such as "Introduction", "Conclusion"..etc
Brief Summary of Research on BMMSCs
BMMSCs reside in the nonhematopoietic components in the postnatalbone marrow and provide a suitable hematopoietic microenvironmentfor the hematopoietic cell population's proliferation and differentiation[2, 3]. In vitro, they are "adherent, clonogenic, nonphagocytic,and fibroblastic in habit (defined as colony-forming units-fibroblastic;CFUFs)" [4, 5]. The BM CFU-Fs are heterogeneous populationswith varying colony sizes, growth rates, immunophenotypes, anddifferentiation abilities [5, 6]. Based on extensive studiesat the level of the unfractionated cell population, immunoselectedcell population, or single cell, it has been widely acceptedthat the BM CFU-Fs are able to differentiate under proper experimentalconditions into bone, cartilage, adipocyte, fibrous tissues,and hematopoietic supporting tissues in vitro and in vivo [6–21].Evidence also showed that BM CFU-Fs could undergo unorthodoxdifferentiation, giving rise to cells with visceral mesoderm,neuroectoderm, and endoderm characteristics when induced [6,22–25], could engraft in bone, muscle, brain, lung, heart,liver, gastrointestinal tract, and hematopoietic system whentransplanted [20, 23, 26, 27], and could even contribute tomost somatic cell types when injected into an early blastocyst. Therefore, adult stem cells (generally defined as clonogeniccells capable of both self-renewal and multilineage differentiation[28, 29]) were presumed to reside in the BM CFU-Fs and thusthe terms: marrow mesenchymal stem cells , mesenchymal stemcells , and multipotent adult progenitor cells  wereproposed. However, a trend was noticed that the term mesenchymalstem cells was gradually improperly being adopted in some literaturewithout applying the stringent criteria for stem cells. Consequently,a position statement  proposed to use "multipotent mesenchymalstromal cells" to designate the plastic-adherent cells isolatedfrom bone marrow or other tissues with multipotent differentiationcapacity. To date, specific markers are still lacking for isolatingthe BMMSC subset with predictably broad or restricted potential; nevertheless, surface markers such as Strol-1+ , SH2+(CD105+), CD34–, CD45, CD14–, and so on  wereused individually or in combination as general markers to defineor purify BMMSCs. In addition to the multipotent differentiationcapacity, BMMSCs were also discovered to be immunologicallyimmature. They do not appear to elicit alloreactive lymphocyteproliferative responses and may modulate immune responses by suppressing the proliferation of T lymphocytes , thusthey are able to survive in a xenogeneic environment [20, 36].As BMMSCs are easy to isolate and expand, they have arousedgreat interest as attractive candidates for cell therapy orcellular vehicles in molecular therapy to deliver genes [37–39].The study of BMMSCs has been extended from bench to bedsidewith many clinical applications such as improving hematopoieticengraftment  and correcting genetic disorders .
Detection of CFU-Fs in Peripheral Blood
The detection of peripheral blood-borne CFU-Fs (PB CFU-Fs) wasactually earlier than the detection of BM CFU-Fs. The observations[42, 43] made in the early 20th century on the transformationof blood leukocytes into fibroblast-like cells and further intoconnective tissues were probably the earliest indications forthe existence of PB CFU-Fs. The existence of PB CFU-Fs withclonogenesis feature and capacity for prolonged passaging waslater confirmed by many other investigators [44–48]. Butit was suspected that these blood-borne fibroblast-like cellsmight be a result of connective tissue fragment contaminationof blood when obtained by the cardiac route [49–51]. Thesuspicion of the contamination with endothelial or connectivetissue cells during blood collection was soon eliminated byindependent experiments conducted in adult male rabbits and adult guinea pig , respectively, by comparing the numberof fibroblastic colonies in the blood collected by multiplepunctures versus fewer or single puncture(s). The investigatorsfound no change in the number of the fibroblastic colonies regardlessthe number of heart punctures applied.
