Multipotent Progenitor Cells Are Present in Human Peripheral Blood

ISSN: 1524-4571 Copyright © 2009 American Heart Association. All rights reserved. Print ISSN: 0009-7330. Online

TX 72514Circulation Research is published by the American Heart Association. 7272 Greenville Avenue, Dallas,

DOI: 10.1161/CIRCRESAHA.109.195859 published online Apr 23, 2009; Circ. Res.

Annarosa Leri, Claudio Schneider, Carlo Alberto Beltrami and Piero Anversa Toffoletto, Stefania Marzinotto, Laura Mariuzzi, Nicoletta Finato, Maura Pandolfi,D'Aurizio, Roberto Verardo, Silvano Piazza, Enio Klaric, Renato Fanin, Barbara

Daniela Cesselli, Antonio Paolo Beltrami, Silvia Rigo, Natascha Bergamin, Federica Multipotent Progenitor Cells Are Present in Human Peripheral Blood

http://circres.ahajournals.org/cgi/content/full/CIRCRESAHA.109.195859/DC1Data Supplement (unedited) at:

http://circres.ahajournals.org

located on the World Wide Web at: The online version of this article, along with updated information and services, is

http://www.lww.com/reprintsReprints: Information about reprints can be found online at

[email protected]. E-mail:

Fax:Kluwer Health, 351 West Camden Street, Baltimore, MD 21202-2436. Phone: 410-528-4050. Permissions: Permissions & Rights Desk, Lippincott Williams & Wilkins, a division of Wolters

http://circres.ahajournals.org/subscriptions/Subscriptions: Information about subscribing to Circulation Research is online at

by on May 21, 2011 circres.ahajournals.orgDownloaded from


http://circres.ahajournals.org/cgi/content/full/CIRCRESAHA.109.195859/DC1

http://circres.ahajournals.org/subscriptions/

mailto:[email protected]

http://www.lww.com/reprints


Multipotent Progenitor Cells Are Present in HumanPeripheral Blood

Daniela Cesselli, Antonio Paolo Beltrami, Silvia Rigo, Natascha Bergamin, Federica D’Aurizio,Roberto Verardo, Silvano Piazza, Enio Klaric, Renato Fanin, Barbara Toffoletto, Stefania Marzinotto,

Laura Mariuzzi, Nicoletta Finato, Maura Pandolfi, Annarosa Leri, Claudio Schneider,Carlo Alberto Beltrami, Piero Anversa

Abstract—To determine whether the peripheral blood in humans contains a population of multipotent progenitor cells(MPCs), products of leukapheresis were obtained from healthy donor volunteers following the administration ofgranulocyte/macrophage colony-stimulating factor. Small clusters of adherent proliferating cells were collected, andthese cells continued to divide up to 40 population doublings without reaching replicative senescence and growth arrest.MPCs were positive for the transcription factors Nanog, Oct3/4, Sox2, c-Myc, and Klf4 and expressed several antigenscharacteristic of mesenchymal stem cells. However, they were negative for markers of hematopoietic stem/progenitorcells and bone marrow cell lineages. MPCs had a cloning efficiency of �3%, and following their expansion, retaineda highly immature phenotype. Under permissive culture conditions, MPCs differentiated into neurons, glial cells,hepatocytes, cardiomyocytes, endothelial cells, and osteoblasts. Moreover, the gene expression profile of MPCs partiallyoverlapped with that of neural and embryonic stem cells, further demonstrating their primitive, uncommitted phenotype.Following subcutaneous transplantation in nonimmunosuppressed mice, MPCs migrated to distant organs and integratedstructurally and functionally within the new tissue, acquiring the identity of resident parenchymal cells. In conclusion,undifferentiated cells with properties of embryonic stem cells can be isolated and expanded from human peripheralblood after granulocyte/macrophage colony-stimulating factor administration. This cell pool may constitute a uniquesource of autologous cells with critical clinical import. (Circ Res. 2009;104:00-00.)

Key Words: stem cells � stem cell plasticity � circulating progenitors � multipotency � gene expression profile

The recognition that the peripheral blood (PB) in adulthumans may contain a population of multipotent progen-

itor cells (MPCs) would have relevant clinical implicationsfor the treatment of multiple diseases in several organs.1 Thispossibility would facilitate the acquisition and expansion ofthis cell pool and their subsequent delivery. To be effectiveclinically, MPCs should have the ability to engraft, grow, anddifferentiate in cell lineages specific of the damaged tissueand eventually lead to a structural and functional recovery ofthe injured organ.2 Whether this undifferentiated cell pool isfound in all individuals or is restricted to a subset of patientsis as critical as the characterization of their phenotypicproperties. Additionally, the gene expression profile andgrowth behavior of circulating MPCs must be compared withthe properties of resident progenitor cells from solid andnonsolid organs such as the bone marrow, heart, and liver.3,4

This information is critical to determine whether they consti-tute an independent cell class or possess features of tissue-specific stem cells that migrated from the organ of origin tothe circulation.

Two populations of progenitor cell– defined colony-forming units fibroblastic cells have previously been identi-fied in human PB5,6: fibrocytes and mesenchymal stromalcells (MSCs). Fibrocytes are immunophenotypically similarto monocytes, grow modestly in culture and show a longpopulation doubling time.5 Conversely, MSCs display themorphology and surface phenotype of bone marrow residentmesenchymal cells and, occasionally, can be expanded invitro.6 Although fibrocytes may generate ectodermal, meso-dermal, and endodermal structures,7 the differentiation poten-tial of MSCs has been restricted to the mesoderm.8 Thelimitation in growth of fibrocytes and the lineage restrictionof MSCs strengthens the need for identification of acirculating MPC that combines the plasticity of fibrocyteswith the growth of bone marrow MSCs. This is relevant forthe heart because there is no documentation yet that MSCsin mammals acquire the cardiomyocyte lineage in vitro orin vivo.

An essential premise of this putative human MPC mustinvolve its ability to migrate to distant organs in an animal

Original received February 16, 2009; revision received March 29, 2009; accepted April 15, 2009.From the Center for Regenerative Medicine (D.C., A.P.B., S.R., N.B., F.D., R.F., B.T., S.M., L.M., N.F., M.P., C.A.B.), University of Udine, Italy;

National Biotechnology Laboratory (R.V., S.P., E.K., C.S.), Padriciano, Trieste, Italy; and Departments of Anesthesia and Medicine (A.L., P.A.),Cardiovascular Division, Brigham and Women’s Hospital, Harvard Medical School, Boston, Mass.

Correspondence to Piero Anversa, MD, Departments of Anesthesia and Medicine, Cardiovascular Division, Brigham and Women’s Hospital, 75Francis St, Boston, MA 02115. E-mail [email protected]

© 2009 American Heart Association, Inc.

Circulation Research is available at http://circres.ahajournals.org DOI: 10.1161/CIRCRESAHA.109.195859

1 by on May 21, 2011 circres.ahajournals.orgDownloaded from


model, home to the tissue, survive in the unfamiliar milieu,differentiate, and ultimately acquire the function of residentparenchymal cells. If this were to be successful, this categoryof human MPCs may promote tissue repair and restore thestructural and physiological integrity of the damaged organ.

Materials and MethodsSamplesHuman specimens were collected in accordance with the Declarationof Helsinki and with approval by the Independent Ethics Committeeof the University of Udine. Bone marrow samples (n�20) wereobtained by iliac crest aspiration. For stem cell donors, granulocyte/macrophage colony-stimulating factor (GCSF) mobilization regimenconsisted of injections of 10 �g/kg GCSF for 4 to 5 consecutivedays. Leukapheresis (n�32) was performed on day 5 using anautomated cell separator. Buffy coats (n�50), derived from nonmo-bilized healthy donors and destined to be discarded, were collected.

MPC and MSC ExpansionMononuclear cells were isolated through density gradient centrifu-gation (Biocoll, Biochrom) (see the expanded Materials and Methodssection in the online data supplement, available at http://circres.ahajournals.org). Bone marrow multipotent adult stem cells(MASCs) were obtained as described previously.4 Online Table I inthe online data supplement illustrates MPC and MSC cultureconditions. Proliferating MPCs were subcultured.4

Fluorescence-Activated Sell-Sorting Analysisand CloningCultured cells and mechanically disaggregated spleen cells werestained using directly conjugated primary antibodies and analyzed byfluorescence-activated sell-sorting (FACS) analysis (CyAn,Beckman-Coulter). To generate single-cell–derived clones, a cellsorter (MoFlo, Beckman-Coulter) was used to automatically depositindividual cells in each well of 96-well Terasaki plates (n�1200wells).

Telomere Repeat Amplification ProtocolTelomerase activity was detected with the TRAPeze kit (Chemicon).

