First published online April 26, 2007
Stem Cells
Vol. 25 No.
7
July 2007, pp.
1807
-1813
doi:10.1634/stemcells.2006-0581; www.StemCells.com
© 2007 AlphaMed Press
TISSUE-SPECIFIC STEM CELLS |
The Late Dividing Population of 
-Retroviral Vector Transduced Human Mobilized Peripheral Blood Progenitor Cells Contributes Most to Gene-Marked Cell Engraftment in Nonobese Diabetic/Severe Combined Immunodeficient Mice
Sebastian Brennera,b,
Martin F. Ryserb,
Narda L. Whiting-Theobalda,
Marcus Gentschb,
Gilda F. Lintona,
Harry L. Malecha
aLaboratory of Host Defenses, National Institute of Allergy and Infectious Diseases, National Institutes of Health, Bethesda, Maryland, USA;
bDepartment of Pediatrics, University Clinic Carl Gustav Carus, Dresden, Germany
Correspondence: Sebastian Brenner, M.D., Department of Pediatrics, University Clinic Carl Gustav Carus, Building 21, Fetscherstr. 74, 01307 Dresden, Germany. Telephone: 49 351 458-6884; Fax: 49 351 458-6333; e-mail: sebastian.brenner{at}uniklinikum-dresden.de
Received September 18, 2006;
accepted for publication April 3, 2007.
First published online in STEM CELLS EXPRESS April 26, 2007.
 |
ABSTRACT
|
|---|
We used the nonobese diabetic/severe combined immunodeficient (NOD/SCID) mouse model to assess the repopulation potential of subpopulations of mobilized human CD34+ peripheral blood progenitor cells (PBPC). First, PBPC were transduced with 
-retrovirus vector RD114-MFGS-CFP, which requires cell division for successful transduction, at 24 hours, 48 hours, and 72 hours to achieve 96% cyan fluorescent protein (CFP)-positive cells. Cells were sorted 12 hours after the last transduction into CFP-positive (divided cells) and CFP-negative populations. CFP-positive cells were transplanted postsort, whereas the CFP-negative cells were retransduced and injected at 120 hours. The CFP-negative sorted and retransduced cells contained markedly fewer vector copies and resulted in a 32-fold higher overall engraftment and in a 13-fold higher number of engrafted transgene positive cells. To assess cell proliferation as an underlying cause for the different engraftment levels, carboxyfluorescein succinimidyl ester-labeling of untransduced PBPC was performed to track the number of cell divisions. At 72 hours after initiation of culture, when 95% of all cells have divided, PBPC were sorted into nondivided and divided fractions and transplanted into NOD/SCID mice. Nondivided cells demonstrated 45-fold higher engraftment than divided cells. Late dividing PBPC in ex vivo culture retain high expression of the stem cell marker CD133, whereas rapidly proliferating cells lose CD133 in correlation to the number of cell divisions. Our studies demonstrate that late dividing progenitors transduced with -retroviral vectors contribute most to NOD/SCID engraftment and transgene marking. Confining the -retroviral transduction to CD133-positive cells on days 3 and 4 could greatly reduce the number of transplanted vector copies, limiting the risk of leukemia from insertional mutagenesis.
Disclosure of potential conflicts of interest is found at the end of this article.
|
INTRODUCTION
|
|---|
Hematopoietic stem cells (HSC) maintain hematopoiesis throughout life. To reconstitute the entire blood system following transplantation, HSC must self-renew, proliferate, and differentiate into all blood cell lineages. HSC transplanted by vein into the peripheral blood exit the blood and migrate into the bone marrow (homing). Active interaction with the bone marrow microenvironment is a prerequisite for the maintenance of the stem cell pool and constant blood formation. The reconstitution potential of HSC is dependent upon the source of these cells, with decreasing engraftment potential observed as one compares fetal liver, cord blood, bone marrow, or mobilized peripheral blood stem cells [1, 2]. No ultimate marker has yet been identified that allows isolation of pure pluripotent HSC from humans. CD34 or CD133 (prominin-1, AC133) [3] expression on the cell surface can be used to isolate a cell population that definitely includes pluripotent HSC, since transplantation of isolated CD34+ or CD133+ cells achieves successful life-long hematopoiesis. CD34+ cells are heterogeneous and consist of lineage restricted committed progenitors and a small subset of HSC, which is divided into short-term and long-term repopulating cells. The CD34+ or CD133+ cell population can further be enriched for stem cells by selection of a CD38 negative population [4]. Ex vivo cultivation of HSC leads to a progressive loss of long-term repopulating cells [5, 6]. Cytokine cocktails enable the preservation of the stem cell potential for short-term cultures while promoting the expansion of committed progenitor cells. Long-term culture or significant ex vivo expansion of pluripotent HSC is not feasible under current culture conditions. The cell cycle status of cultured HSC appears to correspond to