| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
*
Department of Pathology and
Graduate Program in Immunology, University of Texas Southwestern Medical Center, Dallas, TX 75235
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Despite strides in the understanding of NK cell specificity and function, the developmental sequence leading to generation of NK cells expressing a mature cell surface phenotype remains to be fully elucidated. Human CD34+ hematopoietic progenitor cells have been described that will give rise to CD56+ NK cells in vitro under the influence of IL-2 or IL-15 alone (7, 8) or in conjunction with other cytokines (8, 9, 10, 11, 12). Stem cell factor (SCF), for example, generally enhances the expansion of NK cells from CD34+ progenitors when used with IL-15 or IL-2, but alone it has no effects on NK differentiation. However, mice with mutations in SCF or its receptor do not show significant defects in generation of NK cells, perhaps because of redundancy with another factor (13, 14). Interestingly, flt3 ligand (flt3L) has strong homology to SCF, and both are produced by stromal cells (15, 16). Like SCF, flt3L can synergize with a number of other growth factors to support proliferation of primitive mouse and human hematopoietic progenitor cells (17, 18, 19, 20), and administration of flt3L in vivo causes expansion of NK cells (21, 22). Knockout mice lacking flt3L exhibit a deficiency with low to absent mature NK cells (23), an observation that warrants further analysis of the role of this cytokine in NK cell development.
We have previously identified a murine progenitor population characterized as lineage marker (Lin)-, c-kit+, Sca2+ that is unable to respond efficiently to IL-15 alone (24). Preculture of these cells with IL-6, IL-7, SCF, and flt3L induces IL-15 responsiveness in a fraction of the treated cells, and subsequent culture in IL-15 leads to the generation of large numbers of NK1.1+ but Ly49- cells. In this manuscript, a refined characterization of those cytokines that support NK cell generation in vitro is presented. These studies reveal a critical role for flt3L in NK cell differentiation. Additionally, a stromal cell-based system is presented that is capable of generating Ly49+ NK cells from bone marrow-derived progenitors.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
C57BL/6 or (C57BL/6 x DBA.2)F1 mice 7–12 wk old, bred at University of Texas Southwestern Medical Center, were used as the source of bone marrow progenitors. CB17.SCID mice obtained from Taconic Farms (Germantown, NY) were used as recipients for in vivo repopulation studies. The SCID mice were irradiated with 175 rad using a Mark I irradiator (J. L. Shepherd and Associates, San Fernando, CA) and were maintained under pathogen-free conditions with acidified water containing antibiotics (200 mg neomycin sulfate, 200 mg sulfamethoxazole, 50 mg trimethoprim, and 5.2 mg polymixin B in 500 ml H20, pH 2.5, all from Sigma, St. Louis, MO) for 2 wk followed by acidified water only for the remaining 1–2 wk.
Monoclonal Abs
Except as noted below, all mAbs and their isotype controls were
obtained from PharMingen (San Diego, CA). Anti-Ly49G2 (4D11) and
anti-Ly49D (4E5) were provided by Dr. J. Ortaldo (National
Cancer Institute, Frederick, MD), anti-Ly49A (JR9-318) was the gift
of Dr. J. Roland (Institut Pasteur, Paris, France), and anti-Sca2
hybridoma supernatant was provided by Dr. G. Spangrude (University of
Utah, Salt Lake City, UT). Goat anti-rat 
Texas Red (Southern
Biotechnology Associates, Birmingham, AL) or streptavidin-Red670 (Life
Technologies, Grand Island, NY) was used to detect some primary
Abs.
Stromal cells
OP9 stromal cells (25) were the kind gift of Dr. J.C. Zúñiga-Pflücker (University of Toronto, Ontario, Canada). The cells were passaged weekly in MEM (Life Technologies) containing 20% FCS, 200 nM glutamine, 100 µg/ml streptomycin sulfate, and 100 U/ml penicillin.
Cell preparation, isolation, and analysis of precursor cells
Lin-, c-kit+,
Sca2+ progenitors were isolated by a combination
of magnetic bead depletion of Lin+ cells and cell
sorting as previously described (24).
Lin-, c-kit+,
flt3+ or flt3- progenitors
were isolated in a similar fashion. Analysis of cultured cells was
performed by sequential staining with anti-FcIII/II (2.4G2), a
biotinylated mAb, and finally streptavidin-Red670 plus a FITC- and/or
PE-conjugated mAb.
In vitro culture conditions
Sorted Lin-, c-kit+, Sca2+; Lin-, c-kit+, flt3+; or Lin-, c-kit+, flt3- marrow cells were cultured in 96-well U-bottom plates (Falcon, San Jose, CA) at 10,000 cells/well in 0.2 ml of complete RPMI (RPMI 1640 containing 10% FBS, 100 µg/ml streptomycin sulfate, 100 U/ml penicillin, 1 mM sodium pyruvate, 2 mM glutamine, and 1x nonessential amino acids) and a mixture of 0.5 ng/ml murine IL (mIL)-7 (PeproTech, Rocky Hill, NJ), 30 ng/ml mouse SCF (BioSource, Camarillo, CA), and 100 U/ml murine flt3L (a gift from Dr. D. Rennick at DNAX, Palo Alto, CA). The cells were refed with the same media on day 3, and then on day 5 the cultures were harvested, washed, counted, and replated at 15,000/well in complete RPMI containing 30 ng/ml mIL-15 (a gift from Dr. T. Trout, Immunex, Seattle, WA), 150 ng/ml human IL-15 (R&D Systems, Minneapolis, MN), or 5000 IU/ml of recombinant human IL-2 (Chiron, Emeryville, CA). Cells grown on OP9 stroma were plated on day 5 at 200,000–500,000 cells per well in 24-well plates (Falcon) containing a confluent monolayer of OP9 stromal cells in IL-15- or IL-2-containing media. After an additional 3 days, the cultures were refed with the same media, and on day 11 or 12 of total culture time, the cells were harvested for analysis.
