Chondrocyte deformation within compressed agarose constructs at the cellular and sub-cellular levels

*Corresponding author. Tel.: -44-171-975-5274; fax: -44-181-983-1799.

E-mail address: [email protected] (D.L. Bader)

Journal of Biomechanics 33 (2000) 81}95

Chondrocyte deformation within compressed agaroseconstructs at the cellular and sub-cellular levels

David A. Lee!, Martin M. Knight", John F. Bolton#, Bernadine D. Idowu!,Michael V. Kayser#, Dan L. Bader",*

!IRC in Biomedical Materials, University College London Medical School, Brockley Hill, Stanmore, Middlesex HA7 4LP, UK"IRC in Biomedical Materials, Queen Mary and Westxeld College, University of London, Mile End Road, London E1 4NS, UK

#Electron Microscopy Unit, Institute of Orthopaedics, University College London Medical School, Brockley Hill, Stanmore, Middlesex HA7 4LP, UK

Abstract

Mechanotransduction events in articular cartilage may be resolved into extracellular components followed by intracellularsignalling events, which "nally lead to altered cell response. Cell deformation is one of the former components, which has beenexamined using a model involving bovine chondrocytes seeded in agarose constructs. Viable #uorescent labels and confocal laserscanning microscopy were used to examine cellular and sub-cellular morphology. It was observed that cell size increased up to day6 in culture, associated with an increase in the contents of proteoglycan and collagen. In addition, the organisation of the cytoskeletoncomponents, described using a simple scoring scale, revealed temporal changes for actin "bres, microtubules and vimentinintermediate "laments. The constructs on day 1 were also subjected to uncon"ned compressive strains. A series of confocal scansthrough the centre of individual cells revealed a change from a spherical to an elliptical morphology. This was demonstrated bya change in diameter ratio, from a mean value of 1.00 at 0% strain to 0.60 at 25% strain. Using simple equations, the volume andsurface areas were also estimated from the scans. Although the former revealed little change with increasing construct strain, surfacearea appeared to increase signi"cantly. However further examination, using transmission electron microscopy to reveal "neultrastructural detail at the cell periphery, suggest that this increase may be due to an unravelling of folds at the cell membrane. Celldeformation was associated with a decrease in the nuclear diameter, in the direction of the applied strain. The resulting nuclear strainin one direction increased in constructs compressed at later time points, although its values at all three assessment times were less thanthe corresponding values for cell strain. It is suggested that the nuclear behaviour may be a direct result of temporal changes observedin the organisation of the cytoskeleton. The study demonstrated that the chondrocyte}agarose model provides a useful system for theexamination of compression events at both cellular and sub-cellular levels. ( 1999 Elsevier Science Ltd. All rights reserved.

Keywords: Chondrocyte; Mechanotransduction; Confocal microscopy; Transmission electron microscopy; Cytoskeleton; Cell and nuclear deforma-tion; Compressive strain

1. Introduction

During normal walking, articular cartilage is subjectedto a complex state of dynamic loading, with a signi"cantcomponent applied perpendicular to the articular surface(Armstrong et al., 1984). It is well established that thechondrocytes within the tissue can detect and respond tothe applied load by altering their metabolism, througha process known as mechanotransduction (Palmoski etal., 1979; Palmoski and Brandt, 1982,1984; Gray et al.,

1988,1989; Kiviranta et al., 1988; Sah et al., 1989,1991;Parkkinen et al., 1992; Burton-Wurster et al., 1993; Kimet al., 1994). The application of compressive load tocartilage induces cell deformation, changes in hydrostaticor #uid pressure and deformation of the charged ex-tracellular matrix, with associated localised alterations inosmolality and pH (Urban, 1994; Hall et al., 1996a,b)These initial signals may activate speci"c intracellularmechanotransduction pathways.

It is not possible to study cell deformation alone inintact cartilage explants due to the coupling of mechan-ical and physicochemical processes. Several groups have,therefore, utilised a system involving the culture ofisolated chondrocytes embedded in agarose gel(Freeman et al., 1994; Buschmann et al., 1995; Lee and

0021-9290/99/$ - see front matter ( 1999 Elsevier Science Ltd. All rights reserved.PII: S 0 0 2 1 - 9 2 9 0 ( 9 9 ) 0 0 1 6 0 - 8

Bader, 1995,1997; Lee et al., 1998a,b; Knight et al., 1996;Knight, 1997). The application of static or dynamic com-pressive strains to the chondrocyte/agarose system re-sults in cell deformation and frequency dependentalterations in cell metabolism. The initial chondrocytedeformation event will induce strain events at the sub-cellular level, potentially involving the cell membraneand the nucleus, mediated by extracellular matrix and thecytoskeleton. These processes have been demonstrated,previously in intact cartilage. Little is known, however,about the exact nature of the deformation process or thepotential induction of intracellular mechanotransductionpathways in the agarose system.

Cell deformation is typically associated with changesin cell volume and/or surface area (Guilak, 1994,1995;Guilak et al., 1995; Buschmann et al., 1996; Wong et al.,1996,1997; Errington et al., 1997; Knight, 1997). Alter-ations in chondrocyte volume are known to activatevolume regulatory mechanisms involving various mem-brane ion transporters which have been implicated inaltered cellular activity (Urban et al., 1993; Hall, 1995;Hall et al., 1996a,b). Membrane deformation, induced bychanges in surface area, can activate various forms ofstretch sensitive ion channels and induce alterations inmembrane potential (Steck, 1989; Wright et al.,1992,1996; Bell and Holmes, 1994; Guilak et al., 1994;Bowman and Lohr, 1996; Tavernarakis and Driscoll,1997). Deformation of chondrocytes within intact carti-lage is reported to be associated with a decrease in thevolume of chondrocytes due to cellular con"nement andan increase in extracellular osmolality during compres-sion (Guilak, 1994,1995; Guilak et al., 1995; Buschmannet al., 1996; Wong et al., 1997). A previous study, usinglight microscopy, estimated volume strains in chon-drocytes embedded in agarose subjected to compressionand reported that deformation induced a reduction incell volume (Freeman et al., 1994).

Static compression of articular cartilage explants hasbeen shown to result in nuclear deformation and anassociated reduction in the volume of the nucleus(Guilak, 1994,1995; Guilak et al., 1995; Buschmann et al.,1996). Recent studies have suggested that deformation ofthe intracellular organelles including the nucleus, maymediate alterations in matrix synthesis associated withcompression (Buschmann et al., 1996; Quinn et al., 1998;Janmey, 1998). However, the precise signalling pathwaysremain unclear. The mechanism by which gross compres-sion is mediated through the cell to the nucleus is thoughtto involve the cytoskeleton (Ben-Ze'ev, 1991; Ingber,1991; Wang et al., 1993; Ingber et al., 1994; Guilak, 1995;Maniotis et al., 1997; Janmey, 1998). Furthermore thecytoskeletal organisation itself is known to be critical fornormal cell metabolism and may therefore be involved inmechanotransduction. Indeed a number of studies havedemostrated active remodelling of the cytoskeleton inresponse to mechanical stimuli (Franke et al., 1984;

Ookawa et al., 1992; Barbee, 1994; Parkkinen et al., 1995;Meazzini et al., 1998).

In the current study the authors investigate the e!ect ofcompressive strain on chondrocytes seeded withinagarose constructs at both the cellular and sub-cellularlevels. This involved characterising the morphology,cytoskeletal organisation and matrix elaboration withinunstrained agarose/chondrocytes constructs cultured forup to 6 days. The volume and surface area of chon-drocytes within constructs, both unstrained andsubjected to incremental compressive strains, have beenestimated using confocal laser scanning microscopy inconjunction with calcein AM staining. Transmissionelectron microscopy was also employed to study thee!ect of cell deformation on the ultrastructural organisa-tion at the cell periphery. Alterations in chondrocytedeformation at the cell and sub-cellular levels was alsoexamined to compare the cell and nuclear deformation instrained constructs.

