SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Disclosure of Potential Conflicts of Interest
  9. Acknowledgements
  10. References
  11. Supporting Information

Duchenne muscular dystrophy (DMD) is a common X-linked disease resulting from the absence of dystrophin in muscle. Affected boys suffer from incurable progressive muscle weakness, leading to premature death. Stem cell transplantation may be curative, but is hampered by the need for systemic delivery and immune rejection. To address these barriers to stem cell therapy in DMD, we investigated a fetal-to-fetal transplantation strategy. We investigated intramuscular, intravascular, and intraperitoneal delivery of human fetal mesenchymal stem cells (hfMSCs) into embryonic day (E) 14–16 MF1 mice to determine the most appropriate route for systemic delivery. Intramuscular injections resulted in local engraftment, whereas both intraperitoneal and intravascular delivery led to systemic spread. However, intravascular delivery led to unexpected demise of transplanted mice. Transplantation of hfMSCs into E14–16 mdx mice resulted in widespread long-term engraftment (19 weeks) in multiple organs, with a predilection for muscle compared with nonmuscle tissues (0.71% vs. 0.15%, p < .01), and evidence of myogenic differentiation of hfMSCs in skeletal and myocardial muscle. This is the first report of intrauterine transplantation of ontologically relevant hfMSCs into fully immunocompetent dystrophic fetal mice, with systemic spread across endothelial barriers leading to widespread long-term engraftment in multiple organ compartments. Although the low-level of chimerism achieved is not curative for DMD, this approach may be useful in other severe mesenchymal or enzyme deficiency syndromes, where low-level protein expression may ameliorate disease pathology.

Disclosure of potential conflicts of interest is found at the end of this article.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Disclosure of Potential Conflicts of Interest
  9. Acknowledgements
  10. References
  11. Supporting Information

Duchenne muscular dystrophy (DMD) is an X-linked myopathy affecting 1 in 3,500 boys. The main genetic defect leads to the near absence of dystrophin, resulting in muscle damage and wasting. Affected individuals suffer from progressive muscle wasting and weakness, which is apparent by 3–5 years of age, and become wheelchair-bound by 12 years of age. Although improvements in multidisciplinary care have extended survival into the third and fourth decades of life, patients generally succumb to respiratory failure or cardiomyopathy [1, 2]. Currently, there is no cure, and thus experimental therapies, such as transplantation of stem cells capable of regenerating damaged muscles, are under intensive investigation [3, [4], [5], [6]7].

Numerous groups have shown that bone marrow (BM) stem cells can participate in muscle regeneration in injury models [8, 9]. In mdx mice (a model of human DMD), transplantation with adult BM led to dystrophin expression 12 weeks later in 10% of muscle fibers [9], whereas adult BM or fetal liver transplanted into embryonic day (E) 14 (fetal) mdx mice resulted in widespread albeit low-level engraftment and myogenic differentiation [10]. Furthermore, a boy with DMD who underwent bone marrow transplantation showed long-term persistence of donor nuclei in muscle [11]. Although the responsible cell type in bone marrow is not known, mesenchymal stem cells (MSC) are implicated by the greater engraftment and higher levels of dystrophin expression seen with use of the MSC-containing fraction [8, 9]. Transplantation of adult human MSCs is reported to lead to incorporation in myofibers in a murine injury model and to proof-of-concept restoration of dystrophin expression in a dystrophic mouse model [7, 12].

The application of myogenic stem cell transplantation for treatment of DMD is hindered by several factors peculiar to DMD. These include the large muscle mass needing to be treated, which makes local injections impracticable, especially to the less accessible cardiac and respiratory musculature. In addition, the use of allogeneic cell types requires long-term immunosuppression to prevent immune rejection of the transplanted cells [13, [14]15].

Because prenatal diagnosis for DMD is well established [16, 17], one way of addressing these issues is to undertake transplantation both in utero and with fetal myogenic stem cells with the ability to cross endothelial barriers; this fetal-to-fetal approach has several advantages over a postnatal approach. First, the fetal environment is highly conducive to expansion of stem cell compartments; indeed, large-scale migration of stem cells occurs naturally only in fetal life, allowing their systemic distribution. Second, there is a huge stoichiometric advantage—the human fetus is only approximately 30 g in size at 13 weeks [18], allowing delivery of proportionately more cells than can be given postnatally. In addition, the immunological naiveté of the early gestation fetus is thought to be predisposed to achieving tolerance to foreign antigens if presented before a critical window, around 12–14 weeks' gestation [19, 20].