In vivo repopulation studies
Lymphocyte- and blast-sized (LB) Lin-, c-kit+, flt3+, or flt3- progenitors were sorted from (C57BL/6 x DBA.2)F1 bone marrow as described above. The percentage of each population in whole bone marrow was calculated (% = fraction of LB cells x fraction of Lin- cells in LB gate x fraction of c-kit+, flt3+ or c-kit+, flt3- cells in LB, Lin- gate x 100). The cells were washed two times with PBS and resuspended such that the equivalent of 1.4 x 107 whole bone marrow cells for each progenitor was contained in 0.75 ml. CB17.SCID mice were irradiated with 175 rad 0.5–1 h before injection. The mice were then injected with 0.75 ml of cells via the tail vein and maintained as described above. Donor type cells were detected with anti-Kb FITC (PharMingen).
Western blot analysis
IL-2 cultured SCID NK cells (1–2 x 106 cells) or NK cells derived from Lin-, c-kit+, Sca2+ progenitor cells (105 cells) were lysed in a HNTG lysis buffer (10% glycerol, 50 mM HEPES, pH 7.4, 150 mM NaCl, 1 mM EDTA, and 1 mM MgCl2) with 1% Triton X-100 for 30 min on ice. Cleared lysates were separated on a 4–20% SDS-PAGE gel after addition of 2x sample buffer (50 mM Tris, pH 6.8, 2% SDS, 0.1% bromphenol blue, and 5% 2-ME). Separated proteins were transferred onto a nitrocellulose membrane. Western analysis was done after blocking the membrane in TBST (20 mM Tris and 500 mM NaCl) with 3% nonfat dry milk for 30 min. Primary anti-CD94-2 Ab (26) was added at a dilution of 1:200 in TBST for 1 h at room temperature, followed by the secondary Ab (donkey anti-rabbit IgG-HRP conjugate, 1:1000 dilution, Amersham Pharmacia Biotech, Piscataway, NJ) for 1 h. The membrane was developed in enhanced chemiluminescence developing solution (Amersham Pharmacia Biotech). A CD94-GST fusion protein generated in yeast was used as a control. Lysates from T cell lines EL-4 and Bw5147 did not show CD94 expression (data not shown).
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
We previously identified a population of multipotent progenitors
in mouse bone marrow characterized as Lin-
(B220, Gr1, Mac1
, CD2, Ter119, and NK1.1),
c-kit+, Sca2+, that
gives rise to NK1.1+ cells in vitro
(24). Although culture of these progenitors in mIL-15
alone led to generation of small numbers of lytic
NK1.1+ cells, primary culture in a mixture of
early acting cytokines, including IL-6, IL-7, SCF, and flt3L, for 5
days before secondary culture in mIL-15 significantly enhanced by 10-
to 30-fold the number of NK cells generated. We inferred from these
data that primary culture of the progenitor cells in early acting
cytokines gave rise to a population of IL-15-responsive precursors that
subsequently gave rise to NK cells on exposure to IL-15. Indeed, we
previously demonstrated that, whereas the starting population of
progenitors was IL-2/15Rß-, 5–15% of the
cells following the primary culture were
IL-2/15Rß+ (24). More recently, we
found that when the IL-2/15Rß+ and
IL-2/15Rß- cells generated in the primary
culture were sorted and placed in culture in IL-15, only the
IL-2/15Rß+ precursor cells gave rise to NK
cells (data not shown). It was not clear from this analysis, however,
which of the early acting cytokines was critical for expanding and/or
inducing the differentiation of an IL-15-responsive cell that
subsequently could give rise to NK1.1+ cells on
culture in IL-15.
Early experiments in which IL-6 was or was not included indicated that, whereas IL-6 did increase cell yields during the primary culture, the yield of NK cells was lower overall. We hypothesize that this effect was due to the ability of IL-6 to stimulate differentiation and proliferation of non-NK progenitors. Therefore, subsequent cytokine comparisons were performed using only IL-7, SCF, and flt3L.
Progenitors were cultured for 5 days with SCF alone, flt3L alone, IL-7
+ SCF, IL-7 + flt3L, or IL-7 + SCF + flt3L. To determine the relative
ability of these various cytokine combinations to expand and/or induce
the differentiation of an IL-15-responsive precursor, the cells were
then washed and replated in mIL-15 for an additional 6 days. The fold
expansion in mIL-15 on a per-cell basis was then calculated for each
condition (Table I
). Neither SCF nor
flt3L alone generated cells that responded efficiently to IL-15.
Although the addition of IL-7 to SCF did not appear to have an effect,
the addition of IL-7 to flt3L enhanced the ability of the generated
population to respond to IL-15 (p = 0.10,
Mann-Whitney test (27)). This observation suggests that
IL-7 may play an important role in the generation of IL-15-responsive
NK precursors, at least for cells grown in the absence of SCF. However,
cells cultured in IL-7 + flt3L were 2.5-fold better at responding to
IL-15 than were cells grown in IL-7 + SCF. This difference, while not
highly significant (p = 0.05, Mann-Whitney
test), also suggests an important role for flt3L in NK development. In
fact, culture in IL-7 + flt3L stimulated generation of a population of
cells on day 5 that responded more efficiently to IL-15 than those
populations generated by all the other cytokine combinations. Although
the addition of SCF to IL-7 + flt3L did not improve the ability of the
generated cells to respond to IL-15, it did improve the overall yield
of NK cells, largely because of the additional proliferation induced
during the primary culture before culture in IL-15 (data not shown).
