Mesenchymal Stem Cells and Bone Regeneration

INVITED REVIEW

Mesenchymal Stem Cells and Bone Regeneration

KARL H. KRAUS, DVM, MS Diplomate, ACVS & ABVP and CARL KIRKER-HEAD, Vet MB, MRCVS, Diplomate, ACVS & ECVS

Objective—To review the role of mesenchymal stem cells (MSC) in bone formation and regen-eration, and outline the development of strategies that use MSC in bone healing and regeneration.Study Design—Literature review.Methods—Medline review, synopses of authors’ published research.Results—The MSC is the basic cellular unit of embryologic bone formation. Secondary bonehealing mimics bone formation with proliferation of MSC then their differentiation into compo-nents of fracture callus. Bone regeneration, where large amounts of bone must form, mimics bonehealing and can be achieved with MSC combined with strategies of osteogenesis, osteoinduction,osteoconduction, and osteopromotion. MSC based strategies first employed isolated and cultureexpanded stem cells in an osteoconductive carrier to successfully regenerate a critical segmentaldefect in the femur of dogs, which was as effective as autogenous cancellous bone. Because MSCappeared to be immunologically privileged, a study using mismatched allogeneic stem cells dem-onstrated that these cells would regenerate bone without inciting an immunologic response, doc-umenting the possibility of banked allogeneic MSC for bone regeneration. A technique wasdeveloped for selectively retaining MSC from large bone marrow aspirates at surgery for boneregeneration. These techniques utilized osteoconductive and osteoinductive carriers and resulted inbone regeneration that was similar to autogenous cancellous bone.Conclusion—MSC can be manipulated and combined with carriers that will result in bone regen-eration of critically sized bone defects.Clinical Relevance—These techniques can be employed clinically to regenerate bone and serve as analternative to autogenous cancellous bone.r Copyright 2006 by The American College of Veterinary Surgeons

INTRODUCTION

THE EMBRYOLOGIC mechanisms of bone forma-tion entail an orchestration of cellular, humoral, and

mechanical factors resulting in formation of skeletal bonytissues capable of controlled growth, structural responseto stress, and repair of injury, uniquely without scar.Bone repair parallels embryologic formation, reflectingthe integration of the cellular, humoral, and mechanicalfactors of bone formation. The anlage of bone in forma-tion and repair are mesenchymal stem cells (MSC), whichrespond to and produce regenerative cytokines, replicateand differentiate, form structural matrix, and respond tomechanical demands to restore skeletal function. The roleof MSC in bone formation continues to be defined, and

manipulation of MSC has resulted in new strategies forbone regeneration.1–3

BONE HEALING

Repair of many fractures, specifically those with mo-tion or interfragmentary gaps, occur by secondary boneunion, which is a complex progression from the MSC toits progeny: chondroblasts, chondrocytes, fibroblasts,and osteoblasts, which form a fracture callus.4 Fractureof bone disrupts matrix and results in hemorrhage. Cy-tokines from matrix and degranulation of platelets in thefracture clot form a milieu of biologically active proteins,some chemoattractive to MSC, which migrate to the area

Address reprint requests to Dr. Karl H. Kraus, TUSVM, 200 Westboro Road, North Grafton, MA 01536. E-mail: Karl.Kraus@

Tufts.edu.

Submitted September 2005; Accepted October 2005

From the Orthopedic Research Laboratory, Tufts Cummings School of Veterinary Medicine, North Grafton, MA.

r Copyright 2006 by The American College of Veterinary Surgeons

0161-3499/06

doi:10.1111/j.1532-950X.2006.00142.x

232

Veterinary Surgery

35:232–242, 2006

of damage. MSC then proliferate and follow osteoblastic,chondroblastic, or fibroblastic lineages depending on thelocal fracture environment. Favorable biologic and me-chanical environments result in proliferation and differ-entiation of MSC to osteoblasts and chondrocytes.5

These cells produce matrix with formation of osteoid andcartilage creating a fracture callus. Osteoid mineralizesand cartilage undergoes endochondral ossification form-ing bone until it bridges the fracture gap. As the callusremodels, it assumes geometric and physical characteris-tics to reflect loads applied to it. Thus the cellular eventsof bone healing are chemoattraction, migration, prolifer-ation, and differentiation, and the MSC is the funda-mental ancestor of this cellular assemblage.

BONE REGENERATION

The natural processes of bone repair are sufficient toeffect timely restoration of skeletal integrity for most frac-tures when an appropriate mechanical environment existsor is created with internal fixation or coaptation. How-ever, some situations require manipulation or augmenta-tion of natural healing mechanisms to regenerate largerquantities of new bone than would naturally occur toachieve surgical goals. Specific situations that may requireadditional interventions include substantial loss of hostbone from trauma or tumor resection, arthrodesis, orspinal fusion, non- or delayed unions, metabolic disease,arthroplasty, or insufficient healing potential of the hostbecause of local or systemic disease or old age. Materialsand strategies that are employed must duplicate and am-plify the events of secondary bony union to achieve thedesired result.6–8 Bone can be regenerated through thefollowing strategies: osteogenesis—the transfer of cells; os-teoinduction—the induction of cells to become bone; os-teoconduction, providing a scaffold for bone forming cells;or osteopromotion—the promotion of bone healing andregeneration by encouraging the biologic or mechanicalenvironment of the healing or regenerating tissues.6 Themost efficacious strategies use as many of the fundamentalcomponents of bone regeneration as possible, and eachfacet of bone regeneration relies upon MSC.6,9,10

MSCs

MSC are multipotential cells capable of differentiatinginto osteoblasts, chondrocytes, adipocytes, tenocytes, andmyoblasts.11,12 Their origin is either the cambium layer ofperiosteum or bone marrow, although other sources suchas muscle, fat, and synovium provide a limitedsource.4,13–15 Bone marrow and periosteum sources arerichest in young animals with their numbers diminishing,but still present in old age.11,14 MSC numbers and bio-logic activity are greatest in metaphyseal bone and areas

of thick and vascular periosteum, contributing to morerobust healing at these locations.14 The study of MSC hasprimarily utilized those retrievable from bone marrowaspirates.15,16 Isolation techniques generally are based onthe adherent properties of the MSC.2 Density gradientcentrifugation is employed initially to separate nucleatedMSC.15 As the cells are cultured, MSC adhere to the flasksurface.15,17 The non-adherent cells are removed with theculture medium when it is changed, concentrating theMSC.17 The cells are repeatedly passaged, expanding thecell population, until a pure culture is produced.