2. Materials and methods

2.1. Chondrocyte isolation and preparation of agaroseconstructs

The front feet from freshly slaughtered 18 month-oldsteers were obtained from a local abattoir. The metacar-palphalangeal joint was opened, under aseptic condi-tions, within three hours of slaughter. Full-depth slices ofcartilage were removed from the proximal joint surface,washed in Earle's balanced salt solution (EBSS, Gibco,Paisley, UK) and cultured at 373C/ 5% CO

2for 16 h in

Dulbecco's minimal essential medium supplementedwith 20% (v/v) foetal calf serum (DMEM #20% FCS,Gibco, Paisley, UK). The cartilage slices were diced "ne-ly, and incubated at 373C for 1 h in DMEM #20% FCS#700 unit.ml~1 pronase (BDH Ltd., Poole, UK). Thesupernatant was removed and replaced with DMEM#20% FCS #100 unit.ml~1 collagenase type 1a(Sigma, Poole, UK) and incubated at 373C for 16 h. Thesupernatant containing released chondrocytes waspassed through a 70 lm pore size sieve (Falcon, Oxford,UK) to remove undigested tissue, washed twice inDMEM #20% FCS and resuspended at a cell concen-tration of 8]106 cells.ml~1. Cell viability was deter-mined using the trypan blue exclusion test. Thechondrocyte suspension was added to an equal volume of6% agarose (type VII, Sigma, Poole, UK) in EBSS to givea "nal concentration of 4]106 cells.ml~1 in 3% agarose.The agarose/chondrocyte suspension was pipetted intoa specially designed perspex mould and allowed to gel at43C for 20 min. Cylindrical constructs (5 mm in diameterand 5 mm in height) were cored out and cut longitudi-nally into half-core constructs. In some experiments,block constructs, measuring 17]11]5 mm, were used.

82 D.A. Lee et al. / Journal of Biomechanics 33 (2000) 81}95

The constructs were subsequently maintained in DMEM#20% FCS at 373C/ 5% CO2 for up to 6 days. Culturemedia were replaced every 2 days.

2.2. Cell morphology, cytoskeletal organisation and matrixelaboration by chondrocytes in unstrained agarose

On days 0, 1, 3 and 6 representative agarose/chon-drocyte half-core constructs were removed from cultureand incubated in a 5 lM solution of Calcein AM (Cam-bridge Bio Science, Cambridge, UK) in DMEM #20%FCS at 373C/ 5% CO

2for 1 hr. The specimens were

mounted onto the stage of an inverted microscope(Nikon, Kingston-upon-Thames, UK) associated witha Oz confocal laser scanning microscope (Noran Instru-ments, Milton Keynes, UK). A horizontal scan throughthe centre of a representative chondrocyte within theconstruct was produced, using the selected ]40/0.55ELWD objective lens and a laser excitation wavelengthof 488 nm. To de"ne the cell boundary, a half maximumintensity threshold was applied to the resulting confocalimages as adopted in previous studies (Guilak, 1994;Errington et al., 1997; Knight et al., 1996,1998;Knight, 1997). The full-width half-maximum (FWHM)diameters of each cell were measured parallel (x)and perpendicular (y) to the long axis of the constructand a diameter ratio was determined using the followingequation.

Diameter ratio"x

y. (1)

This parameter is equivalent to the deformation indexusing in previous studies (Lee and Bader, 1995; Knight,1997; Knight et al., 1997; Lee et al., 1998b). This processwas repeated for a total of 100 cells.

On days 0, 1, 3 and 6, representative agarose/chon-drocyte core constructs were removed from culture,washed in phosphate bu!ered saline (PBS, Gibco, Pais-ley, UK) and "xed in 4% formaldehyde in PBS #2%(w/v) sucrose, pH 7.2 at 373C for 20 min. Thin discs,1 mm in thickness and 5 mm in diameter were cut fromthe "xed agarose constructs and washed twice in PBS.The cells were permeabilised by incubation in PBS#300 mM sucrose #3 mM MgCl

2#0.5% (v/v) triton

X-100 (all BDH, Poole, UK) for 5 min. at 43C. Thespecimens were washed in PBS #1% (w/v) bovineserum albumin (BSA, Sigma, Poole, UK) and incubatedin a solution containing 1 lg ml~1 #uorescein iso-thiocyanate-conjugated phalloidin (FITC-phalloidin,Sigma, Poole, UK) in PBS #1% BSA (Sigma, Poole,UK) for 1 h at 373C. Further specimens were incubatedin solutions containing either monoclonal anti {-tubulin(clone TUB2-1, Sigma, Poole, UK) or monoclonal anti-vimentin (clone V9, Sigma, Poole, UK) in PBS #1%BSA for 1 h at 373C. Both antibodies were used at 1 : 50

dilution. The specimens were washed in PBS #1% BSAand subsequently incubated for 1 h at 373C in a solutioncontaining #uorescein isothiocyanate-conjugated rabbitanti-mouse immunoglobulin at a 1:30 dilution in PBS#1% BSA. The specimens were washed in PBS #1%BSA and visualised using a ]60 oil immersion lens, NA1.40 (Nikon, Kingston-upon-Thames, UK) and a 15 lmconfocal slit with 32X jump averaging. Single confocalscans were obtained from a population of individual cellsand classi"ed into three groups as follows; nonorganisedcytoskeleton * Amorphous staining within the cellcytoplasm without speci"c organisation, partially or-ganised cytoskeleton* Staining visible as faint structuralfeatures in some areas of the cell cytoplasm and clearlyorganised cytoskeleton * Staining visible as clear struc-tural features within the cell cytoplasm. A numericalscoring system was used to indicate overall levels oforganisation within each sample population by allocat-ing scores of 0, 1 and 2 for non-organised, partiallyorganised and clearly organised cells, respectively. Allclassi"cations were performed by a single observer.

On days 0, 1, 3 and 6, representative agarose/chon-drocyte half-core constructs were removed from cultureand digested with 2.8 unit.ml~1 papain and 10 unit.ml~1

agarase (both Sigma, Poole, UK) as described previously(Lee and Bader, 1997). The digests were assayed for totalsulphated glycosaminoglycan using the DMB method(Farndale et al., 1982) and total hydroxyproline usinga microplate modi"cation of the method reported byEdwards and O'Brien (1980).

2.3. Chondrocyte deformation on day 1

Representative agarose/chondrocyte block constructs,cultured for 1 day in DMEM #20% FCS, were incu-bated in a 5 lm solution of Calcein AM (Cambridge BioScience, Cambridge, UK) in DMEM #20% FCS at373C/ 5% CO

2for 1 h at 373C. The constructs were

placed within a specially designed compression test rig(Knight et al., 1996), mounted on the stage of the invertedmicroscope associated with the confocal laser scanningmicroscopy system. Based on a previously optimisedtechnique (Knight et al., 1996,1998), a horizontal con-focal scan was made through the centre of a representa-tive chondrocyte within the agarose construct. Theindividual cells for examination were restricted to thosewithin the middle of the agarose construct, at least 1 mmaway from the plattens at either end and at a depth ofbetween 50 and 100 lm from the surface closest to thecoverslip. A 5% uniaxial uncon"ned static compressivestrain was applied at a strain rate of approximately20% min~1, by means of a micrometer head attached tothe test rig. A 10 min period of stress relaxation waspermitted, after which a scan through the centre of thesame chondrocyte was produced. Further scans of thischondrocyte were produced at 10, 15, 20 and 25%

D.A. Lee et al. / Journal of Biomechanics 33 (2000) 81}95 83

strains. The process was repeated for a total of 17 cells.The cell boundary was de"ned as in the previous sectionand cell diameters, parallel (x) and perpendicular (y) tothe direction of the applied strain, were determined ateach level of applied strain. Diameter ratios were cal-culated for each cell at each strain level using Eq. (1).

The cross-sectional area of the cells was calculatedusing the area measurement function associated with theconfocal laser scanning microscope and additionallyfrom the measured x and y diameters using the followingequation for an ellipse:

Area of ellipse"p

4xy. (2)

The volume (<) of each cell at each strain level wasestimated as follows.

<"py2x

6. (3)

The surface area (SA) of each cell at each strain level wasestimated using the following standard equations for anoblate ellipsoid (xOy) or for a sphere (x"y):

SA%--*140*$

"Cpy2

2 D#Cpx2

2e DlnCy(1#e)

x D, (4)

where eccentricity e is calculated as follows:

e2"y2!x2

y2, (5)

SA41)%3%

"

p(x#y)2

4. (6)

In a separate experiment representative agarose/chon-drocyte block constructs, cultured for 1 day in DMEM#20% FCS were placed within the test rig and com-pressed to a 20% gross compressive strain, which wasmaintained for 15 min. Control constructs remained inthe unstrained state. The diameter ratios of a populationof cells were determined en block using phase contrastmicroscopy. The medium within the test rig was replacedwith 4% paraformaldehyde/0.4% glutaraldehyde in0.1 M sodium cacodylate bu!er pH 7.2 (all Agar Scient-i"c, Stansted, UK) and the constructs were "xed, in thestrained or unstrained state, for 2 h at room temperature.The constructs were subsequently removed from the testrig. Diameter ratios were monitored en block for up to12 h after removal from strain using phase contrastmicroscopy. Slices, approximately 0.8 mm thick, wereremoved from the "xed constructs in one of three ori-entations, representing the x/y, x/z and y/z planes. Theslices were additionally "xed for 16 h at 43C in 2%glutaraldehyde in 0.1 M sodium cacodylate bu!er pH 7.2(all Agar Scienti"c, Stansted, UK). The slices werewashed in 0.1 M sodium cacodylate bu!er pH 7.2 and

post-"xed in 1% osmium tetroxide (Agar Scienti"c, Stan-sted, UK) in 0.1 M sodium cacodylate bu!er pH 7.2. Theslices were dehydrated using an alcohol series (70, 90, 96and 100%) with 2 changes of 10 min at each concentra-tion. Between the 70 and 90% alcohol incubations, theslices were additionally incubated in 2% uranyl acetate(Agar Scienti"c, Stansted, UK) in 70% alcohol. Celldiameters and diameter ratios (x/y, x/z and y/z for theequivalent plane) were measured for cells within slices ateach alcohol concentration using phase contrast micro-scopy. After dehydration, the slices were in"ltrated andembedded in Spurr's resin (Agar Scienti"c, Stansted,UK). Semi-thin (1 lm) sections were cut, stained with 1%toluidine blue and viewed using light microscopy fordetermination of diameter ratios. In addition, ultra-thin(90 nm) sections were collected onto copper grids, stainedwith 2% aqueous uranyl acetate and Reynolds' leadcitrate (Reynolds, 1963) and viewed using TEM (PhillipsCM12, Phillips, Cambridge, UK).

Micrographs were obtained from a total of 119 cellsvisualised by transmission electron microscopy. Thenegatives were backlit using an X-ray viewing box anda video camera was used to view the image on a com-puter screen. Image Grabber 2.1 (Graftek Imaging, Aus-tin, USA) was used to capture the image. Using Optilab(Graftek Imaging, Austin, USA), the image was thre-sholded to di!erentiate the cell from its background.Using the thinning command, a perimeter one pixel widewas obtained and exported to NIH Image 1.6. The cellperimeter (P

t) was quanti"ed. Additionally, the major (y)

and minor (x) axes were measured from the capturedimage and the perimeter of the best "tting ellipse (P

e)

calculated using the following equation:

p%"

p

J2J(x2#y2).

(7)

An equivalent equation was applied in the x/z and y/zplanes. A perimeter ratio was estimated as follows:

Perimeter ratio"P

tP%

. (8)

A value of 1.0 for the perimeter ratio indicates a perfectlysmooth cell periphery.

A one-way ANOVA was used to compare cell dia-meters and diameter ratios for cells during "xation anddehydration. Unpaired Student's t-tests were used tocompare perimeter ratios from strained cells with thecorresponding unstrained controls. In all cases the 0.05level of signi"cance was used.

2.4. The inyuence of culture on cell and nucleardeformation

Half-core agarose/chondrocyte constructs were pre-pared and cultured in DMEM #20% FCS for 0, 1, 3 or


Table 1Overview of the experimental procedures used

Specimen Half-core construct Block constructCulture period Days 0, 1, 3 and 6 Day 1

Compressed state Unstrained Strained 0 and 20% Strained 0, 5, 10, 15, 20, 25% Strained 0 and 20%Staining Cell (live), CSK j (live) Cell (live), nucleus Cell (live) Cell ("xed)Imaging technique CLSM CLSM CLSM TEMParameters measured Diameter Diameter Diameter Diameter

Diameter ratio Diameter ratio Diameter ratioCSK organisation Volume Perimeter ratioGAG Surface areaHydroxyproline

CLSM Confocal laser scanning microscopy.TEM Transmission electron microscopy.CSK Cytoskeletal elements.GAG Glycosaminoglycan.

Table 2Unstrained cell diameters, diameter ratios, total sulphated glycosaminoglycan content (GAG) and total hydroxyproline content (OH-proline) ofagarose/chondrocyte half-core constructs cultured for up to 6 days in DMEM #20% FCS

Time in culture (days)

0 1 3 6

Cell diameter! (lm) 9.45 [1.5] 9.40 [1.2] 10.2 [1.9] 11.8 [2.8]Diameter Ratio" (x/y) 0.99 [0.05] 1.00 [0.04] 1.00 [0.04] 1.02 [0.03]GAG# (lg ml~1 agarose) 73.0 [15.9] 85.9 [15.1] 265.9 [24.6] 585.8 [82.8]OH-proline# (lg ml~1

agarose)16.7 [8.9] 22.2 [6.4] 27.3 [5.7] 47.3 [9.1]

!Diameter values are presented as sample medians with inter quartile ranges in parenthesis (n"100 at each time point)."Diameter ratios are presented as sample means with standard deviations in parenthesis (n"100 at each time point).#GAG and OH-proline values are presented as sample means with standard deviations in parenthesis (n"4 at each time point).

6 days. At each time point, the specimens were dividedinto two groups and incubated for 1 hr at 373C/ 5%CO

2in either DMEM #20% FCS #5 lM Calcein

AM or DMEM #20% FCS #1 lg ml~1 Syto 64(Cambridge Bio Science, Cambridge, UK) to label thecells or nuclei, respectively. Individual constructs wereimmersed in DMEM#20%FCS within the compres-sion test rig as previously described. Single confocalsections were made bisecting the centre of individual cellsor nuclei in both the unstrained construct and again at20% gross compressive strain. Cell or nuclearboundaries were de"ned as in the previous section andx diameters were measured parallel to the axis of com-pression. Measuring the same cells in both the unstrainedand strained states improved the statistical power of thestudy by providing paired data thus enabling the per-centage changes in cell and nucleus dimensions to becalculated.

An overview of the experimental procedures is present-ed in Table 1.

3. Results

3.1. Cell morphology, cytoskeletal organisation and matrixelaboration by chondrocytes in unstrained agarose

Chondrocytes, freshly isolated from articular cartilagehad a median cell diameter of 9.5 lm (Table 2). With timein culture the median diameter increased, reaching11.8 lm by day 6. Mean diameter ratios were approxim-ately 1.00 throughout the 6 day culture period (Table 2).