The identification of human fetal mesenchymal stem cells (hfMSCs) in the early gestation fetus raises the possibility of using an ontologically related cell type for intrauterine transplantation strategies [21]. hfMSCs have extensive capacity to expand and self-renew every 24–40 hours, reaching 70 population doublings before senescence [21, 22]. In contrast to adult BM-derived MSC, hfMSCs express pluripotency stem cell markers such as Oct-4 and Nanog, have longer telomeres and greater telomerase activity, and senesce later in culture than their adult counterparts [23]. hfMSCs also express several key adhesion molecules, such as CD44, integrins, and CD106, which are implicated in cellular migration across endothelial barriers. In addition, they have recently been shown to differentiate robustly down the myogenic lineage, forming long multinucleated mature myotubes, and to contribute to muscle regeneration in both injured and dystrophic muscle [24], which underlies their potential use for fetal transplantation. Autologous use of hfMSCs may now be possible because ultrasound-guided techniques have already developed to the extent that fetal blood can be sampled in ongoing pregnancies at 12 weeks' gestation with <5% loss rate [25], and advances in imaging and thin-gauge fetoscopy should render hfMSC harvest and intraperitoneal reinfusion feasible in the late first or early second trimester [26, 27]. hfMSCs can be efficiently transduced with integrating vectors without affecting either long-term self-renewal or differentiation capacity [22]. Allogeneic therapy with fetal BM and liver-derived hfMSCs is another promising option because hfMSCs are human leukocyte antigen II (HLA-II)-negative and inhibit lymphocyte proliferation [28]. Moreover, the use of a fetal source of donor cells with relevant developmental ontogeny for intrauterine transplantation may provide an additional advantage over the use of adult cell sources [29, [30], [31]32].

In this report, we investigated a number of routes for transplanting hfMSCs in developing wild-type fetal mice. We followed this up by xenotransplanting hfMSCs into fully immunocompetent dystrophic mdx mice. Our results showed that hfMSCs engrafted in a variety of organs; higher engraftment levels were found in skeletal and cardiac muscle, where degeneration and regeneration cycles are ongoing. In addition, there was evidence of site-specific muscle differentiation and long-term chimerism up to 19 weeks after transplantation. In the context of converging fields of rapid prenatal diagnosis, and advances in gene and cell therapeutic technologies, our observations may have implications for the prenatal treatment of a range of genetic mesenchymal and protein-deficiency disorders.

Materials and Methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Disclosure of Potential Conflicts of Interest
  9. Acknowledgements
  10. References
  11. Supporting Information

Ethics

Fetal blood and tissue sampling was approved by the Institutional Ethics Committee in compliance with national guidelines regarding the use of fetal tissue for research [33]. All women gave written informed consent. All animal procedures were approved in accordance with Home Office Project Licenses (U.K.).

hfMSC Derivation and Differentiation

First trimester (7–13 weeks' gestation) fetal blood was obtained by ultrasound-guided cardiac aspiration before clinically indicated termination of pregnancy. After the procedure, BM was collected from fetal long bones.

Fetal blood was plated at 106 nucleated cells per 100-mm dish and cultured in D10 (10% FBS [Stem Cell Technologies, Canada] in Dulbecco's modified Eagle's medium-high glucose (Sigma, Poole, Dorset, U.K., http://www.sigmaaldrich.com], with 2 mmol/l l-glutamine and 50 IU/ml penicillin/streptomycin (Invitrogen, Paisley, U.K., http://www.invitrogen.com) [24]). Single-cell suspensions of fetal BM were prepared by flushing long bones with 25-gauge needles and plating as for fetal blood. After 3 days, nonadherent cells were removed, and the medium was replaced. Cells were trypsinized at subconfluence and used for subsequent experiments. Immunophenotyping was performed at passage 2 and hfMSCs from passages 3–10 were used in the experiments (at 15 and 40 population doublings). Osteogenic, adipogenic, and myogenic differentiation assays were performed as previously reported [24]. hfMSCs derived from both fetal blood and BM were used in all experiments.

Lentiviral Transduction

Some samples of hfMSCs were stably transduced (98%) with a lentiviral vector to express the enhanced green fluorescent protein (GFP) as previously reported [22].

Animals—MF1 and mdx

MF1 mice (Charles River, Margate, Kent, U.K., http://www.criver.com/) and mdx mice (C57BL10ScSn/DMDmdx) were time-mated and transplanted at E14–16 under inhalational anesthesia (isoflurane). A full-depth midline laparotomy was performed to expose the gravid uterus. Identification of the fetal abdomen, yolk sac vessels, and lower limbs through the translucent uterine wall allowed delivery of hfMSCs by the intraperitoneal, intravascular, and intramuscular routes, respectively. Cells were injected in 20 μl of warm saline using a 33-guage needle (Hamilton, Bonaduz, Switzerland), and the mice were allowed to recover in a warmed cage after closure of the abdominal wound with 6/0 silk sutures.