The isolated cells are confirmed to possess osteogenicpotential by various techniques.13,15,17,18 In vitro culturingof purified MSC in the presence of dexamethasone, ascor-bic acid, and glycerophosphate12 results in the cells pro-gressing although a osteoblastic lineage.19 The cells assumea cuboidal osteoblastic shape and there is a transient in-duction of alkaline phosphatase activity.20 The cells ex-press bone matrix protein mRNAs and the deposition of ahydroxyapatite-mineralized extracellular matrix confirm-ing that the cells isolated become bone forming cells.12,15

In vivo studies that document osteoblastic potential of theMSCs include loading isolated and culture expanded cellsinto porous ceramic carriers and implanting them in sub-cutaneous tissues of a living animal.14 Vascularized boneforms within the confines of the ceramic implants and notin acellular implants. Modification of these techniqueshave demonstrated that MSC also can be induced to fol-low chondroblastic lineages.21–24 In addition, cells can belabeled and have been shown to survive implantation andmaintain their multilineage potential.25 MSC can alsomaintain viability and multilineage potential after cryo-preservation.14 Therefore, MSC can be effectively isolated,culture expanded, preserved, and implanted.

Osteogenesis

A graft that supplies and supports bone forming cellsis termed osteogenic.6 Bone formation requires the cel-lular machinery to fabricate its structural components. Assuch, no strategy of bone regeneration can neglect intro-duction of cells, and the most efficacious strategies nur-ture an early cellular environment.6 Autogenouscancellous bone grafts are the best example of an os-teogenic graft and are the ‘‘gold standard’’ for bone re-generative materials. Cancellous autograft provides amixture of cells from fully differentiated osteoblasts liningthe cancellous bone to undifferentiated MSC in the mar-row component.4 MSC can respond to the local biologicand mechanical environment and differentiate into anycellular component needed in secondary bone union,therefore offer robust bone healing potential for a varietyof situations. However, the cells do not work alone butare combined with matrix and signals provided by the

233KRAUS AND KIRKER-HEAD

cancellous particles, yielding a mixture that contains oth-er facets of bone regenerating strategy.

Autogenous cancellous bone grafts are easily obtainedfrom the iliac crest or greater tubercle of the humerus indogs and cats and wing of the ilium and sternum inhorses, and are the mainstay of osteogenic material inorthopedic surgery. Harvesting bone graft requires a sep-arate surgical site increasing surgical time and cost. Do-nor site morbidity occurs in humans,26,27 but in ourexperience occurs less frequently in veterinary patients.The quantity of autogenous cancellous bone that can beharvested may be insufficient, especially in older or smallpatients, or when a very large amount of graft material isneeded. Thus, although autogenous cancellous bone isprincipally ideal, substitutes are often needed.

Another example of an osteogenic material is bonemarrow.28,29 Bone marrow aspirates contain MSC, cellsalready committed to osteogenic or chondrogenic lineage,and some biologically active proteins that stimulate boneregeneration in similar manner to naturally occurringfracture clot, many from platelet degranulation. Howev-er, use of bone marrow does not yield the magnitude ofosteogenesis observed with cancellous bone for severalreasons. Aspiration results in varying quality of bonemarrow depending on technique and patient.28 Osteo-progenitor cell content can be low and most importantly,bone marrow lacks the necessary scaffold or osteocon-ductive material to be efficacious on its own.

Osteoinduction

Materials that have the capacity to induce bone for-mation when placed into a site where no bone formationwill occur are termed osteoinductive.6,30–32 These mate-rials do not work alone, but recruit MSC or their progenyto infiltrate the material (chemoattraction and migration)then induce the multipotential cells to multiply and be-come cells that comprise the regenerating bony callus(proliferation and differentiation). The best known ex-ample of osteoinductive materials is demineralized bonematrix (DBM). By decalcifying bone (usually allogeneic)in processes that do not inactivate bone’s biologicallyactive components, organic matrix is exposed along witha plethora of bone stimulatory cytokines trapped withinthe organic matrix during bone formation. These arenaturally occurring initiators and mediators of bonehealing, which are released when bone fractures. Bonemorphogenetic proteins (BMP), the most specific familyof bone inducing cytokines, are members of the trans-forming growth factor (TGF-b) superfamily of growthfactors that include cartilage-derived growth factors, andother growth and differentiation factors. Although oftenreferred to as a specific entity, the growth factors ofDBM are a combination of biologically active cytokines,

which may be chemotactic to MSC, induce MSC to fol-low certain osteoblastic or chondroblastic lineages, orstimulate replication of MSC or their progeny.

DBM is most often supplied as particles or fibers, of-ten with carriers to improve handling properties. Carriersoften include glycerol or gelatin, which do not specificallycontribute to bone regeneration. DBM alone does nothave the bone regenerative capacity of autogenous can-cellous bone and it is speculated that a primary reason forthis is its lack of inherent osteogenic capacity, i.e. DBMdoes not directly supply MSC to the graft site. For thisreason DBM is often employed as an extender, mixedwith autogenous cancellous bone when the amount ofgraft material needed is larger than can be harvested.

