CG&D
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
Cancer Research Clinical Cancer Research
Cancer Epidemiology Biomarkers & Prevention Molecular Cancer Therapeutics
Molecular Cancer Research Cell Growth & Differentiation

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Delfino, D. V.
Right arrow Articles by Riccardi, C.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Delfino, D. V.
Right arrow Articles by Riccardi, C.
Cell Growth & Differentiation Vol. 12, 51-60, January 2001
© 2001 American Association for Cancer Research


Articles

Differentiation of Ly49s-positive or -negative Natural Killer Cells Is Inhibited by Anti-H-2b Monoclonal Antibodies Acting at the Level of Bone Marrow Progenitors from B6 Mice1

Domenico V. Delfino2, Margarita Salcedo, Barbara Di Marco, Emira Ayroldi, Giuseppe Nocentini, Stefano Bruscoli, Luigi Brunetti, Hans-Gustaf Ljunggren and Carlo Riccardi

Department of Clinical and Experimental Medicine, Section of Pharmacology, University of Perugia, 06100 Perugia, Italy [D. V. D., B. D. M., E. A., G. N., S. B., C. R.]; Karolinska Institute MTC, Stockholm, Sweden [M. S., H-G. L.]; and Department Scienza del Farmaco, University G. D’Annunzio, 66100 Chieti, Italy [L. B.]

Abstract

To investigate the role of MHC class I on in vitro differentiation of natural killer (NK) cells, a CD44low/-CD2-classIlow population was isolated from mouse bone marrow. This population, which lacked expression of NK-1.1, Ly49A, Ly49C/I, and Ly49G, generated populations of NK-1.1+ NK cells expressing Ly49A, Ly49C/I, or Ly49G when cocultured for 13 days with syngeneic supportive stromal cells in the presence of interleukin 2. Ly49A and Ly49C/I were absent on the progeny of progenitors tested after 7 days of culture but were expressed as a late event together with low-level expression of NK-1.1, from day 8 of culture. The addition of anti-H-2b monoclonal antibody to cultures at day 0 inhibited proliferation of progenitors supported by either syngeneic, allogeneic, or H-2b-deficient stromal cells, thus suggesting that the effect was not exerted on stromal cells. Additional analyses demonstrated that class Ilow progenitors generated class I+ cells on which the anti-H-2b monoclonal antibody exerted its inhibitory effect.

Introduction

The discovery of Ly49 family molecules in mouse and killer inhibitory receptors in humans has highlighted the importance of MHC class I recognition in both the killing of allogeneic and altered self cells and tolerance toward normal syngeneic cells by NK3 effector cells (1, 2, 3, 4) . It is not known in detail how and when NK cells acquire a repertoire of Ly49 molecules, allowing them to discriminate among different target cells. Specifically, the stage at which repertoire is acquired during the differentiation process from progenitors to mature NK cells and whether MHC class I molecules play a role in this process, as has been shown in thymic differentiation of CD8+ T cells, are not known.

Until now, the development of the NK repertoire has largely been limited to in vivo studies. In the course of these studies, Sykes et al. (5) found that expression of Ly49 molecules on NK cells is determined by interaction with class I molecules expressed on BM-derived radiosensitive cells and, to a lesser extent, by class I antigens expressed on radioresistant host elements. The importance of class I molecules in the creation of the NK repertoire was also highlighted by studies of mutant mice lacking ß2-microglobulin or the transporter protein Tap-1. Tap-1-/- mice, which express class I molecules at very low levels, generate a normal number of NK cells with an altered repertoire as compared with their littermates expressing normal levels of class I antigens (6) . Interestingly, in the same Tap-1-/- mice, the absence of class I antigens determines an almost complete lack of CD8+ T cells, indicating that the development of both CD8+ T and NK cells is influenced by class I molecules, albeit through different mechanisms. Of particular interest, in the work of Johansson et al. (7) , there was no evidence of deletion of potentially autoreactive NK cells, and NK cell tolerance development seemed rather to be the result of an adaptation process (7 , 8) due to the presence of cells in the environment in which differentiation of NK cells occurs. However, in these in vivo studies, it was not possible to assess whether the influence of class I molecules and Ly49 expression is exerted at a specific stage of the differentiation pathway of NK cells, e.g., on NK cell progenitors, or on their mature progeny. Dissection of these issues would greatly benefit from isolation of NK progenitors able to give rise to a progeny expressing Ly49s in in vitro cultures, and a recent report indicates that stromal cells are critical for the expression of Ly49 molecules (9) in differentiation of progenitors from adult BM, whereas in fetal NK cells, expression of Ly49E has been detected only at the mRNA level (10) .

In the present study, we isolated a CD44low/-CD2-class Ilow BM population from B6 mice (H-2b haplotype) that generated Ly49A-, Ly49C/I-, or Ly49G-positive and -negative NK cells after culture with IL-2 and irradiated BM stromal cells. The expression of Ly49 molecules was a late event in the differentiation process, and treatment with anti-H-2b class I mAb inhibited the ability of the progenitor population to generate NK cells by acting on progenitors.

Results

Characterization of the NK Precursor Population.
In our previous work, a murine BM subpopulation enriched for NK precursors was isolated by a cell sorter and characterized as CD44low/-TCR-NK-1.1-. After stimulation with IL-2 and cocultivation on a discontinuous layer of supportive stromal cells for 2 weeks, cells in this population were able to give rise to a virtually pure NK-1.1+TCR- cell population with the ability to kill Yac-1 cells. The frequency of the NK cell precursors was calculated by limiting dilution analysis to be around 1:500 (11) . To obtain a similar population by using magnetic bead depletion, mAbs to CD44 and CD2 were used to obtain a CD44low/-CD2- subpopulation. Fig. 1Citation shows a flow cytometry analysis of BM cells before and after the magnetic bead depletion procedure: Fig.1ACitation shows the isotypic control for CD44 and CD2; Fig. 1BCitation shows unfractionated BM stained for CD44 and CD2; and Fig. 1CCitation shows the BM after depletion of CD44highCD2+ cells. Moreover, in Fig. 1, B and CCitation , boxes formed by dashed lines indicate how we discriminated CD44high from CD44low/- cells, and the box formed by solid lines indicates the CD44low/-CD2- population. Utilization of these markers allowed isolation of a NK cell precursor-containing BM population with no contaminating mature NK cells or other lineage-positive cells. This is because CD44 is expressed at high intensity on most BM cells, including the majority of T cells and all mature NK cells (11 , 12) , and CD2 is also expressed on T, NK, B, and pre-B cells (13 , 14) . In addition, MHC class I molecules, also expressed on mature NK cells, was evaluated in the attempt to identify and enrich a precursor population. Because a variable percentage of CD44low/-CD2- cells were class Ilow, they were depleted of the class Ihigh cells (and were also Fc receptor negative as determined using the 2.4G2 anti-Fc receptor mAb) for a subsequent trial.