Transplanted mice were harvested for analysis at various time points until 18 weeks postnatally. At necropsy, murine BM was flushed from the long bones and cultured in D10. Tissues from transplanted MF1 animals were formalin-fixed and paraffin-embedded, with 10-μm sections collected on slides. Tissues from transplanted mdx mice were snap-frozen in cooled isopentane, and 6 μm transverse cryosections were collected on slides. Snap-frozen transplanted mdx muscle was also stored for RNA harvest, as were tissues collected for DNA isolation.

Genomic DNA Isolation

Genomic DNA from either cells or tissue samples was isolated similarly. After incubation overnight in lysis buffer (10 mM Tris, 50 mM NaCl, 5 mM EDTA, 1 mg/ml Proteinase K, and 0.5% SDS) in a microcentrifuge tube, DNA was extracted by standard phenol/chloroform method, precipitated with isopropanol, and quantified spectrophotometrically (Sigma).

Polymerase Chain Reaction and Sequencing

Genomic DNA (20 ng) was added to a polymerase chain reaction (PCR) mixture with specific primers for the detection of human-specific DNA sequences (Table 1) (Invitrogen). Using human/mouse cell dilutions, the sensitivity for detecting human sequences in a murine background was found to be 1 in 10,000 (10−4). Glyceraldehyde-3-phosphate dehydrogenase (GAPDH) primers that amplify both human and murine DNA were used as loading controls, and genomic DNA from hfMSCs and C2C12 murine myoblast cells were used as positive and negative controls, respectively (Table 1). PCR products were sequenced with an automated ABI 3700 sequencer (Applied Biosystems, Foster City, CA, http://www.appliedbiosystems.com).

Table Table 1.. Human and murine specific primers for detection of human DNA sequences and reverse transcription-polymerase chain reaction primers
Thumbnail image of

Reverse Transcriptase PCR

Total RNA from skeletal muscle of intrauterine-transplanted mdx mice was harvested by using RNeasy (QIAGEN, Crawley, West Sussex, U.K., http://www.qiagen.com), according to the manufacturer's instructions. cDNA was obtained by reverse transcription (RT) using 1 mg of total RNA with a kit (Promega, Southampton, U.K.; http://www.promega.com) as described previously [24]. PCRs were carried out using specific primers (Table 1). Negative controls were either RT without enzyme or PCR with Milli-Q (Millipore, Brussels, Belgium, http://www.millipore.com) water instead of cDNA. Amplification of the control cDNA without RT did not generate any products in PCRs (data not shown). Murine GAPDH was used as a loading control, and human fetal muscle (16 weeks' gestation) was used as a positive control.

Immunohistochemistry

Primary antibodies used for immunohistochemical analysis were mouse monoclonal anti-sarcomeric myosin MF20 (Developmental Studies Hybridoma Bank, Iowa City, IA, http://www.uiowa.edu/∼dshbwww), anti-human lamin A/C, anti-human spectrin, anti-human dystrophin DYS3 (Novocastra; Vision BioSystems (Europe) Ltd, Newcastle Upon Tyne, U.K., http://www.vision-bio.com), anti-laminin (Chemicon, Hampshire, U.K., http://www.chemicon.com), and rabbit anti-desmin (Sigma), and anti-GFP (Invitrogen). Immunophenotyping of hfMSCs was performed with mouse monoclonal anti-CD14, CD29, CD31, CD34, CD44, CD45, CD49b/d/e (BD Pharmingen U.K. Ltd. Oxford, U.K.; http://www.bdbiosciences.com) CD105 (SH2), SH3&4, CD106, vimentin, laminin, fibronectin, von Willebrand factor (vWF), HLA-DR and HLA-I (all from Dako UK Ltd., Ely, Cambridgeshire, U.K., http://www.dako.co.uk/, except SH3 and -4 from Osiris Therapeutics Inc., Baltimore, http://www.osiristx.com)

Cells were fixed in methanol/acetone for 5 minutes and blocked with nonserum protein block (Dako UK Ltd.). Thereafter, incubation was with the appropriate primary antibody for 1 hour with subsequent incubations by a secondary fluorophore conjugated antibody (Alexa Fluor 488 & 594; Invitrogen). Slides were mounted with 4,6-diamidino-2-phenylindole (DAPI) (Vector Laboratories, Orton Southgate, Peterborough, U.K., http://www.vectorlabs.com/uk) and analyzed by epifluorescence microscopy.