their stem cell repopulation potential. At initiation of culture, most mobilized CD34+ peripheral blood progenitor cells (PBPC) are in G0 or G1 of the cell cycle. In response to cytokines, the majority of cells go into cell cycle. Compared with CD34+ cells in G0, cells in cell cycle phase G1, and even more so, those in S/G2 phase, demonstrate diminished engraftment potential [7–11]. This is consistent with and equivalent to the observation that rapidly cycling cells are shown to be committed progenitors [5, 6]. Understanding the composition of the HSC compartment during ex vivo culture is critical for clinical applications, where HSC are genetically engineered. In contrast to lentivirus based vectors [12], -retrovirus vectors absolutely require cell division for stable vector integration [13]; therefore, successfully engrafted transgene positive cells must have proliferated during ex vivo transduction. Accordingly, proliferation (defined as at least one cycle of cell division) likely must not itself result in complete loss of engraftment by long-term repopulating cells. For further optimization of gene therapies with -retroviral vectors (and possibly even with lentivectors) in terms of efficiency and safety, it is desirable to enrich a population of cells that proliferates without losing its potential for hematopoietic reconstitution prior to transplantation. To define the relationship among cell proliferation, transduction efficiency, and repopulation potential of human stem cells, we analyzed, separated, and transplanted subsets of carboxyfluorescein succinimidyl ester (CFSE) (5- and 6-carboxyfluorescein diacetate succinimidyl ester) labeled or -retrovirus vector marked PBPC into the nonobese diabetic/severe combined immunodeficient (NOD/SCID) mouse model at different times after initiation of ex vivo culture [14].
|
MATERIALS AND METHODS
|
|---|
Source of Normal CD34+PBPC
After obtaining informed consent (protocols 94-I-0073 and 95-I-0134, approved by the Institutional Review Board of the National Institute of Allergies and Infectious Diseases), three healthy adult volunteers received five daily subcutaneous injections with 10 µg/kg granulocyte colony-stimulating factor (Filgrastim; Amgen, Thousand Oaks, CA, http://www.amgen.com). PBPC were collected by apheresis on day 5 (CS3000 Plus; Baxter Healthcare, Deerfield, IL, http://www.baxter.com), and CD34+ were selected from the apheresis product (ISOLEX 300I; Nexell Therapeutics Inc., Irvine, CA, http://www.miltenyibiotec.com) and aliquots stored in liquid nitrogen. The purity for CD34+ cells after selection was >95%.
Generation of MFGS Vector Encoding Cyan Fluorescent Protein
To generate -retrovirus MFGS-CFP transfer vector, the open reading frame of cyan fluorescent protein (CFP) (Clontech, Palo Alto, CA, http://www.clontech.com) was inserted directionally into the NcoI-BamHI cloning site of the MFGS vector. To obtain a RD114-pseudotyped MFGS-CFP vector, we transfected FLYRD18 packaging cells with MFGS-CFP plasmid, and a high titer FLYRD18 producer clone was selected [5].
Virus Supernatant Production and Ultracentrifugation Concentration
FLYRD18 MFGS-CFP producer cells were plated at 2 x 106 per 185-cm2 flask at 48–60 hours prior to virus supernatant collection. Virus supernatant was collected in serum-free X-VIVO 10 (Cambrex, East Rutherford, NJ, http://www.cambrex.com) with 1% human serum albumin (X-VIVO10/1% HSA). After ultracentrifugation (25,000 rpm [83,000g]; 90 minutes; 4°C), the virus pellet was suspended in 1% of the original volume of fresh X-VIVO10/1% HSA by gently rotating on a horizontal shaker overnight. Virus titer of original neat MFGS-CFP supernatant determined on PBPC was 2 x 106 infectious units.
Transduction of PBPC
Cultures were initiated in RetroNectin (Takara, Otsu, Japan, http://www.takara.co.jp) precoated 6-well plates with 1 x 106 CD34+PBPC in 3 ml of growth medium (X-VIVO10/1% HSA; 50 ng/ml FLT3-ligand [FLT3L], 10 ng/ml stem cell factor [SCF], 10 ng/ml thrombopoietin [TPO], 10 ng/ml interleukin-3 [IL3]) per well. PBPC for NOD/SCID in vivo experiments were transduced at 24 hours, 48 hours, and 72 hours after culture initiation for 12 hours using dilutions of concentrated MFGS-CFP vector equivalent to fivefold concentrated from the original neat supernatant and 5 µg/ml protamine. For a NOD/SCID transplantation study, PBPC were sorted into CFP-positive and -negative cell populations 12 hours after initiation of the third transduction. Although the CFP-positive cells were transplanted directly into NOD/SCID mice, CFP-negative cells were transduced one more time at 96 hours and injected at 120 hours (day 5 of ex vivo culture). Twelve hours after each transduction, cells were transferred to fresh cytokine containing medium. Naïve nontransduced PBPC served as negative control for CFP expression. Cell viability determined by trypan blue exclusion prior to transplantation ranged from 86%–95%.
