Histologic and Dynamic Changes Induced by Chronic Metabolic Acidosis in the Rat Growth Plate
EDUARDO CARBAJO*,
JOSÉ MANUEL LÓPEZ,
FERNANDO SANTOS,
FLOR ANGEL ORDÓÑEZ,
PILAR NIÑO* and
JULIÁN RODRÍGUEZ
*Department of Anatomy, School of Medicine, Hospital Central de Asturias,
University of Oviedo, Oviedo, Asturias, Spain. Department of Cell Biology, School of Medicine, Hospital Central de
Asturias, University of Oviedo, Oviedo, Asturias, Spain. Department of Pediatrics, IUOPA, School of Medicine, Hospital Central de
Asturias, University of Oviedo, Oviedo, Asturias, Spain.
Correspondence to Dr. Eduardo Carbajo,
Anatomía, Facultad de Medicina,
C/Julián
Clavería s/n, 33006 Oviedo, Asturias, Spain.
Phone: 34-98-510-3585; Fax: 34-98-510-3585; E-mail:
ecp3{at}correo.uniovi.es
Abstract. To understand better the pathophysiology of growth
impairmentin persistent metabolic acidosis, the morphology and dynamicsof
the growth plate were studied in young rats grouped as follows:rats that were
made acidotic by oral administration of ammoniumchloride for 14 d (AC),
nonacidotic rats that were fed ad libitum(control [C]), and
nonacidotic rats that were pair-fed withthe AC group (PF). AC rats became
markedly acidotic and growthretarded. The volume of newly formed bone per day
(mean ±SEM) was significantly lowered (P < 0.05) in AC
rats (AC,3.4 ± 0.4; C, 8.4 ± 0.6; PF, 6.4 ± 0.5
mm3/d).Growth plate height was lower in AC rats (303.8 ±
12.7µm) than in either C (478.0 ± 16.0 µm) orPF rats (439.0
± 21.4 µm). The processes of chondrocyteproliferation (assessed by
bromodeoxyuridine labeling) and maturation(assessed by stereologic estimators
of size and shape of chondrocytesand the volume of matrix per cell) were not
impaired by acidosis.By contrast, the dynamics of hypertrophic chondrocytes
werealtered significantly: both cell turnover per column per day(AC, 4.4
± 0.4; C, 8.0 ± 0.8; PF, 6.2 ±0.6) and linear velocity
of advance of chondrocytes (AC, 5.7± 0.5; C, 11.2 ± 0.9; PF,
9.4 ± 0.8 µm/h)were lowered significantly. The study presented here
shows theinhibitory effect of metabolic acidosis on cartilage cell
progressionand endochondral bone formation. Finally, the data show that
metabolicacidosis caused a marked shortening of the growth plate because
chondrocyteturnover was affected to a greater extent than bone tissue
formation.
Chronic metabolic acidosis is a common manifestation of renaldisease, such
as renal failure and renal tubular acidosis. Sustainedmetabolic acidosis
stunts growth
(1,2,3),
but the pathogenesisof this growth retardation is not understood fully.
Reducedfood intake, increased protein catabolism, acidotic state itself,and
growth hormone/insulin-like growth factor-1 dysfunctionhave been implicated
as pathogenic factors
(4,5,6,7,8).
The modifications induced by chronic renal failure and malnutritionon the
growth plate of long bones
(9,10,11),
the effector organsof longitudinal growth
(12,13),
have been studied recently.It is of note that despite that metabolic acidosis
frequentlyis associated with these disorders, no data are available onthe
modifications induced by chronic metabolic acidosis on thestructure and
dynamics of the growth plate.
Longitudinal bone growth is a complex phenomenon that resultsfrom
progressive replacement of growth plate cartilage by osseoustissue at the
metaphysis. Proliferation and hypertrophy of chondrocytesand matrix synthesis
are the main events that sustain growthat the epiphysial plate
(13,14).
Therefore, a detailed studyof the morphology and dynamics of the growth plate
in the presenceof persistent metabolic acidosis will add to our understanding
ofthe pathophysiology of growth impairment in this disorder. Accordingly,we
used stereologic and cytodynamic techniques to analyze thecharacteristics and
activity of chondrocytes in the growth plateof young rats with growth
retardation secondary to metabolicacidosis.
Animals and Experimental Protocol
Female Sprague Dawley rats (65 ± 5 g of body wt and 25± 3 d
of age) were obtained from the breeding area ofthe animal facility building
of the University of Oviedo. Ratswere housed in individual cages in a light-
and temperature-controlledenvironment (12 h light-dark cycle; 22 ±
1°C) andreceived standard 23.9% protein rat chow (AO3, Panlab SL,
Barcelona,Spain) and tap water. After 3 d of adaptation to the experimental
area,the rats were classified into three groups of five rats each(day 0 of
the protocol): acidotic rats (AC), rats that werepair-fed with AC (PF), and
control rats that were fed ad libitum(C). Acidosis was induced by
daily administration of ammoniumchloride (NH4Cl) as drinking
solution at progressively increasedconcentrations of 1.8% (days 0 to 5), 2.0%
(days 6 to 10), and2.2% (days 11 to 14).
