In a previous investigation, we demonstrated that mesenchymal stem cells (MSCs) actively migrated to cardiac allografts and contributed to graft fibrosis and, to a lesser extent, to myocardial regeneration. The cellular/molecular mechanism responsible for MSC migration, however, is poorly understood. This paper examines the role of CD44-hyaluronan interaction in MSC migration, using a rat MSC line Ap8c3 and mouse CD44−/− or CD44+/+ bone marrow stromal cells (BMSCs). Platelet-derived growth factor (PDGF) stimulation of MSC Ap8c3 cells significantly increased the levels of cell surface CD44 detected by flow cytometry. The CD44 standard isoform was predominantly expressed by Ap8c3 cells, accounting for 90% of the CD44 mRNA determined by quantitative real-time polymerase chain reaction. Mouse CD44−/− BMSCs bonded inefficiently to hyaluronic acid (HA), whereas CD44+/+ BMSC and MSC Ap8c3 adhered strongly to HA. Adhesions of MSC Ap8c3 cells to HA were suppressed by anti-CD44 antibody and by CD44 small interfering RNA (siRNA). HA coating of the migration chamber significantly promoted passage of CD44+/+ BMSC or Ap8c3 cells, but not CD44−/− BMSCs, through the insert membranes (p < .01). Migration of MSC Ap8c3 was significantly inhibited by anti-CD44 antibodies (p < .01) and to a lesser extent by CD44 siRNA (p = .05). The data indicate that MSC Ap8c3 cells, in response to PDGF stimulation, express high levels of CD44 standard (CD44s) isoform, which facilitates cell migration through interaction with extracellular HA. Such a migratory mechanism could be critical for recruitment of MSCs into wound sites for the proposition of tissue regeneration, as well as for migration of fibroblast progenitors to allografts in the development of graft fibrosis.
Fibrosis is one of the cardinal histological changes seen in organ allografts experiencing chronic rejection [1, 2]. Recent studies of animal models have indicated that the majority of intragraft fibroblasts come from progenitor cells of the transplant recipients . Of relevance to this, there are reports of myocardial regeneration derived from extracardiac progenitors in human hearts and in animal models [4, 5]. Influx of mesenchymal progenitors of host origin into allografts is believed to represent a wound healing process secondary to rejection injury [3, 6]. We have hypothesized that the progenitor cells are released from the bone marrow stromal cell (BMSC) compartment into the blood stream, from which they migrate into allografts under the influence of cytokines or growth factors .
It is well established that recruitment of inflammatory cells into allografts involves interaction of multiple adhesion molecules expressed by the migrating cells and their cognate ligands expressed on vascular endothelium. One of the most important adhesion molecules is the hyaluronan receptor CD44 [8, 9]. An increase in levels of CD44 is a hallmark of T-cell activation . The CD44-hyaluronic acid (HA) interaction is used for activated T-cell extravasation into sites of inflammation. Migration of lymphocytes to inflammatory tissues is regulated by cytokines, growth factors, and their receptors .
Little has been known about the mechanism responsible for the recruitment of mesenchymal progenitors that give rise to intragraft fibroblasts . We recently conducted a series of in vitro and in vivo studies to test a hypothesis that fibroblast progenitors are activated by platelet-derived growth factor (PDGF) to migrate into allografts via binding of the cell surface adhesion molecule CD44 to HA of the extracellular matrix (ECM). We examined the expression of CD44 by a rat MSC line Ap8c3, which contains fibroblast progenitors. We studied CD44-HA interaction and its relation to MSC migration. We found that MSC Ap8c3 cells were activated by PDGF to increase expression of cell surface CD44. Adhesion and migration of Ap8c3 cells were dependent on CD44-HA interaction, which can be blocked by either anti-CD44 antibody or CD44-specific small interfering RNA (siRNA).
Materials and Methods
Ap8c3 is a rat mesenchymal progenitor cell line derived from fibroblastic cells recovered from an aortic allograft in an F344 rat. Ap8c3 cells are spontaneously immortalized and contain progenitors to several mesenchymal cell types, including fibroblasts, myocytes, osteoblasts, chondrocytes, and adipocytes . Upon stimulation of transforming growth factor-β (TGF-β), the majority (60%) of Ap8c3 cells express myofibroblast phenotype (α-smooth muscle actin [SMA]+) .
