|
|
||||||||
Lund Strategic Research Center for Stem Cell Biology and Cell Therapy, Lund University, Lund, Sweden
| Abstract |
|---|
|
|
|---|
| Introduction |
|---|
|
|
|---|
Careful identification of different cell lineages can be equally important when studying the multilineage differentiation potential of stem cells within a specific tissue or organ. Hemopoietic stem cells (HSC)3 represent the best-characterized somatic stem cells, and despite representing less than 0.01% of total murine bone marrow (BM) cells, long-term repopulating HSC can be highly purified prospectively (3, 4). This has facilitated important studies in which mice have been transplanted with single HSC, to assess frequencies of HSC in purified populations and following manipulations, such as attempts to expand HSC ex vivo. Because HSC are strictly defined by their ability to long-term repopulate all blood cell lineages, such transplantation studies include reconstitution analysis of the B, T, and myeloid cell lineages several months after transplantation. Specific identification of HSC-derived myeloid reconstitution is particularly critical because myeloid progenitors and their progeny have short half-lives, in contrast to lymphoid progeny that can be long-lived (5, 6). Detection of HSC-derived multilineage reconstitution is established using Abs against cell surface Ags thought to be specific to the B (B220), T (CD3), and myeloid (Gr-1/Mac-1) lineages. Typically, this is done separately for each of the different lineages while costaining for two different isoforms of the CD45 (common leukocyte Ag), to distinguish donor and recipient derived reconstitution (7, 8, 9, 10, 11, 12, 13, 14, 15). However, performing simultaneous staining against B220, CD3, Mac-1, and Gr-1 with the Ab clones traditionally used, here we demonstrate that the RB6-8C5 Ab detecting Gr-1 also binds to a subpopulation of CD3+CD8+ T cells present in peripheral blood (PB), constitutively as well as following transplantation, probably because RB6-8C5 also detects high levels of the Ly6C Ag. In addition to having important implications for determination of HSC frequencies and efforts to ex vivo expand HSC, our studies reveal that application of the traditional blood cell lineage analysis on mice reconstituted with single HSC can be deceptive, resulting in faulty identification of long term HSC with apparent lineage-restricted potentials. This reinforces the paramount importance of developing more stringent assays for evaluation of multilineage differentiation as well as transdifferentiation potentials of stem cells.
| Materials and Methods |
|---|
|
|
|---|
C57BL/6-CD45.1 or C57BL/6-CD45.2 mice were bred at B & S (Ry, Denmark) and used after approval from the local ethics committee (Länsstyrelsen, Skåne, Sweden). Mice were used as transplant recipients between 8 and 14 wk of age, and as donors between 10 and 14 wk of age.
Abs
The following mAbs were used: anti- B220 (RA3-6B2), Mac-1 (M1/70), CD4 (L3T4), CD8
(53-6.7), CD3
(145-2C11), Gr-1 (RB6-8C5), CD44 (IM7), CD122 (TM-
1), CD45.1 (A20), CD45.2 (104), CD34 (RAM34), ScaI (E13-161.7), Ly6C (AL21), and c-kit (2B8). All Abs were from BD PharMingen (San Diego, CA) and directly conjugated as indicated with FITC, PE, PE-Cyanin-5 (PE-Cy5/CyChrome/Tricolor), or allophycocyanin. A goat anti-rat Tricolor Ab (Caltag Laboratories, Burlingame, CA) was used to visualize lineage-positive cells in the BM. All Abs were used at optimally titrated concentrations and showed no unspecific binding compared with isotype-control matched control Abs. Identical staining profiles were obtained with the RB6-8C5 Ab conjugated with different fluorochromes (D. Bryder, unpublished data).
Cell isolations, HSC purification, and FACS
Cell preparation techniques for PB and purification of Lin-Sca+kit+CD34- BM cells have been described (16). May-Grunwald Giemsa stainings of cells sorted from PB were performed as described previously (17). For FACS analysis, single cell suspensions were prepared and stained with appropriate Abs and analyzed on a FACSCalibur (BD Biosciences, Mountain View, CA) (16). All sorts were performed on a FACSVantage SE (BD Biosciences) and reanalysis reproducibly showed purity exceeding 96%. Analyses were performed using FlowJo software (Treestar, San Carlos, CA).
