| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
* Department of Experimental Immunohematology and
Department of Immunopathology, Sanquin Research and Landsteiner Laboratory, Academic Medical Center, University of Amsterdam, Amsterdam, The Netherlands;
Department of Hematology, Academic Medical Center, Amsterdam, The Netherlands; and
Department of Immunohematology and Blood Transfusion, Leiden University Medical Center, Leiden, The Netherlands
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
Recently, the identity of the cell responsible for CFU-EC formation became subject of debate. Yoder et al. showed that the CFU-EC develops from hematopoietic cells and not from endothelial progenitor cells, which was confirmed by Rohde et al. In the latter study, it also became clear that the CFU-EC contained both monocytes and T cells (11, 12, 13, 14).
To determine the mechanism that controls formation of this colony, we analyzed the role of several cell types involved in CFU-EC culture. Recent findings about the monocytic origin were extended, and a specific role for activated T cells was demonstrated. Cell-cell contact between the monocyte and the CD4+ T cell was found to be indispensable for CFU-EC formation. Therefore, we hypothesize that CD4+ T cells become activated upon recognition of peptide Ag in the context of MHC class II molecules expressed by monocytes. Subsequently, the CD4+ T cells produce paracrine factors which are shown to be capable to induce colony formation by purified monocytes. Successive gene expression analysis of the CFU-EC suggests that the monocytes have acquired a proangiogenic phenotype during colony formation. This study provides new evidence for the hypothesis that the role of inflammatory cells, such as T cells and monocytes in vascular repair is to recruit and facilitate other cells that are involved in vessel formation.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
Apheresis buffy coats anticoagulated with 0.4% trisodium citrate (pH 7.4) were obtained after informed consent from healthy volunteers of the Sanquin Blood Bank (Amsterdam, The Netherlands). From PBMC, different cell populations were either selected or depleted using magnetic cell sorting according to the manufacturer’s instructions (Miltenyi Biotech). PBMCs were labeled with Ab-coupled beads against CD3, CD4, CD8, CD14, CD19, CD34, or CD56 (Miltenyi Biotech) for direct selection or depletion. For indirect depletion, Abs against CD34 (BD Biosciences), KDR (VEGFR-2; R&D Systems), or CD146 (BD Biosciences) were used; after a washing step, a second incubation was performed with goat anti-mouse IgG microbeads (Miltenyi Biotech). To deplete or select one particular population or combinations of populations, cells were passed through either a selection or a depletion column. To determine the purity of the isolated cell populations, flow cytometry was used.
Culture of CFU-EC
CFU-EC were cultured in EndoCult medium with supplements, according to the manufacturer’s protocol (Stem Cell Technologies). Five million cells were plated in fibronectin (FN)-coated six-well plates (Biocoat Cellware; BD Biosciences). After 48 h, the nonadherent cells were replated in FN-coated 24-well plates (Biocoat Cellware) (1 x 106 cells/well). At days 5–7, colonies were independently counted by two investigators.
The endothelial-like phenotype was confirmed by acetylated low density lipoprotein (acLDL) uptake experiments. Nonadherent cells of day 2 CFU-EC cultures were plated onto FN-coated glass slides. At day 5 of CFU-EC culture, acLDL, conjugated with Alexa Fluor 488 (10 µg/ml; Alexa Fluor 488 AcLDL; Molecular Probes), was added to the medium. After 4 h, the cells were fixed with 4% formaldehyde, and propidium iodide was added to visualize cell nuclei. Cells were examined for uptake of acLDL using a Zeiss LSM510-Meta confocal microscopy system with a x40 oil lens (Carl Zeiss). Images were acquired with the manufacturer’s software.
In some experiments, isolated cells were incubated for 30 min with Abs to activate or block T cells or monocytes. For blocking experiments, anti CD3 (clone CLB.1XA) was used in a standard concentration of 10 µg/ml (adapted from Refs.15 and 16); and to block MHC class II, clone L243 was used (17). Because the L243 Ab was described to induce apoptosis, we chose a lower concentration and tested it for the induction of apoptosis by monocytes (18), which did not occur at a concentration of 540 ng/ml. For activating experiments, anti-CD3 (clone CLB.1XE, 1 µg/ml) in combination with anti-CD28 (clone CLB.15E8, 1 µg/ml) was used. As controls, corresponding isotype Abs were used. Transwell inserts (0.4 µm pore size; Costar; Corning) were used to coculture several cell populations without cell-cell contact.
