Stereological and serial section analysis of chondrocytic enlargement in the proximal tibial growth plate of the rat.

THE ANATOMICAL RECORD 239:255-268 (1994)
Stereological and Serial Section Analysis of Chondrocytic
Enlargement in the Proximal Tibia1 Growth Plate of the Rat
G.J. BREUR, J. TURGAI, B.A. VANENKEVORT, C.E. FARNUM, AND N.J. WILSMAN
Department of Comparative Biosciences, University of Wisconsin, Madison, Wisconsin,
53706; Department of Anatomy, Cornell University, Ithaca, New York 14853 (Farnum)
ABSTRACT
Background: It has been suggested that within the growth
plate, the final volume and shape of hypertrophic chondrocytes are important variables in determining the rate of longitudinal bone growth. To better understand the organization and regulation of chondrocytic hypertrophy as related to longitudinal bone growth, the beginning and end, and the
location and magnitude of chondrocytic volume and shape changes during
the hypertrophic process were defined in the proximal tibia1 growth plate
of 35-day-old rats.
Methods: In this study we used two different approaches, a stereological
analysis of chondrocytes in unbiasedly defined, narrow growth plate
strata, and a serial section reconstruction and measurement of individual
cells. In both experiments chondrocytes were preserved using optimal
chemical fixation. Proliferating chondrocytes were identified using bromodeoxyruidine labelling, and the rate of longitudinal bone growth was
determined using oxytetracycline labelling.
Results: In both studies, immediately following cell division in the proliferative zone, chondrocytic volume gradually increased toward the midpoint of the growth plate. During this phase of about 30 hours, approximately 20% of the final cell volume was obtained. During the following 20
hours the remaining 80% was acquired. The estimated rate of cell volume
increased changed from approximately 50 pm3/hr during the first 30 hours
to about 800 pm3/hr during the last 20 hours. The increase in cell volume
resulted in an increase in both the vertical and the horizontal chondrocytic
diameters. Cell parameters did not change during the final five hours of the
maturation process.
Conclusions: In this study we demonstrated that chondrocytic enlargement starts immediately following cell division in the proliferative zone,
and that chondrocytic enlargement consists of two morphologically distinguishable phases. The transition point between the first and the second
phase of chondrocytic enlargement corresponded with the junction between the proliferative zone and the maturation zone.
0 1994 Wiley-Liss, Inc.
Key words: Growth plate, Chondrocytes,Enlargement, Cell shape, Hypertrophy, Rats
The major portion of the elongation of long bones is
contributed by cartilage discs (growth plates), located
near the end of long bones. During periods of active bone
growth, growth plates form cartilage that is converted
into metaphyseal bone (endochondral ossification). The
formation of new cartilage is a well organized process,
as evidenced by the columnar arrangement of chondrocytes and the polar morphology of the growth plates
(Ham and Cormack, 1984). Chondrocytes at the epiphyseal end of the columns actively divide (proliferation),
while more distally, towards the metaphyseal side of the
growth plate, cells increase in size and diameters (hypertrophy) (Buckwalter e t al., 1986; Cruz-Orive and
0 1994 WILEY-LISS, INC
Hunziker 1986; Hunziker et al., 1987; Hunziker and
Schenk, 1989; Kember, 1960,1972,1973,1978;Kember
and Walker 1971; Walker and Kember, 1972). Finally,
chondrocytes die abruptly during the conversion of the
newly formed cartilage into metaphyseal bone (Brighton et al., 1973, 1982; Farnum and Wilsman, 1987,
1989; Hunziker and Schenk, 1989; Schenk, 1980).
Received November 15, 1993; accepted February 10, 1994.
Dr. Breur’s present address is Department of Veterinary Clinical
Sciences, School of Veterinary Medicine, Purdue University, West
Lafayette, IN 47907-1248. Address reprint requests there.
256
G.J. BREUR ET AL.
The rate of longitudinal bone growth is a function of
the rate of cell division in the columns and the height
(cell diameter parallel to the long axis of the bone) of
juxta-metaphyseal hypertrophic chondrocytes (Hunziker and Schenk, 1989; Kember, 1960, 1972, 1973,
1978; Thorngren and Hansson, 1973a,b, 1981; Sissons,
1955). It has been suggested, based on the large variety
of the number of newly produced chondrocytes in
growth plates growing at different rates and the relatively minor observed changes in the mean height of
juxta-metaphyseal chondrocytes, that the rate of new
cell production in the proliferative zone is the principal
determinant of the rate of longitudinal bone growth
(Kember and Walker, 1971; Kember, 1978; Moss-Salentijn, 1974; Nilsson et al., 1986; Thorngren and Hansson, 1973a,b). However, recent studies have indicated
that the role of chondrocytic hypertrophy may not be as
passive as previously suggested. It has been demonstrated that the volume of hypertrophic chondrocytes
may be up to 10 times larger than the volume of proliferative chondrocytes, and chondrocytic height may
increase four- to fivefold between the proliferative zone
and the hypertrophic zone (Buckwalter et al., 1986;
Hunziker et al., 1987; Hunziker and Schenk, 1989).
Furthermore, recent studies of chondrocyte performance in the proximal tibial growth plate of mice and
rats at different ages suggested that changes in the
rate of bone growth are accompanied by changes in the
size of hypertrophic chondrocytes (Buckwalter and
Mower, 1987; Hunziker and Schenk, 1989). In a recent
study, including 16 growth plates spread among four
locations and two species a t two different ages, we
found a strong correlation (rat = 0.98; pig = 0.83) and
a positive linear relationship between the rate of longitudinal bone growth and the volume of hypertrophic
chondrocytes (Breur et al., 1991). A similar correlation
and relationship was found in rabbits (Kuhn et al.,
1993). These studies suggest that in growth plates,
chondrocytic enlargement plays a major role in the determination of the rate of longitudinal bone growth.
The process of chondrocytic enlargement a s i t relates
to longitudinal bone growth is characterized by a n increase in cell volume and a modulation of cell shape. In
the proliferative zone, cells have a flattened, oblate
spheroid shape, while in the hypertrophic zone enlarged chondrocytes are more rounded and have obtained a more sphere-like shape, or a prolate spheroid
or superspheroid shape (Buckwalter e t al., 1986, 1987;
Cruz-Orive and Hunziker, 1986; Hunziker et al., 1987;
Hunziker and Schenk, 1989). Even though the process
of chondrocytic enlargement is generally referred to as
hypertrophy, several workers have demonstrated that
the increase in cell volume is mainly the result of cellular swelling (absorption of water), rather than cellular hypertrophy (increase in cell organelles) (Brighton
e t al., 1973,1982; Buckwalter et al., 1986; Hunziker et
al., 1987).
The regulation of chondrocytic volume and chondrocytic shape is poorly understood. It has been hypothesized that chondrocytic shape is modulated by transphyseal collagen fibrils in the longitudinal septa of the
growth plate and the periosteal-perichondrial complex
(matrix directed cell swelling) (Buckwalter and Sjolund, 1989). However, experimental evidence and evidence obtained from pseudoachondroplastic growth
plates suggest that other matrix constituents also may
be involved in the modulation of chondrocytic shape
(Breur et al., 1990, 1992; Dietz et al., 1992; Engfeldt,
1969a; Westerborn, 1961). Recently, a positive relationship between the final volume of hypertrophic
chondrocytes and the rate of longitudinal bone growth
was demonstrated (Breur et al., 1991; Kuhn et al.,
1993). We considered this evidence that the final volume of hypertrophic chondrocytes may be regulated.
However, it is unknown where in the growth plate or
how chondrocytic volume is regulated.
A problem in the study of the organization of chondrocytic hypertrophy is the lack of a uniform definition
of the point where cellular enlargement begins. In defining zones of growth plates, investigators often have
used criteria based on cell size or cell shape (Brighton,
1978; Buckwalter et al., 1985, 1986; Holtrop, 1972a,b;
Hunziker e t al., 1987; Hunziker and Schenk, 1989;
Scott and Pease, 1956). Traditionally, i t was thought
that chondrocytic enlargement was restricted to the
growth plate zones containing large rounded cells
(maturation and hypertrophic zone), and that the zone
containing small flattened cells (proliferative zone)
was solely composed of young proliferating chondrocytes (Ham and Cormack, 1984). However, it has been
demonstrated in several studies of growth plates from
different locations that chondrocytic proliferation only
takes place in the proximal portion of the proliferative
zone, at the epiphyseal ends of the columns (Farnum
and Wilsman, 1993; Kember, 1960, 1972, 1973; Kember and Walker, 1971; Walker and Kember, 1972). The
function of chondrocytes in the proliferative zone distal
to the actively proliferating cells is unknown, but in
two recent stereological studies they were, though
largely ignored, implicitly considered a s enlarging, hypertrophying chondrocytes (Buckwalter et al., 1986;
Hunziker et al., 1987; Hunziker and Schenk, 1989).
In previous stereologic studies of the growth plate,
growth plate zones were defined using subjectively applied cell morphological criteria (Buckwalter et al.,
1986; Hunziker et al., 1987; Hunziker and Schenk,
1989). Also, the analyzed zones were relatively thick,
thus perhaps obscuring more subtle changes taking
place within the analyzed zones. Another limitation of
the previous studies is the fact that the reported
changes in cell volume and cell shape were not related
to time, and i t may be that the reported changes in cell
volume and cell shape do not reflect changes in the rate
of cell volume increase and the rate of cell shape modulation. Analysis of the rate of chondrocytic volume
and shape changes is important, because rapid rate
transitions may signal points where mechanisms regulating chondrocytic enlargement act.
