| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
a Frankel Laboratory for Bone Marrow Transplantation, Center for Stem Cell Research, Schneider Children’s Medical Center of Israel, Petach Tikva, Israel;
b McGowan Institute for Regenerative Medicine and
c Department of Pathology, University of Pittsburgh Medical Center, Pittsburgh, Pennsylvania, USA
Key Words. In vivo tracking • Fluorescence microscopy • Laser tweezers • Energy transfer • Bone marrow transplantation • PKH membrane linkers
Nadir Askenasy, M.D., Department of Pediatric Hematology Oncology, Schneider Children’s Medical Center of Israel, 14 Kaplan Street, Petach Tikva 49202 Israel. Telephone: 972-3-641-1475; Fax: 972-3-641-1475; e-mail: anadir{at}012.net.il
 |
ABSTRACT |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
B10.BR mice) preoperated for placement of optical windows over femoral epiphyses. Labeled cells were tracked in vivo by fluorescence microscopy. Cellular adhesion to the BM stroma was tested with laser tweezers, and viability was assayed by the propidium iodide (PI) exclusion test, as determined from energy-transfer measurements of the pair PKH67-PI in freshly excised femurs in situ. At optimal concentrations for in vivo tracking, 1-4 µM PKH dyes neither impaired the viability of BMCs nor the capacity of allogeneic HSCs to reconstitute hematopoiesis in myeloablated recipients. The optical window allowed in vivo visualization of 23%-26% of the PKH-labeled BMCs in the femur. The homing efficiencies at 16 hours posttransplantation were quantified as 1.77% ± 0.15% and 0.21% ± 0.02% for syngeneic and allogeneic BMCs, respectively. In femurs excised 16 hours after transplantation, 70% ± 9% of the cells were adherent to the BM stroma, and two-thirds of the cells were PI negative (viable). In vivo tracking and in situ assessment of labeled HSCs in recipient BM provide important quantitative and qualitative insights into the early stages of engraftment. Correlation of early events and the efficiency of durable engraftment serve as the basis for a systematic approach toward optimization of the conditions for transplantation.
|
INTRODUCTION |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Analysis of hematopoietic cell engraftment performed weeks and months after transplantation provides important functional information on the overall efficiency of the experimental procedure. However, this analysis does not yield specific insights into the early stages of engraftment in a manner that may be used to improve the efficiency of the particular protocol. Homing to the bone marrow (BM), seeding in the stromal microenvironment, and engraftment of short- and long-term reconstituting cells occur within the first hours to days posttransplantation [13-19]. In most studies, these stages are wrapped in a "black box" that cannot be dissected unless an early experimental end point is set. Transplanted cells are presumed to lodge in a hematopoietic niche composed of various stromal elements [20] that provides ligands for adhesion and signals that control the quiescence, self-renewal, proliferation, and differentiation of the hematopoietic progenitors and stem cells [21-26]. Considering that our limited understanding of BM physiology imposes severe limitations on simulation of the conditions in vitro, the self-imposed critique of studies performed in culture is well taken [21, 27-29]. In fact, the interaction of hematopoietic cells with the BM stroma induces reciprocal changes in the expression of ligands and cytokines, which are markedly affected by culture conditions [29-31].
Recently, a variety of imaging techniques have been tested for the study of various aspects of cellular behavior in physiological conditions [32]. Optical imaging has the optimal spatial and temporal resolution for characterization of cell-to-cell interactions and underlying molecular mechanisms [33-38]. These considerations have motivated us to search for alternative modes of investigation of the early stages of hematopoietic cell engraftment. Our primary focus was in vivo characterization of the interactions between transplanted hematopoietic cells and the recipient BM stromal microenvironment. In this study, we present a new approach to optical imaging of cells labeled with membrane linkers at the level of recipient BM in vivo and assessment of adhesion and viability in situ. A series of control experiments was performed to validate the physiological competence of our approach.
|
MATERIALS AND METHODS |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
B10) and allogeneic (B10.BRB10) bone marrow transplantation (BMT) was performed by injection of bone marrow cells (BMCs) suspended in 0.20 ml phosphate-buffered saline (PBS) into the lateral tail vein on day 0. All procedures were reviewed and approved by the Institutional Animal Care Committee of Carnegie Mellon University.