CD44−/− and CD44+/+ BMSCs were obtained from the bone marrow of homozygous CD44-null mice (B6; 129-cd44tmlHbg) and wild-type mice (B6129SF/J), respectively. The adult animals (male, weighing 15–20 g) were purchased from Jackson Laboratory (Bar Harbor, ME, http://www.jax.org). BMSCs were harvested from these animals by flushing large limb bones with a syringe containing Dulbecco's modified Eagle's medium (DMEM) with 15% fetal calf serum (FCS). The recovered cells were cultured in DMEM with antibiotics plus 15% FCS in a 30-mm dish. The cells were periodically washed with DMEM to remove nonadhering cells for the first 72 hours in culture. The resulting dish-adhering, fibroblast-like cells were used for the studies. In differentiation cultures, both CD44+/+ and CD44−/− BMSCs gave rise to multilineages, including adiocytes, chondrocytes, and myofibroblasts (data not shown). Inclusion of mesenchymal stem cells in BMSCs of mouse, rat, and human has been widely reported .
MSC Ap8c3 cells were stimulated with PDGF-AB (20 ng/ml; Chemicon, Temecula, CA, http://www.chemicon.com), monocyte chemotactic protein-1 (MCP-1; 20 ng/ml; Cell Science, Norwood, MA, http://www.cellsciences.com), TGF-β (20 ng/ml; Chemicon), vascular endothelial growth factor (VEGF; 20 ng/ml; Biosource, Camarillo, CA, http://www.biosource.com), fibroblast growth factor-basic (FGF-b; 20 ng/ml; Chemicon), and HA (100 μg/ml; Sigma-Aldrich, St. Louis, http://www.sigmaaldrich.com) for 12 hours in culture medium. The cells were then collected and stained with either a fluorescein isothiocyanate (FITC)-conjugated anti-CD44 antibody (OX50, Chemicon) or an isotype-matched control (Chemicon). Flow cytometric analysis of cell surface CD44 was performed using a FACSCalibur analyzer (Beckon Dickinson Immunocytometry Systems, San Jose, CA, http://www.bd.com).
Quantitative Real-Time Polymerase Chain Reaction
Total RNA from cells was extracted using an RNeasy Mini kit (Qiagen, Valencia, CA, http://www1.qiagen.com). Reverse transcription-polymerase chain reaction (RT-PCR) was performed using 500 ng of total RNA samples with oligo dT primers (Roche Applied Science, Indianapolis, https://www.roche-applied-science.com). Quantitative amplification of the cDNA templates was carried out using a Quantitect SYBR Green PCR kit (Qiagen) in a TaqMan ABI PRISM 7700 real-time PCR system (Applied BioSystems, Foster City, CA, http://www.appliedbiosystems.com). The primer pairs used in this study for amplification of rat CD44 genes encoding the standard and variant isoforms (v1 to v10, respectively) include the following: exons E5 (sense: 5′-AAGACATCGATGCCTCAAAC; anti-sense: 5′-CTCCAGTA GGC TGTGAAGTG; v1 (5′-GCCT-CAACTGTGTACTCAAA; 5′-GGCTATCCTGAGTCT GA GTTG); v2 (5′-GATGACTACCCCTGAAACACC; 5′-TATG AAGATGACTCT TGGACT C); v3 (5′-ACGGAGTC AAAT ACCAACCC; 5′-GGGTATTTGTCTGTTTCAT CTTC); v4 (5′-TGCAACTACTCCATGGGTTT; 5′-GGTGGTTGTCTGAAGTAGTAC); v5 (5′-TATAG ACAGAAACAGCACCA; 5′-CTCTTCATCCTGA TACTCATGG); v6 (5′-TGGGCAGATC-CTAATAGCAC; 5′-TCACATGGGAGTCTTCACTTG); v7 (5′-CTGCCTCAGC CCACA ACAAC; 5′-CGGTCCATGAAACATCCTCTTG); v8 (5′-CCAGTCATAGTAC AACCC; 5′-GTGGTCCTGTCCTGTTCAAGTC); v9 (5′-CAGAACTTC TCTACATTACC; 5′-TAGATGG CAGAACAGAAGTTG); and v10 (5′-GGTCGAAGAAGAGGTGGAAG; 5′-CA AAGACCTC AGTCTTAGCAG). The following conditions were used in real-time PCR: 95°C for 15 minutes and 45 cycles at 94°C for 15 seconds, 55°C for 30 seconds, and 72°C for 30 seconds. Negative control was included in each run in which the template was replaced by an equal volume of water. Amplification of glyceraldehyde-3-phosphate dehydrogenase (GAPDH; 5′-GGCATGGACTGTGGTCATGAGC; 5′-TGCACCACCAA CTGCTTAGC) mRNA was used as the loading control for all the quantitative PCR. The absolute levels of the mRNA expression were normalized with respect to GAPDH mRNA content. Samples per group were run in triplicate. To exclude nonspecific amplification, all amplified samples were examined in electrophoreses on 1% agarose gel.