Taken together, these data imply that IL-7 and flt3L specifically
expand or induce differentiation of IL-15-responsive NK precursors from
Lin-, c-kit+,
Sca2+ progenitors. SCF has an additive effect on
the overall expansion of this population of cells, but it does not
affect the frequency with which IL-15-responsive NK precursors are
generated.
|
It has been reported that flt3L-/- mice
lack NK cells, whereas IL-7-/- or
IL-7R-/- mice show minimal defects in NK
development (23, 28, 29). We chose, therefore, to focus
initially on the role of flt3L in NK development. The ability of the
Lin-, c-kit+,
Sca2+ population to respond to flt3L suggested
that this population may express the receptor for flt3L (flt3). A
four-color flow cytometry analysis revealed that 20–30% of
Lin-, c-kit+,
Sca2+ cells express flt3 (Fig. 1). A similar proportion of the total
Lin-, c-kit+
population also express flt3. Because of the greater ease with which
three-color as compared with four-color sorting can be performed, we
chose to analyze the NK potential of LB Lin-,
c-kit+, flt3+ vs
Lin-, c-kit+,
flt3- populations, which represent respectively
0.2% and 1.0–1.5% of whole bone marrow. Subsequent experiments in
which Lin-, c-kit+,
Sca2+, flt3+ or
flt3- populations were isolated, however,
yielded similar results (data not shown).
|
Lin-, c-kit+,
flt3+ and Lin-,
c-kit+, flt3- cells
were cultured in IL-7, SCF, and flt3L for 5 days followed by IL-15
alone for an additional 6 days. Significantly,
flt3+ progenitors gave rise to, on average,
3-fold more NK cells on a per-cell basis than did
flt3- progenitors (Fig. 2A). This was due in large
part to the fact that the population of NK precursors generated by
culture of flt3+ progenitors in IL-7, SCF, and
flt3L responded more efficiently to IL-15 on a per-cell basis to
generate NK1.1+ cells than did the population of
NK precursors generated in the same manner from
flt3- progenitors (Fig. 2B). To
control for differences in NK generation due to stimulation of the
flt3+ but not flt3-
progenitors with flt3L itself in the primary culture, we attempted to
culture both populations with IL-7 and SCF alone followed by culture in
IL-15. The flt3- population gave rise to similar
numbers of NK cells as when IL-7, SCF, and flt3L were utilized (data
not shown). Surprisingly, however, the flt3+
population failed in two separate experiments to survive with IL-7 and
SCF alone (data not shown). This suggests that somehow signals sent
through flt3 are critical for the survival of
Lin-, c-kit+,
flt3+ cells. SCF does, however, augment
proliferation of these flt3+ cells in the
presence of flt3L, because SCF + IL-7 + flt3L stimulated 2-fold more
proliferation of Lin-,
c-kit+, flt3+
cells than IL-7 and flt3L alone (data not shown).
|
Mature peripheral NK cells are NK1.1bright,
CD2+, and flt3- (Ref.
30 and J. Klem and V. Kumar, unpublished observations).
They are largely c-kit-, although a small
fraction is positive for this receptor (30).
Interestingly, not only did flt3+ progenitors
give rise to more NK cells, the cells derived from these progenitors
more closely resembled mature NK cells than did the cells derived from
flt3- progenitors. NK cells derived from
flt3- progenitors lysed the prototypical NK
target YAC-1 as well as or better than NK cells derived from
flt3+ progenitors (data not shown). In addition,
NK cells derived from both progenitors expressed NK1.1. However, the
level of expression of the NK1.1 receptor on NK cells generated from
the flt3- progenitors was much lower than that
on NK cells generated from the flt3+ cells (Fig. 3). Furthermore, the NK cells derived
from flt3- progenitors failed to express the
lymphoid marker CD2 and continued to express c-kit. A
significant fraction of the flt3+
progenitor-derived NK cells, on the other hand, did express CD2, and
the cells were NK1.1bright and
c-kit–/dim. Both populations of NK
cells were negative for flt3.
|
and data not shown).
Despite this, both populations were able to distinguish class
I+ targets from class I-
targets (data not shown), a property we had previously found in the
Ly49- NK cells derived in vitro from
Lin-, c-kit+,
Sca2+ progenitors (24). That is, the
in vitro-derived NK cells lysed the class I-deficient target RMA-S but
failed to lyse the class I+ target RMA. On the
basis of similar observations for KIR-negative human NK cells and fetal
liver-derived murine NK cells, it seems likely that the receptor
responsible for this activity in developing NK cells is CD94/NKG2
(26, 31). In support of this, CD94 can be detected in
lysates of NK cells derived from Lin-,
c-kit+, Sca2+
progenitors (Fig. 4) and can be detected
on the cell surface of in vitro-derived NK cells using Qa1b tetramers
(N. S. Williams, P. V. Sivakumar, and V. Kumar, unpublished
data). Efforts are currently underway in our laboratory to formally
identify the receptor responsible for this activity.