Surgical Procedure
Microsurgery was performed with a Leica GZ6 Surgical Stereoscope (Northvale, NJ; http://www.leica-camera.com) on day -4, and the mice were allowed to recover before transplantation (day 0). To visualize the recipient BM microenvironment, an optical window was placed over the distal epiphysis of the femur. The bone was exposed by a sharp skin incision, and the quadriceps muscle was detached from its patellar insertion. The muscle was reflected and secured to the skin, and soft tissue was removed with a scalpel blade. Bleeding was controlled by cauterization. The bone cortex was thinned with an electrical drill using 1-mm tip drills. Upon appearance of fissures in the eroded area, bone plates were removed with fine-tip forceps. An area of cortex of approximately 2.5 x 4 mm was removed from the distal epiphysis and a small portion of the diaphysis. Neocryl dental cement (NeoResins; Wilmington, MA; http://www.avecia.com/neoresins) was applied to the rim of the exposed bone cortex and was covered with a 3 x 5-mm glass window (cover slip #0), sterilized and pretreated with heparin sulfate. After the window had been secured with dry cement, the quadriceps was sutured to one of the skin flaps and the skin was closed with 4/0 silk sutures. The limb was casted with gauze and Neocryl in a semiflexed position.
Preparation and Isolation of Cells
Whole BMCs (wBMCs) and lineage-negative cells were harvested from femurs and tibia crushed in Hank’s balanced salt solution (HBSS), as previously reported [38]. Briefly, BMCs were suspended using an 18-gauge needle, filtered with a 30 µm sterile nylon mesh, collected by centrifugation (400 g, 10 minutes, 4°C), and resuspended in HBSS containing 2% fetal calf serum (FCS; GIBCO Laboratories; Grand Island, NY; http://www.invitrogen.com). RBCs were lysed by incubation with ammonium chloride (Sigma) for 4 minutes at room temperature. Excess HBSS was added to arrest lysis, and nucleated cells were counted after two washes with HBSS.
Lineage-negative BMCs were isolated from low-density cells collected by centrifugation (800 g, 20 minutes, 4°C) from a density gradient column at 1.087 g/ml (CedarLane; Hornby, ON, Canada; http://www.cedarlanelabs.com). The cells were gently mixed for 30 minutes at 4°C with saturating amounts of rat anti-mouse monoclonal antibodies (mAbs): CD5 (clone 53-7.3), GR-1 (RB6-8C5), CD45R (RA3-6B2), Ly-76 (TER119), T-cell receptor (TCR)-
ß (H57-597), and TCR-
(GL3) (Pharmingen; San Diego, CA; http://www.pharmingen.com). Antibody-coated cells were washed twice with HBSS containing 1% bovine serum albumin (BSA) and were incubated with sheep anti-rat IgG conjugated to M-450 magnetic beads at a ratio of 4 beads per cell (Dynal Inc.; Lake Success, NY; http://www.dynal.no). Rosetted cells were precipitated by exposure to a magnetic field, and supernatant containing lineage-negative cells was collected. The average yield of lineage-negative cell isolation was 4%, with a viability of 95% as assayed by the trypan blue exclusion test.
T lymphocytes were isolated from low-density cells by incubation with biotinylated anti-CD4 (clone L3T4) and anti-CD8 (53-6.7) mAbs (Pharmingen) for 30 minutes at 4°C. Antibody-coated cells were mixed for 20 minutes at 4°C with magnetic beads (Cellection Biotin Binder Kit, Dynal Inc.). Rosetted T cells were precipitated in a magnetic field, collected, and resuspended in HBSS. Immunomagnetic beads were detached by incubation for 10 minutes at room temperature with deoxyribonuclease, according to the manufacturer’s instructions.
Staining with PKH Membrane Linkers
Labeling was performed by incubation of 107 cells in 1 ml Diluent C with freshly prepared PKH26 or PKH67 membrane linkers (0.5-10 µM) for 10 minutes at room temperature (Sigma). Cells were collected by centrifugation (400 g, 10 minutes, 4°C), resuspended in HBSS supplemented with 10% FCS, and washed twice with HBSS. The average recovery of this procedure was 80%-90% viable cells, as assessed by the trypan blue exclusion test.
