hyaluronanexpressedbythehematopoietic ... · bovine serum, i-inositol, folic acid,...

Hyaluronan Expressed by the HematopoieticMicroenvironment Is Required for Bone MarrowHematopoiesis*□S

Received for publication, April 27, 2012 Published, JBC Papers in Press, May 31, 2012, DOI 10.1074/jbc.M112.376699

Valentina Goncharova‡, Naira Serobyan§, Shinji Iizuka¶, Ingrid Schraufstatter‡, Audrey de Ridder‡�,Tatiana Povaliy‡, Valentina Wacker‡, Naoki Itano**, Koji Kimata‡‡, Irina A. Orlovskaja§§, Yu Yamaguchi¶,and Sophia Khaldoyanidi‡1

From the ‡Torrey Pines Institute for Molecular Studies, San Diego, California 92121, the §La Jolla Institute for Molecular Medicine,San Diego, California 92121, the ¶Sanford-Burnham Medical Research Institute, La Jolla, California 92037, the �University ofNamur, Rue de Bruxelles 61, 5000 Namur, Belgium, the **Kyoto Sangyo University, Kyoto 603-8555, Japan, the ‡‡Aichi MedicalUniversity, Nagakute 480-1095, Japan, and the §§Institute of Clinical Immunology, Novosibirsk 630099, Russia

Background: Hyaluronan (HA) contributes to the extracellular matrix in bone marrow.Results: HA expressed by the hematopoietic microenvironment supports hematopoiesis and is involved in hematopoieticstem/progenitor cell migration by regulating the production of soluble factors.Conclusion: Endogenous HA is an important regulatory element of the hematopoietic microenvironment.Significance: Understanding the biology of HA may help to develop strategies for improving the quality of the stem cellmicroenvironment.

The contribution of hyaluronan (HA) to the regulatory net-work of the hematopoietic microenvironment was studiedusing knock-out mice of three hyaluronan synthase genes(Has1, Has2, and Has3). The number of hematopoietic pro-genitors was decreased in bone marrow and increased inextramedullary sites of Prx1-Cre;Has2flox/flox;Has1�/�;Has3�/� triple knock-out (tKO) mice as compared with wildtype (WT) and Has1�/�;Has3�/� double knock-out (dKO)mice. In line with this observation, decreased hematopoieticactivity was observed in long term bone marrow cultures(LTBMC) from tKO mice, whereas the formation of theadherent layer and generation of hematopoietic cells in WTand dKO cultures was not different. 4-Methylumbelliferone(4MU) was used to pharmacologically inhibit the productionof HA in LTBMC. Treatment with 4MU inhibited HA synthe-sis, decreased expression of HAS2 and HAS3, and eliminatedhematopoiesis in LTBMC, and this effect was alleviated bythe addition of exogenous HA. Exogenous HA also aug-mented the cell motility in LTBMC, which correlated with theHA-stimulated production of chemokines and growth factors.Conditionedmedia fromHA-induced LTBMC enhanced the che-motaxis of hematopoietic stem/progenitor cells (HSPC) inresponse to SDF-1. Exposure of endothelial cells to 4MUdecreased their ability to supportHSPC rolling and adhesion. Inaddition, migration of transplanted HSPC into the marrow of4MU-pretreatedmice was lower than in untreatedmice. Collec-tively, the results suggest that HA depletion reduces the ability ofthemicroenvironment tosupportHSPC,andconfirmarole forHA

as a necessary regulatory element in the structure of the hemato-poieticmicroenvironment.

The bonemarrowmicroenvironment regulates several func-tions of hematopoietic stem/progenitor cells (HSPC),2 includ-ing their recruitment into the bone marrow following trans-plantation. One mechanism by which the microenvironmentregulates HSPC homing is by influencing the production ofchemokines that mediate HSPC migration. The chemokineSDF-1 and its cognate receptor CXCR4 are known to play animportant role in mediating homing of both long-term recon-stituting HSCs (1) and committed hematopoietic progenitorcells (2), although a subset of HSCs has been identified thathomes in a SDF-1/CXCR4-independent manner (3). Otherchemokines, including CCL19 and CCL21, are predominantlychemotactic for committed hematopoietic progenitor cells (4).In addition to chemokines, growth factors produced in thebonemarrow contribute indirectly to the recruitment of HSPCby regulating the expression of CXCR4 on HSPCs (5), or byinducing the release of CCL12 and other chemokines fromintracellular storage (6).Effective recruitment of circulating HSPCs into bone mar-

row depends on the balance of chemokines and growth factorspresent, which is itself influenced by a variety of factors includ-ing the bone marrow cellular composition, stress factors,inflammation, circadian rhythms, disease development, and

* This work was supported, in whole or in part, by National Institutes of HealthGrants R41CA126004, R43AI082759, and R21NS062428 and University ofCalifornia TRDRP Grant 16RT-0134 (to S. K. K.).

□S This article contains supplemental Figs. S1–S7.1 To whom correspondence should be addressed: 3550 General Atomics

Ct., San Diego, CA 92121. Tel.: 858-597-3879; Fax: 858-597-3804; E-mail:[email protected].

2 The abbreviations used are: HSPC, hematopoietic stem/progenitor cell;ECM, extracellular matrix; CRA, competitive reconstitution assay; 4MU,4-methylumbelliferone; LTBMC, long term bone marrow culture; CS, chon-droitin sulfate; CM, conditioned media; bHABP, biotinylated HA-bindingprotein; CCM, control untreated cultures; HMW, high molecular weight;LMW, low molecular weight; MSC, mesenchymal stem cells; BMC, bonemarrow cells; HSCs, hematopoietic stem cells; CFC, colony forming cells;CFU, colony forming unit; HMW, high molecular weight.

THE JOURNAL OF BIOLOGICAL CHEMISTRY VOL. 287, NO. 30, pp. 25419 –25433, July 20, 2012© 2012 by The American Society for Biochemistry and Molecular Biology, Inc. Published in the U.S.A.

JULY 20, 2012 • VOLUME 287 • NUMBER 30 JOURNAL OF BIOLOGICAL CHEMISTRY 25419

by guest on October 12, 2019

http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/cgi/content/full/M112.376699/DC1

http://www.jbc.org/

therapeutic interventions (7–9). Thus, the quality of themicroenvironment has a profound impact on homing of HSPCto the bone marrow and the subsequent efficiency of stem cell-based tissue regeneration. This makes the bone marrowmicroenvironment an important therapeutic target, the healthof which might be improved by a “pretransplant conditioning”prior to stem cell transfer to promote tissue regeneration.Extracellular matrix (ECM) molecules produced by cells

within the bone marrow contribute to the highly complexstructure of the regulatory microenvironment (10, 11).Although some ECM components, such as collagens, fibronec-tin, laminin, and hemonectin, are known to participate in thebone marrow regulatory network (10, 11), the role of otherECM components, including hyaluronan (HA), remainsunclear. HA is amember of the glycosaminoglycan polysaccha-ride family and is amajor component of the bonemarrow ECM(12).HA is a negatively charged polymer consisting of repeatingdisaccharides of N-acetyl-D-glucosamine (GlcNAc) and glucu-ronic acid (GlcA) residues, and is involved in regulating manycellular functions, including proliferation (13, 14), migration(15), cytokine production (16, 17), and expression of adhesionmolecules (18). The effects of HA on cell functions are medi-ated by the HA receptors including CD44, RHAMM, andHARE, all of which have been shown to induce intracellularsignaling and regulate gene transcription (19). Although theinvolvement of HA in normal cell and tumor cell biology isgenerally appreciated, there is a large gap in our understandingof how HA contributes to the regulation of the bone marrowmicroenvironment.We have previously demonstrated that HA can directly stim-

ulate cytokine production from bone marrow-resident cells,and thus is not a passive structural element of the bonemarrowECM, but a necessary and specific signal-inducingmolecule forhematopoiesis (16). In the present study we have extendedthese observations to show that HA stimulates cells of thehematopoietic microenvironment in an autocrine fashion toproduce a range of soluble factors, including cytokines andchemokines that mediate the motility of hematopoietic cells.Moreover, we demonstrate that enforced reduction of bonemarrow HA levels inhibits the ability of the microenvironmentto support hematopoiesis and to recruit circulatingHSPCs intobone marrow in vivo. Our findings support an important regu-latory role for endogenous HA as an integral part of the hema-topoietic microenvironment.

EXPERIMENTAL PROCEDURES

Mice—All animal experiments were conducted according toNIH guidelines and in agreement with the Torrey Pines Insti-tute for Molecular Studies (TPIMS) policy and Sanford-Burn-hamMedical Research Institute (SBMRI) policy on animal use,and were approved by the Institutional Animal Care and UseCommittees (IACUC). C57BL/6 and BALB/c mice were fromthe Jackson Laboratories. The generation of the loxP-modifiedHas2flox allele and Prx1-Cre;Has2flox/flox conditional knock-outmice has been reported previously (20). Constitutive Has1knock-out mice were created by replacing a part of exon 5encoding the catalytic site of the enzyme (21). ConstitutiveHas3 knock-out mice (22) were kindly provided by Dr. John

McDonald (University of Utah). Has1�/�;Has3�/� doubleconstitutive knock-out mice (dKO) and Prx1-Cre;Has2flox/flox;Has1�/�;Has3�/� triple knock-out mice were bred from theselines in 100% C57BL/6 background. Levels of endogenous HAwere reduced by irradiation (11 gray) ofC57BL/6mice (8-week-old females, 10mice per group) followed by immediate intrave-nous injection of 100 �l of 3 mM 4-methylumbelliferone (4MU,an HA synthesis inhibitor (23); Sigma). Animals received threeadditional intravenous injections of 100 �l of 3 mM 4MU with6-h intervals. Control mice were subjected to the same treat-ment except they received injections of an equal volume ofvehicle (phosphate-buffered saline, PBS). To evaluate the effi-ciency of HSC homing, lineage-negative cells (Lin�, 2 � 106cells/mouse) were purified frombonemarrowofC57BL/6mice(8-week-old females) using MACS (Miltenyi Biotec, Auburn,CA) and were injected intravenously into irradiated and PBS-or 4MU-treated mice. Twenty-four hours later, mice were sac-rificed, bone marrow was collected from femurs and tibias, andthe number of HSCs recruited into the bonemarrowwas deter-mined by competitive reconstitution assays (CRA).Competitive Reconstitution Assay—C57BL/6JolaHsd Pep3a