|
|
To determine whether the ability of flt3+
and flt3- progenitors to give rise to somewhat
distinct NK1.1+ cells in vitro could be
recapitulated in vivo under more physiologic conditions,
Lin-, c-kit+,
flt3+ or flt3- bone marrow
progenitors were sorted from (C57BL/6 x
DBA.2J)F1 mice (H-2b/d) and
injected into 175 rad-treated CB17.SCID mice
(H-2d). Semiallogeneic donor cells were utilized
so that their presence in the host could be detected with Abs to
H-2b MHC. The absence of T cells in the host
prevents rejection of the allogeneic marrow cells, and NK-mediated
rejection of (H-2b/d)F1
bone marrow cells by parental H-2d NK cells has
been shown to be insignificant (32). It should be noted
that the numbers of flt3+ and
flt3- progenitor cells injected per mouse were
30,000 and 120,000, respectively. The relative number of each
population utilized was calculated from the relative ratios of
flt3+ and flt3- cells in
normal adult marrow (i.e., 1:4). The mice were then sacrificed on day
27 or 36 after cell transfer. Data from these two time points were
pooled. Despite the fact that four times as many
flt3- cells were injected, the total number of
splenic NK cells generated in mice injected with
flt3+ cells was 2-fold higher than the number in
mice injected with flt3- cells. Thus, the
differences in the NK-generating activity in
flt3+ vs flt3- cells in
vitro could be reproduced in vivo as well. However, NK cells from both
sets of mice expressed a completely mature phenotype:
NK1.1bright, CD2+,
c-kit-. Additionally, near-normal
percentages of Ly49+ cells were detected (Table II). Thus, under more physiologic
conditions both types of progenitors give rise to NK cells with a
mature phenotype, although less efficiently from the
flt3- population.
|
Because both flt3+ and
flt3- progenitor cells gave rise to
Ly49+ NK cells in vivo, we reasoned that their
inability to generate Ly49+ cells in vitro could
be due to inadequacies of the culture system. It is known that NK
development in vivo is bone marrow dependent (33). We
hypothesized, therefore, that additional signals from bone
marrow-derived stromal cells may be necessary for complete NK
maturation, including Ly49 induction. To provide an environment
conducive to lymphoid development, we chose to utilize a stromal cell
line, OP9, derived from the calvaria of a B6C3F2-op/op mouse
deficient in the production of M-CSF (25). The haplotype
of this stromal cell line is H-2k (data not
shown). Lin-,
c-kit+, flt3+ and
Lin-, c-kit+,
flt3- progenitors from C57BL/6 mice, grown for 5
days in IL-7, SCF, and flt3L, were thus plated on confluent monolayers
of this OP9 stroma in the presence of human IL-15 or high-dose human
IL-2 (5000 IU/ml). IL-2 and IL-15 were used interchangeably in these
experiments, because we have noted that 5000 IU/ml human IL-2 gives
results similar to those obtained with 150 ng/ml human IL-15 (data not
shown). After 6 days of culture on stroma with IL-15 or IL-2, both
populations gave rise to NK1.1bright populations
that expressed a high density of CD2, although a small fraction of the
flt3--derived NK cells were still
CD2–/dim at this time point (Fig. 5). Thus, stroma is able to provide the
signals necessary to up-regulate CD2 on NK cells derived from
flt3- progenitors. However, using this protocol,
only the NK cells derived from the flt3+ but not
flt3- population were induced to express Ly49
molecules (Fig. 6
, A and
C). Cells reactive with the Ab 4D11, known to detect Ly49G2
(34), and with the Ab 5E6, known to detect the C57BL/6
forms of Ly49C and Ly49I (35), were present. In addition,
a subset of NK cells derived from flt3+
progenitors stained positive with an Ab, 1F8, recently generated in our
laboratory and known to detect Ly49C, Ly49I, and Ly49H but not Ly49G2
or Ly49D (T. George and M. Bennett, unpublished data). In some
experiments, such as the one shown in Fig. 6, there was some apparent
staining of NK1.1- cells with the 5E6 mAb.
Although there is a remote possibility that this represents true
expression of Ly49 molecules on NK1.1- cells, we
feel it more likely to be nonspecific sticking of the 5E6 mAb to
residual stromal cells in the culture. First and most importantly, if
it were true expression of Ly49C or Ly49I, one would expect to see a
similar staining pattern with the 1F8 mAb, which is also known to
detect Ly49C and Ly49I. Second, the 5E6+
NK1.1- populations are less pronounced when
cleaner preparations, having fewer residual stromal cells, are utilized
(data not shown).
|
E), indicating that at least two distinct Ly49 molecules
are being expressed. Some cells express only Ly49 molecules reactive
with 4D11, others express only Ly49 molecules reactive with 1F8, and
still others express both molecules. No Ly49A (JR9-318 positive) or
Ly49D (4E5 positive) molecules (36, 37) were detected at
this time point (data not shown). With longer culture on stromal cells,
very small numbers of JR9-318- and 4E5-reactive cells (<1.0% of
NK1.1+ cells) were noted; however, the fraction
of 4D11- and 1F8-reactive cells did not increase substantially (data
not shown). In the same experiments, NK cells that differentiated in
the absence of stroma failed to express Ly49 molecules (Fig. 6, B and D). Generation of Ly49+ NK1.1+ cells from Lin-, c-kit+ progenitors
Although coculture of Lin-,
c-kit+, flt3+
progenitors with OP9 stroma did yield NK1.1+,
Ly49+ cells, the frequency of
Ly49+ cells generated in vitro was low. Because
in vivo transfer of this population gave rise to near-normal numbers of
Ly49+ NK cells (Table II), we hypothesized that
interactions with other cells, in addition to stroma, may be important
for generation of NK1.1+,
Ly49+ cells. Therefore, we decided not to
separate flt3+ or flt3-
cells from the Lin-,
c-kit+ marrow population. Thus,
Lin-, c-kit+ cells
were cultured with IL-7, SCF, and flt3L for 5 days as usual, followed
by culture with IL-2 in the absence or presence of stromal cells.