Characterization of Chimeras by Flow Cytometry
Samples were collected by tail bleeding into heparinized serum vials in 200 µl HBSS, and the cells were then washed twice and collected by centrifugation (400 g, 10 minutes, 4°C), as previously described [39]. After RBC lysis, the nucleated cells were incubated with anti-H2k-fluorescein isothiocyanate (FITC) (clone AF6-88.5) and anti-H2b-phycoerythrin (PE) (clone 36-7-5) mAbs (Pharmingen). Excess mAbs were removed by incubation with HBSS containing 2% FCS, and the low-density cells were collected by centrifugation, washed twice with fluorescence-activated cell sorting (FACS) medium (HBSS supplemented with 1 mg/ml BSA, 1 mg/ml sodium azide, and 0.36 mg/ml sodium bicarbonate), and then fixed with 0.5% paraformaldehyde. The fraction of donor peripheral blood lymphocytes was determined by two-color flow cytometry (Coulter Elite; Miami, FL).
In Vivo Microscopy
Direct observation of PKH-labeled BMCs in recipient BM was performed with Nikon Eclipse 800 (Nikon; Melville, NY; http://www.microscopyu.com) and Axioplan (C. Zeiss; Thornwood, NY; http://www.zeiss.com) fluorescence microscopes. Fluorescence resonance energy transfer measurements were performed with a microscope controlled by QED software (QED Imaging; Pittsburgh, PA; http://www.qedimaging.com). Images were acquired with a charged-coupled device camera (Hamamatsu Photonics KK; Hamamatsu, Japan; http://www.hamamatsu.com) and were analyzed either by superposition of fluorescence over brightfield images or by red-green-blue (RGB) reconstruction using three layers of fluorescence microscopy (excitation/emission). RGB image reconstruction was performed in two ways: A) green background fluorescence of BM stromal cells (460/530) and red fluorescence of cells labeled with PKH26 (540/580); B) green fluorescence of PKH67 (485/530) and red fluorescence of propidium iodide (PI) acquired in PKH67-PI fluorescence resonance energy transfer (FRET) measurements (460/625). In both cases, the blue layer consisted of UV-induced autofluorescence of BM osseous elements, the autofluorescence being acquired with a standard DAPI set of filters (360/455). Layers were pseudocolored to simulate the real hues.
In Vivo Cellular Tracking
PKH-labeled BMCs were tracked in recipient bone marrow in vivo at magnifications of 10-100x. The recipients were anesthetized, the cast and sutures were removed, the skin and quadriceps muscle were reflected, and the connective tissue was removed from the window with a scalpel blade. The mice were secured to the microscope stage, and the hind leg was secured proximally and distally to minimize motion artifacts. The bone marrow was examined through the optical window. The average duration of the imaging sessions was 1 hour.
In Situ Determination of BMC Adhesion to the BM Stroma
Adhesion of cells to BM stroma was performed in freshly isolated femurs using an optical trap integrated in a Zeiss fluorescence microscope at magnifications of 40-100x. Briefly, the laser tweezers (Cell Robotics Int.; Albuquerque, NM; http://www.cellrobotics.com) use a focused laser beam at 1,064 nm from a continuous wave diode laser as an energy trap for particles under focus. The cells are trapped by the inward pointing momentum vector of light, which is orders of magnitude stronger in the focused portion. Trapped cells can then be translocated by deflecting the light beam using a joystick-operated set of mirrors. The cells were detected and focused on by fluorescence microscopy, and trapping was performed in brightfield. Nonadhesive cells were those that changed position in reference to the BM matrix.
In Situ Determination of Cellular Viability
Cellular viability was assessed in freshly excised femurs in situ using the PI exclusion test. Cells labeled with PKH67 were excited at 460/10 nm, and fluorescence resonance energy transfer PKH67-PI was determined at the PI emission wavelength of 625/20 nm (Chroma Technology; Brattleboro, VT; http://www.chroma.com) [33]. To measure the viability of transplanted cells, femurs were incubated in PBS containing 5 µM PI for 20 minutes, and excess PI was removed by incubation with PBS. PI accumulated by dead cells with permeable membranes was excited by PKH67 emission (530 nm), and energy transfer resulted in bright red fluorescence. To eliminate the nonselective PI emission, the intensity of the red layer was lowered during RGB image reconstruction.