View larger version (17K):
[in this window]
[in a new window]
 
Fig. 1. Phenotypic characterization of CD44low/-CD2-class Ilow BM cell population. A, isotypic controls for CD44 and CD2 mAbs in unfractionated fresh BM; B, flow cytometry two-color analysis (CD44 versus CD2) of unfractionated BM; C, the CD44low/-CD2- cell population after enrichment. In B and C, boxes formed by dashed lines indicate the CD44high or the CD44low/- cells, whereas the box formed by solid lines indicates the CD44low/-CD2- cell population.

 
As a result of IL-2 stimulation, mature NK cells can be derived not only from proliferation and differentiation of cultured NK cell precursors but also from expansion of mature NK cells (15) ; therefore, to assay for NK cell precursor differentiation, the test population must be free of mature NK cells. The fact that all three mAbs used were directed to surface molecules expressed on NK cells and that they were used in three sequential steps to isolate the CD44low/-CD2-class Ilow cells (the purity at each step was >=95%) should allow the complete elimination of any residual mature NK cells, and this was confirmed by flow cytometry analysis. In Fig. 2, A–CCitation , flow cytometry stainings of unfractionated BM cells with different NK cell markers are shown. In Fig. 2Citation , A—C, the two-color staining shows the NK-1.1+CD44high, DX5+CD44high, or Ly49G mature NK cells, respectively, present in the unfractionated BM. In Fig. 2, D–FCitation , the flow cytometry analysis on the test population obtained after BM depletion is shown. NK-1.1+, DX5+, and Ly49G+ mature NK cells completely disappeared after the depletion steps (Fig. 1, D–FCitation , respectively). However, as an additional test, we took advantage of a feature discriminating NK cells derived from NK cell precursors and mature NK cells. Mature NK cells proliferate after stimulation with IL-2 alone, whereas precursors require not only IL-2 but also the presence of BM stromal cells (9 , 11 , 12 , 16) . In support of this observation, a recent study (9) showed that when a NK progenitor population was deliberately contaminated with mature BM NK cells, these latter cells were detected after stimulation with IL-2 in the presence or absence of stroma, thus demonstrating that proliferation of mature NK cells is not influenced by stromal cells. As shown in Fig. 3Citation , no cells were detected at the end of the culture period when the CD44low/-CD2-class Ilow subpopulation was cultured with only IL-2, without the support of stromal cells. When the same cells were cultured with IL-2 on irradiated supportive stromal cells, the cell number increased greatly starting from day 0 and reached a plateau from day 7 to day 13 (Fig. 3)Citation . These experiments strongly suggest that no mature NK cells were present in the isolated BM subpopulation because they were not detected even after stimulation with high doses of IL-2 that would have induced their expansion.



View larger version (37K):
[in this window]
[in a new window]
 
Fig. 2. Flow cytometry analyses of mature NK cells in the BM before and after magnetic bead depletion. A, NK-1.1 versus CD44. NK-1.1+CD44high NK cells are indicated by the arrow. B, DX5+CD44high NK cells are indicated by the arrow. C, histogram of Ly49G2+ NK cells. In A–C, numbers represent the percentage of positive cells. D—F, flow cytometry analyses of the putative progenitor population after magnetic bead depletion: D, NK-1.1 versus CD44; E, DX5 versus CD44; and F, Ly49G2. Dot plots of D and E represent 30,000 events. In D–F, no positive cells were detected after subtraction of the isotypic control background. The dot plots are representative of three independent experiments.

 


View larger version (13K):
[in this window]
[in a new window]
 
Fig. 3. Cell number time course. •, CD44low/-CD2-class Ilow BM cells stimulated by IL-2 and cocultured with supportive stromal cells; {circ}, CD44low/-CD2-class Ilow BM cells stimulated by IL-2 and cultured without supportive stromal cells. The results from four independent experiments are expressed as the mean ± 1 SE.

 
Flow Cytometric Analysis of the Cells Generated by IL-2 on Stromal Cells.
A two-color (NK-1.1 versus CD3) flow cytometry analysis was performed on day 3, day 7, and day 13 to ascertain whether the generated cells were NK cells. The flow cytometry dot plots shown in Fig. 4Citation indicate that at day 3, almost all cells were negative for both markers with very few (1.3 ± 0.2%) NK-1.1lowCD3- cells. The analysis at day 7 shows that two distinct cell populations were generated: (a) a small NK-1.1+CD3- population (13 ± 11.3%); and (b) a larger NK-1.1-CD3- population (81.2 ± 12.4%). At day 13, virtually all cells (91.9 ± 3.4%) were NK-1.1+CD3- cells. The fact that few contaminating CD3+ cells [Fig. 3Citation , day 13 panel, 5.6 ± 2.4% of CD3+NK-1.1+ cells (top right quadrant) and 1.8 ± 1.2% of CD3+NK-1.1- cells (bottom right quadrant)] were present is a further indirect demonstration that depletion of mature NK cells was effective. Because T cells have a higher IL-2-dependent proliferation rate in the unfractionated BM as compared with mature NK cells due to the expression of high-affinity IL-2 receptor, are about double in number (8–10% versus 3–4%, respectively), in the unfractionated BM and are only about 7% at day 13 after the depletion steps, residual mature NK cells will almost certainly be absent or, if present, will be much less than the residual 7% T cells and thus will not substantially contribute to the final NK cell number. These results suggest that the CD44low/-CD2-class Ilow BM population contains NK cell precursors that gave rise (day 13 dot plot) to a virtually pure population of NK cells.