The goal of this study is (a)to determine where chondrocytic enlargement commences, and (b) to analyze
chondrocytic volume and shape changes in relation to
time. We studied the process of chondrocytic enlargement in the proximal tibial growth plate of 35-day-old
rats and used two different approaches. First, in eight
unbiasedly defined, narrow strata of the growth plate,
using stereological techniques, we measured changes
in mean cellular volume, and mean horizontal and
mean vertical chondrocytic profile diameters. In addition, we calculated the rate by which these parameters
changed. A unique aspect of this study is defining and
ENLARGEMENT OF GROWTH PLATE CHONDROCYTES
257
Measurement of Rate of Longitudinal Growth
dividing the growth plate into many thin strata. As the
width of each stratum becomes narrower, we have the
ability to integrate changes from stratum to stratum.
With the second approach, volumes and diameters of
individual chondrocytes within columns were measured using serial section techniques. Interpretations
of serial section studies are always limited by a small
sample size; however, used in combination with a n integrative stereological analysis, the analysis by serial
sections is complementary. In both experiments, bromodeoxyuridine (BrdU) labelling was used to identify
proliferating chondrocytes, and oxytetracycline (OCT)
bone labelling was used to estimate the rate of longitudinal bone growth. All chondrocytes were preserved
using optimal chemical fixation.
Following thawing of the frozen growth plates,
growth plates were cut in four equal parts. Of each
part, an approximately 250 pm thick vertical slab was
cut with a razor blade. These slabs, representing a systematic sampling of the entire growth plate, were put
on glass slides, mounted in glycerine, and viewed under epifluorescent light. On each of the four 250 mm
thick slabs, the distance between the 2 fluorescent oxytetracycline labels was measured a t 4 unbiasedly determined locations along the growth plate and the
mean of these measurements was considered the longitudinal bone growth per 24 hours (t(bg/day)).
ANIMALS, MATERIALS, AND M E T H O D S
Irnrnunocytochernical Identification of
Proliferating Chondrocytes
Animals, Tissue Collection, and Tissue Processing
For the stereological analysis of unbiasedly defined,
narrow growth plate strata, six 35-day-old male hooded
rats (weight 90-120 g) were selected. Oxytetracycline
hydrochloride (1-2 mgikg IP) was administered 48 and
24 hours before euthanasia. Thirty minutes before
euthanasia, the animals were injected with the thymidine analogue 5-bromo-2’-deoxyuridine (BrdU; 100 mgi
kg; 20 mg/kg). Immediately after anesthesia (pentobarbital 30 mgikg IP) and exsanguination, the left
proximal tibial growth plate was isolated, and the animal was euthanized with a n overdose of barbiturate.
The growth plate was submerged in the primary fixation solution and further trimmed into 1 x 1 x 3 mm
blocks with epiphyseal bone on one end and metaphyseal bone on the other end. The fixation protocol for
these blocks consisted of a primary fixation in 2% glutaraldehyde-2% paraformaldehyde in 0.05 M cacodylate buffer (pH 7.3) with 0.7% ruthenium hexammine
trichloride for 2 hours. The blocks then were rinsed in
0.05 M cacodylate buffer (pH 7.3), followed by secondary fixation in 0.2% OsO,-O.7% ruthenium hexammine
trichloride in 0.05 M cacodylate buffer (pH 7.3) for 2
hours. After a second rinse in 0.05 M cacodylate buffer
(pH 7.31, the blocks were rapidly dehydrated in graded
alcohols, cleared in propylene oxyde, and embedded in
epon-araldite. Growth plate tissue fixed according to
this protocol contains hypertrophic chondrocytes preserved in their native state, which makes meaningful
stereology possible (Farnum and Wilsman, 1987; Hunziker et al., 1982, 1983,1984). An additional advantage
is that this protocol allows the immunocytochemical
identification of cells in the S-phase of the cell cycle
using BrdU. Following the collection of the left proximal tibial growth plate, the right proximal tibial
growth plate was isolated and placed in airtight bags.
The right proximal radial growth plates were stored at
-20°C until used for the measurement of the rate of
longitudinal bone growth. All collections were done between 9:OO and 11:OO A.M.
The proximal radial growth plate of one 35-day-old
male hooded rat was used for the reconstruction and
measurements of individual growth plate chondrocytes. The tissue collection and tissue processing for
this study were as described above, except that secondary fixation was omitted.
One micrometer thick sections were cut, put on acid
cleaned slides, and etched in 50% sodium ethoxide (pH
12.5) for 15 minutes. Etching resulted in both deplastification of the sections and denaturing of the nuclear
DNA. Following etching, sections were passed through
graded alcohols, washed in PBS (0.05 M phosphate
buffered saline a t pH 7.3), and incubated with a n antiBrdU monoclonal antibody (Beckton Dickinson Immunocytometry Systems, Mountain View, CA 94039) a t a
dilution of 25 plilml PBS (1:40 dilution of stock solution), for 12 hours at 4°C in a humidified slide chamber.
The primary antibody was localized with a n avidinbiotin-peroxidase complex (biotinylated horse antimouse IgG and avidin-biotin-peroxidase complex; Vector Laboratories Inc, Burlingame, CA 94010). After
incubation with primary antibody, the sections were
washed in PBS and incubated with the biotinylated
horse anti-mouse IgG (9 pl biotinylated horse antimouse IgG in 1 ml PBS). Following another wash, the
sections were incubated with the avidin-biotin-peroxidase complex (9 p1 avidin stock solution and 9 pl biotinylated peroxidase stock solution per 1ml PBS). Both
incubations were for 30 minutes in a humidified slide
chamber a t 4°C. The BrdU incorporated was demonstrated by a final incubation for 30 minutes in a fresh
solution of 0.05% 3,3‘-diaminobenzidine tetrahydrochloride (Sigma, St. Louis, MO) and 0.02% hydrogen
peroxide in PBS.
Chondrocytes that had incorporated BrdU (chondrocytes in the S-phase of the cell cycle) were demonstrated by a brown precipitate over the nucleus (Fig. 1).
Method controls (deletion of the primary antibody, deletion of the secondary antibody, and staining of unlabelled growth plate tissue a s a negative control) resulted in blocking of immunolabeling. Two specificity
controls were done. Incubation of the primary antibody
(25 pl stock solution per 1 ml PBS) with BrdU (1,5,10,
or 100 pgiml PBS) for 48 hours a t 4°C before the immunocytochemical procedure resulted in complete ablishment of labelling a t BrdU concentrations 2 10
pgiml PBS. In contrast, pre-incubation of the primary
antibody (25 pl stock solution per 1 ml PBS) with thymidine (10, 50, 100, or 1000 pgiml PBS; Sigma, St.
Louis, MO) did not affect the intensity of the immunolabel.
258
G.J. BREUR ET AL.
Fig. 1. Localization of chondrocytes in the S-phase of the cell cycle
using a monoclonal antibody against the thymidine analogue
5-bromo-2'-deoxyuridine (BrdU). a: BrdU containing cells, identified
by their dark nuclei, are demonstrated in the proximal portion of the
proliferative zone. b Immunolabelling is completely abolished if the
primary antibody is incubated with BrdU (210pgiml PBS) before the
immunocytochemical procedure. Bar = 100 pm. c: Pre-incubation of
the primary antibody with thymidine (10,50,100, or 1000 pgiml PBS)
does not affect the intensity of the immunolabel. Bar = 30 km.
Stereological Analysis of Unbiasedly Determined
Growth Plate Strata
the micrographs ( x SOO), growth plates were divided
proximodistally in 8 equal, narrow strata. In each stratum measurements for the following stereological parameters were made: mean cell volume (ik))
chondro,
cytic volume fraction (VdVwri), the mean principal
diameters of chondrocytic profiles (Tf(hor) and X v e r t i ) , and
the percentage of cells in the S-phase of the cell cycle
(Labelling Index).
In defining zones of growth plates, investigators
most often have used subjectively applied criteria,
based on cell size or organization of chondrocytes
(Brighton, 1978; Buckwalter et al., 1985, 1986; Cruz-
Strata analysis
Three blocks of each growth plate were randomly
choosen for stereological analysis. One section cut vertically (parallel to the long axis of the bone) a t 1 pm
was obtained from each block and stained with azure
11, basic fuchsine, and methylene blue (Humphrey and
Pittman, 1974). Two unbiasedly choosen fields per section (including epiphyseal bone, the entire growth
plate, and metaphyseal bone) were photographed. On
259
ENLARGEMENT O F GROWTH PLATE CHONDROCYTES
Orive and Hunziker, 1986; Ham and Cormack, 1984;
Holtrop, 1972a,b; Hunziker et al., 1987; Hunziker and
Schenk, 1989; Scott and Pease, 1956). In this study we
divided the growth plate in eight strata of equal height.
With this approach the definition of growth plate zones
was completely independent from cell size and chondrocytic organization, which ensured a n unbiased approach to the analysis of the process of chondrocytic
enlargement. In preliminary studies we had determined that approximately eight strata were required
to obtain enough sensitivity to detect minor changes in
the parameters of interest.
Reference volumes
Labelling Index per unit area is equal to the Labelling
Index per unit volume.