In Situ Staining of Femoral BM Stroma
The epiphyseal plates of the femurs were removed by sharp incisions, and the bone lumen was then perfused through 22-gauge blunt needles using a peristaltic minipump (P720; Instech Laboratories; Plymouth Meeting, PA; http://www.instechlabs.com) at rates of 0.1-0.2 ml/min. To stain the BM stromal cells, the femoral lumen was perfused at room temperature: 5 minutes with Diluent C, 10 minutes with Diluent C containing 5 µM PKH67, and 20 minutes with PBS containing 2% BSA to remove excess dye. The viability of the cellular lining of the BM stroma was assessed by energy transfer of the pair PKH67-PI.
Determination of the Visibility Factor of the Optical Window
BMCs were collected from crushed femurs in HBSS, RBCs were lysed, and the nucleated cells were then counted and suspended in 2 ml HBSS containing 2% FCS. The entire population of cells from each femur was first analyzed using a spectrofluorimeter (Spex Fluorolog; Edison, NJ). For detection of a small number of cells, a calibration curve was determined from measurements on samples containing 100-1,000 cells labeled with PKH dyes and 106 unlabeled cells. The number of PKH-labeled cells in the femurs was determined from the measured fluorescence intensity normalized against the calibration curve. In the second stage, samples of cells at an approximate concentration of 108 cells/ml were layered on slides equipped with teflon-encircled reservoirs. The number of PKH-labeled cells was determined by screening the samples using a fluorescence microscope.
|
RESULTS |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
presents placement of the transparent window over the eroded cortex of distal femoral epiphysis (A) and a femur excised after surgery for demonstrative purposes (B).
|
The Effect of Surgery on the BM Stroma
We performed a series of control experiments comparing windowed femurs with contralateral naïve femurs to assess the impact of the surgical procedure on the femoral BM. First, the histology of the excised femurs did not present significant differences in the osseous infrastructure of the BM between operated and naïve femurs. Figure 2 presents images of the femoral BM acquired through the optical window in brightfield (A) and emphasis of the osseous trabeculae detected by UV-excited bone autofluorescence (B). Second, we tested whether thermal injury caused by cauterization and erosion of bone cortex had adverse effects on the BM. The cellular lining of the BM was stained with PKH membrane linkers in situ (n = 4). Uptake of the lipophylic dye, PKH67, was homogeneous, indicating the confluence of the stromal cells lining the osseous structures in the windowed femurs (Figs. 3A-3C), as also observed in naïve femurs (Fig. 3D). Third, we assessed the viability of the stromal cells in freshly excised femurs using the PI exclusion test (n = 4). After staining of the stroma with PKH67, the femurs were incubated in medium containing 5 µM PI, and images were acquired at hourly intervals after excess PI was washed out. Energy transfer measurements of the pair PKH67-PI did not present greater PI uptake by stromal cells of operated femurs than of naïve bones within the first 3 hours ex vivo (not shown).
|
|
The Effect of PKH Concentration on Cellular Viability
We first determined the optimal concentration of PKH membrane linkers for BMC staining. Cells were labeled, and fluorescence intensity was plotted against PKH concentration. The deflection point of the curve occurred at a dye concentration of 4 µM, indicating saturation of PKH incorporation in cellular membranes. The viability of the PKH-labeled and naïve BMCs was 91% ± 3% and 96% ± 2%, respectively, as assessed by the trypan blue exclusion test (each experiment consisted of four independent incubations). The viability of PKH-labeled and naïve cells was lower at 87% ± 3% and 91% ± 4% after 3 hours of storage at 4°C (not significant), and at 66% ± 5% and 76% ± 4% after 24 hours of incubation at 37°C, respectively (p < 0.05).