Ex vivo expansion cultures
Expansion cultures were performed as described previously using serum-free X-Vivo 15 medium (BioWhittaker, Walkersville, MD) supplemented with 1% detoxified BSA (StemCell Technologies, Vancouver, British Columbia, Canada) and recombinant cytokines as indicated (16, 18). Cytokines were used at 50 ng/ml except for IL-3, which was used at 20 ng/ml.
In vivo reconstitution experiments
In vivo reconstitution experiments with freshly isolated or ex vivo expanded HSC populations were performed using the congenic CD45.1/CD45.2 system as described previously (16, 18). All recipient mice (CD45.1) were in addition to the specified number of test cells (CD45.2), also transplanted with unfractionated BM cells as competitor cells (of same CD45-type as the recipient mice, CD45.1) to provide radioprotection and to allow quantification of the reconstitution activity (3, 19).
| Results |
|---|
|
|
|---|
A traditional PB lineage analysis of mice transplanted with HSC is performed by gating specifically on donor derived cells (CD45.2) and analyzing these for contribution to the B (B220), T (CD3), and myeloid (Mac-1 and Gr-1) lineages (Fig. 1) (7, 8, 9, 10, 11, 12, 13, 14, 15, 18, 20). However, when using the Ab clones traditionally used for these analyses (RA3-6B2, 145-2C11, M1/70, and RB6-8C5, respectively), we noticed that the sum of the different lineage markers consistently exceeded 100% (Fig. 1), suggesting either unspecific Ab staining, or that some of the markers used as "lineage-specific" might in fact be coexpressed on other lineages, and if so, not entirely specific for the lineage they were intended to detect. To investigate this in more detail, PB cells from untransplanted C57BL/6 mice were first investigated for coexpression of B220 and CD3 on Mac1+RB6-8C5+ cells (Fig. 2). Strikingly, a significant fraction of RB6-8C5+ cells was found to coexpress CD3 in all mice analyzed, whereas no other coexpression was observed in the other combinations investigated (Fig. 2A). Cell sorting and subsequent May-Grunwald Giemsa staining of RB6-8C5+CD3+ cells revealed a lymphoid morphology, in contrast to RB6-8C5+CD3- cells, that virtually all had a granulocytic appearance (Fig. 2B).
|
|
), suggesting that they, to a large extent, represent memory CD8+ T cells (21). In fact, all phenotypically defined memory CD8+CD44highCD122+ cells were positive for RB6-8C5 staining (Fig. 3D).
|
|
Based on our finding of a subset of PB CD3+CD8+ T cells coexpressing RB6-8C5, the use of RB6-8C5 as a myeloid specific Ab could result in an overestimation of levels of myeloid reconstitution following HSC transplantation. Thus, we next investigated the potential significance of using RB6-8C5/Mac-1+CD3- cells as a more stringent parameter for myeloid reconstitution following transplantation of freshly isolated and ex vivo expanded HSC. In mice transplanted with 30 purified CD45.2+Lin-Sca+kit+CD34- cells, in competition with 200,000 unfractionated CD45.1+ BM cells, high levels of CD45.2 reconstitution were observed, that included myeloid cells regardless of adapting the traditional (RB6-8C5/Mac-1+ only) or new (RB6-8C5/Mac-1+CD3-) criteria for myeloid reconstitution. However, the levels of myeloid reconstitution were consistently lower when evaluating myeloid reconstitution as RB6-8C5/Mac-1+CD3- cells (Table I). Mice transplanted with the same number of Lin-Sca+kit+CD34- cells cultured under in vitro conditions promoting 2- to 3-fold HSC expansion (18), gave similar differences when using the two criteria for myeloid reconstitution, thus confirming the enhanced long-term multilineage reconstituting activity of the ex vivo expanded cells (Table I). The importance of adapting a more stringent lineage analysis became particularly evident when purified Lin-Sca+kit+CD34- HSC were transplanted at limiting numbers (most positive mice reconstituted with a single HSC). Under these conditions, if presence of RB6-8C5/Mac-1+CD3- cells was adapted as a requirement for myeloid reconstitution, mice long-term reconstituted exclusively with B and T lymphocytes were observed, that upon traditional lineage analysis (presence of RB6-8C5/Mac-1+ cells) would have appeared as also being myeloid and therefore HSC reconstituted (Table I, Fig. 5A). Strikingly, the frequency of mice long-term multilineage reconstituted with limiting numbers of in vitro expanded HSC was dramatically (3-fold) overestimated if using RB6-8C5/Mac-1+ cells rather than RB6-8C5/Mac-1+CD3- cells as criteria for myeloid reconstitution (Table I). Thus, the standard protocol for evaluation of myeloid reconstitution results in overestimation of HSC numbers.
