Extracellular matrix promotes the growth anddifferentiation of murine hematopoietic cells invitro.

A Campbell, … , M S Wicha, M Long

J Clin Invest. 1985;75(6):2085-2090. https://doi.org/10.1172/JCI111928.

We have developed a long-term culture system in which murine marrow cells are culturedon a complex extracellular matrix (ECM) that is derived from marrow and extracted withguanidine hydrochloride and dithiothreitol. Marrow cultures were established with freshmurine marrow cells and recharged at 2 wk (week 0). Phase microscopy showed adramatically increased adherent cell layer development on ECM compared with controlswithin a week after recharge. By electron microscopy, this adherent layer was composed ofnumerous reticular cells apparently attached to the ECM which extended cytoplasmicprojections to the surrounding hematopoietic cells. Adherent cellularity on ECM-coateddishes increased to 30 times the control values by week 2. Cumulative suspension cells onECM dishes were eight times controls. ECM influenced both hematopoietic progenitor cellproliferation and differentiation. Adherent colony-forming unit-granulocyte/macrophage andcolony-forming unit-megakaryocyte were greater than 30 and 15 times the control values,respectively, by week 2 (P less than or equal to 0.05). There were more mature granulocyticand megakaryocytic cells in ECM-coated dishes than in controls at all time points. This newculture system directly demonstrates that ECM is an important component of thehematopoietic microenvironment.

Research Article

Find the latest version:

http://jci.me/111928-pdf

Rapid Publication

Extracellular Matrix Promotes the Growth and Differentiationof Murine Hematopoietic Cells In VitroAlan Campbell and Max S. WichaDepartment of Internal Medicine, Division of Hematology and Oncology, University of Michigan Medical Center,Simpson Memorial Research Institute, Ann Arbor, Michigan 48109

Michael LongDepartment of Pediatrics and Communicable Diseases, Division of Hematology and Oncology,University of Michigan Medical Center, Ann Arbor, Michigan 48109

Abstract

Wehave developed a long-term culture system in which murinemarrow cells are cultured on a complex extracellular matrix(ECM) that is derived from marrow and extracted with gua-nidine hydrochloride and dithiothreitol. Marrow cultures wereestablished with fresh murine marrow cells and recharged at2 wk (week 0). Phase microscopy showed a dramaticallyincreased adherent cell layer development on ECMcomparedwith controls within a week after recharge. By electron mi-croscopy, this adherent layer was composed of numerousreticular cells apparently attached to the ECMwhich extendedcytoplasmic projections to the surrounding hematopoietic cells.Adherent cellularity on ECM-coated dishes increased to 30times the control values by week 2. Cumulative suspensioncells on ECMdishes were eight times controls. ECMinfluencedboth hematopoietic progenitor cell proliferation and differen-tiation. Adherent colony-forming unit-granulocyte/macrophageand colony-forming unit-megakaryocyte were >30 and 15 timesthe control values, respectively, by week 2 (P < 0.05). Therewere more mature granulocytic and megakaryocytic cells inECM-coated dishes than in controls at all time points. Thisnew culture system directly demonstrates that ECM is animportant component of the hematopoietic microenvironment.

Introduction

The hematopoietic microenvironment is known to influencethe growth and differentiation of hematopoietic cells. Althoughpoorly defined, this microenvironment is composed of bothcellular and extracellular components. In long-term bonemarrow cultures, the maintenance of hematopoiesis is depen-dent on the establishment of an intact layer of adherent cellsthat are derived from the marrow stroma (1-3). Such adherentcell layers are capable of elaborating soluble factors whichinfluence hematopoietic proliferation and/or differentiation

Dr. Long is a scholar of the Leukemia Society of America. Addresscorrespondence to Dr. Campbell, University of Michigan, Division ofHematology/Oncology, 102 Observatory.

Receivedfor publication 21 November 1984 and in revisedform 12March 1985.

(4). Little is known, however, about the function of theextracellular components within the hematopoietic microen-vironment. There is substantial evidence in a variety of otherdevelopmental systems that the extracellular matrix (ECM)'influences cell growth and differentiation (5-7). Recent studiessuggest that the ECM also plays a role in the in vitroproliferation and differentiation of hematopoietic cells. Inhib-itors of matrix formation are known to decrease stem cellproduction (8), while compounds stimulating matrix productionincrease stem cell proliferation (9). Finally, the stromal cellpopulation is known to produce a variety of ECMproteinsincluding fibronectin, laminin, and type IV collagen (10).

