- Email: [email protected]
A new method based on contact surface profilometry for quantitative measurement of resorbed bone volume
Technical Note
Physica Medica - Vot. XXI, N. 1, January-March 2005
A n e w m e t h o d based on contact surface profilometry for quantitative m e a s u r e m e n t of resorbed bone v o l u m e Franco Fusi, t Luca Mercatelli,2 Venere Basile, LM a u r o Pucci, 2Salvatore Siano, 3 Pietro A n t o n i o Bernabei, 4 Monica Monici 5 1. 2. 3. 4. 5.
Dept. of Clinical Physiopathology, Medical Physics Sect. University of Florence, 89 Pieraccini 6, I 50139 Florence (Italy) INOA - Nm ional Inst. for Applied Optics, Largo Enrico Fermi 6, 1 50125 Florence (Italy) Inst. of A!,plied Physics ,Nello Carrara,,, cue, Via Madonna del Piano, 1 50019 Sesto Eno, Florence (Italy) Haematcl,:gy Division, Careggi Hospital, Viale Morgagni 45, 1 50134, Florence (Italy) cEo - Ce~; ".r of Excellence in Optronics, Largo Enrico Fermi 6, I 50125 Florence (Italy)
Manuscript received: April 13, 2004; revised: December 28, 2004 Accepted lk c publication: December 30, 2004
Abstract Bone is a dynamic tissue. Its continuous remodeling depends on the balance between bone formation and bone resorption. These two processes are carried out by specialized cells called osteoblast and ostreoclast respectively. The osteoclastic bone resorption consists in clegradation of the mineral and collagen components of bone. The stud,, of bone turnover requires accurate assessment of osteoclastic bone resorption, that becomes even more important in pathologic bone loss due to the uncoupling between bone formation and bone resorption. Osteoclastic activity is difficult to measure. Many techniques, generally based on the detection of the resorbing lacunae (pits) due to the bone degradation, allow to estimate bone resorption, but none of them quantitatively and directly measures the volume of resorbed bone. We propose a reliable and relatively simple method, based on contact surface profilometry, to evaluate directly and quantitatively the volume of resorbed bone. The method has the following advantages: i. to perf:wm a comparison of the same bone surface before and after the exposure to the osteoclastic activity;
ii. to enhance the sensitivity by utilization of bone slices shaped and polished in order to concentrate the cell activity in a controlled alrea. KEYWORD.*: Surface profilometry, bone resorption. 1. INTROD UCTION
The stud,r of bone turnover and bone loss, due to physiological or pathological conditions [1], requires accurate assessment of osteoclastic bone resorptio~t. Osteoclasts are cells of haemopoietic origin that, by a differentiation process, assume the ability to resorb bone [z~. Bone resorption is a multi-step process consisting in the adhesion of the osteoclast to the bone surface, cell polarization and formation of a subcellular c o m p a r t m e n t in which the degradation of both mineral and collagen components of bone occurs [3, 4]. BoJm erosion by osteoclastic cells is expected to induce surface texture variations at a microscopic scale, leaving resorption pits and exposed collagen fibrils. Osteodastic activity is difficult to measure and many techniques have been used in order to estimate bone resorption, with very different results and reference values. Histomorphometric techniques often report the percentage of bone surface covered by multinuclear giant cells with morphological appearance of mature resorbing osteoclasts. These techniques show significan': variations mostly arising from the methods used for osteoclast identification. They vary from put,ely morphological techniques [5, 6, 7, 8]
to enzymatic methods based on acid phosphatase expression [91. The methods based on the analysis of bone surface characteristics consider areas showing multinucleated cells associated with Howship's lacunae (pits) [10]. Pit assays are the more diffuse methods for the evaluation of in vitro bone resorption [11, 12, 133. In the pit assay, bone resorption is estimated by measuring the pit areas formed by osteoclasts cultured on bone slices [14]. This survey is error-prone since the recognition of the pits is difficult, the lacunae depth is often disregarded and it is not possible to perform comparative measures on the same slide before and after exposure to the osteoclast activity. Scanning electron microscopy (StU), that is widely used to reveal the presence of the pits, can provide a detailed inspection of the bone surface. However, since a conductive coating of the samples is needed for SEM observation, comparative measures on the same surface, before and after exposure to osteoclastic cells, are not possible. Several methods based on optical microscopy have been considered in order to analyse pit shape and dimensions, thus assessing bone resorption. Walsh proposed a technique based on wide-field microscopy and standard latex spheres used as scale markers [151. The profile of reflective surfaces was recovered by illumination with rapid
Address I:orcorrespondence: Franco Fusi: e-marl: [email protected] 41
E Fusi et alii: A new method based on contact surface profilometry
scanning of laser beam [16J. Specimens with fiat surfaces have been analysed by interference fringe m e t h o d too [17]. In practice, the use of this m e t h o d for the assessment of bone resorption is limited owing to the very small depth variations (a few wavelengths) measured. However, the determination of bone resorption by optical microscopy techniques presents the problem of very high light scattering and diffusion due to the irregularity of bone surface and partial transparency to visible light. Moreover, the use of chromophores is an impediment for comparative measures on the same sample. Recently, other advanced surface techniques have been used to inspect bone surface: x-ray photoelectron spectroscopy (xvs), to determine the chemical composition of the bone surface, and optical profilometry, to determine surface roughness [18]. Finally, biochemical evaluations of in vitro bone resorption have been proposed as well. They consist in measuring the a m o u n t of collagen degradation products released in the culture m e d i u m during the bone resorbing process [11, 19, 20, 21] Although all these methods assess osteoclastic bone resorption, none of t h e m allow to measure directly the volume of bone eroded by osteoclast activity. We propose a reliable method, based on contact surface profilometry, to evaluate directly and quantitatively the volume of resorbed bone. In this m e t h o d contact profilometry is used not only to detect surface roughness but also to measure with accuracy the volume of degraded bone tissue, through the comparison of surface profiles monitored on the same bone surface before and after the exposure to the osteoclastic activity. The m e t h o d is also very sensitive, due to the utilization of bone slices shaped and polished in order to concentrate the cell activity in a controlled area. 2. MATERIAL AND METHODS
2. 1. Bone slices Calf leg bone was treated as described elsewhere [22]. Small rectangular slices of bone were cut using a band saw. They were reduced by abrasion to a size of 4 m m x 10 m m and 2 m m thick, taking care to avoid any w a r m i n g of the samples that could change the bone structure. The slice surfaces were slightly polished with emery to obtain uniform roughness. A final polishing treatment was performed in order to have a preferential roughness pattern, using only translational movements parallel to the longest side of the slices. Two cross-shaped ruts with U profile (width g 200 btm and depth g 100 btm) were excavated on one of the larger surfaces of each slice. The ruts had the double function of gathering the cells, thus circumscribing the resorption activity, and representing a geometrical reference for the m e a s u r e s (FIG. 1). It is known that cells 42
FIG. 1. Schematic representation of a bone slice with crossshaped ruts. The slice surfaces were polished with emery in order to obtain a uniform and low roughness number. tend to accumulate close to, or inside any surface discontinuity, t h u s / h e n c e we expected to find an enhancement of resorption inside the ruts. These were dug utilizing a small disk saw with carborundum blade 200 g m thick. The depth of the ruts was about 100 g m to allow a good scan by the profilometer pick-up. This is essential to set the starting point for any profile measurement in order to analyse the same bone area before and after any cell treatment.
2. 2. Osteoclastic cells and preparation of the samples Cell populations derived from the FLG 29.1 h u m a n cell line were used. The FLG 29.1 cell line, derived and stabilized from a patient suffering from acute monoblastic leukemia (FAB M5a), was characterized as a model of osteoclastic precursor [23]. Cells were cultured in RPMI 1640 m e d i u m supplem e n t e d with 10% heat-inactivated fetal calf serum (FCS) and split twice weekly. Cultures were carried out in a fully humidified atmosphere, at 37~ and 5% CO2. Viability was determined by blue trypan dye exclusion. Cells were treated in order to induce differentiation toward the osteoclastic pathway, as elsewhere described [22, 24]. 15 culture flasks, each containing a bone slice, were prepared with a cell density of 3 x 105. Control flasks containing bone slices soaked in culture m e d i u m only were prepared too. The bone slices were maintained in the culture flasks for 72 h, then washed with distilled water, sonicated to remove adherent cells, washed again and dried at r o o m temperature before the surface analysis by contact profilometry.
2. 3. Bone surface analysis The survey of bone surface was performed by using a commercial microprofilometer (Hommelwerke GmbH) equipped with a pick-up terminated by a
Physica Medica 9Vol. XXI, N. 1, January-March 2005 diamond s @us (whose tip has a radius o f 5 ~tm and a full angh: o f 90~ During the surface scanning, the differential vertical displacement y (x) of the stylus respect to a conventional zero value was converted into an eleztric signal, then amplified, digitalised and analysed in order to achieve a n u m b e r o f roughness parameters according to the international standards (ISO). The intrinsic vertical resolution depends on the analog to digital converter (aDC) and on the measuring range. The instrument was equipped with a 12 bit ADC providing a sub-micron resolution (measuring range/4096). After a preliminary optimisation o f the method, we selected the main roughness parameters (the average roughness Ra and the root-mean-square profile height R~) which resulted as the most indicative for the specii~c surface texture modifications associated with bone resorption by osteoclastic cells. T h e y are defined in the following way:
1s L~ R =T2 ly (x)ldx
form error and waviness from the roughness. The cutoff frequency of the low pass filter (according tO DIN 4777) was set at 1/6 o f the scanned profile. The best surface lying on the profiles was calculated and, consequently, the volume of the rut was determined. The volume was calculated as the space portion delimited by the rut surface and the best fitting plane of the slice surface on which the horizontal paths o f the profiles lied. In Figure 2 and Figure 3, an example o f a single profile as well as a tridimentional (3-D) reconstruction of the rut surface are shown. A suitable software (working on MatLab platform) was developed to p e r f o r m these calculations. Profiles o f the bone slices were recorded before and after the exposure to osteoclastic cell activity, and then compared in order to measure changes in surface roughness and rut volume. At least three measures were made on each
q0
iq:(@mfoL'(y(X)) 2 d.x)1/2 where L~ is the length o f assessment. The fi)J:mer represents the average deviation o f the proti!e from the m e a n line, i.e., the average distance fi'om the vertical displacement to the m e a n line over the length of assessment, while the latter is the average o f the square deviation from the m e a n line. R a is suitable to m o n i t o r gradual changes. It does not differentiate b e t w e e n peaks and valleys and is weakly sensitive to surface defects, thus ensuring high repeatability o f the m e a s u r e m e n t s . Rq is sensitive 1:o peaks and valleys and gets meaningful w h e n lo0~:ing at the profile as a statistical function. It represents the standard deviation o f the profile height distribution. The m e a s u r e m e n t s were p e r f o r m e d positioning the sample on a specially designed holder, mechanically connected to a micrometric slit which could be moved perpendicularly to the scanning direction. T w e n t y o n e roughness profiles, with a spacing o f 200 _+ 5 ttm (5 ~m is the reading precision o f the micrometric slit), were acquired for each slice. The first profile was taken at a distance of 100 _+ 5 }am from the edge of the short rut (see Figure 1). Profile paths (1000 _+ 2 g m length) were parallel to the short rut and perpendicular to the long one. The scanned plane surface on each slice resulted a rectangular area o f 4000 _+ 10btm • 1000 __ 2~m, that is 4 m m z. The scanning parameters were decided according to both the accuracy of the measurements and the heavy computation. To optimise both S / N ratio and resolution a scan speed o f 0.05mm/s was selected. A Gaussian low pass filter was used to separate long wavelengths like form,
8O
001
lenghtDin]
Fro. 2. (Top) Singles 2-D profiles of the same bone slice recorded before and after (thick line) exposure to FLC 29.1 cells activit3~ After resorption the shape of the profile is nearly mantained but rut's depth is increased. (Bottom) Microscopic images, obtained with grazing lighting, of a bone slice recorded before (right) and after (left) exposure to differentiated F~C 29.1 cells. A detail of the bottom of a rut is imaged. 43
E Fusi et alii: A new method based on contact surface profilometry
6.
FIG. 3. Tridimensional reconstruction of the rut before (top) and after (bottom) cell activity Bone resorption is evident inside the rut: the particular shows how peaks and discontinuities appear smoothed as a consequence of cell activity
slice. The difference between a single measurement of the rut volume and the average value was found to be about 2%. The reproducibility of the single profile area (a section of the rut volume) was suitable being the max error from the average less than 1%. Surface analysis by tridimentional digital widefield microscopy [25] and SElVtwas performed to have visual evidence of osteoclastic cell activity on the bone slices and, in particular, in the area scanned by surface profilometry. A grazing lighting was necessary to evaluate the surface roughness by imaging. The imaged zone was lighted by an optical fibre placed over the long rut. A home-made holder for slices and optical fibre was coupled to the microscope to attain good reproducibility of lighting and consequently of the imaging. 3. RESULTS AND DISCUSSION
After a 72 h exposure to the cell activity, the bone slices were prepared as above described and their surface was tested by contact profilometry. For each slice, the surface profile obtained was compared with that monitored before the exposure in the culture flasks. The control samples were treated in the same manner. The average roughness Ra and the root-meansquare profile Rq were calculated, for each profile, with reference to the horizontal part of the profile beside the rut. Because the main surface of the slices are not perfectly parallel each other, the main line 44
of the single profile was rotated to superimpose the horizontal zero line. For each bone R a and Rq were calculated as the average of the 21 profle values. The analysis of the planar surface of both slices exposed to cell activity and controls failed to demonstrate any statistical difference in the average roughness Ra and in the root-mean-square profile Rq, monitored before and after the permanence in the flasks. This result was expected because of the presence, on the slice surface, of ruts in which the cells tend to gather and concentrate their activity. On the contrary, the volumes of the ruts obtained by surface profilometry after the exposure to osteoclastic cells appeared larger than those obtained before the treatment. A significant increase was evidentiated in rut's depth by analysing singles 2-D profiles (FIG. 2), while less important changes were found in the width of the rut. The volume of resorbed bone was calculated by comparing the surface profiles monitored before and after cell exposure. The mean value was 2_+ 0.5 p m 3 per pm2with m a x i m u m values up to 20_+0.5 btm3 per p m 2 on the b o t t o m of the ruts. The systematic error in volume determination was calculated to be about 0.2%. The m a x i m u m deviation from average (3 determination for each slice) was less than 1%. The rut volume of the control slices was found to be not significatively different before and after the microgravity treatment because the differences were comparable to the reproducibility error
Physica Medica 9 Vol. XXI, N. 1, January-March 2005
(0.6_+0.5 a m 3 per p m 2, a value very close to that measured ,m the planar surface). These reaults are comparable with data reported in literature. Boyde and Jones (26) summarised and compared ,:he different p h o t o g r a m m e t r i c m e t h o d s used to study shape and dimensions of Howship's lacunae in order to extrapolate the volume o f resorbed bol,e. Laser scanning confocal microscopy resulted fine m o r e reliable method. Measures obtained by ~:onfocal microscopy revealed a depth o f the pits tanging from 5% to 20% o f their diameter. A mear~ resorbed volume of 500 p m 3 for a pit of 20 pm width was estimated. The results we obtained witti the technique here presented are consistent with this value. The m e a n resorbed volume measured with our m e t h o d appears overestimated because the determination was made in a pit-enriched area (the b o t t o m of the rut) with the purpose of increasing the sensitivity. Contrariwise, the other methods generally consider larger areas where the pits cover about 5% o f the surface only. After ttl,: assay by surface profilometry, the slices were obse~':ved by optical microscopy (FIG. 2, bottom) and SEM (FIG. 4). Both the assays gave evidence o f osteoclastic cell activity inside the ruts o f the b o n e slices kept in the culture flasks containing differentiated FLG 29.1 cells, thus supporting the results obtained by profilometry. In fact, the observed changes in rut vo3umes are consistent with the presence of pit-enriched areas on the b o t t o m o f the ruts of the bone slice!; exposed to the cell cultures.
Fro. 4. Scanning Electron Microscopy of a bone slice exposed to FLC 29.1 cells cultured in modelled hypogravity. m e a s u r e m e n t s over the same b o n e sample before and after the exposure to the cell activity. Another important feature o f the m e t h o d is the possibility to increase the cell density in a limited area o f the slices (the b o t t o m of the ruts) and, consequently, to concentrate the resorption activity, thus enhancing the signal-to-noise ratio and the sensitivity. BIBLIOGRAPHY
[1]
[2] [3] [4]
4. CONCI,1 SIONS
There are m a n y techniques allowing the study o f b o n e surface modification to evidentiate resorption pits. Scan1:ing electron microscopy, light microscopy, many kinds o f profilometry (mechanical, laser and interfi:.rometric) are the most c o m m o n . As revealed from literature, all of them give information about the area of resorption pits, often completely disregarding the depth, thus evaluating only the percentage a~ea of degraded b o n e and not supplying direct and quantitative measurements of resorbed b o n e volume. The m o r e advanced m e t h o d s cited in the introduction have the potentiality for estimating the resorbed b o n e with accuracy but, in practice, their utili>.ation appears difficult. In fact, they generally w o r k properly only w h e n applied on unscattering surf~tces, require time-consuming procedures and computationally intensive processes. We approached the p r o b l e m o f bone resorption assessment addressing to the contact microprofilomel:ry because it ensures, in our opinion, the better c o m p r o m i s e a m o n g accuracy, scanning speed and. easy use. The m e t h o d we propose is relativel) 5;imple and able to p e r f o r m directly quantitative measures o f resorbed bone. This is mainly due to th(:- possibility o f carrying o u t profilometric
[5]
Pederson L, Winding B, Foged N T, Spelsberg T C, Oursler M J. Identification of breast cancer cell line-derived paracrine factors that stimulate osteoclast activity. Cancer Res 1999:59 (22); 5849-5855. Suda T, Takahashi N, Martin TJ. Modulation of osteoclast differentiation. Endocr Rev 1992: 13; 66-80. Roodman G D. Cell biology of the osteoclast. Exp Hematol 1999:27 (8); 1229-1241. Everts V, Delaisse J M, Korper W,, Niehof A, Vaes G and Beertsen W. Degradation of collagen in the boneresorbing compartment underlying the osteoclast involves both cysteine-proteinases and matrix metalloproteinases. J Cell Physiol 1992: 150; 221-231. Wronski T J, Lowry P L, Walsh C C, Ignaszewski L A. Skeletal alterations in ovariectomized rats. Calcify Tissue Int 1985: 37; 324-328.
[6]
[7]
[8]
[9]
Shen V, Birchman R, Lindsay R, Dempster D W. Shortterm changes in histomorphometric and biochemical turnover markers and bone mineral density in oestrogen and/or dietary calcium-deficient rats. Bone 1995: 16; 149-156. Westerlind K C, Wronski T J, Pitman E L, Luo Z P, An K N, Bell N H, Turner R T. Estrogen regulates the rate of bone turnover but bone balance in ovariectomized rats is modulated by prevailing mechanical strain. Proc Natl Acad Sci USA 1997: 94; 4199-4204. Bagi C M, Miller S C. Comparison of osteopenic changes in cancellous bone induced by ovariectomy and/or immobilization in adult rats. Anat Rec 1994: 239; 243254. Sims N A, Morris H A, Moore R J, Durbridge T C.
Increased bone resorption precedes increased formation in the ovariectomized rat. Calcif Tissue Int 1996: 59; 121-127.
Chen M M, Jee W S S, Ke H Z, Lin B Y, Li Q N, Li X J. Adaptation of cancellous bone to ageing and immobilization in growing rats. Anat Rec 1992: 234; 317-334. [11] Foged N T, Delaisse J M, Hou P, Lou H, Sato T, Winding B, Bonde M. Quantification of the collagenolytic activity of isolated osteoclasts by enzyme-linked immunosorbent assay. J Bone Miner Res 1996: 11; 226-237. [10]
45
E Fusi et alii: A new method based on contact surface profilometry [12]
[13]
[14] [15]
[16]
[17] [18]
[19]
[20]
46
Sugawara K, Hamada M, Hosoi S, Tamaoki T. A useful method to evaluate bone resorption inhibitors, using osteoclast-like multinucleated cells. Anal Biochem 1998: 255; 204-210. Lorget F, Kamel S, Mentaverri R, Wattel A, Naassila M, Maamer M, Brazier M. High extracellular calcium concentrations directly stimulate osteoclast apoptosis. Biochem Biophys Res Commun 2000: 268; 899-903. Chambers T J, Revell P A, Fuller K, Athanasou N A. Resorption of bone by isolated rabbit osteoclasts. J Cell Sci 1984: 66; 383-399. Walsh C A, Beresford J N, Birch M A, Boothroyd B, Gallagher J A. Application of reflected light microscopy to identify and quantitate resorption by isolated osteoclasts. J Bone Miner Res 1991: 6; 661-671. Moss J P, Linney A D, Grindrod L D, Mosse C A. A laser scanning system for the measurement of facial surface morphology. Optics and Laser in Engineering 1989: 10; 179-190. Virdee M S. Nanometrology of optical fiats by laser autocollimation. Surface Topography 1988: 1; 415-425. Schwartz Z, lohmann C H, Wieland M, Cochran D L, Dean D D, Textor M, Bonewald L E Boyan B D. Osteoblast proliferation and differentiation on dentin slices are modulated by pre-treatment of the surface with tetracycline or osteoclasts. J Periodontol 2000: 71; 586-597. Breuil V, Cosman E Stein L, Horbert W, NievesJ, Shen V, Lindsay R, Dempster D W. Human osteoclast formation and activity in vitro: effects of alendronate. J Bone Miner Res 1998: 13; 1721-1729. James I E, Lark M W, Zembryki D, Lee-Rykaczewski E V, Hwang S M, Tomaszek T A, Belfiore R Gowen
[21]
[22]
[23]
[24]
[25]
[26]
M. Development and characterization of a human in vitro resorption assay: demonstration of utility using novel antiresorptive agents. J Bone Miner Res 1999: 14; 1562-1569. Lorget E Mentaverri R, Meddah B, Cayrolle G, Wattel A, Morel A, Schecroun N, Maamer M, de Vernejoul M C, Kamel S, Brazier M. Evaluation of in vitro bone resorption: High-performance liquid chromatography measurement of the pyridinolines released in osteoclast cultures. Anal Biochem 2000: 284; 375-381. Monici M, Agati G, Fusi F, Paglierani M, Cogoli A, Bernabei P A. Gravitational unloading induces osteoclast-like differentiation of FLC 29.1 cells. J Gravitational Physiology 2002:9 (1); 261-262. Gattei V, Bernabei P A, Pinto A, Bezzini R, Ringressi A, Formigli L, Tanini A, Attadia V, Brandi M L. Phorbol ester induced osteoclast-like differentiation of a novel human leukemic cell line (FLG 29.1). J Cell Biol 1992: 116; 437-447. Monici M, Agati G, Fusi F, Pratesi R, Paglierani M, Santini V, Bernabei P A. Dependence of leukemic cell autofluorescence patterns on the degree of differentiation. Photochem Photobiol Sci 2003: 2; 981-987. Fusi E Monici M, Agati G. Wide-field autofluorescence microscopy for the imaging of living cells in Current laser and optical techniques in biology. Palumbo G, Pratesi REds. London. The Royal Chemical Society, 2004; 357-374. lSBN 0 85404 321 7. Boyde A, Jones S J. Really close up! Surveying surfaces at sub-micrometre resolution: the measurement of osteoclastic resorption lacunae. Photogrammetric Record 1992:14 (79); 59-84.