Confocal scans through the centre of individual chon-drocytes within agarose, after #uorescent localisation forall three cytoskeletal components, could be classi"ed intothree groups as follows; nonorganised cytoskeleton* (Figs. 1A, B and C), partially organised cytoskeleton* (Figs. 1D, E and F) and clearly organised cytoskeleton* (Figs. 1G, H and I). Clear organisation of actin inter-mediate "laments was characterised by punctate stainingaround the periphery of the cell, whilst microtubulesand vimentin intermediate "laments formed a "brous


Fig. 1. Single confocal images illustrating nonorganised (A, B, C), partially organised (D, E, F) and clearly organised (G, H, I) cytoskeletal features inisolated chondrocytes seeded within agarose constructs. Specimens were stained for actin micro"laments (A, D, G), microtubules (B, E, H) or vimentinintermediate "laments (C, F, I). Scale bar"2 lm.

Fig. 2. Mean scores obtained using a numerical scoring system toindicate overall levels of cytoskeletal organisation within a populationof chondrocytes cultured within agarose constructs for up to 6 days.Scores of 0, 1 and 2 were allocated to individual cells demonstratingnon-organised, partially-organised and clearly-organised cytoskeletalfeatures respectively. Each point represents the mean and standarderror for a population of between 60 and 103 cells from two separateexperiments.

network which extended between the plasma membraneand the nuclear envelope. Average scores obtained usingthe numerical scoring system are indicated in Fig. 2. Atday 0 and after 1 day in culture chondrocytes showedlittle organisation of actin micro"laments. However, thisorganisation increased steadily thereafter up to 6 days inculture. By contrast, partial or clearly organised micro-tubule and vimentin intermediate "lament cytoskeletalcomponents were present in a proportion of chon-drocytes at day 0. However, the organisation of both ofthese cytoskeletal components was less marked after1 and 3 days in culture than on initial seeding. There wasa general increase in their organisation between days3 and 6. Separate tests indicated consistent classi"cationby the observer and accuaracy of classi"cation judged bycomparison with blind assessment by other assessors.

Biochemical analysis of matrix elaboration withinchondrocyte/agarose half-core constructs for up to6 days, are presented in Table 1. Over the 6 day cultureperiod the total sulphated glycosaminoglycan content,a measure of synthesised proteoglycan, increased stead-ily. By contrast, total collagen, as indicated by thehydroxyproline assay, altered little over the "rst 3 days in


Fig. 3. Horizontal confocal sections bisecting the centre of a single chondrocyte, stained with Calcein AM within an agarose construct compressed tostrains (e) of up to 25%. The images have been subjected to computer thresholding to delineate the cell boundaries.

culture but increased thereafter to yield a value whichwas threefold greater than on initial cell seeding.

3.2. Chondrocyte deformation on day 1

A series of confocal scans through the centre of a repre-sentative cell, on application of incremental compressivestrains, suggest a change from a spherical to an ellipticalmorphology (Fig. 3). This change was a direct conse-quence of a signi"cant reduction in x diameter and a sig-ni"cant increase in y diameter. The change in cell shape isclearly re#ected in a reduction in mean cell diameterratios from 1.00$0.03 at 0% strain to 0.60$0.05 inagarose constructs subjected to 25% compressive strain.The relatively small variability in estimated diameterratios of the cell population, approximately 8%, suggestfairly uniform deformation in the volume of interest inthe compressed construct.

Percentage changes in both volume and surface area,estimated for cells subjected to increments of gross com-pressive strain up to 25% are presented in Fig. 4 A and B,respectively. Based upon horizontal scans through thecentre of a sample of 17 cells, the cross-sectional areascalculated using the equation for an ellipse were largerthan those measured using the software measurement

function by a mean value of 3.6$3%. There was nosigni"cant alteration in cell volume on application ofstrain (Fig. 4A). By contrast estimated cell surface areaincreased signi"cantly, with a mean increase of 5.1% at20% gross compressive strain (Fig. 4B).

The mean diameter ratios of a population of cells,measured within chondrocyte/agarose block constructs"xed under strain and subsequently released weremonitored throughout processing for transmission elec-tron microscopy. There was no signi"cant change inmean diameter ratio at any stage. Indeed, the meandiameter ratio of cells in the x/y plane determined fromsemi-thin sections prepared from resin embedded speci-mens was similar to that in specimens prior to "xation(0.62$0.05 and 0.66$0.07, respectively). However, themean diameter of cells, measured within "xed and slicedagarose specimens maintained at various alcohol concen-trations, indicated that dehydration from 0 to 100%alcohol induced a 25% reduction in cell diameter.

The diameter ratios of a population of cells withinsemi-thin sections prepared from resin-embedded speci-mens are presented in Fig. 5. For unstrained constructsthere was no signi"cant di!erence between diameter ra-tios from specimens sectioned in orientations x/y, x/z andy/z. Cells from semi-thin sections prepared from agarose


Fig. 4. Percentage changes in cell volume (A) and surface area(B) estimated for 17 cells within agarose constructs compressed tostrains of up to 25%. Correlation coe$cient r"0.26, p'0.05 forvolume and r"0.62, p(0.001 for surface area. The data were ob-tained from two separate experiments.

Fig. 5. Diameter ratios, determined from semi-thin sections preparedfrom resin embedded specimens, for chondrocytes "xed within con-structs either unstrained or subjected to 20% compressive strain andsectioned in the x/y, x/z and y/z planes. Each value represents the meanand standard error of the mean for at least 36 replicates from twoseparate experiments. Unpaired student's t-test results indicate di!er-ences from unstrained control values as follows: *"p(0.05.

constructs strained by 20% exhibited signi"cantly re-duced diameter ratios in the x/y and x/z planes, althoughthe two values were not signi"cantly di!erent. By con-trast, the mean diameter ratio in the y/z plane was1.00$0.09.

Representative transmission electron micrographs ofchondrocytes within unstrained agarose constructs andconstructs strained by 20% and sectioned in the x/y, x/zand y/z planes are presented in Fig. 6. Chondrocytesfrom unstrained constructs exhibited a pro"le which wasapproximately round in all three planes (Figs. 6A, B, C).The periphery of the cells was characterised by numerouscell processes and undulations with sizes in the range0.1}0.5 lm. Chondrocytes from strained agarose con-structs sectioned in the x/y and x/z planes exhibited anelliptical morphology and were markedly smoother inpro"le as demonstrated by the lack of cell processes andinvaginations (Figs. 6D, E). Chondrocytes from strainedconstructs sectioned in the y/z plane were more roundedand exhibited some cell processes and undulations(Fig. 6F).

Perimeter ratios from chondrocytes within agaroseconstructs, unstrained and strained by 20%, are present-ed in Fig. 7. For unstrained cells, mean perimeter ratiosranged from 1.36 to 1.44, with no signi"cant di!erencesbetween sections prepared in the three orthogonalplanes. Perimeter ratios for strained cells were signi"-cantly lower than unstrained cells in all three planes.There was no signi"cant di!erence in the perimeter ratiosfor strained cells in the x/y and x/z planes, whilst ratios inthe y/z plane were signi"cantly greater.

3.3. The inyuence of culture on cell and nucleardeformation

At all time points studied, 20% compression of cell-agarose constructs resulted in deformation of the cellsfrom a spherical to an oblate elipsoid morphology aspreviously described. Representative images of chon-drocytes subjected to compressive strain on day 1 arepresented in Figs. 3 and 8A, B. For the purposes of thisstudy, deformation was characterised by a reduction inx diameter, parallel to the axis of compression. On day 0,the reduction in x diameter was approximately 17%(Fig. 9). However, with increasing time in culture therewas a signi"cant reduction in cell deformation such thatcompression on days 3 and 6 resulted in cell x diameterstrains of 12 and 6%, respectively (Fig. 9).