Cultures were harvested at 3, 6, 10, or 13 days after exposure to IL-2,
and the frequency of NK1.1+ and
Ly49+ cells was determined (Table III). By day 10, the frequency of
NK1.1+ cells had reached close to 100% in the
presence or absence of stromal cells. As expected,
Ly49+ cells were generated only when the cells
were cultured with stroma. However, as compared with cultures initiated
with Lin-, c-kit+,
flt3+ cells, a substantially greater number of
Ly49+ cells was generated. As with the expression
of NK1.1, Ly49 molecules were first detected 3 days after culture with
IL-2 and stroma, and the numbers increased rapidly to peak at 10 days
postculture. Ly49 receptors reactive with four distinct Abs, including
4D11, 1F8, JR9-318, and 4E5, could be detected. The time course
analysis revealed that some Ly49 molecules could be detected earlier
than others. Thus, Ly49 receptors recognized by the mAbs 4D11 and 1F8
were noted as early as 3 days after culture on stroma, whereas Ly49
receptors detected by JR9-318 and 4E5 were not noted until day 6. While
the frequency of 4D11+ and
1F8+ cells at later time points approached that
seen on peripheral C57BL/6 NK cells, the frequency of
4E5+ and JR9-318+ cells did
not. The reason for this discrepancy is not yet clear, but possible
explanations are explored in Discussion.
|
).
Mature peripheral Ly49+ NK cells readily expand
in IL-2 without any stroma; hence, there is no reason to believe that
stroma is necessary for survival of any contaminating mature NK cells.
In support of this, when Thy1.2+ progenitor cells
were deliberately contaminated with 0.5% or 2.5% mature
Thy1.1+NK1.1+ bone marrow
cells, Thy1.1+, Ly49+ NK
cells were detected in the presence or absence of stroma (data not
shown). Second, the pattern of Ly49 expression is selective. For
example, although 40–60% of NK cells in the periphery express Ly49G2
and an approximately equivalent percentage express Ly49D, 15- to
30-fold more Ly49G2+ than
Ly49D+ cells are noted in our in vitro system
(Table III). Third, under the standard assay conditions,
Ly49+ cells are not generated from
flt3- progenitors (Fig. 6C). There is
no reason to believe that NK1.1+ contaminants
would occur selectively in cultures of flt3+
progenitor cells, especially as mature NK cells are
flt3-. |
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
). In the first step, multipotent stem
cells containing NK cell progenitors are driven to become
IL-2/15-responsive NK precursors under the influence of several early
acting cytokines, especially flt3L. We hypothesize that this effect is
due to induction of the IL-2/15Rß (CD122) chain on a fraction of the
progenitor cells. This step can occur in the absence of stroma (at
least in vitro). In the second step, the NK precursors, so generated,
give rise to mature Ly49-expressing NK cells on being exposed to
IL-2/15. In this step, the induction of Ly49 molecules but not NK1.1 is
strictly dependent on the presence of stromal cells. Using this
two-part culture system, we have investigated the mechanisms
responsible for NK cell differentiation.
|
(39) and
IL-15 (40). Likewise, the importance of flt3L in NK development has been suggested by several recent reports. First, NK1.1bright cells are missing in flt3L-/- mice (23). Second, administration of recombinant flt3L induces a striking increase in the absolute number of mature nonactivated NK cells within various tissues (21, 22). This increase in NK cell numbers may provide a possible additional explanation for the potent antitumor effect of in vivo-administered flt3L (41, 42, 43), which was previously attributed solely to an increase in dendritic cells (44).
Although such observations suggest a significant potential for flt3L in
therapy of both cancer and viral infection through expansion of NK
cells, they do not address the mechanism by which this occurs. The data
presented in this paper show that flt3L acts by expanding and/or
inducing the differentiation of a progenitor population into
IL-15-responsive precursors capable of giving rise to NK cells on in
vitro culture in IL-15 (Fig. 7). The progenitor on which flt3L exerts
its effect is identified in the mouse as Lin-,
c-kit+, flt3+.
Somewhat surprisingly, in light of data from the
flt3L-/- mouse, the corresponding
Lin-, c-kit+,
flt3- population also gave rise to NK, albeit
less efficiently than the flt3+ population (Fig. 2). An explanation for this discrepancy may lie in the phenotype of NK
cells derived from each progenitor. NK cells derived from
flt3- progenitors showed a less mature
NK1.1dim, CD2-,
c-kit+ phenotype than NK cells derived from
flt3+ progenitors, which were
NK1.1+, CD2+/-,
c-kit-. It is possible that in
flt3L-/- mice, Lin-,
c-kit+, flt3- cells
inefficiently give rise to NK cells. As these NK cells express quite
dim levels of NK1.1, it could appear that
flt3L-/- mice lack
NK1.1bright cells as previously reported
(23). It would be interesting to further phenotype any NK
cells that develop in these mice for expression of CD2 and
c-kit.