View larger version (30K):
[in this window]
[in a new window]
 
Fig. 4. Phenotypic characterization of cells generated in culture. Two-color flow cytometry analyses (NK-1.1 versus CD3) of cells generated 3 (DAY-3), 7 (DAY-7), or 13 days (DAY-13) after coculture of CD44low/-CD2-class Ilow BM cells with stromal cells and stimulation with IL-2 (bottom panels). Top panels represent the isotypic controls for anti-NK-1.1 and anti-CD3 mAbs on day 3, day 7, and day 13 cells. Numbers represent the percentage of positive cells ± 1 SE. The results are representative of three independent experiments.

 
Expression of Ly49 Molecules.
NK cells that were generated at day 13 from the CD44low/-CD2-class Ilow precursors were analyzed by flow cytometry to examine the expression of Ly49 molecules. mAbs to Ly49A, Ly49C/I, or Ly49G were used, and, as shown in Fig. 5Citation , 10–15% of the NK cells were Ly49A+, 40–50% were Ly49C/I+, and 48 ± 5.6% were Ly49G+. To study at which point during the differentiation pathway Ly49 molecules were expressed, the two cell populations present at day 7 were analyzed. Fig. 6Citation shows that day 7 NK-1.1+ cells express Ly49A (middle panel; top right quadrant) and Ly49C/I (bottom panel; top right quadrant), whereas they were not expressed on day 7 NK-1.1- cells (top panel, bottom right quadrant). The same analysis performed 1 day later (day 8), compared with the day 7 analysis, showed an increased relative number of cells expressing Ly49A [8 ± 0.6% versus 5 ± 1.1% (middle panels, top right quadrants) and 5 ± 0.8% versus 1 ± 0.2% (middle panels, bottom right quadrants)] or Ly49C/I [10.1 ± 1.4% versus 5 ± 0.3% (bottom panels, top right quadrants) and 10.6 ± 3.4% versus 1.8 ± 0.6% (bottom panels, bottom right quadrants)]. MFI values show that although more numerous, cells from day 8 expressed Ly49A and Ly49C/I at low to intermediate levels (as demonstrated by the decrease of MFI values reported in Table 1Citation ) compared with day 7 cells. This suggests that the day 7 NK-1.1- cells were progenitors which, after 1 more day of culture, started to express both NK-1.1 and Ly49 molecules not at the high levels of fully mature NK cells but at low to intermediate levels and, consequently, that expression of Ly49 molecules was a late event in the differentiation pathway of NK cells. The day 7 analysis in Fig. 4Citation shows that a minor population of NK cells was already present (possibly derived from the few NK-1.1low cells detected at day 3), but the large expansion of NK cell number was due primarily to the NK-1.1-CD3- cell population predominant at day 7 and that this double-negative population expressed mature NK cell markers starting from day 8. This further confirms that the generation of NK cells observed here was sustained by proliferation and subsequent differentiation of NK-1.1- progenitors and not by mature NK-1.1+ cells contaminating the original BM progenitor population.



View larger version (19K):
[in this window]
[in a new window]
 
Fig. 5. Ly49A, Ly49C/I, and Ly49G2 expression of day 13-generated cells. Representative one-color flow cytometry analysis of cells generated after 13 days of coculture of CD44low/-CD2-class Ilow BM cells with stromal cells and stimulation with IL-2. The histograms labeled control represent the isotypic control for anti-Ly49A and anti-Ly49C/I (top histogram) or for anti-Ly49G2 (bottom histogram) mAbs. Numbers are the means ± 1 SE from six independent experiments.

 


View larger version (37K):
[in this window]
[in a new window]
 
Fig. 6. Expression of NK-1.1, Ly49A, and Ly49C/I on day 7- and day 8-generated cells. Representative two-color flow cytometry analysis (NK-1.1 versus Ly-49A, middle panels; NK-1.1 versus Ly49C/I, bottom panels) of cells generated after 7 (DAY-7) or 8 (DAY-8) days of coculture of CD44low/-CD2-class Ilow BM cells with stromal cells and stimulation with IL-2. Top panels represent the isotypic controls for anti-Ly49A and anti-Ly49C/I mAbs on day 7 and day 8 cells. Numbers are the means ± 1 SE from four independent experiments.

 

View this table:
[in this window]
[in a new window]
 
Table 1 MFI of Ly49A+ and Ly49C/I+ cells from day 7 and day 8 cultures

 
Effect of Anti-H-2b mAb on Cultures.
Because studies on ß2-microglobulin-/- and Tap-1-/- mice have demonstrated that class I molecules are important in influencing the development of NK cells (6) , we used an anti-H-2b class I mAb to analyze this issue in vitro using the CD44low/-CD2-class Ilow cells. The mAb was added (1 µg/ml) to the cultures at day 0, together with IL-2. In the control groups, IL-2 plus an irrelevant or an anti-CD45.2 (this molecule was expressed on the progenitor population and its progeny as checked by preliminary trials) mAb with the same isotype as the anti-class I mAb were added. After 13 days cells were harvested and counted, and Fig. 6Citation shows that a significantly lower cell number (around 10 times less) was detected in the group to which the anti-H-2b mAb was added compared with controls. Cell counts determined that the inhibition of cell expansion was significant at day 3 (Fig. 7)Citation . To further control the specificity of the anti-class I mAb inhibition and to avoid the possibility that the inhibition was due to an antibody-dependent cell cytotoxicity mechanism, Fc receptors were blocked by the addition of a blocking anti-Fc receptor to the cultures at day 0, and this treatment did not revert the anti-class I mAb-induced inhibition (data not shown). The few cells obtained after 13 days of culture in the group to which the anti-class I mAb was added were pooled to obtain a sufficient number of cells for flow cytometric analysis. They were mature NK cells and expressed Ly49A, Ly49C/I, and Ly49G at the same frequency as controls (moreover, cells from both experimental and control groups expressed Fc receptors; data not shown), thus suggesting that the inhibitory effect was only quantitative and was not directed to a specific Ly49+ NK cell subpopulation.