Osteoconduction

A material that provides a scaffold for MSC and theirprogeny to migrate into, and proliferate within, is termedosteoconductive.6 These are physical materials of three-dimensional shape that offer an appropriate framework,surface characteristics for adherence of MSC, osteoblasts,osteocytes, chondroblasts, and chondrocytes,33,34 and in-terconnecting porosity for cellular proliferation and vas-cular ingrowth.33 Such materials may or may not impartload bearing characteristics during bone regeneration, mayor may not be absorbable, and may be naturally occurringor synthetic. Specific examples of naturally occurring os-teoconductive materials include autogenous cancellousbone, and some forms of DBM or deproteinized bone.Synthetic materials include hydroxyapatite, collagen, tri-calcium phosphate, apatitic calcium phosphate, calciumsulfates, porous coralline ceramics, bioactive glass, calci-fied triglycerides, or the polyhydroxy acid family of pol-ymers (polylactide, polyglycolide).33–35 Bioabsorption andosteoconduction are quite variable between materials andbecause of their relatively passive role in bone regenera-tion, these materials have limited utility alone. Althoughbone can infiltrate and often replace many of these ma-terials, they lack essential osteogenic and osteoinductiveproperties and rely on natural cellular activities of MSC ortheir progeny to infiltrate and provide their own biologicsignals for proliferation and differentiation. Their use hasbeen best demonstrated when there are very active MSCpopulations (e.g. metaphyseal areas)34; however, their po-tential use should not be discounted because many of thesematerials have been shown to be essential in MSC orcytokine delivery in bone regeneration.34,35

Osteopromotion

A material or physical impetus that results in en-hancement of regenerating bone is termed osteopromo-tive.6 Osteopromotion can function at various stages

234 STEM CELLS AND BONE REGENERATION

during bone healing and provide different stimulatorysignals to bone regenerating tissues. Osteopromotion dif-fers from osteogenesis or osteoconduction as bone for-mation is enhanced without cells or a scaffold; however,osteopromotive stimuli alone cannot induce bone forma-tion. Osteopromotion can be achieved by introduction ofsubstances or materials that enhance bone regeneration,or by physical or mechanical strategies that induce pro-liferation and differentiation of MSC and their progeny.The best example of an osteopromotive substance isplatelet-rich plasma (PRP), which is produced by cen-trifugation of the patient’s own blood yielding a suspen-sion of concentrated platelets. Platelets function not onlyin hemostasis, but are essential for initiation of the heal-ing process, including bone healing. The a granules ofplatelets contain a rich source of growth factors includingTGF-b, platelet-derived growth factor (PDGF), vascularendothelial growth factor (VEGF), and insulin-likegrowth factor (IGF). TGF-b is a potent chemotacticagent for MSC and supports differentiation of osteo-blasts35; PDGF and IGF I and II are potent mitogenicagents, and VEGF is a potent angiogenic agent.36 Plate-lets will release the contents of their granules on contactwith fibrin or von Willebrands factor.37 Thus, PRP pro-vides a surgeon with a naturally occurring mixture ofconcentrated growth factors promoting a robust biologicenvironment for bone healing.

Cellular and humoral factors are a common focus ofresearchers studying bone healing, and essential mechan-ical factors can be neglected. MSC and their progenyrespond aggressively to local mechanical forces and thelocal mechanical environment has a strong influence oncellular proliferation and differentiation.5,38 Developingfracture callus or areas of regenerating bone are subjectedto physical forces. If these forces result in physical dis-

ruption of regenerative tissues then bone healing cannotoccur and non-union results; however, forces of lessermagnitude may have a physical effect on regeneratingtissues without cell death. Compression of MSCs result-ing in increasing pressure within the cell will encouragechondroblastic differentiation; however, low-to-moderatemagnitudes of tensile strain and hydrostatic tensile stressmay stimulate new bone formation. This occurs becauseof deformation or elongation of the cells including MSCand early osteoblastic and chondroblastic cells. Cellulardeformation results in intracellular signals, which in turnencourage regenerating cells to proliferate and follow anosteoblastic lineage, provided the biologic environmentand oxygen levels are suitable.

The result of this mechanical environment can ob-served clinically in the fracture callus of a diaphysealfracture healing by secondary bony union (Fig 1). Duringweight bearing, most forces across the diaphysis are axial,healing tissues are compressed and, like a balloon, ten-sion is created on the outside of the fracture callus. Thistension results in low-to-moderate magnitudes of tensilestrain and hydrostatic tensile stress which is the mostfavorable mechanical environment for bone formationand is why in callus formation bone is formed primarilyon the outside of the callus, in areas of tension and cel-lular elongation. Closer to the callus center, MSCs aremore subjected to compressive forces and thereforechondrogenesis is favored.

The same mechanical influences are seen in boneregeneration in distraction osteosynthesis. As the bonesegments are distracted, MSC undergo tensile strain andhydrostatic tensile stress which is primarily in the centralbone area between the fracture ends being distracted.New bone formation is encouraged and, often largeamounts of new bone can be regenerated. Exuberant

Fig 1. The effect of the mechanical environment on bone regeneration. In a fracture callus healing a gap (left), compressive forces

are presented between the fracture ends and chondrogenesis is favored. On the outside of the callus stem cells are subjected to tensile

forces and osteogenesis is favored. In distraction osteogenesis (right), distractive forces are presented between the fracture ends and

osteogenesis is favored.


callus outside the central axis is not usually seen. Severalstrategies have been employed to impart an optimal me-chanical environment for osteopromotion. These includeorthopedic fixation devices with fixation stiffness allow-ing an appropriate fracture strain, axial dynamization,non-linear fixation stiffness,39 and elastic dynamization.

