|Home | About | Journals | Submit | Contact Us | Français|
Stem cells have considerable potential to repair damaged organs and tissues. We previously showed that prenatal transplantation of human first trimester fetal blood mesenchymal stem cells (hfMSCs) in a mouse model of osteogenesis imperfecta (oim mice) led to a phenotypic improvement, with a marked decrease in fracture rate. Donor cells differentiated into mature osteoblasts, producing bone proteins and minerals, including collagen type Iα2, which is absent in nontransplanted mice. This led to modifications of the bone matrix and subsequent decrease of bone brittleness, indicating that grafted cells directly contribute to improvement of bone mechanical properties. Nevertheless, the therapeutic effect was incomplete, attributing to the limited level of engraftment in bone. In this study, we show that although migration of hfMSCs to bone and bone marrow is CXCR4-SDF1 (SDF1 is stromal-derived factor) dependent, only a small number of cells present CXCR4 on the cell surface despite high levels of internal CXCR4. Priming with SDF1, however, upregulates CXCR4 to increase the CXCR4+ cell fraction, improving chemotaxis in vitro and enhancing engraftment in vivo at least threefold in both oim and wild-type bone and bone marrow. Higher engraftment in oim bones was associated with decreased bone brittleness. This strategy represents a step to improve the therapeutic benefits of fetal cell therapy toward being curative.
Stem cell therapy holds much promise to treat a variety of diseases by capitalizing on the capacity of transplanted cells to migrate toward target areas of injury. There they subsequently differentiate into specific lineages, thereby replacing damaged cells by healthy ones, although some therapeutic benefit may also be mediated via their trophic effects . Adult mesenchymal stem cells (MSCs) found in bone marrow or fat are harvestable, proliferate, do not form tumors, and differentiate into a range of mesodermal lineages such as bone and cartilage. We have previously reported that human first trimester fetal blood mesenchymal stem cells (hfMSCs) have advantageous characteristics relevant to cell therapy compared with their adult counterparts: they have higher expansion potential, grow several times faster, senesce later, are telomerase-active and have longer telomeres, and differentiate more readily into osteoblasts [2, 3]. Osteogenesis imperfecta (OI) is a genetic disorder characterized by multiple fractures starting in utero due to a mutation in the collagen type I gene. oim mice (B6C3fe-a/a-oim), a model of severe OI, have a G deletion at nucleotide 3,983 in COL1A2 resulting in absence of normal heterotrimeric collagen α1(I)2α2(I)1, replaced by homotrimeric α1(I)3, which accumulates in the extracellular matrix . As a result, homozygous oim have brittle bones, multiple fractures, and skeletal deformities. We recently showed that prenatal transplantation of hfMSCs led to a two-thirds decrease in long bone fracture rate, with donor cells preferentially migrating to bone marrow and bones, where they differentiated into mature osteoblasts, producing bone proteins . In an independent study to identify the mechanisms linking cell recruitment to bones to the improvement in bone mechanics, we showed that grafted cells produced collagen type Iα2 protein, which is absent in nontransplanted mice, contributing to modifications of the bone matrix, as evidenced by a reduction of hydroxyproline content (indicating the presence of normal collagen) and by changes in bone crystallinity observed by Raman spectroscopy, subsequently leading to a decrease in bone brittleness and increase in bone pasticity . These results indicate that grafted cells directly contribute, at least partially, to the improvement in bone mechanical properties and stress the importance of donor cell recruitment in bones.
The clinical effectiveness of cell therapy, however, is challenged by the low level of engraftment in target organs. Therefore homing and engraftment of donor cells to injured tissues is one of the hurdles to overcome [5–10], and thus optimizing homing and engraftment is a translational priority. For example, the improvement in skeletal phenotype associated with transplantation of hfMSCs in oim mice was associated with only 3%–5% engraftment levels in bone, with most mice still having fractures [5, 6]. Similarly, Li et al. have reported low and variable levels of engraftment following neonatal transplantation of adult murine MSCs in oim mice, with no report of therapeutic benefit . In humans, Horwitz and colleagues reported <2% engraftment in transplanted OI children, with no sustainable long-term improvements of bone quality [7, 8, 11]. Le Blanc et al. found 0.3%–7% engraftment following prenatal hfMSC therapy in a human OI fetus, but the child still presented fractures despite concomitant biphosphonate treatment . Together both experimental evidence and clinical evidence show that, although cellular therapy for OI is promising due to the large effects linked to minimal engraftment, it is not yet curative .