View larger version (30K):
[in this window]
[in a new window]
 
Fig. 7. Effect of anti-class I mAbs on the generation of cells from CD44low/-CD2-class Ilow BM cells. The number of cells generated after 3 (day-3) or 13 (day-13) days of culture of CD44low/-CD2-class Ilow BM cells with stromal cells, IL-2, plus a control mAb ({square}), the anti-H-2b mAb ({blacksquare}), or an anti-CD45.2 mAb (), respectively, is shown. Results are the number of cells/well ± 1 SE of seven independent experiments.

 
Effect of Anti-H-2b mAb on Stromal Cells.
Because stromal cells were present in our culture system to support the differentiation of progenitors to mature NK cells, the mAb could exert its inhibitory effect by binding H-2b molecules either on the progenitor population or on stromal cells. To determine which cells the H-2b mAb was acting on, the possible effect of anti-H-2b on supportive stromal cells was tested as follows. Aliquots of CD44low/-CD2-class Ilow progenitors from B6 mice (H-2b) were plated onto irradiated supportive stromal cells obtained from mice with different haplotypes [B6 (H-2b), Tap-1-/- (very low H-2b), or C3H (H-2k)], and all groups were stimulated with IL-2 plus the control mAb (control) or IL-2 plus the anti-H-2b mAb. As shown in Fig. 8Citation , the mAb exerted the same inhibitory effect in all groups, irrespective of the haplotype of stromal cells. Taken together, these experiments suggest that the inhibitory effect of the anti-class I mAb was not exerted on supportive stromal cells but rather on the CD44low/-CD2-class Ilow progenitor population.



View larger version (17K):
[in this window]
[in a new window]
 
Fig. 8. Effect of anti-class I mAb on different stromal cells. Coculture of CD44low/-CD2-class Ilow cells with stromal cells from mice with different haplotypes. Bars represent the number of cells generated after 13 days of culture of progenitor cells with IL-2 on stromal cells from class I-deficient Tap-1-/- (Tap-1-/- STROMA) or allogeneic C3H (C3H STROMA) mice. In each histogram, Control represents the group in which a control mAb was added, and H-2b represents the group in which the anti-H-2b class I mAb was added together with IL-2. Results are the mean ± 1 SE of four different experiments.

 
Flow Cytometry Analysis of Day 7 Cells Generated after the Addition of Anti-class I mAb.
Because two cell populations were detected by flow cytometry analysis after 7 days of culture, the possibility that the mAb exerted its effect on either type of cells was examined using two-color flow cytometry analysis (NK-1.1 versus CD3) of day 7 cells. Interestingly, Fig. 9Citation shows that in the group in which the anti-class I mAb was added, the immature NK-1.1-CD3- cells were absent, and only the NK-1.1+CD3- cells were detected. These experiments clearly suggest that the inhibitory effect of anti-class I mAb was exerted on the double-negative NK progenitor and exclude the possibility of an effect exerted on the mature progeny generated in culture. The correlation between disappearance of the day 7 NK-1.1-CD3- cells and inhibition of NK cells at day 13 is an additional indirect proof that day 7 NK-1.1-CD3- cells are progenitors ready to give rise to the NK cells present at day 13. It is noteworthy that the relative number of similar day 7 NK-1.1+ cell populations present in Figs. 4Citation , 6Citation , and 9Citation show a certain degree of variability, although the SE analysis did not find evidence of significant differences among them. This variability may be explained by the method used to separate the progenitor cell population from fresh BM. Magnetic bead separation is a technique that allows a certain degree of variation compared with more precise techniques such as cell sorter separation. On the other hand, our method allows the processing of enormous numbers of BM cells in an acceptable period of time.



View larger version (26K):
[in this window]
[in a new window]
 
Fig. 9. Effect of anti-class I mAb on cells generated after 7 days of culture. Two-color flow cytometry analysis (NK-1.1 versus CD3) of cells generated after a 7-day coculture of CD44low/-CD2-class Ilow cells with stromal cells and IL-2. Control and Class I mAb represent the groups in which IL-2 plus a control mAb or IL-2 plus the anti-class I mAb were added to the cultures at day 0, respectively. The flow cytometry analysis is representative of three independent experiments.

 
Effect of Anti-class I mAb during the First Days after Addition to Cultures.
The inhibitory effect of the anti-H-2b mAb was somewhat surprising because the original progenitor population expressed a very low level of class I molecules on their surface. In an attempt to understand the reason why this could have happened, the cells were harvested at day 3 in the group in which IL-2 plus the control mAb (control) were added to the cultures and in the group in which IL-2 plus the anti-class I mAb were added to the cultures and analyzed by flow cytometry for expression of class I and NK-1.1. Fig. 10Citation shows that in the control group, two populations were present at day 3: (a) one expressing little or no class I; and (b) the other expressing a high level of class I. Interestingly, in the group in which the anti-class I mAb was added together with IL-2, only the class Ilow population was detectable. The FITC-conjugated anti-class I mAb used to analyze cells at day 3 was different from the mAb added at day 0, and previous trials have demonstrated that these two mAbs bind to different epitopes and do not interfere with each other’s binding. These data clearly suggest that IL-2 and stroma coculture induced generation of class I+ cells from class Ilow progenitors and that the anti-class I mAb exerted its effect on these class I+ progeny. The possibility that class I+ cells present on day 3 were not derived from class Ilow cells but were a consequence of expansion of contaminating class I+ cells that might not have been completely depleted was investigated by precoating the original CD44low/-CD2-class Ilow cells with the anti-H-2b mAb before placing them in coculture with IL-2 at day 0. By this means, any class I+ cell contaminating the original class Ilow progenitor population would be inhibited by the mAb. Only a very modest (<10%) inhibition was seen after the precoating of day 0 cells, thus suggesting that the effect seen when mAb was added to the cultures was exerted on the class I+ cells derived from class Ilow early precursor cells after coculture and stimulation with IL-2 (data not shown).