Cryopreserved tissues were sectioned at 6 μm, collected on slides, air-dried, and blocked with 5% goat and fetal calf serum and papain-digested whole fragments of unlabeled secondary anti-mouse immunoglobulin [34] for 1 hour. Thereafter, tissues were incubated with monoclonal mouse anti-lamin A/C, anti-human spectrin, anti-laminin, anti-human dystrophin (DYS3), and rabbit anti-desmin antibodies at 4°C overnight. A secondary fluorescein-conjugated anti-rabbit IgG (Vector Laboratories) was then used to label the desmin and laminin, and Alexa Fluor 488 or 594 fluorochrome-linked goat anti-mouse IgG (Invitrogen) was used to label lamin A/C, spectrin, and dystrophin. Slides were mounted with DAPI and analyzed as above.

Formalin-fixed, paraffin-embedded sections were dewaxed and rehydrated before treatment with DAPI, and GFP-labeled hfMSCs were identified by epifluorescence microscopy. For immunohistochemical staining, slides were heat-treated before incubation with mouse anti-human vimentin, as above, and subsequent incubations were with a biotin-labeled secondary antibody, followed by the avidin-biotinylated enzyme complex (Vector Laboratories). Finally, slides were incubated with NovaRed substrate (Vector Laboratories) and analyzed under light microscopy.

Fluorescence In Situ Hybridization

XY fluorescence in situ hybridization (FISH) was performed using human-specific centromeric repeat probes DXZ1 (labeled with SpectrumOrange) and DYZ1 (labeled with SpectrumGreen) (Vysis UK Ltd., Richmond, U.K.; http://www.vysis.com/). Tissue sections were dewaxed and heat-treated before treatment with Proteinase K (100 μg/ml; Sigma). Secondary fixation was with 2:1 (vol/vol) methanol/acetone for 2 minutes on ice before being dehydrated. Slides were then hybridized with the probe for 4 hours and mounted in DAPI before analysis by epifluorescence microscopy.

Statistics

Parametric data were analyzed by unpaired t test. A p value of <.05 was considered statistically significant.

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Disclosure of Potential Conflicts of Interest
  9. Acknowledgements
  10. References
  11. Supporting Information

hfMSC Harvest, Differentiation, and Transduction

hfMSCs from fetal blood and BM were recovered as plastic-adherent spindle shaped cells. They expanded rapidly and immunocytochemical staining confirmed their nonhemopoietic, nonendothelial phenotype, being negative for CD34, CD45, CD14, CD31, and vWF but expressing mesenchymal type markers CD105, SH3, SH4, and vimentin and cell adhesion molecules CD29, α2, α4 and α5 integrin (CD49b, CD49d, CD49e), CD44, CD106 (VCAM-1), laminin, and fibronectin. hfMSCs did not express HLA-II and had low levels of HLA-I. Under permissive media, they differentiated readily into osteoblasts, adipocytes, and myotubes as reported previously [21, 22, 24] (data not shown).

Intrauterine Transplantation in Wild-Type Mice

Initially, intrauterine transplantations were performed in out-bred wild-type MF1 mice to examine the fate of transplanted GFP-labeled hfMSCs when administered by different routes.

Intraperitoneal Transplantation

After intraperitoneal transplantation with various doses of hfMSCs (5 × 103–1 × 106), transplanted fetuses (n = 50) were harvested at 30 minutes (n = 3) and 2–21 days after transplantation (n = 20). Fetal viability beyond 2 days after transplantation was 43% (20 of 47); no relationship was demonstrated between the cell dose injected and fetal demise. GFP-labeled hfMSCs were identified within the fetal abdomen 30 minutes after transplantation (Fig. 1A–1C) and appeared as a collection of cells in the abdomen at day 2 (Fig. 1D). There was evidence of hfMSCs spread into the liver substance at day 7 (Fig. 1E), but at longer time points, only a few cells were found within the liver. Transplanted hfMSCs were identified by their GFP expression, by human specific vimentin, or by FISH for human XY chromosomes (Fig. 1G).

thumbnail image

Figure Figure 1.. GFP-labeled human fetal mesenchymal stem cells (hfMSCs) can be identified by immunostaining for GFP ([A], green, ×16) and human vimentin ([B], red; [C] merged, ×16) 30 minutes after intraperitoneal transplantation into embryonic day 14 MF1 mice. Human fetal mesenchymal stem cells aggregate as a collection of cells around the liver by day 2 ([D], red arrowhead, human vimentin stain in brown, hematoxylin nuclear counterstaining, ×4) and can be found in the substance of the liver by day 7 after intraperitoneal transplantation (green vimentin stain [E], and human XY-FISH showing 2 X chromosomes [F], red arrows, blue 4,6-diamidino-2-phenylindole nuclear staining, ×16). Abbreviations: FISH, fluorescence in situ hybridization; GFP, green fluorescent protein.

Download figure to PowerPoint

Intravenous Transplantation

Transplantation of hfMSCs (5 × 104–2 × 105) into the yolk sac vessel (n = 30) resulted in their distribution to multiple tissue compartments. This was evidenced by the appearance of these GFP-positive cells in all tissues examined (the liver, heart, spleen, lungs, orbit, and spinal cord) (Fig. 2A–2F) 30 minutes after transplantation (n = 6). Transplanted hfMSCs were seen coursing through the vascular compartment (Fig. 2G). However, none of the intravenously transplanted fetuses survived even for a short time; resorption or miscarriage of the transplanted fetuses occurred within 2 days of intrauterine transplantation (n = 24).

thumbnail image

Figure Figure 2.. Green fluorescent protein-labeled human fetal mesenchymal stem cells (hfMSCs) (green) engraft widely after intravascular delivery in the liver ([A], white arrowhead, ×16), heart ([B], green cells, ×16), spleen ([C], white arrowheads, ×16), lungs ([D], white arrowhead, ×16), orbit ([E], white arrowhead, ×40), and spinal cord ([F], white arrowhead, ×60) of transplanted embryonic day 14 MF1 mice. They can also be found in the vascular compartment as seen here in the hind limb vessel ([G], green hfMSCs among murine anucleate red blood cells, ×100). Nuclei were counterstained with 4,6-diamidino-2-phenylindole (blue). Abbreviation: mins, minutes.

Download figure to PowerPoint

Intramuscular Transplantation

Because the hind limbs of the fetus can be seen through the uterine wall, hfMSCs (1 × 105–1 × 106 in 5 μl of PBS) were injected in a small volume of saline directly into the left hind limbs. Of the 13 fetuses transplanted, eight were recovered for analysis. The transplanted cells were found mainly over the gluteal muscles and quadriceps, with very few cells found lower in the tibialis anterior and gastrocnemius. These cells aligned themselves along the axis of the muscle fibers just 7 days after transplantation (Fig. 3) and were found up to the latest time point examined at day 20 (day 15 postnatal). However, no evidence of GFP-positive cell incorporation into the host myofibers was detected.

thumbnail image

Figure Figure 3.. Intramuscular injection of green fluorescent protein-labeled human fetal mesenchymal stem cells (hfMSCs) (green) in embryonic day 14 MF1 mice resulted in hfMSC engraftment, with some cells lining up along the axis of the muscle fibers at day 7 ([A] and [B], ×100). Autofluorescence from muscle fibers is picked up in both green and red channels, appearing yellow-orange. Nuclei were stained with 4,6-diamidino-2-phenylindole (blue).

Download figure to PowerPoint

Intrauterine Transplantation in mdx Mice

Having demonstrated the utility of intraperitoneal transplantation for systemic delivery of hfMSCs, we next transplanted hfMSCs in mdx murine fetuses. A total of 413 mdx fetuses at E14–16 from 73 dams were transplanted i.p. with 106 hfMSCs. Forty-four (11.9%) pups were born alive, of which five were harvested at birth; 26 of the remaining 39 survived past weaning and were harvested between 6 and 18 weeks postnatally (n = 31 in total).

Widespread Engraftment in Multiple Tissue Compartments

Human-specific sequences (β2-microglobulin and HLA-DQα) were positive in all tissues examined at all time points except in lungs harvested at birth (Fig. 4A, Table 2). Specificity was confirmed by sequencing the PCR products (Fig. 4B). Engraftment of human cells was further confirmed by immunostaining for human-specific lamin A/C or vimentin within all PCR-positive tissue sections (Fig. 5). The frequency of human cell chimerism was 1–3 per 1,000 (1.48 ± 0.67/1,000) murine cells per nonmuscle tissue section (brain, lung, liver, and spleen), whereas the frequency of human cells in culture-expanded BM adherent cells was one in 2 × 105. Nontransplanted mdx mice were PCR-negative for human specific sequences and did not demonstrate any human cells on immunohistochemistry.

thumbnail image

Figure Figure 4.. Detection of human-specific sequences by polymerase chain reaction (PCR) for human β2-microglobulin from various organs in two transplanted mdx mice (lanes 6 and 7) 13 weeks after intrauterine transplantation of 106 human fetal mesenchymal stem cells (hfMSCs) demonstrating positive signals (162 base pairs) from almost all the organs assayed. Human DNA was used as the positive control, and a nontransplanted mouse DNA was used as a negative control (A). Direct sequencing of the PCR product confirmed the presence of human β2-microglobulin as matched to the human gene at locus DQ217933, indicated by the box (B). Abbreviation: TA, tibialis anterior.