We hypothesize that bone marrow-derived ECMplays acrucial role in the development of the hematopoietic microen-vironment. In order to test this hypothesis, we have developeda new system of murine long-term marrow cultures establishedon ECMextracted from bone marrow. We report that thisextracellular material promotes rapid formation of an adherentstromal cell layer and, subsequently, proliferation and differ-entiation of hematopoietic cells. This system demonstrates theimportance of ECMin the hematopoietic microenvironment.

Methods

Preparation of ECM. Femurs and tibias removed from 4- to 6-wk-oldrabbits were obtained from Pel-Freeze Biologicals (Rogers, AR). Rabbitswere used to ensure sufficient quantities of marrow extract. Themarrow was removed and immediately placed in 50 mMTris buffer(pH 7.4) containing 3.4 M NaCl, leupeptin (1 ug/ml), and soybeantrypsin inhibitor (10 jsg/ml) (Sigma Chemical Co., St. Louis, MO) toprecipitate ECMproteins, a modification of a previously describedmethod (5). This suspension was homogenized in a Brinkman Polytronhomogenizer for 30 s at high speed. The marrow homogenate wascentrifuged at 1,500 g for 10 min at 40C and the supernatant andfatty layer removed. The pellet was washed exhaustively in 3.4 MNaCl solution, digested with RNase (100 yg/ml) and DNase (25 ug/ml) (Sigma Chemical Co.) in Fisher's medium at 370C for 1 h, andwashed in 3.4 M NaCl solution overnight. The ECMwas extractedfrom the pellet with 20 ml of 50 mMTris buffer containing 4 Mguanidine HCI and 2 mMdithiothreitol (pH 7.4), which is known tosolubilize proteoglycans and other ECMcomponents (11, 12). Theextract was centrifuged at 1,500g for 10 min and the supernatant

1. Abbreviations used in this paper: CFU-GM, colony-forming unit-granulocyte/macrophage; CFU-Mk, colony-forming unit-megakaryocyte;ECM, extracellular matrix; EHS, Engelbreth-Holm-Swarm; Gdn/DTT,guanidine/dithiothreitol.

Murine Hematopoietic Cells In Vitro 2085

J. Clin. Invest.© The American Society for Clinical Investigation, Inc.0021-9738/85/06/2085/06 $ 1.00Volume 75, June 1985, 2085-2090

removed and dialyzed exhaustively against distilled water. The extractionprocedure was repeated four times and the extracts pooled. 2 ml ofthis extract containing -0.4 mg dry weight of insoluble material waspipetted onto 35-mm tissue culture plastic dishes (Becton-Dickinson& Co.) and allowed to dry overnight. The dishes were sterilized with500,000 rad of 'Co irradiation.

Fresh marrow cells were obtained from 4- to 6-wk-old male,syngeneic C57/BL6 mice (Charles River Breeding Labs, Inc., Wilming-ton, MA). Monodispersed nucleated cells (3 X 106) were suspended in3 ml of Fisher's medium containing 5 X l-7 Mhydrocortisone (SigmaChemical Co.), 40 ,ug/ml of gentamicin, and 20% horse serum (HyCloneLaboratories Inc., Logan, UT). Cultures were established on ECM-coated dishes or on control tissue culture dishes. The cultures wereincubated at 330C in a humidified atmosphere with 5% CO2. At 2 wk(week 0), all nonadherent cells were gently removed and the culturesrecharged with 3 X 106 fresh nucleated marrow cells in 3 ml of Fisher'smedium. At weekly intervals thereafter, medium and suspension cellswere removed and replaced with fresh medium. At various intervals,the adherent layer was removed by incubation with trypsin-EDTA1:250 (KC Biologicals) for I min at room temperature. Total nucleatedcell counts were performed by hemocytometer and cell viabilityassessed by trypan blue exclusion. Cytocentrifuged preparations ofadherent and suspension cells were stained with Wright stain oracetylcholinesterase (13).

Progenitor cell assays. Short-term assays for granulocyte progenitorcells (CFU-GM) and megakaryocyte progenitor cells (CFU-MK) wereperformed as previously described (14, 15). Briefly, bone marrow cellswere cultured as previously described using conditioned medium fromWEHI-3 and mouse lung conditioned medium (16, 17). Before plating,the bone marrow cells were generally incubated at 370C for 30-60min at -5 X 106 cells/ml in McCoy's SA medium containing 10%fetal calf serum (FCS). Marrow cells were cultured at 30-100 X 104cells/ml in 35-mm petri dishes (Becton-Dickinson & Co.) in modifiedMcCoy's SA medium containing 10% FCSand 0.25% Bactoagar. Afterincubation (7 d, 370C, 7% C02, 100% humidity), petri dishes wereremoved from the incubator, dried, and stained in situ for acetylcho-linesterase activity (18). Colonies were counted at a magnificationof 40.

Electron microscopy. Representative matrix and control dishes wereharvested at -2 wk postrefeeding and prepared for scanning andtransmission electron microscopy, as described (19).

Protein analysis. SDS polyacrylamide gel electrophoresis, after themethod of Laemmli (20), was performed on samples of bone marrowguanidine/dithiothreitol (Gdn/DTT) extract, high molecular weightstandards (Sigma Chemical Co.), and Gdn/DTT extract of ECMwasobtained from the Engelbreth-Holm-Swarm (EHS) tumor, which isknown to produce large quantities of the ECMcomponents laminin,type IV collagen, fibronectin, and entactin (12). All gels are run underidentical conditions. The gels were stained using silver nitrate asdescribed (21).

Results

Morphologic characteristics of long-term cultures on matrixsubstrata. Bone marrow cells cultivated in control dishes(tissue culture plastic) developed islands of adherent cellswithin 7-10 d. In contrast, ECMsupported a more rapid andextensive development of an adherent layer that becameconfluent within the same time period. ECMderived from thenonhematopoietic tissues EHStumor and cortical bone (devoidof marrow) did not promote adherent layer formation orhematopoiesis (data not shown). This suggests organ specificity.The adherent layer that developed on marrow-derived ECMwas composed of flat, branching cells that formed large areashaving a cobblestone appearance. However, relatively fewhematopoietic cells were seen in either cultures at this time.After recharging at week 0, the control dishes showed gradual

development of the adherent layer while ECM-coated dishesshowed a striking increase in both hematopoietic and stromalcells (Fig. 1, A and B). This difference was even more evidentwhen the cultures were examined by scanning electron mi-croscopy. In ECMcultures, the adherent cell layer was com-posed of large branching cells that extended long filamentousprojections which encircled hematopoietic cells (Fig. 1, CandD). Transmission electron microscopy demonstrated that theadherent phase of the ECMculture was composed of severalcell layers (Fig. 1, E and F). The basal layer consisted of flatstromal cells in direct contact with the matrix extract. Hema-topoietic cells were seen to overlie this cell layer. They wereoften in close apposition to the stromal cells.

Cellular proliferation. Although hematopoietic cells wereproduced throughout the culture period in both ECMandcontrol dishes, there was an eightfold increase in the suspensionphase of ECMcultures compared with controls (Fig. 2). In theadherent phase, there was an even greater difference betweenthe cellularity of ECMand control cultures. Within 2 wk ofrecharging the cultures, there was a 30-fold increase in theadherent layer cellularity of ECM cultures compared withcontrols (Fig. 3). A significant difference in adherent cellularitypersisted throughout the 6 wk of culture. After week 6, adecrease in total cellularity occurred in both ECMand controlcultures. However, the adherent layer cellularity of the ECMbased cultures remained significantly greater than that ofcontrols. In order to establish whether this decrease in cellularitywas due to senescence of the adherent layer or stem celldepletion, another set of cultures was recharged a second timeat week 6. These cultures were capable of another round ofproliferation which lasted -3 wk (data not shown).

Granulocytic and megakaryocytic progenitor cell prolifera-tion. ECM-based cultures sustained greater numbers of hema-topoietic progenitor cells than did control cultures (Table I).By week 3, suspension phase CFU-GM were significantlyhigher in the ECMcultures. In ECM-based adherent layersthere were 30-fold more CFU-GMthan in the adherent layerof control cultures. Megakaryocyte progenitor cells were alsosignificantly elevated in ECMadherent layers from week 3 to4 (Table I).