Multilineage Differentiation In VitroPolyclonal (n�3) and clonogenic MPCs (n�39) were cultured inspecific media4 to induce their differentiation into osteoblasts,endothelial cells, muscle cells, and neuronal cells. In all cases, cellswere fixed and lineage commitment was determined by histologicalanalysis, immunolabeling, and confocal microscopy (Online TableII). Additionally, aliquots of differentiated cells were used forfunctional assays. The supernatant of cells cultured in hepatocyte andneural differentiation media were collected.

Microarray and RT-PCR AnalysisTotal RNA was extracted using TRIzol and treated with DNase I.RNA was reverse-transcribed, and cDNA was amplified using theappropriate primers (Online Table III). Microarray analysis wasperformed as described previously.4

In Vivo Administration of MPCsUndifferentiated MPCs at passage 3 were harvested resuspended inPBS at a final concentration of 4�103 cells per microliter. AdultBALB-C mice (n�6, Harlan) were injected subcutaneously with 106

MPCs suspended in PBS, and the mice were euthanized 2 weekslater. Lungs, liver, spleen, heart, brain, bone marrow, femur, skeletalmuscles, and the subcutis in proximity of the injection site werecollected. Tissues were fixed in formalin or snap-frozen for immu-nolabeling and immunohistochemical studies, respectively (OnlineTable II).

Statistical AnalysisResults are expressed as means�SD. Statistical significance wasdetermined by unpaired Student’s t test; P�0.05 was consideredsignificant.

Experimental protocols are described in detail in the online datasupplement.

Results

Isolation and Growth of Human MPCsThis study included 102 healthy volunteers divided in 3groups: bone marrow (n�20), blood (n�50), and hematopoi-etic stem cell (n�32) donors. In the latter case, stem cellmobilization was induced by GCSF administration, and 1 mLof product of leukapheresis was obtained from each individ-ual to isolate MPCs. Bone marrow samples were used tocollect MASCs, which differentiate in endodermal, mesoder-mal, and ectodermal cell lineages.4 These cells were used forcomparison with MPCs. PB, in the absence of stem cellmobilization, was used to culture fibrocytes and MSCs.Therefore, the objective was to compare the growth behaviorof MASCs,4 fibrocytes,5 MSCs,6 and putative MPCs.

MASCs were consistently obtained in all 20 bone marrowsamples (Figure 1A). Fibrocytes from PB were easily cul-tured but, as repeatedly shown,5,7,8 had minimal growthcapacity and expressed hematopoietic surface antigens (Fig-ure 1B and 1C and Online Figure I). We were unable toisolate MSCs from 50 samples of 50 mL each of human PB.Although 6 different growth media were tested (Online TableI), MSCs were not identified or failed to survive and expandunder these conditions. This negative result questions thesignificance of this cell class for the colonization of the boneby circulating progenitors in humans.6 Age may be a criticaldeterminant of the pool size of MSCs in the blood,9 possibly

Figure 1. Phenotypic properties of MASCs and fibrocytes.Phase-contrast microscopy. A, Single MASCs form multicellularclones. B and C, Fibrocytes shown at low (B) and higher (C)magnification adhere to the plastic substrate but do not spread.

2 Circulation Research May 22, 2009



Figure 2. Phenotypic proper-ties of MPCs. A and B, Cul-tures of MPCs illustrated atlow (A) and higher (B) magnifi-cation. MPCs adhere to theplastic substrate, spread, andform colonies. Arrows definethe edge of the colony. C,Detection of Oct3/4, Nanog,Sox2, c-Myc, and Klf4 inMPCs and fibrocytes byRT-PCR. D and E, Oct3/4(green) (D) and Nanog(magenta) (E) proteins in MPCsare shown by immunolabeling.F, By telomere repeat amplifi-cation protocol assay, prod-ucts of telomerase activity in 2MPC cell lines (1 and 2) dis-play a 6-bp periodicity. Heat-inactivated lysates (i�) wereused as negative control.TSR8 indicates telomerasecontrol template. The 1301 cellline was used as a positivecontrol (c�). Negative control(c�): primer–dimer lane.

Cesselli et al Circulating Human Multipotent Progenitor Cells 3



accounting for the difficulty to compare findings from differ-ent laboratories.

A complex procedure had to be implemented to isolatefrom 1 mL of product of leukapheresis an adherent fibroblast-like cell population (see Online Materials and Methods),which grew rapidly and retained an undifferentiated pheno-type at serial passages (Figure 2A and 2B). Positive resultswere obtained in 5 cases, most likely because of the limitedamount of product of leukapheresis available for analysis. Atmost, 2 to 3 clusters of adherent proliferating cells wereacquired from each sample, documenting that the quantity athand represented a critical determinant of the successfulcollection and growth of MPCs. In these positive examples,however, primitive cells continued to divide up to 40 popu-lation doublings without reaching replicative senescence andgrowth arrest. MPCs had a population doubling time of 38�3hours, which was significantly shorter than that of MASCs(48�2 hours; P�0.0001).

In primary culture and after 3 and 20 population doublings,MPCs were positive for transcription factors typical of humanembryonic stem (ES) cells, Nanog, Oct3/4, Sox2, c-Myc, andKlf4 (Figure 2C through 2E). MPCs manifested high levels oftelomerase activity, which exceeded that present in MASCs(Figure 2F). Telomerase activity was not detected in fibro-

cytes. Additionally, fibrocytes were negative for the pluripo-tency markers, Nanog, Oct3/4, Sox2, c-Myc, and Klf4 (Fig-ure 2C). By FACS analysis, several antigens characteristic ofmesenchymal stem cells were identified in MPCs. Theyincluded CD90, CD29, CD13, CD105, CD44, and CD73(Online Figure II). MPCs were negative for markers ofhematopoietic stem/progenitor cells, including CD45, CD34,CD38, CD117, and CD133 and bone marrow cell epitopessuch as CD14 (Figure 2G), CD4, CD8, CD19, and CD20(data not shown).

MPCs Are Clonogenic and MultipotentStem cells possess 3 fundamental characteristics; they areself-renewing, clonogenic, and multipotent in vitro and invivo.10 To strengthen the notion that MPCs represented astem cell class, individual MPCs were FACS sorted andautomatically deposited in single wells of Terasaki plates.11

Wells containing �1 cell were excluded from the analysis.Over a period of 2 to 3 weeks, 90 single-cell–derivedmulticellular clones were obtained from a total of 1200deposited cells. The number of cells per clone varied from100 to 500 (Figure 3A). Of the 90 clones, 39 were expand-able, whereas 51 reached rapidly growth arrest. Therefore,MPCs had an actual cloning efficiency of �3%, which was

Figure 3. MPCs are self-renewing and clonogenic and form ectodermal deriv-atives. A, Cellular clone formed by the deposition of an individual MPC in asingle well of a Terasaki plate. B, Detection of Oct3/4 (green) and Nanog(magenta) in clonogenic MPCs. C through F, Clonogenic MPCs in neurogenicmedium express synaptophysin (green) (C), neuron-specific enolase (green)(D), �3-tubulin (red) (E and F), and GFAP (green) (F). G, Differentiating clono-genic MPCs generate L-glutamate.




significantly higher than that commonly found with residentprogenitor cells from solid organs.11,12 Conversely, this valuewas lower than that obtained with mesenchymal cells fromthe bone marrow, heart, and liver.4 Clonogenic MPCs weregrown and characterized; they continued to express Nanogand Oct3/4 and displayed surface antigens essentially identi-cal to those of MPCs in primary culture (Figure 3B andOnline Figure III). Thus, during serial expansion, MPCsretained their original identity and immature phenotype,indicating that stem cells can be mobilized into the PB andcollected and amplified in vitro in a selective manner.

To establish the differentiation potential of clonogenicMPCs, cells from single clones were exposed to neurogenic,myogenic, osteogenic, endothelial cell (EC), and hepatocyte-forming media, respectively.4 In neurogenic medium, MPCsacquired neuron-like cell morphology and expressed synap-tophysin, neuron-specific enolase, and �3-tubulin arranged infilaments and bundles (Figure 3C through 3E). A fractionof MPCs maintained a flattened morphology and stainedfor glial fibrillary acidic protein (Figure 3F). Functionalcompetence of MPC-derived neurons was documented byspontaneous release of the neurotransmitter L-glutamate(Figure 3G).

When clonogenic MPCs were exposed to myogenic me-dium, cardiomyocytes were formed; they expressed�-sarcomeric actin and �-cardiac actinin (Figure 4A and 4B).The gap junction protein connexin 43 was detected at sites ofcell-to-cell contact (Figure 4A), and ryanodine receptors weredistributed between filaments of �-actinin (Figure 4B). Spon-taneous intracellular calcium oscillations were recorded in thepresence of the calcium indicator Fluo-4 (Online Movie I).The ability of MPCs to commit to other mesodermal celllineages was shown by culturing these cells in osteogenic and

EC-forming media. Differentiating cells expressed osteopon-tin and incorporated tetracycline in areas of active calciumdeposition (Online Figure IV, A). Following stimulation withvascular endothelial growth factor (VEGF), the majority ofcells stained for von Willebrand factor and were able touptake acetylated LDL (Figure 4C and 4D).