The mechanisms involved in the homing of donor cells to injured tissue are poorly understood. The signals required for the recruitment of donor stem cells to sites of injury are arguably analogous to the process of leukocyte recruitment from blood into tissue in response to inflammatory stimuli, orchestrated by chemokines, cytokines, and growth factors [13, 14], such as stromal-derived factor (SDF1) [15–17], hepatocyte growth factor (HGF) , basic fibroblast growth factor (bGFG) , platelet-derived growth factor (PDGF) [19, 20], bone morphogenic proteins BMP-2 and BMP-4 , insulin-like growth factor I (IGF-1) , and matrix metalloproteinases (MMPs) .
The importance of the CXCR4-SDF1 pathway has been recently documented by Granero-Moltó and colleagues, who showed that migration of MSCs to fracture site was exclusively CXCR4 dependent . A number of studies have reported strategies to manipulate stem cell homing via manipulation of CXCR4 expression to increase migration, for example, using hypoxic preconditioning with desferrioxamine (DFX) ; IGF-I or IGF-II [21, 24], which have also been shown to enhance expression of the HGF receptor c-Met [25, 26]; or PDGF . CXCR4 expression is also regulated by cytokine treatment, such as tumor-necrosis factor α (TNFα)  and interleukin (IL-6) , which, like IL-1, PDGF, and transforming growth factor β (TGFβ) are released during the early stages of fracture . Critically, however, none of these studies have investigated whether such in vitro manipulation enhances MSC homing in vivo and impacts on disease progression.
Here, we used the oim model to develop a strategy to increase donor cell homing to bone. We showed that hfMSCs present high levels of internal CXCR4 in both the cytoplasm and the nucleoli, with only a small fraction of cells expressing CXCR4 on the cell surface. However, priming hfMSCs with SDF1, oim bone, or oim plasma, which are SDF1-rich, led to upregulated CXCR4 expression and increased the number of CXCR4+ cells. This was associated in vitro with increased chemotaxis to oim and wild-type bone and bone marrow, and in vivo to increased engraftment in both oim and wild-type bones when compared with mice transplanted with nonprimed cells. Increased engraftment also translated into therapeutic benefit, as evident from a further decrease in fracture occurrence, improvement in bone mechanical properties, and decrease in bone brittleness.
Human fetal blood collection  (n = 6, 10+4 to 12+5 gestation age) was approved by the Research Ethics Committee of Hammersmith and Queen Charlotte's Hospital. Blood (gestation 10 weeks+4 days) was collected by cardiocentesis under ultrasound guidance  during clinically indicated termination of pregnancy under general anesthesia. Cells were selected by plastic adherence , cultured in Dulbecco's modified Eagle's medium (DMEM-LG) (Invitrogen, Carlsbad, CA, http://www.invitrogen.com) + 10% fetal bovine serum (BioSera, East Sussex, U.K., http://www.biosera.com/), and studied at passages 4–8. Blood hfMSCs were previously fully characterized [5, 32]. Adult MSCs (passage 3) were from Tulane Center for Gene Therapy (Tulane University, New Orleans, LA, http://tulane.edu).
Extracellular CXCR4 was detected using 10 g/ml mouse monoclonal anti-human CXCR4 (MAB172, R&D Systems, Minneapolis, MN, http://www.rndsystems.com), followed by anti-murine IgG 2B-B fluorescein isothiocyanate (R&D Systems). For intracellular staining, extracellular CXCR4 receptor was blocked with MAB172 (1 hour, 4°C). Cells were then fixed with 4% paraformaldehyde (PFA) (Sigma-Aldrich, St. Louis, MO, http://www.sigmaaldrich.com), permeabilized with 0.5% Triton X-100 (Sigma-Aldrich), stained with anti-human CXCR4-phycoerythrin (MAB172-B, R&D Systems), and analyzed by FACSCalibur flow cytometry (Becton Dickinson, Franklin Lakes, NJ, http://www.bd.com).