To date, PB CFU-Fs have been detected in the adult peripheralblood of a variety of mammalian species including guinea pig[43, 52, 53], rabbit [43, 45, 47, 53, 54], dog [55, 56], mouse[53, 57, 58], rat [59 and unpublished data in our laboratory],and human [46, 53, 60–66]. It was also reported that fibroblasticcolonies developed in the macrophage cultures obtained fromchicken cardiac blood, although the authors explained it asconnective tissue fragment contamination , which has alreadybeen refuted . The observations recorded in those old publications[42, 43] also implied that PB CFU-Fs also exist in Rhesus monkeyblood; yet, no conclusive data were documented.
Colony-Forming Efficiency of PB CFU-Fs
PB CFU-Fs are extremely low in frequency. The frequency of CFU-Fsis indicated by CFE, colony-forming efficiency  defined asthe ratio of the number of colonies to the number of cells seeded.Many investigators have calculated the CFE of CFU-Fs in primaryculture. Accordingly, CFE was the ratio between the colony numberand the number of seeded mononuclear cells (MNCs) in cultures.Table 1 shows clearly that the CFE of PB CFU-Fs varies widelyboth among and within species. Compared with BM CFU-Fs [11,67], PB CFU-Fs have a much lower CFE. For example, human bonemarrow generally yields colony numbers in the range one per5 x 103 to 1 x 104 MNCs [33, 68–70], whereas the yieldof PB CFU-Fs is usually poor, and it is quite common to failin detecting any of them [71–73]. Moreover, maintainingPB CFU-Fs seems to be difficult also .
Histochemical Characteristics and Immunophenotype of PB CFU-Fs
PB CFU-Fs shared many phenotypic characteristics with BM CFU-Fsand showed an immunophenotypic profile that was similar overallbetween species (Tables 2 and 3). Human PB CFU-Fs synthesizeda series of collagens and other extracellular matrix molecules[53, 60]. They lacked or possessed low levels of the hematopoieticprogenitor marker CD34 and lacked the macrophage marker CD14and the leukocyte common antigen CD45 [53, 60–63]. Itwas also shown that human PB CFU-Fs had low-level CD117 (c-kit) and did not express human leukocyte antigen (HLA)-DR [61,62], VIII-factor-associated antigen , CD31 , and neurofilament, but exhibited a range of mesenchymal lineage phenotypes[53, 60–63]. Human PB CFU-Fs expressed CD106 (vascularcell adhesion molecule-1) and intercellular adhesion molecule-1[53, 60], which are adhesive molecules "used by stromal cellsto interact with marrow hematopoietic progenitors" [78, 79].However, the human PB CFU-Fs were negative for Stro-1 ,which is a human marrow stromal marker widely used for purifyingthe BMMSCs [33, 80]. Human PB CFU-Fs did not express Muc-18(53)as BM CFU-Fs did . And there were also inconsistent findingsfor endoglin (CD105) expression [53, 60, 62, 63]. More interestingly,PB CFU-Fs could be enriched in CD133+ cell populations ,which has been used as a strategy to isolate cells with hematopoietic,endothelial potential, or hemangioblasts .
Multidifferentiation Potential of PB CFU-Fs
There is accumulating evidence from both in vivo and in vitroexperiments suggesting that the PB CFU-Fs possess mesenchymallineage differentiation capability. A single clone strain clonedfrom canine species indicated that immortalized PB CFU-F cellsgave rise to "bone-lining cells" expressing osteocalcin afterautologous i.v. transplantation . PB CFU-Fs from human,guinea pig, mouse, rabbit, and rat proved to be able to developinto osteoblasts, reticular cells, lipocytes, chondrocytes,myotubes, and fibroblasts [53, 59, 62]. There are also otherstudies showed that the PB CFU-Fs may contain other precursorcells or more immature cells capable of turning into neuronal/glialcells [61, 62], generating hematopoietic progenies in lethallyirradiated animals , forming extensive networks in vitrowhen growing on BD Matrigel (BD Biosciences, San Diego, http://www.bdbiosciences.com), significantly improving the collateral blood flow (arteriogenesis)and neoangiogenesis formation in a murine hind limb ischemiatransplant model after i.v. infusion and integrating into theskeletal muscle in the affected limb . A recent study indicatedthat the differentiation direction of both PB CFU-Fs and BMCFU-Fs to osteogenesis, angiogenesis, and neovessel formationwas milieu-dependent and could be adjusted by modification offavorable conditions . In the studies on BM CFU-Fs, it hasbeen revealed that only a small proportion of single coloniescould form bone marrow organs when grafted heterotopically .Analogously, current data of PB CFU-Fs suggest a heterogeneouscell population probably composed of different types of progenitor/precursorcells or cells at different differentiation stages. It is unknownfor how many passages the multidifferentiation ability of thesecells could be preserved. Although it is possible that thereare circulating multipotent adult progenitor cells or pluripotentstem cells residing in the circulation by analogy with the findingsin bone marrow , stringent evidence is still lacking. Themultipotent differentiation ability of the PB CFU-Fs makes thema potential candidate for application in cell therapy and tissueengineering. The rabbit PB CFU-Fs combined with porous calciumphosphate resorbable substitutes have been demonstrated to enhancebone regeneration in the rabbit ulna critical-sized bone-defectmodel, suggesting allogeneic PB CFU-Fs may be a new source ofcirculating osteogenic cells for bone regeneration . Asthe blood is more accessible than bone marrow, the advantagesof using peripheral blood as a potential source of CFU-Fs areobvious. However the CFE of CFU-Fs is significantly lower inthe PB than that of the BM, which is a major obstacle for theirfuture characterization and clinical application.
Until now, we deliberately used the term "PB CFU-Fs" insteadof the term "PBMSCs". The purpose for doing so is to avoid potentialconfusion and misunderstanding as well as to make the deliveryof information easier, because the terminology for this cellpopulation has not yet been standardized. Since their discovery,dozens of names have been given to this cell population or itssubpopulation: fibrocytic cells , fibrocyte, fibrocyticcolony , fibroblast, fibroblast colony, precursors for fibroblastcolonies , fibroblast-like colonies, stellate colonies ,fibroblastoid stromal cells , stromal cells , CD34–/lowhematopoietic stem cell clones with mesenchymal stem cell characteristics, circulating skeletal stem cells , CD34–CD105+mesenchymal cell lines , CD34-negative fibroblast-like celllines, CD34-negative CD105-positive cell line , and mesenchymalstem cells [54, 59, 62, 63, 65]. Notwithstanding the multifariousnames, the PB CFU-Fs is commonly adopted. However, "CFU-Fs"only emphasizes the morphological feature of the cells in theculture system in vitro. It does not deliver enough functionalconnotations and could not cover the state of the cells in vivo.A more biologically meaningful term is needed. "Fibrocyte" and"fibroblast" are apparently not appropriate. Fibroblast is definedas a stellate or spindle-shaped cell with cytoplasmic processespresent in connective tissue, capable of forming collagen fibers.An inactive fibroblast is sometimes called a fibrocyte. Fibrocyteis also a designation for the leukocyte subpopulation circulatingin PB and capable of entering sites of tissue injury rapidly.Circulating fibrocytes were first identified by Bucala et al.in 1994  and were characterized by surface antigens of TypeI collagen+/Type III collagen+/vimentin+/CD34+/CD45+/CD13+/D11b+/MHCclass II+/CD86+ [88–91]. Although both circulating fibrocytesand PB CFU-Fs are described as "fibroblast-like," "spindle"and "elongated," the former are cells of fusiform  witha slightly plump body and two elongated thin ends, whereas thelatter are more blanket-like, that is, stretched-out in shape,pale in color, and not having two sharp ends. As these two differentcell types may simultaneity exist in the cultures, they needto be carefully discerned. The PB CFU-Fs isolated by using cultureconditions similar to those defined for BM CFUFs showed themorphology, phenotype, and differentiation characteristics thatare mainly apt to suggest that they are mesenchymal lineagecells. The name "mesenchymal stem cells," may be suitable fora subpopulation of the PB CFU-Fs after their mesenchymal stemnesshas been stringently proved. However, according to the positionstatement  on the nomenclature for MSC (for both mesenchymalstem cells and multipotent mesenchymal stromal cells), it isnot suitable for applying to the PB CFU-Fs as a whole. Thuswe adopt the nomenclature in that position statement  anduse peripheral blood-derived multipotent mesenchymal stromalcells, that is, PBMSCs denoting the cells, which, to our knowledgeto date, circulate in low numbers, share, most, although notall, of the surface markers with BMMSCs, are adherent, clonogenicand fibroblast-like, and contain a subpopulation capable ofdifferentiating along and even beyond mesenchymal lineages.Cells termed circulating osteoblast-lineage cells  wererecently isolated by flow cytometry with antibodies to osteocalcinand bone alkaline phosphatase, which were osteogenic both invivo and in vitro . This cell population may also fall inthe category of PBMSCs. However, the circulating osteoblast-lineagecells were much larger in number than the PB CFU-Fs isolatedby the plastic-adherence method.
Kuwana et al.  reported another blood-derived cell populationtermed "monocyte-derived mesenchymal progenitors (MOMP)." Exposureof these MOMP to certain inductive conditions resulted in theexpression of genes and proteins specific for osteoblasts, skeletalmyoblasts, chondrocytes, and adipocytes. However, the MOMP seemedto be morphologically similar to the circulating fibrocytes(based on images presented by the authors) and shared some surfacemarkers with the latter, such as Type I collagen+/CD34+/CD45+.MOMP also expressed CD14. Interestingly, a cell population,which was very similar to the MOMP both in morphology and inphenotype (CD14+/CD34+/CD45+), was isolated from human peripheralblood by another independent research group, demonstrated todifferentiate into mature macrophages, T lymphocytes, epithelialcells, endothelial cells, neuronal cells, and liver cells, andhence was termed "pluripotent stem cells (PSC)" . As ithas been documented more than once by independent research groupsthat PBMSCs are CD14–/CD45–/CD34–/low cellpopulation, the MOMP/PSC and the PBMSCs are likely to be distinctcell types. There was also a paper published in 2000  andcommonly cited by other investigators as evidence for circulatingmesenchymal stem cells . This paper named a cell population"mesenchymal precursor cells" isolated from the blood of normalindividuals. These cells had a phenotypic profile (CD105+/vimentin+/TypeI collagen+/CD34–/CD45–) similar to the PBMSCs,yet took a morphology analogous to the circulating fibrocytes.Further investigations into the MOMP, the PSC and the so-calledmesenchymal precursor cells are needed before we are able toaccurately classify these cell populations. And we need to bewareof any confusion and misunderstanding that may be caused bythe names tagged with "mesenchymal."
Do PBMSCs Migrate from Bone Marrow?
With the suspected contamination during sample collection beingexcluded, how the PBMSCs enter blood circulation is still amystery. A straightforward speculation is that they are migrantsfrom bone marrow or other organs. Accumulating data showed thatex vivo expanded BMMSCs achieved engraftment in various normaland damaged tissues as well as homed to the bone marrow aftersystemic infusion [98–106]. However, the migration ofthe infused BMMSCs to extravascular tissues or homing to thebone marrow does not directly support the conjecture that theBMMSCs in situ could spontaneously leave the marrow cavity andenter the bloodstream, or migrate in response to systemic signalstowards to tissues in need of repair. There was an observationmade of mice in parabiosis that phenylhydrazine-induced hemolyticanemia resulted in a threefold increase in the PBMSCs, and partner-derivedPBMSCs could be found in spleens and femoral bone marrow ofboth mice . Real-time migration pattern of tail vein-injectedBMMSCs in response to a tibia fracture revealed that the cellsresided in the lungs for 1 day, moved to liver and brain onday 2, migrated to the fracture site by day 4, and remainedthere . Intravenously infused rat PBMSCs homed to the bonemarrow and migrated into the lesions of chronic rejection inthe cardiac grafts in heart transplant recipients . Thesedata suggested that the mobility of the multipotent mesenchymalstromal cells (MSCs) between bloodstream and organs. But theystill only provide one-directional evidence (bloodstream tobone marrow); when and why the PBMSCs in the phenylhydrazine-inducedhemolytic anemia mice increased are still unclear.