View larger version (21K):
[in this window]
[in a new window]
 
Fig. 10. Effect of the anti-class I mAb during the first 3 days of culture. Two-color flow cytometry analysis (NK-1.1 versus class I) of cells generated after 3 days of culture with IL-2 plus a control mAb (Control) or IL-2 plus the anti-class I mAb (Anti-Class I) is shown. The flow cytometry analysis is representative of three independent experiments.

 
Discussion

The purpose of this work was to examine the role of MHC class I molecules on proliferation and differentiation of NK progenitors able to generate in vitro mature NK cells expressing Ly49 molecules.

To isolate a progenitor population, we took advantage of a previously isolated BM population of NK progenitors characterized as CD44low/-NK-1.1-TCR- that were able to proliferate and differentiate in vitro in the presence of IL-2 on supportive stromal cells (10) . We obtained a CD44low/-CD2- cell population, which is basically a lineage-negative population because most mature hematopoietic cells are CD44high, and T, B, and NK cells also express CD2 on their surfaces. It has been shown previously (17) that hematopoietic stem cells express high levels of class I molecules but that class I molecules expression is down-regulated as they differentiate to more committed progenitors. To see whether this was also the case for NK precursors, CD44low/-CD2-class Ilow cells were tested for the presence of NK progenitors. Furthermore, because all mature BM cells, including NK cells, express class I molecules on their surface, depletion by anti-class I mAb represents an additional step for eliminating mature NK cells or other lineage-positive cells still present in the isolated population. Thus, contamination of mature NK cells in the CD44low/-CD2-class Ilow population that would have given false positive results was avoided by the combination of three mAbs directed against CD44, CD2, and class I and highly expressed on mature NK cells. The use of these mAbs in three sequential steps eliminated any NK cells remaining after the previous depletion step. The result of this procedure was tested in two ways using a phenotypic assay and a functional assay. Flow cytometric analyses showed that cells were NK-1.1-, DX5-, and Ly49G2-. Furthermore IL-2-induced expansion of mature NK cells undetectable by flow cytometry did not occur because our population did not grow after stimulation with IL-2 alone but only after stimulation with IL-2 and stroma, and the stroma dependence is a feature of progenitors but not of mature NK cells (9) . Thus, unlike mature NK cells, our precursor population was stroma dependent.

The tested progenitor population gave rise to mature NK cells expressing Ly49A, Ly49C/I, and Ly49G. In a recent work by Williams et al. (9) , Ly49s+ NK cells were obtained after two subsequent steps of culture. During the first step, their progenitor population gave rise to Ly-49- cells after stimulation with different cytokines such as IL-7, SCF, and flt3L, but without stroma. The obtained Ly49- cells completed their differentiation by expressing Ly49s molecules when stimulated with either IL-2 or IL-15 (used indifferently) and stroma. Compared with this work, our progenitors seem to be in a later step of differentiation because they required IL-2 and stroma for differentiation [which were used by Williams et al. (9) in the second step of differentiation], but not IL-7 and other "first-step" growth factors. Of interest, NK cells generated in our culture system display a similar proportion of Ly49s as fresh splenic NK cells, in contrast with the Ly49s proportion of NK cells generated in the study of Williams et al. (9) . In our opinion, the mentioned contrast is only apparent and may be due to the different types of stroma used. We used syngeneic stroma (both progenitors and stroma derived from mice with a H-2b haplotype) that would represent a more physiological model, whereas Williams et al. (9) used an allogeneic stroma (H-2K from OP mice cocultured with H-2b progenitors) that may alter the Ly49 repertoire as hypothesized in the "Discussion" section of the same work (9) . Moreover, the possibility that BM stromal cells are able to influence the NK cell repertoire is also hypothesized in a study by Sykes et al. (5) and was recently demonstrated by Roth et al. (18) .

The time course flow cytometric analysis (day 3, 7, and 13) showed that after 7 days, about 10–20% of cells were mature NK-1.1+CD3- cells, whereas the large majority (80–90%) of cells did not express NK-1.1, CD3, Ly49A, Ly49C/I, or other differentiation markers. This suggested that the major population was derived from progenitors that underwent an extensive proliferation, but not differentiation. The undifferentiated day 7 population expressed Ly49A, Ly49C/I, and low levels of NK-1.1 after 1 additional day of culture (on day 8), demonstrating their ability to differentiate to mature NK cells. Expression of Ly49A and Ly49C/I on day 8 suggests that these molecules are expressed as a late event in the differentiation pathway and expressed simultaneously with or immediately before NK-1.1 molecules. The presence of at least two cell types with different characteristics has also been shown previously (19) in humans in a lineage-negative progenitor population that seems to be similar to the one isolated by us in the mouse.