The endodermal potential of MPCs was tested using amedium specific for hepatocyte differentiation. MPCs pro-gressively assumed a globular shape and stained positive forcytokeratins 18 and 19 (Figure 4E). Developing hepatocytesaccumulated glycogen, released albumin, and, in response tophenobarbital, upregulated the expression and activity ofcytochrome-CYP2B6 (Figure 4F and 4G and Online FigureIV, B). Similar differentiation properties were identified innonclonogenic MPCs that formed cells of the 3 germ layers(Figure 5A through 5G and Online Figure V).

Gene Expression Profile of MPCscDNA microarray technology was used to characterize globalgene expression of MPCs. Subsequently, the transcriptionalprofile of MPCs was compared with that of PB fibrocytes,mononuclear blood cells, and tissue resident MASCs. Unsu-pervised hierarchical clustering analysis was performed ondifferentially expressed genes (see the expanded Materialsand Methods section in the online data supplement). A clusterdendrogram revealed a clear separation between MPCs andMASCs, on the one hand, and fibrocytes and mononuclearblood cells, on the other (Figure 6A). Quantitatively, fibro-cytes and mononuclear blood cells showed an almost identi-cal transcriptional profile. However, significant differenceswere detected between these cell classes and MPCs andMASCs, raising questions about the undifferentiated, multi-potent state of fibrocytes.5,7 Conversely, the similarity in gene

Figure 4. MPCs differentiate into mesodermal and endodermal lineages. A and B, MPCs form cardiomyocytes expressing �-sarcomericactin (red) (A), cardiac �-actinin (red) (B), connexin 43 (green) (A), and ryanodine receptors (green) (B). C and D, MPCs commit to theEC lineage and express von Willebrand Factor (green) (C) and uptake DiI-labeled acetylated LDL (red) (D). E through G, MPCs form he-patocytes that show cytokeratins 8 to 18 to 19 (red) (E), accumulate glycogen granules (PAS staining) (F), and produce albumin (G).




expression of MPCs and MASCs appeared to reflect theclonogenicity and multipotentiality of these cell categories(Figure 6B and Online Table IV).

In comparison with fibrocytes and mononuclear bloodcells, MPCs and MASCs showed 6560 differentially ex-pressed genes (adjusted P�0.001). Gene ontology–basedfunctional annotation and EASE analysis13,14 were used todetect the biological function of the listed genes. Multipotentcells were enriched in transcripts for extracellular matrixproteins and intracellular protein trafficking, whereas fibro-cytes and mononuclear blood cells were characterized bythe expression of genes implicated in RNA processing,immune response, antigen presentation, immunoglobulins,and major histocompatibility complex domains (OnlineTables V and VI).

MPCs and MASCs expressed a core of common genes,consistent with the stem cell properties of both cell types.

However, a set of differentially expressed genes was alsoidentified. To establish the biological import of these find-ings, analysis of functional annotation was performed.15 Withrespect to MASCs, genes characteristic of neural and ES cellswere upregulated in MPCs (Online Figure VI), suggestingthat this cell pool possessed a more immature embryonic-likephenotype.

Behavior of MPCs In VivoTo determine whether MPCs can enter the systemic circula-tion, migrate to distant organs, and acquire mesodermal,endodermal and ectodermal cell lineages, these cells wereinjected subcutaneously in nonimmunosuppressed mice, andthe presence of human structures was evaluated 2 weeks laterin the spleen, heart, liver, brain, and skin. Nonimmunosup-pressed mice were chosen in view of the immunomodulatoryproperties of bone marrow mesenchymal cells. Human-specific anti-mitochondria and anti–nuclear lamin A/C anti-bodies were used to identify human cells in treated mice.With the exception of the heart, human cells were found in allother organs examined. These cells were positive for thehematopoietic makers CD34 and CD45 and were locatedunderneath the capsule of the spleen in close proximity offoci of murine hematopoiesis (Figure 7A through 7F). Col-lectively, they comprised 0.05% of the entire cell pool(Figure 7G). Clusters of developing human hepatocyteswhich expressed cytokeratin 7 were detected in the mouseliver (Figure 7H through 7K). Additionally, scattered humancells resembling neurons were seen in the brain (Figure 8Athrough 8C); in the skin, vessels of human origin wererestricted to the dermis (Figure 8D through 8G). In all cases,fluorescence in situ hybridization for human X and Ychromosome, chromosome 17, and human HER2 gene wasimplemented to confirm the human origin of the structuresgenerated by engraftment and commitment of MPCs.

DiscussionIn the present study, we have identified a novel population ofcells which circulate in the PB and have properties of stemcells: clonogenicity, self-renewal, and multipotentiality.MPCs manifest surface epitopes similar to those present inMSCs but display a much more immature phenotype withmolecular and functional characteristics restricted to humanES cells.16 MPCs express the pluripotent state-specific tran-scription factors, Oct3/4, Nanog, and Sox2 together withc-Myc and Klf4. Additionally, they show high levels oftelomerase activity, proliferate intensively in vitro, and,under permissive culture conditions, form cells of the 3germ layers.17 Following transplantation in vivo, MPCsmigrate to distant organ and integrate within the newparenchyma, acquiring the identity of resident cells struc-turally and functionally.

The membrane immunophenotype of MPCs favors theirmesenchymal ancestry, which is consistent with the complex-ity to detect these cells in the PB.8,18 The relatively low yieldof MPCs after GCSF mobilization obtained in our study maybe attributable to the small samples available for cell collec-tion, inherent resistance of primitive cells of mesenchymalorigin to migrate out of the organ of residence,19 or both.

Figure 5. Differentiation of nonclonogenic MPCs into the 3 germlayers. A, MPCs form glial cells which express GFAP (green)and �3-tubulin (red). B through D, MPCs form bone, as shownby von Kossa staining (brown) (B), and differentiate into smoothmuscle cells that express smooth muscle actin (green) (C) andcardiomyocytes that are positive for �-sarcomeric actin (red) (D)and connexin 43 (green) (D). E, Commitment to the EC lineageis documented by the uptake of acetylated LDL (red). F and G,MPCs give rise to hepatocytes that express cytokeratins 8 to18 to 19 (green) (F) and accumulate glycogen (PAS staining,purple) (G).




However, the critical determinant of the difficulty encoun-tered in the isolation of MPCs may relate to the protocol usedfor the mobilization of bone marrow cells. Although GCSF isrelatively effective in mobilizing CD34� cells in the circula-tion,20 this strategy has not been optimized for nonhemato-poietic progenitors, which show an inverse relationship withthe number of mobilized CD34� cells.21 Age and disease ofthe donor may also affect the egress of cells from the bonemarrow.22 Importantly, it has recently been emphasized thatselective recruitment of bone marrow cell subsets is largelydependent on the cytokines used to promote cell migration,23

providing the basis for the limited number of positive samplesobtained here. Unfortunately, the demographic and clinicalcharacteristics of volunteer donors and patients were notavailable to us; therefore, we could not conclude that age, sex,and the clinical conditions did not influence the number ofcirculating progenitors and their successful collection.

The retention of progenitor cells in bone marrow niches isprimarily dictated by the interaction between CXCR4 ex-pressed on the surface of the primitive cells and the stromalcell–derived factor-1 ligand released by neighboring support-ing cells.24,25 The mechanism by which GCSF favors thetranslocation of progenitors to PB involves the disruption ofthe stromal cell–derived factor-1/CXCR4 axis by reducingthe expression of both receptor and ligand.26 This process isachieved more effectively by CXCR4 antagonists. Withrespect to GCSF, CXCR4 antagonists result in a more than10-fold higher number of hematopoietic progenitors (HPCs)and endothelial progenitor cells in PB. However, GCSF aloneor in combination with CXCR4 antagonists does not inducemobilization of a class of mesenchymal stem cells,27 thestromal progenitor cells (SPCs). Conversely, the simulta-neous administration of CXCR4 antagonists and VEGFsuppresses migration of HPCs but favors the egress of SPCsfrom the bone marrow, although the number of SPC-derivedcolonies per milliliter of blood remains several orders of

magnitude lower than that of effectively mobilized HPCs andendothelial progenitor cells.27 These observations demon-strate that various subsets of bone marrow progenitor cellsrequire the administration of different growth factors andreiterate the difficulty to disengage mesenchymal cells fromtheir microenvironment. Optimal protocols for the mobiliza-tion of bone marrow cells into the PB remain to bedeveloped.23

The products of leukapheresis used in this work derivedfrom GCSF treatment, which is the most effective protocolapproved for bone marrow cell mobilization in patients.28

However, mRNA levels for CXCR4 are significantly lower inMPCs than in the other circulating bone marrow cellsexamined here, weakening the mobilizing action of GCSFand preventing the use of CXCR4 antagonists. Similarly,MPCs do not possess VEGF receptor (VEGFR)2 and are,therefore, not responsive to VEGF. Careful analysis of thepattern of surface receptors present on MPCs may provideuseful insights for optimizing MPC migration and harvesting.MPCs have relatively high levels of VEGFR1/flt1 that, onbinding with placental growth factor, induces rapid chemo-taxis of bone marrow repopulating and progenitor cells andattracts VEGFR1-positive cells at sites of tumor growthforming premetastatic niches.29,30 MPCs express IGFBP5,which in combination with insulin-like growth factor (IGF)-1and IGF-2, promotes migration of undifferentiated mesenchy-mal cells.31 However, MPCs possess a group of adhesionmolecules including integrin �3, the integrin-related IT-GBL1, VCAM1, L1CAM, and OBCAM that may favor theretention of these cells in the bone marrow niches, furthercomplicating their access to PB.