Cells grown exponentially were fixed in 4% PFA in 125 mM HEPES (pH 7.6; 10 minutes, 4°C), then in 8% PFA in the same buffer (50 minutes, 4°C), and permeabilized in 0.5% Triton X-100 in phosphate-buffered saline (PBS) (30 minutes, gentle rocking) as previously described . After fixation and permeabilization, cells were rinsed (three times) in PBS, incubated (30 minutes) with 20 μM glycine in PBS, blocked (1 hour) with PBS+ (PBS supplemented with 1% bovine serum albumin [BSA], 0.2% fish skin gelatin, 0.1% casein; pH 7.6), incubated (2 hours) with primary antibodies in PBS+, washed (five times over 1.5 hours) in PBS+, incubated (1 hour) with secondary antibodies in PBS+, washed (overnight, 4°C) in PBS+, and rinsed (three times) in PBS before being mounted in VectaShield labeled with 4′,6-diamidino-2-phenylindole (DAPI; Vector Laboratories, Burlingame, CA, http://www.vectorlabs.com) and visualized immediately. For fluorescence microscopy (Zeiss Axioscope I microscope [Hertfordshire, U.K., http://www.zeiss.com/], equipped with charge-coupled device [CCD] camera and iPlab software [Exton, Pennsylvania, http://iplab.net/]) images were collected sequentially (TIFF files). Raw TIFF images were collected without saturation of the intensity signal and analyzed in Adobe Photoshop (Adobe Systems, San Jose, CA, http://www.adobe.com/) without further thresholding or filtering (e.g., no background subtraction). The cells were immunostained with primary mouse CXCR4 (MAB172, clone 44716, R&D Systems) at 1:50 dilution. Secondary antibodies for immunofluorescence were donkey anti-mouse conjugated with fluorescein isothiocyanate (FITC) (1:100, multiple-labeling grade; Jackson ImmunoResearch Laboratories, West Grove, PA, http://www.jacksonimmuno.com).
Cells were trypsinized, washed (PBS), and resuspended in DMEM + 0.5% BSA (107 cells/ml). hfMSC suspension (100 μl) was placed in the upper compartment of a chemotaxis chamber. Chemoattractants (SDF1, bone marrow, bone cells or plasma from oim or wild-type mice, or DMEM-0.5% BSA) were placed in the lower compartment, separated by an 8-μm polycarbonate filter (Neuroprobe, Gaithersburg, MD, http://www.neuroprobe.com/). Cells were allowed to undergo chemotaxis (1 hour), and the filter was then removed, washed, fixed, and stained (1% crystal violet) (Sigma-Aldrich). Ten random fields were counted at ×40 magnification by a blinded observer (triplicates). The migration index was calculated as the ratio of the number of cells migrating toward the chemoattractant to the number of cells migrating toward media alone. For blocking controls, cells were preincubated with AMD3100 (Sigma-Aldrich) at 100 μg/ml for 30 minutes at 37°C and then used in the chemotaxis assay. For priming before chemotaxis, cells were incubated with either oim or wild-type plasma for 30 minutes at 37°C and then used in the chemotaxis assay.
All experimental protocols complied with Home Office guidelines (PPL 70/6857). Heterozygous male and female (B6C3Fe a/a-Col1a2oim/ Col1a2oim) mice (Jackson Laboratory) were housed in filter cages with a 12/12-hour light/dark cycle (21°C), with water and chow (Purina, Croydon, U.K., http://www.purina.com) ad libitum. Offspring were genotyped by sequencing the oim fragment , and homozygous colonies were established. Progeny were weaned at 30 ± 1 days and culled at 8 weeks of age.
Blood hfMSCs (106 cells, nonprimed or primed with either SDF1 or oim or wild-type plasma for 1 hour) were injected intraperitoneally in oim (n = 20) or wild-type neonates (n = 20) 2–3 days after birth. The mice were culled at 8 weeks of age for analysis.