To confirm that for colony formation proliferation of cells was necessary, PBMC were irradiated with 40 Gy before culture to damage DNA, thereby inhibiting cell division. To ensure that cells were still viable after irradiation, cells were analyzed by flow cytometry for 7-aminoactinomycin D (firma) and annexin V (Bender Med Systems) expression. No significant difference in cell death or apoptosis was seen directly and 48 h after irradiation (data not shown). In addition, [3H]TdR incorporation experiments were performed. At day 4 of culture, [3H]TdR was added to the CFU-EC culture. After 24 h, cells were lysed and measured for uptake of [3H]TdR. To visualize the contribution of cells to the colonies formed, cells were labeled with PKH2 green-fluorescent linker dye specific for cells with phagocytic activity. Cells were examined using a Zeiss LSM510-Meta confocal microscopy system with a x40 oil lens (Carl Zeiss). Images were acquired with the manufacturer’s software.
Quantitative PCR array for angiogenesis
A RT2 Profiler PCR Array (SuperArray Bioscience) was performed to check for expression of genes involved in facets of angiogenesis, like vascular remodeling and growth factor production. Per experiment, a set of 5 housekeeping genes (HKG) was included. The relative gene expression of the genes was calculated as 
Ct sample = (Ct sample GENE) – (Ct sample HKG). Then, the relative gene expression (RGE) = 2 power (Ct sample1 – Ct sample 2) (19).
Cytokine and growth factor analysis
The cytokine and growth factor array test was performed on the Evidence Investigator from Randox Laboratories (Crumlin).
Statistical analysis
Statistical differences were assessed using the paired t test. Significance was assumed at p < 0.05. Data are shown as mean ± SEM, unless mentioned otherwise.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
Colonies formed by PBMC were defined as a central core of round cells with more elongated sprouting cells at the periphery (4). The endothelial-like phenotype of the colonies was confirmed by uptake of acLDL (Fig. 1A) and by real-time quantitative RT-PCR for endothelial markers (data not shown). An average number of 20 ± 5 (mean ± SEM) CFU-EC per 10–6 PBMC was counted; which equals a number of 29 ± 7 CFU-EC per milliliter of peripheral blood. Colonies are extremely adhesive to the substrate and therefore representative phenotypic analysis in cell suspension was not possible.
|
The CFU-EC count is largely influenced by the relative number of monocytes present in the culture. The addition of CD14+ cells (2, 5, 10, 20%) at day 0 of culture to a CD14+-depleted cell population reconstituted the colony formation to the level of total PBMC (Fig. 1C). Apparently an optimum exists in the concentration of monocytes needed for CFU-EC formation, because the concentration-dependent increase is lost when cultured with relatively high (i.e., 50%) percentages of monocytes.
To demonstrate the direct participation of monocytes in CFU-EC formation, CD14+ cells were labeled with a phagocyte-specific cell linker dye (PKH2, green) and added to nonlabeled CD14– cells in a ratio of 1:4. When colonies were visualized by fluorescence microscopy, cells in the center as well as in the periphery of the colony were bright green (Fig. 1D). Most cells participating in the colony were PKH2-green, but some nonstained cells were visible. This again indicates that CD14– cells contribute to the outgrowth of CFU-EC from monocytes.
Because of the extremely adhesive capacity of the colony, [3H]TdR uptake experiments were performed to investigate whether the colony formation is indeed the result of cell division. PBMC were cultured as described and at day 4, [3H]TdR was added to the culture. Indeed, an uptake of the radioactively labeled thymidine was measured at day 5, indicating that cells must proliferate to form the CFU-EC (Fig. 1E1). Cells that did not form colonies were used as a negative control. In addition, when PBMC were irradiated to induce DNA damage, no colonies were formed in contrast with controls (23 ± 8.6), which again indicates that colony formation is proliferation dependent (Fig. 1E2).
Direct cell-cell interaction with CD3+ cells is needed for CFU-EC formation of CD14+ cells
Although CD14+ cells are the major constituent of the CFU-EC, it is clear from the low number of colonies formed by purified CD14+ cells that other cells are involved.
By coculturing experiments, we showed that this supporting role is mainly mediated by direct cell-cell contact between CD14+ and CD14– cells. When 4 x 106 CD14+ were cultured in the lower compartment of a transwell system with 2 x 10–6 CD14– cells in the upper compartment, an increase in colony formation of the purified CD14+ cell population could be observed (Fig. 2A; 8.6 ± 2.7 CFU-EC in the transwell system vs 2.0 ± 0.6 CFU-EC of CD14+ cells only; p < 0.05). However, when CD14+ and CD14– cells were cocultured without separation by a filter, many more colonies were observed (36 ± 13) (Fig. 2A). This indicates that the support for CD14+ cells can be only partly ascribed to paracrine factors. Therefore, we determined which cell type interacts with the monocytes by depleting various cell types from PBMC and how to measure the effect on CFU-EC formation.
|
T cells are a heterogeneous population; therefore, both CD4+ and CD8+ T cells were tested for their contribution to colony formation by CD14+ cells. CD8+ T cells had no effect on colony formation during coculture in transwells or when cell-cell contact between CD8+ T cells and monocytes was established. In contrast, CD4+ T cells were able to induce some colony formation during transwell cultures and a significant increase was observed when CD4+ T cells were cocultured with CD14+ cells (10.9 ± 2.9 vs 1 ± 0.3; p < 0.05; Fig. 2C).