|
|
Due to short half-lives of myeloid progenitors and their differentiated progeny, the ability to long-term reconstitute myelopoiesis has not only been considered a required but frequently also sufficient property defining HSC (23). Based on this assumption, cells capable of long-term reconstituting myelopoiesis without simultaneous reconstitution of the B and/or T cell lineages could be interpreted as representing novel HSC with long-term lineage-restricted reconstitution potentials. In line with this, following transplantation of limiting numbers of purified Lin-Sca+kit+CD34- HSC we identified a number of mice with robust long-term reconstitution of CD3+ and RB6-8C5/Mac-1+ cells, but lacking B220+ cells. If relying on traditional lineage analysis this would imply the existence of long-term HSC with a T and myeloid lineage restricted reconstitution potential. However, when analyzing expression of CD3 against RB6-8C5/Mac-1, all such mice proved to be reconstituted by T cells only (Fig. 5B), and thus lacking the myeloid component defining HSC.
| Discussion |
|---|
|
|
|---|
Such multiparameter lineage reconstitution analysis is usually performed in PB, combining CD45.1 and CD45.2 Abs with Abs against the B (B220), T (CD3), and myeloid (RB6-8C5/Mac-1) lineages. The standardized approach, adapted by leading HSC groups (3, 7, 8, 9, 10, 12, 13, 14, 15, 18, 20), has been to investigate each of the lineages separately, with the assumption that the Abs specifically detect each of the lineages investigated. As for long-term HSC, evaluation of the myeloid lineage is particularly critical, as assessed by the presence of RB6-8C5+ and/or Mac-1+ cells. Investigating the specificity of this approach, we here demonstrate in steady state as well as following transplantation, that a sizeable fraction of RB6-8C5+ cells in PB, display a lymphocyte morphology, and do in fact represent CD3+CD8+ T cells.
Importantly, this was not due to unspecific binding of the RB6-8C5 Ab, because this was carefully titrated and matched isotype control Abs showed no unspecific binding to CD3+CD8+ cells. Furthermore, CD3+CD4+ T cells did not show binding of the RB6-8C5 Ab. Retrospectively however, it is not surprising that the RB6-8C5 Ab also labels nonmyeloid cells to some extent, as it has been demonstrated to potentially cross-react with the Ly-6C Ag, although Ly-6G is the main target for the RB6-8C5 Ab (22). In support of this, preincubation of spleen cells with a high concentration of Ly6C specific Abs, competed out binding of RB6-8C5 to CD8+ T cells. Ly6C is expressed on memory and activated T cells and thus adoptive transfer experiments like those described in this study, would likely enhance the number of CD8+RB6-8C5+ cells in these situations due to homeostatic proliferation of newly generated T cells.
Although we believe the impact of this observation will be limited in studies where large numbers of unmanipulated HSC are transplanted, we demonstrate that the increased focus on transplanting candidate HSC populations at limiting numbers can result in overestimation of HSC frequency when using the traditional lineage analysis. More importantly, we show that the use of the RB6-8C5 Ab as myeloid specific, can be deceptive and implicate the existence of a novel long-term HSC with myeloid and T cell, but not B cell reconstitution potential, in mice reconstituted with limiting numbers of HSC. However, when performing simultaneous expression analysis for RB6-8C5 and CD3 expression, it became clear that all RB6-8C5+ cells in these mice coexpressed CD3, and were therefore not derived from long-term HSC but rather from lymphoid-restricted progenitors. In previous studies in which RB6-8C5 has been used as a myeloid specific marker, the existence of long-term HSC with robust myeloid and T cell, but little or no B cell potential, has been proposed (12).