Effect of PKH Dyes on HSC Functionality
The functional competence of labeled BMCs to reconstitute hemopoiesis in myeloablated recipients was evaluated by allogeneic BMT (B10.BRB10). B10 recipients (n = 10) were conditioned with busulfan (145 µg/g, i.p.) 36 hours before transplantation. In a control group of conditioned mice injected with cell-free medium, 90% of the mice died within 18 days (Fig. 4). Injection of 107 naïve or PKH-labeled allogeneic wBMCs rescued 80%-90% of the myeloablated recipients at 30 days. After 12 weeks, the fraction of donor lymphocytes in peripheral blood was 82% ± 5% and 73% ± 5% in mice injected with naïve and PKH-stained BMCs (p = 0.05), respectively (n = 5). There were no significant differences in survival rates and levels of donor hematopoietic chimerism after injection of BMCs labeled with either PKH26 or PKH67. These data indicated that stem cells labeled with PKH dyes had the capacity to reconstitute hemopoiesis in myeloablated recipients.
|
), whereas T lymphocytes repelled each other (Fig. 5C). The behavior of these cells in suspension was used to determine whether PKH dyes affected cell-to-cell contact. The presence of either PKH67 (Fig. 5B) or PKH26 dyes (Fig. 5D) on the cellular surface did not affect the aggregation of lineage-negative cells and the tendency of T lymphocytes to repel each other, respectively (five independent incubations).
|
). To determine the visibility factor of the bone window, the recipient BM was imaged 24 hours after transplantation of 106 PKH-labeled syngeneic BMCs (n = 5). The number of labeled cells was determined in vivo, and then the total number of labeled cells was measured in freshly excised femurs ex vivo using fluorescence microscopy and flow cytometry. Table 1 presents the measurements used for the estimated 23%-26% visibility factor of the optical window. Additional experiments showed relatively small variations in the visibility factor of the window during the first 5 days posttransplantation (range 22%-27%).
|
|
In Situ Imaging of the Bone Marrow
Although in vivo imaging provided detailed information, the resolution of optical imaging in situ allowed further assessment of the interaction of transplanted cells with recipient BM stroma. Adhesion of the transplanted cells to the stromal microenvironment was tested with the aid of laser tweezers. A series of measurements, performed in phantoms, was used to determine the threshold force of the laser tweezers needed to detach adherent BMCs. Applications of subthreshold forces were used to determine adhesion. After placement of the bone windows, busulfan-myeloablated recipients were injected with 106 syngeneic and 5 x 106 allogeneic wBMCs (n = 4). In femurs excised after 16 hours after transplantation, we assessed average samples of 33 cells for adhesion to the BM stroma. We estimated that 70% ± 9% of the PKH-labeled cells were adherent to the BM stroma, without a significant difference between syngeneic and allogeneic BMCs.
The main limitations of this technique were: A) Because the maximal force of the optical trap was applied at magnifications of 100x, the number of cells that could be trapped in a femur was relatively small (25-49 cells) and was restricted to the areas close to the bone window. B) The design of the specific microscopy system used in this study imposed detection of cells by fluorescence of the PKH dyes and trapping in brightfield. Nonadherent cells were those that changed position in reference to the BM matrix. C) Focused illumination of the laser tweezers increased the rate of photobleaching of the fluorochromes and caused cellular death when continuously applied over 40-120 seconds, as assessed by the trypan blue exclusion test.
The viability of PKH67-labeled transplanted cells was assessed with the PI exclusion test in situ. Briefly, the epiphyseal plates of freshly excised femurs were punctured and the bones were then incubated with a solution containing 5 µM PI. Excess PI was removed by incubation in PBS with gentle mixing, and three fluorescence layers were acquired for RGB image reconstruction (Fig. 7). The red layer was acquired using fluorescence resonance energy transfer of the pair PKH67-PI, excitation at PKH67 at 460 nm, and detection of PI emission at 625 nm. Accumulation of PI by dead cells with disrupted membranes resulted in bright red fluorescence. The fact that not all cells emitted red fluorescence suggests that direct excitation of PI was not a significant cause of false-positive assignment of cellular death. Furthermore, in a control experiment, mice were injected with unstained cells (without PKH membrane linkers), and femurs were prepared using the same procedure (n = 3). Under the energy transfer acquisition conditions and image processing used in this study, there was no red fluorescence of dead cells, indicating that PKH67 emission caused PI excitation. In femurs excised 16 hours after intravenous injection of PKH67-labeled cells, one of three cells was PI positive (an average of 25 cells were sampled in each of five femurs).