Although various hematopoietic and nonhematopoieticmultipotent primitive cells have been recognized in humanPB,8,18,32 MPCs have unique characteristics in terms ofgrowth, differentiation, and gene expression profile. At pres-ent, MPCs are the only circulating nonhematopoietic cells

Figure 6. Gene expression analysis. A, With respect to fibrocytes and blood mononuclear cells, MPCs and tissue-derived MASCsexpress different genes. Colored bars identify the different sample groups with respect to the dendrogram. B, Heat map representationof Pearson correlation coefficients among the 4 different cell types. C, Heat map illustration showing analysis of functional annotationresults. “MPC UP” and “MPC DW” indicate, respectively, functional sets enriched for genes upregulated or downregulated in MPCswith respect to MASCs. Color-coded values; log10 (raw probability values of the Wilcoxon rank sum test).




that form expandable clones composed of a homogenous poolof stem cells with broad differentiation potential. Because thisis a fundamental feature of human ES cells, it suggests thatMPCs are present during embryogenesis and persist in adultlife. Multipotency is a highly desired property of adult somaticcells. Effort has been made to reprogram adult fibroblasts intoa pluripotent state by forcing the expression of 4 or fewertranscription factors, including Oct3/4, Sox2, c-Myc, andKlf4.33 On injection in the blastocysts, induced-pluripotentstem cells display the ability to form ectodermal, mesoder-mal, and endodermal derivatives, together with germ cells.

However, the efficiency of reprogramming is low underscor-ing the significance of the cell category identified in thepresent report; MPCs are a class of multipotent cells that arepresent in the PB and do not require retroviral gene transduc-tion. The expression of pluripotent state–specific transcrip-tion factors has previously been detected in adult MSCsisolated directly from the bone marrow.34

The transcriptional profile of MPCs differs markedly fromthat of fibrocytes and mononuclear blood cells but displayssimilarities with MASCs. The distinct molecular signature ofthese cell categories points to differential biological proper-

Figure 7. In vivo engraftment of MPCs. Athrough F, MPCs form hematopoietic cells inthe spleen. The human origin of the differenti-ated cells was documented by human mito-chondrial antigen (brown) (A and B), humanlamin A/C (green) (C), and human chromo-some 17 (white dots in nuclei) (D). E and Fcorrespond to consecutive sections illustratingthe same cell cluster shown in A and B.Hematopoietic cells are positive for humanCD45 (brown) (E) and for human CD34 (brown)(F). G, FACS analysis of human CD45-positivecells from the mouse spleen. H through K,MPCs engraft in the liver forming hepatocytesthat are positive for human lamin A/C (green)(H), human mitochondrial antigen (brown) (I), human X (red) and Y (green) chromosome (J), and human cytokeratin 7 (brown) (K).




ties of each cell pool, which may have independent functionin organ and organism homeostasis and repair. MASCs andMPCs have a superior and broader differentiation potentialthan MSCs. Importantly, adult MPCs may be ontologicallyrelated to ES cells; they share with ES cells multiple genesand behavioral phenotypes related to self-renewal, multipo-tentiality, and lineage specification.35 In fact, MPCs arehighly primitive cells that express pathways characteristicallypresent in neural and ES cells.36 MPCs and MASCs have apartially overlapping gene expression profile. However, theset of differentially expressed genes between these 2 cellclasses includes highly immature markers that predominate inMPCs. This finding suggests that MPCs may represent theancestors of MASCs in the hierarchy of bone marrow cellswith mesenchymal surface immunophenotype.

The motile phenotype of circulating MPCs reflects theirability to migrate to distant tissues and successfully engraftand differentiate into the cell lineages of the recipient organ.Consistent with their remarkable cellular plasticity in vitro,MPCs can form ECs in the vessel wall, hepatocytes in the

liver, and neurons in the brain. Additionally, the in vivoadministration of MPCs unmasked their latent ability tointegrate in sites of extramedullary hematopoiesis. Thepresence of clusters of CD45-positive CD34-positive hu-man cells in the mouse spleen has raised 2 possibilitiesconcerning the blood-forming potential of nonhematopoi-etic progenitors or the participation of MPCs in the “stem cellcycle.”37,38 The latter may mirror a chronic oscillation from aquiescent mesenchymal-like state to an active hematopoieticphenotype.

Sources of FundingSupported by Ministero dell’Universita e della Ricerca Scientifica eTecnologica Progetti di Ricerca di Interesse Nazionale grant2006060854; Friuli-Venezia Giulia Regional Grant LR 11/2003, art.11; and grants from the NIH.

DisclosuresNone.

References1. Dimmeler S, Zeiher AM, Schneider MD. Unchain my heart: the scientific

foundations of cardiac repair. J Clin Invest. 2005;115:572–583.

Figure 8. MPCs engraft in the brain and accumulate at the site of injection. A through C, MPCs homed to the brain have the morpho-logical characteristics of neurons. These cells are human in origin, as documented by human mitochondrial antigen (brown). D throughG, Human MPCs are positive for the human mitochondrial antigen (brown) (D), human chromosome 17 (red) and HER2 (white) (E), andhuman X (red) and Y (green) chromosomes (F). G, Engrafted MPCs differentiate into ECs (CD31, green) in dermal vessels.




2. Anversa P, Kajstura J, Leri A, Bolli R. Life and death of cardiac stemcells. A paradigm shift in cardiac biology. Circulation. 2006;113:1451–1463.

3. Cho HJ, Lee N, Lee JY, Choi YJ, Ii M, Wecker A, Jeong JO, Curry C, QinG, Yoon YS. Role of host tissues for sustained humoral effects afterendothelial progenitor cell transplantation into the ischemic heart. J ExpMed. 2007;204:3257–3269.

4. Beltrami AP, Cesselli D, Bergamin N, Marcon P, Rigo S, Puppato E,D’Aurizio F, Verardo R, Piazza S, Pignatelli A, Poz A, Baccarani U,Damiani D, Fanin R, Mariuzzi L, Finato N, Masolini P, Burelli S,Belluzzi O, Schneider C, Beltrami CA. Multipotent cells can be generatedin vitro from several adult human organs (heart, liver and bone marrow).Blood. 2007;110:3438–3446.

5. Bucala R, Spiegel LA, Chesney J, Hogan M, Cerami A. Circulatingfibrocytes define a new leukocyte subpopulation that mediates tissuerepair. Mol Med. 1994;1:71–81.

6. Kuznetsov SA, Mankani MH, Gronthos S, Satomura K, Bianco P, RobeyPG. Circulating skeletal stem cells. J Cell Biol. 2001;153:1133–1140.

7. Zhao Y, Glesne D, Huberman E. A human peripheral blood henobarb-derived subset acts as pluripotent stem cells. Proc Natl Acad Sci U S A.2003;100:2426–2431.

8. He Q, Wan C, Li G. Concise review: multipotent mesenchymal stromalcells in blood. Stem Cells. 2007;25:69–77.

9. Dimmeler S, Leri A. Aging and disease as modifiers of efficacy of celltherapy. Circ Res. 2008;102:1319–1330.

10. Verfaille CM. ‘Adult’ stem cells: tissue specific or not? In: Lanza R, BlauH, Melton DA, Moore M, Thomas ED, Verfaillie C, Weissman I, WestM, eds. Handbook of Stem Cells. Burlington, Mass: Elsevier AcademicPress; 2004:13–20.

11. Beltrami AP, Barlucchi L, Torella D, Baker M, Limana F, Chimenti S,Kasahara H, Rota M, Musso E, Urbanek K, Leri A, Kajstura J, Nadal-Ginard B, Anversa P. Adult cardiac stem cells are multipotent and supportmyocardial regeneration. Cell. 2003;114:763–776.

12. Bearzi C, Rota M, Hosoda T, Tillmanns J, Nascimbene A, De Angelis A,Yasuzawa-Amano S, Trofimova I, Siggins RW, Lecapitaine N, CascaperaS, Beltrami AP, D’Alessandro DA, Zias E, Quaini F, Urbanek K, MichlerRE, Bolli R, Kajstura J, Leri A, Anversa P. Human cardiac stem cells.Proc Natl Acad Sci U S A. 2007;104:14068–14073.