MSC-BASED STRATEGIES FOR BONE

REGENERATION

Different strategies have been reported for using MSCto regenerate bone in a clinical environment. The objec-tive is to deliver stem cells in appropriate carriers fulfillingthe tenants for bone regeneration. Bone marrow-derivedgrafts, being the easiest retrievable source of stem cells,have been the primary source of MSC; however, bonemarrow aspirates alone are not consistent or rich enoughin MSC to use unmodified as a primary source. Becausethe cells can be isolated and cultured, the historical firstgraft preparation strategy was to use a patient’s own stemcells, culture expand the cells, then re-introduce them withan appropriate carrier.17 Stem cells can be immunogenic-ally privileged, so these same techniques were employed insubsequent studies to investigate allogeneic implantationin hopes of developing techniques for clinically usablestem cell cultures or banks of cell based implants. A thirdstrategy is based on the principle that a prerequisitenumber of stem cells may be present in large bone mar-row aspirates so that aspirates could be concentrated toeffect a bone graft substitute without cell culture. Thiswould allow use of autogenous stem cells and operatingroom table side preparations, obviating the time, incon-venience, and expense of cell culture. These 3 strategieswere investigated as potential clinical applications for re-generation of appendicular skeletal defects using an ap-propriate segmental defect model. Research using MSC isdiverse with numerous studies on the biology and tech-niques of stem cell research. The number of studies usingthese techniques in clinically applicable models in larger(larger than rodent) animals is far more limited. Themodel and studies subsequently summarized cover muchof the published research that has clinical application inveterinary surgery, and offers a concise picture of thecurrent concepts and strategies for use of MSC clinically.

LARGE DIAPHYSEAL SEGMENT DEFECT

MODEL FOR STUDYING BONE REGENERATION

There is little value in attempting to enhance healing offractures that would heal normally. It is hubris to assumethat addition of more stem cells or more cytokines willaccelerate an already honed healing mechanism. Giventhat the biologic and mechanical environments of the

fracture environment are adequate, the natural mecha-nisms of fracture healing result in a rapid and strongrepair. Most fracture failures that should have otherwisehealed are typically technical errors in fixation. This is nottrue when large quantities of bone need to be regeneratedlike in spinal fusion or segmental defects. Thus models forstudy of bone regeneration should mimic conditionswhere the natural mechanisms for healing are exceededand regeneration of sufficient new bone will not occurwithout intervention. Spinal fusion models have beenemployed in dogs for study of MSC application.40 Largesegmental bone defect models have been described insheep femur and tibia.41–43 A segmental defect model ofthe canine femur with plate fixation fulfills most criteriaand was utilized for investigations that encompass basicstrategies for MSC use in regeneration of large segmentaldefects in long bones.45

This model uses the largest canine bone, thus ap-proaching the human condition in size and form, resultingin direct relevance to dogs and humans. The technique,defect size, and fixation type were refined so that subjectswere ambulatory immediately after surgery and walkingnormally within 1 week. Rigid internal fixation provided aknown and consistent mechanical environment betweensubjects and would not fail even if the bone did not healfor 16 weeks. Further, the defects will not heal withoutintervention and the region of bone regeneration is easilyidentified for study (radiology, computed tomography[CT], DEXA, histology, mechanical testing).

Dogs

Purpose bred and closely related intact female red tickhounds, retired from breeding were utilized because theywere a homogenous, readily available source for consec-utive studies. Dogs were housed in separate runs withrubber matting and wood shavings as flooring.

Diaphyseal Segment Defect

Femurs in these dogs were � 20 cm long and 14mmdiameter at mid-diaphysis. A 4.5mm lengthening plate(Synthes Inc, Paoli, PA) permitted 4 screws to be insertedproximal and distal to the segmental defect. Axial loadsclose to the weight of the dogs would result in mild de-formation of the fixation, primarily bending with moststrain contralateral to the plate, without plastic defor-mation. A 21mm segmental defect that included removalof periosteum was sufficiently large to result in atrophicnon-union without intervention. Sufficient autogenouscancellous bone could be harvested from the ipsilateraliliac wing to fill the defect. Autogenous cancellous bone-filled defects healed in a consistent pattern of consolida-tion, incorporation, and remodeling.


Assessment of Healing

Cranial to caudal radiographs were employed to assesshealing. Two previously reported ordinal scales employedto gauge healing within the defect revealed uniform in-creases to near complete radiographic healing by 16weeks. The femoral cortex opposite the bone plate hadthe most mature remodeling, both radiographically andhistologically, and reflected the bony response to milddeformation within the segmental defect that occurredduring normal weight bearing. Using ex vivo mechanicaltesting in torsion, unoperated femurs failed at13.61 � 3.88Nm and grafted femurs at 2.96� 1.3Nm,so grafted femurs had a torsional strength of 23% ofcontrols. Histology and ex vivo CT with the plate re-moved showed incomplete bone regeneration directly un-der the plate. These observations and the low mechanicalstrength of the defects healed with autogenous cancellousbone suggested that the fixation imparted some stressprotection to the healing defect directly under the plate.However, this fixation strength was needed to supportweight bearing if the segmental defect did not heal over16 weeks.

Conclusion

This approach yielded a consistent model of atrophicnon-union without intervention that would consistentlyheal with autogenous cancellous bone grafting. Boneregeneration could be studied during healing withconventional radiography, and ex vivo with radiogra-phy, CT, histology, and mechanical testing. The modelcould be employed to study bone regeneration after useof MSC-based strategies allowing direct comparison withautogenous cancellous bone grafting.

Cultured Expanded Autologous MSC

Because MSC were isolated then studied using culturetechniques, an initial series of studies utilized the modeland culture expansion to test the potential of MSC inbone regeneration.17 After preliminary studies, a studywas designed to test bone regenerative capacity in dogsand to mimic a clinical situation where bone marrowwould be collected in a hospital setting, transported to aremote isolation and culture laboratory, then returned toa hospital setting for final preparation and surgical im-plantation (Fig 2). Defects were filled with a ceramiccarrier, ceramic carrier loaded with MSC, or autogenouscancellous bone.

MSC Collection and Expansion

Bone marrow was collected from the iliac wing of dogs16 days before segmental defect creation. Under an-esthesia and using a 15G Jamshidi needle, 9mL bonemarrow was aspirated into a syringe containing 1mL-heparinized saline solution. Specimens were sent by over-night courier service for delivery to the cell-culture facilitythe next day. MSC were isolated and enriched usingdensity gradient centrifugation and cultured in flasks.Non-adherent cells were removed along with the culturemedium and fresh medium was added to the adherentcells twice weekly. Cells were passaged on or about the10th day. Between the 13th and 15th days, culture flaskswere returned by commercial airline and ground trans-portation to the surgical team.