To confirm the contribution of CD4+ T cells, depletion experiments were performed. Indeed, a decrease in colony formation was observed after depletion of CD4+ T cells (5.3 ± 4.3) from the PBMCs (19.9 ± 6.4), whereas depletion of CD8+ T cells did not show a decrease in colony formation (25.3 ± 4.3).
Cell-cell contact-dependent support of T cells is TCR-MHC class II mediated
Because the support to monocytic CFU-EC formation by T cells was a CD4+ T cell-specific event, we hypothesized that the cell-cell contact-derived effect might be due to an Ag-presenting-like interaction between MHC class II on the monocyte and the TCR-CD3 complex on the CD4+ T cell. To test whether this interaction occurs, an anti-CD3 Ab that specifically inhibits TCR-CD3 signaling (by blocking clustering of the TCR-CD3 complex) was added to a coculture of CD4+ T cells and monocytes. Indeed, colony formation was inhibited after blocking the TCR-CD3 complex on CD4+ T cells (7.7 ± 4.8 vs 45.4 ± 16.1 for controls; p < 0.05; Fig. 3A).
|
The decrease in colony formation after blocking the CD3-MHC class II interaction was confirmed by [3H]TdR uptake experiments. Cocultures of CD4+ T cells and monocytes show a decrease in proliferation when the TCR-CD3 complex or MHC class II were blocked (Fig. 3C).
Conversely, when CD4+ T cells were activated before coculture with CD3-CD28-activating Abs, a massive increase in colony formation was observed compared with controls (394 ± 118 vs 45.4 ± 16.2; p < 0.05; Fig. 3, A and B). Onset of colony formation was already visible in the first 2 days of culture.
T cells produce factors that induce colony formation by monocytes
Because CD4+ T cells become activated upon recognition of a peptide Ag in the context of MHC class II molecules expressed by monocytes, we hypothesized that the paracrine factors produced by activated CD4+ T cells subsequently stimulated CFU-EC formation by monocytes (Fig. 4). To test this hypothesis, CFU-EC cultures were performed with preactivated CD4+ T cells in a transwell insert above purified monocytes. Indeed, coculturing of CD3CD28 activated CD4+ T cells with CD14+ cells, with separation by a transwell insert, showed similar results as cocultures with activated CD4+ T cells (226 ± 33.2 vs 394 ± 118; p < 0.001; Fig. 3, A and B). The paracrine effect was confirmed by replacing the medium from monocytes-only cultures with supernatant of cultures with activated CD4+ T cells, and again the colony formation was substantial and fast (753 ± 30.8 vs 0 ± 0, p < 0.001, Fig. 3, A and B). The induction of colony formation by the supernatant of activated T cells showed a dose-dependent response, because replacement of 10% of the medium induced already colony formation by purified monocytes, but a maximal effect was reached when 75–80% of the medium was replaced by T cell supernatant (data not shown). Because 25% supernatant with 75% EndoCult medium gave the most distinguishable colonies, we decided to use this ratio for further experiment involving T cell supernatant.
|
The CFU-EC colonies derived from monocytes show an angiogenesis-supporting phenotype, rather than an endothelial phenotype
To confirm the angiogenic properties of the CFU-EC formed by the monocytes stimulated with paracrine factors from activated CD4+ T cells, a real-time quantitative PCR-array for genes involved in angiogenesis was performed. Interestingly, as is shown in Table I, using CD4+ supernatant-stimulated CD14+ cell-derived CFU-EC, the genes up-regulated in the CFU-EC culture are mostly those that are involved in supporting angiogenesis, like proteases (matrix metalloproteinases 2 and 9, plasminogen activator) and cytokines (angiopoietin 2, fibroblast growth factor A, IL-8, MCP-1; Ref. 20). In addition, also factors involved in inflammation were up-regulated, like IL-8, IFN-
, and TNF-
(21, 22). Comparable results were obtained when arrays were performed on RNA isolated from standard PBMC cultures or from monocyte cocultures with activated CD4+ T cells.