Once we standardized the experimental model, the second set of experiments was done to examine the influence of MHC class I antigens on in vitro differentiation of CD44low/-CD2-class Ilow cells. The anti-H-2b mAb was strongly inhibitory and did not exert its effect by acting on stromal cells because inhibition was seen even when the generation of NK cells was supported by allogeneic stromal cells derived from C3H mice (H-2k) or by H-2b-deficient stromal cells from Tap-1-/- mice that could not have been recognized by the anti-H-2b mAb. Taken together, these data suggest that the anti-H-2b mAb exerted its effect on the progenitor population, and flow cytometric analysis on day 7 cells clearly indicated that only its undifferentiated NK-1.1-CD3- but not the differentiated NK-1.1+CD3- progeny was absent after the addition of mAb to cultures. The data obtained with the mAb preadhered to precursors suggest the presence in the original CD44low/-CD2- cells of a class Ilow population that can differentiate within 3 days to a class I+ subpopulation. We do not know what the mechanism of the inhibitory effect of anti-class I mAb was. It is not probable that it blocked the binding to cells expressing class I ligands on their surface because the only known ligands for class I molecules are expressed on mature NK cells (Ly49 molecules) or on T cells (CD8 molecules), and no such cells were present in the first 3 days of culture after the addition of IL-2, when the anti-H-2b mAb had already exerted its inhibitory effect. Instead, we favor the hypothesis of a stimulatory effect of the mAb on class I molecules. This hypothesis is based on previous data showing that class I molecules are not only ligands that trigger T lymphocytes through the TCR or inhibit/stimulate NK cells trough Ly49 molecules but are also able to mediate different functions in the cell on whose surface they are expressed. For example, MHC class I molecules, when stimulated, may transduce signals (20 , 21) , regulate cell adhesion and proliferation (22) , and induce apoptosis (23) . This potential opens the possibility for class I molecules to interfere with cell functions and, as a consequence, with physiological processes not sufficiently analyzed up to now. For example, stimulation of class I molecules may determine the inhibition demonstrated in this work by inducing either apoptosis, growth arrest, or both of NK cell progenitors, thus controlling the NK cell number in the BM and in the periphery. Moreover, because it has also been shown that class I is linked with the receptor of IL-2 (24) , it may be possible that these molecules, when stimulated by the anti-H-2b mAb, interfered with signals triggered by stimulation of the IL-2 receptor and, as a consequence, inhibited the triggered IL-2-dependent proliferation program. The aforementioned hypothesis translated in a more general context would also suggest that the development of NK cell progenitors to mature NK cells is under the control of mature BM cells expressing ligands for H-2 (CD8 or Ly49s on the surface of T lymphocytes or NK cells, respectively) that mimic the inhibitory function of class I mAb. An additional support for this possibility is our previous data showing that BM T cells are able to inhibit NK cell differentiation (25) . Although appealing, this hypothesis would require additional testing, which is currently in progress.

Although the normal number of NK cells in class I-deficient mice such as Tap-1-/- or ß2-microglobulin-/- mice may appear to be in contrast with the inhibitory role of class I molecules shown in the present study, this is only an apparent discrepancy. Genetic absence of class I molecules cannot necessarily be mimicked by binding of a specific anti-class I mAb. In class I-deficient mice, the absence of class I molecules may trigger compensatory mechanisms, and the consequences of the lack of their expression may affect NK cell differentiation by acting from the beginning of the hematopoietic process (e.g., on totipotent hematopoietic stem cell) until the ultimate events; on the other hand, the anti-class I mAb effect shown by us is exerted specifically on IL-2-responsive, stroma-dependent progenitors ready to differentiate to mature NK cells within 1–2 weeks. In addition to this, the mAb may be either stimulating or blocking (as discussed above) class I molecules, and this cannot be seen in class I-deficient mice because the protein is not expressed. Finally, the role of control in NK cell differentiation hypothesized by us would explain why changes in expression of MHC class I antigens occur on hematopoietic progenitors during their differentiation as shown in the present study or in previous reports (17) . Because of this, we think that the in vitro model used by us is complementary to the knockout mice models and analyzes a different function of class I molecules.

Our studies suggest that modulation of class I molecule expression on surface of progenitors may represent a mechanism by which progenitors can control their number before completing the differentiation program. Furthermore, these data suggest the possibility that class I molecules expressed on progenitors play a role in NK cell differentiation. Likewise, class I molecules expressed on surrounding cells play a role in cellular adaptation mechanisms.

In conclusion, our data, in addition to the work on class I-deficient mice, highlight class I molecules as keys to the control of NK cell development by different mechanisms and/or in different phases of their developmental process. We think that our experimental model is a useful tool to consider a possible mechanism to modulate the number of NK cells by acting on class I molecules of specific progenitors.

Materials and Methods

Animals.
C57BL/6 and C3H/HEN mice were purchased from Charles River (Calco, Lecco, Italy). Tap-1-/- mice were purchased from The Jackson Laboratory (Bar Harbor, ME). Mice were used at 2–4 months of age. They were housed in an isolated colony and fed laboratory chow and acidified (pH 2.4) water ad libitum.

Culture Medium.
The standard culture medium was RPMI 1640 (Life Technologies, Inc., Grand Island, NY) supplemented with 2 mM L-glutamine, 100 units/ml penicillin, and 100 mg/ml streptomycin. For NK-long term BM cultures, 5% FCS (Life Technologies, Inc.) and 5 x 10-5 M 2-Mercaptoethanol were added to the standard culture medium (complete medium).

BM Cell Harvest.
BM was obtained from the femurs and tibias of mice killed by cervical dislocation that were disinfected by immersion in 70% ethanol. After the bones were removed and cleaned of skin and muscle, the BM was exposed by cutting the ends of the bones and expelled by inserting a needle and forcing medium through the bone shaft. The cell clumps were then broken by vortexing, and an aliquot of the cell suspension was counted with a hemocytometer.

Cultures.
BM cells were diluted in complete medium at a concentration of 2.5 x 106 cells/ml. Aliquots of 2.5 x 105 cells/well in 0.1 ml of complete medium were plated in each of the 60 innermost wells of round-bottomed 96-well plates (Costar; Corning Inc., Corning, NY), and the remaining external wells were filled with sterile water. Cultures were incubated at 37°C in 5% CO2 in humidified air without further medium change for 6 weeks. To obtain pure supportive stromal cells, after 6 weeks, the cultures were exposed to 20 Gy to kill the hematopoietic cells (including NK precursors) while sparing a sparse layer of viable and functional stromal cells. The cells containing the NK cell precursors (see below) were seeded at a concentration of 1 or 5 x 103 onto the stroma, and IL-2 (500 IU/ml) was added (day 0). The generated cells were then harvested and tested after various days (day 3, day 7, day 8, and day 13).