13. Smyth GK. Linear models and empirical Bayes methods for assessingdifferential expression in microarray experiments. Stat Appl Genet MolBiol. 2004;3:Article3.

14. Hosack DA, Dennis G, Sherman BT, Lane HC, Lempicki RA. Identifyingbiological themes within lists of genes with EASE. Genome Biol. 2003;4:R70.

15. Schaeffer EM, Marchionni L, Huang Z, Simons B, Blackman A, Yu W,Parmigiani G, Berman DM. Androgen-induced programs for prostateepithelial growth and invasion arise in embryogenesis and are reactivatedin cancer. Oncogene. 2008;27:7180–7191.

16. Boyer LA, Lee TI, Cole MF, Johnstone SE, Levine SS, Zucker JP,Guenther MG, Kumar RM, Murray HL, Jenner RG, Gifford DK, MeltonDA, Jaenisch R, Young RA. Core transcriptional regulatory circuitry inhuman embryonic stem cells. Cell. 2005;122:947–956.

17. Wagers AJ, Weissman IL. Plasticity of adult stem cells. Cell. 2004;116:639–648.

18. Roufosse CA, Direkze NC, Otto WR, Wright NA Circulating mesen-chymal stem cells. Int J Biochem Cell Biol. 2004;36:585–597.

19. Fox JM, Chamberlain G, Ashton BA, Middleton J. Recent advances intothe understanding of mesenchymal stem cell trafficking. Br J Haematol.2007;137:491–502.

20. Cashen AF, Lazarus HM, Devine SM. Mobilizing stem cells from normaldonors: is it possible to improve upon G-CSF? Bone Marrow Transplant.2007;39:577–588.

21. Ripa RS, Haack-Sorensen M, Wang Y, Jorgensen E, Mortensen S,Bindslev L, Friis T, Kastrup J. Bone marrow derived mesenchymal cellmobilization by granulocyte-colony stimulating factor after acute myo-

cardial infarction: results from the Stem Cells in Myocardial Infarction(STEMMI) trial. Circulation. 2007;116(suppl I):I-24–I-30.

22. Hill JM, Zalos G, Halcox JP, Schenke WH, Waclawiw MA, QuyyumiAA, Finkel T. Circulating endothelial progenitor cells, vascular function,and cardiovascular risk. N Engl J Med. 2003;348:593–600.

23. Kolonin MG, Simmons PJ. Combinatorial stem cell mobilization. NatBiotechnol. 2009;27:252–253.

24. Aiuti A, Webb IJ, Bleul C, Springer T, Gutierrez-Ramos JC. The che-mokine SDF-1 is a chemoattractant for human CD34� hematopoieticprogenitor cells and provides a new mechanism to explain the mobili-zation of CD34� progenitors to peripheral blood. J Exp Med. 1997;185:111–120.

25. Peled A, Kollet O, Ponomaryov T, Petit I, Franitza S, Grabovsky V, SlavMM, Nagler A, Lider O, Alon R, Zipori D, Lapidot T. The chemokineSDF-1 activates the integrins LFA-1, VLA-4, and VLA-5 on immaturehuman CD34� cells: role in transendothelial/stromal migration andengraftment of NOD/SCID mice. Blood. 2000;95:3289–3296.

26. Levesque JP, Hendy J, Takamatsu Y, Simmons PJ, Bendall LJ. Dis-ruption of the CXCR4/CXCL12 chemotactic interaction during hemato-poietic stem cell mobilization induced by GCSF or cyclophosphamide.J Clin Invest. 2003;111:187–196.

27. Pitchford SC, Furze RC, Jones CP, Wengner AM, Rankin SM. Differ-ential mobilization of subsets of progenitor cells from the bone marrow.Cell Stem Cell. 2009;4:62–72.

28. Spitzer G, Adkins D, Mathews M, Velasquez W, Bowers C, Dunphy F,Kronmueller N, Niemeyer R, McIntyre W, Petruska P. Randomizedcomparison of G-CSF 1 GM-CSF vs G-CSF alone for mobilization ofperipheral blood stem cells: effects on hematopoietic recovery afterhigh-dose chemotherapy. Bone Marrow Transplant. 1997;20:921–930.

29. Hattori K, Heissig B, Wu Y, Dias S, Tejada R, Ferris B, Hicklin DJ, ZhuZ, Bohlen P, Witte L, Hendrikx J, Hackett NR, Crystal RG, Moore MA,Werb Z, Lyden D, Rafii S. Placental growth factor reconstitutes hema-topoiesis by recruiting VEGFR1(�) stem cells from bone-marrow micro-environment. Nat Med. 2002;8: 841�849.

30. Kaplan RN, Riba RD, Zacharoulis S, Bramley AH, Vincent L, Costa C,MacDonald DD, Jin DK, Shido K, Kerns SA, Zhu Z, Hicklin D, Wu Y,Port JL, Altorki N, Port ER, Ruggero D, Shmelkov SV, Jensen KK, RafiiS, Lyden D. VEGFR1-positive haematopoietic bone marrow progenitorsinitiate the pre-metastatic niche. Nature. 2005;438:820–827.

31. Fiedler J, Brill C, Blum WF, Brenner RE. IGF-I and IGF-II stimulatedirected cell migration of bone-marrow-derived human mesenchymalprogenitor cells. Biochem Biophys Res Commun. 2006;345:1177–1183.

32. Kucia M, Halasa M, Wysoczynski M, Baskiewicz-Masiuk M, MoldenhawerS, Zuba-Surma E, Czajka R, Wojakowski W, Machalinski B, RatajczakMZ. Morphological and molecular characterization of novel population ofCXCR4� SSEA-4� Oct-4� very small embryonic-like cells purifiedfrom human cord blood: preliminary report. Leukemia. 2007;21:297–303.

33. Takahashi K, Yamanaka S. Induction of pluripotent stem cells frommouse embryonic and adult fibroblast cultures by defined factors. Cell.2006;126:663–676.

34. Greco SJ, Liu K, Rameshwar P. Functional similarities among genesregulated by Oct4 in human mesenchymal and embryonic stem cells.Stem Cells. 2007;25:3143–3154.

35. Doherty JM, Geske MJ, Stappenbeck TS, Mills JC. Diverse adult stemcells share specific higher-order patterns of gene expression. Stem Cells.2008;26:2124–2130.

36. Ramalho-Santos M, Yoon S, Matsuzaki Y, Mulligan RC, Melton DA.“Stemness”: transcriptional profiling of embryonic and adult stem cells.Science. 2002;298:597–600.

37. Huss R, Lange C, Weissinger EM, Kolb HJ, Thalmeier K. Evidence ofperipheral blood-derived, plastic-adherent CD34(-/low) hematopoieticstem cell clones with mesenchymal stem cell characteristics. Stem Cells.2000;18:252–260.

38. Quesenberry PJ, Colvin GA, Lambert J-F. The chiaroscuro stem cell: aunified stem cell theory. Blood. 2002;100:4266–4271.




Supplement Material

Legends to Online Figures

Online Figure I. Surface phenotype of fibrocytes. Histogram plots of FACS data: fibrocytes

express several hematopoietic markers. Red, isotype controls; green, specific antibody

labeling; MFI, mean fluorescence intensity.

Online Figure II. Surface phenotype of MPCs. Histogram plots of FACS data: MPCs express

multiple mesenchymal surface antigens. Red, isotype controls; green, specific

antibody labeling; MFI, mean fluorescence intensity.

Online Figure III. Surface phenotype of clonogenic MPCs. Histogram plots of FACS data:

clonogenic MPCs express multiple mesenchymal surface antigens. Red, isotype

controls; green, specific antibody labeling; MFI, mean fluorescence intensity.

Online Figure IV. Mesodermal and endodermal differentiation. A, In osteogenic medium,

MPCs express osteopontin (green) and incorporate tetracyclines (red) at sites of

calcification. B, Differentiating MPCs upregulate cytochrome CYP2B6 in response to

phenobarbital. Panels i, ii and iii correspond to resorufin red fluorescence images of

the same cell aggregates which are shown in pseudo-colors (Q-LUT, scale bar) in

panels iv, v and vi. Panels i and iv depict cells not exposed to pentoxyresorufin (PR,

negative control); panels ii and v correspond to cells exposed to PR only and panels iii

and vi correspond to cells exposed to phenobarbital and PR.

Online Figure V. Differentiation of non-clonogenic MPCs into the three germ layers. A,

MPCs generate neurons which express NeuroD (green). B, MPCs form bone as shown

by tetracycline incorporation (red). C, Commitment to the EC lineage is documented

by the expression of von Willebrand factor (green).

Online Figure VI. Gene expression analysis. Heat map illustration showing AFA results.

“MPC UP” and “MPC DW” indicate, respectively, functional sets enriched for genes

up-regulated or downregulated in MPCs with respect to MASCs. Color-coded values;

Log10 (raw P values of the Wilcoxon rank sum test).