|
We tested the pooled supernatant from activated T cells from four donors for a number of molecules. Results are shown in Table II. It is clear that hardly any vascular endothelial growth factor (VEGF) is formed by the T cells, but IL-6, IL-8, IL-10, TNF-, and MCP-1 are present in the supernatant.
|
, IFN-, and MCP-1, also the protein levels are increased. |
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
Recently, also others have described the cell types involved in CFU-EC formation. Yoder et al. (11) showed that CFU-EC possess myeloid progenitor cell activity, differentiate into macrophages, and are not clonally related to the more immature EPC called late outgrowth endothelial cells. In addition, Rohde et al. (13) demonstrated that the colony consisted mainly but not solely of monocytes.
From our experiments, it is clear that CD14+ cells need a supporting cell in CFU-EC formation. This is also obvious from Fig. 1C; the increase in colony formation by PBMC is dependent on monocyte concentrations, although the concentration-dependent increase in colony formation is lost when the cell population consists of >50% of monocytes. Furthermore, it was found that activated T cells, and in particular activated CD4+ T cells, support CFU-EC formation. We show that the paracrine factors produced by CD4+ T cells following cell-cell interactions with CD14+ cells contribute to CFU-EC formation. Because this support is CD4 dependent, we hypothesized that the interaction between monocytes and T cells in this culture might be via MHC class II molecules. This hypothesis, as shown in Fig. 4, is supported by the observation that blocking or activating MHC class II on CD14+ or CD3 on CD4+ T cells strongly modulates colony formation. Furthermore, when T cells were activated via the TCR-CD3 complex before culture (thus by mimicking the interaction with the monocyte), the coculture showed a massive and faster outgrowth of colonies from monocytes. Contact-induced T cell stimulation could be replaced by addition of activating CD4+ T cells in the upper compartment of a transwell system or just by adding activated T cell supernatant. Bypassing the Ag-presenting mechanism in this way thus enables a more rapid growth of more CFU-EC. Nevertheless, it must be emphasized that only a subset of monocytes is able to form colonies, because the numbers of colonies are still far lower than the number of CD14+ cells in the starting population. It remains to be investigated which MHC-presented Ags can support the monocyte-induced T cell activation; the Ags presented in the in vitro assay, however, might be completely different from those that are relevant at sites of inflammation.
Additionally, the exact soluble factors that are involved in CFU-EC formation cannot be determined from our studies. Rohde et al. described TNF- as a potential candidate, although from their study it is not clear whether the TNF- is derived from the monocyte after T cell contact or from the T cell itself. Also other cytokines, released when the T cells are activated in the EndoCult medium (e.g., IL-8, MCP-1) are nominated candidates. Interestingly, we detected hardly any VEGF in the EndoCult medium conditioned with activated T cells. VEGF is a known inducer of endothelial-like differentiation of monocytes (23, 24, 25) and can be produced by T cells (26); however, in our experiments this growth factor seems not of importance.
Our PCR array data on the CFU-EC cultures from monocytes stimulated with CD4+ T cell supernatant showed an increase in proinflammatory genes and genes involved in vascular remodeling. Interestingly, among the decreased genes are several known inhibitors of angiogenesis as well. Thus, the activated CD4+ T cells, present at inflammatory sites, induce monocytes to differentiate into angiogenesis-supporting cells.
From our data, it can be concluded that the clinical studies that have observed a correlation between CFU-EC numbers and several diseases (4) have in fact measured the response of a CD14+ cell population to activated T cells. In this respect, the variable(s) that might have an association with several diseases might be 1) the frequency of a certain CD14+ monocyte subset, 2) the in vitro (exogenous) Ag presentation of monocytes, 3) the capacity of T cells to become locally activated and produce paracrine factors, 4) the sensibility of CD14+ monocytes to these factors, or 5) a combination of these factors. Combining these facts, we hypothesize that the number of CFU-EC more likely reflects the ability of inflammatory cells to participate in neovascularization at sites that require (collateral) vessel formation. In agreement, the importance of T lymphocytes for in vivo occurrences like wound healing, tumor growth, and atherosclerotic plaque development are well known. A common mechanism in these processes is angiogenesis (27). The previously recognized association between angiogenesis and chronic inflammation in for example cancer and atherosclerosis shows that both processes have similar initiating events and characteristics. In both events, the influx of immune cells and the production of cytokines and growth factors play an important role (22, 28, 29).