Delivery of MSC into Segmental Defect

Blocks of porous hydroxyapatite and b-tricalciumphosphate (Zimmer, Warsaw, IN), with pore size of 200–

Fig 2. One model of fabrication of a bone graft substitute. Bone marrow is harvested from a patient. The bone marrow is

transported to a culture facility where the mesenchymal stem cells (MSCs) are isolated and culture expanded. The MSCs are then

transported back to the hospital and operating theater where they are combined with a osteoconductive matrix to provide an

osteogenic composite for implantation.


450mm, were fabricated into cylinders (14mm diameter,21mm length, with a central canal 8mm diameter). MSCwere loaded onto the ceramic cylinders and implantedinto the specific dog from which they had been harvested.Cell-free implants were treated in identical manner andimplanted into other dogs operated similarly includingprevious bone marrow harvest. Another group had au-togenous iliac cancellous bone implanted.


Cranial to caudal radiographs were taken before andimmediately after surgery, and at 4, 8, 12, and 16 weeks.After 16 weeks, femora were analyzed by quantitativehistology of non-decalcified sections.

From radiographs, new bone formed in the pores ofthe MSC loaded implants by 4 weeks but not in cell-freeimplants. New bone formed beyond the confines of theceramic implant on the medial aspect of the femur (i.e.away from the plate). New bone also formed in parentdiaphyseal bone that interfaced with the implant. Boneformation was greatest at 12 weeks and declined some-what because of remodeling by 16 weeks. The ceramicfractured in several implants, but these fractures ap-peared to be bridged by new bone formation by the studyend. Restoration of the segmental defect was comparablewith regeneration resulting from the implantation of au-togenous cancellous bone graft. New bone growth out-side the confines of the ceramic implant was not seen incell-free implants.

Histologic evaluation confirmed ingrowth of new bonewithin the ceramic implant and bridging of the segmentaldefect. There was stimulation of new bone formationfrom host diaphysis at the interfaces of the implants, andproliferation of bone outside the ceramic implant on thecaudal medial aspect of the regenerated femur. New boneformation was the rarest, directly under the plate. Cell-free implants had only modest infiltration of new boneformation within the pores of the ceramic implant, withfibrous tissue infiltrating most of the ceramic’s pores.

Conclusion

A critically sized diaphyseal bone defect could be re-generated using MSC and an osteoconductive material.The gap was bridged and the implant infiltrated with newbone. New bone was formed outside the ceramic implantin areas of highest mechanical stress, suggesting that thesecells were responding and proliferating in response tomild mechanical stresses. Further, addition of MSCs re-sulted in stimulation of new bone formation in areas ad-jacent the implant site, suggesting that either these cellswere infiltrating the adjacent host bone, or stimulatingthe host bone to generate new bone. Of practical impor-

tance, the study was performed in a clinically relevantsituation (hospital and off-site culture laboratory).

A similar study was performed using a segmental de-fect in the tibia of sheep.43,44 A critically sized segmentaldefect was created and stabilized with an external fixator.A hydroxyapatite ceramic carrier was employed with(n¼ 2) and without (2) culture expanded autologousbone marrow-derived MSC. Results were similar to dogs,with bone formation occurring more completely withinthe internal macropore spaces of the cell-loaded hydro-xyapatite implants.

CULTURE EXPANDED ALLOGENEIC MSC

MSC lack certain hematopoietic cell surface mark-ers,11,46,47 and possess other characteristics that suggestthat they can elude T-cell-mediated cell rejection andtherefore are immune-privileged.48–53 The suggestion thatMSC are sufficiently non-immunogenic raised the possi-bility that allogeneic MSC could be utilized as effectivelyas autogenous MSC in bone regeneration. This prospectwould imply the development of ‘‘off-the-shelf’’ MCS-based bone regenerative materials, obviating the time de-lay and cost of harvesting and culture expanding an in-dividual’s own MSC. This hypothesis was tested toconfirm the feasibility of allogeneic MCS-banked boneregeneration products in a study using the previous mod-el (Fig 3).54

Dogs

Donor and recipient dogs were paired based on ped-igree differences, leukocyte antigen typing, and mixedlymphocyte reaction assay; donor and recipient dogs

Fig 3. Fabrication of mesenchymal stem cell (MSC)-based

bone graft substitutes from allogeneic cell cultures. Allogeneic

cell cultures are maintained in a stem cell banking facility.

When needed the cells can be loaded into an appropriate ma-

trix and transported to the hospital or operating theater for

implantation. MSCs are immunologically privileged and will

not incite an immunologic host response.


were from different lineages and sources. Dogs adminis-tered MSC were determined to be sufficiently immuno-logically different based on a complete mismatch of themajor histocompatability complex Class I and II mole-cules, and mixed lymphocyte reaction assays. The modelmethodology was identical to the autologous MSC study(size, weight, breed, age, and sex of the dogs; bone mar-row harvest, cell isolation and culturing, ceramic osteo-conductive carrier, size of defect, housing and care of thedogs, bone marrow harvest technique, surgeon, radio-graphic technique, and histologic evaluation) to allowdirect comparison between studies without need for ad-ditional controls (empty defects, cell free implants, orconcurrent autogenous MSC groups).


Immunologic response to the implant was evaluatedhistologically and by alloantibody detection. Non-decal-cified bone sections were analyzed for bone regenerationand one half of the explanted segmental defect was sub-mitted for decalcification and qualitative analysis of cel-lular reaction. Antibody production in recipient serumagainst the implanted allogeneic MSC and the donor wasanalyzed by the allo-antibody assay for the detection ofperipheral blood mononuclear cells suggesting tissue re-jection and included secondary goat anti-canine immuno-globulin M (IgM) and sheep anti-canine IgG. So that atransient immunologic reaction was not missed, femurswere harvested and immunologic assays were performedat 4, 8, and 16 weeks.

Radiographically, hydroxyapatite–tricalcium phos-phate implants were infiltrated with new bone and inter-faced with the diaphyses of the host bone similarly tofindings in autogenous MSC studies. Bone formed withinthe porous spaces of the ceramic implant. New bone re-generated from the implant forming a healing callus thatwas larger than the ceramic, primarily on the medial andcaudal aspect of the femur, indicating that allogeneic cellswere responding appropriately to the mechanical envi-ronment. At no time was an adverse host response(lymphocytic infiltration) observed and no systemic allo-antibody production was evident at any sampling time inany subject.

The added histologic time points provided informationof how MSCs acted in the formation of new bone, spe-cifically early vascular ingrowth and cellular proliferationwhich was then followed by progressive increases in wo-ven and lamellar bone. New bone formation was by in-tramembranous rather than endochondral ossification.Cellular events included migration, proliferation, anddifferentiation. Chemoattraction was not needed as anappropriate number of MSC were present in the initialosteoregenerative material. There was progressive re-es-

tablishment of the medullary cavity and reformation ofthe contour of the femoral diaphysis suggesting appro-priate remodeling was occurring.

Conclusion

This study revealed the remarkable finding that all-ogeneic MSCs can function similar to autogenous MSCin bone regeneration without demonstrable immunologicreaction. Such results strongly support the feasibility ofallogeneic MSC-based products when bone regenerationis needed surgically. Because MSC can be cryopreserved,materials could be readily available without the expenseand delay of culture expansion of a patient’s own MSCfrom bone marrow aspirates.

SELECTIVE RETENTION OF AUTOGENOUS MSC

The carrier materials utilized in the segmental defectsupplied MSC in a substrate capable of supporting thesecells through an osteogenic lineage, fulfilling the criterionof osteogenesis. The hydroxyapatite tricalcium phosphatesupported ingrowth of osteogenic cells and new boneformation as well as vascular ingrowth. Although thematerial lacked strength and absorbability, it sufficientlysupported bone regeneration through osteoconduction.Proliferation of stem cells occurred in areas known tohave an appropriate mechanical environment, thereforeboth autogenous and allogeneic MSC responded to me-chanical osteopromotive stimuli. Still, osteoinductive ac-tivity was indirect, and relied on the molecular machineryof the MSC to induce and fabricate the appropriate cy-tokines.

Bone marrow aspirates alone or combined with anosteoconductive matrix have been reported to promotebone regeneration through osteogenesis in some specificsituations.10,28,29 More recently, bone marrow aspirateshave been processed so that the osteogenic cells can beconcentrated from the non-osteogenic cells, enriching theresulting material with MSC, and other osteogenic cells.The principle of selective cell retention (SCR) relies onisolation of MSC, specifically cells that are adherentwhile other non-osteogenic nucleated cells of bone mar-row aspiration, such as the hematopoietic cells are gen-erally non-adherent.55 Yet, even with these newtechniques and using large aspirates, the cellular densityis not comparable with that obtained with cell culturingtechniques. In normal secondary bone union however,relatively few MSC are chemotactically recruited to mi-grate into the healing callus. Osteoinductive influencesresult in proliferation and differentiation, finishing outthe essential cellular events of bone healing.

The principle of cell retention-based bone regenerationwould be to provide smaller but prerequisite numbers of


MSCs, then add a carrier matrix with osteoinductive aswell as osteoconductive properties so that these cellswould adequately proliferate and differentiate. In addi-tion, osteopromotive materials could be combined ifneeded to produce a bone graft substitute that would becomparable with autogenous cancellous bone graft. Thepractical advantage would be that simple bone marrowaspirates could be obtained without the morbidity ofharvesting autogenous cancellous bone. The bone graftsubstitute could be manufactured table side at surgeryusing the patient’s own cells, circumventing the need forcell culture, expansion, or preservation.

Experimental Approach

A series of studies were performed to investigate theutility of SCR technology to regenerate bone using thesame dog model (Fig 4).55 For an osteoinductive andosteoconductive matrix material, a custom combinationof demineralized cortical fibers (DBM) and mineralizedcancellous chips, an allomatrix, was developed (Veteri-nary Transplant Services, Kent, WA). Conventional au-tograft was employed as control in 1 group. Theallogeneic matrix material was implanted alone andmixed with a simple untreated bone marrow aspirate in2 groups. SCR was performed by collecting 30mL he-parinized bone marrow from both iliac wings. The bonemarrow was passed several times over the allogeneic ma-trix allowing the adherent osteogenic cells to combinewith the matrix to concentrate them. A calcium–throm-bin clotting agent was added resulting in a material withhandling properties similar to autogenous cancellousbone. This material alone was placed into the 21mm

femoral segment defect in a fourth group. In 1 additionalgroup, PRP was added to the allogeneic matrix andcell retention material as a biologic osteopromotiveagent.


Regeneration was followed radiographically every 4weeks for 16 weeks. Micro-CT and 3D reconstruction aswell as histology was performed on the regenerated fem-oral segments from explanted specimens. In all cases au-togenous cancellous bone, allomatrix treated with SCR,and allomatrix, SCR, and PRP treated defects completelybridged the segmental defect and restored the continuityof the femoral diaphysis. The radiographic progression ofhealing between autograft, and SCR and allomatrixpreparations were not distinguishable. However, defectstreated with allomatrix alone or allomatrix with simpleuntreated bone marrow aspirates resulted in non-unionof the defect in 50% and 33% of the segmental defectsrespectively assuring that the SCR techniques were trulyresponsible for the robust bone regeneration.

Conclusion

These studies confirmed the hypothesis that SCRtechniques, when combined with an osteoinductive andosteoconductive matrix, resulted in bone graft substituteswith regenerative capacity comparable with autogenouscancellous bone. These same techniques have been sub-sequently employed in a spinal fusion model in dogs withsimilar results,40 and then refined to commercially avail-able devices and techniques for the fabrication of SCR-based bone graft substitutes (Cellect, De Puy Osteobio-logics, Raynham, MA).

MSC is the seed from which bone is formed. The studyand understanding of the MSC has led to hypothesesabout how these cells could be manipulated to regeneratebone for clinical problems. Keeping in mind all facets ofbone regeneration and using appropriate matrix materi-als, MSC have been manipulated effectively by cultureexpanding autogenous or allogeneic cells, or by concen-trating them using SCR techniques. Large amounts ofbone can be regenerated in clinically applicable models.These techniques can now be further refined for applica-tion in specific clinical situations.

ACKNOWLEDGMENT

The authors thank Sudha Kadiyala and Scorr P. Bruder

for long-term collaborative research support and Beth

Mellor for the illustrations.

Fig 4. One step fabrication of a bone graft substitute using

selective cell retention (SCR). As mesenchymal stem cells

(MSCs) are adherent to certain surfaces, they can be concen-

trated from large bone marrow aspirates in the operating

theater at the time of surgery. Although the numbers of cells

are still relatively small, they can be added to osteoinductive

and osteoconductive matrix materials to produce a suitable

osteogenic bone graft substitute.


REFERENCES

1. Caplan AI, Bruder SP: Mesenchymal stem cells: building

blocks for molecular medicine in the 21st century. Trends

Mol Med 7:259–264, 2001

2. Caplan AI: Mesenchymal stem cells. J Orthop Res 9:641–650,

1991

3. Khan SN, Cammisa FP, Sandhu HS, et al: The biology of

bone grafting. J Am Acad Orthop Surg 13:77–86, 2005

4. Yoo JU, Johnstone B: The role of osteochondral progenitor

cells in fracture repair. Clin Orthop 355:S73–S81, 1998

5. Carter DR, Beaupre GS, Giori NJ, et al: Mechanobiology of

skeletal regeneration. Clin Orthop 355:S41–S55, 1998

6. Attawia M, Kadiyala S, Fitzgerald K, et al: Cell-based ap-

proaches for bone graft substitutes, in Laurencin CT (ed):

Bone Graft Substitutes. West Conshohocken, PA, ASTM

International pp 126–141

7. Cancedda R, Dozin B, Giannoni P, et al: Tissue engineering

and cell therapy of cartilage and bone. Matrix Biol 22:81–

91, 2003

8. Cancedda R, Mastrogiacomo M, Bianchi G, et al: Bone

marrow stromal cells and their use in regenerating bone.