Extensive efforts aiming to expand long-term HSC ex vivo have been disappointing, usually resulting in loss rather than gain of HSC numbers (25). However, a few recent studies have suggested that long-term HSC can undergo self-renewing divisions in vitro with a concomitant, albeit limited increase in long-term reconstituting HSC activity (18, 26). In the present studies we demonstrate that the use of RB6-8C5 as a specific marker of myeloid reconstitution can result in erroneous calculation of HSC numbers following such manipulations, emphasizing the importance of adapting a stringent lineage analysis when comparing different sources of HSC, as they might differently contribute to reconstitution of cells coexpressing different lineage markers.
The existence of a subfraction of RB6-8C5+ cells coexpressing CD3 was reported in a previous study where double staining for RB6-8C5 and Mac-1 was explored to enable separation of the granulocyte and macrophage lineages (27). However, the CD3+RB6-8C5+ cells were not further characterized and hypothesized to represent a rare cell population. As a consequence, the potential impact of this finding on lineage analysis following transplantation was not entertained, and accordingly has not resulted in any modification of the traditional blood lineage analysis (7, 9, 10, 12, 13, 14, 15, 18, 20, 26). The results in the present studies clearly emphasize the importance of introducing a more stringent and reliable lineage analysis. This will be paramount for the increasing efforts exploring potentially altered lineage reconstitution potentials of HSC following genetic modifications or manipulation in vitro. Thus, we propose a new standard for the critical analysis of myeloid reconstitution in PB, in which staining for the myeloid markers RB6-8C5 and Mac-1 are combined with staining for T and B lymphoid Ags. Thus, myeloid reconstitution should be stringently defined as donor-derived Gr-1/Mac-1+ (or Mac-1+) cells negative for expression of B and T cell markers.
We believe that the observed cross lineage expression of markers previously believed to be highly specific for one cell lineage is unlikely to be unique to blood cells. On the contrary, in times of exceptional claims of stem cell plasticity (1, 2), it should rather be assumed to be the rule. Consequently, any claims regarding unexpected cell fates should be substantiated not only by acquisition of lineage-specific markers for the translineage, but also through absence of markers of the lineages normally derived from the somatic stem cell investigated.
| Acknowledgments |
|---|
| Footnotes |
|---|
2 Address correspondence and reprint requests to Dr. Sten-Eirik W. Jacobsen, Department of Stem Cell Biology, BMC B12, Klinikgatan 26, 221 84, Lund, Sweden. E-mail address: Sten.Jacobsen{at}stemcell.lu.se ![]()
3 Abbreviations used in this paper: HSC, hemopoietic stem cell; BM, bone marrow; EE, expansion equivalent; PB, peripheral blood. ![]()
Received for publication May 14, 2003. Accepted for publication November 11, 2003.
| References |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
T. A. Zola, E. S. Lysenko, and J. N. Weiser Mucosal Clearance of Capsule-Expressing Bacteria Requires Both TLR and Nucleotide-Binding Oligomerization Domain 1 Signaling J. Immunol., December 1, 2008; 181(11): 7909 - 7916. [Abstract] [Full Text] [PDF] |
||||
![]() |
K. A. Matthias, A. M. Roche, A. J. Standish, M. Shchepetov, and J. N. Weiser Neutrophil-Toxin Interactions Promote Antigen Delivery and Mucosal Clearance of Streptococcus pneumoniae J. Immunol., May 1, 2008; 180(9): 6246 - 6254. [Abstract] [Full Text] [PDF] |
||||
![]() |
A. L. Nelson, A. M. Roche, J. M. Gould, K. Chim, A. J. Ratner, and J. N. Weiser Capsule Enhances Pneumococcal Colonization by Limiting Mucus-Mediated Clearance Infect. Immun., January 1, 2007; 75(1): 83 - 90. [Abstract] [Full Text] [PDF] |
||||
![]() |
J. Laurent, N. Bosco, P. N. Marche, and R. Ceredig New insights into the proliferation and differentiation of early mouse thymocytes Int. Immunol., August 1, 2004; 16(8): 1069 - 1080. [Abstract] [Full Text] [PDF] |
||||
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |