v6 itself or its influence on associating molecules that accounts for the impaired reconstitutive capacity. Irrespective of this open question, the findings supported functional activity of CD44v6/v7 in hematopoiesis.
To attack the question of an impact of CD44v6/v7 on early hematopoiesis versus lymphopoiesis and T-cell maturation, the nonlethally irradiated, allogeneic host was reconstituted with CD44v6/v7-/- BMC or received anti-CD44v6 treatment concomitantly with the graft. We had described that an antipanCD44 treatment has no major influence on syngeneic reconstitution but is lethal in the allogeneic host because of a gut-associated GvH reaction [61 ]. Instead, an antibody blockade of CD44v6 had no negative impact on the survival rate of the allogeneic host, and a deficit of CD44v6/v7 on the bone marrow graft led only to a weak impairment of the reconstitutive capacity. Similar findings accounted for the sublethally irradiated, allogeneic host.
Two in vivo features of the allogeneic reconstitution should be discussed. First, the number of BMC required for reconstitution of the lethally irradiated allogeneic, as compared with the syngeneic, host was increased when transferring CD44v6/v7+/+ BMC, unaltered when transferring CD44v6/v7-/- BMC, and even slightly decreased when the host received anti-CD44v6. We interpret these findings in the sense that because of the failure to activate killer inhibitor receptors, donor natural killer (NK)/NK-T cells may have sufficed to induce lethal GvH reactions in the gut. With increasing the number of transferred CD44v6/v7+/+ BMC, there might have been sufficient donor-derived hematopoietic cells to interfere with the cytotoxic activity of NK/NK-T cells. In line with this interpretation was the finding that even after the transfer of high numbers of allogeneic BMC, mice went through a crisis of GvH reactions with rough fur and weight loss. However, the vast majority of mice finally recovered. Why has no increased number of BMC been required when transferring CD44v6/v7-/- BMC or when treating the host with anti-CD44v6? One possible explanation could be that NK/NK-T cells are CD44v6+, their activity being blocked by anti-CD44v6, respectively, mitigated in the absence of CD44v6/v7. With respect to the transfer of CD44v6/v7-/- BMC, an alternative explanation should be mentioned. CD44v7 inhibits activation-induced cell death of gut-associated T cells [41 ]. Thus, in the absence of CD44v6/v7, activated T cell will become eliminated more readily. Such a mechanism could also explain the transient feature of gut-associated GvH reactions in the allogeneically reconstituted mouse.
Our second concern relates to the decreased survival rate of the nonmyeloablatively pretreated host after the transfer of intermediate numbers of allogeneic BMC. Although to a varying degree, this feature has been observed, irrespective of anti-CD44v6 treatment and irrespective of CD44v6/v7 expression on host or donor cells. Because in this setting GvH as well as HvG reactivities play an important role, it is possible that by increasing the number of donor cells, an unfavorable balance between donor and host lymphocytes has been approached, such that by mutual killing, the number of host hematopoietic cells becomes insufficient for survival, and concomitantly, the remaining host hematopoietic cells are insufficient, too. The transient decrease in donor as well as host cells would be in line with this interpretation. Dissecting moribund mice also provided evidence that mice were anemic, and their bone marrow, spleen, and thymus were rather empty or completely devoid of lymphocytes. There have been signs of GvH reactivity (destruction of the gut), too, but depletion of hematopoietic cells has been the dominating feature. Whether the supposed mutual killing is exclusively T-cell-mediated or supported by NK/NK-T cells remains to be explored. Recent experiments transferring NK-cell-depleted BMC or depleting NK cells in the host support a major role of NK cells in accceptance of an allogeneic graft in the sublethally irradiated host (unpublished results).