Granulocytic and megakaryocytic differentiation. In orderto assess the degree of hematopoietic differentiation, cytocen-trifuged preparations were stained for granulocyte morphology(Wright's stain) or for megakaryocytes (acetylcholinesterasestain) and differential counts were performed. The results areshown in Table II. Both suspension and adherent cells fromECM-based cultures showed a higher total number and higherproportion of mature granulocytes (metamyelocytes, bands,polymorphonuclears) than controls. Megakaryocytes were alsoelevated in both adherent and suspension phases of ECM-coated dishes. By week 3, ECM-based cultures contained>4,500 mature megakaryocytes (80% of which were in theadherent layer), whereas control cultures contained <100megakaryocytes (data not shown).

Erythroid proliferation and differentiation. No evidence ofsignificant terminal maturation of erythroid cells was seen onmicroscopic examination of Wright stained suspension oradherent cells. This is consistent with previous work (22, 23),which demonstrated that certain manipulations, such as addi-tion of exogenous erythropoietin, anemic mouse serum, ormechanical agitation, are required to induced terminal eryth-ropoiesis in long-term cultures.

2086 A. Campbell, M. S. Wicha, and M. Long

F

Figure 1. Photomicrographs of long-term cultures. (A, B) Phase-contrast photomicrograph of cultures at week 2. (A) Tissue culturecontrols. Note relatively sparse hematopoietic islands. X 100. (B)ECM-based cultures. Note dense confluent layer of well spread, flat-tened, adherent cells underlying large aggregations of round hemato-poietic cells. X 100. (C, D) Scanning electron microscopy of culturesat week 2. (C) Tissue culture controls. Note scant adherent layer plussmall aggregations of hematopoietic cells. X 517. (D) ECMsubstra-tum. The ECMis covered by a confluent layer of adherent cells.Associated with these layers are large reticular (Re) and blanket (Bl)

cells that are closely associated with hematopoietic cells. X 517. (E,F) Transmission electron micrographs of week 3 cultures. (E) Sec-tion cut perpendicularly to adherent cell layer. ECMis seen at thelower left hand border. Re, Reticular adventitial cell attached toECMand extending cytoplasmic processes toward a neighboring cell.Similar cytoplasmic processes are seen surrounding the hematopoieticcells in a manner reminiscent of blanket cells. X 1,650. (F) Highermagnification view of the boxed area in E, showing several layers ofcytoplasmic processes in contact with each other and covering thelower hematopoietic cells. X 16,500.

Murine Hematopoietic Cells In Vitro 2087

so Figure 2. Cumulative sus-pension cell counts. Long-o term marrow cultures were

Z

/ established as described inMethods, and weekly total

z 0 / __ nucleated suspension cells3 / --- were determined. Values0o. ~/,,represent mean cumulative

@ /,/' suspension counts of 3-5!i individual cultures. *, ma-<|0 2 3 4 5 6 trix extract; o, tissue cul-

WEEKS ture plastic.

Molecular characterization of ECMcomponents. In orderto characterize components possibly involved in the observedeffects of ECMin long-term bone marrow cultures, marrow-derived ECMwas subjected to electrophoretic separation on a5% SDS polyacrylamide gel. A silver stained gel is shown inFig. 4, and for comparison a gel of EHS-derived ECM isshown. In contrast to the EHS extracellular matrix, bonemarrow extracellular matrix contained two major proteinbands of M, 116,000 and 135,000.

Discussion

Long-term bone marrow cultures established on a substrate ofextracellular matrix show enhanced proliferation and differ-entiation compared with those established on plastic. Thesecultures are most striking for the rapid development of adensely cellular adherent layer. When these adherent layersare recharged, hematopoietic cells rapidly proliferate in theadherent cell layer, achieving greater cellularity than on plastic.Comparison of ECMwith controls shows up to a 30-foldgreater adherent cellularity on ECMthan on tissue cultureplastic. ECM-based cultures contain greater total numbers ofboth progenitor cells and their differentiated progeny. Com-parison of the suspension and adherent phases indicates that

iO.

l 10e\

W~~~~~EK

U,~1_

cutuewer havstd At wee 2 anhratrthifrne

bewe C - anxrc)adtissueculture plasticorE n(twelnevl ota nuceated

significant (P < 0.05).

Table I. Effect of ECMon Hematopoietic Progenitor Cells

Adherent phase* Suspension phase*

Tissue culture Tissue cultureWeek ECM plastic ECM plastic

Granulocyte progenitor cells per culture§

0 NT NT 1,780+329 1,274±4962 3,632±60" 109±78 496±109 408±1423 3,606±1,541" 105±10 828±2861 227±314 1,344±353 7811 626±217 1141

Megakaryocyte progenitor cells per culture**

0 NT NT 78±64 24±382 169±811 10±3 23±18 27±93 190±1641 8±3 77±96 18±44 206±2491 1±1 45±19 51

Values represent mean±SDof 3-5 replicate cultures.* Adherent phase refers to cells attached to the bottom of the long-term culturedish, either on ECMor plastic, and removed by trypsinization.t Suspension phase refers to cells found suspended in the long-term culture me-dium.§ Colony-forming units/granulocyte macrophage were assayed by plating 50,000cells from either the adherent phase or suspension phase of ECMbased or con-trol long-term cultures. 3-5 individual long-term cultures were harvested ateach time point, and for each of these 3-5 replicate CFU-GMassays weredone.1" P - 0.05 (I test).11 Values represent pooled data from I to 2 individual progenitor cell assays.** Colony-forming units-megakaryocyte were grown in semisolid agar culturesmodified as described (15) so that the cultures were biased toward the forma-tion of megakaryocyte colonies. 3-5 individual long-term cultures were har-vested at each weekly time point, and for each of these 3-5 replicate CFU-Mkassays were done.

Table II. Granulocyte Differentiationin ECMand Control Cultures

Total granulocytes per culture X 105

Adherent phase Suspension phase

Tissue culture Tissue cultureWeek ECM plastic ECM plastic

0 Immature 0.6 0.1 NT NTMature 1.4 0.1

1 Immature NT NT 4.7 7.1Mature 14.8 10.5

2 Immature 26.1 0.7 6.4 2.5Mature 24.0 0.8 14.8 1.6

3 Immature 21.2 0.3 3.0 0.6Mature 22.8 0.0 12.3 0.7

4 Immature 4.3 0.1 3.2 0.2Mature 6.1 0.0 8.2 0.3

Cultures were established as in Table I. The degree of granulocyticmaturation was determined by removing at each time point, adher-ent phase, or suspension phase cells from 3 to 5 ECM- or plastic-based long-term cultures. The cells were cytocentrifuged, Wrightstained, and 400-cell differential counts were performed. Immaturecells are defined as blasts, progranulocytes, or myelocytes, and ma-ture cells as metamyelocytes, megakaryocytes, bands, and polymor-phonuclears. NT, not tested.

2088 A. Campbell, M. S. Wicha, and M. Long

kDa205,

A B CFigure 4. SDSpolyacrylamide gel electrophoresis of ECMderivedfrom bone marrow or EHS tumor. Bone marrow and EHSextracel-lular matrix were electrophoresed on a 5% polyacrylamide gel underreducing conditions, according to Laemmli (20), and stained withsilver nitrate as described (21). The EHS tumor has previously beenshown (12) to contain the basement membrane components laminin,type IV collagen, fibronectin, and entactin. (A) High molecularweight standards: myosin (MT 205,000), fl-galactosidase (Mr 116,000),phosphorylase b (M, 97,400), bovine albumin (Mr 66,000), egg albu-min (M, 45,000), carbonic annhydrase (M, 29,000). (B) Bone marrowextracellular matrix Gdn/DTT extract. Note major protein compo-nents at Mr 116,000 and Mr 135,000. (C) EHSextracellular matrixGdn/DTT extract. Note absence of 116,000- and 135,000-M, proteinbands. kDa, molecular weight.

the ECM-induced response was primarily restricted to theadherent cell layer. Wehypothesize that ECMis involved inthe development of a stromal cell layer that promotes prolif-eration and differentiation of hematopoietic stem cells to agreater extent than stromal cell layers based on plastic. Also,we have found that the stromal cell layer induced by ECMcan be recharged a second time with fresh marrow cellsresulting in renewed hematopoietic cell proliferation.

Electrophoretic analysis of marrow-derived ECMshowsthe presence of major protein constituents which are notpresent in the basement membrane containing EHS ECM.The functional significance of these bands is currently thesubject of investigation.

A major emphasis in the study of the hematopoieticmicroenvironment in long-term cultures has been the exami-nation of stromal cells and soluble factors. Allen and Dexter(24) have recently emphasized the importance of blanket cellsand macrophagelike cells in regulating stem cell proliferation

in culture. The former appear to be flat, large, well-spreadcells that directly overlie developing hematopoietic cells. Songand Quesenberry (4) have shown that the adherent cell layeris capable of elaborating soluble factors that regulate hemato-poiesis. Little is known about the role of ECMin the hema-topoietic microenvironment. Sampath and Reddi (1 1) showedthat ECMderived from cortical bone could induce the for-mation of a bony ossicle containing fully functional hemato-poietic tissue when implanted subcutaneously in rats. Thisinductive property was lost when the ECMwas extracted with4 M guanidine, but regained again when the residue andextract were reconstituted. We previously demonstrated thatin long-term cultures, stromal cells are capable of synthesizingand depositing extracellular components such as fibronectin,laminin, and type IV collagen (10). Furthermore, the additionof cis-hydroxyproline (which specifically inhibits collagen syn-thesis) resulted in a decrease of stem cell number in thesecultures (8). Keating et al. (25) found that proteoglycans arealso a major component of the adherent layer of long-termcultures. Spooncer and co-workers (9) have shown that stim-ulation of glycosaminoglycan synthesis with #-xylosides resultsin increased total cellularity and stem cell numbers in long-term cultures. These studies emphasize the importance ofextracellular matrix in the hematopoietic microenvironment.

Our data suggest the ECMis an important component ofthe hematopoietic environment. Wehypothesize that the roleof extracellular matrix is to stimulate soluble factor productionby stromal cells. Thus, a complete understanding of thehematopoietic microenvironment must take into account itsthree major components: the ECM, cellular elements, andsoluble factors. The use of extracellular matrix in our culturesrepresents a new system for the long-term culture of hemato-poietic cells. This system provides a powerful tool to study therole of extracellular matrix in the hematopoietic microenviron-ment.

Acknowledgments

The authors wish to thank Linda Gragowski and Connie Heffher fortheir research assistance in this project.

These studies were supported in part by HL-31568 and HD-16721from the National Institutes of Health, 83-1169 from the AmericanHeart Association, BC-357 from the American Cancer Society, and agrant from the Pardee Foundation.

References

1. Dexter, T. M., M. A. S. Moore, and A. P. C. Sheridan. 1977.Maintenance of hematopoietic stem cells and production of differen-tiated progeny in allogeneic and semiallogeneic bone marrow chimerasin vitro. J. Exp. Med. 145:1612-1616.

2. Moore, M. A. S., and T. M. Dexter. 1978. Stem cell regulationin continuous hematopoietic cell culture. Transplant Proc. 10:83-90.

3. Dexter, T. M. 1982. Stromal cell associated hematopoiesis.J. Cell. Physiol. I(Suppl.):87-94.

4. Song, Z. X., and P. J. Quesenberry. 1984. Radioresistant murinemarrow stromal cells: a morphologic and functional characterization.Exp. Hematol. 12:523-533.

5. Wicha, M. S., G. Lowrie, E. Kohn, P. Bagavandoss, and T.Mahn. 1982. Extracellular matrix promotes mammaryepithelial growthand differentiation in vitro. Proc. Natl. Acad. Sci. USA. 79:3213-3217.

6. Enat, R., D. M. Jefferson, N. Ruiz-Opazo, Z. Getmaitin, L.Leinwand, and L. Reid. 1984. Hepatocyte proliferation in vitro: its

Murine Hematopoietic Cells In Vitro 2089

."ll.dA16 -g-Ab

__A -is,

97

66i--

jorg',

4 5

dependence on the use of a serum-free hormonally defined mediumand substrata of extracellular matrix. Proc. Nall. Acad. Sci. USA. 81:1411-1415.