Chondrocytic volume fraction. The chondrocytic volume fraction of a stratum V(c)/V(stri
was estimated from
micrographs generated for the stereological analysis
using point counting techniques (Cruz-Orive and Hunziker, 1986). Vw is the total volume of chondrocytes in
that stratum, and V(str)the total volume of the stratum.
Mean cell volume. The mean cell volume i i c ) of a stratum was calculated using the following equation (CruzOrive and Hunziker, 1986; Gundersen and Jensen,
1985; Gundersen, 1986):
vw
=
(11/3)
x 31,
In each animal, the volume of the proximal tibial where iic)is the volume weighted mean chondrocytic
the mean cubed intercept length. The
growth plate V(ppii, the volume of unbiasedly deter- volume, and
mined strata V(stri (see strata analysis), and the volume point sampled intercept length was measured from the
of newly produced bone per day V(bg/dayiwere measured. micrographs with the aid of a n image analysis system
Reference volumes were calculated by multiplying the (Bioquant IV, R&M Biometrics Inc, Nashville, TN). Of
growth plate area in a plane perpendicular to the long each growth plate stratum, at least 300 intercepts unaxis of the bone by respectively, the mean-growth plate der 16 randomly choosen angles were measured (Cruzheight tippi), the mean stratum height tiw, and the Orive and Hunziker, 1986). This technique is suitable
length growth per day h g i d a y ) . The growth plate area for volume estimations of anisotropic objects, and can
perpendicular to the long axis of the bone was esti- be used on vertical sections (Baddely et al., 1986; Cruzmated by measuring the principal diameters of the in- Orive and Hunziker, 1986; Gundersen and Jensen,
tact growth plates with a sliding mechanical caliper 1985; Gundersen, 1986).
Chondrocytic numerical density. Assuming minimal
(accuracy 10 p.m). The estimate of the projected area
cell
size variability, the chondrocytic numerical denwas obtained by using the ellipse formula (Cruz-Orive
sity Nic)/V(str),
the number of chondrocytes per mm3 was
and Hunziker, 1986; Howell et al., 1960):
calculated using the relationship (Cruz-Orive and
area = (d4)x longest diameter x shortest diameter
Hunziker, 1986):
Nici/V(str)= [V(c)/V(str)]
/ UICJ]
Growth plate height was measured on 1 pm thick sections cut from each of 3 blocks per growth plate. On where Nw is the total number of chondrocytes and
each section, measurements were made a t 2 unbiasedly the mean chondrocytic volume in that stratum.
chosen locations along the growth plate from microNumber of chondrocytes per stratum. The relationship
graphs used for the stereological analyses. MeasureN(c) = [N(ci/Vistri] X Vistri
ments were made from the epiphyseal to the metaphyseal bone. The estimated mean height of each was used to calculate Nw, the number of chondrocytes
individual growth plate was obtained as the mean of 6 in a stratum.
measurements. The mean stratum height in the proxNumber of hypertrophic chondrocytes leaving the growth
imal tibial growth plate was equal to the growth plate plate per day. The number of hypertrophic chondrocytes
height divided by eight (number of strata).
leaving the growth plate per day N(out/day)
was estimated
by
Basic estimators
N(out/day) = [N~ci/V(stri]~~,~
X V(bg/day)
Labelling index.
where
[N~ci/V(str~]~~,~
is
the
numerical
density of the terTo identify chondrocytes in the S-phase of the cell
cycle, 3 serial 1 p.m thick sections were cut of each of 3 minal, juxta-metaphyseal stratum, and V(bgiday)the volrandomly selected blocks of each growth plate. The me- ume of newly formed bone per day.
dian section was stained with azure 11, basic fuchsine,
and methylene blue (Humphrey and Pittman, 19741, Chondrocytic morphology
Principal cell profile diameters.
while the two surrounding sections were used for the
In each stratum, the mean principal profile diameidentification of S-phase cells. Photographs of the median section served as a reference section and for the ters X ( h o r l and X i v e r t i were measured. The X ( h o r l of a
recording of cell profiles containing nuclei. The preced- chondrocytic profile was defined as the caliper diameing and the succeeding sections were used to determine ter of that profile measured perpendicular to the long
whether nucleated cell profiles of the median section axis of the bone. The X v e r t i was defined a s the length of
were BrdU-labelled or not. The Labelling Index was the intercept bisecting the cell profile measured paraldefined as the ratio of BrdU-labelled cell profiles and lel to the long axis of the bone. These definitions were
nucleated cell profiles from the median section. At choosen under the assumption that cells have a n ellipleast 200 nucleated cell profiles per stratum were an- soid shape with principal axes perpendicular and paralyzed. In this study, we assumed that all cells, includ- allel to the long axis of the bone. In addition, such
ing cells in the S-phase of the cell cycle, had the same measurements describe the principal cell profile diamnuclear diameter measured perpendicular to the long eters in relation to longitudinal bone growth (Kember,
axis of the bone. Under these conditions, the obtained 1960, 1972, 1973, 1978; Cruz-Orive and Hunziker,
260
G.J. BREUR ET AL.
1986; Hunziker et al., 1987; Hunziker and Schenk,
1989; Thorngren and Hansson, 1973a,b, 1981; Sissons,
1955).
Axial ratio of chondrocytic profiles. The axial ratio of
chondrocytic profiles was calculated as the ratio X(vert):
Xlhorl and describes how flat a cell profile is relative to
the long axis of the bone (Buckwalter et al., 1985). An
axial ratio smaller than 1represents a n oblate ellipsoid
cell profile; a n axial ratio of 1, a circular profile; and a n
axial ratio larger than 1, a prolate ellipsoid cell profile.
Temporal relationships of the process of
chondrocytic enlargement
The spatial relationships of chondrocytes within a
growth plate reflect the temporal changes that take
place during the process of chondrocytic enlargement
(Ham, 1952; Ham and Cormack, 1984; Hunziker et al.,
1987). Assuming that each chondrocyte in a column
represents a later stage of development than the preceding, more proximal chondrocyte, the strata not only
represent different stages of the enlargement process,
but also represent a certain time period. Because
growth plates, over a short time period, can be considered a steady state system where the rate of cell gain in
the proliferative zone is equal to the rate of cell loss at
the metaphyseal junction (Hunziker and Schenk, 1989;
Kember, 1960, 1972, 1973, 1978, 1983; Kember and
Kirkwood, 1987; Thorngren and Hansson, 1973a,b,
1981), definitions of kinetic parameters in cell renewal
systems can be applied (Cleaver, 1967).
Turnover time. The turnover time of a stratum Tlturn)
is the time required for the replacement of the total
number of cells in that stratum. This time is equal to
the mean amount of time each cell spends in this compartment if no cell divisions takes place. Thus, the cell
turnover time of a stratum can be estimated by
(Cleaver, 1967; Quastler, 1959, 1960; Quastler and
Sherman 1959)
T(turnlstr = N~c~,tJK~out),t,
where K(out),tris the mean number of cells leaving the
stratum per hour. This study was restricted to the process of cellular enlargement, and therefore only the
most distal stratum containing proliferating cells
(Stratum 3) and strata without proliferating cells
(Strata 4 through 8) were included. Because of these
restrictions and the fact that the growth plate is a
steady state system, the efflux K(outIst, of each of the
included strata is approximately the same and can be
estimated as follows (Cleaver, 1967; Quastler, 1959,
1960; Quastler and Sherman 1959):
Klout)str = K(outIterm= “out!
X 11/24]
where K(out)te,m
is the chondrocytic efflux from the terminal, juxta-metaphyseal stratum, and N(nutithe total
number of chondrocytes leaving the growth plate per
day.
Rate of gain or loss. The rate of gain or loss per hour
of parameters related to cell morphology (Om, X(hor1
and Xlvert!) were calculated for the strata for which
Tlturn) was estimated. The derivation of this parameter
is illustrated in this example of the estimation of the
rate of mean cell volume change Aih,,,/hr = [U (C)str~
(
c str]
i / ~T(turn),t,
~
~
~
In this equation, ~ ( cst, )is the
~
~mean
~
~cell volume of the
stratum preceding the stratum of interest. u ( ~and
) ~ ~ ~
T(turn),trare, respectively, the mean cell volume and the
turnover time of the stratum of interest.
Statistical analysis
The arithmetic means and standard deviations of all
parameters in identical strata of the 6 animals were
calculated and compared with the means of adjacent
strata using one-way analysis of variance and Fisher’s
analysis for least significant differences (protected
LSD) (Snedecor and Cochran, 1980).
Serial Section Reconstruction and Measurement of
Individual Chondrocytes
One tissue block was randomly choosen for the analysis. One hundred vertical (parallel to the long axis of
the bone) serial sections were cut. Sections were 1 pm
thick, except every fourth section, which was cut a t 2
pm. The accuracy of the microtome was verified in a
preliminary study in which the actual thickness of sections cut at 1 pm was measured using a Watson interference objective. In this study, the actual mean section
thickness of 36 unbiasedly choosen sections cut at 1pm
was 0.995 pm (20.06). One micron thick sections were
stained with azure 11, basic fuchsine, and methylene
blue (Humphrey and Pitmann, 1974). The 2 pm thick
sections were used for the immunocytochemical identification of cells in the S-phase of the cell cycle.