Cell Adhesion Assay
The in vitro binding of MSC Ap8c3 cells and CD44−/− and CD44+/+ BMSCs to several ECM components was evaluated in a cell adhesion assay. The procedure is adopted with minor modifications from a previous report . Briefly, 96-well assay plates (Corning, Corning, NY, http://www.corning.com) were coated for 24 hours in 37°C with HA (5 mg/ml) or fibronectin (10 μg/ml) (Sigma-Aldrich). The coating was allowed to air-dry and then gently washed twice with phosphate-buffered saline (PBS). Cell cultures were treated with PDGF-AB (20 ng/ml) for 12 hours and were then harvested by treatment of the cells with Trysin/EDTA (Invitrogen, Carlsbad, CA, http://www.invitrogen.com), washed twice with DMEM, and resuspended (5 × 105 cells per ml) in DMEM containing 1% fetal bovine serum (FBS). One hundred microliters of the cell suspension (5 × 104 cells) was fed onto each substrate-coated well and incubated for 30 minutes at 37°C. The nonadherent cells were removed by gentle washes three times with PBS, and the number of adherent cells was estimated indirectly by adding 20 μl of Cell Titer 96 AQ One solution reagent (Promega, Madison, WI, http://www.promega.com). The absorbance at 492 nm was determined using an HTS 7000 Bioassay reader (Perkin Elmer Life and Analytical Sciences, Bucking-hamshire, U.K., http://www.perkinelmer.com). The inhibitory effect of anti-CD44 antibody on the binding of the cells to HA was examined in an antibody blocking experiment in which the cells were pretreated with an anti-CD44 antibody (OX50; Chemicon) for 30 minutes at 37°C before plating on the substrates. Series of antibody concentration (0, 0.01, 0.1, 1, 5, 10, 20, and 30 μg/ml) were tested. A mouse monoclonal antibody (mAb) immunoglobulin G1 (IgG1) (Chemicon) was used as isotype control.
The effect of PDGF on adhesion of Ap8c3 cells and CD44+/+ and CD44−/− BMSCs to HA was studied using the following procedure. Cells in cultures were treated with or without PDGF-AB (20 ng/ml) for 12 hours. The cells were then procured and seeded to HA-coated wells as described above. After incubation of the cells at 37°C for 3 hours, the nonadherent cells were removed by gentle washing. Cell counting was performed under inverted microscope by two researchers separately. Ten randomly selected high-power fields for each sample were examined.