It should be noted, however, that immature NK1.1dim, CD2-, c-kit+ NK cells do not exist in the spleens of normal mice in vivo. Such NK cells may represent a developmental artifact that occurred because in vitro the flt3- progenitors did not progress through a normal developmental pathway important for the up-regulation of NK1.1 and CD2 and down-regulation of c-kit. We therefore assessed the NK potential of flt3+ and flt3- progenitors in vivo. Significantly, both flt3+ and flt3- progenitors gave rise to NK cells with a mature CD2+, c-kit+, NK1.1bright phenotype, albeit less efficiently from the flt3- population. Thus, signals necessary to up-regulate NK1.1 and CD2 and to down-regulate c-kit were provided to the flt3- progenitors in vivo. One possible explanation for the ability of flt3- cells to give rise to mature NK cells in vivo but not in vitro is that flt3+ and flt3- cells are in the same lineage pathway, with flt3- cells being more immature than flt3+ cells. Thus, in vivo but not in vitro, flt3- cells receive sufficient signals to complete their normal developmental pathway. In support of this, preliminary data have suggested that NK1.1+ cells are detectable in the peripheral blood of mice injected with flt3+ progenitors earlier than they are detectable in mice injected with flt3- progenitors. Furthermore, flt3- cells have a greater capacity than flt3+ cells to self-renew following serial transfers into irradiated recipients, again suggesting that flt3- cells are more immature than flt3+ cells (N. S. Williams and V. Kumar, unpublished data). Because flt3L-/- mice lack NK1.1bright cells, it is possible that flt3- progenitors normally must differentiate into flt3+ cells before entering the NK differentiation pathway and that interaction of flt3L with flt3+ cells is essential for the development of mature NK cells. In support of this, we found in vitro that flt3L was absolutely essential for the survival of Lin-, c-kit+, flt3+ cells. These cells failed to survive when cultured in IL-7 and SCF alone. Together, these data reinforce the hypothesis that flt3L plays a critical role in NK cell differentiation and suggest an explanation for the NK cell-deficient phenotype of flt3L-/- mice.
Although we have not pursued the observation here, it was interesting
to note that IL-7 also seemed to play an important role in the
development of IL-15-responsive precursors from
Sca2+ progenitors. This is in contrast to the
observation that IL-7R-/- mice show no
defects in NK cells. It is possible that in vivo, signaling through
this receptor can be compensated for by other cytokines not present in
our in vitro system. On the other hand, IL-7-/-
mice were found to have a 3-fold decrease in the absolute number of NK
cells. It is therefore a formal possibility that IL-7 can exert its
effects, at least on NK cells, independently of the IL-7R receptor.
We are currently examining Lin-,
c-kit+, flt3+ NK
progenitors for the presence of the IL-7R-chain in an effort to
better understand the role of IL-7 in NK cell development.
In addition to allowing maturation of NK cells from both flt3+ and flt3- populations to a NK1.1bright, CD2+, c-kit- stage, in vivo transplantation also allowed for expression of Ly49 molecules on these NK cells. We show here for the first time that this environment can be recreated in vitro by growing flt3+ progenitors in IL-7, SCF, and flt3L and then placing them on a confluent monolayer of the bone marrow stromal cell line, OP9, in the presence of IL-15 or IL-2. The NK cells generated expressed Ly49 molecules reactive with three distinct Ly49 Abs, 4D11 (Ly49G2), 1F8 (Ly49C/I/H), and 5E6 (Ly49C/I). It is tempting to speculate, on the basis of the larger percentage of cells reactive with 1F8 vs 5E6, that Ly49H may be expressed on these cells, but definitive evidence of this awaits more extensive analysis. It is clear, however, that at least three distinct Ly49+ subsets are generated, because two-color analysis of 4D11 vs 1F8 revealed the presence of two single-positive populations in addition to a double-positive population.
Because the frequency of Ly49+ cells was lower
than that observed on mature peripheral NK cells, we speculated that
additional cellular interactions may be necessary for inducing the
normal frequency of Ly49 molecules. Therefore, we utilized a less
refined Lin-, c-kit+
progenitor population to initiate the cultures. Interestingly, this
strategy did yield a higher frequency of Ly49+
cells for reasons we do not yet understand. Using this protocol, Ly49
molecules reactive with four distinctive Abs were detectable. In
addition, the time course analysis performed revealed that Ly49
molecules reactive with the mAbs 4D11 and 1F8 appeared at least 3 days
earlier than molecules reactive with the mAbs 4E5 and JR9-318. Such
findings mirror observations made of the onset of Ly49 expression on NK
cells isolated from neonatal spleens. In these studies, expression of
Ly49G2, as detected with 4D11, preceded expression of Ly49A (45, 46). It is tempting to speculate that subsets of Ly49 molecules
may be expressed with different kinetics, although definitive evidence
of such regulation awaits further analysis of individual Ly49 gene
expression in developing NK cells. Finally, it is significant that at
the earliest time points examined, individual
Ly49+ NK cells could express more than one Ly49
molecule (Fig. 6E and data not shown). Several models have
been put forth to explain regulation of the repertoire of Ly49
molecules in NK cells. In one model, NK cells stochastically express
individual Ly49 molecules one at a time and continue to do so until an
interaction of sufficient magnitude between an expressed Ly49 and
self-class I MHC occurs (47). In a second model,
individual NK cells begin to express one, two, or multiple Ly49
molecules at one time. Those NK cells that express at least one
self-MHC-reactive Ly49 molecule but not so many self-reactive receptors
as to be insensitive to virus- or tumor-induced changes in class I are
then selected for expansion (47). Our data at this point
do not allow us to distinguish between these two hypotheses. Neither
model seems to adequately explain our observations, because we observe
certain Ly49 molecules being expressed in concert (i.e., 4D11- and
1F8-reactive molecules), whereas others require additional time before
detection (i.e., 4E5- and JR9-318-reactive molecules). Furthermore, the
majority of Ly49+ NK cells fail to express Ly49A,
the Ly49 molecule known to bind the H-2k MHC of
the OP9 stroma (48). Both models would predict the need
for at least one self-reactive Ly49 receptor for survival and/or
expansion. In this regard, it is interesting to note that both Ly49C/I
(T. George and M. Bennett, unpublished data) and Ly49G2 (N. S.