The following two experiments may provide some direct evidencefor PBMSCs' origin under pathological conditions. In the firstexperiment, labeled BMMSCs were injected into the femurs ofosteogenesis imperfecta mice, and they were later detected inthe contralateral femurs, lung, and liver besides the localbone cavity . The other study  tested the hypothesisthat following a bone fracture there is systemic recruitmentof bone-forming cells to a fracture site by using a rabbit ulnarosteotomy model. In this study, labeled BMMSCs were reimplantedinto the remote tibial bone marrow cavity 48 h after the osteotomy,and the labeled cells were detected in the callus of the ulnarfracture site after 3 weeks. Inert beads were also used in theexperiment to rule out the possibility of passive leakage ofthe labeled cells into peripheral circulation. We know thathematopoietic stem cells do circulate. Whether or not BMMSCsnaturally migrate into the circulation is an important questionto address the existence/origin of PBMSCs. But this argumentwill only be meaningful when the view holds up that the marrowstromal cells and hematopoietic cells are of separate origin.
Are There Common Precursors in Adults for Mesenchymal and Hematopoietic Lineage Cells?
Whether there is a common hematopoietic-mesenchymal stem cellin adults has been a long-time debate. Historical views consideredthe hematopoietic and mesenchymal cells to be two histogeneticallyindependent cell lines. This was grounded on the failure ofdetecting both recipient BMMSCs in heterotopic bone marrow transplants[110, 111] and donor BMMSCs after systemic bone marrow transplantation[7, 111–114]. These observations were challenged by datafrom other research groups, which showed that the BMMSCs ofdonor origin enhanced hematopoietic recovery and engrafted inmarrow sinuses after bone marrow transplantation [115–117].However, the evidence of donor-originated BMMSCs in recipientbone marrow is not necessarily in support of the common hematopoietic-mesenchymalstem cell idea because it could be an outcome of the donor BMMSCscompeting with the host BMMSCs, as bone marrow contains bothhematopoietic cells and BMMSCs, and they are inseparable.
In recent years, studies on the extensive plasticity of cellsfrom bone marrow emerged continuously. Single bone marrow-derivedstem cells had multiorgan, multilineage engraftment includinghematopoietic lineage epithelial cells of the liver, lung, gut,and skin on transplantation either into irradiated hosts or into a non-irradiated host . Numerous reports also describedbone marrow stromal cells turning into neural cells [119, 120],cardiac and skeletal muscle [121–123], hepatocytes [124–126],epithelia and endothelia in lung , and epithelia of thegastrointestinal tract . And similar cross-lineage differentiationwas also demonstrated on the peripheral blood-derived hematopoieticstem cells into hepatocytes and epithelia . The versatilebehavior of both the bone marrow stromal cells and the circulatinghematopoietic stem cells seems to make a common precursor conceivablefor hematopoietic and mesenchymal stromal cells. In addition,several CD34– cell populations, such as CD34–/c-Kit+/Sca-1+,CD34–/Lin–/CD38–, and CD34–/Lin–subsets, were identified and demonstrated to differentiate intoCD34+ progenitors, initiate multilineage hematopoiesis, andreconstitute the lymphohematopoietic system [130–132],suggesting that the CD34– cell population may containmore primitive cells. Singer et al. described adherent cellsfrom bone marrow contain cells with hematopoietic as well asstroma-like features [133, 134], whereas Dominici et al. reportedthat plastic nonadherent population from bone marrow can generateboth functional osteoblasts/osteocytes and hematopoietic cells. Are we very close to a common hematopoietic-mesenchymalprecursor? Some investigators argued that all these versatilestem cells are in fact different subpopulations of tissue-committedstem cells .