(CD45.2) and C57Bl/6.SJL.Ptprca Pep3b/BoyJ (CD45.1) mice(referred to here as B6/CD45.2 and B6/CD45.1mice, respectively,Jackson Laboratories) were used for this assay. B6/CD45.2 micewereusedas “donors,” and sex- andage-matchedB6/CD45.1micewere used as “competitors.” Two-month-old recipient B6/CD45.1micewere lethally irradiated (11 gray), then reconstitutedwith bone marrow cells purified from B6/CD45.2 donors (2 �105/mouse) and B6/CD45.1 competitors (106/mouse). Eightweeks after cell transplantation, peripheral bloodwas harvestedfrom the recipientmice, andmaturemononuclear cells derivedfrom donor or competitor HSCs were detected by FACS anal-ysis using antibodies specific for CD45.1 (clone A20, eBiosci-ence, San Diego, CA) or CD45.2 (clone 104, Invitrogen). Thepercentage of donor-derivedCD45.2-positive cells generated inthe recipients was then calculated.Cell Culture and Reagents—Mouse stromal fibroblast-like

cell lines M10B4 and MS-5 were cultured in DMEM supple-mented with 10% FCS, and the S-17 cell line was cultured inRPMI (Invitrogen) with 5% FCS. Themouse progenitor cell lineFDCP-mix was cultured in Iscove’s medium supplementedwith 20% horse serum (StemCells Technologies, Vancouver,Canada) and 10%WEHI-3B-conditioned medium. The mousebone marrow-derived endothelial cell line STR-12 and lungmicrovasculature-derived LEISVO endothelial cell line are agift from Dr. Kobayashi, Hokkaido University, School of Med-icine, Sapporo, Japan, and were cultured in RPMI supple-mented with 10% FCS (24). Human mesenchymal stem cells(MSC) were cultured in �-minimal essential medium (Invitro-gen) supplemented with 16.5% FCS and 2 mM L-glutamine(Sigma). Mouse MSCs were cultured in RPMI supplementedwith 10% FBS, 50 �g/ml of penicillin, and 50 �g/ml ofstreptomycin.Long Term BoneMarrow Culture (LTBMC) and CFUAssays—

LTBMCwere established according to standard protocols usingbone marrow from C57BL/6 or BALB/c mice. Briefly, bonemarrow cells (106 cells/ml) were cultured for 8 weeks in �-min-imal essential medium supplemented with horse serum, fetal

Microenvironment-associated Hyaluronan

25420 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 287 • NUMBER 30 • JULY 20, 2012


http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

bovine serum, i-inositol, folic acid, 2-mercaptoethanol, L-glu-tamine, and 10�6 M hydrocortisone (MyeloCult,� StemCellTechnologies) in 6-well tissue culture plates (VWR, Brisbane,CA) at 37 °C in a humidified atmosphere containing 5% CO2.During each weekly feeding, nonadherent cells were collectedfrom the culturemedium, counted, and used inCFU assays. Forthe CFU assay, 104 cells/ml of nonadherent cells were culturedin methylcellulose media supplemented with cytokines (Meth-oCult, StemCell Technologies). Colonieswere counted under aninverted microscope after 7 and 14 days of culture. Where indi-cated, bone marrow cells were collected from control wild type(WT)mice, dKO (Has1�/�;Has3�/�) mice, and tKOmice (Prx1-Cre;Has2flox/flox;Has1�/�;Has3�/�). Endogenous synthesis of HAin LTBMCwas inhibited by the addition of 300�M 4MU, whichwas added at day 1, then every week thereafter. Where indi-cated, the cultures were treated with 100 �g/ml of endotoxin-free HA (600–1,500 kDa, purified from umbilical cord (�5%protein and chondroitin sulfate (CS) total impurities, Sigma),recombinant HA (200 and 15 kDa, Lifecore Biomedical,Chaska, MN), 10 units of bovine hyaluronidase (testicularhydrolase, Sigma), 100 �g/ml of CS (Sigma) or CD44 specificantibody (clone KM201, Abcam). To prepare conditionedmedia (CM), the adherent layers of LTBMC were cultured inserum-free DMEM supplemented with 4.5 g/liter of glucoseand sodium pyruvate for 12 h, followed by stimulation with 100�g/ml of HA. CMwas collected 24 h later and used for cytokineand chemokine, ELISA, and Transwell assays.Time Lapse Video Recording—Microscopy and video record-

ing of the adherent cell layer of LTBMC was performed on aLeica DM IRBEmicroscope using a�20 objective. Images wereobtained with a Hamamatsu digital camera (Hamamatsu Pho-tonics, Hamamatsu City, Japan) and analyzed with ImprovisionOpenLab 3.0 software (Improvision, Boston, MA), whichallows the determination of motility of individual cells. Dis-tance and time of travel of 10 individual cells from each culturewas tracked, calculated, and normalized to mm/s.Flow Cytometry—Detection of CD45.1 and CD45.2 on

murine peripheral blood mononuclear cells was determinedwith a standard FACS protocol using anti-mouseCD45.1-FITC(clone A20, eBioscience, San Diego, CA) and anti-mouseCD45.2-phycoerythrin (clone 104, Invitrogen) antibodies. Phy-coerythrin- and FITC-conjugated isotype control antibodieswere from BD Pharmingen and Invitrogen, respectively. Thepresence of HA on the STR-12 cell surface was detected byincubating cells with 2 �g/ml of biotinylated HA-binding pro-tein (bHABP, Sigma) for 2 h. Staining was visualized by incu-bating the STR-12 cells with 10 �g/ml of Alexa Fluor 546-con-jugated streptavidin (Invitrogen) for 15 min. As a negativecontrol, the STR-12 cells were pretreated with 200 units/ml ofhyaluronidase for 2 h prior to staining with bHABP. Fluores-cence intensity was analyzed on a FACScalibur (Becton Dick-inson, San Jose, CA) according to standard procedures.In VitroMicrocapillary Flow Assay—Rolling and adhesion of

hematopoietic cells under physiological shear stress wasassessed in vitro using microcapillary tubes coated with amonolayer of STR-12 endothelial cells. Briefly, glass capillaries(Fisher Scientific, Pittsburgh, PA) were coated with 2%3-aminopropyltriethoxysilane in acetone (Sigma), washed

twice with PBS, dried, and sterilized. Thereafter, the capillarieswere coatedwith 5mg/ml of gelatin typeB (Sigma) for 30min at37 °C. STR-12 endothelial cells were grown in the glass capil-laries until 100% confluent.Where indicated, STR-12 cells weregrown in the presence of 300 �M 4MU. Defined levels of flow(wall shear stress) were applied to the capillaries by perfusingwarm media (RPMI containing 0.75 mM Ca2� and Mg2� and0.2%HSA) through a constant infusion syringe pump (HarvardApparatus, Holliston, MA). The capillaries were then perfusedwith 10 ml of FDCP-mix (1 � 105 cells/ml) at various levels ofshear stress. At least five STR-12-coated capillaries were run ineach experimental group. The interactions of the injectedFDCP-mix cells with the endothelial layer were observed in thecentral sector of each capillary using an inverted phase-contrastmicroscope, and the images were recorded. Rolling FDCP-mixcells demonstrated multiple discrete interruptions and flowedslowly, whereas adherent cells remained stationary at a givenpoint for extended periods of time (�30 s). All results areexpressed as the number of rolling or adherent cells/field, rep-resenting the mean � S.D. from 5 capillaries.Cytokine, Chemokine, and Growth Factor Assays—The pro-

duction of a panel of cytokines, chemokines, and growth factorsin murine BM cultures was quantified using the RayBio MouseCytokine Antibody Array III&3.1 and Quansys Biosciencesplatform, according to the manufacturer’s recommendations.ConfocalMicroscopy—STR-12 cells were cultured on poly-D-

lysine-coated glass coverslips until 50% confluent. The cellswere fixed with 4% paraformaldehyde (Electron MicroscopySciences, Hatfield, PA) in PBS (Invitrogen) for 30 min. Afterwashing and blockingwith 2%FCS for 2 h at room temperature,the cells were treated with bHABP (Sigma) for 2 h at 4 °C. Afterwashing, the cells were incubated with FITC-conjugated avidin(BD Pharmingen) in PBS containing 2% FCS for 1 h at roomtemperature. Negative controls were treated identically exceptbHABPwas omitted.Afterwashing and staining the nuclei withDAPI (4�-6-diamidino-2-phenylindole) (Sigma) for 10 min, thecells were washed and covered with a drop of AntiFade (Molec-ular Probes, Invitrogen). Images were taken on an OlympusFluoview FV1000 confocal microscope.Transwell Chemotaxis Assay—A single cell suspension of

bone marrow was loaded into the upper wells of Matrigel-coated Transwells (Corning, NY, 5-�m pore size, 106 cells/in-sert). The lower wells contained media alone, or media supple-mented with 50 ng/ml of SDF-1, control CM, or CM fromHA-stimulated LTBMC. The assembled wells were incubatedfor 4 h in a 37 °C incubator, then the upper compartments wereremoved, and the cells present in the lower compartments werecollected, enumerated, and subjected to CFU assays.Immunoblotting—Cell monolayers were lysed with modified

RIPAbuffer (50mMTris-HCl, pH7.4, 10% glycerol, 1%NonidetP-40, 150mMNaCl, 5mMMgCl2, 2mMEDTA, 0.2mMPMSF, 2�g/ml of leupeptin, 2 �g/ml of aprotinin, 2 mM sodium pyro-phosphate, 2mM sodiumvanadate, and 10mM sodium fluoride)and clarified by centrifugation. The cell lysates were resolved bySDS-PAGE and transferred to nitrocellulose membranes.Membranes were blocked with 4% dry milk in TBS-Tween andexposed to goat polyclonal HAS-1, HAS-2, or HAS-3 specificantibodies (Santa Cruz Biotechnology, Santa Cruz, CA). Anti-




http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

body binding was detected using horseradish peroxidase(HRP)-conjugated donkey anti-goat secondary antibody (SantaCruz Biotechnology) and revealed by enhanced chemilumines-cence (ECL Plus, Amersham Biosciences Bioscience/GEHealthcare, Piscataway, NJ).Detection of HAConcentrations—CMand cell lysate samples

collected from LTBMC and STR-12 cultures were tested forHA concentrations by an ELISA-like assay (Echelon, Salt LakeCity, UT) according to the manufacturer’s instructions.Statistical Analysis—Statistical analyses were carried out

using Student’s t test.

RESULTS

HAS Activity in BoneMarrow Cells of Mesenchymal Origin IsRequired for Hematopoiesis in Vitro—To investigate whetherHA synthesis in the bone marrow hematopoietic microenvi-ronment is important for supporting hematopoiesis in vitro, weinitiated LTBMC frombonemarrow cells isolated from controlwild type (WT) mice, double Has1/3 knock-out (KO) mice(Has1�/�;Has3�/�) and triple Has1/2/3 KO mice (Prx1-Cre;Has2flox/flox;Has1�/�Has3�/�). We found that bone marrowcells isolated from control and dHAS1/3 KOmice were able toform an adherent layer that supported formation of “cobble-stone” areas and generation of hematopoietic cells (Fig. 1A). Incontrast, bonemarrow cells isolated from triple tHAS1/2/3 KOmice formed adherent layers that failed to support formation ofhematopoietic foci (Fig. 1A). Production of hematopoietic cells(Fig. 1B) and hematopoietic progenitors (Fig. 1C) in LTBMC

was compared between the groups. Although generation ofhematopoietic cells and their progenitors in dHAS1/3 KO cul-tures was not different from those in control cultures, littlehematopoietic activity was detected in tHAS1/2/3 KO cultures.The expression of HAS1, HAS2, and HAS3 in bone marrow-

derived cells and in the adherent layer of LTBMCwas examinedby Western blot using HAS-specific antibodies (Fig. 2A). Wefound that HAS2 is expressed in bone marrow-derived endo-thelial cells, MSC, and its derivatives stromal fibroblast-likecells (S17, M10B4, and MS5) to various degrees. In addition,HAS2 was detected in the adherent layer of LTBMC (Fig. 2B).In contrast, levels of HAS1 and HAS3 (not shown) were similarin all tested cells.HAS expression correlated with HA production in LTBMC

that had been cultured in specialized LTBMC media (Myelo-Cult) that supports hematopoietic activity in vitro (Fig. 2C).Interestingly, the production of HA in LTBMC was signifi-cantly decreased when instead of hematopoiesis-supportiveMyeloCult media cultures were grown in DMEM, which doesnot support hematopoiesis in vitro. To further investigatewhether the production of endogenous HA is required forhematopoiesis, LTBMC cultured in MyeloCult media weretreated with 4MU, an inhibitor of HA synthesis (23) (Fig. 2D).4MU Inhibits Hematopoiesis in LTBMC—4MU reduces the

UDP-uronic acid availability in cells, and subsequentlydecreases synthesis of HA (25, 26). Because 4MU was reportedas toxic to cells at a concentration of 1 mM (27), we first tested

FIGURE 1. Hematopoietic activity in dKO HAS1/3 and tKO HAS1/2/3/bone marrow cultures. A, LTBMC were started using bone marrow cells isolated fromcontrol wild type (WT) mice, dHAS1/3 KO (Has1�/�;Has3�/�) mice, and tHAS1/2/3 KO (Prx1-Cre;Has2flox/flox;Has1�/�;Has3�/�) mice. Images of the adherentlayers were taken on week 3 of culture (�20 objective). Arrows in WT and dKO cultures indicate areas with formed hematopoietic foci (cobblestone areas). Nocobblestone areas were observed in tKO cultures. Representative images out of 12 similar cultures per group are shown. Nonadherent cells generated incultures were collected in week 3 (B) and the number of committed progenitors investigated using the cfu assay (C). Results are mean concentrations � S.D.of triplicate cultures, from one of three similar experiments. *, p � 0.05.




http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

the effect of different concentrations of 4MU on survival ofnonadherent and adherent cells in LTBMC. The addition of4MU to established steady state LTBMC for 24 h demonstrateda significant increase (p � 0.05) in the number of dead cells, asmeasured by trypan blue exclusion, which was detected at 700�M 4MU and higher concentrations, whereas no effect wasdetected at 100 and 300�M4MU (supplemental Fig. S1A). Sim-ilar results were obtained when the number of apoptotic cellswas evaluated by measuring annexin V binding to the cell sur-face using FACS (supplemental Fig. S1B). Interestingly, expo-sure of established steady state LTBMC to 4MU for 7 days (onefeeding period of LTBMC) had no or only a mild effect on sur-vival of cells at low (100 and 300 �M) and intermediate (500 �M

and 700 �M) concentrations, but dramatically (9–10-fold, p �0.05) increased the percentage of dead cells at 1 mM 4MU (sup-plemental Fig. S1, A and B). However, exposure of the estab-lished steady state LTBMC to all tested concentrations of 4MUfor 7 days resulted in a significant (p � 0.05) decrease in thenumber of hematopoietic cells and progenitors generated intreated cultures as compared with nontreated control (supple-mental Fig. S2, A and B), suggesting that the process of hema-topoiesis is affected by 4MU. In contrast, the addition of 4MUdirectly to theCFU assays did not influence colony formation atconcentrations of 100 and 300�M,whereas at concentrations of500, 700, and 1000 �M, both the number and size of colonieswere significantly (p� 0.05) decreased (supplemental Fig. S2C).Together, these findings suggest that 300�M is the highest non-toxic concentration of 4MU that can be used to study the effectof pharmacological inhibition of the HA synthesis on hemato-poiesis in LTBMC.Treatment of LTBMC with 300 �M 4MU did not change the

expression of HAS1 (not shown), whereas the expression ofHAS2 and HAS3 proteins was decreased in the adherent layer

of LTBMC (Fig. 2B), which correlatedwith a 20-fold decrease inlevels of HA secreted into the culture supernatant (Fig. 2D).This magnitude of reduction had a striking physiological effecton LTBMC, as the addition of 300 �M 4MU on the day of initi-ation of cultures completely abrogated hematopoiesis as mea-sured by the number of nonadherent cells produced in 4MU-treated LTBMC compared with control during the wholeperiod of culture (Fig. 3A). To examine the influence of 4MUonthe number of committed progenitors generated in cultureduring treatment, nonadherent cells were harvested from con-trol and 4MU-treated LTBMC at each weekly feeding, washed,and cultured in methylcellulose in the presence of hematopoi-etic growth factors. Committed progenitor cells were detecta-ble in the 4MU-treated cultures during the first week of culture,but there was a sharp decrease in the number of colony formingcells (CFC) over the subsequent weeks. As expected, the num-ber of CFC present in control, untreated LTBMC increasedsteadily over the first 5 weeks of culture, forming a plateauduring weeks 5 and 6 (Fig. 3B).Consistent with the negative effect of 4MU treatment onHA

production and generation of CFC in LTBMC, reconstitutingHA levels by addition of exogenous high molecular weight(HMW) (�1,500 kDa) HA to LTBMC significantly improvedhematopoiesis, as demonstrated by the increased number ofboth nonadherent mature cells (Fig. 3C) and their committedprogenitors (Fig. 3D) measured at week 3 of culture. Interest-ingly, the addition of CS to the 4MU-treated LTBMC alsoimproved hematopoietic activity in these cultures by increasingthe levels of nonadherent cells to those observed in controls(Fig. 3,C andD), although to a lesser degree than seen withHA.As expected, the effect of exogenous HA in 4MU-treatedLTBMC was blocked by the CD44-specific HA-binding block-

FIGURE 2. Expression of HAS2, HAS3, and HA in bone marrow cells. A, the expression of HAS2 in cells was examined by Western blot analysis. The presenceof HAS2 proteins in endothelial cell lines (STR-12, lane 1, and LEISVO, lane 2), bone marrow-derived stromal cell lines (S-17, lane 3, M10B4, lane 4, and MS-5, lane5), and bone marrow-derived mesenchymal stem cells (MSC; mouse: lane 6, and human, lane 7) was detected, followed by re-probing with �-actin antibody.One representative blot is shown of five similar experiments. B, the expression of HAS2 and HAS3 proteins in control and 4MU-treated adherent layers of LTBMCwas detected by Western blot. Protein products for HAS2, HAS3, and �-actin are shown on representative images of three similar experiments. C, theconcentration of HA in control and conditioned media was measured by ELISA. For control samples MyeloCult and DMEM were used. To prepare conditionedmedia, LTBMC were cultured in MyeloCult for 4 weeks to steady state and then cultured either with MyeloCult or DMEM for 72 h (“Experimental Procedures”).D, the production of HA polymers in control (vehicle (PBS)-treated) and 4MU-treated (300 �M) LTBMC was measured by an ELISA-like assay. Results are meanconcentrations � S.D. of triplicate cultures, from one of two similar experiments. *, p � 0.05.




http://ww

w.jbc.org/

Dow

nloaded from








http://www.jbc.org/

ing antibody, which is in accordance with previously reportedfindings (16, 28–30) (supplemental Fig. S2).We next examined LTBMC microscopically to directly

visualize the effect of 4MU-induced inhibition of endoge-nous HA synthesis on hematopoiesis. Examination of thecontrol LTBMC revealed the formation of foci of hemato-poiesis, the so called “cobblestone areas,” within the adher-ent layer of control untreated LTBMC (Fig. 3E). By contrast,this process was not efficient in the 4MU-treated cultures,where the adherent layer was formed but no discrete focicould be seen (Fig. 3F).HA-induced Factors Stimulate Cell Motility in Vitro—Next,

time lapse photography of the adherent layerswas performed todetermine whether HA deprivation affected the motility ofhematopoietic cells in LTBMC. Because 4MU prevents only denovo synthesis of HA, but does not degrade existing HA, thelevel of HA in the established steady state cultures was reducedusing hyaluronidase. The results showed that both the speed ofmovement and the distance traveled by hematopoietic cells

within the adherent layers was significantly decreased underconditions of HA deprivation. The effect of hyaluronidase wasspecific, as the addition of exogenous HA back to the culturessignificantly increased hematopoietic cell movement. In linewith this observation, the addition of HA into the control cul-tures increased cell motility (Fig. 4A).The stimulating effect of HA on the motility of hematopoi-

etic cells in LTBMC could result directly from HA-stimulatedintracellular signaling for motility (31), or could arise second-arily to HA-stimulated production of soluble chemotactic fac-tors. To test these possibilities, we first examined the effect ofHA on the SDF-1-mediated chemotaxis of hematopoietic cellsusing Transwell assays in which HA was added into the lowerwells. The results demonstrated that, in the absence of SDF-1,HA did not stimulate chemotaxis of either mature hematopoi-etic cells or progenitors (supplemental Fig. S4). These findingssuggested that the HA-induced increase in motility of hemato-poietic cells in LTBMC is likely not a direct effect of HA onhematopoietic cells, but rather results from HA-stimulated

FIGURE 3. Effect of 4MU-induced HA deprivation on hematopoiesis in murine LTBMC. A, LTBMC were treated with 300 �M 4MU or vehicle (PBS, controlcultures) during weekly feeding. The number of nonadherent cells is expressed as mean � S.D. of triplicates from three similar experiments. Significantdifferences (p � 0.05) in the number of nonadherent cells was observed from week 3 through 7 of culture. B, nonadherent cells collected during each feedingwere tested in quadruplicate for the number of committed progenitors using a cfu assay. The number of progenitors produced in each LTBMC is shown asmean � S.D. Significant differences (p � 0.05) in the number of progenitors was observed from week 1 through 7 of culture. C and D, the number ofnonadherent cells (C) and progenitors (D) in control vehicle (PBS)-treated LTBMC and in LTBMC treated with 300 �M 4MU was measured during week 3 ofculture. Where indicated, 100 �g/ml of HMW HA (umbilical cord, 600 –1,500 kDa) or 100 �g/ml of CS were added. The results of one experiment representativeof three similar experiments are shown as mean � S.D. Significant differences between the groups is indicated by an asterisk (*, p � 0.05). Images (�20) ofcontrol vehicle (PBS)-treated (E) and 4MU-treated (F) LTBMC were taken during week 3 of culture. Representative images from one of three similar experiments areshown.




http://ww

w.jbc.org/

Dow

nloaded from



http://www.jbc.org/

production of soluble chemotactic factors by the accessorymicroenvironmental cells.To determine whether hematopoietic progenitors respond

to theHA-induced chemokines, we used staticTranswell assaysto examine the effect of CM from HA-treated adherentLTBMC layers on SDF-1-mediated chemotaxis. We found thatin contrast to HA alone (supplemental Fig. S4), the presence ofHA-CM, but not CM from control untreated cultures (CCM),significantly increased the number of progenitors thatmigratedtoward SDF-1 (Fig. 4B). The low level of progenitor cell migra-tion towardHACMwas also detectedwithout SDF-1 (1.6� 0.2colony forming cells per well), whereas plain media and CCMhad no effect (supplemental Fig. S5). Treatment of CCM orHA-CM with hyaluronidase did not influence the chemotacticability of hematopoietic progenitors toward SDF-1 supple-mented with CM. Similarly, the addition of exogenous HA intothe CCMandHA-CMdid not influence the chemotactic abilityof hematopoietic cells toward SDF-1 (Fig. 4B). These findingssuggest that SDF-1-mediated chemotaxis of hematopoieticprogenitors can be enhanced by soluble factors induced by HAin accessory bone marrow cells.HA Stimulates Production of Soluble Factors by the Adherent

Layer of LTBMC—To screen for HA-induced factors, we usedprotein microarray technologies to measure 62 distinct growthfactors in individual LTBMC samples (Quansys Biosciences,UT, and RayBiotech, Inc., CA). Established adherent layersfrom steady state LTBMC were precultured for 12 h in serum-free medium, and then stimulated with serum-free mediumsupplemented with 100 �g/ml of HMW HA (umbilical cord,

1,500–600 kDa) or LMW HA (15 kDa). Control culturesreceived only serum-free media. After 24 h of stimulation, CMwas harvested and analyzed for the presence of cytokines usingtwo proteinmicroarray platforms (fromRayBiotech andQuan-sys Biosciences). As a negative control, plain culture mediumnot exposed to cells and supplemented with HMW HA orLMW HA was used. The results of both the RayBiotech plat-form (Fig. 4C) and Quansys Biosciences (Table 1) assays wereconsistent, and confirmed our previous findings that HA up-regulates IL-1 and IL-6 in vitro (16) and in vivo (32). The anal-

FIGURE 4. Effect of HA on the production of chemotactic factors and cell motility in vitro. A, steady state LTBMC (4 weeks) were untreated (control, PBS) ortreated with hyaluronidase (10 units/ml) or HMW HA (100 �g/ml, umbilical cord, 600 –1,500 kDa) alone. Where indicated, hyaluronidase was washed away andHMW HA was added. After 72 h the velocity of 10 representative motile cells in control and treated LTBMC was calculated in time lapse videos and expressedas mean � S.D. The results of one of two similar experiments are shown. B, the effect of HMW HA CM from LTBMC on SDF-1-mediated chemotaxis ofhematopoietic cells was tested using Matrigel-coated Transwells. Bone marrow cells that migrated toward plain media (C: control), control CM, or HMW HA CMsupplemented with SDF-1 (50 ng/ml), were tested in triplicate. Where indicated, HMW HA (100 �g/ml) or hyaluronidase (10 units/ml) was added to the CM.Migrated bone marrow cells were collected from the lower wells and further assayed for the number of migrated progenitors using methylcellulose cultures.Asterisk (*) indicates a statistically significant difference (p � 0.05) between Transwells containing SDF-1 alone versus Transwells containing SDF-1 plus HMW HACM. C, LTBMC were established in 6-well plates and cultured for 3 weeks. Cells were starved for 12 h then stimulated with 100 �g/ml of HMW or LMW HA dilutedin serum-free media. Control cultures were incubated with serum-free media alone. After 24 h of incubation, supernatants were collected and analyzed for thepresence of 62 different cytokines and chemokines using the RayBio Mouse Cytokine Antibody Array III and 3.1. The effect of HMW and LMW HA on theproduction of chemokines and cytokines compared with control, nonstimulated LTBMC is shown. The manufacturer’s positive (P.C.) and negative (N.C.)controls for the assay are indicated, and the results of one test of two similar experiments are shown.

TABLE 1The effect of HMW HA on cytokine and chemokine production inLTBMCThe control and HA CM samples were analyzed at dilutions of 1:5 and 1:25 usingQuansys Q-Plex Mouse Cytokine Array. Quantification of images was performedusing the Quansys Array software. The experiment has been performed twice withsimilar results.

Cytokines Control CM HA CM

pg/mlIL-1� 0 204IL-1� 0 160IL-2 0 29IL-3 0.1 19IL-4 0 4.8IL-5 0 6.2IL-6 6 4012.7IL-9 122 65009.4IL-10 0 2171.2IL-12 0.1 172.9MCP-1 783 179665.3INF� 24 6709TNF� 0 14412.5MIP-1� 0 49221.6GM-CSF 1.6 102.7RANTES 1.7 370.9




http://ww

w.jbc.org/

Dow

nloaded from



http://www.jbc.org/

yses of HA-CM from LTBMC extend these results to show thatHMW HA also increased production of several chemokines,includingMIP-1�,MCP-1, CXCL16, SDF-1, andRANTES (Fig.4C, Table 1). In addition, HMWHA stimulated the productionof growth factors including G-CSF, IGFBP-3, IL-12, LIX, IL-9,KC, and soluble VCAM-1 (Fig. 4C, Table 1). Control culturemedium with or without HA that were not exposed to cellsshowed negative results (not shown), indicating that the testedHA samples were not contaminated with detectable levels ofsoluble factors tested by the protein microarray.In Vivo Model of Pharmacologically Induced HA Depletion—

Having shown that HA-stimulated factors augmented migra-tion of HSPC in vitro, we next determined if endogenous HA isimportant for HSPCmotility and homing in vivo. Because con-stitutive knock-out of HAS2 results in embryonic lethality ofmice (33), we tested an alternate in vivo model in which HAsynthesis is transiently inhibited and the levels of HA in bonemarrow are reduced. To ensure that the bone marrow is ascompletely depleted as possible, we combined two approaches:1) lethal irradiation of mice to ablate hematopoietic tissue andto degrade endogenous HA produced by host bone marrow

cells, and 2) pharmacological inhibition of HA synthesis totransiently prevent de novo production of HA by host bonemarrow cells. A schematic of the experimental protocol for thein vivo experiments is shown in Fig. 5A.We first assessed the ability of lethal irradiation and 4MU

treatment to deplete endogenous HA. To do this, mice werelethally irradiated with 11 gray followed by four consecutiveinjections with 4MU (referred as “treated” in Fig. 5B) or PBS(referred as “control”) with 6-h intervals. Both control andtreated mice were sacrificed after the fourth injection of 4MUor PBS, and bone marrow cells and peripheral blood cells werecollected. The cells were lysed, and lysates were tested for thepresence ofHAby anELISA-like assay. As shown in Fig. 5B, thistreatment protocol was very effective in reducing endogenousHA levels, with a 20.5-fold reduction of HA in serum, a 20-folddecrease in bone marrow cells, and a 14-fold decrease inperipheral blood cells in treatedmice, comparedwithHA levelsin control mice. Thus, this approach is effective in markedlyreducing HA expression in vivo, and the mice can be used as anexperimental model in which HA levels in bone marrow aretransiently decreased.

FIGURE 5. The effect of HA depletion in an in vivo mouse model. A, schematic representation of a transiently induced HA depletion model is shown. Lethallyirradiated mice received four injections of PBS (control mice) or 4MU (treated mice). B, after the fourth 4MU or PBS injection, control and treated mice (n �3/group) were sacrificed and HA levels in serum, bone marrow cells (BMC) and peripheral blood cells (PBC) collected from each mouse were measured using anELISA-like assay. C, after the fourth 4MU or PBS injection, control and treated mice (n � 10/group, CD45.2 mice) received intravenous injection of lineage-negative (Lin�) bone marrow cells enriched with HSCs as illustrated in supplemental Fig. S4. After 24 h following cell administration the control and treatedmice were sacrificed (donor mice), bone marrow cells were isolated, and the number of HSC that had homed into the recipients’ bone marrow was tested usingCRA as illustrated in supplemental Fig. S5. The percent ratio between donor (CD45.2) and competitor (CD45.1) peripheral blood cells in recipients is shown.D, the survival of the CRA recipients (n � 10/group, illustrated in supplemental Fig. S7) that received competitor bone marrow cells mixed with donor bonemarrow cells from either control mice (PBS) or treated mice (4MU) is shown. This experiment was repeated twice.




http://ww

w.jbc.org/

Dow

nloaded from




http://www.jbc.org/

Transient Depletion of Endogenous HA Reduces Homing ofTransplanted HSPC into Murine Bone Marrow—To test theeffect of HA reduction on recruitment of transplanted HSPCintomouse bonemarrow, we used a combination of two assays:1) a standard homing assay to assess the ability of intravenouslyinjected cells that circulate in blood to enter bone marrow, fol-lowed by 2) a classical CRA to evaluate the number of long-termreconstituting HSCs that entered the ablated bone marrow.Bl6/CD45.2 mice (C57BL/6JolaHsd Pep3a) were the source ofdonor cells, and Bl6/CD45.1 (C57Bl/6.SJL.Ptprca Pep3b/BoyJ)mice were both the source of competitor cells and the recipientmice for the reconstitution assays.To test the role of HA associatedwith themicroenvironment

in HSC homing, we used a model of transient inhibition of HAin bone marrow (Fig. 5A). HSPC isolated from healthy Bl6/CD45.2 donors (step number 1) were injected into lethally irra-diated Bl6/CD45.2 recipients pre-treated with 4MU or PBS(step number 2). In this setting, lethal irradiation not onlydegrades endogenousHA, but also eliminates endogenousHSCin the recipients resulting in hematopoietic ablation. Thisallows detecting donor HSCs injected after irradiation, whichmigrate into “empty” bone marrow. Twenty-four hours later,all Bl6/CD45.2 recipients were sacrificed, bone marrow cellswere collected and used to further evaluate the number of thelong term reconstituting HSC (Bl6/CD45.2) of the donor thathad entered the bone marrow of the recipient (supplementalFig. S6).To evaluate the number of HSCs in collected bone marrow

samples, CRAwas used (supplemental Fig. S7) (34). Competitorbonemarrowwas isolated fromhealthy nontreated Bl6/CD45.1mice. Thereafter, competitor Bl6/CD45.1 bone marrow cellsand donor #2 Bl6/CD45.2 bone marrow cells were mixed andinjected into lethally irradiated recipient Bl6/CD45.1 mice(recipient #2), after which the mice were allowed to recover for8 weeks. The 8-week period allows HSC from the Bl6/CD45.2donor #2 bonemarrowcell suspension to engraft in Bl6/CD45.1recipient #2 bone marrow and generate mature hematopoieticcells that can be detected by FACS asCD45.1� orCD45.2� cells(34). Recipient #2 mice were then sacrificed after 8 weeks andperipheral blood samples were analyzed for the presence ofCD45.1� and CD45.2� leukocytes by FACS analysis (supple-mental Fig. S5). We found that the number of CD45.2� cells inrecipients #2 receiving bone marrow cells from HA-depleted4MU-treated donor #2 mice was significantly lower than inrecipients #2 receiving bone marrow from control donor #2mice (Fig. 5C). In addition, recipient #2mice that received bonemarrow from4MU-treated donor #2mice had reduced survivalas compared with recipient #2 mice reconstituted with bonemarrow from control donor #2 mice (Fig. 5D). Together, theseresults suggest that a lower number of HSCs was recruited intobone marrow of the HA-depleted mice (lethally irradiated4MU-treated recipients #1) as compared with controls (lethallyirradiated PBS-treated recipients #1).4MU Treatment of Endothelial Cells Decreases Adhesive

Interactions between Hematopoietic Progenitors and Endothe-lial Cells under Conditions of Shear Stress—In vivo, the processof homing includes rolling and adhesion of circulating cells tothe luminal surface of endothelial cells, which is followed by

chemokine-mediated transmigration. Therefore, we nextinvestigated whether exposure to 4MU influences the level ofthe cell surface-associated HA in bone marrow-derived endo-thelial cells. We found that incubation of the bone marrow-derived endothelial cell line STR-12 with 4MU significantlydecreased the amount ofHAon the cell surface, as examined bybinding ofHABP (Fig. 6A). To quantitativelymeasure the levelsof HA in STR-12, cells were cultured in the presence of 300 �M

4MU. In control cultures 4MU was omitted. FACS analysisdemonstrated that 55% of cells stopped production of HA (Fig.6B). Because autofluorescence of cells may change followingtreatment with 4MU and interfere with quantitative measure-ments of HA levels, we also used an ELISA-like assay. WhenSTR-12 cells reached confluence, supernatant from control and4MU-treated cultures was collected, cells were lysed, and thelevels of HA in the collected samples were evaluated by theELISA-like assay. We found that 4MU treatment significantlydecreased both secreted HA and cell-associated HA in STR-12cells (Fig. 6C). Western blot analysis demonstrated thatdecreased levels of HA in 4MU-treated STR-12 cells is associ-ated with decreased expression of HAS2 and HAS3 (Fig. 6D).Numerous reports have shown that rolling and adhesion of

various cell types on endothelial cells under physiological flowinvolves interactions of CD44withHA expressed on themicro-vasculature (35–37). Because CD44 has been previouslyreported tomediate homing of HSPC (28), we next investigatedwhether the reduced expression of HA on bone marrow endo-thelial cells influenced rolling and adhesion of HSPC undershear stress conditions. The STR-12 bone marrow-derivedendothelial cell line was cultured in glass microcapillaries andtreated with 4MU to reduce cell surface-associated HA. Due tothe large number of cells required for this assay, we used FDCP-mix cells as a source of hematopoietic progenitors instead ofprimary HSCs. FDCP-mix cells showed reductions in both roll-ing and adhesion in microcapillaries lined with 4MU-treatedendothelial cells, compared with control, untreated endothelialcells (Fig. 6E). This result suggests that HA on the endothelialcell surface is involved in the initial steps of the HSC homingcascade, i.e. for rolling and subsequent adhesion. Althoughimportant, this finding indicates that the in vivo 4MU experi-ments cannot differentiate between effects of HA depletion inthe hematopoietic microenvironment and effects of HA reduc-tion on the luminal surface of vasculature that may causedecreased HSC homing and engraftment.Abnormal Distribution of Hematopoietic Progenitors in

tHAS1/2/3 KOMice—Because HAS2 in tHAS1/2/3 KOmice isconditionally knocked out in cells of mesenchymal origin, andbecause these cells contribute to the structure of the hemato-poietic microenvironment (38, 39), the tHAS1/2/3 KO animalmodel might be a better tool to study the role of HA in regula-tion of the hematopoietic microenvironment specifically. Mal-functioning of the hematopoietic microenvironment might bereflected by abnormal distribution of hematopoietic progeni-tors within the hematopoietic tissues. Therefore, the CFU assaywas used to test the number of hematopoietic progenitors inbone marrow, peripheral blood, spleen, and liver isolated fromtHAS1/2/3 KO, dHAS1/3 KO, and WT mice. A decreasednumber of progenitors in bone marrow and an increased num-




http://ww

w.jbc.org/

Dow

nloaded from






http://www.jbc.org/

ber of granulocyte-macrophage progenitors in extramedullarysites such as peripheral blood, spleen, and liver were found intHAS1/2/3 KO mice, as compared with WT and dHAS1/3 KO(Fig. 7). Similarly, an abnormal distribution of erythroid andmultipotent hematopoietic progenitors in tHAS1/2/3 KOmiceas comparedwith dHAS1/3KOmice andWTcontrolwas dem-onstrated, which confirms an important role for HA associatedwith the microenvironment in regulating hematopoietichomeostasis.

DISCUSSIONAlthough the quantity and quality of transplanted HSC are

important for the recovery of hematopoiesis, the functional sta-tus of the regulatory hematopoieticmicroenvironment is a crit-ical parameter that determines the regenerative function ofHSCs. The ability of the regulatory microenvironment to sup-port hematopoiesis, i.e. its quality, may be compromised underpathological circumstances such as during disease develop-ment or as a result of therapeutic interventions. Thus, thehematopoietic microenvironment is a potentially importanttherapeutic target and should be allowed to recover prior to

HSC transplantation. To effectively prepare the marrowmicroenvironment for HSC transplantation it is important tounderstand which of the molecular pathways regulating thefunction of themicroenvironment are disrupted under the spe-cific pathological condition. In this study we have shown thatendogenous HA, the level of which could be decreased by irra-diation, chemotherapy, hormone therapy, aging, and otherconditions, is an important component of the bone marrowhematopoietic microenvironment. Disruption of HA produc-tion by pharmacological inhibition of HAS activity decreasedhematopoietic activity in vitro and reduced the ability of thebone marrow microenvironment to recruit HSCs to the mar-row in vivo.4MU has been reported to be a potent inhibitor of HA syn-

thesis (23), and we confirmed this by showing that levels of HAwere decreased in 4MU-treated LTBMC. Interestingly, wefound that protein levels of HAS2 and HAS3 were decreased inLTBMC adherent layers, whereas the level of HAS1 was notchanged. These effects of 4MU correlated well with the dra-matic decrease in production of hematopoietic cells in LTBMC.

FIGURE 6. The effect of 4MU treatment on hematopoietic progenitor-endothelial cell interactions. A, the bone marrow-derived endothelial cells STR-12were nontreated (positive control, PC) or treated with 300 �M 4MU. The cell surface expression of HA in control and treated cultures was examined using bHABPand visualized by FITC-conjugated avidin (green). For the negative control (NC), the cells were stained with FITC-conjugated avidin. Nuclei were stained by DAPI(blue). Representative images of three similar staining are shown. B, STR-12 endothelial cells were grown in the presence or absence of 300 �M 4MU in culturemedia for 48 h. Thereafter, the cells were collected and the presence of HA on the cell surface was evaluated by flow cytometry using bHABP. For the negativecontrol, the cells were pre-treated with hyaluronidase prior to staining. Overlays of negative control and HA staining is show for control nontreated and4MU-treated cells. C, STR-12 cells were grown in culture media supplemented with 300 �M 4MU. In control cultures 4MU was omitted. After 48 h of culture,supernatants (SN) and lysed cells samples (CS) were tested for the levels of HA using an ELISA-like assay. The experiment was repeated twice. Asterisks (*)indicate a statistically significant difference (p � 0.05) between groups. The expression of HAS2 and HAS3 was tested in these cells by Western blot (D). Arepresentative image of two similar experiments is shown. Actin was used as a loading control. E, STR-12 endothelial cells were grown in microcapillaries in theabsence (control) or presence of 300 �M 4MU (treated). Hematopoietic cells (FDCP-mix cell line) were allowed to interact with endothelial cells grown inmicrocapillaries (n � 5/group) under conditions of physiological shear stress. The number of rolling and adherent FDCP-mix cells is shown as mean � S.D. isshown. The experiment was repeated twice.




http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

In line with these findings, lack of hematopoiesis in LTBMCfrom tHAS1/2/3 KO bone marrow was observed, whereashematopoietic activity in dHAS1/3 KO cultures was compara-ble with that from control WT cultures. Together, these find-ings suggest thatHAassociatedwith the stromalmicroenviron-ment is required for hematopoiesis.The lack of hematopoietic activity in the 4MU-treated

LTBMC could be mediated by several distinct mechanisms.However, it is not due to the toxic effect of 4MU on bone mar-row cells. By using annexinV binding and trypan blue exclusiontests we have shown that at 300 �M 4MU does not induce apo-ptosis and cell death, whereas higher concentrations are toxic.Furthermore, the addition of 300 �M 4MU in methylcellulosecultures did not change the number and size of colonies, sug-gesting that at this concentration 4MU does not interfere withcolony formation. Because the cobblestone areas and hemato-poietic foci within the adherent layer of 4MU-treated LTBMCare not formed, it is likely that the pool of multipotent HSCsand committed progenitors is affected indirectly, throughinterfering with the hematopoiesis supportive function of thehematopoietic microenvironment.Consistent with the effect of 4MUon hematopoiesis, HA-de-

pleted cultures treated with exogenous HA showed a corre-sponding increase in the number of progenitors asmeasured bythe CFU assay. These stimulatory effects of HA in 4MU-treated cultures are mediated, at least in part, by CD44,which is in line with our previously published findings (16,28–30). However, the role of other glycosaminoglycans,

such as CS, and other receptors (TLR4 (40, 41), RHAMM(42), and HARE (43)) cannot be excluded and requires addi-tional investigation.The analyses of CM from LTBMC demonstrated a striking

increase in the concentrations of a variety of proliferation-stim-ulating cytokines in the HMW HA-stimulated cultures. Nota-ble concentration changes occurred in factors that stimu-late proliferation of HSCs and progenitors directly, such asIGFBP-3 (44) and LIX/CXCL5 (45), as well as those that actsynergistically with other cytokines to induce proliferation,such as IL-12 (46). In addition to positive regulators, HMWHAstimulated production of some negative regulators of HSPCproliferation such as IL-8, which has been shown to be a nega-tive regulator of myeloid progenitor cell proliferation (47, 48).However, the suppressive effects of IL-8 can be blocked byMIP-2� (47), which was also up-regulated by HMWHA in ourcultures. We have also detected increased concentrations ofMIP-1� and MIP-1� in HMWHA-CM, of which MIP-1� wasreported to inhibit proliferation and colony formation of mye-loid progenitor cells (47). However, Broxmeyer and colleagues(47) have also shown that MIP-1� can block the suppressiveeffects of MIP-1� on progenitor cell proliferation. We alsodetected thatHMWHAstimulates the production ofMCP-1 inLTBMC. The effects of MCP-1 on HSPC proliferation are con-troversial: several groups have reported that MCP-1 inhibitsprogenitor cell proliferation (49, 50), whereas others havedescribed its hematopoiesis-promoting activity (51).

FIGURE 7. Distribution of hematopoietic progenitors in mice with HA deficiency. The number of CFU in BM (A), blood (B), spleen (C), and liver (D) in wild type(WT, n � 3), double Has knock-out (KO) mice (dHAS1/3 KO, Has1�/�;Has3�/�, n � 6), and triple Has KO mice (tHAS1/2/3 KO, Prx1-Cre;Has2flox/flox;Has1�/�;Has3�/�, n � 5) was determined by colony forming assay. The number of progenitors in each sample is shown as mean � S.D. Asterisks (*) indicate a statisticallysignificant difference (p � 0.05) between groups.




http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

Interestingly, HMW HA stimulated the production of solu-ble VCAM-1 in LTBMC. Soluble VCAM-1 inhibits the VLA-4/VCAM-1 cell adhesion pathway and might contribute to thedetachment of progenitor cells from the niche (52). Whendetached, the progenitor cells becomemore sensitive to stimu-latory factors (53), and thismight contribute to the overall stim-ulatory effects of HA on hematopoiesis in LTBMC. HMWHAis known to have a stimulatory effect on cell surface expressionand turnover of VCAM-1 (18), whichmay explain the increasedsVCAM-1 concentrations observed in LTBMC. Collectively,the results of the cytokine analyses demonstrate the complexityof the effects of HA on hematopoiesis. Positive and negativefactors form a network that regulates proliferation of stemcells and committed progenitors according to physiologicaldemands. HA appears to be one of the key molecules that con-trol the balance between positive and negative regulators ofHSC proliferation.Given that HA regulates the production of cytokines and

chemokines involved in cell trafficking, it was important todeterminewhetherHA affected themotility of cells in LTBMC.Indeed, analysis of the time lapse microscopy recordings dem-onstrated that cells in HA-treated LTBMC moved faster, andcorrespondingly, their velocity was decreased when HA isdegraded. HA has been previously reported to influence cellmotility of fibroblasts (15), and solubleHA is reported to inhibittransmigration of human HSPC (54). Therefore, we investi-gated whether the motility effects observed in HA-treatedLTBMC were mediated by HA or by the HA-induced chemo-kines. We found that HA did not directly influence SDF-1-me-diated chemotaxis of hematopoietic cells, but HA-CM actedsynergistically with SDF-1 to increase chemotaxis. Proteinmicroarray analysis demonstrated that HA regulates produc-tion of several chemokines, including MCP-1, RANTES, andMIPs, which alone have no or low effect of migration of hema-topoietic progenitors, whereas in the presence of SDF-1 wereshown to exhibit a synergistic effect (55). Thus, as a componentof the local microenvironment, HA regulates the production ofchemokines that may contribute to the recruitment of circulat-ing HSCs and progenitors into the bone marrow, as illustratedin Fig. 8. However, the effect of HA in marrow is not limited toregulation of soluble factor production. HA is also known tocontribute to soluble factor sequestration, retention, and pres-

entation,mediating cell crawling, and interactionswith the vas-culature (56–58).Until now the lack of appropriate mouse models has made it

challenging to study the role of endogenous HA in vivo.Although all three HA synthases, HAS1, HAS2, and HAS3, arepresent in bonemarrow, HAS2 appears to be the dominant HAsynthase producing HMW HA. HAS1 and HAS3 knock-outmice are both viable (59, 60) becauseHAS2 can compensate theproduction of HA; by contrast, HAS2 knock-out mice demon-strate embryonic lethality (33). This prompted us to develop analternate approach to inhibitHAS2 inmarrow. Previous studieshave demonstrated that in addition to hematopoietic ablation,whole body irradiation sharply decreases the concentration ofglycosaminoglycans, including HA, in mouse spleen and bonemarrow (61). During exposure to irradiation, HA undergoeschemical degradation and depolymerization (62). IrradiationinducesHA chain fragmentation, and dose-dependently affectsthe three-dimensional polymeric structure of HA (63, 64).Thus, irradiation induces a synchronous disruption and degra-dation ofHA, but does not interferewithHA synthesis (65–67).Importantly, 4MU exerts inhibitory effects on HA synthesis byeffectively depleting cellular UDP-GlcA, one of the two sub-strates needed for HA synthesis (23) and causes a profounddisruption of HA synthesis in vitro (68). Previous studies haveshown that injection ofmice with 4MUprofoundly inhibits HAsynthesis for at least 3 h post-injection (69). Thus, for our in vivomodel we used a combination of irradiation and 4MU treat-ment to induce a transient but marked decrease in marrow HAconcentrations.The homing and CRA experiments demonstrated that

recruitment of transplanted HSCs into the marrow is less effi-cient in mice that were treated with 4MU in addition to lethalirradiation. The recruitment of circulating HSCs into marrow(i.e. homing) is a multistep process, which involves rolling ofHSCs on the luminal surface of endothelial cells (low affinityadhesive interactions), subsequent firm adhesion (high affinityadhesive interactions), followed by transmigration through theendothelial cell layer. BecauseHA is expressed on the surface ofbone marrow-derived endothelial cells, it is possible that thereduced homing efficiency in 4MU-treated animals is in partdue to the lack of HA on the endothelial cell surface. We testedthis hypothesis by growing nontreated and 4MU-treated bonemarrow-derived endothelial cells in microcapillary tubes, andtesting their ability to support rolling and adhesion of FDCP-mix cells under conditions of physiological shear stress. Con-sistent with our hypothesis, we found that endothelial cells withreduced surface expression of HA are significantly less able tosupport adhesive interactions with hematopoietic cells underconditions of physiological shear stress. Thus, tHAS1/2/3 KOmice, in which HAS2 is conditionally deleted in cells of mesen-chymal origin only might be a better experimental model toaddress the role of microenvironment-associated HA in regu-lating functions of stem cells.Interestingly, lack ofHAproduction by cells ofmesenchymal

origin correlates with re-distribution of progenitors in tHAS1/2/3KOmice as comparedwith dHAS1/3KOandWTmice.Wefound a decreased number of progenitors in bone marrow andincreased number of progenitors in blood, spleen, and liver of

FIGURE 8. Schematic representation of HA-mediated recruitment ofHSCs. HA secreted by bone marrow cells activates cells of the hematopoieticmicroenvironment and stimulates them to produce cytokines and chemo-kines that participate in the regulatory network in the marrow.




http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

tHAS1/2/3 KO mice, whereas in dHAS1/3 KO mice the num-ber of progenitors in tissues was similar to that in WT mice.Interestingly, increased levels ofmarrow-associatedHA in hya-luronidase 2KOmice also resulted in hematopoietic abnormal-ities (70). Thus, a complex machinery that involves HA syn-thesis, binding, retention, accumulation, degradation, andclearance should be orchestrated to maintain optimal levels ofHA in healthy tissues. Together, our findings and observationspublished by others suggest that the tissue-associated HA inbonemarrowmight be clinically relevant and reflect the biolog-ical health of the hematopoietic microenvironment.Collectively, our results strongly suggest that HA is a biolog-

ically active component of the hematopoietic microenviron-ment and is involved in regulating hematopoietic homeostasis.Because some treatments or compounds reduceHAconcentra-tions in tissues (32) and some conditions are associated withincreased levels of HA (71), it may prove clinically useful tomonitor the dynamics of endogenous HA recovery to aid inidentifying the optimal time for stem cell transplantation. Ourdata also suggest that biologically active exogenous HA poly-mers of the correct size, source, and conformation aswell asHAsynthesis inhibitors may have potential use in clinical hematol-ogy to correct misbalanced HA levels.

Acknowledgment—We thank John McDonald for providing Has3knock-out mice.

REFERENCES1. Wright, D. E., Bowman, E. P.,Wagers, A. J., Butcher, E. C., andWeissman,

I. L. (2002) Hematopoietic stem cells are uniquely selective in their migra-tory response to chemokines. J. Exp. Med. 195, 1145–1154

2. Kim,C.H., and Broxmeyer, H. E. (1998) In vitro behavior of hematopoieticprogenitor cells under the influence of chemoattractants. Stromal cell-derived factor-1, steel factor, and the bone marrow environment. Blood91, 100–110

3. Fukuda, S., Bian, H., King, A. G., and Pelus, L. M. (2007) The chemokineGRO� mobilizes early hematopoietic stem cells characterized by en-hanced homing and engraftment. Blood 110, 860–869

4. Broxmeyer,H. E. (2001) Regulation of hematopoiesis by chemokine familymembers. Int. J. Hematol. 74, 9–17

5. Kollet, O., Petit, I., Kahn, J., Samira, S., Dar, A., Peled, A., Deutsch, V.,Gunetti, M., Piacibello, W., Nagler, A., and Lapidot, T. (2002) HumanCD34(�)CXCR4(�) sorted cells harbor intracellular CXCR4, which canbe functionally expressed and provide NOD/SCID repopulation. Blood100, 2778–2786

6. Weidt, C., Niggemann, B., Kasenda, B., Drell, T. L., Zänker, K. S., andDittmar, T. (2007) Stem cell migration. A quintessential stepping stone tosuccessful therapy. Curr. Stem Cell Res. Ther. 2, 89–103

7. Askari, A. T., Unzek, S., Popovic, Z. B., Goldman, C. K., Forudi, F., Kied-rowski, M., Rovner, A., Ellis, S. G., Thomas, J. D., DiCorleto, P. E., Topol,E. J., and Penn, M. S. (2003) Effect of stromal cell-derived factor 1 onstem-cell homing and tissue regeneration in ischaemic cardiomyopathy.Lancet 362, 697–703

8. Ratajczak, M. Z., Kucia, M., Reca, R., Majka, M., Janowska-Wieczorek, A.,and Ratajczak, J. (2004) Stem cell plasticity revisited. CXCR4-positive cellsexpressing mRNA for early muscle, liver and neural cells “hide out” in thebone marrow. Leukemia 18, 29–40

9. Stumm, R. K., Rummel, J., Junker, V., Culmsee, C., Pfeiffer,M., Krieglstein,J., Höllt, V., and Schulz, S. (2002) A dual role for the SDF-1/CXCR4chemokine receptor system in adult brain. Isoform-selective regulation ofSDF-1 expression modulates CXCR4-dependent neuronal plasticity andcerebral leukocyte recruitment after focal ischemia. J. Neurosci. 22,

5865–587810. Klein, G. (1995) The extracellular matrix of the hematopoietic microen-

vironment. Experientia 51, 914–92611. Yoder, M. C., and Williams, D. A. (1995) Matrix molecule interactions

with hematopoietic stem cells. Exp. Hematol. 23, 961–96712. Fraser, J. R., Laurent, T. C., and Laurent, U. B. (1997) Hyaluronan. Its

nature, distribution, functions and turnover. J. Intern. Med. 242, 27–3313. Brecht, M., Mayer, U., Schlosser, E., and Prehm, P. (1986) Increased hya-

luronate synthesis is required for fibroblast detachment and mitosis.Biochem. J. 239, 445–450

14. Hamann, K. J., Dowling, T. L., Neeley, S. P., Grant, J. A., and Leff, A. R.(1995) Hyaluronic acid enhances cell proliferation during eosinopoiesisthrough the CD44 surface antigen. J. Immunol. 154, 4073–4080

15. Andreutti, D., Geinoz, A., and Gabbiani, G. (1999) Effect of hyaluronicacid onmigration, proliferation and �-smoothmuscle actin expression bycultured rat and human fibroblasts. J. Submicrosc. Cytol. Pathol. 31,173–177

16. Khaldoyanidi, S., Moll, J., Karakhanova, S., Herrlich, P., and Ponta, H.(1999) Hyaluronate-enhanced hematopoiesis. Two different receptorstrigger the release of interleukin-1� and interleukin-6 from bone marrowmacrophages. Blood 94, 940–949

17. Noble, P. W., Lake, F. R., Henson, P. M., and Riches, D. W. (1993) Hyalu-ronate activation of CD44 induces insulin-like growth factor-1 expressionby a tumor necrosis factor-�-dependent mechanism in murine macro-phages. J. Clin. Invest. 91, 2368–2377

18. Oertli, B., Beck-Schimmer, B., Fan, X., and Wüthrich, R. P. (1998) Mech-anisms of hyaluronan-induced up-regulation of ICAM-1 and VCAM-1expression by murine kidney tubular epithelial cells. Hyaluronan triggerscell adhesionmolecule expression through amechanism involving activa-tion of nuclear factor-�B and activating protein-1. J. Immunol. 161,3431–3437

19. Entwistle, J., Hall, C. L., and Turley, E. A. (1996) HA receptors. Regulatorsof signalling to the cytoskeleton. J. Cell Biochem. 61, 569–577

20. Matsumoto, K., Li, Y., Jakuba, C., Sugiyama, Y., Sayo, T.,Okuno,M., Dealy,C. N., Toole, B. P., Takeda, J., Yamaguchi, Y., and Kosher, R. A. (2009)Conditional inactivation of Has2 reveals a crucial role for hyaluronan inskeletal growth, patterning, chondrocyte maturation and joint formationin the developing limb. Development 136, 2825–2835

21. Kobayashi, N., Miyoshi, S., Mikami, T., Koyama, H., Kitazawa, M.,Takeoka,M., Sano, K., Amano, J., Isogai, Z., Niida, S., Oguri, K., Okayama,M., McDonald, J. A., Kimata, K., Taniguchi, S., and Itano, N. (2010) Hya-luronan deficiency in tumor stroma impairs macrophage trafficking andtumor neovascularization. Cancer Res. 70, 7073–7083

22. Bai, K. J., Spicer, A. P., Mascarenhas, M. M., Yu, L., Ochoa, C. D., Garg,H. G., and Quinn, D. A. (2005) The role of hyaluronan synthase 3 inventilator-induced lung injury. Am. J. Respir. Crit. Care Med. 172, 92–98

23. Kakizaki, I., Kojima, K., Takagaki, K., Endo, M., Kannagi, R., Ito, M.,Maruo, Y., Sato, H., Yasuda, T.,Mita, S., Kimata, K., and Itano, N. (2004) Anovel mechanism for the inhibition of hyaluronan biosynthesis by4-methylumbelliferone. J. Biol. Chem. 279, 33281–33289

24. Imai, K., Kobayashi,M.,Wang, J., Ohiro, Y., Hamada, J., Cho, Y., Imamura,M.,Musashi,M., Kondo, T.,Hosokawa,M., andAsaka,M. (1999) Selectivetransendothelial migration of hematopoietic progenitor cells. A role inhoming of progenitor cells. Blood 93, 149–156

25. Jokela, T. A., Jauhiainen, M., Auriola, S., Kauhanen, M., Tiihonen, R.,Tammi, M. I., and Tammi, R. H. (2008) Mannose inhibits hyaluronansynthesis by down-regulation of the cellular pool of UDP-N-acetylhexo-samines. J. Biol. Chem. 283, 7666–7673

26. Vigetti, D., Ori, M., Viola, M., Genasetti, A., Karousou, E., Rizzi, M., Pal-lotti, F., Nardi, I., Hascall, V. C., De Luca, G., and Passi, A. (2006) Molec-ular cloning and characterization of UDP-glucose dehydrogenase fromthe amphibian Xenopus laevis and its involvement in hyaluronan synthe-sis. J. Biol. Chem. 281, 8254–8263

27. Vigetti, D., Rizzi, M., Moretto, P., Deleonibus, S., Dreyfuss, J. M., Karou-sou, E., Viola, M., Clerici, M., Hascall, V. C., Ramoni, M. F., De Luca, G.,and Passi, A. (2011) Glycosaminoglycans and glucose prevent apoptosis in4-methylumbelliferone-treated human aortic smooth muscle cells. J. Biol.Chem. 286, 34497–34503




http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

28. Khaldoyanidi, S., Denzel, A., and Zöller, M. (1996) Requirement for CD44in proliferation and homing of hematopoietic precursor cells. J. LeukocyteBiol. 60, 579–592

29. Khaldoyanidi, S., Karakhanova, S., Sleeman, J., Herrlich, P., and Ponta, H.(2002) CD44 variant-specific antibodies trigger hemopoiesis by selectiverelease of cytokines from bone marrow macrophages. Blood 99,3955–3961

30. Moll, J., Khaldoyanidi, S., Sleeman, J. P., Achtnich,M., Preuss, I., Ponta, H.,and Herrlich, P. (1998) Two different functions for CD44 proteins in hu-man myelopoiesis. J. Clin. Invest. 102, 1024–1034

31. Pilarski, L. M., Masellis-Smith, A., Belch, A. R., Yang, B., Savani, R. C., andTurley, E. A. (1994) RHAMM, a receptor for hyaluronan-mediated motil-ity, on normal human lymphocytes, thymocytes and malignant B cells. Amediator in B cell malignancy? Leuk. Lymphoma 14, 363–374

32. Matrosova, V. Y., Orlovskaya, I. A., Serobyan, N., and Khaldoyanidi, S. K.(2004) Hyaluronic acid facilitates the recovery of hematopoiesis following5-fluorouracil administration. Stem Cells 22, 544–555

33. Camenisch, T. D., Spicer, A. P., Brehm-Gibson, T., Biesterfeldt, J., Augus-tine, M. L., Calabro, A., Jr., Kubalak, S., Klewer, S. E., and McDonald, J. A.(2000) Disruption of hyaluronan synthase-2 abrogates normal cardiacmorphogenesis and hyaluronan-mediated transformation of epitheliumto mesenchyme. J. Clin. Invest. 106, 349–360

34. Chang, H., Jensen, L. A., Quesenberry, P., and Bertoncello, I. (2000) Stand-ardization of hematopoietic stem cell assays. A summary of a workshopand working group meeting sponsored by the National Heart, Lung, andBlood Institute held at the National Institutes of Health, Bethesda, MD onSeptember 8–9, 1998 and July 30, 1999. Exp. Hematol. 28, 743–752

35. Milinkovic, M., Antin, J. H., Hergrueter, C. A., Underhill, C. B., and Sack-stein, R. (2004) CD44-hyaluronic acid interactions mediate shear-resis-tant binding of lymphocytes to dermal endothelium in acute cutaneousGVHD. Blood 103, 740–742

36. Siegelman, M. H., DeGrendele, H. C., and Estess, P. (1999) Activation andinteraction of CD44 and hyaluronan in immunological systems. J. Leuko-cyte Biol. 66, 315–321

37. Siegelman, M. H., Stanescu, D., and Estess, P. (2000) The CD44-initiatedpathway of T-cell extravasation uses VLA-4 but not LFA-1 for firm adhe-sion. J. Clin. Invest. 105, 683–691

38. Park, D., Sykes, D. B., and Scadden, D. T. (2012) The hematopoietic stemcell niche. Front. Biosci. 17, 30–39

39. Schraufstatter, I.U.,Discipio,R.G., andKhaldoyanidi, S. (2012)Mesenchymalstem cells and their microenvironment. Front. Biosci. 17, 2271–2288

40. Jiang, D., Liang, J., Fan, J., Yu, S., Chen, S., Luo, Y., Prestwich, G. D.,Mascarenhas, M. M., Garg, H. G., Quinn, D. A., Homer, R. J., Goldstein,D. R., Bucala, R., Lee, P. J., Medzhitov, R., and Noble, P. W. (2005) Regu-lation of lung injury and repair by Toll-like receptors and hyaluronan.Nat.Med. 11, 1173–1179

41. Taylor, K. R., Trowbridge, J. M., Rudisill, J. A., Termeer, C. C., Simon, J. C.,and Gallo, R. L. (2004) Hyaluronan fragments stimulate endothelial rec-ognition of injury through TLR4. J. Biol. Chem. 279, 17079–17084

42. Pilarski, L. M., Pruski, E., Wizniak, J., Paine, D., Seeberger, K., Mant, M. J.,Brown, C. B., and Belch, A. R. (1999) Potential role for hyaluronan and thehyaluronan receptor RHAMM inmobilization and trafficking of hemato-poietic progenitor cells. Blood 93, 2918–2927

43. Harris, E. N., Kyosseva, S. V., Weigel, J. A., and Weigel, P. H. (2007) Ex-pression, processing, and glycosaminoglycan binding activity of the re-combinant human 315-kDa hyaluronic acid receptor for endocytosis(HARE). J. Biol. Chem. 282, 2785–2797

44. Liu, L. Q., Sposato, M., Liu, H. Y., Vaudrain, T., Shi, M. J., Rider, K.,Stevens, Z., Visser, J., Deng,H. K., andKraus,M. (2003) Functional cloningof IGFBP-3 from human microvascular endothelial cells reveals its novelrole in promoting proliferation of primitive CD34�CD38- hematopoieticcells in vitro. Oncol. Res. 13, 359–371

45. Choong, M. L., Yong, Y. P., Tan, A. C., Luo, B., and Lodish, H. F. (2004)LIX. A chemokine with a role in hematopoietic stem cells maintenance.Cytokine 25, 239–245

46. Jacobsen, S. E., Veiby, O. P., and Smeland, E. B. (1993) Cytotoxic lympho-cyte maturation factor (interleukin 12) is a synergistic growth factor forhematopoietic stem cells. J. Exp. Med. 178, 413–418

47. Broxmeyer, H. E., Sherry, B., Cooper, S., Lu, L.,Maze, R., Beckmann,M. P.,Cerami, A., and Ralph, P. (1993) Comparative analysis of the humanmacrophage inflammatory protein family of cytokines (chemokines) onproliferation of human myeloid progenitor cells. Interacting effects in-volving suppression, synergistic suppression, and blocking of suppression.J. Immunol. 150, 3448–3458

48. Lu, L., Xiao, M., Grigsby, S., Wang, W. X., Wu, B., Shen, R. N., andBroxmeyer, H. E. (1993) Comparative effects of suppressive cytokines onisolated single CD343� stem/progenitor cells from human bone marrowand umbilical cord blood plated with and without serum. Exp. Hematol.21, 1442–1446

49. Broxmeyer, H. E., Cooper, S., Li, Z. H., Lu, L., Sarris, A., Wang, M. H.,Chang, M. S., Donner, D. B., and Leonard, E. J. (1996) Macrophage-stim-ulating protein, a ligand for the RON receptor protein-tyrosine kinase,suppresses myeloid progenitor cell proliferation and synergizes with vas-cular endothelial cell growth factor andmembers of the chemokine family.Ann. Hematol. 73, 1–9

50. Cashman, J. D., Eaves, C. J., Sarris, A. H., and Eaves, A. C. (1998) MCP-1,notMIP-1�, is the endogenous chemokine that cooperates with TGF-� toinhibit the cycling of primitive normal but not leukemic (CML) progeni-tors in long-term human marrow cultures. Blood 92, 2338–2344

51. Xu, Y. X., Talati, B. R., Janakiraman,N., Chapman, R. A., andGautam, S. C.(1999) Growth factors. Production of monocyte chemotactic protein-1(MCP-1/JE) by bone marrow stromal cells. Effect on the migration andproliferation of hematopoietic progenitor cells. Hematology 4, 345–356

52. Südhoff, T., and Söhngen,D. (2002) Circulating endothelial adhesionmol-ecules (sE-selectin, sVCAM-1, and sICAM-1) during rHuG-CSF-stimu-lated stem cell mobilization. J. Hematother. Stem Cell Res. 11, 147–151

53. Verfaillie, C.M. (1993) Soluble factor(s) produced by humanbonemarrowstroma increase cytokine-induced proliferation and maturation of primi-tive hematopoietic progenitors while preventing their terminal differen-tiation. Blood 82, 2045–2053

54. Avigdor, A., Goichberg, P., Shivtiel, S., Dar, A., Peled, A., Samira, S., Kollet,O., Hershkoviz, R., Alon, R., Hardan, I., Ben-Hur, H., Naor, D., Nagler, A.,and Lapidot, T. (2004) CD44 and hyaluronic acid cooperate with SDF-1 inthe trafficking of human CD34� stem/progenitor cells to bone marrow.Blood 103, 2981–2989

55. Liesveld, J. L., Rosell, K., Panoskaltsis, N., Belanger, T., Harbol, A., andAbboud, C. N. (2001) Response of human CD34� cells to CXC, CC, andCX3C chemokines. Implications for cell migration and activation. J. He-matother. Stem Cell Res. 10, 643–655

56. Christophis, C., Taubert, I., Meseck, G. R., Schubert, M., Grunze, M., Ho,A. D., and Rosenhahn, A. (2011) Shear stress regulates adhesion and roll-ing of CD44� leukemic and hematopoietic progenitor cells on hyaluro-nan. Biophys. J. 101, 585–593

57. Gordon, M. Y., Riley, G. P., Watt, S. M., and Greaves, M. F. (1987) Com-partmentalization of a haematopoietic growth factor (GM-CSF) by gly-cosaminoglycans in the bone marrow microenvironment. Nature 326,403–405

58. Roberts, R., Gallagher, J., Spooncer, E., Allen, T. D., Bloomfield, F., andDexter, T.M. (1988) Heparan sulfate bound growth factors. Amechanismfor stromal cell mediated hemopoiesis. Nature 332, 376–378

59. Spicer, A. P., andNguyen, T. K. (1999)Mammalian hyaluronan synthases.Investigation of functional relationships in vivo. Biochem. Soc. Trans. 27,109–115

60. Spicer, A. P., Tien, J. L., Joo, A., and Bowling, R. A., Jr. (2002) Investigationof hyaluronan function in the mouse through targeted mutagenesis. Gly-coconj. J. 19, 341–345

61. Noordegraaf, E. M., Erkens-Versluis, E. A., and Ploemacher, R. E. (1981)Studies of the hemopoietic microenvironments. V. Changes in murinesplenic and bone marrow glycosaminoglycans during postirradiation he-mopoietic regeneration. Exp. Hematol. 9, 326–331

62. Reháková, M., Bakos, D., Soldán, M., and Vizárová, K. (1994) Depolymer-ization reactions of hyaluronic acid in solution. Int. J. Biol. Macromol. 16,121–124

63. Shin da, Y., Hwang, E., Cho, I. H., and Moon, M. H. (2007) Molec-ular weight and structure characterization of sodium hyaluronate andits � radiation degradation products by flow field-flow fractionation




http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

and on-line multiangle light scattering. J. Chromatogr. A 1160,270–275

64. Srinivas, A., and Ramamurthi, A. (2007) Effects of �-irradiation on phys-ical and biologic properties of cross-linked hyaluronan tissue engineeringscaffolds. Tissue Eng. 13, 447–459

65. Matsumoto, T., Iwasaki, K., and Sugihara, H. (1994) Effects of radiation onchondrocytes in culture. Bone 15, 97–100

66. Pye, D. A., Kumar, S., Wang, M. J., and Hunter, R. D. (1994) Irradiationof bovine aortic endothelial cells enhances the synthesis and secretionof sulphated glycosaminoglycans. Biochim. Biophys. Acta 1220,266–276

67. Wegrowski, J., Lefaix, J. L., and Lafuma, C. (1992) Accumulation of gly-cosaminoglycans in radiation-induced muscular fibrosis. Int. J. Radiat.Biol. 61, 685–693

68. Nakamura, T., Takagaki, K., Shibata, S., Tanaka, K., Higuchi, T., and Endo,M. (1995) Hyaluronic acid-deficient extracellular matrix induced by addi-tion of 4-methylumbelliferone to the medium of cultured human skinfibroblasts. Biochem. Biophys. Res. Commun. 208, 470–475

69. Yoshihara, S., Kon, A., Kudo, D., Nakazawa, H., Kakizaki, I., Sasaki, M.,Endo, M., and Takagaki, K. (2005) A hyaluronan synthase suppressor,4-methylumbelliferone, inhibits liver metastasis of melanoma cells. FEBSLett. 579, 2722–2726

70. Jadin, L., Wu, X., Ding, H., Frost, G. I., Onclinx, C., Triggs-Raine, B., andFlamion, B. (2008) Skeletal and hematological anomalies in HYAL2-defi-cient mice. A second type of mucopolysaccharidosis IX? FASEB J. 22,4316–4326

71. Laurent, T. C., Laurent, U. B., and Fraser, J. R. (1996) Serum hyaluronan asa disease marker. Ann. Med. 28, 241–253




http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

Orlovskaja, Yu Yamaguchi and Sophia KhaldoyanidiRidder, Tatiana Povaliy, Valentina Wacker, Naoki Itano, Koji Kimata, Irina A.

Valentina Goncharova, Naira Serobyan, Shinji Iizuka, Ingrid Schraufstatter, Audrey deBone Marrow Hematopoiesis

Hyaluronan Expressed by the Hematopoietic Microenvironment Is Required for

doi: 10.1074/jbc.M112.376699 originally published online May 31, 20122012, 287:25419-25433.J. Biol. Chem.

10.1074/jbc.M112.376699Access the most updated version of this article at doi:

Alerts:

When a correction for this article is posted•

When this article is cited•

to choose from all of JBC's e-mail alertsClick here

Supplemental material:

http://www.jbc.org/content/suppl/2012/06/06/M112.376699.DC1

http://www.jbc.org/content/287/30/25419.full.html#ref-list-1

This article cites 71 references, 27 of which can be accessed free at


http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/lookup/doi/10.1074/jbc.M112.376699

http://www.jbc.org/cgi/alerts?alertType=citedby&addAlert=cited_by&cited_by_criteria_resid=jbc;287/30/25419&saveAlert=no&return-type=article&return_url=http://www.jbc.org/content/287/30/25419

http://www.jbc.org/cgi/alerts?alertType=correction&addAlert=correction&correction_criteria_value=287/30/25419&saveAlert=no&return-type=article&return_url=http://www.jbc.org/content/287/30/25419

http://www.jbc.org/cgi/alerts/etoc

http://www.jbc.org/content/suppl/2012/06/06/M112.376699.DC1

http://www.jbc.org/content/287/30/25419.full.html#ref-list-1

http://www.jbc.org/

hyaluronanexpressedbythehematopoietic ... · bovine serum, i-inositol, folic acid,...

Documents