Physica Medica - Vot. XXI, N. 1, January-March 2005
A n e w m e t h o d based on contact surface profilometry for quantitative m e a s u r e m e n t of resorbed bone v o l u m e Franco Fusi, t Luca Mercatelli,2 Venere Basile, LM a u r o Pucci, 2Salvatore Siano, 3 Pietro A n t o n i o Bernabei, 4 Monica Monici 5 1. 2. 3. 4. 5.
Dept. of Clinical Physiopathology, Medical Physics Sect. University of Florence, 89 Pieraccini 6, I 50139 Florence (Italy) INOA - Nm ional Inst. for Applied Optics, Largo Enrico Fermi 6, 1 50125 Florence (Italy) Inst. of A!,plied Physics ,Nello Carrara,,, cue, Via Madonna del Piano, 1 50019 Sesto Eno, Florence (Italy) Haematcl,:gy Division, Careggi Hospital, Viale Morgagni 45, 1 50134, Florence (Italy) cEo - Ce~; ".r of Excellence in Optronics, Largo Enrico Fermi 6, I 50125 Florence (Italy)
Manuscript received: April 13, 2004; revised: December 28, 2004 Accepted lk c publication: December 30, 2004
Abstract Bone is a dynamic tissue. Its continuous remodeling depends on the balance between bone formation and bone resorption. These two processes are carried out by specialized cells called osteoblast and ostreoclast respectively. The osteoclastic bone resorption consists in clegradation of the mineral and collagen components of bone. The stud,, of bone turnover requires accurate assessment of osteoclastic bone resorption, that becomes even more important in pathologic bone loss due to the uncoupling between bone formation and bone resorption. Osteoclastic activity is difficult to measure. Many techniques, generally based on the detection of the resorbing lacunae (pits) due to the bone degradation, allow to estimate bone resorption, but none of them quantitatively and directly measures the volume of resorbed bone. We propose a reliable and relatively simple method, based on contact surface profilometry, to evaluate directly and quantitatively the volume of resorbed bone. The method has the following advantages: i. to perf:wm a comparison of the same bone surface before and after the exposure to the osteoclastic activity;
ii. to enhance the sensitivity by utilization of bone slices shaped and polished in order to concentrate the cell activity in a controlled alrea. KEYWORD.*: Surface profilometry, bone resorption. 1. INTROD UCTION
The stud,r of bone turnover and bone loss, due to physiological or pathological conditions [1], requires accurate assessment of osteoclastic bone resorptio~t. Osteoclasts are cells of haemopoietic origin that, by a differentiation process, assume the ability to resorb bone [z~. Bone resorption is a multi-step process consisting in the adhesion of the osteoclast to the bone surface, cell polarization and formation of a subcellular c o m p a r t m e n t in which the degradation of both mineral and collagen components of bone occurs [3, 4]. BoJm erosion by osteoclastic cells is expected to induce surface texture variations at a microscopic scale, leaving resorption pits and exposed collagen fibrils. Osteodastic activity is difficult to measure and many techniques have been used in order to estimate bone resorption, with very different results and reference values. Histomorphometric techniques often report the percentage of bone surface covered by multinuclear giant cells with morphological appearance of mature resorbing osteoclasts. These techniques show significan': variations mostly arising from the methods used for osteoclast identification. They vary from put,ely morphological techniques [5, 6, 7, 8]
to enzymatic methods based on acid phosphatase expression [91. The methods based on the analysis of bone surface characteristics consider areas showing multinucleated cells associated with Howship's lacunae (pits) [10]. Pit assays are the more diffuse methods for the evaluation of in vitro bone resorption [11, 12, 133. In the pit assay, bone resorption is estimated by measuring the pit areas formed by osteoclasts cultured on bone slices [14]. This survey is error-prone since the recognition of the pits is difficult, the lacunae depth is often disregarded and it is not possible to perform comparative measures on the same slide before and after exposure to the osteoclast activity. Scanning electron microscopy (StU), that is widely used to reveal the presence of the pits, can provide a detailed inspection of the bone surface. However, since a conductive coating of the samples is needed for SEM observation, comparative measures on the same surface, before and after exposure to osteoclastic cells, are not possible. Several methods based on optical microscopy have been considered in order to analyse pit shape and dimensions, thus assessing bone resorption. Walsh proposed a technique based on wide-field microscopy and standard latex spheres used as scale markers [151. The profile of reflective surfaces was recovered by illumination with rapid
Address I:orcorrespondence: Franco Fusi: e-marl: [email protected] 41
E Fusi et alii: A new method based on contact surface profilometry
scanning of laser beam [16J. Specimens with fiat surfaces have been analysed by interference fringe m e t h o d too [17]. In practice, the use of this m e t h o d for the assessment of bone resorption is limited owing to the very small depth variations (a few wavelengths) measured. However, the determination of bone resorption by optical microscopy techniques presents the problem of very high light scattering and diffusion due to the irregularity of bone surface and partial transparency to visible light. Moreover, the use of chromophores is an impediment for comparative measures on the same sample. Recently, other advanced surface techniques have been used to inspect bone surface: x-ray photoelectron spectroscopy (xvs), to determine the chemical composition of the bone surface, and optical profilometry, to determine surface roughness [18]. Finally, biochemical evaluations of in vitro bone resorption have been proposed as well. They consist in measuring the a m o u n t of collagen degradation products released in the culture m e d i u m during the bone resorbing process [11, 19, 20, 21] Although all these methods assess osteoclastic bone resorption, none of t h e m allow to measure directly the volume of bone eroded by osteoclast activity. We propose a reliable method, based on contact surface profilometry, to evaluate directly and quantitatively the volume of resorbed bone. In this m e t h o d contact profilometry is used not only to detect surface roughness but also to measure with accuracy the volume of degraded bone tissue, through the comparison of surface profiles monitored on the same bone surface before and after the exposure to the osteoclastic activity. The m e t h o d is also very sensitive, due to the utilization of bone slices shaped and polished in order to concentrate the cell activity in a controlled area. 2. MATERIAL AND METHODS
2. 1. Bone slices Calf leg bone was treated as described elsewhere [22]. Small rectangular slices of bone were cut using a band saw. They were reduced by abrasion to a size of 4 m m x 10 m m and 2 m m thick, taking care to avoid any w a r m i n g of the samples that could change the bone structure. The slice surfaces were slightly polished with emery to obtain uniform roughness. A final polishing treatment was performed in order to have a preferential roughness pattern, using only translational movements parallel to the longest side of the slices. Two cross-shaped ruts with U profile (width g 200 btm and depth g 100 btm) were excavated on one of the larger surfaces of each slice. The ruts had the double function of gathering the cells, thus circumscribing the resorption activity, and representing a geometrical reference for the m e a s u r e s (FIG. 1). It is known that cells 42
FIG. 1. Schematic representation of a bone slice with crossshaped ruts. The slice surfaces were polished with emery in order to obtain a uniform and low roughness number. tend to accumulate close to, or inside any surface discontinuity, t h u s / h e n c e we expected to find an enhancement of resorption inside the ruts. These were dug utilizing a small disk saw with carborundum blade 200 g m thick. The depth of the ruts was about 100 g m to allow a good scan by the profilometer pick-up. This is essential to set the starting point for any profile measurement in order to analyse the same bone area before and after any cell treatment.
2. 2. Osteoclastic cells and preparation of the samples Cell populations derived from the FLG 29.1 h u m a n cell line were used. The FLG 29.1 cell line, derived and stabilized from a patient suffering from acute monoblastic leukemia (FAB M5a), was characterized as a model of osteoclastic precursor [23]. Cells were cultured in RPMI 1640 m e d i u m supplem e n t e d with 10% heat-inactivated fetal calf serum (FCS) and split twice weekly. Cultures were carried out in a fully humidified atmosphere, at 37~ and 5% CO2. Viability was determined by blue trypan dye exclusion. Cells were treated in order to induce differentiation toward the osteoclastic pathway, as elsewhere described [22, 24]. 15 culture flasks, each containing a bone slice, were prepared with a cell density of 3 x 105. Control flasks containing bone slices soaked in culture m e d i u m only were prepared too. The bone slices were maintained in the culture flasks for 72 h, then washed with distilled water, sonicated to remove adherent cells, washed again and dried at r o o m temperature before the surface analysis by contact profilometry.
2. 3. Bone surface analysis The survey of bone surface was performed by using a commercial microprofilometer (Hommelwerke GmbH) equipped with a pick-up terminated by a
Physica Medica 9Vol. XXI, N. 1, January-March 2005 diamond s @us (whose tip has a radius o f 5 ~tm and a full angh: o f 90~ During the surface scanning, the differential vertical displacement y (x) of the stylus respect to a conventional zero value was converted into an eleztric signal, then amplified, digitalised and analysed in order to achieve a n u m b e r o f roughness parameters according to the international standards (ISO). The intrinsic vertical resolution depends on the analog to digital converter (aDC) and on the measuring range. The instrument was equipped with a 12 bit ADC providing a sub-micron resolution (measuring range/4096). After a preliminary optimisation o f the method, we selected the main roughness parameters (the average roughness Ra and the root-mean-square profile height R~) which resulted as the most indicative for the specii~c surface texture modifications associated with bone resorption by osteoclastic cells. T h e y are defined in the following way:
1s L~ R =T2 ly (x)ldx
form error and waviness from the roughness. The cutoff frequency of the low pass filter (according tO DIN 4777) was set at 1/6 o f the scanned profile. The best surface lying on the profiles was calculated and, consequently, the volume of the rut was determined. The volume was calculated as the space portion delimited by the rut surface and the best fitting plane of the slice surface on which the horizontal paths o f the profiles lied. In Figure 2 and Figure 3, an example o f a single profile as well as a tridimentional (3-D) reconstruction of the rut surface are shown. A suitable software (working on MatLab platform) was developed to p e r f o r m these calculations. Profiles o f the bone slices were recorded before and after the exposure to osteoclastic cell activity, and then compared in order to measure changes in surface roughness and rut volume. At least three measures were made on each
q0
iq:(@mfoL'(y(X)) 2 d.x)1/2 where L~ is the length o f assessment. The fi)J:mer represents the average deviation o f the proti!e from the m e a n line, i.e., the average distance fi'om the vertical displacement to the m e a n line over the length of assessment, while the latter is the average o f the square deviation from the m e a n line. R a is suitable to m o n i t o r gradual changes. It does not differentiate b e t w e e n peaks and valleys and is weakly sensitive to surface defects, thus ensuring high repeatability o f the m e a s u r e m e n t s . Rq is sensitive 1:o peaks and valleys and gets meaningful w h e n lo0~:ing at the profile as a statistical function. It represents the standard deviation o f the profile height distribution. The m e a s u r e m e n t s were p e r f o r m e d positioning the sample on a specially designed holder, mechanically connected to a micrometric slit which could be moved perpendicularly to the scanning direction. T w e n t y o n e roughness profiles, with a spacing o f 200 _+ 5 ttm (5 ~m is the reading precision o f the micrometric slit), were acquired for each slice. The first profile was taken at a distance of 100 _+ 5 }am from the edge of the short rut (see Figure 1). Profile paths (1000 _+ 2 g m length) were parallel to the short rut and perpendicular to the long one. The scanned plane surface on each slice resulted a rectangular area o f 4000 _+ 10btm • 1000 __ 2~m, that is 4 m m z. The scanning parameters were decided according to both the accuracy of the measurements and the heavy computation. To optimise both S / N ratio and resolution a scan speed o f 0.05mm/s was selected. A Gaussian low pass filter was used to separate long wavelengths like form,
8O
001
lenghtDin]
Fro. 2. (Top) Singles 2-D profiles of the same bone slice recorded before and after (thick line) exposure to FLC 29.1 cells activit3~ After resorption the shape of the profile is nearly mantained but rut's depth is increased. (Bottom) Microscopic images, obtained with grazing lighting, of a bone slice recorded before (right) and after (left) exposure to differentiated F~C 29.1 cells. A detail of the bottom of a rut is imaged. 43
E Fusi et alii: A new method based on contact surface profilometry
6.