Analysis of Human Cell Engraftment and CFP-Transgene Expression by Flow Cytometry
Anti-human fluorochrome-conjugated monoclonal antibodies were used to identify human hematopoietic cells (CD45-CyChrome), human neutrophils (CD13-phycoerythrin [PE]), human B cells (CD19-PE), and human CD34+ cells (CD34-fluorescein isothiocyanate). CFP expression was analyzed on a Vantage cell sorter (Becton, Dickinson and Company, Franklin Lakes, NJ, http://www.bd.com) equipped with a Krypton-UV laser (excitation at 413 nm).
TaqMan Analyses
MFGS-CFP vector copy number in genomic DNA of transduced cells was determined by real-time quantitative TaqMan polymerase chain reaction (Applied BioSystems, Foster City, CA, http://www.appliedbiosystems.com). The following primers and probe were used: forward primer GTGAAGGCTGCCGACCC, 6FAM labeled probe TGGACCATCCTCTAGACTGCCATGGC, and reverse primer CTCGCCCTTGCTCACCAT.
PBPC Cell Proliferation Assay and CFSE-Based Cell Sorting
CD34+PBPC were thawed and cell cultures initiated in RetroNectin precoated 6-well plates in growth medium (X-VIVO10/1% HSA; 50 ng/ml FLT3L, 10 ng/ml SCF, 10 ng/ml TPO, 10 ng/ml IL3). Since CFSE labeling resulted in enhanced cell death when performed directly after thawing, CD34+PBPC were pulse labeled with the cell tracker dye CFSE 2 hours after initiation of culture. This latency did not result in any difference with regard to cell division when compared with labeling directly after thawing. Briefly, 5 x 106 cells were incubated with a final concentration of 20 µM CFSE at 37°C [15]. During incubation time, cells were gently vortexed every 2 minutes. After 10 minutes, cells were washed twice with serum containing medium and finally resuspended in serum-free growth medium. A subset of CFSE labeled cells were treated with colchicine (100 ng/ml) and served as a reference for nondividing cells. To improve tracking of cell divisions on subsequent days, for most experiments, cells were sorted on a linear scale 24 hours after initiation of culture to yield a uniformly CFSE labeled cell population [15]. Analyses for cell division were performed each day by flow cytometry until day 5, and cell viability was determined by trypan blue exclusion. Some CFSE labeled cells were also stained for CD34, CD133, and CD38 expression (anti-CD34-allophycocyanin (APC), anti-CD133-APC, anti-CD38 PerCP-Cy5.5) and analyzed by flow cytometry for both cell proliferation and loss of CD34 and CD133 surface markers. For the study with CFSE labeled PBPC, cells were not transduced with the MFGS-CFP vector. To compare the engraftment capabilities of divided and nondivided PBPC, CFSE labeled PBPC were sorted into divided and nondivided cell fractions at 48 hours initiation of culture. Cells that had divided at 48 hours were transplanted into NOD/SCID mice postsort, and the nondivided cells were put back into culture and resorted at 72 hours. Now, both the nondivided and divided cells were separately transplanted into NOD/SCID mice postsort.
Transplantation of PBPC into NOD/SCID Mice
NOD/SCID mice were received from Jackson Laboratory (Bar Harbor, ME, http://www.jax.org) and housed in microisolator cages provided with autoclaved food and acidified water. For transplantation of each mouse, PBPC were injected via tail vein into sublethally irradiated (300cGy) 6–8-week-old NOD/SCID mice. For all studies, a cell aliquot was retained in ex vivo liquid culture for further analysis.
Harvest of Bone Marrow from Mice
Mice were sacrificed 9 weeks post-transplant and bone marrow from tibias and femurs flushed into X-VIVO10/1% HSA. Red cells were lysed with ACK Lysing Buffer (Quality Biological Inc., Gaithersburg, MD, http://www.qualitybiological.com).