A growth plate region 80 pm wide and 100 pm deep
(83sections), spanning the growth plate from the proximal end of the growth plate columns to metaphyseal
bone, was analyzed. Columns within this region were
unbiasedly selected for reconstruction using the forbidden plane, or “brick” rule (Howard et al., 1985). Camera lucida drawings of cell profiles of interest were
made at ~ 7 4 8 and
,
the individual cell profiles were
digitized. The columns were reconstructed and volumes
of individual cells were measured using a Eutectics
3-dimensional image analysis system (Eutectics Inc.,
Raleigh, NC). With this system, the estimated volume
of individual cells is equal to the cumulative total of
the product of the section thickness and the cell profile
area of all sections containing profiles of that cell. A
correction factor for the 2 pm thick sections was included. The principal caliper cell diameters X(vert)and
Xlhorl, the caliper cell diameters parallel and perpendicular to the long axis of the bone, respectively, were
measured on reconstructions magnified at x 1,000. The
caliper cell diameter in the third dimension was calculated a s the product of the section thickness and the
number of sections containing profiles of that cell. Adjustments for 2 pm thick sections and incompletely
filled sections (minus one section thickness) were
made. The mean of this diameter and Xlhor) is the reported X(hori. Because the spatial arrangement of chondrocytes within columns reflects their development in
a temporal fashion, and because growth plates can be
considered a steady state system (Hunziker et al., 1987;
Hunziker and Schenk, 1989; Kember 1960,1972,1973,
1978,1983; Kember and Kirkwood, 1987; Kember and
Walker, 1971; Thorngren and Hansson, 1973a,b, 1981),
the time that a cell remains in a certain position of the
column (or remains in a certain stage of development)
can be estimated. This time, T(celli,is equal to the time
a hypertrophic cell remains in the most distal position
ENLARGEMENT OF GROWTH PLATE CHONDROCYTES
of the column, and was estimated by (Gray and Scholes,
1951; Hunziker et al., 1987; Hunziker and Schenk,
1989; Sissons, 1955)
261
Chondrocytic morphology
During the process of cellular enlargement, the
mean cell volume ixC) increased approximately 10
Tlcell) = 24 x IXlvert)term / tlbgiday)]
times. The smallest mean chondrocytic volumes were
where tibg/day) is the longitudinal bone growth per day, found in the strata with active cell proliferation, Strata
and Xivewterm the mean vertical diameters of the eight 1,2, and 3 (1,050, 1,250, and 1,590 mm3, respectively).
most distal hypertrophic chondrocytes in this recon- Mean cell volume slowly increased to 2,530 ~m~ in
struction. The cell maturation time Tim), the time be- Stratum 5, and then increased with significant steps
tween cell proliferation, and cell death was estimated towards its final volume (12,840 pm3) in Stratum 7 and
by (Gray and Scholes, 1951; Hunziker and Schenk, Stratum 8. The mean vertical and horizontal chondrocytic profile diameters,Jhe profile diameters measjred,
1989; Kember, 1960; Sissons, 1955)
respectively, parallel (Xlvert))and perpendicular (X[hor))
R m ) = n X T(c~II)
to the long axis of the bone, followed a similar pattern.
In strata 1, 2, and 3,
was 5.4, 5.3, and 6.4 pm,
where n is the number of cells between the last BrdU- respectively. In the more distal strata %ert) increased
labelled cell of a column and metaphyseal bone.
significantly, and the maximum value (25.5 pm) was
Finally, the position of each cell in a column was obtained in Stratum 8. In the most proximal three
determined by measuring the distance from each cell to strata W(hor) was 12.7, 14.0, and 13.9 pm, respectively,
the chondro-osseous junction with the epiphysis.
and i t increased to 21.0 pm in Stratum 8. The smallest
axial ratio was found in Stratum 2 (0.42). It increased
RESULTS
with significant steps from 0.50 in Stratum 3 to 1.26 in
Stereological Analysis
Stratum 7, and obtained its final value 1.31 in Stratum
8. The axial ratio was approximately 1 (1.08) in Stratum 6. Significant differences (P < 0.01) related to ih,
Reference volumes
The mean growth plate height t(gpi), the mean Xvert)and X f h o r ) were found between Stratum 4 and 5, 5
stratum height j ( s t r ) , and the mean longitudinal bone and 6, and 6 and 7. In addition Xcvert) was significantly
growth per day t(bg/day) were 577 pm (+30), 72 pm (54, larger in Stratum 4 than in Stratum 3 ( P < 0.01).
and 331 pmiday (?28), respectively. The estimated Twenty percent of the final mean cell volume was obmean values of the corresponding 3-dimensional pa- tained from Stratum 2 to Stratum 4 ( 2 3 3 hours), which
volume increase mainly resulted in a n increase in
rameters growth plate volume V(gpi),stratum volume X(vert).
The remaining 80%of the final mean cell volume
Vlstr), and the volume of newly produced bone per day
V(bgiday)
were 13.66 mm3 (k1.231, 1.71 mm3 (+0.16), and was obtained from Stratum 5 to Stratum 7 (520 hours);
this volume incre_ase resulted in significant increases
7.81 mm3/day (_t0.72), respectively.
in both Xtvert) and Xlbor).During the last 5 hours of chondrocytic enlargement, the mean cell volume and the
Basic estimators
mean principal cell diameters did not change signifiThe mean Labelling Index, the fraction of cells in the cantly. All results, including the statistical analyses,
S-phase of the cell cycle, was 0.15 in Stratum 1(epiph- are summarized in Table 1. The relationship between
yseal end of the growth plate). 0.19 in Stratum 2, 0.08 OW,Wivert),and Xihor) in the different strata is depicted in
in Stratum 3, and 0 in all other strata. These data Figure 2A.
The pattern of these parameters after time normalsuggest that active cell proliferation only took place in
the proximal 3 strata of the growth plate. The chondro- ization reflected the pattern before time normalization.
cytic volume fraction V<c)/V(str),
the fraction of a stratum A phase with moderate rate increases in Strata 3,4, and
occupied by chondrocytes, increased from 0.21 in Stra- 5 is followed by a phase with major rate increases from
tum 1 to 0.67 in Stratum 8 (metaphyseal end of the Stratum 5 to 6. No significant gains or losses were made
growth plate). The highest number of cells per stratum in Stratum 7. The rate of ain in mean cell volume
Nw and the highest number of chondrocytes per mm3 gradually rose from 18 pm5/hr in Stratum 3 to 230
(N<c)/Vlstr))
was found in Stratum 2 (474,200 cells/stra- pm3/hr in Stratum 5. Then, it abruptly increased to 807
tum and 277,500 cells/mm3) and this number decreased pm3/hr in Stratum 6 (P < 0.01). This rate of cell volume
towards the metaphysis (Stratum 8: 90,200 cellsistra- increase was maintained in Stratum 7 (840 pm3/hr) and
tum and 53,200 cells/mm3). The total number of hyper- then dropped to -11 km3/hr in Stratum 8 (P < 0.05).
trophic chondrocytes leaving the growth plate per day The rate of increase of Xivert) in Stratum 3 was 0.06
N(outiday1and per hour K(out)was 414,600 (269,200) and pm/hr, which gradually increased to 0.39 pm/hr in Stra17,280 (?2,880), respectively. The turnover time Titurn), tum 5. The rate significantly increased to 1.14 pmihr in
the time required for the replacement of the total num- stratum 6 (P < 0.01) and 1.18 pm/hr in stratum 7, and
ber of cells, was the highest in Stratum 3 (19.7 hours) then significantly decreased to 0.16 ym/hr in Stratum
and decreased towards Stratum 7 and 8. The turnover 8 ( P < 0.01). The rate of change OfX(h0r)followedthe rate
time of Stratum 8 was 5.3 hours. The estimated matu- of change of the mean cell volume t v c ) and Xvert). A
ration time, the time from the end of the last cell divi- significant increase was found between Stratum 5 and
sion in the proliferative zone to cell death in the jux- 6 (0.15 pm/hr to 0.43 pmlhr; P < 0.05), and a significant
tametaphyseal cartilage (the total of all turnover decrease between Stratum 7 and Stratum 8 (0.42 pm/hr
times), was 57.8 hours (59.3). All results, including the to 0.13 pm/hr; P < 0.01). The relationship between these
results of analyses for comparison of significant differ- 3 parameters is displayed in Figure 2B. All results are
ences between adjacent strata, are listed in Table 1.
summarized in Table 1.
262
G.J. BREUR ET AL.
TABLE 1. Stereological estimators of the proximal tibia1 growth plate of the rat (standard deviations are given
in parentheses)
Strata
Parameters
1
2
3
4
5
6
7
8
0.00
(0.00)
0.00
io.001
0.63
(0.05)
0.67
(0.031
Basic estimators
LI
0.15
(0.071
VicjNistri
0.21
(0.031
Nid
NiciNistri
(mm
2
0.19
(0.031
2
0.08
(0.04)
0.34
(0.031
348,800
(71,480)
2 474,200
3l 206,700
__*I_ 277,500
(51,950)
(37,980)
0.32
(0.02)
5
(80,360)
Kiouti (cellsihrl
5
5
0.33
(0.031
1,250
(1801
(2101
-
X i v d (pm)
-
5.4
(0.3)
Xihori (pml
12.7
(1.31
Axial ratio
0.49
(0.061
2
2
0.00
(0.001
0.04
(0.061
0.51
(0.05)
152,200
(21,9701
5
(17,2401
88,900
(14,840)
17,280
(2,8801
17,280
(2,8801
5 225,600
205,700
(23,1801
5 131,700
17,280
(2,8801
(38,870)
13.4
(2.81
1,590
(180)
0.00
(0.001
5
333,800
(57,5601
19.7
(4.01
Chondrocytic morphology
oici (pm3i
1,050
0.00
(0.00)
2,530
(3801
5
5
8.9
(1.7)
4,550
6.4
(0.41
8.6
(0.71
5
12.2
(1.41
14.0
(1.21
13.9
(0.91
14.9
(1.41
2
16.2
(1.21
0.42
(0.031
0.50
(0.041
imihr (pm3ihri
18
(101
2
0.62 5
(0.061
75
(311
97,000
i10.7301
83,800
(8,6201
90,200
(9,5701
57,200
(7,8501
49,400
(6,5601
53,200
(7,960)
17,280
(2,8801
17,280
(2,880)
17,280
(2,8801
2
5
5
( 1,0001
5.3
(0.31
0.80
(0.08)
""
5
0.18
(0.051
0.39
(0.09)
-0.01
(0.04)
0.07
(0.051
0.15
(0.05)
5.7
(0.91
5
5.0
(1.01
9,040
f 1,560)
5.3
(0.51
12,840
(1,250)
12,840
(1,7401
18.5
(1.51
5
24.0
(2.2)
25.5
(2.21
18.4
(0.91
5
20.4
(1.0)
21.0
(0.3)
1.08 _T_
(0.11)
807
(2101
230
(75)
0.06
(0.031
5
1.26
1.31
(0.11)
(0.101
840
(5211
5
-11
(272)
1.14
(0.261
1.19
(0.601
0.16
(0.231
0.43
(0.24)
0.42 _II_
(0.281
0.13
(0.211
*Significantly different a t the 95% confidence level.
**Significantly different a t the 99% confidence level.
yseal bone, chondrocytic volume rapidly increased. The
rapid cell volume increase was accompanied by rapid
Our reconstruction consisted of nine columns, a total increases in both X(verti and X(horl (caliper diameter perof 168 chondrocytes. Only two of those columns were pendicular to the long axis of the bone). Approximately
complete columns (originating in the zone of actively 100 pm away from the metaphysis, the maximum valproliferating cells delineated by immunolocalization of ues of uw, X(vertJ, and X ( h o r ) were obtained. All spheroid
BrdU and continuing uninterrupted to the chondro- cells (axial ratio 0.9-1.1) and prolate spheroid cells (axosseous junction with the metaphysis). Six columns ial ratio > 1.1) were found in the phase of rapid volume
were incomplete distally (epiphyseal columns), and one increase (Fig. 3). The variability of the studied paramcolumn was incomplete proximally (metaphyseal col- eters a t any arbitrary level of the growth plate (for
umn). The total reconstruction and the individual col- example 300 pm away from the epiphysis) indicated
substantial cell morphological differences between
umns are displayed in Figure 3.
The results of the serial section reconstruction con- chondrocytes in adjacent columns of the growth plate.
firmed the findings of the stereological analysis of nar- The relationship of uiCi, X(vert1, and X(hor1, and their porow strata. BrdU labelled cells (cells in the S-phase of sition in the columns is demonstrated in Figure 4.
Within a column, the development of the mean cell
the cell cycle) were confined to a narrow zone a t the
proximal ends of the columns, approximately 28-1 19 volume and both principal cell diameters reflected the
hm away from epiphyseal bone. Distally, from the zone pattern described in the stereological analysis of narcontaining proliferating cells, cells gradually doubled row strata. All three parameters increased towards the
their volume towards a transition point approximately metaphysis regardless of whether columns were com215 hm away from the epiphysis. During this gradual plete or incomplete. Again, changes in X(verti most
increase in cell volume cells maintained their oblate closely reflected changes in cell volume. Finally, all
spheroid shape, and their caliper vertical chondrocytic cells in the proximal portion of a column were oblate
diameter (X(vert1, the chondrocytic diameter parallel to spheroid, but more distally oblate and prolate cells ofthe long axis of the bone) increased only moderately. ten succeeded each other (Figs. 3, 5).
The transition from small flattened chondrocytes to
Beyond the transition point 215 pm away from epiphAnalysis of Individual Cells Using Serial
Section Reconstruction
263
ENLARGEMENT OF GROWTH PLATE CHONDROCYTES
30
b
1200
--1.2
=
Ai(c)lhour
A
Ax(vert) I hour
25
800-
20
1
Ax(hor)lhour
1
I
-0.8
L
3
0
15
4000
-E,
10
4
{
3,'
1
2
3
4
5
-0.6
400-
-0.4
.
3
T
0
5
- 0.2
oJ
0
t
600-
6
7
-2ooJ
0
8
1
STRATUM
2
3
4
5
6
7
8
1-0.2
STRATUM
Fig. 2. Changes in chondrocytic morphology during the process of cellular enlargement as measured
stereologically in eight unbiasedly defined, narrow growth plate strata. a: Mean cell volume G(c) (m),
mean vertical cell diamet_er X(vert1 (A),and mean horizontal cell diameter X(hor1(0).
b Rate of change
per hour A u d h r (m), A X(vertdhr (A)and A X(hodhr (0).Standard deviations are listed in Table 1.
B
metaphyseal column
epiphyseal columns
complete columns
Fig. 3. a: Serial section reconstruction of nine growth plate columns. Bar = 50 pm. b: Individual
columns. Two columns were continuing uninterrupted from the epiphyseal end of the columns to the
chondro-ossoeous junction with the metaphysis, six columns were incomplete distally (epiphyseal columns), and one column was incomplete proximally (metaphyseal column).
more spheroid or prolate spheroid chondrocytes at a
point approximately 250 pm away from epiphyseal
bone was very rapid. In the six transitions analyzed,
cells in this region nearly doubled their volume (1.82
0.66) within two sequential positions, and tripled their
*
volume (2.96 2 1.15)within three sequential positions.
The rate of metaphyseal bone growth was 290 pmlday
and the mean X(vert1 of the eight most distal hypertrophic chondrocytes was 33.2 pm. Thus, each chondrocytic position represented 2.8 hours (T(cel1)). Therefore,
G.J. BREUR ET AL.
-45
18000
-40
16000
-35
~
12000
f
+
10000-
m
+
t
0
+
8000-
8000
6000-
9 .
6000
40002000
0
0
50
100
150
200
250
300
30
.
350
- 20
if
-10
2000
5
0
F
3
-15
4000
- ,
400
.....
.
-25
,
50
100
150
200
250
300
350
400
450
DISTANCE FROM EPIPHYSEAL BONE (prn)
DISTANCE (prn)
Fig. 5. Changes in chondrocytic morphology during the process of
cellular enlargement as measured in individual cells within a column
using serial section reconstruction. The position of each cell in a column was determined by measuring the distance from each cell to the
chondro-osseousjunction with the epiphysis. Chondrocytic volume v(c)
(El,vertical cell diameter Xlvert) (A),and horizontal cell diameter
X(hor) (0).
.40-
3530-
and the metaphyseal bone was 17.2. Thus, the estimated mean maturation time T m was 47.3 hours.
DISCUSSION
0
50
100
150
200
250
300
350
400
450
DISTANCE (prn)
40
,
I
351
30
I
l
l
0
0
50
,
,
,
,
,
,
,
,
100
150
200
250
300
350
400
1
450
DISTANCE (pm)
Fig. 4. Changes in chondrocytic morphology during the process of
cellular enlargement as measured in 168 individual cells using serial
section reconstruction. The position of each cell in a column was determined by measuring the distance from each cell to the chondroosseous junction with the epiphysis. a: Chondrocytic volume VW. Oblate spheroid cells (O),
spherical cells (01,and prolate spheroid cells
( + ) . b: Vertical cell diameter Xlvert). c: Horizontal cell diameter
Xlhor).
this rapid increase in volume took place within a time
span of 5.5 to 8.5 hours. In addition, the mean number
of cells in a column between the last BrdU-labelled cell
In this study, using two different and independent
techniques, and using the physiological criterion of loss
of BrdU labelling as a n indicator of where all growth
plate chondrocytes have left the cell cycle, we demonstrated that (a) cell volume increase starts immediately following cell division in the proximal portion of
the proliferative zone, (b) the process of chondrocytic
enlargement as i t relates to longitudinal bone growth
consists of two morphologically distinguishable phases,
(c) the rate of cell volume increase and the rate of cell
shape modulation is significantly higher during the
second than during the first phase, and (d) that cell
volume increase during the first phase results mainly
in an increase in vertical chondrocytic diameter, while
cell volume increase during the second phase results in
a large, significant increase of the vertical chondrocytic
diameter, and a smaller but significant increase of both
horizontal chondrocytic diameters. The transition
point between the first and second phase of chondrocytic enlargement corresponded with the natural transition from flattened chondrocytes to more rounded
chondrocytes, often described by other investigators as
the junction between the proliferative zone and the hypertrophic zone (Brighton, 1978; Buckwalter et al.,
1985,1986; Engfeldt, 196913; Ham and Cormack, 1984;
Holtrop, 1972a,b; Moss-Salentijn, 1974; Scott and
Pease, 1956).
Stereological, Serial Section, and Kinetic Analysis
In the present studies we used three new morphological approaches to study chondrocytic hypertrophy.
