Mesenchymal stem cells from cortical bone demonstrate increased clonal incidence, potency, and developmental capacity compared to their bone marrow-derived counterparts

Abstract

In this study, we show that matrix dense cortical bone is the more potent compartment of bone than bone marrow as a stromal source for mesenchymal stem cells as isolated from adult rats. Lineage-depleted cortical bone-mesenchymal stem cells demonstrated >150-fold enrichment of colony forming unit-fibroblasts per cell incidence. compared to lineage-depleted bone marrow-mesenchymal stem cells, corresponding to a 70-fold increase in absolute recovered colony forming unit-fibroblasts. The composite phenotype Lin(-)/CD45(-)/CD31(-)/VLA-1(+)/Thy-1(+) enriched for clonogenic mesenchymal stem cells solely from cortical bone-derived cells from which 70% of clones spontaneously differentiated into all lineages of bone, cartilage, and adipose. Both populations generated vascularized bone tissue within subcutaneous implanted collagen scaffolds; however, cortical bone-derived cells formed significantly more osteoid than bone marrow counterparts, quantified by histology. The data demonstrate that our isolation protocol identifies and validates mesenchymal stem cells with superior clonal, proliferative, and developmental potential from cortical bone compared to the bone marrow niche although marrow persists as the typical source for mesenchymal stem cells both in the literature and current pre-clinical therapies.

Keywords: Stem cell; bone marrow; bone regeneration; colony forming unit–fibroblasts; cortical bone; mesenchymal stem cell; prospective isolation; tissue engineering.

Figure 1.

The BM and CB compartments and incidence of primary and secondary CFU-F. (a) H&E stain of femur showing adjacent CB and BM compartments of bone. The BM is dense with nuclei, contrasted with the matrix dense CB, including a row of endosteal cells (e) contiguous with the BM and perivascular cells surrounding the blood vessels (BV) (20×). (b) Crushed bone chip pieces pre- and post-enzymatic digestion; arrow points to the abraded bone edge on a post-digestion piece (20×, scale bar: 20 µm). (c) Incidence of CFU-F from BM-MNC versus CB-derived cells at varying seeding doses 5000–50,000 per well (six-well plate). Representative images of CB CFU-F stained with toluidine blue are shown. (d) CFU-F count per 10⁶ cells isolated from each cell fraction. A total of 500,000 cells were seeded in six-well plates and results are reported as mean value ± SEM (n = 3). (e) Incidence of secondary colonies from re-plated primary CFU-F. Counts represent mean value ± SEM (n = 8) per six-well plate comprising 25% of the primary CFU-F (statistical significance: ✱p < 0.01, p < 0.05, p < 0.001). WBM: whole (total) bone marrow; BM-MNC: Ficoll bone marrow mononuclear cells; Lin-BM: antibody cocktail lineage–depleted bone marrow-mononuclear cells; TCB: total cortical bone; Lin-CB: antibody cocktail lineage–depleted cortical bone.

Figure 2.

Primary CFU-F from CB are larger and spontaneously differentiate to mesodermal cell lineages as compared to their BM-derived counterparts. Low power magnification (1.5×, scale bars: 500 µm) images of (a) CFU-F of CB and (b) BM stained with toluidine blue. The emergence of spontaneous differentiation of CB colonies to mesodermal cell lineages: (c) phase microscopy of (L-R) bone, cartilage, adipose (20×, 40×, 40×, scale bars: 20 µm). (d) Stained cultures for bone; von Kossa (brown) and alkaline phosphatase (blue) (4×, scale bar: 100 µm), cartilage; alcian blue and nuclear fast red (20×, scale bar: 20 µm), adipose; Oil Red O and hematoxylin (40×, scale bar: 20 µm). (e) Relative percent incidence of spontaneously differentiating CFU-F in BM-MSCs and CB-MSCs (lower dark bars). Percentages are taken from the mean count of scored CFU-F at seeding densities of fresh cells at 5000–500,000 per six-well plate (n = 3).

Figure 3.

CB-derived cells demonstrate enhanced tri-lineage differentiation potential under inductive conditions. Comparison of passage one cells from CB and BM. (a) Osteogenic assay run for 20 days. The panels indicate mineral deposition by the cells (L-R) with a phase contrast image (10×), Von Kossa stain with hematoxylin (20×), and combined Von Kossa stain with AP in blue (20×, scale bars: 40 µm). (b) Adipogenic assay run for 20 days. (L-R) Phase contrast images at 10 and then 20 days (20×). At right is an Oil Red O stain for the lipid vesicles with hematoxylin (20×, scale bars: 40 µm). (c) Chondrogenic assay run for 21 days with sections of representative 3D pellets. (L-R) Hematoxylin and eosin stain (10×, scale bar: 80 µm), toluidine blue stain (10×), and alcian blue with nuclear fast red (20×, scale bar: 40 µm).

Figure 4.

Immunophenotyping with surface markers reveals a clonogenic MSC population restricted to CB. (a) Representative FACS plots of CD45, CD31, VLA-1, and Thy-1 expression on freshly isolated CB cells, prior to any depletion. (b) Representative FACS plot for CB demonstrating the full “Lin⁻CD45⁻/CD31⁻” depletion of the combined lineage cocktail (10 hematopoietic lineage markers; (CD2, CD3, CD4, CD8, CD18, CD11b/c, CD45RA, CD71, Gr(RP-1), Mono/Mac) + CD45 + CD31), with depleted cells gated on the left in green. (c) FACS plots of Thy-1 and VLA-1 labeling in the Lin⁻CD45⁻/CD31⁻ fractions of CB and BM, respectively. (d) Example CFU-F from CB-MSC of the Lin⁻/CD45⁻/CD31⁻/VLA-1⁺/Thy-1⁺ phenotype with spontaneous tri-lineage differentiation stained with von Kossa’s reagent, alcian blue, and Oil Red O. Higher magnifications show three distinct stained regions (20×, scale bar: 40 µm) and the outer-edge adipose-cartilage boundary (40×, scale bar: 20 µm).

Figure 5.

CB-derived cells contribute more to ectopic de novo bone tissue than BM-derived cells. (a) Example of the final gelated cell and biomaterial composite implant (1 cm diameter) ready for subcutaneous implantation and (b) an implant before excision, 4 weeks post-surgery. (c) H&E stain on a CB cell implant; arrow indicates an endogenously infiltrated blood vessel within the scaffold (40×, scale bar: 20 µm). Representative stains of implants, demonstrating formation of bone tissue in scaffolds: (d) H&E, (e) OPN immunostain. (L-R): CB, BM, no cells (20×, scale bar: 50 µm). (f) Representative Goldner’s Trichrome stain on implants. (L-R): CB, BM (20×, scale bar: 40 µm). (g) Control stains (L) osteopontin positive and isotype negative on scaffolds and (R) Goldner’s Trichrome on a femur (4×, scale bar: 100 µm). (h) Quantitation of bone tissue determined by osteopontin and Goldner’s Trichrome stains as percentage of image area ± SEM (n = 3) (statistical significance: p < 0.05, p < 0.001).