Cell migration was examined in a QCM chemotaxis 96-well migration assay (Chemicon) according to a modification of the manufacturer's instructions. Briefly, the migration chambers (8-μm pore size) were coated with 50 μl of HA (5 mg/ml) and air-dried overnight. Test cells at 70%–80% confluence were procured and resuspended (5 × 105 cells per ml) in DMEM containing 1% FBS and incubated with or without PDGF (20 ng/ml) for 2 hours. One hundred μl (5 × 104 cells) was added to the migration chamber. The lower chamber contained 150 μl of DMEM containing 5% bovine serum albumin. The plates were incubated at 37°C in 5% CO2 for 4 hours. After incubation, cells suspended in media in the migration chamber were gently removed by flipping out the medium. The cells adhering to the top side of the membrane were removed by scratching with a cotton applicator. The migration chamber plate was then placed onto a new 96-well feeder tray containing 150 μl of prewarmed cell detachment solution in the wells. After 30 minutes of incubation at 37°C, 50 μl of a lysis buffer/dye solution was added to the feeder tray and incubated 15 minutes at room temperature. The mixture (150 μl) was then transferred to a new 96-well plate and the plate was read with a fluorescence plate reader using a 480/520-nm filter set (HTS 7000 Bioassay reader). The inhibitory effect of anti-CD44 antibody on the motility of the MSCs was examined in an antibody blocking experiment in which the test cells were pretreated with an anti-CD44 antibody (OX50; Chemicon) for 30 minutes at 37°C before plating into the migration chambers. Isotype-matched IgG1 (Chemicon) was used as a control.
siRNA for rat CD44 was designed from the sequence of the rat CD44 gene obtained from the database of the National Center for Biotechnology Information (NCBI; Bethesda, MD, http://www.ncbi.nih.gov), using programs available online (NCBI accession numbers U96138 [GenBank]). A sequence, 5′-UAG UUAUGGUAACUGGUCCTT-3′ (sense strand), that is complementary to the exon 5 of the rat CD44 gene was chosen and verified in a BLAST (basic local alignment search tool) search of the database. The double-stranded CD44 siRNA and a scrambled control siRNA were purchased from Integrated DNA Technologies (Coralville, IA, http://www.idtdna.com).
Ap8C3 cells at 70%–90% confluence were transfected with CD44 siRNA in 2 ml of complete medium in six-well plates. Transfections were performed with 200 pmol of siRNA using transmessenger transfection reagent (Qiagen) according to the manufacturer's instructions. The cells were then incubated at 37°C in 5% CO2 for 24 hours, replated in 150-mm dishes, and allowed to grow for 24 hours in complete medium. The cells were then harvested and processed for quantitative real-time RT-PCR, flow cytometry, adhesion assay, and migration assay.
Results of experimental data were reported as mean ± SEM. Significance levels were determined by Mann-Whitney, Kruskal-Wallis, or Scheffe test as appropriate (Stat View 5.0; Abacus Concepts, Berkeley, CA, http://www.sas.com). Differences were considered significant at p < .05.
PDGF-Upregulated CD44 Expression by MSC Ap8c3
We hypothesized that mesenchymal progenitors of fibroblasts are activated by pro-inflammatory factors presented in chronic lesions of allografts. The activated progenitors express high levels of cell surface CD44 that mediate the migration of the cells to allograft via interaction with hyaluronan of the ECM. In an initial study, we examined the influence of a number of cytokines and growth factors (including soluble HA, MCP-1, TGF-β, bFGF, VEGF, and PDGF-AB) on cell surface CD44 expression (Table 1). We found that PDGF was the most potent stimulus for CD44 expression. Hence, we focused our studies on CD44 expression by MSC Ap8c3 with stimulation of PDGF.
Ap8c3 cells were cultured in DMEM with 1% FBS with or without addition of PDGF-AB overnight. The cells were stained with an FITC-conjugated mouse anti-rat CD44 mAb (OX50). As shown in Figure 1, Ap8c3 cells expressed low levels of surface CD44 in culture without PDGF (Fig. 1A). The cells increased levels of surface CD44 upon PDGF stimulation (Fig. 1B, 1C). In flow cytometry (Fig. 1D, 1E), the cells treated with PDGF exhibited a 17-fold increase in the levels (mean fluorescent channel shift [MFS]: 539.3 ± 48.23) of surface CD44 when compared with the cells without PDGF stimulation (MFS: 32.31 ± 6.77, p < .01).