First, in our stereological study, unlike previous stereological studies by other workers (Brighton et al.,
1973, 1982; Buckwalter et al., 1985, 1986; Buckwalter
and Mower, 1987; Cruz-Orive and Hunziker, 1986;
ENLARGEMENT OF GROWTH PLATE CHONDROCYTES
265
Hunziker et al., 1987; Hunziker and Schenk, 19891, Therefore, they may challenge one of the main undergrowth plate regions were defined independently from lying assumptions of previously used kinetic models
natural or physiological strata, thus ensuring a n unbi- describing the performance of growth plates (Hanson,
ased approach to chondrocytic hypertrophy. Further- 1967; Hunziker and Schenk, 1989; Kember, 1960,
more, the obtained stereological parameters were used 1972, 1973, 1978, 1983; Tayler et al., 1987a,b; Thornfor a newly developed kinetic analysis of growth plate gren and Hanson, 1973a,b, 1981).
behavior. This kinetic analysis is independent from the
Chondrocytic Enlargement
columnar arrangement of chondrocytes, and indepenDuring the first phase of chondrocytic enlargement,
dent from the parameters mean chondrocytic height
and mean septum thickness, parameters that only can in the present study approximately 30 hours, and corbe estimated accurately using time consuming serial responding to Strata 3 and 4 or a zone 120 to 215 pm
away from the epiphyseal bone, cell volume increased
section reconstructions.
As a second approach, we used computer-aided serial gradually, and 20% of the final cell volume was obsection reconstruction and measurement of chondro- tained. During the following 20 hours, corresponding to
cytes within columns. With this technique sampling is Strata 5, 6, and 7 or a zone 215 to 350 pm away from
also independent from the natural and physiological epiphyseal bone, the remaining 80% of the final cell
growth plate strata. In addition, this technique com- volume was acquired.
The change in chondrocytic morphology from a flatplements the stereological analysis; the data obtained
with the stereological analysis reflect average tened shape in the first phase of chondrocytic enlargechanges, while the data obtained with the serial recon- ment to a rounded shape during the second phase of
struction analysis provides very precise measure- chondrocytic enlargement is accompanied by changes
ments, however, of a limited number of chondrocytes. in cellular activity. Alterations in chondrocytic activity
Finally, this is the first study in which chondrocytic are reflected by changes in proliferative activity, inproliferation was studied on the same tissue as used for termediary metabolism, synthesis of matrix constituthe stereological analysis of chondrocytic enlargement. ents, and mechanisms of cellular enlargement. For exApplication of stereological techniques for the study of ample, active cell division has been demonstrated in
growth plate performance requires fixation protocols the proliferative zone, but not in the maturation and
that preserve chondrocytes in a fully expanded state hypertrophic zones (this study; Farnum and Wilsman,
(Hunziker et al., 1982). Using these fixation protocols, 1993; Kember and Walker, 1971). Also, chondrocytes
chondrocytes closely resemble the morphology of chon- change from a n aerobic metabolism in the proliferative
drocytes fixed using rapid freezing and freeze substitu- zone to a n anaerobic metabolism in the hypertrophic
tion, and living chondrocytes in explants (Hunziker et zone (Brighton and Heppenstall, 1971; Brighton, 1978).
The synthesis of matrix constituents characteristic
al., 1982, 1984; Farnum and Wilsman, 1987; Farnum
et al., 1990). Although recently optimal chemical fixa- for the hypertrophic zone is further evidence that the
tion in combination with BrdU labelling was used for change in cell morphology reflects changes in cell functhe study of chondrocytic proliferation (Farnum and tion. Carbonic anhydrase (Gay et al., 1982; KumpuWilsman, 1993), chondrocytic enlargement was not ad- lainen and Vaanaanen, 1982; Vaanaanen 1984), Type
X collagen (Adams e t al., 1989; Gibson and Flint, 1985;
dressed in that study.
Grant et al., 1985; Schmidt and Linsenmayer, 1985),
Incomplete Columns
and Calpain I1 (Shimizu et al., 1991) have been idenIn addition to confirming the results of the stereolog- tified exclusively in the hypertrophic zone, both intraical study, our serial reconstruction study demon- cellularly and extracellularly. Other matrix constitustrated the existence of columns incomplete at the ents like collagenase (Blair et al., 1989; Brown et al.,
epiphyseal or metaphyseal end. This observation may 1989; Dean et al., 1985, 19891, chondrocalcin (Poole et
suggest that incomplete columns, as observed in the al., 1984; van der Rest et al., 1986), osmium-ferrocyarabbit by other workers (Moss-Salentijn et al., 19871, nide positive electron dense pericellular substance
are not species dependent, and that they may be rec- (Farnum and Wilsman, 19831, and a periodic acid reognized in other growth plates and in growth plates sistant lectin Ricinus Communis I binding pericellular
from other species a s well. This may be further support glycoconjugate (Farnum and Wilsman, 1988) have
for the concept of regular replacement of chondrocytic been localized in the hypertrophic zone, but not in the
columns (Rigal, 1962) and the hypothesis that epiphy- proliferative zone.
There is also evidence that the change in chondroseal columns represent young, newly developing columns, and metaphyseal columns the final segments of cytic morphology is accompanied by a change in the
exhausted columns (Moss-Salentijn et al., 1987, 1991). mechanisms by which chondrocytes enlarge. Several
Recently, i t was suggested that columns even may be workers have demonstrated that the main portion of
further subdivided in smaller groups of 4 to 16 cells cellular enlargement is the result of cell swelling (ab(Wilsman et al., 1993). The combined observations of sorption of water) rather than a n increase of cell orthese studies are important because they imply a pre- ganelles (hypertrophy), as evidenced by a volume inviously unrecognized regulatory level in the growth crease in cytoplasmic space and nucleoplasms between
plate, namely, the regulation of the transient column proliferative and hypertrophic chondrocytes (Brighton
or isogenous group. Furthermore, the finding of incom- et al., 1973, 1982; Buckwalter et al., 1986; Hunziker et
plete columns and the hypothesis of transient columns al., 1987). However, i t has been suggested that in the
or isogenous groups appear to be incompatible with the costochondral junction of the rat, cell swelling a s a
concept that the functional unit for longitudinal bone mechanism for chondrocytic enlargement is restricted
growth within the growth plate is the complete column. to the hypertrophic zone (Brighton et al., 1973). Qual-
266
G.J. BREUR ET AL.
itative ultrastructural studies of the growth plate support this hypothesis. In flattened chondrocytes of the
proliferative zone, organelles are tightly packed together, while in rounded chondrocytes of the hypertrophic zone, cell organelles are spread and the cytoplasm has a more loose appearance (Holtrop, 1972a,b;
Scott and Pease, 1956). Furthermore, re-analysis of
data presented by Buckwalter (Buckwalter et al., 1986)
confirms this hypothesis. In the proximal radial growth
plate of 15-day-old mice, 65% of the cell volume increase between the upper and lower proliferative zone
was the result of a n increase in the volume of the Golgi
apparatus and the rough endoplasmic reticulum; the
balance was mainly provided by a n increase in cytoplasmic ground substance. Eight-five percent of the
chondrocytic volume increase between the distal portion of the proliferative zone and the distal portion of
the hypertrophic zone was contributed by a n increase
in cytoplasmic space and nucleoplasm, while the balance again was mainly provided by a n increase in cytoplasmic ground substance. These studies together
would suggest that cellular enlargement, during what
we have described as the first phase of cellular enlargement and corresponding to Strata 3 and 4, is due to true
hypertrophy, while the cell volume increase in the second phase, Strata 5,6, and 7, is mainly the result of cell
swelling. Consequently, the transition from flattened
cell to a more rounded cell would have to be accompanied by synthesis or activation of membrane structures
that facilitate active transport of water and electrolytes against the high osmotic pressure of the matrix
(Hunziker et al., 1987; Rothstein, 1989). Recently, i t
was hypothesized that potassium channels, responsible
for a large outward-directed K' current in chondrocytes from the hypertrophic zone, may play a n important role during the swelling of growth plate chondrocytes (Walsh et al., 1992).
In a previous study we demonstrated a positive linear relationship between the final volume of hypertrophic chondrocytes and the rate of growth of a growth
plate (Breur et al., 1991). We saw this as evidence that
the final volume of hypertrophic chondrocytes may be
regulated independently from physiological processes
associated with endochondral ossification. The data of
this study indicated that the final volume of hypertrophic chondrocytes was obtained a t least 100 pm
proximally from the metaphysis, thus a t least 5 hours
before cell death. These two studies together would
suggest that a control mechanism for the final volume
of hypertrophic chondrocytes must act at, or proximally
from, this point. The functional significance of a period
without substantial cell morphologic changes at the
end of chondrocytic enlargement is unclear. However,
we speculate that it is related to preparations for the
upcoming conversion of cartilage into bone and the imminent cell death.
This study's contribution to the understanding of
chondrocytic enlargement in the growth plate is twofold. It demonstrates that cell volume increase starts
immediately following cell division in the proximal
portion of the proliferative zone, and that chondrocytic
enlargement consists of a phase of slow and a phase of
rapid cellular enlargement. Although this study clarifies several aspects of chondrocytic enlargement, i t also
raises new questions in relation to the kinetics of the
growth plate. For instance, the finding that chondrocytic enlargement starts immediately following cell division in the proximal portion of the proliferative zone
may imply cellular heterogeneity in this portion of the
growth plate. Thus, in addition to actively proliferating
chondrocytes, this stratum may contain cells that have
left the cell cycle and have started the phase of slow
chondrocytic enlargement. Circumstantial evidence for
such a heterogeneity has been reported in epiphyseal
cartilage (Wilsman and Vansickle, 1971) and in
growth plate cartilage (Farnum and Wilsman, 1993;
Rigal, 1962; Simon and Cooke, 19881, but definitive
proof of functional heterogeneity has not been given.