T cells participate in these processes by modulating the trafficking of other cellular components of the immune system (such as monocytes and macrophages) and in producing cytokines (30, 31, 32). In addition, Mor et al. (26 have described the ability of T cells to produce VEGF at inflammatory sites and respond to VEGF by polarization into a Th1 phenotype).
Additionally, studies suggest that T cells are necessary for the development of collateral blood vessels (33). For example, Stabile et al. have shown that CD4 knockout mice have an impaired arteriogenic response after ligation of the femoral artery in the hind limb. This so-called ischemic hind limb model confirms the important role of CD4+ cells in collateral development. The collateral-enhancing effects of CD4+ T cells appear to reside in the classic immune response activity of these T cells inducing monocyte-macrophage accumulation in the ischemic muscle. The accumulating macrophages then secrete a broad array of cytokines and growth factors, which facilitate collateral development (34). These results were confirmed by Van Weel et al. (35), who described a role for NK cells and CD4+ T cells in arteriogenesis in knockout mice. Their data show that both NK cells and CD4+ T cells modulate arteriogenesis, and they suggest that promoting lymphocyte activation may represent a promising method to treat ischemic disease.
Direct evidence for the actual in vivo function of the CFU-EC-forming cell is still required, because the only evidence for a biological significance are the clinical studies with the correlations between vascular diseases and CFU-EC numbers. Before these experiments can be performed, a number of uncertainties must be solved, one of which is the identity of the colony-forming cell. Future experiments to identify the CFU-EC-forming cell within the pool of CD14+ cells are necessary, because the CFU-EC can be recognized thus far only by its progeny. Without this knowledge, it remains difficult to design in vivo experiments that will elucidate the real function of the CFU-EC.
In this study, we demonstrated that a widely used and commercially available CFU-EC assay is not an assay on endothelial progenitor cells, but reflects the T cell-mediated induction of monocytes toward angiogenesis-supporting cells. These immune cells are not incorporated into the new vessels but are mainly involved in the recruitment and facilitation of cells that do so. We now provide for the first time evidence that a TCR-MHC class II-specific T cell response is involved in this process. Therefore, this study justifies the attention for inflammatory cells and in particular T cells and monocytes in therapeutic angiogenesis, and makes these cells interesting targets for the development of regenerative cellular therapies.
|
Acknowledgments |
|---|
|
Disclosures |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
|
Footnotes |
|---|
1 Address correspondence and reprint requests to Dr. C. E. van der Schoot, Sanquin Research, Department of Experimental Immunohematology, Plesmanlaan 125, 1066 CX Amsterdam, The Netherlands. E-mail address: e.vanderschoot{at}sanquin.nl 
2 Abbreviations used in this paper: EPC, endothelial progenitor cell; CFU-EC, CFU-endothelial cell; FN, fibronectin; acLDL, acetylated low density lipoprotein; Alexa Fluor 488 AcLDL, acLDL conjugated with Alexa Fluor 488; VEGF, vascular endothelial growth factor.
Received for publication May 4, 2007. Accepted for publication January 30, 2008.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
-anti-CD3. J. Immunol. 139: 2873-2879. [Abstract]This article has been cited by other articles:
|
|
 |
|
  B. J. Wu, R. G. Midwinter, C. Cassano, K. Beck, Y. Wang, D. Changsiri, J. R. Gamble, and R. Stocker Heme Oxygenase-1 Increases Endothelial Progenitor Cells Arterioscler Thromb Vasc Biol, October 1, 2009; 29(10): 1537 - 1542. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Desai, A. Glaser, D. Liu, N. Raghavachari, A. Blum, G. Zalos, M. Lippincott, J. P. McCoy, P. J. Munson, M. A. Solomon, et al. Microarray-Based Characterization of a Colony Assay Used to Investigate Endothelial Progenitor Cells and Relevance to Endothelial Function in Humans Arterioscler Thromb Vasc Biol, January 1, 2009; 29(1): 121 - 127. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
20% colony formation (mean ± SEM, n = 5; *, p < 0.05; **, p < 0.01). C, Addition of different percentages (2–50%) of CD14+ cells reconstituted colony formation (mean ± SEM, n = 4, *, p < 0.05). D, CFU-EC cultured from 4 x 106 CD14+ cells, labeled with PKH2 (green) and 1 x 106 unlabeled CD14– cells. A representative CFU-EC colony is shown, consisting of mainly green cells (CD14+) and small numbers of unlabeled cells (x400). E1, [3H]TdR uptake of CFU-EC cultures vs a negative control (by using an Ab that blocks CFU-EC formation). E2, Number of CFU-EC colonies at day 5 cultured from radiated (40Gy) and nonradiated PBMC (5 x 106; mean ± SD; n = 3; *, p < 0.05).