2

Online Tables

Online Table I. Culture conditions employed for the isolation of MPCs and MSCs

from steady state and mobilized PB.

Online Table II. Antibodies and staining protocols.

Online Table III. Primers and amplicon length.

Online Table IV. Pearson correlation coefficients of the correlations in gene expression

behavior among the four cell populations (MPCs, MASCs, Fibrocytes and

mononuclear blood cells).

Online Table V. Gene Ontology (GO) terms of transcripts upregulated in

fibrocytes/mononuclear blood cells with respect to MPCs/MASCs.

Online Table VI. Gene Ontology (GO) terms of transcripts upregulated in MPCs/MASCs

with respect to fibrocytes/mononuclear blood cells.

Online Movie I. Spontaneous calcium transients in cardiomyocytes derived from

differentiation of clonogenic MPCs cultured in myogenic medium. Cells were loaded

with Fluo-4 (green).

Online Materials and Methods

Stem Cell and Blood donors

In healthy donors, bone marrow (BM) samples (n=20) were obtained by iliac crest

aspiration. For stem cell donors, GCSF mobilization regimen consisted of injections

of 10 µg/kg GCSF – filgrastim (Amgen Inc., Thousand Oaks, CA, USA) once daily,

for 4-5 consecutive days. Leukapheresis (n=32) was performed on day 5 using an

automated cell separator. If the target yield of ≥4x106 CD34+ cells/kg recipient weight

was not achieved, an additional dose of filgrastim was administered on day 5, and

leukapheresis was repeated on day 6. Buffy coats (n=50), deriving from non-

mobilized healthy donors and destined to be discarded, were collected.

Primary cultures

BM Cells



3

BM cells were diluted in Hank’s balanced salt solution (HBSS, Sigma-Aldrich, St

Louis, MO, USA), layered on top of Ficoll (Biocoll, Biochrom AG, Berlin,

Germany), centrifuged, and washed twice with HBSS; 1.5x106 freshly isolated cells

were plated in Mesencult (Stem Cell Technologies, Vancouver, BC).

Mononuclear Blood Cells and Leukapheresis Products

The different culture conditions that we tested are summarized in Table I. Two

distinct enrichment strategies were followed to isolate and expand MPCs. At first, we

utilized the same method described for BM-MSC. Cells were isolated by density

gradient centrifugation (Biocoll) and subsequently seeded. Alternatively,

mesenchymal progenitors were enriched using a depletion cocktail tailored to remove

cells expressing Glycophorin A, CD3, CD14, CD19, CD66b, and CD38 (RosetteSep®

Human Mesenchymal Stem Cell Enrichment Cocktail, StemCell Technologies).

Different plating densities, varying from 5.2x104 to 1x106 cells/cm2, were utilized for

leukapheresis products and peripheral blood mononuclear cells.

Freshly isolated cells were cultured in six different media: Myelocult

(StemCell Technologies, Vancouver, BC), Mesencult (StemCell Technologies),

Mesencult added with 20ng/µl IL6 (Peprotech EC, London, UK), IMDM (Sigma-

Aldrich, St. Louis, MO, USA) added with either 2% or 25% FBS (StemCell

Technologies), and MASC expansion medium. The MASC expansion medium is

composed as follows:1 60% low glucose DMEM (Invitrogen, Carlsbad, CA, USA),

40% MCDB-201, 1mg/mL linoleic acid-BSA, 10-9M dexamethasone, 10-4M ascorbic

acid-2 phosphate, 1X insulin-transferrin-sodium selenite (all from Sigma-Aldrich),

2% fetal bovine serum (Invitrogen or StemCell Technologies), 10ng/mL hPDGF-BB,

10ng/mL hEGF (both from Peprotech). As shown in Table I, cells were seeded onto

non-coated plastic dishes, 0.001% fibronectin-coated dishes or irradiated MS5 feeder

cells.

The two-step method involved culture of cells for one week in Mesencult

(StemCell Technologies) followed by recovery of the non attached cells which were

then seeded onto 0.001% fibronectin coated dishes in MASC medium. Subcultures



4

were performed maintaining constant the seeding density (2x103 cells/cm2) as well as

the population doublings1 before further passaging.

Multilineage Differentiation

Osteoblast differentiation was obtained by plating cells at high density (2-3x104/cm2),

in DMEM supplemented with 5% FBS, 10mM β-glycerophosphate, 10-7M

dexamethasone, and 0.2 mM ascorbic acid (all from Sigma-Aldrich); medium was

changed every 3-4 days for 28 days. In a set of experiments, tetracyclines (Sigma-

Aldrich), at a final concentration of 40 µg/ml, were added to the differentiation

medium. Endothelial cell differentiation was obtained by plating cells at high density

on fibronectin coated dishes in a serum-free medium containing 10ng/mL VEGF

(Peprotech).1 Hepatocytic differentiation was induced by growing cells for two weeks

at high density onto fibronectin coated dishes in a medium containing 0.5% FBS, 10

ng/ml FGF-4 and 20 ng/ml HGF (both from Peprotech). After this period, FGF-4 and

HGF were substituted for 20 ng/ml OncostatinM (2) for other 14 days (Peprotech).

Muscle cell differentiation was achieved by plating 0.5-1x104/cm2 cells in

expansion medium containing 5% FBS (Sigma-Aldrich), 10 ng/mL bFGF, 10 ng/mL

VEGF, and 10 ng/mL IGF-1 (all from Peprotech), but not EGF. Cells were allowed to

reach confluence and cultured for up to 4 weeks with medium exchanges every 4

days.1

For neurogenic differentiation, cells were plated in DMEM-high glucose

(Invitrogen), 10% FBS (Sigma-Aldrich). After one day, medium was replaced with

DMEM-high glucose, 10% FBS containing B27 (Invitrogen), 10 ng/ml EGF and 20

ng/ml bFGF (both from Peprotech EC). Five days later, cells were washed and

incubated with DMEM containing 5 µg/ml insulin, 200 µM indomethacin and 0.5

mM IBMX (all from Sigma-Aldrich), in the absence of FBS for 5 hours-5 days.

In all cases, cells were fixed with 4% buffered paraformaldehyde or with

methanol/acetone (1/1). The supernatants of cells cultured in hepatocyte

differentiation medium an in neural differentiation medium were collected at 21 and

11 days, respectively, and snap-frozen.



5

FACS Analysis

To perform FACS analysis, a single cell suspension was obtained from cells at P3

grown in expansion medium and trypsinized (Trypsin-EDTA solution; Sigma-

Aldrich) or from mechanically disaggregated mouse spleen. Cells were stained with

primary antibodies against CD117, CD90, CD49a, CD45, CD38, CD34, CD13,

CD14, CD73, CD44, CD29, CD33, HLA-DR (BD-Biosciences, Two Oak Park,

Bedford, USA), CD133 (Miltenyi Biotec GmbH, Bergisch Gladbach, Germany),

HLA-A,B,C (Novocastra, Newcastle upon Tyne, UK), KDR, CD105 (Serotec,

Oxford, UK). Isotype matched antibodies were used as a negative control. The

analysis was performed by CyAn (Beckman Coulter, Harbor Boulevard, Fullerton,

CA, USA).

TRAP Assay

Telomerase activity was detected utilizing the TRAPeze kit (Chemicon International,

USA), following the manufacturer’s instructions. We performed a radiolabeling

detection method utilizing 32P labeled samples. Impressed radiographic films were

digitalized utilizing a setup constituted by a film-scanner (Agfa SnapScan 1236, Agfa-

Gevaert, Mortsel, Belgium) and a G3 Macintosh computer (Apple, Cuppertino,

California). Telomerase activity was evaluated on digital pictures by ImageJ software

and measured as total product generated (TPG) units by employing the formula:

TPG (units)=[(x-x0)/c x 100] / [(r-r0)/cR]

in which x = signal of the region of the gel lane corresponding to the TRAP product

ladder bands from non-inactivated samples, x0 = signal of the region of the gel lane

corresponding to the TRAP product ladder bands from inactivated samples, r = signal

of the region of the gel lane corresponding to the TRAP product ladder bands from

TSR8 quantitation control, r0 = signal of the region of the gel lane corresponding to

the TRAP product ladder bands from 1 X CHAPS lysis buffer only control, c = signal

from the internal standard (S-IC) in non-inactivated samples, and cR= signal from the

internal standard (S-IC) in TSR8 quantitation control. Each unit of TPG corresponds

to the number of TS primers (in 1 x 10-3 amole or 600 molecules) extended with at

least 4 telomeric repeats by the telomerase enzyme present in the extract.



6

Immunofluorescence and Confocal Microscopy

Primary antibodies and protocols utilized in cell and tissue immunostaining are

indicated in Table II. AlexaFluor Dyes-labeled secondary antibodies (Molecular

Probes/Invitrogen) were employed. Nuclei were stained by DAPI (Vector

Laboratories, Inc, Burlingame, CA, USA. Image acquisition was carried out by a

Confocal Laser Microscope (Leica TCS-SP2, Leica Microsystems, Wetzlar,

Germany), utilizing a 63X oil immersion objective (numerical aperture: 1.40) or a

40X oil immersion objective (numerical aperture: 1.25).

