Bone Acidic Glycoprotein-75 Delineates the Extracellular Sites of Future Bone Sialoprotein

Accumulation and Apatite Nucleation in Osteoblastic Cultures1,2. Ronald J. Midura, Aimin Wang, Dinah Lovitch*, Douglas Law**, Kimerly Powell, and Jeff P. Gorski * Department of Biomedical Engineering and the Orthopaedic Research Center, Lerner Research Institute of The Cleveland Clinic and Foundation, Cleveland, OH; *Division of Biochemistry and Molecular Biology, School of Biological Sciences, and Department of Oral Biology, School of Dentistry, University of Missouri-Kansas City, Kansas City, MO, and **Division of Cell Biology and Biophysics, School of Biological Sciences, University of Missouri-Kansas City, Kansas City, MO Keywords: Bone acidic glycoprotein-75, bone sialoprotein, biomineralization, primary bone, hydroxyapatite, extracellular matrix Please address all correspondence to: Dr. Ronald J. Midura, Dept. Biomedical Engineering, ND20, The Cleveland Clinic Foundation, 9500 Euclid Ave., Cleveland, OH 44195 Phone: 216-445-3212 Fax: 216-445-4383 Email: midura@bme.ri.ccf.org

JBC Papers in Press. Published on March 5, 2004 as Manuscript M312409200

Copyright 2004 by The American Society for Biochemistry and Molecular Biology, Inc.

by guest on April 6, 2018

http://ww

w.jbc.org/

Dow

nloaded from

2

ABSTRACT

Addition of an organophosphate source to UMR osteoblastic cultures activates a mineralization

program in which BSP localizes to extracellular matrix sites where hydroxyapatite crystals are

subsequently nucleated (Wang, A., Martin, J.A., Lembke, L.A., and Midura, R.J. 2000.

Reversible Suppression of in vitro Biomineralization by Activation of Protein Kinase A. J. Biol.

Chem. 275:11082-11091). This study identifies for the first time novel extracellular spherical

structures, termed biomineralization foci (BMF), containing bone acidic glycoprotein-75 (BAG-

75), bone sialoprotein (BSP), and alkaline phosphatase that are the exclusive sites of initial

nucleation of hydroxyapatite crystals in the UMR model. Importantly, in the absence of added

phosphate, UMR cultures after reaching confluency contain two size populations of

morphologically identifiable BMF precursors enriched in BAG-75 (15-25 and 150-250 microns

in diameter). The shape and size of the smaller population are similar to structures assembled in

vitro through self-association of purified BAG-75 protein (Gorski, J.P., Kremer, E.A., Chen, Y.,

Ryan, S., Fullenkamp, C., Delviscio, J., Jensen, K., and McKee, M.D. 1997. Bone acidic

glycoprotein-75 self-associates to form macromolecular complexes in vitro and in vivo with the

potential to sequester phosphate ions. J. Cellular Biochem. 64:547-564). After organophosphate

addition, BSP accumulates within these BAG-75-containing BMF precursors, with

hydroxyapatite crystal nucleation occurring subsequently. In summary, BAG-75 is the earliest

detectable biomarker that accurately predicts the extracellular sites of de novo biomineralization

in UMR cultures. We hypothesize that BAG-75 may perform a key structural role in the

assembly of BMF precursors and the recruitment of other proteins such as alkaline phosphatase

and BSP. Furthermore, we propose a hypothetical mechanism in which BAG-75 and BSP

function actively in nucleation of apatite within BMF.

by guest on April 6, 2018

http://ww

w.jbc.org/

Dow

nloaded from

3

Introduction

The mechanisms responsible for biomineralization of mammalian bone remain

unresolved (1-8). One theory proposes that biomineralization reactions are initiated within

phospholipid membrane-delimited vesicles (matrix vesicles) released into the extracellular

matrix by osteoblasts (1,3,5,8). Another proposes that mineralization is initiated within the “hole

zone” of collagen fibrils at the mineralization front of osteoid in a process requiring

noncollagenous phosphoproteins (1,2,4). In vitro studies have shown that some of these

phosphoproteins can nucleate the formation of apatite crystals when located within collagen or

agarose gels (6,9,10). However, most prior studies of biomineralization have not seriously

considered whether their models were consistent with the kinetics or spatial patterning with

which bone mineralizes in vivo.

Several noncollagenous bone matrix proteins are believed to play an active role in the

biochemical reactions involved in bone biomineralization. One of these proteins, bone

sialoprotein (BSP), is a sulfated, phosphorylated matrix glycoprotein containing an RGD integrin

receptor-binding site (11). BSP has been proposed to function as an apatite nucleator because it

is found in the mineralizing boundaries of bone, dentin and calcifying cartilage tissues (12-17)

co-localizing with the smallest detectable foci of newly forming mineralized matrix in osteoid

(18), and is expressed prior to and during the active mineralization periods of osteoblastic

cultures that produce apatite crystals of biological proportions (19,20). Furthermore, purified

BSP nucleates a small amount of apatite in vitro (9) likely mediated by its polyglutamate

domains (10,21).

Another noncollagenous bone matrix protein, bone acidic glycoprotein-75 (BAG-75), is an

acidic glycoprotein of 75 kDa in apparent size that contains elevated contents of Asp/Glu (29%),

by guest on April 6, 2018

http://ww

w.jbc.org/

Dow

nloaded from

4

sialic acid, and covalently bound phosphate (44 res./mol) (22). Research to date indicates that

BAG-75 is restricted in its expression to actively forming primary or woven bone and dentin (23-

26). Purified BAG-75 displays a strong propensity to self-associate into supramolecular

spherical complexes (10-20 µm diameter) composed of 10-12 nm diameter microfibrils (24,27).

Electronegative complexes composed of BAG-75 can sequester millimolar quantities of

phosphate ions that are available for crystal nucleation reactions (24); BAG-75 also binds up to

135 atoms of Ca+2/molecule at saturation (28). Multimeric BAG-75 complexes exist in vivo at

sites of new primary bone formation as demonstrated by staining frozen tissue sections with a

BAG-75 aggregation-specific monoclonal antibody (24).

While conclusions from in vitro models require validation in vivo [see Gorski et al.,

companion paper (29)], culture systems are readily controlled and reproducible. The UMR 106-

01 BSP cell line constitutively expresses a mature osteoblastic phenotype and is able to deposit

ample amounts of apatite crystals within discrete focal sites over a 24 h incubation period after

addition of an organophosphate source (19,30). Several findings from our laboratory support the

physiological relevancy of the UMR culture model. Most notably, Wang et al. (30) have shown

that the biomineralization processes in both UMR and primary osteoblast cultures are similarly

regulated by parathyroid hormone treatment. Furthermore, the UMR biomineralization process

is an active metabolic process requiring continuous protein synthesis and secretion (19). Lastly,

X-ray diffraction analyses demonstrate that mineral crystals formed in mineralizing UMR

cultures are apatite and have a similar C-axis length as that of apatite crystals isolated from

pediatric bone samples (19). Altogether, our prior work indicates that UMR cultures deposit

apatite crystals at focal sites in a physiologically relevant manner.

by guest on April 6, 2018

http://ww

w.jbc.org/

Dow

nloaded from

5

BSP quantitatively accumulates within biomineralization foci (BMF) only after treatment of

UMR cultures with organophosphates (30). BSP is first detected within BMF after 4-8 h, well

before the first appearance of apatite crystals within BMF (30). Thus, in this model system, BSP

fulfills the minimal necessary temporal-spatial requirements of a protein involved in apatite

nucleation. However, it is not known how BSP is targeted quantitatively to BMF. The present

study was undertaken to determine whether mineralizing UMR cultures, like initial sites of

mineral nucleation in primary bone [see Gorski et al., companion paper (29)], accumulate BSP

and then hydroxyapatite within BAG-75 containing BMF precursors.

by guest on April 6, 2018

http://ww

w.jbc.org/

Dow

nloaded from

6

Materials and methods

Antibodies

The sources of antibodies used in the study were as follows: monoclonal anti-BSP, WV1D1

(9C5) antibody was obtained from the Developmental Studies Hybridoma Bank (Univ. of Iowa,

Iowa City, IA); anti-BAG-75 protein antibodies (#504) were provided by Jeff Gorski (23).

Monospecific polyclonal antiserum raised against rat BSP (LF-87) was obtained from Larry

Fisher (NICDR, NIH) (31). The 504 polyclonal antiserum for BAG-75 (23,24,26,27), and the

WVIDI (9C5) monoclonal antibody for BSP (30) have been shown to be monospecific in light

level immunolocalization, electron microscopic ultrastructural studies, and Western blot

analyses.

Cell culture and treatment with β-glycerophosphate

MC3T3-E1 (subclone M4) osteoblastic cells were cultured and treated with 5 mM β-

glycerophosphate (β-GP) as previously described (32,33). UMR 106-01 BSP cells were passaged

and cultured as described by Wang et al. (30). Prior to immunochemical analyses, cells were

seeded onto either 25 mm Aclar® 33 C .005 thick plastic cover slips (Allied Signal's medical

grade) in 6-well dishes or serum-coated Fisher Plus glass slides in petri dishes (both for confocal

analyses), or Permanox® tissue culture dishes (for electron microscopy analyses, Fisher

Scientific), at a density of 1000 cells/mm2 in growth medium [Eagles minimum essential

medium plus nonessential amino acids, 2 mM L-glutamine, 20 mM HEPES (pH 7.2), and 10%

fetal bovine serum]. At 64 h of incubation, the medium was replaced with fresh growth medium

for an additional 24 h (biomineralization period) supplemented with or without β-GP at a final

concentration of 7 mM. β-GP was prepared as sterile 0.7 M stocks in Nanopure water adjusted

to a final pH of 7.0; aliquots were added directly to fresh medium immediately before initiating

by guest on April 6, 2018

http://ww

w.jbc.org/

Dow

nloaded from

7

biomineralization. Unless otherwise stated, cells were fixed by immersion in cold 70% ethanol

and stored as such at 4° C until processed further.

Wang et al. (30) have previously demonstrated that UMR cultures treated with as little as

1 nM parathyroid hormone expressed high alkaline phosphatase activity levels, and a near

complete conversion of β-GP into inorganic phosphate, yet did not precipitate calcium phosphate

crystals. Chang et al. (34) also reported that spontaneous crystallization of calcium and

phosphate did not occur in UMR culture medium even when the concentrations of these ions

were elevated. Therefore, chemical precipitation is not the primary mechanism governing

mineral formation in this model system.

Culture of fetal rat calvarial cells and treatment with β-GP

As described previously (35,36), calvaria were removed from 19-day fetal rats and sequentially

digested with 0.2% collagenase/0.05% trypsin in HBSS 6 times for 20 minutes each. The

supernatants from the first two digests were discarded and the supernatants from the remaining

digests, rich in osteoblast precursor cells, were pooled for culture. Cells were plated at 2-3 x 108

cells per T75 flasks in α-minimal essential media supplemented with glutamine and nucleosides

(Mediatech, Inc.), 10% FBS, and 0.5% penicillin/streptomycin (Pen/Strep) antibiotic. Cells were

grown to confluence (3-4 days), trypsinized and then plated for experiments in chamber slides at

8000 cells /chamber. After reaching confluency, first passage cultures were re-fed every three

days with medium supplemented with 5% FBS, 0.5% Pen/Strep, 100 µg/ml ascorbic acid, and

with 5 mM β-GP to promote the formation and mineralization of bone mineralization foci.

Controls did not receive ßGP. At 7 and 17 days after reaching confluence, culture slides were

fixed in 4% paraformaldehyde for 2.5 h at room temperature and were then stored 70% ethanol

at 4°C until stained with Alizarin red S.

by guest on April 6, 2018

http://ww

w.jbc.org/

Dow

nloaded from

8

Visualization of apatite mineral using Alizarin red-S

After removal of medium, UMR cultures were briefly rinsed with TBS followed by fixation in

ice-cold 70% ethanol for 1 h. Fixed samples were washed three times with TBS, rinsed with de-

ionized water, and then stained for 10 min with either 40 mM or 0.4 mM Alizarin red-S dye, pH

4.2, at room temperature for brightfield (19) or epifluorescent microscopy (30), respectively.

Cultures were then rinsed five times with water followed by a 15 min wash with TBS to reduce

nonspecific AR-S stain. When necessary, fixed mineralized cultures were decalcified using one

of three methods: (1) treatment with Tris-buffered saline (pH 8.0) containing 50 mM EDTA

overnight at 4° C, (2) treatment with 10% trichloroacetic acid for 10 min. at 4° C, or (3)

treatment with 0.1 M HCl in 70% ethanol for 30 min. at 4° C (30).

Fluorescent immunolabeling

Fixed cells on glass slides (FisherPlus slides, Fisher Scientific Co.), chamber slides (Lab-Tek,

Inc.), or Aclar® coverslips were briefly rehydrated in Tris-buffered saline, pH 7.5, and then

blocked by incubation in Tris-buffered saline, pH 7.5 containing 1% (w/v) bovine serum albumin

(Sigma Immunologicals) for 60 min. Slides were then rinsed five times with Tris-buffered

saline, pH 7.5 and incubated for 1-3 h with primary antibodies diluted in the above blocking

buffer (typically 1:100 to 1:200 dilutions). Slides were then washed five times as before and

treated with either anti-mouse or anti-rabbit IgG Fc-domain secondary antibodies conjugated to

FITC or TRITC fluorochromes diluted 1:1000 in blocking buffer (Jackson Immunoresearch and

Molecular Probes, Inc.). After washing off unbound secondary antibody, specimens were

stained with 1 µg/ml 4',6-diamidino-2-phenylindole (DAPI), mounted with VectaShield (Vector

Labs.), and slides cover slipped prior to imaging with a Leica TCS-SP AOBS (True Confocal

Scanner-SpectroPhotometer Acousto-Optic Beam Splitter) equipped with 4 lasers (Krypton,

by guest on April 6, 2018

http://ww

w.jbc.org/

Dow

nloaded from

9

Argon, UV, and HeNe), or a Nikon Microphot Fx epifluorescence microscope outfitted with a

SPOT-RT CCD digital camera (Diagnostic Instruments). Digital images were post-processed

with Adobe Photoshop (v 6.0).

Cell death assay

Apoptotic cells were detected in UMR cultures using a cell death detection kit (Boehringer

Mannheim). Cultures were fixed in 2% paraformaldehyde in phosphate buffered saline and

processed according to the manufacturer's instructions. Samples were incubated with a Terminal

deoxynucleotidyl transferase dUTP Nick End Labeling (TUNEL) reaction mixture that adds

fluorescein-conjugated dUTP to the free 3'-OH groups in single- and double-strained DNA

fragments within apoptotic nuclei. After washing, the specimens were stained with alizarin red S

(see above) and then 1 µg/ml DAPI, followed by mounting in VectaShield. All fluorochromes

were detected using confocal microscopy. TUNEL specificity was determined by omitting the

terminal deoxynucleotidyl transferase necessary for adding dUTP to the free 3'-OH ends of DNA

strands.

Electron microscopy

For conventional electron microscopy, cultures were fixed for 24 hours with 2% glutaraldehyde

in 150 mM sodium cacodylate, pH 7.2, at 4° C. Samples were post-fixed in 1% OsO4,

dehydrated, embedded, sectioned and stained according to published protocols (19,30). For

ultrastructural immunocytochemistry, tissue samples were fixed with 2% formaldehyde in PBS,

overnight at 4° C. Samples were cryo-protected, frozen in liquid nitrogen and sectioned as

described previously (37), using a method adapted from that of Tokuyasu (38). Frozen sections

80-100 nm thick were processed for indirect immunolabeling using secondary antibodies

coupled to 6 nm or 12 nm colloidal gold (Jackson Immunoresearch, West Grove, PA), under

by guest on April 6, 2018

http://ww

w.jbc.org/

Dow

nloaded from

10

conditions designed to minimize cross-reactivity and non-specific binding of antibodies (37). In

some cases, sections were decalcified by floating them on a drop of ice-cold 10% trichloroacetic

acid for 60 min. prior to immunolabeling. Sections were examined using a JEOL 1200EX

transmission electron microscope at an accelerating voltage of 100 kV.

by guest on April 6, 2018

http://ww

w.jbc.org/

Dow

nloaded from

11

Results

Mineralizing UMR cultures produce focal deposits of apatite that also contain BSP and

alkaline phosphatase activity

When given an organophosphate stimulant like β-GP, UMR cultures deposit an apatite mineral

phase by a utilization of phosphate ions released from β-GP by alkaline phosphatase (19,30).

Figure 1A demonstrates the focal nature of the biomineralization process in UMR cultures after

β-GP exposure, and also reveals the close spatial relationship between BSP and apatite deposits

in this biomineralization model system. Figure 1B shows that these focal areas of apatite in

UMR cultures also stain positive for alkaline phosphatase activity. BSP has been reported to

accumulate quantitatively in these focal areas prior to the detection of an apatite phase, thereby

meeting the necessary temporal-spatial requirements of a matrix protein involved in apatite

formation (30). However, it is not clear how BSP is targeted to these biomineralization foci

(BMF) prior to mineral deposition.