Predominance of CD44s Expressed by MSC Ap8c3
A feature of CD44 expression is that the cell surface CD44 could be expressed as a standard form (CD44s) that lacks any variant exons, or variant isoforms, that result from alternative splicing . Because expression of certain CD44 variant isoforms has been implicated in cell adhesion and migration, we examined the expression of CD44 standard and variant isoforms by Ap8c3, using quantitative real-time RT-PCR. As demonstrated in Figure 2A, a 2.8-fold increase in E5 (exon 5) expression (which represents total CD44) was observed in the cells stimulated with PDGF when compared with the control group (p = .0047). Analysis of the expression of variant exons v1 to v10 (Fig. 2B) demonstrated that mRNA copies of v3, v6, v9, and v10 were significantly increased in the cells stimulated with PDGF when compared with the control (p = .042, .009, .014, and .037, respectively). The majority of CD44 was expressed as standard form (Fig. 2C), whereas 13% of the CD44 expressed by unstimulated Ap8c3 cells and 9% in PDGF-treated Ap8c3 cells contained variant exons. No significant differences in distribution of CD44 isoform expression were present between unstimulated and PDGF-stimulated cells (Fig. 2D).
CD44 Mediated MSC Ap8c3 Adhesion to Hyaluronan
To examine whether CD44-HA interaction is involved in MSC adhesion, we tested the adhering ability of Ap8c3 cells onto matrix proteins hyaluronan and fibronectin. As demonstrated in Figure 3, approximately 60% of the cells remained attached to HA or matrix protein fibronectin. Anti-CD44 antibody (OX50) significantly blocked the cell adhesion to HA (p = .017) and fibronectin (p = .02) in a dose-dependent manner. Cell adhesion assay was also conducted on BMSCs derived from CD44-null mouse and its background wild-type strain (Fig. 3C–3E). CD44−/− BMSCs displayed a reduction in adherence to HA, exhibiting a significant difference from that of CD44+/+ marrow stromal cells (p = .003). Anti-CD44 antibody effectively blocked the adhesion of CD44+/+ BMSCs to HA (p = .001).
The effect of PDGF on adhesion of Ap8c3 cells and CD44+/+ and CD44−/− BMSCs to HA is presented in Figure 4. PDGF significantly increased binding of Ap8c3 cells (PDGF vs. nontreatment; p < .001) and CD44+/+ BMSCs (PDGF vs. nontreatment: p < .001) to HA-coated wells while having no effect on adhesion of CD44−/− BMSCs to HA (p = .169). Microscopic observation showed that the majority of MSC Ap8c3 cells or CD44+/+ BMSCs with PDGF stimulation became well attached to HA gel (flat cells). In contrast, only a small proportion of CD44−/− BMSCs attached to HA and became flat cells, indicating that CD44 molecules play a primary role in MSC adhesion to matrix HA.
Next, we assessed whether CD44-HA interaction is involved in MSC migration. The motility of Ap8c3 cells was examined in a migration chamber in which the migrating cells in the insert chamber went through the poles on the membrane precoated with HA gel. The cells landing in the lower chamber or attaching on the other side of the membrane were quantified using fluorescent optical density reader after staining of the cells with a CyQuant GR dye (Chemicon). The results (Fig. 5A) showed that HA coating of the insert chamber significantly enhanced migration of Ap8c3 cells when compared with the cells in the uncoated insert chamber (p = .0017). Treatment of the cells with PDGF promoted cell migration in HA-coated (p = .05) but not in uncoated chambers (p = .12). Anti-CD44 antibody significantly blocked migration of MSC Ap8c3 cells treated with PDGF in the HA-coated chamber (p = .0001) but not in the uncoated chamber (p = .2). As shown in Figure 5B, CD44−/− BMSCs migrated less effectively in the HA-coated chamber when compared with that of CD44+/+ stromal cells (p = .016). PDGF promoted migrations of CD44+/+ significantly (p = .008) but not of CD44−/− BMSCs (p = .09) in HA-coated transverse wells. PDGF did not increase migration of CD44+/+ and CD44−/− BMSCs through uncoated chambers, indicating that PDGF stimulation of CD44 expression by MSC plays an active role in cell migration through HA matrix.