|
|
DISCUSSION |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Physiological Considerations
Engraftment of hematopoietic cells comprises a complex sequence of interactions with stromal cells and the extracellular matrix of the host BM [21-26]. Previous studies have demonstrated that stromal cell lines support HSC survival in long-term cultures, emphasizing a collaborative mode of interaction between donor cells and the recipient BM microenvironment [27-29]. However, numerous studies have pointed out that the expressions of cytokines and cell-surface ligands of both HSCs and stromal cells were altered by the interaction itself and varied in different experimental conditions [29-31]. Thus, some of the of the HSC-BM stroma interactions observed in culture pertain to the specific experimental conditions, and their physiological significance is not always evident. Development of novel tools for mechanistic characterization of the early stages of engraftment offers the advantage of unbiased observation of these processes in their natural environment in real time and overcomes some of the difficulties associated with simulation of physiological conditions in culture. A better characterization of the early stages of homing and seeding could promote efforts to optimize the conditions of transplantation to achieve high levels of durable engraftment.
Cellular tracking in recipient BM was made possible by application of an optical window on the femur. The window allows visualization of ~25% of the femoral contents of labeled cells, the femur accounting for 6%-7% of the hematopoietic space in the mouse [40]. Because we observed ~25% of the PKH-labeled cells in the femur, and seeding occurs preferentially in the femoral epiphyses early after transplantation [38], it can be estimated that the optical window allows visualization through half of the depth of the femoral lumen. Thus, it is possible to quantify the efficiency of cellular homing and seeding, and track individual cells along the time axis of engraftment. For example, after injection of 106 wBMCs, the number of labeled cells visualized through the bone window was ~70, and the homing efficiencies to the femur and the total marrow space could be estimated as 0.03% and 0.42%, respectively. Considering that the affinity of lineage-negative HSCs for seeding in the femoral BM of nonconditioned mice is 12-fold higher than that of wBMCs [38], significant numbers of hematopoietic stem and progenitor cells may be individually tracked and analyzed in vivo.
In situ analysis of the femurs revealed that, of the 280 transplanted cells in the femur, only 196 cells were adherent and 185 cells were viable. These data posit questions on the validity of some aspects of the ex vivo analysis of hematopoietic bones. On the one hand, some cells might be falsely included in the analysis. Although elegant in vitro tracking experiments may be performed [41] for identification of key molecules involved in early engraftment [34], some nonadherent cells may be trapped and included in the sample by the fixation procedure itself. Selective molecular labeling should also be performed in vivo to assess the physiological significance of observations made in cultures. On the other hand, some cells are excluded from the sample. For example, we have observed that primary seeding occurs mainly in the femoral epiphyses [38]. If the epiphyseal plates are removed during extraction of femoral BMCs for flow cytometric analysis, a significant fraction of engrafting cells are not included. Thus, quantitative assessment of homing and seeding should be performed on cells harvested from crushed femurs.
A prevalent notion postulates that transplanted cells seek a vacant hematopoietic niche, and thus, the success of engraftment reflects a delicate equilibrium between the number of homing cells and availability of niches [20, 42, 43]. Cellular homing to the host BM is mediated by a variety of molecular interactions and is influenced by the phenotype of the transplanted cells and their stage in the cell cycle [11-16, 19, 42-45]. Then, the cells interact with the BM stromal microenvironment and the extracellular matrix in a complex manner, as suggested by the large number of ligands and receptors shown to affect seeding and engraftment [11, 16, 21-29]. Despite the identification of several important molecular mechanisms of homing and seeding, much remains unknown about these early events of engraftment. We believe that introduction of an in vivo assay for direct observation of labeled cells in the marrow space provides an important tool for experimental elaboration of various functions of hematopoietic stem cells. The high spatial and temporal resolution of optical imaging [34-36] may be further extended to achieve molecular resolution in vivo using our experimental approach. Other recent technological developments present very promising perspectives for whole-body imaging to demonstrate the distribution of cells and the efficiency of gene therapy [32, 37].