7. Bissel, M. J., G. M. Hall, and G. Parry. 1982. How does theextracellular matrix direct gene expression? J. Theor. Biol. 99:31-68.

8. Zuckerman, K. S., R. K. Rhodes, D. D. Goodrum, U. R. Patel,B. Sparks, J. Wells, M. S. Wicha, and L. May. 1985. Inhibition ofcollagen deposition in the extracellular matrix prevents the establishmentof a stroma supportive of hematopoiesis in long-term murine bonemarrow cultures. J. Clin. Invest. 75:970-975.

9. Spooncer, E., J. T. Gallagher, F. Krisna, and T. M. Dexter.1983. Regulation of hematopoiesis in long-term bone marrow cultures.IV. Glycos-aminoglycan synthesis and the stimulation of hematopoiesisby fl-D-xylosides. J. Cell. Biol. 96:510-514.

10. Zuckerman, K. S., and M. S. Wicha. 1983. Extracellular matrixproduction by the adherent cells of long-term murine bone marrowcultures. Blood. 61:540-547.

11. Sampath, T. K., and A. M. Reddi. 1981. Dissociative extractionand reconstitution of extracellular matrix components involved inlocal bone differentiation. Proc. Natl. Acad. Sci. USA. 78:7599-7603.

12. Kleinman, M. K., M. L. McGarvey, L. A. Liotta, P. G. Robey,K. Trygvasson, and G. R. Martin. 1982. Isolation and characterizationof type IV procollagen, laminin, and heparan sulfate proteoglycanfrom the EHSsarcoma. Biochemistry. 21:6188-6193.

13. Jackson, C. W. 1973. Cholinesterase as a possible marker ofearly cells of the megakaryocytic series. Blood. 42:413-421.

14. Pelus, L. M., M. E. Broxmeyer, J. I. Korland, and M. A. S.Moore. 1979. Regulation of macrophage and granulocyte proliferation.J. Exp. Med. 150:277-292.

15. Williams, N., H. Jackson, P. Ralph, and I. Nakoinz. 1981. Cellinteractions influencing murine marrow megakaryocytes: nature of thepotentiator cell in bone marrow. Blood. 57:157-163.

16. Ralph, P., M. A. S. Moore, and K. Nilsson. 1976. Lysosymesynthesis by established human and murine histocyte lymphoma celllines. J. Exp. Med. 143:1528-1533.

17. Williams, N., R. R. Eger, H. M. Jackson, and D. J. Nelson.1982. Two factor requirement for murine megakaryocyte colonyformation. J. Cell. Physiol. 110:101-104.

18. Williams, N., H. M. Jackson, R. R. Eger, and M. W. Long.1981. The separate roles of factors in murine megakaryocyte colonyformation. In Megakaryocytes In Vitro: Biology and Precursors. B.Evatt, R. Levine, and N. William, editors. Elsevier Science PublishingCo., Inc., New York. 59-75.

19. Hayat, M. A. 1981. Fixation for Electron Microscopy. AcademicPress, Inc., New York. 321.

20. Laemmli, V. K. 1970. Cleavage of structural protein duringthe assembly of the head of bacteriophage T-4. Nature (Lond.). 227:680-685.

21. Morrissey, J. H. 1981. Silver stain for proteins in polyacrylamidegels: a modified procedure with enhanced uniform sensitivity. Anal.Biochem. 117:307-310.

22. Dexter, T. M., N. G. Testa, T. Allen, T. Rutherford, and E.Sedwick. 1981. Molecular and cell biologic aspects of erythropoiesis inlong-term bone marrow cultures. Blood. 58:699-707.

23. Eliason, J. F., T. M. Dexter, and N. G. Testa. 1982. Theregulation of hemopoiesis in long-term bone marrow cultures. III. Therole of burst forming activity. Exp. Hematol. 10:444-450.

24. Allen, T. D., and T. M. Dexter. 1984. The essential cells of thehematopoietic microenvironment. Exp. Hematol. 12:517-521.

25. Keating, A., T. Wright, M. Kinsell, and J. W. Singer. 1984.Proteoglycan synthesis by hematopoietic microenvironmental cells.Clin. Res. 320:311 (Abstr.)

2090 A. Campbell, M. S. Wicha, and M. Long