Williams and V. Kumar, unpublished data) can receive weak negative
signals from H-2k. If there are true kinetic
differences in the expression pattern of the Ly49 molecules, with
Ly49C/I and Ly49G2 being expressed before Ly49A and Ly49D, it is
possible that the weak signals sent by H-2k
through Ly49G2 or Ly49C/I are sufficient in many cases to shut off
further Ly49 expression. In this regard it is important to note that
the progenitors remain in close proximity to the stromal cells during
their entire development in vitro. Weak signals may be more effective
here than in vivo, where developing NK cells may interact only briefly
with stromal cells. These and other issues relating to the development
of the Ly49 repertoire can now be addressed by utilizing progenitor and
stromal cells of different MHC types and by interrupting contact of the
developing NK cells with the stroma in this in vitro culture
system.
In conclusion, our experiments have demonstrated a role for flt3L in expanding and/or inducing the differentiation of Lin-, c-kit+, flt3+ progenitors into cells capable of responding to IL-15 to give rise to NK cells. The steps of NK differentiation from the flt3+ progenitor can be largely mimicked in vitro by culture with IL-7, SCF, and flt3L followed by culture with IL-15. However, induction of Ly49 is critically dependent on additional signals provided by bone marrow-derived stromal cells.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
2 Address correspondence and reprint requests to Dr. Noelle S. Williams, Department of Pathology, University of Texas Southwestern Medical Center, 5323 Harry Hines Boulevard, Dallas, TX 75235-9072. E-mail address:
3 Abbreviations used in this paper: KAR, killer activatory receptor; KIR, killer inhibitory receptor; Lin, lineage marker; SCF, stem cell factor; flt3L, flt3 ligand; LB, lymphocyte and blast sized; mIL, murine IL.
Received for publication March 12, 1999. Accepted for publication June 28, 1999.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
is essential for the development of 
+ T cells, but not for natural killer cells. J. Exp. Med. 184:289. T cell development and early thymocyte maturation in IL-7-/- mice. J. Immunol. 157:2366.[Abstract]
ß TCR population of i-IEL with an extrathymic maturation. Int. Immunol. 4:699.This article has been cited by other articles:
|
|
 |
|
  I.-K. Park, C. Giovenzana, T. L. Hughes, J. Yu, R. Trotta, and M. A. Caligiuri The Axl/Gas6 pathway is required for optimal cytokine signaling during human natural killer cell development Blood, March 12, 2009; 113(11): 2470 - 2477. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Cheng, H. N. Charoudeh, P. Brodin, Y. Tang, T. Lakshmikanth, P. Hoglund, S. E. W. Jacobsen, and E. Sitnicka Distinct and Overlapping Patterns of Cytokine Regulation of Thymic and Bone Marrow-Derived NK Cell Development J. Immunol., February 1, 2009; 182(3): 1460 - 1468. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. W. Chung, M. S. Kim, Z.-H. Piao, M. Jeong, S. R. Yoon, N. Shin, S. Y. Kim, E. S. Hwang, Y. Yang, Y. H. Lee, et al. Osteopontin Promotes the Development of Natural Killer Cells from Hematopoietic Stem Cells Stem Cells, August 1, 2008; 26(8): 2114 - 2123. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S.-i. Ohno, T. Sato, K. Kohu, K. Takeda, K. Okumura, M. Satake, and S. Habu Runx proteins are involved in regulation of CD122, Ly49 family and IFN-{gamma} expression during NK cell differentiation Int. Immunol., January 1, 2008; 20(1): 71 - 79. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Giroux, E. Yurchenko, J. St.-Pierre, C. A. Piccirillo, and C. Perreault T Regulatory Cells Control Numbers of NK Cells and CD8{alpha}+ Immature Dendritic Cells in the Lymph Node Paracortex J. Immunol., October 1, 2007; 179(7): 4492 - 4502. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. Ceredig, M. Rauch, G. Balciunaite, and A. G. Rolink Increasing Flt3L availability alters composition of a novel bone marrow lymphoid progenitor compartment Blood, August 15, 2006; 108(4): 1216 - 1222. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Iizuka, T. Kaifu, A. Nakamura, M. Obinata, and T. Takai Establishment and Functional Characterization of Novel Natural Killer Cell Lines Derived from a Temperature-Sensitive SV40 Large T Antigen Transgenic Mouse J. Biochem., August 1, 2006; 140(2): 255 - 265. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
F. Stevenaert, K. Van Beneden, V. De Colvenaer, A. S. Franki, V. Debacker, T. Boterberg, D. Deforce, K. Pfeffer, J. Plum, D. Elewaut, et al. Ly49 and CD94/NKG2 receptor acquisition by NK cells does not require lymphotoxin-{beta} receptor expression Blood, August 1, 2005; 106(3): 956 - 962. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. Takeda, E. Cretney, Y. Hayakawa, T. Ota, H. Akiba, K. Ogasawara, H. Yagita, K. Kinoshita, K. Okumura, and M. J. Smyth TRAIL identifies immature natural killer cells in newborn mice and adult mouse liver Blood, March 1, 2005; 105(5): 2082 - 2089. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. S. Choi, V. S. Chhabra, Q. H. Nguyen, B. J. Ank, E. R. Stiehm, and R. L. Roberts Interleukin-15 Enhances Cytotoxicity, Receptor Expression, and Expansion of Neonatal Natural Killer Cells in Long-Term Culture Clin. Vaccine Immunol., September 1, 2004; 11(5): 879 - 888. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. I. Samson, S. Memet, C. A. J. Vosshenrich, F. Colucci, O. Richard, D. Ndiaye, A. Israel, and J. P. Di Santo Combined deficiency in I{kappa}B{alpha} and I{kappa}B{epsilon} reveals a critical window of NF-{kappa}B activity in natural killer cell differentiation Blood, June 15, 2004; 103(12): 4573 - 4580. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. Kawamura, R. Koka, A. Ma, and V. Kumar Differential Roles for IL-15R {alpha}-Chain in NK Cell Development and Ly-49 Induction J. Immunol., November 15, 2003; 171(10): 5085 - 5090. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