Download figure to PowerPoint

Table Table 2.. Engraftment of human fetal mesenchymal stem cells in various tissues at various time points after intrauterine transplantation in mdx mice, as detected by polymerase chain reaction (PCR) for human-specific human leukocyte antigen-DQα sequences (positive PCR signal/total numbers examined [percentage])
Thumbnail image of
thumbnail image

Figure Figure 5.. Immunohistochemical staining for human-specific antigens was performed to confirm the presence of human cells in polymerase chain reaction-positive tissues from transplanted mdx mice. Human cells were found in adherent cells from the bone marrow, with lamin A/C immunostaining of human nuclei ([A], green, ×100). Human cells were also stained with lamin A/C nuclear staining in the spleen ([B], ×60), lungs ([C], ×16), and liver ([D], ×16) (red nuclei and white arrowhead; laminin staining of the basement membrane in green was used to delineate the architecture of the respective tissues). Evidence of human cell engraftment was also detected by immunostaining for vimentin (cytosolic stain) in the meninges ([E], red, ×100, laminin staining in green). Nuclei were all counterstained with 4,6-diamidino-2-phenylindole (blue). Abbreviation: FITC, fluorescein isothio- cyanate.

Download figure to PowerPoint

Myogenic Differentiation in Cardiac and Skeletal Muscle

Tissues from the skeletal muscle, diaphragm, and heart had PCR evidence of human cell engraftment in 30%–89% of transplanted mice (Table 2), with no association between the time of harvest and the pattern of engraftment. Furthermore, RT-PCR for human-specific myosin heavy chain transcripts demonstrated evidence of human muscle differentiation in transplanted mdx skeletal muscle, but not in nontransplanted mdx skeletal muscle (Fig. 6A). This was confirmed by immunohistochemical staining, demonstrating the co-location of human nuclei (positive for lamin A/C) within newly regenerating muscle fibers, adopting a central location within the fiber (Fig. 6B, 6C). However, immunostaining of the fibers for human dystrophin expression was negative in a number of sections. The frequency of human cells within the tibialis anterior ranged from 5–10 per 1,000 murine cells in PCR-positive muscle and was significantly higher than in nonmuscle tissues (7.05 ± 1.80/1,000 compared with 1.48 ± 0.67/1,000, unpaired t test, p < .0001).

thumbnail image

Figure Figure 6.. Evidence of hfMSC engraftment and differentiation in skeletal and cardiac muscle. (A): Reverse transcription-polymerase chain reaction for human myosin heavy chain and murine glyceraldehyde-3-phosphate dehydrogenase demonstrated presence of human specific myogenic gene transcripts from the tibialis anterior muscles of transplanted mdx mice at 6 weeks of age (lanes 1, 2, and 3) and 12 weeks of age (lanes 4, 5, and 6). Positive control from human fetal muscle (lane 7), negative control from nontransplanted mdx skeletal muscle (lane 8), and a water blank (lane 9). (B, C): Engraftment of human fetal mesenchymal stem cells (hfMSCs) in the skeletal muscle was confirmed by immunostaining transverse sections of the tibialis anterior muscle of transplanted mdx mice which revealed centrally located human nuclei ([B], lamin A/C in red, white arrowhead, ×16, laminin in green counterstaining) within regenerating myofibers. Triple labeling with lamin A/C (red, human nuclei), laminin (red, basement membrane). and desmin (green, myofibrils) demonstrates the central location of human nuclei within the myofibril ([C], ×40). Engraftment of hfMSC in the cardiac muscle can be seen by the incorporation of human nucleus within the cardiac myofibers ([D], human lamin A/C in red, desmin in green, and 4,6-diamidino-2-phenylindole [DAPI] in blue, ×40), and participation in muscle regeneration resulted in the expression of human dystrophin within the heart ([E], human dystrophin DYS3 in red and DAPI in blue, ×40). Abbreviations: bp, base pairs; HC, heavy chain.

Download figure to PowerPoint

The presence of human cells was confirmed in cardiac musculature as nuclei of human origin (human lamin A/C positive nuclei) within cardiac muscle fibers (counterstained with desmin). Immunostaining with the human-specific dystrophin antibody showed several clusters of fibers present within the myocardium (Fig. 6D, 6E). It is noteworthy that not all fibers containing human nuclei expressed human dystrophin.