FIG. 3. Tridimensional reconstruction of the rut before (top) and after (bottom) cell activity Bone resorption is evident inside the rut: the particular shows how peaks and discontinuities appear smoothed as a consequence of cell activity
slice. The difference between a single measurement of the rut volume and the average value was found to be about 2%. The reproducibility of the single profile area (a section of the rut volume) was suitable being the max error from the average less than 1%. Surface analysis by tridimentional digital widefield microscopy [25] and SElVtwas performed to have visual evidence of osteoclastic cell activity on the bone slices and, in particular, in the area scanned by surface profilometry. A grazing lighting was necessary to evaluate the surface roughness by imaging. The imaged zone was lighted by an optical fibre placed over the long rut. A home-made holder for slices and optical fibre was coupled to the microscope to attain good reproducibility of lighting and consequently of the imaging. 3. RESULTS AND DISCUSSION
After a 72 h exposure to the cell activity, the bone slices were prepared as above described and their surface was tested by contact profilometry. For each slice, the surface profile obtained was compared with that monitored before the exposure in the culture flasks. The control samples were treated in the same manner. The average roughness Ra and the root-meansquare profile Rq were calculated, for each profile, with reference to the horizontal part of the profile beside the rut. Because the main surface of the slices are not perfectly parallel each other, the main line 44
of the single profile was rotated to superimpose the horizontal zero line. For each bone R a and Rq were calculated as the average of the 21 profle values. The analysis of the planar surface of both slices exposed to cell activity and controls failed to demonstrate any statistical difference in the average roughness Ra and in the root-mean-square profile Rq, monitored before and after the permanence in the flasks. This result was expected because of the presence, on the slice surface, of ruts in which the cells tend to gather and concentrate their activity. On the contrary, the volumes of the ruts obtained by surface profilometry after the exposure to osteoclastic cells appeared larger than those obtained before the treatment. A significant increase was evidentiated in rut's depth by analysing singles 2-D profiles (FIG. 2), while less important changes were found in the width of the rut. The volume of resorbed bone was calculated by comparing the surface profiles monitored before and after cell exposure. The mean value was 2_+ 0.5 p m 3 per pm2with m a x i m u m values up to 20_+0.5 btm3 per p m 2 on the b o t t o m of the ruts. The systematic error in volume determination was calculated to be about 0.2%. The m a x i m u m deviation from average (3 determination for each slice) was less than 1%. The rut volume of the control slices was found to be not significatively different before and after the microgravity treatment because the differences were comparable to the reproducibility error
Physica Medica 9 Vol. XXI, N. 1, January-March 2005
(0.6_+0.5 a m 3 per p m 2, a value very close to that measured ,m the planar surface). These reaults are comparable with data reported in literature. Boyde and Jones (26) summarised and compared ,:he different p h o t o g r a m m e t r i c m e t h o d s used to study shape and dimensions of Howship's lacunae in order to extrapolate the volume o f resorbed bol,e. Laser scanning confocal microscopy resulted fine m o r e reliable method. Measures obtained by ~:onfocal microscopy revealed a depth o f the pits tanging from 5% to 20% o f their diameter. A mear~ resorbed volume of 500 p m 3 for a pit of 20 pm width was estimated. The results we obtained witti the technique here presented are consistent with this value. The m e a n resorbed volume measured with our m e t h o d appears overestimated because the determination was made in a pit-enriched area (the b o t t o m of the rut) with the purpose of increasing the sensitivity. Contrariwise, the other methods generally consider larger areas where the pits cover about 5% o f the surface only. After ttl,: assay by surface profilometry, the slices were obse~':ved by optical microscopy (FIG. 2, bottom) and SEM (FIG. 4). Both the assays gave evidence o f osteoclastic cell activity inside the ruts o f the b o n e slices kept in the culture flasks containing differentiated FLG 29.1 cells, thus supporting the results obtained by profilometry. In fact, the observed changes in rut vo3umes are consistent with the presence of pit-enriched areas on the b o t t o m o f the ruts of the bone slice!; exposed to the cell cultures.
Fro. 4. Scanning Electron Microscopy of a bone slice exposed to FLC 29.1 cells cultured in modelled hypogravity. m e a s u r e m e n t s over the same b o n e sample before and after the exposure to the cell activity. Another important feature o f the m e t h o d is the possibility to increase the cell density in a limited area o f the slices (the b o t t o m of the ruts) and, consequently, to concentrate the resorption activity, thus enhancing the signal-to-noise ratio and the sensitivity. BIBLIOGRAPHY
[1]
[2] [3] [4]
4. CONCI,1 SIONS
There are m a n y techniques allowing the study o f b o n e surface modification to evidentiate resorption pits. Scan1:ing electron microscopy, light microscopy, many kinds o f profilometry (mechanical, laser and interfi:.rometric) are the most c o m m o n . As revealed from literature, all of them give information about the area of resorption pits, often completely disregarding the depth, thus evaluating only the percentage a~ea of degraded b o n e and not supplying direct and quantitative measurements of resorbed b o n e volume. The m o r e advanced m e t h o d s cited in the introduction have the potentiality for estimating the resorbed b o n e with accuracy but, in practice, their utili>.ation appears difficult. In fact, they generally w o r k properly only w h e n applied on unscattering surf~tces, require time-consuming procedures and computationally intensive processes. We approached the p r o b l e m o f bone resorption assessment addressing to the contact microprofilomel:ry because it ensures, in our opinion, the better c o m p r o m i s e a m o n g accuracy, scanning speed and. easy use. The m e t h o d we propose is relativel) 5;imple and able to p e r f o r m directly quantitative measures o f resorbed bone. This is mainly due to th(:- possibility o f carrying o u t profilometric
[5]
Pederson L, Winding B, Foged N T, Spelsberg T C, Oursler M J. Identification of breast cancer cell line-derived paracrine factors that stimulate osteoclast activity. Cancer Res 1999:59 (22); 5849-5855. Suda T, Takahashi N, Martin TJ. Modulation of osteoclast differentiation. Endocr Rev 1992: 13; 66-80. Roodman G D. Cell biology of the osteoclast. Exp Hematol 1999:27 (8); 1229-1241. Everts V, Delaisse J M, Korper W,, Niehof A, Vaes G and Beertsen W. Degradation of collagen in the boneresorbing compartment underlying the osteoclast involves both cysteine-proteinases and matrix metalloproteinases. J Cell Physiol 1992: 150; 221-231. Wronski T J, Lowry P L, Walsh C C, Ignaszewski L A. Skeletal alterations in ovariectomized rats. Calcify Tissue Int 1985: 37; 324-328.