The following observations directly made on the interleukin(IL)-6-mediated CFU-Fs from both BM and PB held great interest.Huss et al. [26, 56, 137–139] found that round cells developedas clusters from the CD34– adherent fibroblastic celllayer during culture especially when the adherent cells reached80%–90% confluency. The round cells expressed specifichematopoietic markers such as CD34, HLA-DR, c-kit, myeloid antigenDM5, MHC class II antigens, and so on, in different degrees.Although adherent cells were negative or had very low expressionlevels for these markers . In addition, the round cellwould reattach and proliferate in an adherent fashion if increasingnumbers of cells detached . Similar phenomena were alsoobserved by Rogers and Berman . After they treated thehematopoietic cell-eliminated stromal layer of murine cultureswith tumor necrosis factor-, cells showed bursts of hematopoieticactivity. In vivo study indicated that IL-6-mediated CFU-Fsdifferentiated into osteocalcin-positive bone-lining cells besides achieving the hematopoietic reconstruction .
Thus the "stem cell cycle" was postulated by Huss [27, 141].In his stem cell cycle model, CD34– fibroblast-like cells"contain hematopoietic pluripotency and a certain number ofthose cells circulate in the peripheral blood" ; "theycan still return to their setting environment within the marrowstroma" , becoming the quiescent stem cells; the quiescentstem cells, once activated, could differentiate into both hematopoieticstem cells and mesenchymal stem cells. The quiescent stem cellin this proposal is obviously a common hematopoietic mesenchymalstem cell. Although this is a fascinating paradigm, the ideaof a common hematopoietic-mesenchymal stem cell still remainscontroversial. Neither sufficient mesenchymal characteristicwere proved in those IL-6-mediated CFU-Fs nor were abundanthematopoietic characteristics demonstrated in the classic media-developedCFU-Fs. Furthermore, undetected hematopoietic stem cell contaminationof mesenchymal stromal cells cannot be ruled out in those resultson which the stem cell cycle proposal was based. More experimentaldata are needed to confirm if there is a common precursor inadult for mesenchymal and hematopoietic lineage cells.
Multipotent Mesenchymal Stromal Cells from Other Sources
Besides the BM and PB, other organs and tissues in adults werealso shown to be sources of MSCs, including pleural cavity,spleen, thymus, peritoneal cavity, lymph node, adipose tissue,muscle, brain, and exfoliated deciduous teeth [1, 8, 57, 86,142–145]. Although Wexler et al.  failed to identifyMSCs in umbilical cord blood (UCB) from full-term deliveries,the isolation of MSCs from UCB was successful in other laboratories[62, 146–149]. It was found that human first-trimesterfetal blood contained more MSCs than blood from the second andthird trimesters . Naruse et al.  isolated MSCs notonly from the UCB but also from the entire circulating bloodof fetal rat. It was reported that the embryonal circulatingMSCs can be retrovirally transduced with 99% efficiency withoutselection . A cell therapy using embryonal circulatingMSCs has also been demonstrated .
MSCs from other sources shared many characteristics with BMMSCsor PBMSCs but still showed some differences from them or betweeneach other in phenotype, proliferation, and differentiationabilities. In addition, the experimental results from differentlaboratories are far from consistent. For example, adipose tissue-derivedMSCs did not express CD106 ; Mareschi et al.  failedto confirm the adipocytic, osteogenic, and chondrocytic differentiationability of MSCs from the UCB using similar inductive conditionsfor BMMSCs.
At the current stage of investigation, neither those additionalorgans/tissues nor the adult peripheral blood can serve as areliable source for MSCs. The bone marrow is the richest andmost reliable reservoir for MSCs. But the exploration of theexistence of MSCs in other sources is important for understandingmesenchymal cell biology. The transit of MSCs in the embryonalcirculation and their distribution in other organs and tissuesin adulthood may help answer many unsolved questions such asthe origin and destination of the PBMSCs and their relationshipwith BMMSCs.
Enter the code exactly as it appears. All letters are case insensitive, there is no zero.