BMF are located within the extracellular matrix between UMR cells

Figure 1C shows X-Y and X-Z views from a three-dimensional stack of confocal images each

representing an optical slice of 0.5 µm depth. BMF are identified by their strong immuno-

staining for BSP, while nuclei are identified by their strong staining with DAPI. These 3D

images demonstrate that BMF are located in the extracellular matrix space between cells, and

that nearly all BMF are devoid of staining for the presence of double-stranded DNA. Further,

BMF neither stain positively with the TUNEL reaction nor exhibit evidence of DAPI-stained

apoptotic bodies (Fig. 1D). Taken together, the absence of DNA remnants, of TUNEL-positive

signal, and of apoptotic bodies indicates that BMF are not apoptotic or necrotic cell debris. These

by guest on April 6, 2018

http://ww

w.jbc.org/

Dow

nloaded from

12

results strengthen the conclusions from our previous studies (19,30) that the biomineralization

process in UMR cultures is not dystrophic.

The 3D confocal dataset in Figure 1C shows that an average BMF is spherical in shape

with typical dimensions of 10-20 µm in X-Y and 10-15 µm in the X-Z viewing planes. Notably,

not all negatively charged macromolecules secreted into the extracellular space are deposited in

BMF as indicated by the lack of overlap between the Alizarin red/BSP signals and hyaluronan

staining (Fig. 1C). Thus, the incorporation of BSP into BMF is unlikely to result from

nonspecific electrostatic interactions. Rather, the data suggest that the strict BSP deposition in

BMF is due to more specific interactions.

Bone Acidic Glycoprotein-75 is a biomarker for BMF precursors prior to active

mineralization

Figures 2 and 3 demonstrate for the first time that spherical BMF precursors are recognizable as

BAG-75 containing supramolecular complexes prior to the addition of β-GP into UMR cultures.

Briefly, Figures 2B and 2D depict the representative brightfield appearance of UMR monolayer

cultures at 64 h of incubation, just after confluence has been reached, but prior to

organophosphate addition (“zero-time”). At this time, the cultures consist of a confluent

monolayer of UMR cells containing two size populations of refractile, roughly spherical BMF

precursors (15-25 µm and 150-250 µm in diameter), which project upward from the cell layer at

apparently random intervals. After immunostaining, BAG-75 was detected in both size

populations of BMF precursors prior to initiating mineralization (Figs. 2A and 2C). The

immunofluorescent signal for BAG-75 within the larger BMF precursors consisted of several

bright focal areas each 10-20 µm in diameter embedded within a slightly lower level of diffuse

by guest on April 6, 2018

http://ww

w.jbc.org/

Dow

nloaded from

13

staining (Fig. 2C). This appearance suggests that the larger sized BMF precursors may represent

a cluster of smaller ones.

Both sizes of BMF precursors were able to nucleate calcium phosphate crystals when

UMR cultures were incubated for an additional 24 h with medium containing β-GP (i.e., 64 h to

88 h after plating). This is evident by the strong staining obtained with alizarin red-S dye

binding to large (Figs. 2E, 2F, and 5E) and small BMF (Fig. 5E). Several points can be made

regarding the relative staining patterns observed when BAG-75 immunofluorescence and alizarin

red-S-stained mineral crystals are imaged together. At low magnification, the stronger alizarin

red-S signal displays a punctate appearance comprised of numerous 10-20 µm diameter round

deposits within the interior domain of the larger BMF (Fig. 2E). In contrast, the BAG-75

immunofluorescence signal appears as a layer surrounding the alizarin red-S-stained deposits

and, internally, in junctional areas separating individual alizarin red-S-stained deposits (Fig. 2F).

In these undecalcified cultures, the regions of strongest alizarin red-S staining are seemingly

devoid of apparent BAG-75 staining, while BAG-75 staining is readily visible in the outer layer

surrounding the mineral deposits and within the internal areas physically separating alizarin red-

S-stained round deposits. This apparent lack of BAG-75 immunogenicity deeper within these

larger BMF may be due to the coverage of antigenic sites by the hydroxyapatite crystalline

phase. This point is documented upon decalcification of mineralized cultures as seen in the next

section.

Thus, in UMR cultures, BAG-75 is a biomarker both for BMF precursors prior to

initiating mineralization, and for mature BMF containing mineral crystals regardless of size.

Within the larger BAG-75-enriched mature BMF, the apatite mineral deposits seem to be

organized into small domains (10-20 µm) having sizes similar to those of the smaller BMF.

by guest on April 6, 2018

http://ww

w.jbc.org/

Dow

nloaded from

14

Again, this appearance suggests that the larger sized BMF precursors may represent a cluster of

smaller ones.

BAG-75-enriched focal deposits represent the initial sites of mineralization in primary

calvarial osteoblastic cultures

The presence of BMF in the UMR osteoblastic culture system raises the possibility that

structures like these may also exist in primary osteoblastic cultures. Figures 3D and 3E indicate

that initial focal mineral deposits in fetal rat calvarial osteoblastic cell cultures have structural

features and dimensions similar to those of UMR BMF. When partially decalcified by EDTA,

these focal sites stain positively for the presence of BAG-75 (Fig. 3F and 3G). Two other

characteristics of mineralization in primary osteoblastic cultures are consistent with that

occurring in BMF in UMR cultures (Figure 2): (a) the focal area for BAG-75 immunostaining is

larger than the focal area measurement for alizarin red-S staining (compare Figs. 3D with 3F,

and Figs. 3E with 3G), and (b) focal BAG-75 staining temporally precedes that of apatite

deposition (for example, compare Figs. 3A and 3B with Figs. 3D and 3F). Additional support

for the concept that BMF are the sites of mineral nucleation in many osteoblastic model systems

is shown in Figure 3H-J. Non-transformed osteoblastic cell line MC3T3-E1 exhibits focal

alizarin red-S stained mineral deposits similar to those observed in UMR cultures (Figs. 3H and

3I) that also stain positively for the presence of BAG-75 after partial decalcification using EDTA

(Fig. 3J).

Decalcification reveals that the BAG-75 content increases within BMF during active

mineralization

Since the results above suggest that BAG-75 epitopes might be partially blocked by

hydroxyapatite mineral deposits, it was important to show that this antigenicity could be

by guest on April 6, 2018

http://ww

w.jbc.org/

Dow

nloaded from

15

recovered upon decalcification. To this end, replicate UMR cultures at 64 h of incubation were

treated with or without β-GP supplementation for 24 h. Mineralized cultures were fixed and then

decalcified using EDTA chelation (Figure 4) or other decalcification methods (30). Atomic

absorption measurement of elemental calcium determined that these treatments released more

than 98% of the original calcium content of fully mineralized cultures (30). In the absence of

confounding mineral crystals, a large increase in BAG-75 signal is observed indicating that the

BAG-75 content of BMF is substantially increased over the 24 h mineralization period (Fig. 4).

This accumulation of BAG-75 occurs only after addition of β-GP to the UMR cultures, as

control cultures not exposed to β-GP over the same time frame do not exhibit this increase in

BAG-75 immunofluorescence (data not shown). Thus, as shown previously for BSP (30), the

content of BAG-75 epitope within BMF increases during active mineralization.

When decalcification solutions were recovered and processed for Western blotting with

anti-BAG-75 (data not shown) or BSP antibodies (30), very little BAG-75 or BSP protein was

detectable in these fractions. These findings indicate that most of the BAG-75 and BSP epitopes

are not solubilized by decalcification of mineralized cultures. This suggests that most of the

BAG-75 and BSP accumulated within BMF is not the result of adsorption to pre-existing

hydroxyapatite crystals. A similar conclusion was made with respect to the BAG-75 present at

sites of initial mineral nucleation in primary bone [see Gorski et al., companion paper (29)].

BSP co-localizes to large and small BMF containing BAG-75

Without β-GP, BSP is predominantly secreted into the media by UMR cells, though after β-GP is

added, BSP rapidly accumulates within supramolecular complexes that subsequently mineralize

(30). We demonstrate in Figure 5 that these extracellular sites in mineralizing UMR cultures

seem to be the same as BMF precursors enriched in BAG-75 and recognizable in zero-time

by guest on April 6, 2018

http://ww

w.jbc.org/

Dow

nloaded from

16

cultures without β-GP (Figs. 2 and 4). In fully mineralized cultures, BSP was localized within

both large (Fig. 5G,H) and small (Fig. 5C,H) mature BMF, which also stained strongly for BAG-

75 (Figs. 4 and 5A). Note that the shape and boundaries of the areas of BSP staining

corresponded exactly with those delimited by BAG-75 immunostaining (Fig. 5D, overlay

image). The physical characteristics and BAG-75 content of the small BMF precursors seen in

Figs. 2 and 4A-D identify them as similar structures. Interestingly, although many BMF

exhibited a strong immunofluorescence signal for both proteins (Fig. 5D, arrows), some small

BMF structures displayed noticeably lower levels of co-staining. These findings could be due to

differences in mineral content, which may lead to blocking of antigens.

Altogether, these results demonstrate that BSP begins to accumulate in BAG-75-enriched

BMF after the addition of β-GP, but prior to nucleation of hydroxyapatite crystals (30).

Ultrastructural analysis of BMF during active mineralization of UMR cultures

The X-Z view from the confocal imaging shown in Figure 1C indicates that small BMF are

roughly spherical in shape with a diameter range of 10-20 µm. The X-Z ultrastructural image of

a BMF in a mineralizing UMR culture (Fig. 6A) indicates that these confocal estimates are quite

accurate. The inset scanning EM image in Figure 6A also shows that the overall dimension and

shape of supramolecular complexes formed by purified BAG-75 protein are similar to those of

small BMF observed in mineralized UMR cultures.

BMF are filled with both thin fibrillar and granular substances (Fig. 6B). In addition,

BMF contain what appear as translucent round structures of 300-800 nm diameter, and electron

opaque round structures of 50-200 nm in diameter (Figs 6, A-C). BMF did not, however, exhibit

evidence of opaque, rounded structures having a classic appearance of condensed chromatin or

apoptotic bodies (39,40). This finding is consistent with DAPI- and TUNEL-staining analyses

by guest on April 6, 2018

http://ww

w.jbc.org/

Dow

nloaded from

17

shown in Figures 1C and 1D, respectively. A comparison of BMF during mineralization (Fig.

6C) as compared to unmineralized controls (Fig. 6D) indicates that the only visible differences

are that the former are completely filled with fibrillar, granular and vesicular structures with little

void space (Fig. 6C, inset), whereas the latter exhibit more void space and are less granular in

appearance (Fig. 6D, inset).

Figures 6B and 7C (arrowhead) show clearly that there are two distinct size classes of

spherical shaped structures within mineralized BMF. One population, ranging in diameters of 50-

200 nm, has a highly opaque core with a discernable 6-8 nm thick trilaminar-delimiting

boundary (i.e., “dark-light-dark”) consistent with the ultrastructural characteristics of a

phospholipid membrane (Fig. 7C, arrowhead). These smaller sized structures resemble the

ultrastructural appearance of matrix vesicles (1,3,5,8). Another larger sized population (Figs. 6B

asterisks, and 7C), ranging in diameters from 300-800 nm, has a relatively translucent core and a

10-20 nm thick boundary surrounding the entire structure. Within mineralizing BMF, both the

small and large spherical structures appear to come in close contact with each other at their

surfaces (Figs. 6B, inset and Fig. 7C).

Localization of BAG-75 and BSP within BMF at the ultrastructural level

Figures 2E and 2F demonstrate that the inner core of the BMF contains more apatite mineral than

its outer boundaries. Within the inner core of BMF, the immunogold particles indicating the

locations of BSP were generally distributed on 10-nm thin-filament fibrils (Fig. 7A, arrows).

Towards the outer surface of BMF, BSP immunogold particles were generally associated with

larger vesicles, but not the smaller ones (Figs. 7B and 7C, arrows versus arrowheads,

respectively). Close inspection of the gold-particle staining pattern on these larger structures

gives the impression that BSP epitopes are on the surface. This is particularly evident for the

by guest on April 6, 2018

http://ww

w.jbc.org/

Dow

nloaded from

18

structure shown in Figure 7C (arrow) where gold particles were only observed in a location that

appeared to cut tangentially across the vesicle surface. Non-immune antibody control sections

typically contained a few immunogold particles, usually present in groups of one or two, but

never in clusters as large as those shown here (see Figure 3C in this paper, and Figure 7E in the

Gorski et al. companion paper for images of non-immune antibody staining).

Double staining for BAG-75 and BSP using different sized immunogold particle

antibodies (12 nm and 6 nm, respectively) revealed that these two bone matrix proteins are co-

localized in the outer boundary region of BMF (Fig. 8). BAG-75 epitopes were enriched in the

outermost layer while BSP epitopes were underneath this layer within submicron spatial

resolution of each other. What is interesting about this outer region of BMF is that it represents

the boundary between unmineralized and mineralized matrix areas (Figs. 2E and 2F). BAG-75

immunogold particles were difficult to detect within the core of BMF possibly reflecting the

effect of protein-protein interactions on epitope accessibility. by guest on April 6, 2018

http://ww

w.jbc.org/

Dow

nloaded from

19

Discussion

This study has revealed that BAG-75 is expressed very early during the biomineralization

process of UMR 106-01 BSP osteoblastic cultures. Furthermore, this study provides evidence

that focal extracellular deposition sites enriched in BAG-75 accurately predict the location of

subsequent active mineral nucleation in mineralizing osteoblastic cultures. This finding may be

a general feature of all osteoblastic cultures undergoing in vitro biomineralization since UMR,

primary fetal calvarial, and MC3T3-E1 osteoblastic model systems share this property. While

sharing a common enrichment in BAG-75, mineralized BMF sites in these models display some

structural differences in overall size (smaller in MC3T3-E1 than in UMR or FRC cultures) and

shape (more 3-dimensional in UMR than in MC3T3-E1 and FRC cultures). We hypothesize that

relative differences in secretion rates of individual BMF components (vesicles, BSP, BAG-75) or

in the local distribution of cells committed to osteogenic differentiation may contribute to such

distinctions.

BAG-75 is a biomarker for locating the future positions of BMF, and we refer to these

BAG-75-enriched areas prior to mineral formation as BMF precursors. Two size populations of

these spherical supramolecular extracellular complexes were recognized morphologically and by

virtue of their enrichment in BAG-75. Shortly after the addition of β-GP, an inducer of

mineralization, UMR cultures begin to quantitatively accumulate BSP within these BAG-75-

enriched BMF precursors (30). A few hours later, these BAG-75- and BSP-enriched BMF then

deposit hydroxyapatite within, thereby converting them into mature BMF. Regardless of mineral

content, BAG-75 is consistently present within the outer layer of mineralized BMF suggesting a

role in defining the dimensions of the forming mineralized matrix. A similar expression pattern

by guest on April 6, 2018

http://ww

w.jbc.org/

Dow

nloaded from

20

of BAG-75 in developing bone implies it participates in setting the boundary limits of new

mineralized trabeculae [see Gorski et al., companion paper (29)].

As the earliest recognized biomarker of BMF, we hypothesize that BAG-75 serves two

roles, one as a structural scaffold and the other as a source of phosphate ions for apatite

nucleation. With regard to its first function, we propose that BAG-75 plays a key role in the

initial assembly of the structural scaffold of BMF precursors. This is supported by evidence that

purified BAG-75 protein alone is able to self-assemble into thin linear strands of up to 1 micron

in length (24,27,41). These BAG-75 microfibrils can coalesce to form supramolecular spherical

structures of up to 10-15 microns in diameter (Fig. 5A, inset), which have the capability to retain

large quantities of phosphate ions (24). These structures are remarkably similar in shape and

dimension to the smaller BAG-75-enriched BMF identified here in UMR cultures.

We further rationalize that such a BAG-75 framework could provide key secondary

binding sites to recruit other proteins, such as BSP or alkaline phosphatase, required for

subsequent nucleation of hydroxyapatite. Despite excellent morphological descriptions of

membranous bone formation during embryogenesis (42), few matrix protein biomarkers have

been identified which uniquely define future sites of osteogenesis. Zhu et al. (43) found that

BSP message was weakly expressed in the center of mesenchymal cartilaginous condensations at

14 days of gestation in the mouse. In the current study, we demonstrate that BAG-75

accumulation in the extracellular matrix precedes the deposition of even BSP, and represents one

of the earliest matrix protein markers for sites of biomineralization.

Alkaline phosphatase also appears to be expressed very early during osteogenesis (44-

46), and an elevated expression of alkaline phosphatase has been used to define committed

proliferating osteogenic cells (14). As described in our BMF model of mineralization below,

by guest on April 6, 2018

http://ww

w.jbc.org/

Dow

nloaded from

21

alkaline phosphatase is an important component of matrix vesicles (47,48), where it is believed

to produce inorganic phosphate ions from complex phosphate sources to feed the process of

crystal nucleation. Addition of alkaline phosphatase inhibitors such as levamisole blocks UMR

biomineralization reactions initiated by β-GP (19). With regard to its second functional role in

UMR biomineralization, we hypothesize that the de-phosphorylation of BAG-75 protein itself

could also provide a potential source of phosphate ions for apatite nucleation reactions since this

protein contains an estimated 44-phosphoryl residues/mole (22).

Debate continues regarding the underlying mechanism(s) of vertebrate biomineralization.

One view proposes that phospholipid membrane-delimited vesicles (matrix vesicles) released

into the extracellular matrix by osteoblasts mediate biomineralization reactions (1,3,5,8,49).

They are reported to exist in nearly all mineralizing tissues, but are best described in calcifying

cartilage. They are small (50-200 nm diameter) vesicles containing high levels of

phosphatidylserine and cardiolipin (50,51). They contain several enzymes including alkaline

phosphatase and pyrophosphatase (52), and have cytoskeletal (53) and matrix proteins (54)

associated with their inner and outer membrane surfaces, respectively. Another

biomineralization theory proposes that noncollagenous protein structures assembled within the

type I collagen-rich, extracellular matrix of osteoid directly mediates the heterogeneous

nucleation of apatite crystals (2,7,55). Based on the presence of polyacidic domains and a

capacity to nucleate hydroxyapatite in vitro, BSP has been proposed to act as such a nucleator in

bone and cartilage (21).

Yet another biomineralization theory proposes that calcified spherically shaped structures

with a size range of 300-600 nm in diameter, termed crystal ghost aggregates, nucleate the initial

apatite crystals within osteoid and then subsequently seed a heterogeneous nucleation reaction

by guest on April 6, 2018

http://ww

w.jbc.org/

Dow

nloaded from

22

within the fibrillar collagen matrix (3,4,56,57-59). These mineralized structures in bone tissue

stain positive with acridine orange or ruthenium red suggesting the presence of sulfate groups

(4,60). This staining was originally interpreted as evidence for sulfated proteoglycans in these

structures (61). However, it is possible that a sulfated glycoprotein could also contribute to these

sulfate groups. BSP has been shown to co-localize with crystal ghost aggregates in osteoid (18)

and in osteoblastic cultures (20,30); it also contains several sulfate groups bound to

oligosaccharides and tyrosine residues (62,63). Interestingly, BSP has not yet been reported to

be a component of matrix vesicles (1,47,49,64). Our findings indicate that BSP is associated

with a population of large vesicle-like structures, but not with the smaller matrix vesicles. In this

way, BSP may define a biochemically distinct population of vesicles, which are larger than

matrix vesicles. Since apatite mineral forms in association with these “bioapatite vesicles” (65),

we hypothesize that these larger vesicles may be related to, and likely an early structural stage of,

the fully mineralized BSP-enriched particles referred to as crystal ghost aggregates (3,4).

It is suggested that the large size, amorphous character, and prior lack of an authentic

biomarker for BMF precursor and mature BMF complexes may have prevented their routine

detection with ultrastructural methods, which tends to emphasize discrete electron-opaque

structures of uniform appearance (e.g., matrix vesicles). Despite this limitation, similar

spherical, interfibrillar regions containing crystal ghost aggregates enriched in BSP and an

amorphous extracellular matrix have been identified in vitro (66,67) and in vivo (68) as sites of

subsequent initial mineral nucleation. In addition, Irie et al. (67) have noted the apparent

similarity between these nucleation sites in vitro and those present in membranous or primary

bone. The above-mentioned studies suggest that BMF-like structures have been observed before,

but seem to have been considered obscure entities lacking a clear context with respect to

by guest on April 6, 2018

http://ww

w.jbc.org/

Dow

nloaded from

23

mineralization reactions. It is our contention that only with data such as that provided by our

current study can one properly interpret the significance of these structures to the

biomineralization process.

We propose a 5-step biomineralization model (Fig. 9) in which BMF, typically 15-25 µm

in diameter, serve as a supramolecular domain that facilitates the initial sequestration of calcium

and phosphate ions in an extracellular matrix environment. Step 1 involves the determination of

the boundaries, shape, and size of future mineralized matrix areas by BAG-75 expressing cells.

Step 2 involves the assembly of extracellular BMF precursor structures via BAG-75 self-

assembly reactions. Step 3 involves the accumulation of matrix vesicles and bioapatite vesicles

within the BMF precursor matrix. We envision that the large amount of membrane surface area

within BMF packed with vesicles could facilitate heterogeneous nucleation reactions through

adsorption and concentration of required proteins and ions. Upon the introduction of an

organophosphate stimulant, Step 4 involves the recruitment of required accessory proteins such

as BSP, and the continued accumulation of BAG-75 into the BMF. We suggest that BMF-bound

BSP would function to sequester calcium ions in proximity to copious amounts of phosphate ions

associated with BAG-75. Together these events could lead to the supersaturation conditions

necessary to nucleate hydroxyapatite crystals on the vesicle surfaces within BMF. Finally, Step

5 involves the growth of these seed crystals and the formation of mineralized aggregates.

Continued self-assembly of BAG-75 at the outer boundaries could foster new matrix growth

within individual BMF. Future studies will investigate potential protein-protein interactions of

monomeric and oligomeric BAG-75 complexes with BSP. In addition, identification, isolation

and characterization of the structure and function of BMF precursors are now feasible by

following their BAG-75 biomarker.

by guest on April 6, 2018

http://ww

w.jbc.org/

Dow

nloaded from

24

Acknowledgements

RJM wishes to acknowledge the support of grants from the NIH (AR-45171) and The Lerner

Research Institute of The Cleveland Clinic Foundation. JPG wishes to acknowledge the support

of grants from the NIH (DE-14619 and DE-11197) and University of Missouri Research Board.

The authors acknowledge the expert assistance with confocal imaging by Dr. Judith Drazba

(Imaging Facility at The Lerner Research Institute), and with electron microscopy by Jessica

Kueker (Univ. of Missouri-Kansas City). We thank Dr. Mark McKee, McGill University, and

Dr. Sarah Dallas, Univ. of Missouri School of Dentistry for providing the MC3T3-E1 and fetal

rat calvarial cell cultures, respectively, and Dr. Bjorn R. Olsen, Harvard University for his

careful critique of the manuscript prior to submission.

Footnotes

1 Abbreviations: β-GP, beta-glycerophosphate; BMF, biomineralization foci; BSP, bone

sialoprotein; BAG-75, bone acidic glycoproteins-75; FITC, fluorescein isothiocyanate; TRITC,

Texas red isothiocyanate; TUNEL, Terminal deoxynucleotidyl transferase dUTP Nick End

Labeling; DAPI, 4',6-diamidino-2-phenylindole; EM, electron microscopy; bioapatite vesicles,

300-600 nm diameter translucent vesicular structures within BMF; matrix vesicles, 50-200 nm

diameter opaque vesicular structures exhibiting a trilaminar membrane and accumulating within

BMF, and FRC, fetal rat calvarial.

2 Presented in preliminary form at the 25th Annual Meeting of the American Society for Bone

and Mineral Research, Minneapolis, MN, September 2003.

by guest on April 6, 2018

http://ww

w.jbc.org/

Dow

nloaded from

25

References 1. Anderson, H.C. 1995. Molecular biology of matrix vesicles. Clin. Orthop. Relat. Res. 314:266-

280.

2. Bonucci, E. 1992a. Role of collagen fibrils in calcification. In: Calcification in Biological Systems, Bonucci, E. (ed), CRC Press, Boca Raton, FL, pp. 19-39.

3. Bonucci, E. 1992b. Comments on the ultrastructural morphology of the calcification process: an attempt to reconcile matrix vesicles, collagen fibrils, and crystal ghosts. Bone & Mineral 17:219-222.

4. Bonucci, E. 1995. Ultrastructural organic-inorganic relationships in calcified tissues: cartilage and bone versus enamel. Connect. Tissue Res. 33:157-162.

5. Boskey, A.L. 1992. Mineral-matrix interactions in bone and cartilage. Clin. Orthop. Relat. Res. 281:244-274.

6. Boskey, A.L. 1995. Osteopontin and related phosphorylated sialoproteins: effects on mineralization. Ann. New York Acad. Sci. 760:249-256.

7. Glimcher, M.J. 1989. Mechanism of calcification: role of collagen fibrils and collagen-phosphoprotein complexes in vitro and in vivo. Anat. Rec. 224:139-153.

8. Wuthier, R.E. 1989. Mechanism of de novo mineral formation by matrix vesicles. Conn. Tissue Res. 22:27-33.

9. Hunter, G.K. and Goldberg, H.A. 1993. Nucleation of hydroxyapatite by bone sialoprotein. Proc. Natl. Acad. Sci. USA 90:8562-8565.

10. Hunter, G.K. and Goldberg, H.A. 1994. Modulation of crystal formation by bone phosphoproteins: role of glutamic acid-rich sequences in the nucleation of hydroxyapatite by bone sialoprotein. Biochem. J. 302:175-179.

11. Midura, R.J., and Hascall, V.C. 1996. Bone sialoprotein - a mucin in disguise? Glycobiology 6:677-681.

12. Chen, J. Shapiro, H.S., Wrana, J.L., Reimers, S., Heersche, J.N.M., and Sodek, J. 1991. Localization of bone sialoprotein (BSP) expression to sites of mineralized tissue formation in fetal rat tissues by in situ hybridization. Matrix 11:133-143.

13. Chen, J., Shapiro, H.S., and Sodek, J. 1992a. Developmental expression of bone sialoprotein mRNA in rat mineralized connective tissues. J. Bone Mineral Res. 7:987-997.

14. Ikeda, T., Nomura, S., Yamaguchi, A., Suda, T., and Yoshiki, S. 1992. In situ hybridization of bone matrix proteins in undecalcified adult rat bone sections. J. Histochem. Cytochem. 40:1079-1088.

15. Ingram, R.T., Clarke, B.L., Fisher, L.W., and Fitzpatrick, L.A. 1993. Distribution of noncollagenous proteins in the matrix of adult human bone: evidence of anatomic and functional heterogeneity. J. Bone Mineral Res. 8:1019-1029.

16. Hultenby, K., Reinholt, F.P., Norgard, M., Oldberg, Å., Wendel, M., and Heinegård, D. 1994. Distribution and synthesis of bone sialoprotein in metaphyseal bone of young rats show a distinctly different pattern from that of osteopontin. Eur. J. Cell Biol. 63:230-239.

by guest on April 6, 2018

http://ww

w.jbc.org/

Dow

nloaded from

26

17. Shen, Z., Heinegard, D., and Sommarin, Y. 1995. Distribution and expression of cartilage oligomeric matrix protein and bone sialoprotein show marked changes during rat femoral head development. Matrix Biol. 14:773-781.

18. Bianco, P., Riminucci, M., Silvestrini, G., Bonucci, E., Termine, J.D., Fisher, L.W., and Gehron Robey, P. 1993. Localization of bone sialoprotein (BSP) to golgi and post-golgi secretory structures in osteoblasts and to discrete sites in early bone matrix. J. Histochem. Cytochem. 41:193-203.

19. Stanford, C.M., Jacobson, P.A., Eanes, E.D., Lembke, L.A., and Midura, R.J. 1995a. Rapidly forming apatitic mineral in an osteoblastic cell line (UMR 106-01 BSP). J. Biol. Chem. 270:9420-9428.

20. McQuillan, D.J., Richardson, M.D., and Bateman, J.F. 1995. Matrix deposition by a calcifying human osteogenic sarcoma cell line (SAOS-2). Bone 16:415-426.

21. Tye, C.E., Rattray, K.R., Warner, K.J., Gordon, J.A., Sodek, J., Hunter, G.K., and Goldberg, H.A. 2003. Delineation of the hydroxyapatite-nucleating domains of bone sialoprotein. J. Biol. Chem. 278:7949-7955.

22. Gorski, J.P., and Shimizu, K. 1988. Isolation of a new phosphorylated glycoprotein from the mineralized phase of bone that exhibits limited homology to adhesive protein osteopontin. J. Biol. Chem. 263:15938-15945.

23. Gorski, J.P., Griffin, D., Dudley, G., Stanford, C., Thomas, R., Huang, C., Lai, E., Karr, B., and Solursh, M. 1990. Bone acidic glycoprotein-75 is a major synthetic product of osteoblastic cells and localized as 75- and/or 50-kDa forms in mineralized phases of bone and growth plate and in serum. J. Biol. Chem. 265:14956-14963.

24. Gorski, J.P., Kremer, E.A., Chen, Y., Ryan, S., Fullenkamp, C., Delviscio, J., Jensen, K., and McKee, M.D. 1997. Bone acidic glycoprotein-75 self-associates to form macromolecular complexes in vitro and in vivo with the potential to sequester phosphate ions. J. Cellular Biochem.. 64:547-564.

25. Gorski, J.P. 1998. Is all bone the same? Distinctive distributions and properties of non-collagenous matrix proteins in lamellar vs. woven bone imply the existence of different underlying osteogenic mechanisms. Crit. Rev. Oral Biol. Med. 9:201-223.

26. Qin, C., Brunn, J.C., Jones, J., George, A., Ramachandranand, A., Gorski, J.P., and Butler, W.T. 2001. A comparative study of sialic acid-rich proteins in bone and dentin. Eur. J. Oral Sci. 109:133-141.

27. Gorski J.P., Kremer E.A., and Chen Y. 1996. Bone acidic glycoprotein-75 self-associates to form large macromolecular complexes. Connect. Tissue Res. 35:137-43.

28. Chen, Y., Bhajanit, B.S., and Gorski, J.P. 1992b. Calcium and collagen binding properties of osteopontin, bone sialoprotein, and bone acidic glycoprotein-75 from bone. J. Biol. Chem. 267:24871-24878.

29. Gorski, J.P., Wang, A., Lovitch, D., Law, D., Powell, K., and Midura, R.J. 2004. Extracellular bone acidic glycoprotein-75 defines condensed mesenchyme regions to be mineralized and localizes with bone sialoprotein during intramembranous bone formation. J. Biol. Chem. 279: in press.

by guest on April 6, 2018

http://ww

w.jbc.org/

Dow

nloaded from

27

30. Wang, A., Martin, J.A., Lembke, L.A., and Midura, R.J. 2000. Reversible Suppression of in vitro Biomineralization by activation of protein kinase A. J. Biol. Chem. 275:11082-11091.

31. Fisher, L.W., Stubbs III, J.T., and Young, M.F. 1995. Antisera and cDNA probes to human and certain animal model bone matrix noncollagenous proteins. Acta Orthop. Scand. (suppl. 66) 66:61-65.

32. Stanford CM. Morcuende JA. Brand RA. 1995b. Proliferative and phenotypic responses of bone-like cells to mechanical deformation. Journal of Orthopaedic Research. 13:664-670.

33. Wang D. Christensen K. Chawla K. Xiao G. Krebsbach PH. Franceschi RT. 1999. Isolation and characterization of MC3T3-E1 preosteoblast subclones with distinct in vitro and in vivo differentiation/mineralization potential. J. Bone Mineral Res. 14:893-903.

34. Chang YL. Stanford CM. Keller JC. 2000. Calcium and phosphate supplementation promotes bone cell mineralization: implications for hydroxyapatite (HA)-enhanced bone formation. Journal of Biomedical Materials Research. 52:270-278.

35. Winchester SK, Bloch SR, Fiacco GJ, Partridge NC. 1999. Regulation of expression of collagenase-3 in normal, differentiating rat osteoblasts. J Cell Physiol. 181:479-488.

36. Bonewald LF, Harris SE, Rosser J, Dallas MR, Dallas SL, Camacho NP, Boyan B, Boskey A. 2003. von Kossa staining alone is not sufficient to confirm that mineralization in vitro represents bone formation. Calcif Tissue Int. 72:537-547.

37. Herrera, A.H., Elzey, B., Law, D.J. and Horowits, R. 2000. Terminal regions of mouse nebulin: Sequence analysis and complementary localization with N-RAP. Cell Motil. Cytoskel. 45:211-222.

38. Tokuyasu K.T. 1989. Use of poly(vinylpyrrolidone) and poly(vinyl alcohol) for cryo-ultramicrotomy. Histochem. J. 21:163-171.

39. Thiry M. 1999. Ultrastructural methods for nucleic acid detection by immunocytology. Progress in Histochemistry & Cytochemistry. 34(2):87-159.

40. Kimura K. Sasano H. Shimosegawa T. Mochizuki S. Nagura H. Toyota T. 2000. Ultrastructure of cells undergoing apoptosis. Vitamins & Hormones. 58:257-266.

41. Sato, M., Grasser, W., Fullenkamp, C. and Gorski, J.P. 1992. BAG-75 is a potent inhibitor of rat and chicken osteoclastic activity, FASEB J. 6:2966-2976.

42. Hall BK. 1987. Earliest evidence of cartilage and bone development in embryonic life. Clin. Orthop. 225:255-272.

43. Zhu, J.X., Sasano, Y., Takahashi, I., Mizoguchi, I., and Kagayama, M. 2001. Temporal and spatial gene expression of major bone extracellular matrix molecules during embryonic mandibular osteogenesis in rats. Histochem. J. 33:25-35.

44. Zhang, J.Q., Elzey, B., Williams, G., Lu, S., Law, D.J. and Horowits, R. 2001. Ultrastructural and biochemical localization of N-RAP at the interface between myofibrils and intercalated disks in the mouse heart. Biochemistry 40:14898-14906.