|
RESULTS
|
|---|
CFP-Negative Sorted Human PBPC at 84 Hours Have High Engraftment Potential Compared with CFP-Positive Sorted PBPC
MFGS-CFP transduced human PBPC expressed CFP in 96% of cells at 84 hours after initiation of culture (12 hours after the third transduction). In order to express CFP, cells must have divided during the transduction period. A single round of infection of CD34+ cells with this -retrovirus vector was so effective that it transduces almost every dividing cell (i.e., in double labeling experiments with CFSE and MFGS-CFP transduction, virtually all cells in the "divided" group also expressed CFP; data not shown), allowing us to assume that only very few cycling cells escape vector integration. Transduced PBPC were sorted based on CFP expression into CFP-positive and -negative populations (Fig. 1A). The postsort analyses demonstrated CFP expression in all cells that had been sorted for CFP (Fig. 1C), whereas 3% of cells expressed CFP in the CFP-negative sorted group (Fig. 1B). CFP-positive cells were injected into five mice (13 x 106 cells per mouse) directly after the sort. The CFP-negative sorted cells were subsequently transduced a fourth time at 96 hours and injected into five mice at 120 hours (0.6 x 106 cells per mouse). When a small aliquot of the CFP-negative sorted cells without fourth transduction was analyzed at culture day 11, these cells expressed CFP in 54% of cells, likely a result of delayed expression of transgene by cells in which vector integration had not yet been completed or had not yet produced sufficient CFP transgene to be detected (Fig. 1D). CFP-negative cells that received a fourth transduction expressed CFP in 85% of cells (Fig. 1E), whereas CFP-positive sorted cells at 84 hours expressed CFP in 99.7% of cells (Fig. 1F) when analyzed on day 11 of ex vivo culture. Since 99.7% in the CFP-positive sorted cells expressed CFP on day 11, the proportion of CFP-positive cells from nonintegrated vector during the sort must have been low. Average vector copy number per transgene positive cell was 1.4 for CFP-negative sorted cells, 3 for those CFP negative sorted cells after the subsequent fourth transduction, and 11 for the initially CFP-positive sorted cell population, all as analyzed on day 11. The mean fluorescence intensity of CFP expression correlated with the average vector copy number and was 381 for cells with 1.4 vector copies per CFP+ cell, 689 for cells with 3 vector copies per CFP+ cell, and 1,649 for cells with 11 vector copies per CFP+ cell, respectively (Fig. 1D–1F).
View larger version (49K):
[in this window]
[in a new window]
|
Figure 1. CFP-negative sorted cells at 84 hours have high NOD/SCID engraftment potential. Transduced human peripheral blood progenitor cells (PBPC) were sorted at 84 hours based on CFP expression (A). The gray overlay represents nontransduced control PBPC. The postsort analyses are shown to the right and left (B, C). Panels (D–F) show analyses of the progeny of cell aliquots not transplanted but retained in liquid culture and analyzed on day 11. Flow cytometric analyses of CFP expression in CFP-negative sorted cells (54%, mean fluorescence intensity [MFI] 381, [D]), CFP expression in CFP-negative sorted cells with a subsequent additional fourth transduction (85%, MFI 689, [E]), and CFP expression in CFP-positive sorted cells (99,7%, MFI 1,649, [F]) are given. CFP-negative sorted cells with an additional fourth transduction were transplanted at 0.6 x 106 cells per NOD/SCID mouse (G), and CFP-positive sorted cells were transplanted at 13 x 106 cells per NOD/SCID mouse postsort (H) and engraftment analyzed 9 weeks later. In panels (G) and (H), representative analyses are presented, and the percent of human cell engraftment in bone marrow of CFP-negative and CFP-positive cells is shown in the upper quadrants. Although the number of CFP-positive sorted cells transplanted (H) was >20-fold the number of CFP-negative sorted cells transplanted at 120 hours (G), the human cell engraftment (CD45+) was similar (2.3% [range, 0.7%–4.2%] vs. 3.5% [range, 0.9%–5.4%]), demonstrating the engraftment potential for the latter cells. Abbreviations: CFP, cyan fluorescent protein; hrs, hours; NOD/SCID, nonobese diabetic/severe combined immunodeficient; SCC, side scatter; Tx, transduction.
|
|
For the engraftment studies with MFGS-CFP transduced cells, we transplanted the total subgroup of cells obtained postsort subdivided into 5 NOD/SCID mice for each group rather than equalizing the number of sorted cells prior to injection. Human cell engraftment after 9 weeks in mice transplanted with CFP-positive sorted cells averaged 2.3% (range, 0.7%–4.2%), and in this group CFP expression in the engrafted human cells averaged 67% (range, 63%–72%) (Fig. 1H). Human cell engraftment in mice transplanted with CFP-negative sorted cells with a subsequent fourth transduction averaged 3.5% (range, 0.9%–5.4%), whereas CFP expression in these cells averaged 26% (range, 21%–38%) (Fig. 1G). It is important to note that, in this second group (late dividing in culture group), despite the injection of 21.7 times fewer cells per mouse than in the first group (0.6 x 106 vs. 13 x 106 cells), the total human cell engraftment was 1.5-fold greater than that in the first group (3.5% vs. 2.3%). Taking the unequal number of transplanted cell populations into account, the percentage of human cells in the mouse bone marrow was 32-fold higher for the group of late dividing cells (21.7 [0.6 x 106 late dividing vs. 13 x 106 early dividing cells transplanted] x 1.5 [3.5% vs. 2.3%]). There were no significant differences with regard to lineage reconstitution between the two transplanted groups. The average percentage of CFP-positive human CD45 cells in the mouse bone marrow was 0.9% (26% of the 3.5% engrafted human CD45+ cells) when 0.6 x 106 late dividing cells were transplanted compared with 1.5% CFP-positive human CD45 cells in the mouse bone marrow (67% of the 2.3% human CD45+ cells) when 13 x 106 early dividing cells were transplanted. Taking the unequal number of transplanted cell populations into account, the percentage of human CFP+ cells in the mouse bone marrow was 13-fold higher for the group of late dividing cells (21.7 [0.6 x 106 late dividing vs. 13 x 106 early dividing cells transplanted] divided by 1.7 [0.9% vs. 1.5%]).