CD44 RNA Silencing Inhibited MSC Ap8c3 Adhesion and Migration to HA
Ap8c3 cells were transfected with 200 pmol of CD44 siRNA or scrambled control RNA. Our initial studies showed that inhibition of CD44 expression by siRNA occurred at 24 hours with treatment, peaked at 48 hours, and lasted for a maximum of 72 hours. The effects of siRNA were then evaluated at 48 hours with the treatment using quantitative real-time RT-PCR, flow cytometry, and cell adhesion and migration assays. As demonstrated by Figure 6, expression of the CD44 gene by the target cells was significantly reduced (Fig. 6A pFigure 6., < .01). Correspondingly, levels of cell surface CD44 detected by flow cytometry (Fig. 6B) were significantly lowered in the siRNA-treated group (n = 6, MFS: 206 ± 30.3) when compared with scrambled controls (n = 6, MFS: 400.6 ± 60.5, p < .001). CD44 siRNA also significantly blocked adhesion of Ap8c3 cells (n = 12) onto HA (Fig. 6C) when compared with controls (n = 12, p < .01). The inhibitory effect of siRNA on cell migration was observed, although to a lesser extent (p = .05).
One of the interesting findings of this study was the upregulatory effect of PDGF on CD44 expression by MSC Ap8c3 cells. PDGF is a potent fibrogenic growth factor that plays critical roles in the induction and progression of fibroproliferative diseases such as atherosclerosis, liver cirrhosis, and allograft fibrosis, with the fibroblasts as its target cell [16–18]. High titers of intragraft PDGF are associated with chronic rejection of cardiac allografts [3, 19]. The current study demonstrated that PDGF is able to activate mesenchymal progenitors as indicated by up-regulation of CD44 expression by MSC Ap8c3 cells after PDGF stimulation. Increase in levels of CD44 is a hallmark of T-cell activation, and the CD44-HA interaction is used for activated T-cell extravasation into sites of inflammation .
A homing mechanism of CD34+ stem cells to bone marrow involving CD44-HA interaction and upregulation by stromal-derived factor-1 was recently described . Similarly, increased CD44 expression by MSC Ap8c3 cells may signal a commitment of the cells to a migratory phenotype.
Upregulation of CD44 expression on MSC Ap8c3 cells stimulated by PDGF was characterized by a preferential use of the CD44 standard form (87% in cells without PDGF stimulation and 91% with PDGF stimulation). Only a small proportion of the CD44 mRNA derived from cells before (13%) and after (9%) PDGF stimulation contained variant exons. Interestingly, copies of CD44 mRNA containing variant exons v3, v6, v9, and v10 were significantly increased in the cells in response to PDGF stimulation, whereas the frequencies (percentage) of each variant exon used by stimulated or unstimulated cells remained statistically unchanged (p > .05). This is because a proportional increase in CD44 standard form coincided with increases in CD44 variant isoforms containing exons v3, v6, v9, and v10 in Ap8c3 cells after PDGF stimulation. The data suggest that increase in CD44 variant isoform expression occurs only in a small subpopulation of the Ap8c3 cells. The numbers of CD44 variant isoforms expressed by Ap8c3 cells in reality may be smaller than that determined by quantitative real-time RT-PCR, in which the primers were designed to amplify individual exons. A CD44 variant isoform may contain one or more spliced variant exons . Expression of certain CD44 variant isoforms was reported as being implicated in tumor cell metastasis . Despite its small proportion, CD44 variants might be important for the support of MSC migration. Our current study can neither include nor exclude this possibility. The biological function of CD44 variant isoforms expressed by the MSC is poorly understood. In view of the protein structure of the extracellular domain of CD44 molecule, it seems likely that expression of variant exons by Ap8c3 cells might have biological functions other than HA binding . Further studies are needed to clarify the role of individual CD44 variant isoform in MSC biology.
Several studies have demonstrated that PDGF can stimulate a variety of target cells to migrate, including smooth muscles and fibroblasts [17, 18]. The role of CD44 in PDGF-promoted cell migration, however, had not been explored. Data derived from cell adhesion experiments indicate that CD44-HA interaction indeed plays a positive role in promoting cell adhesion, which is a critical step for cells to migrate. Anti-CD44 antibody partially blocked the adhesion of Ap8c3 cells to HA or matrix protein fibronectin, which also binds to CD44. The role of CD44 in cell adhesion to HA was also confirmed in the experiments using mouse CD44−/− and wild-type CD44+/+ MBSC.