Epifluorescence and phase contrast images were obtained with a live cell

imaging dedicated system consisting of a Leica DMI 6000B microscope connected to

a Leica DFC350FX camera (Leica Microsystems); 10X (numerical aperture: 0.25),

40X oil immersion (numerical aperture: 1.25) and 63X oil immersion (numerical

aperture: 1.40) objectives were employed. Bright field images were captured utilizing

an Olympus AX70 microscope connected to an Olympus DP50 camera (Olympus,

Tokyo, Japan); 10X (numerical aperture: 0.40), 20X (numerical aperture: 0.70) and

40x (numerical aperture: 0.95) objectives were used. Adobe Photoshop software was

utilized to compose and overlay the images, and adjust contrast (Adobe, USA).

Functional Assays

For neuroectodermic differentiation, extracellular L-glutamic acid was measured

using the Amplex Red Glutamic Acid/Glutamate Oxidase Assay kit (Molecular

Probes/Invitrogen). For mesodermic differentiation, Von Kossa staining was

performed. After fixation in 4% paraformaldehyde for 20 minutes, cells were exposed

to 2% silver nitrate (Sigma-Aldrich) in a clear glass coplin jar placed directly in front

of a 60-W lamp for 1 hour. Slides were rinsed in distilled water, fixed with 2.5%

sodium thiosulphate (Sigma-Aldrich) for 5 minutes and washed. Cells were

counterstained with Nuclear Fast Red (Sigma-Aldrich) for 1 minute and rinsed in

water.

For osteogenic differentiation, tetracycline incorporation was assessed. Fixed

cells were mounted on slides utilizing Vectashield mixed with 1µM DAPI (Vector

Laboratories). Leica TCS-SP2 350nm laser was used to excite tetracyclines and



7

DAPI. Tetracycline emission fluorescence was collected in the 500-600 nm interval,

whereas DAPI emission fluorescence in the 400-500 nm interval. The Leica Dye-

Finder algorithm was utilized to create a compensation matrix able to discriminate

and resolve DAPI and tetracyclin emission spectra overlapping. Osteogenic

differentiation was subsequently confirmed by immunolabeling for osteopontin (Santa

Cruz Biotechnology, Inc., Santa Cruz, CA, USA).

For endothelial cell differentiation, the active uptake of acetylated LDL was

evaluated by incubating the cells for 4 hours in 10µg/ml DiI-Acetylated LDL

(Molecular-Probes/Invitrogen). Cells were then washed and the red fluorescence of

acetylated LDL red fluorescence was collected with an inverted fluorescence

microscope. Endothelial cell commitment was subsequently confirmed by

immunolabeling for von Willebrand Factor (Sigma-Aldrich).

To establish the functional competence of differentiated myocytes, calcium

transients were measured with the calcium indicator Fluo-4. Cells were incubated for

10 minutes at room temperature with 1mM of the ester form of the fluo-4 dye (Fluo 4-

AM, Molecular Probes/Invitrogen) previously mixed with 25% Pluronic F-127

(Molecular Probes/Invitrogen) in HBSS solution (Sigma-Aldrich). Subsequently, cells

were extensively washed and kept in HBSS at room temperature for 30 minutes to

allow de-esterification. Low ester concentration and low incubation temperature were

chosen to minimize the compartmentalization of the indicator. Images were acquired

serially with a Leica DMI 6000 B inverted microscope (LeicaMicrosystems) at a 400x

magnification (HCX PL Fluotar objective). Images were collected every 5.6 seconds

for a period of 60 seconds.

For endodermic differentiation, the production of albumin was measured at 17

days. At this time interval, edium was removed, cells were washed with HBSS and

fresh medium was added. Four days later, the supernatant was collected. Albumin

concentration was determined by ELISA assay (DRG International Inc.,

Mountainside, NJ, USA). Albumin production was expressed as pg/h/cell.

Additionally, MPCs-hepatocyte aggregates were formed by the hanging drop method

to perform the pentoxyresorufin assay (PROD assay). Briefly, 103 differentiated



8

MPCs were placed into 100µl drops.1 After 5 days, half of the drops were treated with

Phenobarbital, 1mM final concentration (Sigma-Aldrich). Four days later, aggregates

were collected and incubated for one hour in 120 µM pentoxyresorufin (Molecular-

Probes/Invitrogen) buffer at 37°C. Pentoxyresorufin is O-dealkylated by the human

hepato-specific CYP2B6 cytochrome P450, changing a non-fluorescent compound

into the fluorescent compound resorufin. Thus, the fluorescence intensity caused by

PROD metabolism estimates CYP activity. Assessment and detection of resorufin in

situ was performed collecting pictures with a confocal microscope set with fixed

acquisition parameters. Aggregates incubated in the absence of pentoxyresorufin were

utilized to establish baseline fluorescence. Human hepatocytes were utilized as

positive controls.

Periodic acid-Schiff (PAS) staining was employed to detect glycogen

accumulation. Slides were oxidized in 1% periodic acid for 5 minutes, rinsed three

times in distilled water and treated with Schiff’s reagent for 15 minutes. After

extensive washing, slides were stained with Mayer’s hematoxylin for 1 minute.

Diastase digestion was obtained with 1µg/ml preheated diastase solution (Sigma-

Aldrich) at 37 °C for 15 minutes.

PCR and RT-PCR

Total RNA was extracted from non-confluent human cells at P3 with TRIzol

(Invitrogen). After treatment with DNase I (Ambion, Foster City, CA, USA), reverse

transcription was performed with 1 µg total RNA, random hexanucleotides and M-

MLV reverse transcriptase (Invitrogen). PCR amplification was carried out in a final

volume of 50 µl. Quantities of cDNA varying from 80-150 ng were incubated with 10

mM Tris-HCl pH 9.0, 1.5 mM MgCl2, 0.2 mM dNTPs, 25 pmol of each primer and

2U Taq I polymerase (Amersham, Little Chalkont Bukinghamshire, UK). PCR

cycling conditions were as follows: 94°C for 2 minutes; 30-45 cycles at 94°C for 60

sec, 60°C for 60 and 72°C for 90 sec. PCR products were run onto 1.8% agarose gels.

Primer pairs are listed in Online Table III.

RNA Extraction, cDNA Labeling and Microarray Hybridization



9

For this assay, MPCs (n=6), heart-derived MASCs (n=9), liver-derived MASCs (n=8),

bone marrow derived MASCs (n=4), adipose tissue-derived MASCs (n=4), blood

mononuclear cells (n=5) and fibrocytes (n=5) were analyzed. RNA was extracted as

described above. After RNA quality control, microarray cDNA targets were obtained

from 10µg total RNA with the two-step indirect fluorescent labeling procedure.2 First-

strand cDNA synthesis reaction was performed using unmodified random primers in

the presence of aminoallyl-dUTP. The resulting primary amino groups were then

coupled to the NHS-ester of the fluorescent dye. Each of the 41 samples was labeled

using the Cy5 fluorophore and hybridized to the cDNA microarray slide along with

the reference RNA, labeled with the Cy3 fluorophore. Reference RNA was prepared

by mixing together RNA extracted from Placenta, CaCo-2, OVCAR-3, U-118 MG

and THP-1 (stimulated with LPS), mixed in 2:1:1:1:1 proportions, respectively.

Two different versions of the LNCIB cDNA microarray slides were used in

the large-scale transcript expression profiling studies: 18K and 28K features cDNA

microarray slides. LNCIB 18K features cDNA microarray slides were prepared by

spotting on Amersham Type 7 Star Reflective Microarray Slides the purified PCR

products of 9,465 LNCIB human cDNA clones3 and 8,160 cDNA clones selected

from the human sequence-verified I.M.A.G.E. clone collection (Research

Genetics/Invitrogen, Carlsbad, CA, USA).4 This clone set consists of ~10,620 unique

human genes. LNCIB 28K features cDNA microarray slides were prepared by

spotting the purified PCR products of 19,741 LNCIB human cDNA clones and 8,160

I.M.A.G.E. cDNA clones as above. This clone set consists of ~13,620 unique human

genes.

Purified cDNA inserts, resuspended in 50% DMSO (Sigma-Aldrich) were

printed on the slides using the SDDC-2 ChipWriter Pro microarrayer (Virtek Vision

International, Waterloo, Ontario). The 18K purified PCR products were printed in

duplicate and the 28K products as single copy. GAPDH and β-actin cDNAs were

employed as positive controls; the corresponding purified PCR products printed at

different concentrations on the slides. Plant cDNAs and printing buffer (50% DMSO)

alone were used as negative controls.