Fractures in femur, tibia, and humeri of both body sides were assessed at 8 weeks of age from whole-body x-rays and confirmed after dissection by two independent observers blinded as to whether or not each particular mouse was transplanted. Fractures were delineated by callus formation, and each bone was graded visually as fractured or intact. The number of mice with at least one fracture and the fracture incidence (number of fractured bones/total bones assessed) were calculated.
Total RNA was extracted using TRIzol (Invitrogen), and cDNA was synthesized using Pd(N)6 random hexamers (Amersham, Piscataway, NJ, http://www.amersham.com) and 1 μl of 200 U M-MLV reverse transcriptase (RT) in the presence of deoxyribonucleotide triphosphates (dNTPs) (Promega Corp., Madison, WI, http://www.promega.com) (10 minutes, 25°C; 60 minutes, 42°C; and 10 minutes, 75°C). Gel based RT polymerase chain reaction (RT-PCR) was carried out for the detection of CXCR4 expression. The PCR mix (20 μl per sample) consisted of 12.7 μl diethylpyrocarbonate (DEPC)-treated water (Invitrogen), 2 μl 10× NH4 buffer (Bioline, Tauton, MA, http://www.bioline.com), 0.6 μl 50 mM MgCl2 (Bioline), 2 μl 2.5 mM dNTP (Promega), 20 μM each of forward and reverse primers as described elsewhere (5′-ACGTCAGTGAGGCAGATG-3′; 5′-GATGACTGTGGTCTTGAG-3′) , 0.2 μl BioTaq DNA polymerase (Bioline), and 2 μl cDNA. The reaction was run in a Peltier Thermal cycler (MJ Research, St. Bruno, Quebec, Canada, http://www.mj-research.com/). The amplified products (202 bp) were run on a 2% agarose gel in 1× tris-base, acetic acid, ethylenediaminetetraacetic acid (TAE) buffer containing ethidium bromide for 1 hour at 100 V, followed by visualization under UV light.
Quantitative RT-PCR (QRT-PCR) was performed with the ABI Prism 7700 Sequence Detector (Applied Biosystems, Foster City, CA, http://www.appliedbiosystems.com), and all reactions were carried out in duplicate in a total volume of 25 μl. We used primers amplifying sequences of the β-actin gene (GenBank accession number: NM_001101) present in humans and not in mouse to determine the amount of human cellular cDNA in samples (primer specificity confirmed by absence of amplification of mouse cDNA) and primers common to both human and mouse to determine the total cDNA in each sample, as previously described [5, 6, 32]. For both sets, the absence of dimer formation was confirmed using Dissociation Curves 1.0 software (Applied Biosystems). Human/mouse chimerism was estimated as a ratio. Serial dilution of human in mouse cells formed the calibration curves. The primer sequences were as follows: h-specific, 5′- CTGGAACGGTGAAGGTGACA-3′ and 5′- AAGGGACTTCCTGTAACAATGCA-3′; nonspecific, 5′-GCTCCTCCTGAGCGCAAGTA-3′ and 5′-GATGGAGGGGCCGGACT-3′.
Bones were fixed for 48–72 hours in 10% neutral buffered formalin and decalcified in 10% formic acid and 10% neutral buffered formalin at room temperature. Bones were embedded in paraffin, and 3-μm sections were cut onto Superfrost slides (Sigma-Aldrich), deparaffinized in xylene, and rehydrated. Heat-induced epitope retrieval was performed in a steamer (Dako, Glostrup, Denmark, http://www.dako.com). Donor cells were visualized using a mouse monoclonal vimentin (Abcam, Cambridge, U.K., http://www.abcam.com) primary antibody. Staining specificity was verified using appropriate negative controls (vimentin on nontransplanted and isotype antibody controls on transplanted tissues) and positive controls.