Further definition of the functional heterogeneity may
permit a more accurate estimation of cell cycle time
and enhance our understanding of the process of chondrocytic maturation. Another, more general question is
related to the role of matrix synthesis in bone length
growth. It is generally understood that bone length
growth is a function of chondrocytic proliferation, chondrocytic enlargement, and matrix synthesis (Seinsheimer and Sledge, 1981; Buckwalter and Sjolund, 1989),
while the rate of bone length growth is only determined
by the number of newly produced chondrocytes per column per day and the mean height of terminal hypertrophic chondrocytes (Hunziker and Schenk, 1989;
Kember, 1960, 1972, 1973, 1978; Thorgren and Hansson, 1973a,b, 1981; Sissons, 1955). Consequently, many
recent studies related to bone length growth are restricted to chondrocytic proliferation and volume and
shape changes during chondrocytic enlargement (e.g.,
Breur et al., 1991; Buckwalter et al., 1985, 1986; Farnum and Wilsman, 1993; Hunziker et al., 1987; Hunziker and Schenk, 19891, even though data of several
studies have implied that the cartilage matrix contribution to longitudinal bone growth may vary from 30%
to as much a s 60% (Buckwalter et al., 1986, Breur et
al., 1992; Hunziker e t al., 1987; Hunziker and Schenk,
1989). It seems unlikely that the important contribution from cartilage matrix to bone length growth is
completely inconsequential for the determination of
the rate of bone length growth. Thus, revisiting the role
of cartilage matrix formation in bone length growth
may be justified.
LITERATURE CITED
Adams, S.L., K.M. Pallante, and M. Pacifici 1989 Effects of cell shape
on type X collagen gene expression in hypertrophic chondrocytes.
Connect. Tissue Res., 20t223-232.
Baddeley, A.J., H.J.G. Gunderson, and L-M. Cruz-Orive 1986 Estimation of surface area from vertical sections. J. Microscopy, 142:
259-276.
Blair, H.C., D.D. Dean, D.S. Howell, S.L. Teitelbaum, and J.J. Jeffrey
1989 Hypertrophic chondrocytes produce irnrnunoreactive collagenase in vivo. Connect. Tissue Res., 23.65-73.
Breur, G.J., G.A. Padgett, C.E. Farnum, and N.J. Wilsrnan 1992 Cellular basis of decreased rate of longitudinal growth of bone in
pseudoachondroplastic dogs. J. Bone Joint Surg., 74-At516-528.
Breur, G.J., B.A. VanEnkevort, C.E. Farnum, and N.J. Wilsman 1991
Linear relationship between the volume of hypertrophic chondrocytes and the rate of longitudinal bone growth in growth plates.
J. Orthop. Res., 9t348-359.
Breur, G.J., N.J. Wilsrnan, C.E. Farnum and G.A. Padgett 1990 Abnormal chondrocytic hypertrophy in growth plates of pseudoachondroplastic Scottish deerhound puppies. Proc. 36th Orthop. Res. Soc.. New Orlenas, LA, D. 90.
Brighton, C.T. 1978 Structure and function of the growth plate. Clin.
Orthop., 136t22-32.
Brighton, C.T., and R.B. Heppenstall 1971 Oxygen tension in zones of
ENLARGEMENT OF GROWTH PLATE CHONDROCYTES
the epiphyseal plate, the metaphysis and diaphysis. J. Bone Joint
Surg., 53-A:719-728.
Brighton, C.T., Y. Sugioka, and R.M. Hunt 1973 Cytoplasmic structure of epiphyseal plate chondrocytes. J . Bone Joint Surg., 55-At
771-784.
Brighton, C.T., Y. Sugioka, and R.M. Hunt 1982 Quantitative zonal
analysis of cytoplasmic structures of growth plate chondrocytes in
vivo and in vitro. J. Bone Joint Surg., 64-At1336-1349.
Brown, C.C., R.M. Hembry, and J.J. Reynolds 1989 Immunolocalization of metalloproteinases and their inhibitors in the rabbit
growth plate. J Bone Joint Surg., 71-At580-593.
Buckwalter, J.A., and D. Mower 1987 Alterations of growth plate
associated with slowing of bone growth. Proc. 33rd Orthop. Res.
SOC.,
San Francisco, CA, p. 490.
Buckwalter, J.A., D. Mower, J. Schafer, R. Ungar, B. Ginsberg, and K.
Moore 1985 Growth-plate-chondrocvteprofiles and their orientation. J. Bone Joint Surg., 67-A:942-955.
Buckwalter, J.A., D. Mower, R. Ungar, J . Schaeffer, B. Ginsberg 1986
Morphometrical analysis of chondrocyte hypertrophy. J . Bone
Joint Surg., 68-At243-255.
Buckwalter, J.A., and R.D. Sjolund 1989 Growth of cornstalks and
longbones. Iowa Orthop. J., 9:25-31.
Cleaver, J.E. 1967 Population kinetics in animal tissue. In: Thymidine Metabolism and Cell Kinetics. Wiley & Sons, New York, pp.
184-234.
Cruz-Orive, L-M, and E.B. Hunziker 1986 Stereology for anisotropic
cells: application to growth plate cartilage. J . Microscopy, 143:
47-80.
Dean, D.D., O.E. Muniz, I. Berman, J.C. Pita, and M.R. Carreno 1985
Localization of collagenase in the growth plate of rachitic rats. J .
Clin. Invest., 76:716-722.
Dean, D.D., O.E. Miniz, and D.S. Howell 1989 Association of collagenase and tissue inhibitor of metalloproteinases (TIMP) with hypertrophic cell enlargement in the growth plate. Matrix, 9:366375.
Dietz, F.R., J. Maynard, B. Fritz, and L. Trombino 1992 Comparison
of physeal morphology in several growth disorders. Proc. 38th
Orthop. Res. SOC.,Washington, DC, p. 636.
Engfeldt, B. 1969a Studies in the epiphyseal growth zone. I11 The
effect of papain on the ultrastructure of epiphyseal cartilage.
Acta Pathol. Microbiol. Scand., 75:220-228.
Engfledt, B. 1969b Studies in the epiphyseal growth zone. 111Electron
microscopic studies on the normal epiphyseal growth zone. Acta
Pathol. Microbiol. Scand., 75t201-219.
Farnum, C.E., and N.J. Wilsman 1983 Pericellular matrix of growth
plate chondrocytes: a study using postfixation with osmium-ferrocyanide. J . Histochem. Cytochem., 31r765-775.
Farnum, C.E., and N.J. Wilsman 1987 Morphologic stages of the terminal hypertrophic chondrocyte of growth plate cartilage. Anat.
Rec., 219:221-232.
Farnum, C.E., and N.J. Wilsman 1988 Lectin-binding histochemistry
of intracellular and extracellular glycoconjugates of the reserve
cell zone of growth plate cartilage. J . Orthop. Res., 6t166-179.
Farnum, C.E., and N.J. Wilsman 1989 Condensation of hypertrophic
chondrocytes a t the chondro-osseous junction of growth plate cartilage in Yucatan swine: Relationship to long bone growth. Am. J.
Anat., 186r346-358.
Farnum, C.E., J. Turgai and N.J. Wilsman 1990 Visualization of living terminal chondrocytes of growth plates cartilage in situ by
differential interference contrast microscopy and time-lapse cinematography. J . Orthop. Res., 8:750-763.
Farnum, C.E., and N.J. Wilsman 1993 Determination of proliferative
characteristics of growth plate chondrocytes by labeling with bromodeoxyuridine. Calcif. Tissue Int., 52t110-119.
Gay, C.V., R.E. Anderson, H. Schaer, and D.S. Howell 1982 Identification of carbonic anhydrase in chick growth plate cartilage. J.
Histochem. Cytochem., 30:391-394.
Gibson, G.J., and M.H. Flint 1985 Type X collagen synthesis by chick
sternal cartilage and its relationship to endochondral development. J. Cell. Biol., 101t277-284.
Grant, W.T., M.D. Susmann, and G. Balain 1985 A disulfide-bonded
short chain collagen synthetized by degenerative and calcifying
zones of growth plate cartilage. J . Biol. Chem., 260:3798-3803.
Gray, L.H., and M.E. Scholes 1951 The effect of ionizing radiation on
the broad bean root: Part VIII, Growth rate studies and histolog. ical analysis. Br. J . Radiol., 24:82-92.
Gundersen, H.J.G. 1986 Stereology of arbitrary particles. A review of
unbiased number and size estimators and the presentation of
some new ones, in memory of William R. Thompson. J . MicrosCODY. 143:3-45.
Guhdersen, H.J.G., and E.B. Jenssen 1985 Stereological estimation of
267
the volume-weighted mean volume of arbitrary particles observed on random sections. J . Microscopy, 138t127-142.
Ham, A.W. 1952 Some histophysiological problems peculiar to calcified tissue. J. Bone Joint Surg., 34-At701-728.
Ham, A.W., and D.H. Cormack 1984 Histology. Lippincott, Philadelphia, pp. 422-462.
Holtrop, M.E. 1972a The ultrastructure of the epiphyseal growth
plate. I, The flattened chondrocyte. Calcif. Tiss. Int., 9.131-138.
Holtrop, M.E. 19721, The ultrastructure of the epiphyseal growth
plate. 11,The hypertrophic chondrocyte. Calcif. Tiss. Int., 9.140151.