95% of Human PBPC Divide Within 72 Hours of Ex Vivo Culture
Twenty-four hours after initiation of culture in X-VIVO10/1% HSA supplemented with 50 ng/ml FLT3L, 10 ng/ml SCF, 10 ng/ml TPO, and 1 ng/ml IL3, no cell division based on CFSE labeling was detected. After 48 hours of ex vivo culture, 44% of human PBPC had divided once. After 72 hours and 96 hours, 95% and 99.6% of these cells, respectively, had divided at least once, whereas after 120 hours, almost all cells had divided, suggesting that, after 5 days in ex vivo culture, there are very low numbers of cells that are not eventually responsive to the proliferative stimulus provided by the cytokine cocktail used. The percentage of dividing cells was reproducible with different donor derived CD34+ PBPC and differed only minimally when increasing concentrations of the cytokine cocktail were used. Checkerboard analyses demonstrated that, within the given cytokine cocktail, the presence of IL3 (0.5–50 ng/ml) was the main determinant for cell proliferation, and that CFSE-based cell proliferation was only seen to lower when the IL3 concentration was dropped below 1 ng/ml (data not shown).
Slow Dividing Progenitors in Contrast to Rapidly Dividing Cells Remain CD133-Positive
Whereas both stem cell surface markers, CD34 and CD133, decline on proliferating hematopoietic cells in ex vivo culture, CD34 and in particular CD133 expression remains high on slowly dividing cells even after 5 days of culture (Fig. 2). The slow/late dividing cell fraction correlates best with the CD133+CD38dim/neg expression throughout the 5-day ex vivo culture period. Interestingly, although the CD34+CD38dim/neg expression discriminates well the slow dividing cells from the rapidly dividing cells within the first 3 days of ex vivo culture, the discrimination is lost thereafter. As shown in Figure 2, some CD133-negative cells are also CD38-negative, suggesting that evaluation of CD38 expression during ex vivo culture of PBPC should, if at all, only be done in combination with a second stem cell marker. To illustrate the correlation of CD133 expression and proliferation status in more detail, we present CD133/CD38 expression on 4-day-cultured PBPC with regard to their proliferation status (Fig. 3). Nondivided or once divided cells express high levels of CD133, whereas cells that have divided more than twice have lost CD133 expression.
View larger version (69K):
[in this window]
[in a new window]
|
Figure 2. CD133/CD38 surface staining is superior to CD34/CD38 staining in order to phenotypically discriminate slow dividing from rapidly dividing cells. To analyze the phenotype of proliferating and nondividing hematopoietic progenitors, peripheral blood progenitor cells were CFSE labeled on day 0 and analyzed for CD34/CD38 and CD133/CD38 surface expression, respectively, on days 2–5 of ex vivo culture. Representative analyses are presented. (A): Upper panel shows flow cytometric dot blot analyses of CD133 and CD38 surface expression on hematopoietic progenitors. The percentage of CD133+CD38dim/neg cells declines from 22.7% to 4% between day 2 and day 5. Lower panel shows the corresponding CFSE based proliferation analyses. Filled histograms show the proliferation status of CD133+CD38dig/neg gated cells. The nonfilled histograms represent the nongated cells seen in the respective dot plots. (B): Upper panel shows flow cytometric dot blot analyses of CD34 and CD38 surface expression on hematopoietic progenitors. The percentage of CD34+CD38dim/neg cells declines from 23% to 4.7% between day 2 and day 5. The histogram panel shows the corresponding CFSE based proliferation analyses. Filled histograms show the proliferation status of CD34+CD38dim/neg gated cells. The nonfilled histograms represent the nongated cells seen in the respective dot plots. Abbreviations: APC, allophycocyanin; CFSC, carboxyfluorescein succinimidyl ester; hrs, hours.
|
|
View larger version (55K):
[in this window]
[in a new window]
|
Figure 3. Flow cytometric analysis of CD133 and CD38 expression of CFSE labeled peripheral blood progenitor cells on culture day 4. The CD133+ progenitor phenotype correlates with slow dividing cells. Once cells have divided more than twice, the progenitor phenotype is lost. Abbreviations: APC, allophycocyanin; CFSE, carboxyfluorescein succinimidyl ester.