In this study, we focused on two important parameters of early engraftment: adhesion to the BM stroma and viability of the transplanted cells. It is possible that, in hosts conditioned with busulfan, an agent with predominant cytoreductive activity [46], residual cytotoxic activities of T and natural killer cells in the host BM contribute to the 8.5-fold higher efficiency of homing of syngeneic versus allogeneic donor cells [47, 48]. Attempts to detect apoptosis, track changes in the expression of adhesion molecules on the surfaces of cells, and map the patterns of expression of ligands by the BM stroma are under way. We aim to combine a series of techniques for characterization of a number of functions of the engrafting cells in vivo or in situ and in real time. This approach is expected to yield new insights into various aspects of HSC engraftment. For example, homing and seeding of highly purified BMC populations may be assessed quantitatively and may be related to the success of durable engraftment. In addition, the impact of conditioning on short- and long-term hematopoietic reconstitution, as well as the topological and chronological patterns of HSC engraftment in the host femoral microenvironment, may be assayed in detail.
Methodological Considerations
Optical imaging of labeled cells in recipient BM through a transparent window in vivo is a complex task and requires consideration of several biological and technical aspects. Surgical placement of the bone window for visualization of femoral BM raised the concern of possible side effects caused by bone repair. The chemical environment in the granulation tissue and bone callus may be different from that in the naïve BM microenvironment. Bone repair starts immediately after surgery, and growth of the eroded cortex is completed within ~4 weeks in the mouse. As a result of callus formation, the field of view is gradually reduced in a centripetal manner, and finally the optical window is displaced. We have assessed several parameters of the BM structure, integrity, and viability, and its competence to host engraftment of transplanted cells. However, more subtle changes in cytokine and chemokine expression are likely to occur. For example, inflammatory cytokines in the granulation tissue at the edge of the eroded bone cortex may alter expression of chemokines involved in homing, such as stromal-derived factor [49]. These molecules might affect the efficiency of homing because they act as chemoattractants and modulate the binding affinities of cells to the BM stroma and endothelium [27, 50-53]. Our confidence in the physiological competence of the animal preparation used in this study is based on several considerations. First, the surgical procedure had a minimal effect on BM infrastructure and femoral cellularity. Second, in a femur with high rates of blood supply, we do not expect accumulation of cytokines and chemokines. Notably, the surgical procedure was performed with minimal interference to the femoral blood supply. Third, the BM stromal microenvironment supports engraftment after injury caused by preconditioning with irradiation and chemical agents, such as cyclophosphamide [54-56]. Injury to BM stroma may even promote engraftment, as suggested by inefficient engraftment of BMCs transplanted after repair of irradiation-induced injury, or when irradiation was applied in multiple subtoxic doses [55-57]. Finally, the competence of surgically manipulated femurs in hosting hemopoiesis was evident from engraftment in displaced femurs at ectopic sites [58].
PKH membrane linkers have been used for cellular tracking in previous studies [15, 17, 18]. We demonstrate that the dye was stable on the surface of quiescent cells for periods exceeding 3 weeks, did not compromise cellular viability, and did not impair the capacity of stem cells to reconstitute hematopoiesis in myeloablated recipients. At the optimal dye concentration for in vivo imaging, we observed incorporation of the membrane linkers in the cellular membrane and intracellular organelles, delineating the nucleus, endoplasmic reticulum, and vacuoles. The cellular contents of the BM are heterogeneous in reference to uptake of PKH dyes, and in some cells, the intensity of fluorescence was threefold higher than in others. Therefore, determination of cell division based on dilution of the membrane linkers should be carefully performed. Because some hematopoietic progenitors proliferate early after transplantation [42, 44, 45], PKH dyes are optimal for short-term tracking of BMCs and for detection of quiescent cells over longer periods.