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Disclosure of Potential Conflicts of Interest
  9. Acknowledgements
  10. References
  11. Supporting Information

Intrauterine hemopoietic stem cell therapy has already been used successfully to treat inherited immune deficiencies with good effect [35, [36], [37]38]. Because of the ability of prenatally transplanted cells to distribute systemically within the host, resulting in prenatal tolerance to foreign antigens [39, 40], it has been suggested that MSC can be similarly transplanted to treat inherited mesenchymal diseases or enzyme deficiency states [41]. Such an approach, which could use either wild-type allogeneic MSC or genetically corrected autologous MSC, should be suitable for DMD, where transplanted myogenic cells could migrate to multiple muscle compartments, engraft, and differentiate into functional muscle, expressing dystrophin with a survival advantage over dystrophin-null myofibers.

Initial transplantation into wild-type MF1 mice was performed to determine the most appropriate route for delivery of myogenic stem cells. The intravascular delivery route resulted in immediate widespread distribution along the vascular tree, with transplanted cells reaching all muscle groups examined. However, none of the transplanted mice survived beyond 30 minutes after transplantation. Although it is not clear what caused this, we speculate that the relatively large size of the ex vivo expanded hfMSCs may have led to plugging of the placental vasculature, resulting in fetal demise. In support of this theory, a recent report documented the intrauterine intravascular transplantation of up to 20 × 106 murine BM cells, which are smaller than culture-expanded hfMSCs, with resulting viable live births [42].

With an intramuscular approach, the transplanted cells were found stretched over a large area around the gluteal and quadriceps region, having migrated with the lengthening limb during fetal growth. There was, however, no evidence of hfMSC participation in muscle formation in this developmental noninjury model, with cells aligning themselves along skeletal muscle but failing to incorporate into myofibers up to 20 days after transplantation. This contrasted with our previous finding of hfMSC participation in muscle regeneration in both postnatal muscle injury and dystrophic models [24]. Furthermore, no transplanted cells were found in the contralateral noninjected legs; hence, this route would lead only to local delivery of cells and would not achieve the systemic delivery needed to treat DMD.

Intraperitoneal delivery for cell therapy has been the most explored route to date; successful passage of hemopoietic cells through the peritoneum into the vascular compartment has been described in several animal models including mice [10, 43], goats [44], baboons [45, 46], and sheep [47, 48], in addition to its historical clinical use in anemic human fetuses [49, 50]. After intraperitoneal delivery into the MF1 fetus, hfMSCs were found distributed as a collection of cells, with subsequent passage into the liver substance after 7 days. We therefore chose this route for transplantation of hfMSCs into the mdx fetus.

After intrauterine transplantation of either hfMSCs derived from either fetal blood or fetal BM in mdx fetuses, we detected engraftment in all transplanted animals using sensitive PCR techniques and immunostaining for human-specific antigens. Human DNA and cells were found in all tissue compartments examined, including the brain, lungs, liver, spleen, BM, tibialis anterior muscle, diaphragm, and myocardium. It is noteworthy that engraftment levels in muscle tissues were significantly higher than that those found in nonmuscle tissues, presumably because of increased recruitment of progenitor cells from the BM or other reservoirs via the circulation in response to ongoing muscle regeneration/degeneration. Such mechanisms of systemic recruitment of stem cells in response to tissue damage have only begun to be elucidated [51, [52]53]. Selective homing of transplanted donor cells to tissues with ongoing injury has also been observed in a report where intrauterine transplantation of un-fractionated whole BM or fetal liver resulted in higher engraftment rates in muscle compared with nonmuscle tissues in dystrophic mdx mice [10]. In addition, Bittner and colleagues demonstrated the recruitment and homing of BM cells to dystrophic but not to normal muscle [54]. More recently, Francois et al. [55] also demonstrated increased recruitment and homing of donor cells to tissues subjected to irradiation injury in immunodeficient mice transplanted systemically with human adult BM-derived MSCs.

We found consistent engraftment of human cells through 19 weeks after transplantation, with all animals at each time point displaying evidence of human cell engraftment in at least one tissue (Table 2). This contrasted with a recent report by another group who transplanted human adult BM-derived MSC into wild-type fetal mice and found that PCR evidence of engraftment dropped from 56% of animals at 1 month to 13% and 0% at 4 and 5 months, respectively, after transplantation [56]. This may be due to differences in the engraftment kinetics of ontologically different primitive hfMSCs versus adult-derived MSCs; fetal cells have an engraftment advantage over adult cells, at least in fetal recipients. A differential engraftment potential of fetal versus adult cells has also been observed with intrauterine transplantation of fetal compared with adult hemopoietic stem cells [29, [30], [31]32]. An alternative reason may be the use of different mouse strains in the two studies.