Rats' weight as well as chow and water consumption were measureddaily with
an electronic balance Ohaus GT 2001 (Ohaus ScaleCorp., Florham Park, NJ).
Nose to tip tail length was measuredunder anesthesia on days 0 and 14. Food
efficiency was calculatedas grams of weight gained per gram of food consumed.
On day14, rats were killed by exsanguination under anesthesia with
methoxyflurane.All rats received intraperitoneal injections of calcein
(Sigma,St. Louis, MO; 15 mg/kg body wt) and 5-bromo-2'-deoxyuridine
(BrdU;Sigma; 100 mg/kg body wt) 4 d and 1 h before being killed,
respectively.Blood acid-base equilibrium was determined with a Ciba-Corning
855gas analyzer (Ciba-Corning Diagnostics Corp., Medfield, MA).
Tissue Collection and Processing
Tibiae were isolated immediately after death. Soft tissues wereremoved,
and tibia lengths and frontal and sagittal diametersat the level of the upper
growth plates were measured with theuse of a sliding mechanical caliper
(accuracy, 0.01 mm). Frontaland sagittal diameters were used for estimation
of the growthplate area perpendicular to the long axis, horizontal projection
area,according to the ellipse formula
(15). Blocks from proximal
tibialgrowth plates were chosen by a systematic random-sampling process
reportedpreviously (10). Two
blocks were used for measurement of therate of longitudinal bone growth, two
blocks were used for histomorphometricanalysis, and one block was used for
immunocytochemical identificationof proliferating chondrocytes.
The tissue processing procedure has been described
(10). Briefly,tissue blocks
for determination of longitudinal bone growthrate were fixed in 40% ethanol,
dehydrated in ethanol, and embeddedin Durkupan-ACM (Sigma). Tissue blocks for
histomorphometrywere transferred immediately to 2% glutaraldehyde solution
containing0.7% ruthenium hexamine trichloride (Strem Chemicals, Newburyport,
MA)for 3 h, transferred to 1% osmium tetroxide and 0.7% rutheniumhexamine
trichloride for 2 h, dehydrated in ascending concentrationsof acetone, and
embedded in Durkupan-ACM. Tissue blocks forimmunocytochemical identification
of proliferating chondrocyteswere fixed in a mixture of 0.5% glutaraldehyde
and 4% paraformaldehyde,dehydrated in ascending concentrations of acetone,
and embeddedin Durkupan-ACM.
Embedded tissues were cut on a Reicher Ultracut E ultramicrotome(Leica
Microsistemas S.A., Barcelona, Spain) parallel to thetibial vertical axis,
with the section angle oriented randomlyrelative to the horizontal plane for
each block. The mean thicknessof the sections, analyzed under a JEOL JEM-2000
EX II electronmicroscope (Izasa-Rego, Barcelona, Spain), was 1.08 ±
0.04(SD) µm.
Determination of Longitudinal Bone Growth Rate
Sections were examined under a Leitz incident light fluorescencemicroscope
(Leica Microsistemas S.A.), and the distance betweenthe zone of vascular
invasion and the proximal part of the calceinlabel was measured with an
eyepiece micrometer (16).
Measurementswere obtained at three locations, determined in an unbiased
manner,on each of the four sections per rat, and the mean of these
measurementsdivided by 4 (d) was considered as the longitudinal bone growth
perday in each rat. The volume of newly formed bone per day alsowas
calculated by multiplying the longitudinal bone growth perday by the growth
plate area perpendicular to the long axisof the tibia
(17).
Immunocytochemical Identification of Proliferating Chondrocytes
As described elsewhere
(10), semithin sections were
etched in50% sodium ethoxide and incubated overnight with an anti-BrdU
monoclonalantibody (1:20; Dakopatts, Glostrup, Denmark). This was followedby
incubation with biotinylated anti-mouse IgG for 60 min (BiomedaCorp., Foster
City, CA) and streptavidin-peroxidase complex(45 min at room temperature).
The final reaction product wasdeveloped by incubation with
3,3'-diaminobenzidine. Proliferativeactivity was quantified by the
estimation of the labeling index,which was defined as the percentage of
BrdU-labeled cells withinthe proliferative stratum.
Histomorphometry
Sections were stained with toluidine blue and photographed underthe light
microscope first at low magnification (x230). On theselow-magnification
prints, growth plates were divided into stemcell, proliferating, and
hypertrophic zones defined accordingto morphologic criteria
(15). The height of the
proliferativezone was calculated as the mean of three different measurements
performedat three locations chosen randomly on each section. The heightsof
the growth plate and its different zones were estimated bypoint counting from
the height of the proliferative zone. Thevolume of the growth plate and the
volume of the different zoneswere estimated by multiplying the horizontal
projection areaby the mean growth plate height and the mean height of the
differentzones, respectively.