Howard, V., S. Reid, A. Baddely, and A. Boyde 1985 Unbiased estimation of particle density in the tandem scanning reflected light
microscope. J . Microscopy, 138:203-212.
Howell, D.S., E.E. Delchamps, W. Reimer, and I. Kiem 1960 A profile
of electrolytes in the cartilagenous plate of growing ribs. J . Clinic.
Invest., 39t919-929.
Humphrey, C.D., and F.E. Pittman 1974 A simple methylene blueAzure 11-basic fuchsine stain for epoxy embedded tissue sections.
Stain Technol., 491t9-14.
Hunziker, E.B., W. Herrmann, and R.K. Schenk 1982 Improved cartilage fixation by ruthenium hexammine trichloride (RHT); a
prerequisite for morphometry in growth cartilage. J . Ultrastruct.
Res., 81:l-12.
Hunziker, E.B., W. Hermann, and R.K. Schenk 1983 Ruthenium hexammine trichloride (RHTkmediated interaction between plasmalemmal components and pericellular matrix proteoglycans is
responsible for the preservation of chondrocytic plasma membranes in situ during cartilage fixation. J. Histochem. Cytochem.,
31:717-727.
Hunziker, E.B., W. Herrmann, R.K. Schenk, M. Mueller, and H. Moor
1984 Cartilage structure after high pressure freezing, freeze substitution and low temperature embedding. I. Chondrocyte ultrastructure-implications for the theories of mineralization and
vascular invasion. J. Cell. Biol., 98t267-276.
Hunziker, E.B., and R.K. Schenk 1989 Physiological mechanisms
adopted by chondrocytes in regulating longitudinal bone growth
in rats. J. Phys., 414:55-71.
Hunziker, E.B., R.K. Schenk, and L-M. Cruz-Orive 1987 Quantitation
of chondrocyte performance in growth-plate cartilage during longitudinal bone growth. J. Bone Joint Surg., 69-A.162-173.
Kember, N.F. 1960 Cell division in endochondral ossification: A study
of cell proliferation in rat bone by the method of tritiated thymidine autoradiography. J. Bone Joint Surg., 42-Bt824-839.
Kember, N.F. 1972 Comparative patterns of cell division in epiphyseal cartilage plates in the rat. J . Anat., lllt137-142.
Kember, N.F. 1973 Patterns of cell division in the growth plate of the
rat pelvis. J. Anat., 116.445-452.
Kember, N.F. 1978 Cell kinetics and the control of growth in long
bones. Cell Tiss. Kinet., 11t477-485.
Kember, N.F. 1983 Cell kinetics in cartilage. In: Cartilage. B.K. Hall
ed. Academic Press, New York, pp. 149-180.
Kember, N.F., and J.K. Kirkwood 1987 Cell kinetics and longitudinal
bone growth in birds. Cell Tiss. Kinet., 20t625-629.
Kember, N.F., and K.V.R. Walker 1971 Control of bone growth. Nature, 229t428-429.
Kuhn, J.L., J.D. Hornovich, and E.E. Lee 1993 The relationship between bone growth rate and hypertrophic chondrocyte volume in
the New Zealand white rabbit of varying age. Proc. 39th Orthop.
Res. SOC.,San Francisco, CA, p. 695.
Kumpulainen, T., and H.K. Vaanaanen 1982 Immunocytochemical
demonstration of extracellular carbonic anhydrase in epiphyseal
growth cartilage. Calcif. Tissue Int., 34.428-430.
Moss-Salentijn, L. 1974 Studies of long bone growth I: Determination
of differential elongation in paired growth plates of the rat. Acta
Anat., 90t145-160.
Moss-Salentijn, L., N.F. Kember, M. Shinozuka, W.F. Wu, and Anjana
Bose. 1991 Computer simulations of chondrocytic clone behaviour
in rabbit growth plates. J. Anat., 175t7-17.
Moss-Salentijn, L., M.L. Moss, M. Shinozuka, and R. Skalak 1987
Morphological analysis and computer-aided, three dimensional
reconstruction of chondrocytes columns in rabbit growth plates.
J. Anat., 151:157-167.
Nilsson, A., J. Isgaard, A. Lindahl, A. Dahlstrom, A. Skottner, and
O.C.P. Isaksson 1986 Regulation of growth hormone of the number of chondrocytes containing IGF-I in rat growth plate. Science,
233t571-574.
Poole, A.R., I. Pidoux, A. Reiner, H. Choi, and L.C. Rosenberg 1984
Association of an extracellular protein (chondrocalcin) with the
calcification of cartilage in endochondral bone formation. J. Cell.
Biol., 98t54-65.
268
G.J. BREUR ET AL
Quastler, H. 1959 The description of steady-state kinetics. In: The
kinetics of cellular proliferation. F. Stohlman, ed. Grune & Stratton, New York, pp. 431-433.
Quastler, H. 1960 Cell population kinetics. Ann. N. Y. Acad. Sci.,
9Ot580 -591.
Quastler, H., and F.G. Sherman 1959 Cell population kinetics in intestinal epithelium of the mouse. Exp. Cell Res., 17:420-438.
Rigal, W.M. 1962 The use of tritiated thymidine in studies of chondrogenesis. In: Radioisotopes and Bone. F. McLean, P. Lacroix,
and A.M. Budy, eds. Blackwell, Oxford, pp. 197-225.
Rothstein, A. 1989 The Na + /H + exchange system in cell pH and
volume control. Rev. Physiol. Biochem. Pharmacol., 112t235257.
Schenk, R. 1980 Basic histomorphology and physiology of skeletal
growth. In: Treatment of Fractures in Children and Adolescents.
B.G. Weber, Ch. Brunner, and F. Freuler, eds. Springer, New
York, pp. 3-19.
Schmidt, T.M., T.F. Linsenmayer 1985 Developmental acquisition of
type X collaEen in the embryonic chick tibiotarsus. Dev. Biol.,
iom73-381‘:
Scott, B.L., and D.L. Pease 1956 Electron microscopy of the epiphyseal
apparatus. Anat. Rec., 126r465-482.
Seinsheimer, F., and C.F. Sledge 1981 Parameters of longitudinal
growth rate in rabbit epiphyseal growth plates. J. Bone Joint
Surg., 63 -A:627-631.
Shimizu, K., T. Hamamoto, T. Hamakubo, W.J. Lee, K. Suzuki, Y.
Nakagawa, T. Murachi, and T. Yamamuro 1991 Immunohistochemical and biochemical demonstration of calcium-dependent
cysteine proteinase (calpain) in calcifying cartilage of rats. J. Orthop. Res., 9t26-36.
Simon, M.R., and P.S. Cooke 1988 Cellular heterogeneity and insulinlike growth factor I immunoreactivity among epiphyseal growth
plate chondrocytes in the pig. Acta Anat., 133r66-69.
Sissons, H.A. 1955 Experimental study on the effect of local irradiation on bone growth. In: Progress in Radiobiology. J.S. Mitchell,
B.E. Holmes, and C.L. Smith, eds. Edinburgh, Oliver and Boyd,
Edinburgh, pp. 436-448.
Snedecor, G.W., and W.G. Cochran 1980 Statistical Methods. Iowa
State University Press, Ames, 10,pp. 83-106.
Taylor, J.F., E. Warrell, and R.A. Evans 1987 Response of the growth
plates to tibia1 osteotomy in rats. J . Bone Joint Surg., 69-B:664669.
Thorngren, K-G, and L.I. Hansson 1973a Cell kinetics and morphology of the growth plate in normal and hypophysectomized rats.
Calcif. Tissue Res., 13:113-129.
Thorngren, K-G, and L.I. Hansson 197313 Cell kinetics of the growth
plate in hypophysectomized rats treated with GH and thyroxine.
2. Zellforschung, 142:431-442.
Thorngren, K-G, and L.I. Hansson 1981 Cell production in different
growth plates in the rabbit. Acta Anat., IIOt121-127.
Vaanaanen, H.K. 1984 Immunocytochemical localization of carbonic
anhydrase isoenzymes I and I1 in human bone, cartilage and giant cell tumor. Histochemistry, 81t485-487.
van der Rest, M., L.C. Rosenberg, B.R. Olson, and A.R. Poole 1986
Chondrocalcin is identical to the C-propeptide of type I1 collagen.
Biochem. J., 297t923-925.
Walker, K.V.R., and Kember, N.F. 1972 Cell kinetics of growth cartilage in the rat tibia. I Measurements in young male rats. Cell
Tissue Kinet., 5t401-408.
Walsh, K.B., S.D. Shawna, and R.E. Wuthier 1992 Characterization of
a delayed rectifier potassium current in chicken growth plate
chrondrocytes. Am. J. Physiology, 262(5Pti):1335-1340.
Westerborn, 0. 1961 The effect of papain on epiphyseal cartilage. A
morphological and biochemical study. Acta Chir. Scand. Suppl.,
270:l-84.
Wilsman, N.J., C.E. Farnum, G.J. Breur, and E.M. Leiferman 1993
Spatial arrangement of growth plate chondrocytes and a proposed
functional unit as represented by an “average equivalent column” of cells. Proc. 39th Orthop. Res. SOC.,San Francisco, CA, p.
694.
Wilsman, N.J., and D.C. VanSickle 1971 Histochemical evidence of a
functional heterogeneity in neonatal canine epiphyseal chondrocytes. Histochem. J., 3t311-318.