45. Pockwinse, S.M., Wilming, L.G., Conlon, D.M., Stein, G.S., and Lian, J.B. 1992. Expression of cell growth and bone specific genes at single cell resolution during development of bone tissue-like organization in primary osteoblast cultures. J. Cell Biochem. 49:310-323.

by guest on April 6, 2018

http://ww

w.jbc.org/

Dow

nloaded from

28

46. Sugawara, Y., Suzuki, K., Koshikawa, M., Ando, M., and Iida, J. 2002. Necessity of enzymatic activity of alkaline phosphatase for mineralization of osteoblastic cells. Jpn. J. Pharmacol. 88:262-269.

47. Kirsch, T., Nah, H.-D., Shapiro, I.M., and Pacifici, M. 1997. Regulated production of mineralization-competent matrix vesicles in hypertrophic chondrocytes. J. Cell Biol. 137:1149-1160.

48. Wurthier, R.E., and Register, T.C. 1985. Role of alkaline phosphatase, a polyfunctional enzyme, in mineralizing tissues. In: Chemistry and biology of mineralized tissues, Butler, WT (ed), Ebsco Media Inc., Birmingham, AL, pp. 113-124.

49. Sela, J., Schwartz, Z., Swain, L.D., and Boyan, B.D. 1992. Role of matrix vesicles in calcification. In: Calcification in Biological Systems, Bonucci, E. (ed), CRC Press, Boca Raton, FL, pp. 73-105.

50. Peress, N.S., Anderson, H.C., and Sajdera, S.W. 1974. The lipids of matrix vesicles from bovine fetal epiphyseal cartilage. Calcif. Tissue Res. 14:275-281.

51. Wuthier, R.E. 1975. Lipid composition of isolated epiphyseal cartilage cells, membrane and matrix vesicles. Biochim. Biophys. Acta 409:128-143.

52. Majeska, R.J., and Wuthier, R.E. 1975. Studies on matrix vesicles isolated from chick epiphyseal cartilage: association of pyrophosphatase and ATPase activities with alkaline phosphatase. Biochim. Biophys. Acta 391:51-60.

53. Hale, J.E., and Wuthier, R.E. 1987. The mechanism of matrix vesicle formation. Studies on the composition of chondrocyte microvilli and on the effects of microfilament-perturbing agents on cellular vesiculation. J. Biol. Chem. 262:1916-1925.

54. Wu, L.N., Genge, B.R., Lloyd, G.C., and Wuthier, R.E. 1991. Collagen-binding proteins in collagenase-released matrix vesicles from cartilage. Interaction between matrix vesicle proteins and different types of collagen. J. Biol. Chem. 266:1195-1203.

55. Eanes, E.D. 1992. Dynamics of calcium phosphate precipitation. In: Calcification in Biological Systems, Bonucci, E. (ed), CRC Press, Boca Raton, FL, pp. 1-17.

56. Satomura, K., and Nagayama, M. 1991. Ultrastructure of mineralized nodules formed in rat bone marrow stromal cell culture in vitro. Acta Anat. 142:97-104.

57. Bonnuci, E., and Reurink, J. 1978. The fine structure of decalcified cartilage and bone: a comparison between decalcification procedures performed before and after embedding. Calcif. Tissue Res. 25:179-190.

58. Bonnuci, E, Silvestrini, G, and DiGrezia, R. 1988. The ultrastructure of the organic phase associated with the inorganic substance in calcified tissues. Clin. Orthop. Relat. Res. 233:243-261.

59. Bonnuci, E, Silvestrini, G, and DiGrezia, R. 1989. Histochemical properties of the "crystal ghosts" of calcifying epiphyseal cartilage. Connect. Tissue Res. 22:43-50.

60. Shepard, N. 1992. Role of proteoglycans in calcification. In: Calcification in Biological Systems, Bonucci, E. (ed), CRC Press, Boca Raton, FL, pp. 41-58.

by guest on April 6, 2018

http://ww

w.jbc.org/

Dow

nloaded from

29

61. Shepard, N., and Mitchell, N. 1981. Acridine orange stabilization of glycosaminoglycans in beginning endochondral ossification. A comparative light and electron microscopic study. Histochemistry 70:107-114.

62. Midura, R.J., McQuillan, D.J., Benham, K.J., Fisher, L.W., and Hascall, V.C. 1990. A rat osteogenic cell line (UMR 106-01) synthesizes a highly sulfated form of bone sialoprotein. J. Biol. Chem. 265:5285-5291.

63. Ecarot-Charrier, B., Bouchard, F., and Delloye, C. 1989. Bone sialoprotein II synthesized by cultured osteoblasts contains tyrosine sulfate. J. Biol. Chem. 264:20049-20053.

64. Wu, L.N.Y., Genge, B.R., Dunkelberger, D.G., LeGeros, R.Z., Concannon, B., and Wuthier, R.E. 1997. Physicochemical characterization of the nucleational core of matrix vesicles. J. Biol. Chem. 272:4404-4411.

65. Wang, A., Gorski, J.P., and Midura, R.J. 1999. Bioapatite vesicles, a novel biomineralization structure. Trans. 45th Ann. Mtg. Orthop. Res. Soc. 24:209.

66. Nefussi, J.R., Brami, G., Modrowski, D., Oboeuf, M, and Forest, N. 1997. Sequential expression of bone matrix proteins during rat calvaria osteoblast differentiation and bone nodule formation in vitro. J. Histochem. and Cytochem. 45:493-503.

67. Irie, K., Zalzal, S., Ozawa, H., McKee, M.D., and Nanci, A. 1998. Morphological immunocytochemical characterization of primary osteogenic cell cultures derived from fetal rat cranial tissue. Anat. Rec. 252:554-567.

68. Riminucci, M., Silvestrini, G., Bonucci, E., Fisher, L.W., Gehron Robey, P., and Bianco, P. 1995. The anatomy of bone sialoprotein immunoreactive sites in bone as revealed by combined ultrastructural histochemistry and immunochemistry. Calcif. Tiss. Int. 57:277-284.

by guest on April 6, 2018

http://ww

w.jbc.org/

Dow

nloaded from

30

Figure Legends

Figure 1. Mineralizing UMR cultures deposit BSP and calcium mineral in extracellular

matrix complexes referred to as BMF. Panel A shows a UMR culture that had been treated

with β-GP for 24 h, and immunostained with anti-BSP monoclonal WV1D1(9C5) followed by

alizarin red S staining. BSP is denoted by the green color, while calcium mineral is denoted by

the red color (AR-S). DIC, differential interference contrast view of the same field of view.

Color overlay panel shows both the green and red color bands represented in a full RGB color

spectrum. Panel B shows a UMR culture that had been treated with β-GP for 24 h, and

immunostained with anti-BSP monoclonal WV1D1(9C5) followed by Vector Red alkaline

phosphatase activity staining. BSP is denoted by the green color, while alkaline phosphatase

activity is denoted by the red color. The white arrowheads point to BMFs that stain positive for

both BSP and alkaline phosphatase activity. Panel C depicts a 3D confocal analysis of a

mineralizing UMR culture. BSP is denoted by the green color, hyaluronan is denoted by the red

color, and DAPI-stained nuclei are indicated by the blue color. The X-Y view is a single 0.5-µm

slice at the indicated depth in the X-Z view panel. Panel D shows the results of a FITC-TUNEL

assay on a 24 h mineralized UMR culture. Top panel shows a color overlay of the TUNEL

(green) and alizarin red-S-stained calcium mineral (red) fluorescence. Bottom panel shows a

color overlay of the TUNEL (green), calcium mineral (red), and DAPI-stained nuclei (blue).

Note that calcium mineral deposits do not co-localize with either DAPI-stained or TUNEL-

stained nuclei, and that only about 1-2% of the nuclei are TUNEL-positive in this field of view.

Figure 2. BAG-75 is a biomarker for precursors of BMF prior to nucleation of mineral

crystals. Panels A-D are representative of confluent UMR cultures at 0 h prior to the addition of

β-GP; E and F are representative of cultures 24 h after the addition of β-GP. A and B, C and D,

by guest on April 6, 2018

http://ww

w.jbc.org/

Dow

nloaded from

31

and, E and F, respectively, represent images of the same field of view. Panel A, Anti-BAG-75

antibodies specifically label small 15-25 micron diameter populations of BMF precursors (white

arrows). Panel B, Brightfield view of small BMF precursors (black arrows). Panel C, Anti-

BAG-75 antibodies immunostain a large 150-250 micron diameter BMF precursor (white

arrow). Panel D, Brightfield image of large 150-250 micron diameter BMF precursor (black

arrow). Panels E and F, Large mineralized BMF stained with Alizarin Red S dye (white

arrows) still stains positive for BAG-75 on its surface and within its structure (white

arrowheads). Scale bars : 200 microns for A-D and 100 microns for E and F.