|
|
Nondivided Human PBPC at 72 Hours Have Higher NOD/SCID Engraftment Potential than Early Dividing Cells
Human cell engraftment in NOD/SCID mice 9 weeks after injection of 0.5 x 106 PBPC that had divided as determined by CFSE labeling at 48 hours or 72 hours after initiation of culture averaged 0.7% (five mice; range, 0.2%–1.2%) and 1.1% (six mice; range, 0.3%–2.2%), respectively (Fig. 4). Transplantation of an equivalent number of cells (0.5 x 106) that had not divided at 72 hours, however, resulted in an average engraftment of 45% (six mice; range, 31%–57%). It is important to note that, for the experiment with CFSE labeled PBPC, cell numbers were equalized in the three groups before transplantation, whereas in the previously discussed experiment (MFGS-CFP transduced PBPC), the proportional ratios of cells in each subgroup were preserved to the time of transplantation. It is of note that those cells that remain undivided until 72 hours based on CFSE labeling correspond well to the cells that were negative for CFP expression at 84 hours in the previous experiment (Fig. 1). Thus, it is not surprising that both populations share high engraftment potential. Consistently, the engraftment level of the CFP-positive sorted cells is similar to the engraftment level of the group of divided cells based on CFSE labeling. There were no significant differences with regard to lineage reconstitution (CD19, CD13) between the transplanted groups.
View larger version (41K):
[in this window]
[in a new window]
|
Figure 4. Peripheral blood progenitor cells that do not divide until 72 hours in ex vivo culture contain most of the NOD/SCID engrafting cells. To analyze the relationship between cell division and engraftment potential, human CFSE labeled peripheral blood progenitor cells were sorted at 48 hours (A). The undivided cells were resorted at 72 hours for those that remained undivided and those that had divided (B). We transplanted 0.5 x 106 cells from each source into NOD/SCID mice. After 9 weeks, engraftment was assessed. Representative analyses are shown. Mice that had received undivided cells demonstrated human cell engraftment in 49% (F). Mice that had been transplanted with cells that divided during the first 72 hours of ex vivo culture showed human cell engraftment of approximately 1% (D, E). Human CD45 staining of wild-type NOD/SCID bone marrow was used to set the gate for human cell engraftment (C). Abbreviations: CFSE, carboxyfluorescein succinimidyl ester; hrs, hours; NOD/SCID, nonobese diabetic/severe combined immunodeficient; SCC, side scatter.
|
|
|
DISCUSSION
|
|---|
There are no phenotypic markers that allow in vitro isolation of a pure and homogenous stem cell population, and there is no ultimate cytokine cocktail that allows preservation or even expansion of human hematopoietic stem cells in prolonged ex vivo liquid culture. We used a well established combination of four cytokines (FLT3L/SCF, TPO, and IL3) at low concentrations under serum free conditions to cultivate PBPC for up to 5 days. The chosen cytokine combination has been shown to preserve cell viability and allowed cell proliferation and maintenance of a cell fraction with SCID-repopulating potential [5]. We used an extremely efficient RD114-pseudotyped -retrovirus MFGS-CFP vector to reproducibly transduce human PBPC to greater than 95% and demonstrate that CFP-negative sorted cells at 84 hours show a 32-fold higher engraftment potential in NOD/SCID mice than CFP-positive sorted cells based on the number of cells injected (Fig. 1). Based on the observation that CFP-positive cells (divided cells) sorted at 84 hours display a markedly reduced engraftment potential compared with CFP-negative sorted cells, we performed a subsequent engraftment study with divided versus nondivided cells. For this experiment, we did not transduce the PBPC but labeled the cells with the cell tracker dye CFSE, which allowed us to investigate the engraftment potential independently from potentially confounding virus vector transductions and transgene expression [16]. We demonstrate that 95% of CD34+ PBPC derived cells divide within the first 72 hours after initiation of culture, and that these cells contribute only minimally to NOD/SCID engraftment (Fig. 4). In contrast, the small fraction of PBPC that divides after 72 hours has high engraftment and reconstitution potential (Fig. 4). These results are consistent with the high engraftment levels of CFP-negative sorted (late dividing) cells at 84 hours and the low engraftment level of CFP-positive sorted (early dividing) cells in the first study. The data presented here corroborate and extend recent studies, which have shown that the more primitive subset of the CD34+ cell population, such as CD34+CD38neg cells, cycles late [17–19]. The CD34+ cell population can be divided into committed cells that divide early, being highly susceptible for transduction, and are hence highly transgene positive at 84 hours with high transgene copy numbers but have low NOD/SCID engraftment potential, and into late cycling cells with high engraftment potential and low copy numbers. Increased failure for CFP-positive and early cycling cells to engraft is most likely due to the committed phenotype of this subset of PBPC. In addition, rapidly cycling cells show enhanced apoptosis upon transplantation [11, 20]. Although transduction rates of greater than 95% of cultured PBPC appear very promising for gene therapy applications, the main determinant for the long-term outcome using -retroviral vectors is not the transduction of the early dividing cells but rather the transduction of the few most primitive cells that divide late. With short-term cultures below 72 hours, these cells would mostly escape -retroviral vector gene transfer.