Novartis Found Symp 249:133–143, 2003, discussion 143–

137, 170–134, 239–141

9. Bruder SP, Fox BS: Tissue engineering of bone. Cell based

strategies. Clin Orthop 367:S68–S83, 1999

10. Muschler GF, Midura RJ: Connective tissue progenitors:

practical concepts for clinical applications. 395:Clin Ort-

hop 66–80, 2002

11. Pittenger MF, Mackay AM, Beck SC, et al: Multilineage

potential of adult human mesenchymal stem cells. Science

284:143–147, 1999

12. Jaiswal N, Haynesworth SE, Caplan AI, et al: Osteogenic

differentiation of purified, culture-expanded human me-

senchymal stem cells in vitro. J Cell Biochem 64:295–312,

1997

13. Yoo JU, Barthel TS, Nishimura K, et al: The chondrogenic

potential of human bone-marrow-derived mesenchymal

progenitor cells. J Bone Jt Surg Am 80:1745–1757, 1998

14. Bruder SP, Jaiswal N, Haynesworth SE: Growth kinetics,

self-renewal, and the osteogenic potential of purified hu-

man mesenchymal stem cells during extensive subcultiva-

tion and following cryopreservation. J Cell Biochem

64:278–294, 1997

15. Kadiyala S, Young RG, Thiede MA, et al: Culture expanded

canine mesenchymal stem cells possess osteochondrogenic

potential in vivo and in vitro. Cell Transplant 6:125–134,

1997

16. Bruder SP, Fink DJ, Caplan AI: Mesenchymal stem cells in

bone development, bone repair, and skeletal regeneration

therapy. J Cell Biochem 56:283–294, 1994

17. Bruder SP, Kraus KH, Goldberg VM, et al: The effect of

implants loaded with autologous mesenchymal stem cells

on the healing of canine segmental bone defects. J Bone Jt

Surg Am 80:985–996, 1998

18. Cooper LF, Harris CT, Bruder SP, et al: Incipient analysis of

mesenchymal stem-cell-derived osteogenesis. J Dent Res

80:314–320, 2001

19. Haynesworth SE, Goshima J, Goldberg VM, et al: Charac-

terization of cells with osteogenic potential from human

marrow. Bone 13:81–88, 1992

20. Bruder SP, Jaiswal N, Ricalton NS, et al: Mesenchymal stem

cells in osteobiology and applied bone regeneration. Clin

Orthop 355:S247–S256, 1998

21. Barry F, Boynton RE, Liu B, et al: Chondrogenic differen-

tiation of mesenchymal stem cells from bone marrow: dif-

ferentiation-dependent gene expression of matrix

components. Exp Cell Res 268:189–200, 2001

22. Barry FP: Biology and clinical applications of mesenchymal

stem cells. Birth Defects Res Part C Embryo Today

69:250–256, 2003

23. Barry FP: Mesenchymal stem cell therapy in joint disease.

Novartis Found Symp 249:86–96, 2003, discussion 96–102,

170–104, 239–141

24. Barry FP, Murphy JM: Mesenchymal stem cells: clinical ap-

plications and biological characterization. Int J Biochem

Cell Biol 36:568–584, 2004

25. Quintavalla J, Uziel-Fusi S, Yin J, et al: Fluorescently labeled

mesenchymal stem cells (MSCs) maintain multilineage po-

tential and can be detected following implantation into ar-

ticular cartilage defects. Biomaterials 23:109–119, 2002

26. Joshi A, Kostakis GC: An investigation of post-operative

morbidity following iliac crest graft harvesting. Br Dent J

196:167–171, 2004

27. Silber JS, Anderson DG, Daffner SD, et al: Donor site mor-

bidity after anterior iliac crest bone harvest for single-level

anterior cervical discectomy and fusion. Spine 28:134–139,

2003

28. Muschler GF, Nitto H, Boehm CA, et al: Age- and gender-

related changes in the cellularity of human bone marrow

and the prevalence of osteoblastic progenitors. J Orthop

Res 19:117–125, 2001

29. Muschler GF, Nitto H, Matsukura Y, et al: Spine fusion

using cell matrix composites enriched in bone marrow-de-

rived cells. Clin Orthop 102–118, 2003

30. Kirker-Head CA: Development and application of bone

morphogenetic proteins for the enhancement of bone heal-

ing. J Orthopaed Traumatol 6:1–9, 2005

31. Kirker-Head CA: Potential applications and delivery strate-

gies of bone morphogenetic proteins. Adv Drug Delivery

Rev 43:65–92, 2000

32. Kirker-Head CA, Gerhart TN, Armstrong R, et al: Healing

bone using recombinant human bone morphogenetic pro-

tein-2 and copolyme. Clin Orthop 349:205–217, 1998

33. Ishaug SL, Crane GM, Miller MJ, et al: Bone formation by

three-dimensional stromal osteoblast culture in biodegrad-

able polymer scaffolds. J Biomed Mater Res 36:17–28,

1997

34. Bucholz RW: Nonallograft osteoconductive bone graft sub-

stitutes. Clin Orthop Rel Res 395:44–52, 2002

35. Fleming JE Jr, Cornell CN, Muschler GF: Bone cells and

matrices in orthopedic tissue engineering. Orthop Clin

North Am 31:357–374, 2000

36. Cassiede P, Dennis JE, Ma F, et al: Osteochondrogenic po-

tential of marrow mesenchymal progenitor cells exposed to


TGF-beta 1 or PDGF-BB as assayed in vivo and in vitro. J

Bone Miner Res 11:1264–1273, 1996

37. Carlevaro MF, Cermelli S, Cancedda R, et al: Vascular end-

othelial growth factor (VEGF) in cartilage neovascularizat-

ion and chondrocyte differentiation: auto-paracrine role

during endochondral bone formation. J Cell Sci 113(Part

1): 59–69, 2000

38. Kraus KH, Turrentine MA, Jergens AE, et al: Effect of des-

mopressin acetate on bleeding times and plasma von Wille-

brand factor in Doberman pinscher dogs with von

Willebrand’s disease. Vet Surg 18:103–109, 1989

39. Carter DR: Mechanobiology in rehabilitation science. J Re-

habil Res Dev 37:vii–viii, 2000

40. Kowaleski MP, Marston MT, Kraus KH: Nonlinear increas-

ing axial gap stiffness in type II external skeletal fixation: a

mechanical study. Vet Surg 32:120–127, 2003

41. Muschler GF, Matsukura Y, Nitto H, et al: Selective reten-

tion of bone marrow-derived cells to enhance spinal fusion.

Clin Orthop Rel Res 432:242–251, 2005

42. Kirker-Head CA, Herhart T, Hennig GE, et al: Long term

evaluation of healing segmental femoral defects in sheep

using recombinant human bone morphogenetic protein-2.

Clin Orthop 318:222–230, 1995

43. Marcacci M, Kon E, Zaffagnini S, et al: Reconstruction of

extensive long-bone defects in sheep using porous hydro-

zyapatite sponges. Calcif Tissue Int 64:83–90, 1999

44. Kon E, Muraglia A, Corsi A, et al: Autologous bone

marrow stromal cells loaded onto porous hydroxy-

apatite ceramic accelerate bone repair in critical-sized

defects of sheep bones. J Biomed Mater Res 49:328–337,

2000

45. Kraus KH, Kadiyala S, Wotton H, et al: Critically sized os-

teo-periosteal femoral defects: a dog model. J Invest Surg

12:115–124, 1999

46. Haynesworth SE, Baber MA, Caplan AI: Cell surface anti-

gens on human marrow-derived mesenchymal cells are de-

tected by monoclonal antibodies. Bone 13:69–80, 1992

47. Bruder SP, Horowitz MC, Mosca JD, et al: Monoclonal an-

tibodies reactive with human osteogenic cell surface anti-

gens. Bone 21:225–235, 1997

48. Cottler-Fox MH, Lapidot T, Petit I, et al: Stem cell mobi-

lization. Hematology (Am Soc Hematol Educ Program)

419–437, 2003

49. Chohan R, Vij R, Adkins D, et al: Long-term outcomes of

allogeneic stem cell transplant recipients after calcineurin

inhibitor-induced neurotoxicity. Br J Haematol 123:110–

113, 2003

50. Devine SM: Mesenchymal stem cells: will they have a role in

the clinic? J Cell Biochem Suppl 38): 73–79, 2002

51. Devine SM, Peter S, Martin BJ, et al: Mesenchymal stem

cells: stealth and suppression. Cancer J 7(Suppl 2): S76–

S82, 2001

52. Bartholomew A, Sturgeon C, Siatskas M, et al: Mesenchymal

stem cells suppress lymphocyte proliferation in vitro and

prolong skin graft survival in vivo. Exp Hematol 30:42–48,

2002

53. Devine SM, Hoffman R, Verma A, et al: Allogeneic blood

cell transplantation following reduced-intensity condition-

ing is effective therapy for older patients with myelofibrosis

with myeloid metaplasia. Blood 99:2255–2258, 2002

54. Arinzeh TL, Peter SJ, Archambault MP, et al: Allogeneic

mesenchymal stem cells regenerate bone in a critical-sized

canine segmental defect. J Bone Jt Surg Am 85-A:1927–

1935, 2003

55. Brodke DS, Kapur TA, Pedrozo HA, et al: A bone graft

prepared with selective cell retention technology heals ca-

nine segmental defects as effectively as autograft. J Orthop

Res, accepted 2005


Mesenchymal Stem Cells and Bone Regeneration

Documents

Transcript of Mesenchymal Stem Cells and Bone Regeneration