Confocal microscopy of cells labelled with Syto 64indicated that agarose compression resulted in nucleusdeformation which was characterised by a reduction innucleus x diameter. Images of a representative chon-drocyte, subjected to 0 and 20% strain on day 1, arepresented in Fig. 8C and D and illustrate nuclear defor-mation. At each time point, however, there was consider-able variability in the levels of nucleus deformation. The


Fig. 6. Representative transmission electron micrographs of chondrocytes "xed within unstrained agarose constructs and sectioned in the x/y(A), x/z(B) and y/z (C) planes and chondrocytes "xed within constructs subjected to 20% compressive strain and sectioned in the x/y(D), x/z (E) and y/z (F)planes. All micrographs are at the same magni"cation, scale bar"2 lm. The direction of the applied strain is indicated by the horizontal arrow in (D)and (E) and a crossed circle in (F).

Fig. 7. Perimeter ratios for chondrocytes "xed within unstrainedagarose constructs or constructs subjected to 20% compressive strainand sectioned in the x/y, x/z and y/z planes. Each value represents themean and standard error of the mean for at least 36 replicates from twoseparate experiments. Unpaired student's t-test results indicate di!er-ences from unstrained control values as follows: *"p(0.05.

mean values of nuclear x diameter strain on days 0 and1 were 7.2 and 7.7%, respectively, di!erences which werenot statistically signi"cant (Fig. 9). However, by day3 there was a signi"cant increase in nucleus deformationwith a mean x diameter strain of 10.6%. Nucleus defor-mation was not measured at day 6 since compression atthis time point resulted in minimal cell deformation(Fig. 9).

4. Discussion

The present study has examined the e!ects of compres-sive strain on chondrocytes seeded in agarose and cul-tured, unstrained, for up to 6 days. The elaboration ofextracellular matrix and the remodelling of the cytos-keleton during the culture period was investigated. Twodi!erent microscopic techniques, real-time confocal laserscanning microscopy and transmission electron micros-copy were subsequently employed to elucidate strain


Fig. 8. Confocal sections bisecting the centre of two cells at 0% (A, C) and 20% (B, D) agarose compressive strain. Cells were labelled with Calcein AM(A, B) or Syto 64 (B, C) on day 1 of culture. Scale bar"2 lm. The images represented in (A) and (B) have been subjected to computer thresholding todelineate the cell boundaries.

Fig. 9. Reduction in cell and nuclear diameter parallel to the axis ofcompression for chondrocytes cultured in agarose constructs for up to6 days and subjected to 20% agarose compressive strain. Each valuerepresents the mean and standard error for n"50}100 replicates fromtwo separate experiments.

events at the cellular and sub-cellular levels. Confocallaser scanning microscopy, as used in this study, hastheoretical resolutions of between 0.21 and 0.58 lm in thex- and y-axis and between 0.44 and 3.44 lm in the z-axis,depending on the objective lens chosen (Knight, 1997).

The resolution of the transmission electron microscopysystem used is approximately 0.3 nm. The measurementof live cells by confocal laser scanning microscopy inconjunction with the dyes Calcein AM or Syto-64, for theentire cell and nucleus, respectively, overcomes artefactsassociated with the processing of "xed specimens.

Chondrocytes, freshly isolated from bovine articularcartilage and seeded in agarose were found to be round inpro"le, indicated by a diameter ratio close to unity(Table 2). As the chondrocytes were seeded in agaroseconstructs which were totally uncon"ned, these "ndingssuggest that the cells are spherical in morphology. Overthe 6 day culture period the median diameter of the cellsin unstrained constructs increased from 9.45 to 11.80 lm,an increase of approximately 25%. This "nding probablyre#ects either culture-induced upregulation of cell meta-bolism, with associated increase both in cytoplasmic vol-ume and organelle number and size, and/or an increasein cell volume due to reduced extracellular osmolality(Hall et al., 1996a,b; Lee et al., 1998b). There was nosigni"cant alteration in the diameter ratio over the cul-ture period (Table 2) indicating that the cells remainedspherical in morphology when cultured in unstrainedagarose constructs for up to 6 days.

The synthesis of extracellular matrix was assessed bymeasuring the increase in sulphated glycosaminoglycan


and hydroxyproline. The former, a measure of proteog-lycan content, increased consistently between days 1 and6. By contrast, the latter, a measure of collagen contentexhibited a small increase during the "rst 3 days inculture, followed by a signi"cant increase. Thus duringthe "rst 3 days in culture, elaborated extracellular matrixis predominantly proteoglycanaceous and becomes en-riched with collagen at a later stage. These "ndingssupport recent studies investigating the elaboration ofextracellular matrix by chondrocytes within agarose us-ing either biochemical or immunolocalisation techniques(Chang and Poole, 1997; Chang et al., 1997; Knight et al.,1998). By day 6 the elaborated pericellular matrix, sur-rounding chondrocytes in agarose, is known to exhibitsu$cient mechanical integrity to prevent chondrocytedeformation on application of compressive strain(Knight et al., 1998).

Fluorescent imaging techniques were employed to vis-ualise cytoskeletal organisation within chondrocytesseeded in agarose constructs and cultured for up to6 days and it appeared broadly similar to that describedin situ. (Benjamin et al., 1994; Durrant et al., 1999; Lange-lier et al., 1999). It was evident, however, that withina population of chondrocytes cultured in agarosecytoskeletal components were present at varying degreesof organisation (Fig. 1). It was, therefore, in-appropriate to represent the population by the use ofa single image. Chondrocytes cultured in agarose were,therefore, allocated into three groups according tothe organisation of cytoskeletal features within indi-vidual cells and quanti"ed using a numerical scoringsystem (Fig. 1). Mean population values have aminimum of 0.00 and maximum of 2.00, indicating entirepopulations of nonorganised and clearly organised cells,respectively.

Changes in the cytoskeletal components, associatedwith time in culture are shown in Fig. 2. The organisationof actin mico"laments, which was minimal over the "rst24 h in culture, increased markedly between days 3 and 6.By contrast, microtubule and vimentin intermediate"lamentous organisation was present during the initialculture period but diminished after approximately3 days. The initial presence of microtubules and vimentinintermediate "laments may represent remnants of struc-tures present in situ, which were retained during theisolation procedure. The organisation of microtubulesand vimentin intermediate "laments became more pro-nounced between days 3 and 6. These data indicate thedynamic nature of organisation for all three cytoskeletalcomponents. In addition, an association is suggestedbetween cytoskeletal remodelling and the elaboration ofa pericellular matrix. As has been previously mentioned,cytoskeletal organisation per se may also be involved inmechanotransduction (Franke et al., 1984; Ookawa et al.,1992; Barbee, 1994; Parkkinen et al., 1995; Meazzini etal., 1998). Indeed there is considerable evidence linking

the cytoskeleton with activation of ion channels andintracellular calcium signalling (Janmey, 1998). Thereforethe temporal changes observed in cytoskeletal organisa-tion may be directly associated with changes inmechanotransduction pathways.

The application of incremental compressive strains toagarose block constructs cultured for 1 day and stainedwith Calcein AM indicated a clear relationship betweenthe level of construct strain and deformation of the chon-drocytes within the construct (Fig. 3). Deformation wascharacterised by a decrease in the x-diameter, an increasein the y-diameter and an associated reduction in thediameter ratio.

Using confocal laser scanning microscopy, estimationsof volume and surface area of chondrocytes were deter-mined by extrapolation from a single 2-D scan throughthe centre of individual cells thereby overcoming errorsdue to z-axis resolution and axial distortion in the z-plane associated with 3-D reconstruction (Guilak, 1994;Knight, 1997). The cross-sectional area of cells calculatedusing the equation for an ellipse was not signi"cantlydi!erent from that calculated by the measurement func-tion of the system. The mean diameter ratios of cellswithin semi-thin sections prepared from constructs com-pressed to 20% were identical in the x/y and x/z planesand were approximately 1 in the y/z plane. These "ndingscon"rm that deformation is symmetrical and that thecells adopt an oblate ellipsoid morphology, thereby jus-tifying the use of 3-D extrapolation for estimation ofvolume and surface area from a single 2-D scan throughthe centre of individual cells.

Estimation of volumetric behaviour of cells duringdeformation indicated a slight increase in mean cell vol-ume with increasing construct strain (Fig. 4A), althoughthe Pearson's population coe$cient and the unpairedstudent' t-test indicated that volume changes were notsigni"cant. Chondrocyte deformation was, however, as-sociated with a signi"cant increase in the estimated sur-face area of the cell, with a mean percentage value of4.9% at 15% construct strain (Fig. 4B). These "ndingscontradict a previous study which reported a reductionin volume under compression based on an assumption ofcomplete con"nement in the z axis (Freeman et al., 1994).The present "ndings suggest that deformation is symmet-rical and thus cell volume may have been underestimatedat high strains in the previous study.

Cell membranes are highly resistant to alterations intheir equilibrium surface area due to molecular packingand hydrophobic interactions (Steck, 1989). Indeed themaximum increase in surface area prior to rupture isestimated to be approximately 3% (Evans et al., 1976).Thus it may be postulated that the cell membrane inunstrained chondrocytes is folded in such a manner thatcell deformation induces the folds to unravel. Indeed, thisprocess has been demonstrated to allow chondrocytes toincrease their volume by as much as 220% when placed


in a hypotonic solution, suggesting 200}250% excessmembrane area (Guilak and Ting-Beall, 1999). The res-olution of the confocal laser scanning microscope, how-ever, is insu$cient to resolve ultrastructural detail at thecell surface. This "ne detail could be resolved by usingtransmission electron microscopy, which requires com-plicated processing of specimens which could introduceartefacts. Indeed, there was a progressive reduction inmean chondrocyte diameter due to dehydration-inducedspecimen shrinkage. Diameter ratios were, however,maintained throughout processing. It is reasonable toconclude, therefore, that the transmission electronmicroscopy processing protocol adopted does not in#u-ence the deformation of the chondrocytes, whilst otherartefacts, such as shrinkage, would be similar in bothstrained and unstrained chondrocytes. Thus, the relativechanges in chondrocyte pro"le will be maintained.

Figs. 6A, B and C indicate that the periphery of chon-drocytes in unstrained constructs is characterised by nu-merous short cell processes. Perimeter ratios forunstrained cells were similar in all three planes and maybe pooled to produce a population range of 1.15 to 1.83with a mean of 1.37. Electron micrographs prepared inthe x/y and x/z planes from strained specimens indicatea clear smoothing of the cell periphery (Figs. 6D and E).The corresponding perimeter ratios yielded a combinedpopulation with a mean of 1.19 and a range of 1.09}1.30,which was considerably smaller than that associated withthe unstrained state. Thus, cell deformation in strainedspecimens is associated with an unravelling of membranefolds and localised alterations in the radius of curvatureof the membrane. These data suggest that unstrainedcells exhibiting a large perimeter ratio undergo signi"-cant smoothing when subjected to compressive strain.Micrographs from strained specimens prepared in the y/zplane exhibited processes although their numbers andsize were less than for unstrained cells (Fig. 6F). Theperimeter ratio values, with a range of 1.14 and 1.50 anda mean of 1.27, lay between that for unstrained cells andstrained cells prepared in the x/y and x/z planes. Thussome smoothing had occurred in this plane but to a lesserextent than in the other planes. It is evident that onapplication of compressive strain unravelling of the cellmembrane is heterogeneous in nature and oriented whichmight provide a mechanism by which cells detect theirorientation with respect to the applied strain.

This study indicates that intracellular mechanotran-sduction pathways associated with alterations in cellvolume are unlikely to be activated within chondrocytesembedded in agarose and subjected to mechanical com-pression. This is in contrast to the situation for chon-drocytes in situ. Pathways related to localised membranebending may, however, be activated in chondrocyte inagarose and in situ. In particular stretch-activated ionchannels, sensitive to membrane bending, are known toexist (Bell and Holmes, 1994; Bowman and Lohr, 1996;

Tavernarakis and Driscoll, 1997). Their presence onchondrocytes has yet to be con"rmed although certainforms of stretch sensitive channel have been reported onchondrocytes due to the ability to block cell responsewith gadolinium (Wright et al., 1992,1996; Guilak et al.,1994). Moreover, membrane bending is also likely tocause cytoskeletal deformation mediated via speci"ctransmembrane receptor molecules.

Confocal laser scanning microscopy was used to exam-ine cell and nucleus deformation in isolated chondrocytescompressed in agarose gel after culture for up to 6 days.Chondrocyte deformation was reduced over the 6 dayculture period. These "ndings are in agreement withprevious studies and are believed to be due to the syn-thesis of a mechanically functional matrix which formsa sti! shell around the cell (Lee and Bader, 1995; Knightet al., 1996,1998; Knight, 1997).

Cells labelled with Syto 64, a viable nucleic acid stain,exhibited distinct #uorescent staining of the nucleus withbrighter staining around the nuclear membrane (Figs. 8Cand D). Low-intensity cytoplasmic staining was also ob-served which was useful for monitoring the level of celldeformation. There was considerable variability in theunstrained size and shape of the chondrocyte nuclei con-"rming the necessity of measuring the same nuclei inboth an unstrained and compressed state. Using thisapproach deformation of isolated chondrocytes in com-pressed agarose was associated with nucleus deformationcharacterised by a reduction in nucleus x diameter. At alltime points, the levels of nucleus deformation were lessthat that experienced by the cells suggesting that thenucleus is several times sti!er than the cytoplasm inagreement with previous studies (Guilak, 1995; Maniotiset al., 1997; Jones et al., 1997, 1999).

Nucleus deformation may be mediated through thecytoskeleton. Intermediate "laments in particular areknown to form a link between the cell membrane and thenucleus and are known to interact with sequence-speci"cDNA and histones, potentially causing alterations ingene expression (Traub and Shoeman, 1994). However,the di!erent cytoskeletal components are extensively in-terlinked and thus perturbation of one system may in#u-ence the structure of another. Therefore, the largevariability in nucleus deformation at any time point maybe caused by the observed variability in cytoskeletalorganisation (Figs. 1 and 2). In addition, nucleus defor-mation may also be in#uenced by changes in nuclearmatrix composition, such as hyaluronan and biglycancontent, which are known to occur during the cell cycle(Kan, 1990; Liang et al., 1997).

A signi"cant increase in the levels of nucleus deforma-tion was demonstrated over a 3 day culture period des-pite a reduction in the level of cell deformation (Fig. 9). Inaddition, this study has shown temporal changes incytoskeletal organisation over the same culture period(Fig. 2). Thus it is tempting to suggest that the increased


Table 3Summary of experimental results and possible activation of mechanotransduction pathways

Time in culture

Involvement inmechanotransduction

Day 0 Day 1 Day 3 Day 6

Cell strain ### ### ## #

Volume strain ND * ND NDSurface area strain ND ### ND NDNuclear strain ## ## ### NDProteoglycan ECM * # ## ###

Collagen ECM * * # ##

Micro"laments * # ## ###

Microtubules ## * * ##

Vimentin Intermediate"laments

## # * ##

! No change.ND Not determined.ECM Extracellular matrix.#, ##, ### indicate increasing levels of involvement.

nucleus deformation may be a direct result of alteredcytoskeletal organisation within the cytoplasm.