B. T. Wilhelm, J.-R. Landry, F. Takei, and D. L. Mager Transcriptional Control of Murine CD94 Gene: Differential Usage of Dual Promoters by Lymphoid Cell Types J. Immunol., October 15, 2003; 171(8): 4219 - 4226. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
V. Ioannidis, B. Kunz, D. M. Tanamachi, L. Scarpellino, and W. Held Initiation and Limitation of Ly-49A NK Cell Receptor Acquisition by T Cell Factor-1 J. Immunol., July 15, 2003; 171(2): 769 - 775. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 F. Stevenaert, K. Van Beneden, A. De Creus, V. Debacker, J. Plum, and G. Leclercq Ly49E expression points toward overlapping, but distinct, natural killer (NK) cell differentiation kinetics and potential of fetal versus adult lymphoid progenitors J. Leukoc. Biol., June 1, 2003; 73(6): 731 - 738. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. Koka, P. R. Burkett, M. Chien, S. Chai, F. Chan, J. P. Lodolce, D. L. Boone, and A. Ma Interleukin (IL)-15R{alpha}-deficient Natural Killer Cells Survive in Normal but Not IL-15R{alpha}-deficient Mice J. Exp. Med., April 21, 2003; 197(8): 977 - 984. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. Kouro, V. Kumar, and P. W. Kincade Relationships between early B- and NK-lineage lymphocyte precursors in bone marrow Blood, November 15, 2002; 100(10): 3672 - 3680. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Minagawa, H. Watanabe, C. Miyaji, K. Tomiyama, H. Shimura, A. Ito, M. Ito, J. Domen, I. L. Weissman, and K. Kawai Enforced Expression of Bcl-2 Restores the Number of NK Cells, But Does Not Rescue the Impaired Development of NKT Cells or Intraepithelial Lymphocytes, in IL-2/IL-15 Receptor {beta}-Chain-Deficient Mice J. Immunol., October 15, 2002; 169(8): 4153 - 4160. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Sivori, M. Falco, E. Marcenaro, S. Parolini, R. Biassoni, C. Bottino, L. Moretta, and A. Moretta Early expression of triggering receptors and regulatory role of 2B4 in human natural killer cell precursors undergoing in vitro differentiation PNAS, April 2, 2002; 99(7): 4526 - 4531. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
O. Shimozato, J. R. Ortaldo, K. L. Komschlies, and H. A. Young Impaired NK Cell Development in an IFN-{gamma} Transgenic Mouse: Aberrantly Expressed IFN-{gamma} Enhances Hematopoietic Stem Cell Apoptosis and Affects NK Cell Differentiation J. Immunol., February 15, 2002; 168(4): 1746 - 1752. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
B. Kunz and W. Held Positive and Negative Roles of the Trans-Acting T Cell Factor-1 for the Acquisition of Distinct Ly-49 MHC Class I Receptors by NK Cells J. Immunol., May 15, 2001; 166(10): 6181 - 6187. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. B. Pawlowska, S. Hashino, H. McKenna, B. J. Weigel, P. A. Taylor, and B. R. Blazar In vitro tumor-pulsed or in vivo Flt3 ligand-generated dendritic cells provide protection against acute myelogenous leukemia in nontransplanted or syngeneic bone marrow-transplanted mice Blood, March 1, 2001; 97(5): 1474 - 1482. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Q. Wu, Y. Sun, J. Wang, X. Lin, Y. Wang, L. E. Pegg, A. Futterer, K. Pfeffer, and Y.-X. Fu Signal Via Lymphotoxin-{{beta}}R on Bone Marrow Stromal Cells Is Required for an Early Checkpoint of NK Cell Development J. Immunol., February 1, 2001; 166(3): 1684 - 1689. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. A. Fehniger and M. A. Caligiuri Interleukin 15: biology and relevance to human disease Blood, January 1, 2001; 97(1): 14 - 32. [Full Text] [PDF]
|
|
|
|
|
|
J. R. Ortaldo, R. Winkler-Pickett, and G. Wiegand Activating Ly-49D NK receptors: expression and function in relation to ontogeny and Ly-49 inhibitor receptors J. Leukoc. Biol., November 1, 2000; 68(5): 748 - 756. [Abstract] [Full Text]
|
|
|
|
|
|
M. Lohoff, G. S. Duncan, D. Ferrick, H.-W. Mittrucker, S. Bischof, S. Prechtl, M. Rollinghoff, E. Schmitt, A. Pahl, and T. W. Mak Deficiency in the Transcription Factor Interferon Regulatory Factor (IRF)-2 Leads to Severely Compromised Development of Natural Killer and T Helper Type 1 Cells J. Exp. Med., July 31, 2000; 192(3): 325 - 336. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
F. Colucci and J. P. Di Santo The receptor tyrosine kinase c-kit provides a critical signal for survival, expansion, and maturation of mouse natural killer cells Blood, February 1, 2000; 95(3): 984 - 991. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Sivori, M. Falco, E. Marcenaro, S. Parolini, R. Biassoni, C. Bottino, L. Moretta, and A. Moretta Early expression of triggering receptors and regulatory role of 2B4 in human natural killer cell precursors undergoing in vitro differentiation PNAS, April 2, 2002; 99(7): 4526 - 4531. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