Isolation of the NK Cell Precursor-containing Population.
BM cells from six mice were depleted of RBCs by incubation for 5 min at room temperature with 4 ml of autoclaved RBC lysing buffer (8.32 grams of NH4Cl, 0.84 gram of NaHCO3, and 0.043 gram of EDTA per liter of deionized distilled water). The cells were then pelleted, and the FITC-conjugated anti-CD44 mAb (0.006 µg/million BM cells) was added. The cells were incubated at 4°C for 20 min in the dark, washed twice, and resuspended in 1 ml of 1x PBS with 10% FCS. The cells were then added to anti-FITC-conjugated magnetic beads (PerSeptive Diagnostic, Cambridge, MA) contained in 5 ml tubes after the suspension medium was washed out and incubated for 30 min in a rotor at 4°C. After incubation, the tube of cells was exposed to a Dynal MPC-1 magnet (Dynal, Great Neck, NY), and the cells remaining in suspension were collected and incubated for a second step with beads. The anti-CD44 mAb was used at a very low concentration (0.006 µg/million target cells) to deplete the CD44high but not the CD44low BM cells. Alternatively, CD44high BM cells can be eliminated by a step of magnetic bead depletion using the anti-Gr-1 mAb (1 µg/ml target cells), which is expressed on myeloid cells that also express CD44 at high intensity. The CD44low/- BM cells were then depleted of CD2+ by a similar step of depletion using a high concentration of FITC-conjugated anti-CD2 mAb (1 µg/ml target cells) to achieve the complete elimination of CD2+ BM cells. The third step was modified slightly: the FITC-conjugated anti-class I mAb (1 µg/ml target BM cells) was preincubated for 20 min at 4°C with 1 ml of washed magnetic beads. At the end of the incubation period, the uncoated mAb was washed out, and the CD44lowCD2- cell suspension was added to the anti-class I mAb-coated magnetic beads and incubated as described for the previous steps to eliminate only class Ihigh cells.

Cell Count.
The cells were counted with a hemocytometer, and live cells were identified by using the trypan blue exclusion assay.

Cytokines.
Human recombinant IL-2 was generously provided by Hoffmann La Roche (Nutley, NJ) and used at a concentration of 500 IU/ml.

Antibodies.
The following mAbs were used: (a) FITC-conjugated rat antimouse CD44 (IgG2b; clone IM7); (b) FITC-conjugated rat antimouse CD2 (IgG2b; clone RM2-5); (c) FITC-conjugated mouse antimouse H-2Kb (IgG2a; clone AF6-88.5); (d) purified mouse antimouse H-2Kb/H-2Db (IgG2a; clone 28-8-6); (e) purified mouse antimouse isotypic control (IgG2a, {kappa}; clone G155-178); (f) purified mouse antimouse CD45.2 (IgG2a; clone 104); (g) FITC-conjugated hamster antimouse CD3{epsilon} (CD3; IgG; clone 145-2C11); (h) FITC-conjugated hamster isotypic control (IgG; clone UC8-4B3); (i) PE-conjugated mouse antimouse NK-1.1 (IgG2a; clone PK136); (j) PE-conjugated mouse isotypic control (IgG2a; clone G155-178); (k) PE-conjugated rat antimouse Pan-NK cells (IgM; clone DX5); (l) FITC-conjugated rat antimouse Ly49G2 (LGL-1; IgG2a; clone 4D11); (m) FITC-conjugated mouse antimouse Ly49A (IgG2a; clone A1); and (n) FITC-conjugated mouse antimouse Ly49C/I (IgG2a; clone 5E6). All mAbs were purchased from PharMingen (San Diego, CA).

Flow Cytometry.
Approximately 0.5–1 x 105 cells (<5% of dead cells/sample) were pelleted in a round-bottomed centrifuge tube at 200 x g for 5 min. The pellet was resuspended in 10 µl of the predetermined dilution of the mAb and incubated in the dark at 4°C for 20 min. The cells were washed twice and resuspended in 1x PBS. The samples were then analyzed using a FACScan flow cytometer (Becton Dickinson, Sunnyvale, CA). Dead cells were gated out by size (forward scatter). The percentage of positive cells was calculated after subtraction of the background present in the isotypic control sample.

Statistics.
Data are presented as mean ± 1 SE. Student’s t test was used for all experiments. Values of P > 0.05 were considered to be not significantly different (significance was defined as P < 0.05).

Acknowledgments

We thank Sallie S. Boggs and Kenneth D. Patrene for helpful criticism and for reviewing the manuscript. This study is dedicated to the memory of Rosalba Moraca.

Footnotes

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1 Supported in part by the Ministero dell’Università e Ricerca Scientifica e Tecnologica, the Associazione Italiana Ricerca sul Cancro, and Progetlo Finalizzato Biotecnologie. Back

2 To whom requests for reprints should be addressed, at Department of Clinical and Experimental Medicine, Section of Pharmacology, University of Perugia, Via del Giochetto, 06100 Perugia, Italy. Phone: 39-75-5857493; Fax: 39-75-5857405; E-mail: farmaco{at}unipg.it Back

3 The abbreviations used are: NK, natural killer; BM, bone marrow; IL, interleukin; mAb, monoclonal antibody; TCR, T-cell receptor; MFI, mean fluorescence intensity. Back

Received for publication 8/15/00. Revision received 11/15/00. Accepted for publication 11/16/00.