In summary, the present study demonstrates that theapplication of compressive strain to chondrocytes seededwithin agarose construct induces a number of strainevents at the subcellular level which could potentially actas initiators of intracellular mechanotransduction path-ways (Table 3). In addition it is clear that the agarosesystem is highly dynamic in nature, such that the import-ance of various sub-cellular strain events in mediatingmechanotransduction processes may alter during cul-ture, associated with the elaboration of a cartilaginousextracellular matrix and the remodelling of the cytos-keleton.

Acknowledgements

This work was funded by the Engineering and PhysicalSciences Research Council, UK.

References

Armstrong, C.G., Lai, W.M., Mow, V.C., 1984. An analysis of theuncon"ned compression of articular cartilage. Journal of Bi-omechanical Engineering 108, 165}173.

Barbee, K.A., Davies, P.F., Lal, R., 1994. Shear stress-induced reorgan-ization of the surface topography of living endothelial cells imagedby atomic force microscopy. Circulation Research 74, 163}171.

Bell, J., Holmes, M.H., 1994. A note on modelling mechano-chemicaltransduction with an application to a skin receptor. Journal ofMathematical Biology 32, 275}285.

Benjamin, M., Archer, C.W., Ralphs, J.R., 1994. Cytoskeleton of carti-lage cells. Microbial Research Technology 28, 372}377.

Ben-Ze'ev, A., 1991. Animal cell shape changes and gene expression.BioEssays 13, 207}212.

Bowman, C.L., Lohr, J.W., 1996. Mechanotransducing ion channels inC6 glioma cells. Glia 18, 161}176.

Burton-Wurster, N., Vernier-Singer, M., Farquar, T., Lust, G., 1993.E!ect of compressive loading and unloading on the synthesis oftotal protein, proteoglycan and "bronectin by canine cartilage ex-plants. Journal of Orthopaedic Research 11, 717}729.

Buschmann, M.D., Gluzband, Y.A., Grodzinsky, A.J., Hunziker, E.B.,1995. Mechanical compression modulates matrix biosynthesis inchondrocyte/agarose culture. Journal of Cell Science 108, 1497}1508.

Buschmann, M.D., Hunziker, E., Kim, Y.-J., Grodzinsky, A., 1996.Altered aggrecan synthesis correlates with cell and nucleus structurein statically compressed cartilage. Journal of Cell Science 109,499}508.

Chang, J., Poole, C.A., 1997. Confocal analysis of the molecular hetero-geneity in the pericellular microenvironment produced by adultcanine chondrocytes cultured in agarose gel. Histochemical Journal29, 515}528.

Chang, J., Nakajima, H., Poole, C.A., 1997. Structural colocalisation oftype VI collagen and "bronectin in agarose cultured chondrocytesand isolated chondrons extracted from adult canine tibial cartilage.Journal of Anatomy 190, 523}532.

Durrant, L.A., Archer, C.W., Benjamin, M., Ralphs, J.R., 1999. Articularchondrocytes reorganise their cytoskeleton in response to changingmechanical conditions in organ culture. Journal of Anatomy 194,343}353.

Edwards, C.A., O'Brien, W.D., 1980. Modi"ed assay for determinationof hydroxyproline in a tissue hydrolyzate. Clinica Chemica Acta104, 161}167.

Errington, R.J., Fricker, M.D., Wood, J.L., Hall, A.C., White, N.S.,1997. Four dimensional imaging of living chondrocytes in cartilageusing confocal microscopy: a pragmatic approach. American Jour-nal of Physiology 272, 1040}1051.

Evans, E.A., Waugh, R., Melnik, N., 1976. Elastic area compressibilitymodulus of red cell membrane. Biophysical Journal 6, 585}595.

Farndale, R.W., Sayer, C.A., Barrett, A.J., 1982. A direct spectro-photometric microassay for sulfated glycosaminoglycans in carti-lage culture. Connective Tissue Research 9, 247}248.


Franke, R.P., Grafe, M., Schnittler, H., Sei!ge, D., Mittermayer, C.,Drenckhahn, D., 1984. Induction of human vascular endothelialstress "bres by #uid shear stress. Nature 307, 648}649.

Freeman, P.M., Natarajan, R.N., Kimura, J.H., Andriacchi, T.P., 1994.Chondrocyte cells respond mechanically to compressive loads.Journal of Orthopaedics Research 12, 311}320.

Gray, M.L., Pizzanelli, A.M., Grodzinsky, A.J., Lee, R.C., 1988. Mech-anical and physiochemical determinants of the chondrocyte biosyn-thetic response. Journal of Orthopaedics Research 6, 777}792.

Gray, M.L., Pizzanelli, A.M., Lee, R.C., Grodzinsky, A.J., Swann, D.A.,1989. Kinetics of the chondrocyte biosynthetic response to compres-sive load and release. Biochimica et Biophysica Acta 991, 415}425.

Guilak, F., 1994. Volume and surface area measurement of viablechondrocytes in situ using geometric modelling of serial confocalsections. Journal of Microscopy 173, 245}256.

Guilak, F., 1995. Compression-induced changes in the shape and vol-ume of the chondrocyte nucleus. Journal of Biomechanics 28,1529}1541.

Guilak, F., Donahue, H.J., Zell, R., Grande, D., McLeod, K.J, Rubin,C.T., 1994. Deformation induced calcium signalling in articularchondrocytes. In: Mow, V.C., Hochmuth, R.M., Guilak, F., Tran-Son-Tay, R. (Eds.), Cell Mechanics and Cellular Engineering.Springer, New York, pp. 380}397.

Guilak, F., Ratcli!e, A., Mow, V.C., 1995. Chondrocyte deformationand local tissue strain in articular cartilage: a confocal microscopystudy. Journal of Orthopaedics Research 13, 410}421.

Guilak, F., Ting-Beall, H.P., 1999. The e!ects of osmotic pressure on theviscoelastic and physical properties of articular chondrocytes.ASME Advances in Bioengineering, in press.

Hall, A.C., 1995. Volume-sensitive transport in bovine articular chon-drocytes. Journal of Physiology (London) 484, 755}766.

Hall, A.C., Horowitz, E.R., Wilkins, R.J., 1996a. The cellular physiologyof articular cartilage. Experimental Physiology 81, 535}545.

Hall, A.C., Starks, I., Soults, C.L., Rashidbigi, S., 1996b. Pathways forK# transport across the bovine articular chondrocyte membraneand their sensitivity to cell volume. American Journal of Physiology270, 1300}1310.

Ingber, D.E., 1991. Integrins as mechanochemical transducers. CurrentOpinions in Cell Biology 3, 841}848.

Ingber, D., Dike, L., Hansen, L., Karp, S., Liley, H., Maniotis, A.,McNamee, H., Mooney, D., Plopper, G., Sims, J., Wang, N.,1994. Cellular tensegrity: Exploring how mechanical changes in thecytoskeleton regulate cell growth, migration and tissue patternduring morphogenesis. International Review Cytology 150,173}224.

Janmey, P.A., 1998. The cytoskeleton and cell signalling: componentlocalization and mechanical coupling. Physiology Review 78,763}781.

Jones, W.R., Ting-Beall, H.P., Lee, G.M., Kelley, S.S., Hochmuth, R.M.,Guilak, F., 1997. Mechanical properties of human chondrocytesand chondrons from normal and osteoarthritic cartilage. Transac-tion of the Orthopaedics Research Society 22, 199.

Jones, W.R., Ting-Beall, H.P., Lee, G.M., Kelley, S.S., Hochmuth, R.M.,Guilak, F., 1999. Alterations in the Young's modulus and volumet-ric properties of chondrocytes isolated from normal and osteo-arthritic human cartilage. Journal of Biomechanics 32, 119}127.

Kan, F.W., 1990. High-resolution localization of hyaluronic acidin the golden hamster oocyte-cumulus complex by use of ahyaluronidase-gold complex. Anatamical Record 228, 370}382.