To further elucidate the mechanisms responsible for the improved reconstitution of the allogeneic as compared with the syngeneic host by CD44v6/v7-/- BMC and by anti-CD44v6 treatment, an analysis of donor and host cells and of frequencies of donor- and host-reactive lymphocytes was performed in the nonlethally irradiated host reconstituted with an intermediate number (2.5x104) of allogeneic BMC. As in the lethally irradiated syngeneic host, there was evidence for a relative weakness in hematopoiesis of CD44v6/v7-/- BMC and in the anti-CD44v6-treated host. However, this weakness became balanced by an additional defect in T-cell maturation. The latter interpretation has been deduced from the following observations: i) A lower percentage of donor cells was recovered in the thymus than in the bone marrow. ii) The number of donor-derived CD4+ and CD8+ cells was strongly reduced in the spleen. iii) There was a steady decrease of donor-derived cells in thymus and spleen; i.e., early after transfer, the majority of cells in the thymus and spleen were of donor origin (unpublished results), and an excess of host cells was seen later on. This accounted for roughly 25% of the mice. In the remaining mice, an equal distribution of donor- and host-derived or a dominance of host-derived hematopoietic cells was observed. Similar variations in the degree of hematopoietic chimerism have been shown for allogeneically reconstituted dogs and miniature swines after nonmyeloablative pretreatment [62 , 63 ]. Yet, it is remarkable that a dominance of donor cells was seen only in 25% of mice receiving CD44v6/v7-/- BMC, and it was observed in over 90% of mice reconstituted with CD44v6/v7+/+ BMC. iv) CD44v6/v7-/- TC displayed reduced responsiveness toward TCR engagement; i.e., after cross-linking CD3, the percentage of apoptotic TC was significantly lower than that of CD44v6/v7+/+ TC. The observation that HvG reactivity was reduced when the host was CD44v6/v7-/- also supported our hypothesis that the CD44v6/v7 deficiency hampers T-cell maturation in the thymus.
Despite this disadvantage in maturation of CD44v6/v7-/- thymocytes, a deficiency of CD44v6/v7 on host antigen-presenting cells had no influence on the process of negative selection/tolerance induction. Twelve weeks after reconstitution, hardly any host-reactive T cells were detected, even in mice that were repopulated by CD44v6/v7-/- donor cells. The recovery of only low numbers of host-reactive T cells was also independent of whether the host expressed CD44v6/v7 (unpublished results).
Clinically more relevant than the knockout model will be the modulation of engraftment and GvH/HvG reactivities by anti-CD44v6 treatment of the nonmyeloablatively pretreated, allogeneic host. Basically, the effects of anti-CD44v6 treatment corresponded to the ones observed after the transfer of CD44v6/v7-/- BMC in the CD44v6/v7+/+ host and vice versa, i.e., a negative impact on early hematopoiesis, mitigation of GvH and HvG reactivities, and no influence on tolerance induction. However, one peculiarity should be discussed. When BALB/c mice received F1 BMC, anti-CD44v6 treatment had a stronger impact on antidonor than antihost reactivity. When F1 mice received BALB/c BMC, antihost reactivity was more strongly impaired in anti-CD44v6-treated mice; i.e., in both settings, the BALB/c anti-F1 reactivity was more strongly affected. We had supposed we would see an equal reduction on GvH and HvG reactivity in all instances. One possible explanation could have been the different levels of CD44 expression on hematopoietic cells of BALB/c versus F1 mice [64 ]. Irrespective of these minor strain-dependent differences, anti-CD44v6 treatment mitigated GvH and HvG reactivities.
There are first studies on the favorable outcome of allogeneic reconstitution of the nonmyeloablatively conditioned cancer patient [52 , 65 66 67 ]. Furthermore, several protocols, such as infusion of presensitized donor lymphocytes [68 ], anti-CD3 treatment [69 ], and preactivation of donor-reactive lymphocytes of the host accompanied by CTLA4-Ig treatment [70 ], have been explored to convert a mix toward a complete donor chimerism. Based on our observation of a transient reduction in GvH and HvG reactivity in the CD44v6/v7-/- donor and host, respectively, as well as in the anti-CD44v6-treated host, and taking into account the restricted expression of CD44v6, we consider a transient anti-CD44v6 treatment as most appropriate to cope with graft rejection and with acute episodes of GvH reactions.
Received May 16, 2001; revised September 30, 2001; accepted October 11, 2001.
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