Three-point bending tests were performed on 8-week-old unfractured femurs, fresh-frozen and thawed prior to testing using a standard materials testing machine (5866, Instron, Norwood, MA, http://www.instron.com). Bones were placed on their posterior side on two supports 9 mm apart and tested until fracture at the mid-diaphysis with a loading rate of 50 μm/second. Force-deflection curves were analyzed with a custom program (Matlab, MathWorks, Natick, MA, http://www.mathworks.com) to measure the bending stiffness (S, slope of the linear elastic deformation), the yield force (F yield, limit between the elastic and plastic deformation), the ultimate force (Fult, maximum force sustained), the total work to fracture (total area under the curve), and the plastic work to fracture (area under the curve from the yield point to fracture). The plastic (postyield) behavior was assessed by the ratio of plastic work to total work to fracture.
Data were expressed as mean ± SEM (standard error) or median and range. Parametric statistics were applied after confirming normal distributions on histograms and unpaired two-tailed Student t-testing used for comparison between groups. Nonparametric data were compared by the Mann-Whitney test, whereas the two-tailed 2 × 2 Fisher exact was used for categoric comparisons. p < .05 was considered significant.
CXCR4 was expressed by hfMSCs at the transcript level but was below detection levels in adult MSCs, as reported elsewhere (Fig. 1A) . At the protein level, 78 ± 5.2% of the cell population had high levels of intracellular expression of CXCR4 (Fig. 1B), with the protein localizing in both the cytoplasm and the nucleoli (Fig. 1C). In contrast, only 23 ± 2.1% of the cells expressed CXCR4 on the cell surface (Fig. 1B). Cells showed a robust chemotactic response (relative to DMEM alone) toward gradients of SDF1 (Fig. 1D), with maximal migration after 1 hour exposure to 30 ng/ml SDF1 (Fig. 1E), indicating functional CXCR4 receptors.
Since we previously found that transplanted blood MSC preferentially migrated to bones rather than to other organs in oim mice with higher engraftment in oim bones compared with wild-type bone but similar engraftment in both oim and wild-type bone marrow , we first determined SDF1 expression levels in oim and wild-type organs using real-time QRT-PCR. SDF1 mRNA levels were higher in bone compared with all other organs analyzed (kidney, testes, spleen, brain, liver, heart, lung, and bone marrow, p < .01) (Fig. 2A). Compared with wild-type, SDF1 was more expressed in oim bones (p < .01) but showed similar levels in oim and wild-type bone marrow (Fig. 2A). Consistent with this, the in vitro chemotactic response of hfMSCs was greater toward oim bone than wild-type bone (p < .01), whereas chemotaxis to bone marrow was similar for oim and wild-type (Fig. 2B). Incubation of hfMSCs with the CXCR4-blocker AMD3100 at levels blocking the SDF1 response partially abrogated migration to oim and wild-type bone (p < .01) and bone marrow (p < .01), confirming that the SDF1-CXCR4 pathway was integrally involved in donor cell homing to bone and bone marrow (Fig. 2B).
Priming for 1 hour with SDF1, plasma, or SDF1-rich tissue (bone) from oim, but not from wild-type mice, led to substantial upregulation of CXCR4 at the transcript level (p < .01) (Fig. 3A) and an increase in the cell fraction expressing CXCR4 at the cell surface (Fig. 3B). Stimulation with wild-type bone led to a lesser albeit still significant increase in CXCR4 mRNA expression, with minimal effect on increasing the number of cells expressing CXCR4 on the cell surface (Fig. 3B).
As expected from the effects of SDF1 priming on CXCR4 expression, chemotaxis to SDF1 increased when the cells were primed with SDF1-rich oim plasma or SDF1 (p < .01) but not with wild-type plasma (Fig. 3C). Similarly, migration to both oim and wild-type bone and bone marrow increased (p < .01 for all) when the cells were primed with either SDF1 of oim plasma but not with the plasma of noninjured wild-type mice (Fig. 3C). Interestingly, migration to wild-type bone with SDF1-primed cells was increased to a level of chemotaxis even significantly higher than that obtained to oim bones using nonprimed cells (p < .05) (Fig. 3C).