In contrast to other groups who described the preferential pulmonary trapping of systemically infused MSCs in neonatal [57] or adult mice [58], we found no engraftment in the lungs at birth, even though it was found at all later time points. This may be due to the intrauterine shunting of up to 90% of the blood from the pulmonary artery to the descending aorta via the ductus arteriosus during fetal life, bypassing the lungs [59]. Otherwise, it could be due to the small numbers examined at birth (n = 4).

hfMSCs engrafted in muscle participated in muscle regeneration, demonstrated by both RT-PCR and immunostaining. However, the expression of human dystrophin, a late marker of muscle differentiation, was found only in small clusters within the myocardium but not in skeletal muscle. This may be due to incomplete differentiation of engrafted hfMSCs, as described previously by other authors [60].

Our data demonstrate the ability of hfMSCs to migrate out of the peritoneal cavity and engraft in multiple organ compartments after intrauterine transplantation. In humans, this is supported by the recent report of finding donor hfMSCs in the BM and bone of a prenatally transplanted fetus with osteogenesis imperfecta [61]. Such migration is likely to occur via interactions of various integrins and cell adhesion molecules, such as VLA-4, CD44, and CD106, which are expressed by hfMSCs.

Even though hfMSCs have immunomodulatory properties and do not express HLA-II [28, 62], the possibility of immune rejection of xenogeneic human dystrophin-positive fibers cannot be completely excluded and hence may be responsible for the low levels of chimerism found in muscle. In addition, xenogeneic differences in receptor-ligand interactions may result in suboptimal homing and differentiation cues. Despite this, the level of chimerism achieved here is comparable with that in other reports of transplantation with either hemopoietic [43, 63] or mesenchymal cell populations [10, 64].

In contrast to the intrauterine hfMSC transplantation for osteogenesis imperfecta, where a significant pathologic condition exists during fetal life, and levels of chimerism in the BM reached up to 7.4% in the single clinical case reported thus far [61], the levels of chimerism reported here in the mdx model are very low. This itself could relate to the relatively mild course of DMD during fetal life, with dystrophic processes being established only after a high level of ambulation is ongoing, such as at the time of weaning in mice [65]. Moreover, mdx mice have little clinical muscle weakness and have a near-normal lifespan. The use of a model that approximates the clinical progress of human DMD, such as the utrophin-dystrophin–null mouse [66], or the canine muscular dystrophy model [67], may result in more robust homing, engraftment, and participation in muscle regeneration of transplanted stem cells.

Conclusion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Disclosure of Potential Conflicts of Interest
  9. Acknowledgements
  10. References
  11. Supporting Information

This is the first report of intrauterine transplantation of ontologically relevant hfMSCs into fully immunocompetent dystrophic fetal mice, with their systemic spread leading to widespread long-term engraftment in multiple organ compartments. In addition, there was a selectively higher homing and engraftment frequency in regenerating dystrophic muscle tissues compared with other nonmuscle tissues. Although the level of engraftment (0.5%–1.0%) is unlikely to result in any functional improvement in DMD, where levels of 20%–30% are necessary to ameliorate dystrophic pathologic lesions [68, 69], it might still prove therapeutic in protein deficiency disorders such as lysosomal storage disorders or hemophilia, where the production of low levels of proteins/enzymes may be beneficial [70, 71], or where the intrauterine pathology is severe, such as for osteogenesis imperfecta [61]. Alternatively, a strategy of transplanting hfMSCs predifferentiated down the myoid lineage may lead to higher myogenic engraftment rates as seen after postnatal transplantation strategy as previously reported [24].

Acknowledgements

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Disclosure of Potential Conflicts of Interest
  9. Acknowledgements
  10. References
  11. Supporting Information

This work was funded by the Institute of Obstetrics and Gynaecology Trust at Imperial College London. S.W. received salary support by a Philip Gray Fellowship awarded by the Katharine Dormandy Trust. K.O.D. and P.G. received salary support from Action Research, U.K. C.G. was supported by the MRC, U.K. J.M. is funded by the Medical Research Council (Collaborative Career Development Fellowship in Stem Cell Research) and the Muscular Dystrophy Campaign.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Disclosure of Potential Conflicts of Interest
  9. Acknowledgements
  10. References
  11. Supporting Information

Supporting Information

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Disclosure of Potential Conflicts of Interest
  9. Acknowledgements
  10. References
  11. Supporting Information
FilenameFormatSizeDescription
Collins_Table_1.pdf20KSupplemental Table

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.