Figure 3. BAG-75-enriched focal deposits represent the initial sites of mineralization in

osteoblastic cultures from primary fetal rat calvarial (FRC) cells and MC3T3-E1 cells.

Refer to the Materials and Methods section for culture details. Panel A, phase contrast view of

FRC culture after 7 days of β-GP treatment; no alizarin red-S staining was detected at this time.

Panel B, BAG-75 immunostaining of FRC culture after 7 days of β-GP treatment; same field of

view as shown in panel A. Panel C, non-immune IgG control staining of FRC culture after 7

days of β-GP treatment. Panels D and E, alizarin red-S staining of apatite deposits in FRC

culture after 17 days of β-GP treatment. Panels F and G, BAG-75 immunostained FRC culture

after 17 days of β-GP treatment after decalcification using EDTA; same fields of view as shown

in panels D and E, respectively. Panel H, brightfield view of alizarin red-S stained MC3T3-E1

(subclone M4) culture after 12 days of β-GP treatment. Panel I, higher magnification brightfield

view of an individual alizarin red-S stained foci from panel H after partial decalcification. Panel

J, BAG-75 immunostaining of MC3T3-E1 culture (subclone M4) after 12 days of β-GP

treatment; same field of view as shown in panel I. Scale bars in panels A-H are 100 microns,

while those shown in panels I and J are 2 microns.

by guest on April 6, 2018

http://ww

w.jbc.org/

Dow

nloaded from

32

Figure 4. BAG-75 content of BMF increases during mineralization. Fully mineralized

cultures (24 h β-GP) were decalcified overnight with Tris-buffered saline (pH 8.0) containing 50

mM EDTA. Cultures were then processed for immunofluorescence microscopy with anti-BAG-

75 and then with Alexa-green conjugated anti-rabbit IgG antibody. For comparison, a parallel

fully mineralized culture without decalcification in shown in Figure 2E,F. Left panels, Paired

anti-BAG-75 and brightfield images of 24 h (β-GP minus) culture. Right panels, Paired anti-

BAG-75 and brightfield images of 24 h (β-GP plus) culture after decalcification. Scale bar: 20

microns.

Figure 5. BSP co-localizes to large and small BMF complexes containing BAG-75.

Confluent UMR cultures were incubated for 24 h in the presence of β-GP, fixed in ethanol, and

then processed for immunofluorescence microscopy with anti-BSP monoclonal WV1D1(9C5)

(red) and anti-BAG-75 (green) antibodies as described in Materials and Methods. Panels A-D

represent images of the same field of view containing small BMF. White and black arrows mark

BMF that stain strongly for both BAG-75 and BSP antigens. Panel D, Color overlay image for

both BSP (red) and BAG-75 (green). Panel E, Brightfield view of an Alizarin Red-S stained

culture showing the presence of many small BMF (black arrowheads) and two large BMF (black

arrows). Panels F-H are images from the same field of view containing a large and several small

BMF. Upper (Panel G) and lower (Panel H) confocal optical planes of anti-BSP stained large

BMF. Approximate displacement of images in G and H is about 15 microns (4 µm versus 19 µm

from the basal aspect of the culture, respectively). Note the presence of BSP-stained small BMF

(white arrows) and intracellular BSP staining of individual cells within the surrounding

monolayer (white arrowheads) in this lower optical plane only. Scale bars: panel D, 50

microns; panels E-G, 200 microns.

by guest on April 6, 2018

http://ww

w.jbc.org/

Dow

nloaded from

33

Figure 6. Ultrastructural views of BMF. Panel A shows a view cut perpendicular to the

culture surface through a single BMF from a mineralizing UMR culture incubated for 12 h in the

presence of β-GP, then fixed and processed for electron microscopy as described in Materials

and Methods. The black line shown roughly 2 µm above the basal aspect of the culture

represents the relative position of a 100 nm thin-section along a plane parallel to the culture

surface representing the viewing planes shown in panels B-D. Inset shows a scanning EM view

of two supramolecular complexes formed spontaneously from purified BAG-75 (24). Panel B

shows a view of small and large spherical structures within a mineralizing BMF. The

arrowheads indicate the locations of 50-200 nm diameter opaque structures having an

ultrastructural appearance similar to matrix vesicles. The asterisks indicate the locations of 300-

600 nm diameter translucent spheres that we refer to as bioapatite vesicles. The inset in the right

panel is a higher magnification view of the small and large vesicles indicated by the arrowhead

and asterisk at the top of the right panel. Note how both vesicular structures appear to touch each

other’s surfaces. Panels C and D show views cut parallel to the culture surface through a

mineralizing BMF (panel C), and a BMF precursor from a non-mineralizing control (panel D);

note the healthy looking cells near the BMF areas. Boxes in panels C (left) and D (left) are areas

shown at higher magnification in their respective right panels. Arrows in panel C (left) indicate

positions of other BMF not shown at higher magnification in panel C (right). Insets are higher

magnification views of the boxed areas in panels C (right) and D (right).

Figure 7. Ultrastructural view of a mineralizing BMF stained with anti-BSP antibody LF-

87. Panel A, the arrows indicate the positions of matrix areas containing clusters of more than

ten 12-nm gold particles. The immunogold particles appear associated with thin filament fibrils

and not banded fibrillar collagen structures. Panel B, the arrows indicate the positions of clusters

by guest on April 6, 2018

http://ww

w.jbc.org/

Dow

nloaded from

34

of immunogold particles associated with the larger translucent bioapatite vesicles. Arrowheads

indicate the positions of structures that resemble matrix vesicles, and do not contain associated

gold particles. Panel C, a higher magnification view of both large and small vesicular structures.

Arrow indicates a region of the larger structure containing a cluster of immunogold particles that

appears to be at its surface. Arrowhead indicates a clear view of the trilaminar-delimiting

boundary surrounding the smaller vesicle. Note how both larger and smaller vesicles appear to

touch each other’s surfaces.

Figure 8. Ultrastructural view of a mineralizing BMF stained with both anti-BAG-75 and

anti-BSP WV1D1 (9C5) antibodies. Panel A, the box indicates an area of the BMF near its

outer boundary shown at higher magnification in panel B. Panel B, the area labeled “BAG-75”

indicates the position of a large cluster of 12-nm immunogold particles identifying BAG-75

epitopes associated with thin filament fibrils. The box labeled “BSP” indicates the position of a

cluster of 6-nm immunogold particles identifying BSP epitopes. Inset shows a higher

magnification view of the 6-nm gold particles. Note that submicron distances separate both 12-

nm and 6-nm gold particles.

Figure 9. Diagram depicting the UMR biomineralization model. Times shown in steps 4 and

5 refer to organophosphate exposure times. Note that BSP is not retained within the BMF

precursor containing BAG-75 (steps 2 and 3) until an organophosphate stimulant is added to the

culture (step 4), at which time BSP is then retained in the BMF matrix.

by guest on April 6, 2018

http://ww

w.jbc.org/

Dow

nloaded from

36

FIGURES

Figure 1

DC

A B

by guest on April 6, 2018

http://ww

w.jbc.org/

Dow

nloaded from

37

Figure 2

by guest on April 6, 2018

http://ww

w.jbc.org/

Dow

nloaded from

38

Figure 3 (new)

by guest on April 6, 2018

http://ww

w.jbc.org/

Dow

nloaded from

39

Figure 4

by guest on April 6, 2018

http://ww

w.jbc.org/

Dow

nloaded from

40

Figure 5

.

by guest on April 6, 2018

http://ww

w.jbc.org/

Dow

nloaded from

41

Figure 6A

Figure 6B

X-Z view

BMF

A

B

by guest on April 6, 2018

http://ww

w.jbc.org/

Dow

nloaded from

42

Figure 6C and 6D

C

D

by guest on April 6, 2018

http://ww

w.jbc.org/

Dow

nloaded from

43

Figure 7

by guest on April 6, 2018

http://ww

w.jbc.org/

Dow

nloaded from

44

Figure 8

by guest on April 6, 2018

http://ww

w.jbc.org/

Dow

nloaded from

45

Figure 9

by guest on April 6, 2018

http://ww

w.jbc.org/

Dow

nloaded from

GorskiRonald J. Midura, Aimin Wang, Dinah Lovitch, Douglas Law, Kimerly Powell and Jeff P.

sialoprotein accumulation and apatite nucleation in osteoblastic culturesBone acidic glycoprotein-75 delineates the extracellular sites of future bone

published online March 5, 2004J. Biol. Chem. 

  10.1074/jbc.M312409200Access 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

by guest on April 6, 2018

http://ww

w.jbc.org/

Dow

nloaded from