As demonstrated in our study, flow cytometric cell sorting of late dividing cells is an excellent means to enrich long-term repopulating cells from ex vivo cultivated PBPC, and the sorted cells would be a suitable target population for genetic engineering. As CFSE labeling and fluorescence-activated sorting of late dividing cells is not practical for clinical use, we sought to further define this population with respect to surface markers. Up to day 3 of ex vivo culture, costaining of CD34 and CD38 in CFSE labeled cells showed that late dividing cells are mostly expressing high levels of CD34 and low levels of CD38 (Fig. 2A). But on days 4 and 5, correlation of CD34/CD38 expression and the number of cell divisions are nearly lost. However, costaining of CD133 and CD38 on PBPC that had initially been selected by their CD34 expression enables a far better discrimination of slow and fast dividing cells (Fig. 2B). Even on day 5 of ex vivo culture, we were able to gate a population of CD133+CD38dim/neg cells that contains most of the slow dividing cells. Our presented data are consistent with data from Wagner et al., who reported that CD133 is highly upregulated in slow dividing hematopoietic progenitors [21]. Importantly, after day 4 of ex vivo culture, also the fast dividing CD133-negative population contains CD38dim/neg cells, underlining that, after prolonged ex vivo culture, CD38 expression can only be used in combination with other stem cell markers (Fig. 2). We have shown that the slow dividing cell fraction contains most of the SCID repopulating cells. Since most of the slow dividing cells remain CD133-positive in ex vivo culture in contrast to the rapidly dividing cells, transgene independent magnetic sorting of CD133-positive cells on day 4 or 5 prior to transplantation might be a feasible way to enrich for potent hematopoietic progenitors.
There are several important points to be made from our study, which impacts maneuvers that have been suggested or used to enhance long-term gene marking following transplant of ex vivo transduced HSC. First, our studies confirm that there is a subpopulation of cells with high engraftment potential (the late dividers) that nonetheless do retain relatively high engraftment potential even after going through the one cell division required for successful -retroviral vector transduction. However, our studies also begin to define the biological characteristics of these very important transgene positive cells with high repopulation potential. Moreover, despite the fact that published studies show that lentivirus vectors have the capability of transducing some nondividing cells, at least in the case of human HSC, the presence of growth factors and early phases of progression into the cell cycle clearly have a considerable effect on lentivirus transduction efficiency in HSC. Therefore, our findings likely pertain to some extent to lentivectors as well, although specific studies of this kind with lentivectors need to be done. A second point from our studies is that the heterogeneity of the CD34+ cell population underlines the risk of graft failure and/or poor long-term gene marking when -retrovirus vector transduced transgene positive cells are transgene-dependently enriched/selected ex vivo prior to transplantation at a time when late dividing cells would not have yet expressed the selection gene, eliminating many of those cells most capable for repopulation. It might be conceivable to enhance long-term transduction efficiency with -retrovirus vectors by extending the transduction period until more of the most primitive cells go into cell division. However, extensive cell proliferation using current culture conditions leads to cell differentiation followed by reduced repopulating capability and engraftment [5]. Furthermore, in our transduction experiments we demonstrate that transgene positive late dividing cells contain far fewer integration sites than the early and rapidly dividing cell population. To limit the risk of leukemia from insertional mutagenesis, our results suggest to confine transductions with -retroviral vectors to days 3 and 4 of ex vivo culture to reduce the number of transgene copies in rapidly, early dividing cells. Separation of the late dividing cells from the early dividing cells by CD133 expression levels on day 4 or 5 of ex vivo culture and transplantation of only the late dividing cells that contain low vector copy numbers might further minimize the number of transplanted vector integration sites and thus reduce the risk for side effects related to insertional mutagenesis without substantial loss of engraftment efficiency.
|
DISCLOSURE OF POTENTIAL CONFLICTS OF INTEREST
|
|---|
The authors indicate no potential conflicts of interest.
|
ACKNOWLEDGMENTS
|
|---|
The authors thank Dr. Kevin L. Holmes for his excellent support in flow cytometric cell sorting. S.B. and M.F.R. contributed equally to this manuscript.