10

Competitive hybridizations were performed under coverslips at 63°C,

overnight, in the presence of 3X SSC, 0.2% SDS, 100ng/µl human Cot-1 DNA and

330ng/µl yeast tRNA, using a semi-automated hybridization station (ArrayBooster,

Advalytix, Munich, Germany). After washing, slides were scanned, at a resolution of

10µm per pixel, using an Axon GenePix 4000B microarray scanner (Axon

Instruments Molecular Devices, Union City, USA). Images were acquired for the Cy3

and Cy5 channels in a 16-bit TIFF format and analyzed using GenePix 4.0 analysis

software (Axon Instruments Molecular Devices, Union City, USA). All images were

manually flagged to exclude artifacts and bad spots (see below).

Microarray Data Analysis

Low-level analysis of microarray data was performed in the R/Bioconductor

environment using the limma microarray package (http://www.R-project.org).5

Expression data was loaded from the GenePix image analysis output files. Features

displaying median fluorescence intensity lower than the local background plus 2

standard deviations, at least in one channel, were flagged as bad spots. Fluorescence

intensities of the red and green channels were background-subtracted,6 and the

corresponding expression ratios (Cy5/Cy3) were log-transformed (base 2).

Normalization was then carried out with the print-tip lowess method, independently

for each cDNA collection within each array. Normalized ratios from replicate

expression measurements within the same array were averaged to produce the final

expression matrix (FEM) used in the subsequent analyses.

Initially, we generated FEM1, composed of 27,713 features and 41 samples

(25 MASCs derived from adult tissues, 6 MPCs, 5 mononuclear blood cell samples

and 5 peripheral blood-derived fibrocytes). Unsupervised hierarchical cluster analysis

(average-linkage method; Cluster) was applied to the filtered expression matrix 1.7

Flagged features consisting of features not present in at least 80% of samples and

features in which the maximum minus minimum log values were less than 1.5 were

filtered out (7,266 features passed these filtering criteria). The highly stringent

filtering procedure applied (removing all genes with missing values in more than 20%



11

of the columns) was dictated by the use of two different versions of the LNCIB cDNA

microarray slides (18K and 28K). By this approach, it was excluded the possibility of

bias in the clustering procedure. For both genes and samples, uncorrelated distance as

the dissimilarity measure was used. Cluster results were then visualized using Java

TreeView.8

To determine the statistical correlation between the transcriptional profiles of

the four cell types (MPCs, MASCs, fibrocytes and buffy coats), a new filtered

expression matrix (FEM2) was then generated by using the biological samples

expression profiled with the LNCIB 28K features cDNA microarray slides (25

samples: 9 hMASCs derived from adult tissues, 6 MPCs, 5 peripheral blood

mononuclear cell samples and 5 peripheral blood-derived fibrocytes). A correlation

matrix was obtained for each of the four cell types (MPCs, MASCs, Fibrocytes and

Buffy Coats). Features not present at least four times for each cell type or features in

which variance was less than lower hinge were filtered out (18.179 features passed

these filtering criteria). Median vectors specific for each cell type were then calculated

and the corresponding Pearson correlation matrix was computed.

To identify the maximum number of differentially expressed genes in the

various comparisons, we used limma package to build two linear models: 1.

mononuclear blood cell and fibrocyte samples were compared to MPCs and MASCs;

and 2. MPCs were compared to MASCs. For statistics, significance of p-value was

estimated by the FDR method (Benjamini-Hochberg; selected p for Significant Genes:

1x10-3). In the first comparison, 5,408 features (corresponding to 3,336 unique

annotated genes) were upregulated in mononuclear blood cells/fibrocytes vs.

MPCs/MASCs and 5,315 (corresponding to 3,224 unique annotated genes) were

down-regulated. The two gene lists were then subjected to functional analysis using

the DAVID/EASE web tool (LNCIB 28K 10,032 unique annotated genes as a

background list (Online Tables V and VI). For the analysis of functional annotation (AFA),

the Wilcoxon rank-sum test was employed to search for and compare functional and

biological concepts (Functional Gene Sets, FGS). The Wilcoxon rank-sum test

computes a p-value to test the hypothesis that a FGS, defined by a functional



12

annotation, tends to be more highly ranked in an ordered list. Individual genes on the

arrays were ranked by their absolute moderate t-statistics and statistical tests were

performed for the MPCs vs. MASCs comaparison. This analysis was done in two

distinct reference populations: 1. all the non-redundant genes (according to Entrez

Gene identifiers) present on the microarray platform; and 2. all the non-redundant

genes annotated in each specific functional theme (i.e., KEGG, GO, etc.). After

statistical tests were completed, control of false discovery rate (correction for multiple

hypothesis testing) was obtained by applying the Benjamini and Hochberg method as

implemented in the multitest R/Bioconductor package. Functional annotation themes

(FGS) were defined by gene set collections from the complete Molecular Signatures

Database (MSigDB). AFA results were then displayed in a heatmap in which the level

of significance arises from p-values (log 10 of the Wilcoxon rank sum test results).

In Vivo Studies

Animal suffering and number of mice used were minimized in accordance with the

Directive 86/609/EEC on the protection of animals used for experimental and other

scientific purposes. All experiments were performed using 6-week-old BALB/c mice

(n=6, Harlan Europe). Undifferentiated MPCs at P3 were harvested by trypsinization,

washed and resuspended in PBS at a final concentration of 4x103 cells/µl; 106 cells

were injected in the hind limb through a 22G needle and, 2 weeks later, mice were

sacrificed. Lungs, liver, spleen, heart, brain, bone marrow, femur, skeletal muscles

and skin were collected and fixed in formalin or snap-frozen. Formalin-fixed tissues

were then paraffin-embedded and 4 µm thick tissue sections were obtained for

immunolabeling and FISH. Snap-frozen tissue sections, 4 µm in thickness, were post-

fixed in acetone or 4% buffered paraphormaldeyde.

Immunohistochemistry

Protocols and primary antibodies are listed in Table II. To identify human cells

engrafted in murine tissue, immunohistochemistry was performed using a polymer-

based immunohistochemical detection system (EnVision™ Detection

SystemsPeroxidase/DAB, Rabbit/Mouse, Dako, Denmark) in an automated horizontal

slide-processing system (Dako Autostainer). Sensitivity and specificity of the human



13

specific antibodies and probes was determined in human and murine tissue sections,

respectively. Human tissue was used as positive control. Non-injected murine tissues

and omission of the primary antibody were used as negative controls. Immunolabeling

in frozen sections and in formalin-fixed paraffin embedded sections consisted of

incubation with the primary antibodies overnight at 4°C and with secondary

antibodies for 1 hour at 37°C.

References

1. Beltrami AP, Cesselli D, Bergamin N, Marcon P, Rigo S, Puppato E,

D'Aurizio F, Verardo R, Piazza S, Pignatelli A, Poz A, Baccarani U, Damiani

D, Fanin R, Mariuzzi L, Finato N, Masolini P, Burelli S, Belluzzi O,

Schneider C, Beltrami CA. Multipotent cells can be generated in vitro from

several adult human organs (heart, liver, and bone marrow). Blood.

2007;110:3438-3446.

2. Xiang CC, Kozhich OA, Chen M, Inman JM, Phan QN, Chen Y, Brownstein

MJ. Amine-modified random primers to label probes for DNA microarrays.

Nat Biotechnol. 2002;20:738-742.

3. Dalla E, Mignone F, Verardo R, Marchionni L, Marzinotto S, Lazarevic D,

Reid JF, Marzio R, Klaric E, Licastro D, Marcuzzi G, Gambetta R, Pierotti

MA, Pesole G, Schneider C. Discovery of 342 putative new genes from the

analysis of 5'-end-sequenced full-length-enriched cDNA human transcripts.

Genomics. 2005;85:739-751.

4. Lennon G, Auffray C, Polymeropoulos M, Soares MB. The I.M.A.G.E.

Consortium: an integrated molecular analysis of genomes and their expression.

Genomics. 1996;33:151-152.

5. Smyth GK. Limma: linear models for microarray data. In Bioinformatics and

Computational Biology Solutions using R and Bioconductor. R.C. Gentleman,

V.; Dudoit, S.; Irizarry, R.Huber, W., editor Springer, New York, 2005.

6. Edwards D. Non-linear normalization and background correction in one-

channel cDNA microarray studies. Bioinformatics. 2003;19:825-833.



14

7. Eisen MB, Spellman PT, Brown PO, Botstein D. Cluster analysis and display

of genome-wide expression patterns. Proc Natl Acad Sci USA.

1998;95:14863-14868.

8. Saldanha AJ. Java Treeview--extensible visualization of microarray data.

Bioinformatics. 2004;20:3246-3248.




michelle

Text Box

Online


by on May 21, 2011

circres.ahajournals.orgD

ownloaded from


Multipotent Progenitor Cells Are Present in Human Peripheral Blood

Documents

Transcript of Multipotent Progenitor Cells Are Present in Human Peripheral Blood