References

  1. Karlhofer F. M., Ribaudo R. K., Yokoyama W. M. MHC class I alloantigen specificity of Ly-49+ IL-2-activated natural killer cells. Nature (Lond.), 358: 66-70, 1992.[Medline]
  2. Lanier L. L. Natural killer cell receptors and MHC class I interactions. Curr. Opin. Immunol., 9: 126-131, 1997.[Medline]
  3. Yokoyama W. M. Natural killer cell receptors specific for major histocompatibility complex class I molecules. Proc. Natl. Acad. Sci. USA, 92: 3081-3085, 1995.[Free Full Text]
  4. Moretta L., Ciccone E., Mingari M. C., Biassoni R., Moretta A. Human natural killer cells: origin, clonality, specificity, and receptors. Adv. Immunol., 55: 341-380, 1994.[Medline]
  5. Sykes M., Harty M. W., Karlhofer F. M., Pearson D. A., Szot G., Yokoyama W. Hematopoietic cells and radioresistant host elements influence natural killer cell differentiation. J. Exp. Med., 178: 223-229, 1993.[Abstract/Free Full Text]
  6. Ljunggren H-G., Van Kaer L., Ploegh H. L., Tonegawa S. Altered natural killer cell repertoire in Tap-1 mutant mice. Proc. Natl. Acad. Sci. USA, 91: 6520-6524, 1994.[Abstract/Free Full Text]
  7. Johansson M. H., Bieberich C., Jay G., Karre K., Hoglund P. Natural killer cell tolerance in mice with mosaic expression of major histocompatibility complex class I transgene. J. Exp. Med., 186: 353-364, 1997.[Abstract/Free Full Text]
  8. Hoglund P., Sundback J., Olsson-Alheim M. Y., Johansson M., Salcedo M., Ohlen C., Ljunggren H-G., Sentman C. L., Karre K. Host MHC class I gene control of NK-cell specificity in the mouse. Immunol. Rev., 155: 11-28, 1997.[Medline]
  9. Williams N. S., Klem J., Puzanov L. J., Sivakumar P. V., Bennet M., Kumar V. Differentiation of NK1. 1+, Ly-49+NKcellsfromflt3+multipotentmarrowprogenitorcells.J.Immunol.,163: 2648-2656, 1999.
  10. Toomey J. A., Shrestha S., de la Rue S. A., Gays F., Robinson J. H., Chrzanowska-Lightowlers Z. M. A., Brooks C. G. MHC class I expression protects target cells from lysis by Ly-49-deficient fetal NK cells. Eur. J. Immunol., 28: 47-56, 1998.[Medline]
  11. Delfino D. V., Patrene K. D., DeLeo A., DeLeo R., Herberman R. B., Boggs S. S. Natural killer cell precursors in the CD44neg/dim TCRneg population of mouse bone marrow. Blood, 87: 2394-2400, 1996.[Abstract/Free Full Text]
  12. Delfino D. V., Patrene K. D., DeLeo A. R., DeLeo R., Herberman R. B., Boggs S. S. Role of CD44 in the development of NK cells from precursors in long-term cultures of mouse bone marrow. J. Immunol., 152: 5171-5179, 1994.[Abstract]
  13. Yagita H., Nakamura T., Karasuyama H., Okumura K. Monoclonal antibodies specific for murine CD2 reveal its presence on B as well as T cells. Proc. Natl. Acad. Sci. USA, 86: 645-649, 1989.[Abstract/Free Full Text]
  14. Nakamura T., Takahashi T., Fukazawa M., Koyanagi M., Yokoyama A., Kato H., Yagita H., Okumura K. Relative contribution of CD2 and LFA-1 to murine T and natural killer cell functions. J. Immunol., 145: 3628-3634, 1990.[Abstract/Free Full Text]
  15. Trinchieri G. Biology of natural killer cells. Adv. Immunol., 47: 187-376, 1989.[Medline]
  16. Vecchini F., Delfino D., Patrene K. D., DeLeo A., Lu L., Herberman R. B., Boggs S. S. Generation of natural killer cells from long-term cultures of mouse bone marrow. Nat. Immunol., 12: 1-16, 1993.
  17. Okamoto T., Kanamaru A., Hara H., Nagai K. Changes in expression of major histocompatibility complex (MHC) class-I antigen on hematopoietic progenitors during murine development. Exp. Hematol., 15: 190-194, 1987.[Medline]
  18. Roth C., Carlyle J. R., Takizawa H., Raulet D. H. Clonal acquisition of inhibitory Ly49 receptors on developing NK cells is successively restricted and regulated by stromal class I MHC. Immunity, 13: 143-153, 2000.[Medline]
  19. Bennet I. M., Zatsepina O., Zamai L., Azzoni L., Mikheeva T., Perussia B. Definition of a natural killer NKR-P1A+/CD56-/CD16- functionally immature human NK cell subset that differentiates in vitro in the presence of interleukin 12. J. Exp. Med., 184: 1845-1856, 1996.[Abstract/Free Full Text]
  20. Geppert T. D., Nguyen H., Lipsky P. E. Engagement of class I major histocompatibility complex molecules by cell surface CD8 delivers an activation signal. Eur. J. Immunol., 22: 1379-1383, 1992.[Medline]
  21. Tscherning T., Claesson M. H. Signal transduction via MHC class-I molecules in T cells. Scand. J. Immunol., 39: 117-121, 1994.[Medline]
  22. Pettersen R. D., Hestdal K., Lie S. O., Gaudernack G. Role of the TCR binding region of the HLA class I {alpha}2 domain in regulation of cell adhesion and proliferation. J. Immunol., 156: 1415-1424, 1996.[Abstract]
  23. Skov S., Bregenholt S., Claesson M. H. MHC class I ligation of human T cells activates the ZAP70 and p56lck tyrosine kinases, leads to an alternative phenotype of the TCR/CD3 {zeta}-chain, and induces apoptosis. J. Immunol., 158: 3189-3196, 1997.[Abstract]
  24. Waldmann T. A. The IL-2/IL-2 receptor system: a target for rational immune intervention. Immunol. Today, 14: 264-270, 1993.[Medline]
  25. Delfino D. V., Lepri E., Ayroldi E., Migliorati G., Boggs S. S., Riccardi C. Suppression of natural killer cell differentiation by activated T lymphocytes in long-term cultures of mouse bone marrow. Exp. Hematol., 26: 2-9, 1998.[Medline]




This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Delfino, D. V.
Right arrow Articles by Riccardi, C.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Delfino, D. V.
Right arrow Articles by Riccardi, C.


HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
Cancer Research Clinical Cancer Research
Cancer Epidemiology Biomarkers & Prevention Molecular Cancer Therapeutics
Molecular Cancer Research Cell Growth & Differentiation