|
REFERENCES
|
|---|
- Ueda T, Yoshida M, Yoshino H et al. Hematopoietic capability of CD34+ cord blood cells: A comparison with CD34+ adult bone marrow cells. Int J Hematol 2001;73:457–462.[Medline]
- Holyoake TL, Nicolini FE, Eaves CJ. Functional differences between transplantable human hematopoietic stem cells from fetal liver, cord blood, and adult marrow. Exp Hematol 1999;27:1418–1427.[CrossRef][Medline]
- Corbeil D, Roper K, Hellwig A et al. The human AC133 hematopoietic stem cell antigen is also expressed in epithelial cells and targeted to plasma membrane protrusions. J Biol Chem 2000;275:5512–5520.[Abstract/Free Full Text]
- Yao CL, Feng YH, Lin XZ et al. Characterization of serum-free ex vivo-expanded hematopoietic stem cells derived from human umbilical cord blood CD133(+) cells. Stem Cells Dev 2006;15:70–78.[CrossRef][Medline]
- Brenner S, Whiting-Theobald NL, Linton GF et al. Concentrated RD114-pseudotyped MFGS-gp91phox vector achieves high levels of functional correction of the chronic granulomatous disease oxidase defect in NOD/SCID/beta -microglobulin-/- repopulating mobilized human peripheral blood CD34+ cells. Blood 2003;102:2789–2797.[Abstract/Free Full Text]
- Glimm H, Oh IH, Eaves CJ. Human hematopoietic stem cells stimulated to proliferate in vitro lose engraftment potential during their S/G(2)/M transit and do not reenter G(0). Blood 2000;96:4185–4193.[Abstract/Free Full Text]
- Horwitz ME, Malech HL, Anderson SM et al. Granulocyte colony-stimulating factor mobilized peripheral blood stem cells enter into G1 of the cell cycle and express higher levels of amphotropic retrovirus receptor mRNA. Exp Hematol 1999;27:1160–1167.[CrossRef][Medline]
- Berardi AC, Wang A, Levine JD et al. Functional isolation and characterization of human hematopoietic stem cells. Science 1995;267:104–108.[Abstract/Free Full Text]
- Gothot A, van der Loo JC, Clapp DW et al. Cell cycle-related changes in repopulating capacity of human mobilized peripheral blood CD34(+) cells in non-obese diabetic/severe combined immune-deficient mice. Blood 1998;92:2641–2649.[Abstract/Free Full Text]
- Habibian HK, Peters SO, Hsieh CC et al. The fluctuating phenotype of the lymphohematopoietic stem cell with cell cycle transit. J Exp Med 1998;188:393–398.[Abstract/Free Full Text]
- Jetmore A, Plett PA, Tong X et al. Homing efficiency, cell cycle kinetics, and survival of quiescent and cycling human CD34(+) cells transplanted into conditioned NOD/SCID recipients. Blood 2002;99:1585–1593.[Abstract/Free Full Text]
- Roesler J, Brenner S, Bukovsky AA et al. Third-generation, self-inactivating gp91(phox) lentivector corrects the oxidase defect in NOD/SCID mouse-repopulating peripheral blood-mobilized CD34+ cells from patients with X-linked chronic granulomatous disease. Blood 2002;100:4381–4390.[Abstract/Free Full Text]
- Lewis PF, Emerman M. Passage through mitosis is required for oncoretroviruses but not for the human immunodeficiency virus. J Virol 1994;68:510–516.[Abstract/Free Full Text]
- Larochelle A, Vormoor J, Hanenberg H et al. Identification of primitive human hematopoietic cells capable of repopulating NOD/SCID mouse bone marrow: Implications for gene therapy. Nat Med 1996;2:1329–1337.[CrossRef][Medline]
- Oostendorp RA, Audet J, Eaves CJ. High-resolution tracking of cell division suggests similar cell cycle kinetics of hematopoietic stem cells stimulated in vitro and in vivo. Blood 2000;95:855–862.[Abstract/Free Full Text]
- Hall KM, Horvath TL, Abonour R et al. Decreased homing of retrovirally transduced human bone marrow CD34(+) cells in the NOD/SCID mouse model. Exp Hematol 2006;34:433–442.[CrossRef][Medline]
- Holm M, Thomsen M, Hoyer M et al. Dynamic cell cycle kinetics of normal CD34+ cells and CD38+/– subsets of haemopoietic progenitor cells in G-CSF-mobilized peripheral blood. Br J Haematol 1999;105:1002–1013.[CrossRef][Medline]
- Punzel M, Zhang T, Liu D et al. Functional analysis of initial cell divisions defines the subsequent fate of individual human CD34(+)CD38(-) cells. Exp Hematol 2002;30:464–472.[CrossRef][Medline]
- Reems JA, Torok-Storb B. Cell cycle and functional differences between CD34+/CD38hi and CD34+/38lo human marrow cells after in vitro cytokine exposure. Blood 1995;85:1480–1487.[Abstract/Free Full Text]
- Traycoff CM, Orazi A, Ladd AC et al. Proliferation-induced decline of primitive hematopoietic progenitor cell activity is coupled with an increase in apoptosis of ex vivo expanded CD34+ cells. Exp Hematol 1998;26:53–62.[Medline]
- Wagner W, Ansorge A, Wirkner U et al. Molecular evidence for stem cell function of the slow-dividing fraction among human hematopoietic progenitor cells by genome-wide analysis. Blood 2004;104:675–686.[Abstract/Free Full Text]
Top
Abstract
Introduction