Kim, Y.J., Sah, R.L.Y., Grodzinsky, A.J., Plaas, A.H.K., Sandy, J.D.,1994. Mechanical regulation of cartilage biosynthetic behaviour:physical stimuli. Archives of Biochemistry and Biophysics 311,1}12.

Kiviranta, I., Tammi, M., Jurvelin, J., Saamanen, A.-M., Helminen, H.J.,1988. Moderate running exercise augments glycosaminoglycans andthickness of articular cartilage in the knee joint of young beagledogs. Journal of Orthopaedics Research 6, 188}195.

Knight, M., 1997. Deformation of isolated articular chondrocytescultured in agarose constructs. In: Ph.D Thesis. University of Lon-don.

Knight, M., Lee, D.A., Bader, D.L., 1996. Distribution of chondrocytedeformation in compressed agarose using confocal microscopy.Cellular Engineering 1, 97}102.

Knight, M.M., Lee, D.A., Bader, D.L., 1998. The in#uence of elaboratedpericellular matrix on the deformation of isolated chondrocytescultured in agarose. Biochemica Biophysica Acta 1405, 67}77.

Langelier, E., Suetterlin, R., Aebi, U., Buschmann, M.D., 1999.Zonal dependence of the chondrocyte cytoskeleton and in vitroresponse to load. Transaction of Orthopaedics Research Society 24,631.

Lee, D.A., Bader, D.L., 1995. The development and characterisation ofan in vitro system to study strain-induced cell deformation inisolated chondrocytes. In vitro Cell Development Biology inAnimals 31, 828}835.

Lee, D.A., Bader, D.L., 1997. Compressive strains at physiologicalfrequencies in#uence the metabolism of chondrocytes seeded inagarose. Journal of Orthopaedics Research 15, 181}188.

Lee, D.A., Frean, S.P., Lees, P., Bader, D.L., 1998a. Dynamic mechan-ical compression in#uences nitric oxide production by articularchondrocytes seeded in agarose. Biochemical and Biophysical Re-search Communications 251, 580}585.

Lee, D.A., Noguchi, T., Knight, M.M., O'Donnell, L., Bentley, G.,Bader, D.L., 1998b. The response of chondrocyte sub-populationscultured within unloaded and loaded agarose. Journal of Ortho-paedics Research 16, 726}733.

Liang, Y., Haring, M., Roughley, P.J., Margolis, R.K., Margolis, R.U.,1997. Glypican and biglycan in the nuclei of neurons and gliomacells: presence of functional nuclear localization signals and dy-namic changes in glypican during the cell cycle. Journal of CellBiology 139, 851}864.

Maniotis, A.J., Chen, C.S., Ingber, D.E., 1997. Demonstration of mech-anical connections between integrins, cytoskeletal "laments andnucleoplasm that stabilize nuclear structure. Proceedings of theNational Academy of Science of the United States of America 94,849}854.

Meazzini, M.C., Toma, C.D., Scha!er, J.L., Gray, M.L., Gerstenfeld,L.C., 1998. Osteoblast cytoskeletal modulation in response to mech-anical strain in vitro. Journal of Orthopaedics Research 16,170}180.

Ookawa, K., Sato, M., Ohshima, N., 1992. Changes in the microstruc-ture of cultured porcine aortic endothelial cells in the early stageafter applying a #uid-imposed shear stress. Journal of Biomechanics25, 1321}1328.

Palmoski, M.L., Brandt, K.D., 1982. Immobilisation of the knee pre-vents osteoarthritis after anterior cruciate ligament transection.Arthritis and Rheumation 25, 1201}1208.

Palmoski, M.L., Brandt, K.D., 1984. E!ects of static and cyclic com-pressive loading on articular cartilage plugs in vitro. Arthritis andRheumatism 27, 675}681.

Palmoski, M.J., Perricone, E., Brandt, K.D., 1979. Development andreversal of a proteoglycan aggregation defect in normal canine kneeafter immobilisation. Arthritis and Rheumatism 22, 508}517.

Parkkinen, J.J., Lammi, M.J., Helminen, H.J., Tammi, M., 1992. Localstimulation of proteoglycan synthesis in articular cartilage explantsby dynamic compression in vitro. Journal of Orthopaedics Research10, 610}620.

Parkkinen, J.J., Lammi, M.J., Inkinen, R., Jortikka, M., Tammi, M.,Virtanen, I., Helminen, H.J., 1995. In#uence of short term hydros-tatic pressure on organisation of stress "bres in cultured chon-drocytes. Journal of Orthopaedics Research 13, 495}502.

Quinn, T.M., Grodzinsky, A.J., Buschmann, M.D., Kim, Y.J., Hunziker,E.B., 1998. Mechanical compression alters proteoglycan depositionand matrix deformation around individual cells in cartilage ex-plants. Journal of Cell Science 111, 573}583.


Reynolds, E.S., 1963. The use of lead citrate at high pH as an electron-opaque stain for electron microscopy. Journal of Cell Biology 17,208.

Sah, R.L.Y., Doong, J.Y.H., Grodzinsky, A.J., Plaas, A.H.K., Sandy,J.D., 1991. E!ects of compression on the loss of newly synthesisedproteoglycans and proteins from cartilage explants. Archives ofBiochemistry and Biophysics 286, 20}29.

Sah, R.L.Y., Kim, Y.J., Doong, J.Y.H., Grodzinsky, A.J., Plaas,A.H.K., Sandy, J.D., 1989. Biosynthetic response of cartilage ex-plants to dynamic compression. Journal of Orthopaedic Research 7,619}636.

Steck, T.L., 1989. Red cell shape. In: Stein, W.D., Bronner, F. (Eds.), Cellshape: Determinants, Regulation and Regulatory Role. AcademicPress, London, pp. 205}246.

Tavernarakis, N., Driscoll, M., 1997. Molecular modelling ofmechanotransduction in the nematode caenorhabditis elegans. An-nual Review of Physiology 59, 659}689.

Traub, P., Shoeman, R.L., 1994. Intermediate "lament proteins: cytos-keletal elements with gene regulatory function?. International Re-view of Cytology 154, 1}103.

Urban, J.P., 1994. The chondrocyte: a cell under pressure. BritishJournal of Rheumatology 33, 901}908.

Urban, J.P., Hall, A.C., Gehl, K.A., 1993. Regulation of matrix synthesisrates by the ionic and osmotic environment of articular chon-drocytes. Journal of Cell Physiology 154, 262}270.

Wang, N., Butler, J.P., Ingber, D.E., 1993. Mechanotransduction acrossthe cell surface and through the cytoskeleton. Science 260,1124}1127.

Wong, M., Wuethrich, P., Eggli, P., Hunziker, E.B., 1996. Zone-speci"ccell biosynthetic activity in mature bovine articular cartilage: a newmethod using confocal microscopic stereology and quantitativeautoradiography. Journal of Orthopaedics Research 14, 424}432.

Wong, M., Wuethrich, P., Buschmann, M.D., Eggli, P., Hunziker, E.B.,1997. Chondrocyte biosynthesis correlates with local tissue strain instatically compressed adult articular cartilage. Journal of Ortho-paedics Research 15, 189}196.

Wright, M.O., Stockwell, R.A., Nuki, G., 1992. Response of plasmamembrane to applied hydrostatic pressure in chondrocytes and"broblasts. Connective Tissue Research 28, 49}70.

Wright, M.O., Jobanputra, P., Bavington, C., Salter, D.M., Nuki, G.,1996. E!ects of intermittent pressure-induced strain on the elec-trophysiology of cultured human chondrocytes: evidence for thepresence of stretch-activated membrane ion channel. ClinicalScience 90, 61}71.


Documents

Chondrocyte deformation within compressed agarose constructs at the cellular and sub-cellular levels