[6]
[7]
[8]
[9]
Shen V, Birchman R, Lindsay R, Dempster D W. Shortterm changes in histomorphometric and biochemical turnover markers and bone mineral density in oestrogen and/or dietary calcium-deficient rats. Bone 1995: 16; 149-156. Westerlind K C, Wronski T J, Pitman E L, Luo Z P, An K N, Bell N H, Turner R T. Estrogen regulates the rate of bone turnover but bone balance in ovariectomized rats is modulated by prevailing mechanical strain. Proc Natl Acad Sci USA 1997: 94; 4199-4204. Bagi C M, Miller S C. Comparison of osteopenic changes in cancellous bone induced by ovariectomy and/or immobilization in adult rats. Anat Rec 1994: 239; 243254. Sims N A, Morris H A, Moore R J, Durbridge T C.
Increased bone resorption precedes increased formation in the ovariectomized rat. Calcif Tissue Int 1996: 59; 121-127.
Chen M M, Jee W S S, Ke H Z, Lin B Y, Li Q N, Li X J. Adaptation of cancellous bone to ageing and immobilization in growing rats. Anat Rec 1992: 234; 317-334. [11] Foged N T, Delaisse J M, Hou P, Lou H, Sato T, Winding B, Bonde M. Quantification of the collagenolytic activity of isolated osteoclasts by enzyme-linked immunosorbent assay. J Bone Miner Res 1996: 11; 226-237. [10]
45
E Fusi et alii: A new method based on contact surface profilometry [12]
[13]
[14] [15]
[16]
[17] [18]
[19]
[20]
46
Sugawara K, Hamada M, Hosoi S, Tamaoki T. A useful method to evaluate bone resorption inhibitors, using osteoclast-like multinucleated cells. Anal Biochem 1998: 255; 204-210. Lorget F, Kamel S, Mentaverri R, Wattel A, Naassila M, Maamer M, Brazier M. High extracellular calcium concentrations directly stimulate osteoclast apoptosis. Biochem Biophys Res Commun 2000: 268; 899-903. Chambers T J, Revell P A, Fuller K, Athanasou N A. Resorption of bone by isolated rabbit osteoclasts. J Cell Sci 1984: 66; 383-399. Walsh C A, Beresford J N, Birch M A, Boothroyd B, Gallagher J A. Application of reflected light microscopy to identify and quantitate resorption by isolated osteoclasts. J Bone Miner Res 1991: 6; 661-671. Moss J P, Linney A D, Grindrod L D, Mosse C A. A laser scanning system for the measurement of facial surface morphology. Optics and Laser in Engineering 1989: 10; 179-190. Virdee M S. Nanometrology of optical fiats by laser autocollimation. Surface Topography 1988: 1; 415-425. Schwartz Z, lohmann C H, Wieland M, Cochran D L, Dean D D, Textor M, Bonewald L E Boyan B D. Osteoblast proliferation and differentiation on dentin slices are modulated by pre-treatment of the surface with tetracycline or osteoclasts. J Periodontol 2000: 71; 586-597. Breuil V, Cosman E Stein L, Horbert W, NievesJ, Shen V, Lindsay R, Dempster D W. Human osteoclast formation and activity in vitro: effects of alendronate. J Bone Miner Res 1998: 13; 1721-1729. James I E, Lark M W, Zembryki D, Lee-Rykaczewski E V, Hwang S M, Tomaszek T A, Belfiore R Gowen
[21]
[22]
[23]
[24]
[25]
[26]
M. Development and characterization of a human in vitro resorption assay: demonstration of utility using novel antiresorptive agents. J Bone Miner Res 1999: 14; 1562-1569. Lorget E Mentaverri R, Meddah B, Cayrolle G, Wattel A, Morel A, Schecroun N, Maamer M, de Vernejoul M C, Kamel S, Brazier M. Evaluation of in vitro bone resorption: High-performance liquid chromatography measurement of the pyridinolines released in osteoclast cultures. Anal Biochem 2000: 284; 375-381. Monici M, Agati G, Fusi F, Paglierani M, Cogoli A, Bernabei P A. Gravitational unloading induces osteoclast-like differentiation of FLC 29.1 cells. J Gravitational Physiology 2002:9 (1); 261-262. Gattei V, Bernabei P A, Pinto A, Bezzini R, Ringressi A, Formigli L, Tanini A, Attadia V, Brandi M L. Phorbol ester induced osteoclast-like differentiation of a novel human leukemic cell line (FLG 29.1). J Cell Biol 1992: 116; 437-447. Monici M, Agati G, Fusi F, Pratesi R, Paglierani M, Santini V, Bernabei P A. Dependence of leukemic cell autofluorescence patterns on the degree of differentiation. Photochem Photobiol Sci 2003: 2; 981-987. Fusi E Monici M, Agati G. Wide-field autofluorescence microscopy for the imaging of living cells in Current laser and optical techniques in biology. Palumbo G, Pratesi REds. London. The Royal Chemical Society, 2004; 357-374. lSBN 0 85404 321 7. Boyde A, Jones S J. Really close up! Surveying surfaces at sub-micrometre resolution: the measurement of osteoclastic resorption lacunae. Photogrammetric Record 1992:14 (79); 59-84.