To validate our in vitro data, unprimed or SDF1-primed hfMSCs (106 cells) were transplanted intraperitoneally in oim neonates (n = 18 and 22, respectively), and engraftment in epiphysal femur was quantified in 8-week-old mice using quantitative real-time RT-PCR with human-specific and human/mouse-nonspecific primers. When the donor cells were not primed, engraftment was higher in oim bone than in wild-type bone (p < .01) and did not significantly differ for bone marrow, as previously reported  (Fig. 4A). However, when the cells were SDF1-primed, engraftment in both oim and wild-type bone increased (p < .01), with engraftment in wild-type bone reaching similar levels to those reached in oim bone transplanted with nonprimed cells (Fig. 4A). In bone marrow, engraftment to oim and wild-type bone marrow was increased with SDF1-primed cells (p < .01) but was not significantly different from each other (Fig. 4A). Increased donor cell engraftment in oim and wild-type bone and marrow using nonprimed or SDF1-primed cells is illustrated in Figure 4B.
The percentage of mice with at least one fracture in the long bones tibia, femur, and humerus on both sides was decreased from 100% (16/16 mice) in nontransplanted oim to 27.8% in oim mice transplanted neonatally with unprimed cells (hfMSC-UP) (5/18 mice, p = .0001, two-tailed 2 × 2 Fisher exact) and 18.2% in oim mice transplanted with SDF1-primed cells (hfMSC-CXCR4+) (4/22 mice, p = .0001) (Fig. 5A). Although the percentage of mice with fracture was lower in hfMSC-CXCR4+ cells compared with unprimed cells, the difference was not significant (p = .7). Long bone fracture rate (number of fractured bones/total number of bones) was significantly decreased in oim mice transplanted either with hfMSC-UP (8.33% fracture rate, n = 108) or hfMSC-CXCR4+ (4.54% fracture rate, n = 132) compared with nontransplanted mice (27.08% fracture rate, n = 96) (p < .0001 for both), corresponding to a 69.23% decrease in fracture rate for hfMSC-UP and 83.22% for hfMSC-CXCR4+ (Fig. 5B). Although fracture reduction was greater in oim mice transplanted with hfMSC-CXCR4+ compared with hfMSC-UP, the difference did not reach significance (Fig. 5B).
Comparison of the bone mechanical properties in mice transplanted neonatally revealed a benefit of the transplantation with hfMSC-CXCR4+ compared with hfMSC-UP. Both total work to fracture and work from yield to fracture, which we previously reported as being significantly lowered in oim compared with wild-type mice , were higher in hfMSC-CXCR4+ compared with hfMSC-UP treated mice (1.7 ± 0.2 J, n = 14 vs. 1.0 ± 0.1 J, n = 15; p < .05 for total work to fracture; and 1.4 ± 0.2 J vs. 0.8 ± 0.1 J; p < .01 for work from yield) (Fig. 5C, C,5D),5D), indicating that transplantation of hfMSC-CXCR4+ led to better bone quality and improved bone plasticity as the bones require more energy to fracture and are more resistant to fracture after bone deformation. In terms of other parameters, ultimate force, yield force, and stiffness, all parameters of bone strength, were not different between hfMSC-CXCR4+, hfMSC-UP treated mice, and nontransplanted mice, similar to our previous findings after in utero transplantation of hfMSCs .
We previously showed that the therapeutic benefits of hfMSC transplantation in a mouse model of osteogenesis imperfecta were mediated, at least partially, by the recruitment of donor cells to bones where they differentiated into osteoblasts producing bone matrix proteins, thereby modifying bone matrix composition and subsequently improving bone mechanical properties, decreasing brittleness, and ultimately reducing fractures. Addressing one of the main challenges of stem cell therapy, we developed a strategy to increase donor cell homing and engraftment in bone. Priming donor cells with SDF1 not only increased CXCR4 expression on the cell surface and enhanced in vitro migration to SDF1-rich tissues including bones but translated in vivo into enhanced bone quality, enhanced bone plasticity, and a trend to lower fracture rate in oim mice.