The experimental conditions used in this study allowed in vivo observation of cells labeled with a concentration of 2 µM PKH at magnifications of 5–100x. Quantitative analysis at low magnifications may result in underestimation of the efficiency of homing because some cells are not detected, in particular if they crawl under the stromal cells [38]. Although a higher magnification improves the detection efficiency of fluorochromes, it shortens the area and the depth of the field of view. At high magnifications, it is difficult to avoid overlap or gaps between the fields of view, which may result in over- or underestimation of homing efficiency, respectively. During screening of the entire surface of the optical window, the BM is exposed to repeated and prolonged illumination, which induces photobleaching of the fluorochromes in oxygenated physiological environments. Photobleaching remains an experimental factor that limits the number of repeated imaging sessions performed in each recipient, and hence, the temporal resolution of engraftment that can be achieved using our approach.
In situ measurements of adhesion and viability were performed in freshly excised femurs. The main limitation of this preparation is ischemia of the femur and its cellular contents ex vivo. We have not observed uptake of PI by stromal cells within the first 3 hours after excision of the femurs; however, more subtle changes in behavior and function of the BM microenvironment cannot be ruled out. For this reason, we limited the duration of the in situ experiments to 1 hour, during which we could assay samples of 20-50 cells. Ischemic injury may be delayed by lowering the temperature of the in situ bone preparation.
In summary, we have established an experimental procedure for optical tracking of fluorochrome-labeled cells in host BM in a murine model of bone marrow transplantation. The approach is based on placement of a transparent window over the distal femoral epiphysis for in vivo optical imaging of fluorochrome-labeled cells for periods of 3-4 weeks. PKH membrane linkers had minor effects on cellular adhesion, viability, and engraftment, and did not interfere with the capacity of HSCs to reconstitute hematopoiesis in myeloablated recipients. Adhesion of transplanted cells to the BM stroma can be tested in situ with an optical trap, and viability can be assayed with the PI exclusion test using fluorescence resonance energy transfer of the pair PKH67-PI. The minimally invasive procedure described in this study provides the basis for a systematic in vivo characterization of the early stages of hematopoietic cell engraftment at the level of the recipient BM microenvironment with cellular and molecular resolution, and of the relationship between hematopoietic cell engraftment and durable engraftment.
|
ACKNOWLEDGMENT |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
REFERENCES |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  Y. Hayashi, S. Tsuji, M. Tsujii, T. Nishida, S. Ishii, H. Iijima, T. Nakamura, H. Eguchi, E. Miyoshi, N. Hayashi, et al. Topical Implantation of Mesenchymal Stem Cells Has Beneficial Effects on Healing of Experimental Colitis in Rats J. Pharmacol. Exp. Ther., August 1, 2008; 326(2): 523 - 531. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Y. Hayashi, S. Tsuji, M. Tsujii, T. Nishida, S. Ishii, H. Iijima, T. Nakamura, H. Eguchi, E. Miyoshi, N. Hayashi, et al. Topical transplantation of mesenchymal stem cells accelerates gastric ulcer healing in rats Am J Physiol Gastrointest Liver Physiol, March 1, 2008; 294(3): G778 - G786. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. N. Haylock, B. Williams, H. M. Johnston, M. C.P. Liu, K. E. Rutherford, G. A. Whitty, P. J. Simmons, I. Bertoncello, and S. K. Nilsson Hemopoietic Stem Cells with Higher Hemopoietic Potential Reside at the Bone Marrow Endosteum Stem Cells, April 1, 2007; 25(4): 1062 - 1069. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. Zhou, P. D. Acton, and V. A. Ferrari Imaging Stem Cells Implanted in Infarcted Myocardium J. Am. Coll. Cardiol., November 21, 2006; 48(10): 2094 - 2106. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. Kanamori, G. Watanabe, T. Yasuda, H. Nagamine, H. Kamiya, and Y. Koshida Hybrid Surgical Angiogenesis: Omentopexy Can Enhance Myocardial Angiogenesis Induced by Cell Therapy Ann. Thorac. Surg., January 1, 2006; 81(1): 160 - 167. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. Lapidot, A. Dar, and O. Kollet How do stem cells find their way home? Blood, September 15, 2005; 106(6): 1901 - 1910. [Abstract] [Full Text] [PDF]
|
|