The results of migration assays demonstrated that CD44-HA interaction is involved in MSC migratory activities. The motility of Ap8c3 cells was significantly promoted by HA coating of the insert membrane. CD44 blocking antibody significantly decreased the numbers of cells that passed through the insert membrane. CD44+/+ BMSCs displayed active migration in HA-coated chambers but reduced migration in uncoated chambers (p < .05), whereas CD44−/− BMSCs exhibited reduced motile activities in both HA-coated and uncoated chambers. These data suggest that CD44-HA interaction facilitates MSC migration. However, our experiments were not able to determine whether CD44-HA interaction directly promotes cell motility or indirectly promotes cell migration by facilitating cell adhesion to HA [22, 23]. HA has been implicated in biological process such as cell adhesion, migration, and proliferation . Within the ECM, HA acts both as a structural element and as a signaling molecule, which binds to more than one cell surface receptor, including CD44, TSG-6, and receptor for HA-mediated motility (RHAMM) .
The results of siRNA experiments further confirm that hyaluronan receptor CD44 plays a role in MSC migration. The siRNA primers we used in this study were designed to target to exon 5 transcripts of the CD44 gene, therefore, silencing all the CD44 transcripts, including standard and variant isoforms. Inhibition of CD44 gene expression by siRNA resulted in reduction in the levels of cell surface CD44 and limitation in MSC adhesion onto HA matrix. Depletion of CD44 expression using RNA interference has been reported to decrease the binding ability of tumor cells to HA . Inconsistent with the inhibitory effect of CD44 antibody on cell migration, CD44 siRNA exhibited a marginal effectiveness on MSC Ap8c3 cells when compared with scrambled control (p = .05). This result suggests that CD44 plays a lesser role in cell motility than in cell adhesion to HA.
Although our data highlight the importance of CD44-HA interaction in MSC migration, the results also indicate that cell surface CD44 is not the sole molecule involved in MSC migration. All the cells tested, including CD44−/− BMSCs, were able to migrate in various degrees in HA-coated as well as uncoated chambers, suggesting the co-action of other migratory mechanism(s). PDGF-stimulated actin rearrangement and edge ruffle formation have been shown to be dependent on activation of PI 3 kinase that interacts with SH2 domain of tensin, leading to cell motility [26–28]. It is also possible that HA promotes MSC migration via interaction with RHAMM [24, 29]. Migrating MSC also use chemotactic mechanisms to promote cell homing to the site of tissue damage .
Taken together, the data presented in this paper illustrate a mechanism that may be responsible for MSC recruitment from blood circulation into tissues. The mechanism may also control the homing of mesenchymal progenitors of fibroblasts to allografts. Fibroblast progenitor cells could be activated by PDGF produced by the graft tissue during chronic inflammation or rejection [3, 19]. PDGF-activated MSCs express high levels of cell surface CD44, which binds to extracellular HA exposed after denudation of capillary endothelium. Thus, CD44-HA binding facilitates MSC migration. This migratory mechanism is further supported by findings from immunohistochemical studies of the chronically rejecting allografts (supplemental online data). Intensive deposit of extracellular HA in allografts was associated with chronic rejection, reflecting a dynamic change in the ECM. Blocking CD44 with low molecular weight HA has been shown to reduce lymphocyte infiltration and prolong cardiac allograft survival in rat models [19, 31]. Molecular or genetic targeting of cell surface CD44 expression by MSC may also have an inhibitory effect on migration of fibroblast progenitors to allografts, thereby reducing graft fibrosis.
Table Table 1.. Differential expression of CD44 by Ap8c3 cells stimulated by selected cytokines
This work was supported by an NIH/National Institute of Allergy and Infectious Diseases grant (1 R01AI05320 0A2) to G.D.W.
The authors indicate no potential conflicts of interest.