It has been recently hypothesized that the SDF1-CXCR4 axis, as implicated in the engraftment of hemopoietic stem/progenitor cells , may also be involved in recruitment of MSCs to damaged tissues [15, 22, 35]. A number of studies have shown lack of CXCR4 expression in adult MSCs , although this has not been a consistent finding . In this study, we show that, unlike adult MSCs, all hfMSCs express CXCR4 and have intracellular CXCR4, with only a small proportion of cells functionally expressing the receptor on the cell surface. Interestingly, priming of the cells with SDF1 increased the proportion of cell surface CXCR4+ cells, which correlated with enhanced chemotaxis to SDF1 and oim tissues. This result is especially relevant in the context of prenatal cell therapy, which aims at not only repairing de novo fractures but also preventing fracture occurrence by transplanting the cells into fetuses or neonates to contribute to chimeric bone formation. We have previously shown in vivo that engraftment was almost nonsignificant in the absence of injury and very low when injury was present [5, 38]. Here we show that SDF1 priming allows transplanted cells to migrate and engraft in wild-type bones in the absence of injury, thereby being able to contribute to de novo bone formation when injected during development. Altogether, our results implicate the SDF1-CXCR4 axis in the chemotactic response of fetal blood MSC to bone and bone marrow and indicate that SDF1 and/or other factor(s), possibly cytokines, increases migration of cells to healthy or damaged bone cells and bone marrow by externalization of the CXCR4 epitope. This is in line with recent data showing inhibition of bone formation upon SDF1 blockade in mice  and increased bone marrow engraftment of adult MSCs after forced overexpression of CXCR4 . Moreover, in a stabilized fracture model, Granero-Moltó and colleagues recently showed that migration of mouse adult bone marrow MSCs to fracture sites was CXCR4 dependent .
These findings of increased engraftment after physicochemical priming have yet to be tested in other sources of fetal or adult MSCs, but we note that adult MSCs did not express CXCR4 transcript. We chose fetal blood hfMSCs for two reasons. Pragmatically, these had been used in both our previous studies investigating hfMSC transplantation in oim mice. More importantly, the first trimester is the only time in human life when hfMSCs circulate with frequency, when they are implicated in the developmental shift from hepatic to myeloid hemopoiesis . Thus, circulating blood hfMSCs may be functionally susceptible to strategies to facilitate engraftment. Notably our approach did not involve untranslatable strategies such as viral transduction.
Our previous studies on hfMSC transplantation in the oim model have shown that the therapeutic effects of fetal stem cell therapy were mediated by the contribution of donor cells to bone formation and their differentiation into mature osteoblasts, producing the protein absent in nontransplanted mice and thereby modifying bone structure to reduce bone brittleness [5, 6]. In conclusion, the present data confirm these results and show that the therapeutic benefits of cell therapy for OI in terms of fracture rate, bone quality, and plasticity might be improved by increasing donor cell migration and homing to intact and injured bones by in vitro priming of the cells with either SDF1 or autologous plasma.
Mesenchymal stem cells transplanted in early life may prevent organ injury before irreversible damage occurs. Human first-trimester fetal blood MSCs transplanted pre- or neonatally in a mouse model of osteogensis imperfecta reduce fracture rate and improve bone mechanics by contributing to chimeric bone formation during early development, but it is not curative, possibly related to the low levels of donor cell engraftment in the absence of injury. In this study, we show that donor cell priming with SDF1 upregulates CXCR4 on the cell surface of donor cells, increasing engraftment in bones and bone marrow and improving bone mechanics in oim mice. This strategy represents a step to improve the therapeutic benefits of fetal cell therapy.
This research was funded by the Genesis ResearchTrust, the Henry Smith Charity, and Action Medical Research. P.V.G. was supported by Research Council United Kingdom. D.M. was supported by Kidney Research United Kingdom. G.N.J. was supported by the Medical Research Council. H.A. was supported by Action Medical Research. P.d.C. was supported by Great Ormond Street Hospital Children's Charity. J.P. was supported by the Rosetrees Trust.
D.M., G.N.J., K.L., H.A., and M.V.: collection and/or assembly of data, data analysis and interpretation; S.J.S., J.P., P.D.C.: financial support, final approval of manuscript; N.M.F.: provision of study material, manuscript writing, final approval of manuscript; P.V.G.: conception and design, financial support, manuscript writing, final approval of manuscript. G.N.J. and D.M. contributed equally to this work.
The authors declare no competing financial interests.