From each section, two quadrants from the proliferating zoneand one from
the hypertrophic zone were sampled, photographed,and printed on paper with a
final magnification factor of x700.The following stereologic estimators
were determined withineach zone: volume fraction of chondrocytes, numeric
densityof chondrocytes, total number of chondrocytes, mean chondrocyte
volume,mean matrix volume per chondrocyte, mean projected horizontal
diameter,mean chondrocyte height, number of chondrocytes in a verticalcell
column, and growth fraction. Estimation of these parameterswas performed as
described in detail elsewhere
(10). Examinationof the
sections and photographs always was performed by thesame investigator in a
blinded manner (unaware of to which groupthe material under study belonged).
All measurements were donethree times, and means and SD were calculated. SD
always wasless than 10%. The mean values so obtained were used to calculate
thefinal estimates. The minimum number of points to be analyzedwas
determined according to Cruz-Orive and Hunziker
(15).
Cell Kinetics
Growth plate over a short time period (such as 4 d, the timeinterval
established here to calculate longitudinal growth)can be considered a steady
state system in which cell productionand elimination rates as well as growth
plate height are constant
(18).On this basis, the mean
cell turnover per column per day, theturnover time, and the linear velocity
of advance of chondrocytesin the columns were estimated as described
previously (10).
Statistical Analyses
Values of each group are given as mean ± SEM. For thegrowth plate
data, a mean value for each of the tested parameterswas calculated on a
per-rat basis in the group. Then, each ratwas considered a sample for
statistical purposes. The data obtainedwere shown to follow a normal
distribution with homogeneityof variances and independence. Comparison among
the three groupswas performed by ANOVA, with the use of a significance level
of95%, followed by the Newman-Keuls multiple range test.
The administration of NH4Cl induced a severe metabolic acidosis
asdocumented by the significant (P < 0.05) reduction in theserum
total CO2 of AC rats (13.4 ± 0.7 mM/L), comparedwith C rats
(24.4 ± 0.9 mM/L) and PF rats (25.2 ±1.1 mM/L). Mean daily food
intake of AC rats (9.2. ±0.3 g) and, therefore, of PF rats was
approximately 60% of thatof C rats (16.2 ± 0.2 g). The metabolic
acidosis wasassociated with growth retardation as demonstrated by weightand
length gains, tibial length, rate of longitudinal growth,and volume of newly
formed bone per day (Table 1).
All of theseparameters were was significantly lower in AC rats than in anyof
the groups of nonacidotic rats (C, PF). In turn, PF ratsgrew less than C
rats.
Table 1. Body and bone growth data (± SEM) in acidotic and nonacidotic
rats, either fed ad libitum or pair fed with the acidotic
groupa
Under the light microscope, the growth plate of AC rats wasshorter in
height than that of either PF or C rats (Figures
1and
2). The columnar organization
of the growth plate was wellpreserved in the three groups of rats, the
interface betweenthe growth plate cartilage and the metaphysial bone was well
defined,and a common pattern of invading capillary vessels was clearly
identifiable.
Figure 1. Parallel sections of the proximal tibial growth plates of control (C) rats
(a) and acidotic (AC) rats (b) examined by bright field microscopy after
toluidine blue staining (left) and by incident-light fluorescence
microscopy (right). The distance between the lower border of the
growth plate and the fluorochrome-labeled front clearly is longer in C than in
AC rats. Magnification, x50.
Figure 2. Light micrographs of vertical semithin sections of the proximal tibial
growth plate of C (a), AC (b), and pair-fed (PF) rats (c). The growth plate
clearly was shorter in AC rats than in any other group, mainly as a result of
a relative shortening of the hypertrophic zone. Magnification, x160
(toluidine blue staining).
Stereologic parameters describing the features of the growthplate
cartilage and chondrocytes are given in Tables
2 and
3,respectively. The height of
the entire growth plate was significantlylower in AC rats than in either of
the nonacidotic groups ofrats. The shortening of the growth plate in AC rats
was duemainly to a decreased height of the hypertrophic zone (approximately
50%of that of C rats) and, to a lesser extent, the reduction inheight of the
proliferative zone (approximately 70% of thatof C rats). No differences
between groups were observed in thevolume fraction or the numeric density of
chondrocytes, butthe total number of chondrocytes in the proliferative and
hypertrophiczones of the AC group was lower than in either of the othertwo
groups.
Table 3. Growth plate chondrocyte structural parameters (± SEM) in AC
rats and C or PF rats
Structural parameters (Table
3), related either to size (meancell volume) or shape (mean
projected horizontal diameter, meanheight of cell profile) of chondrocytes,
did not differ amonggroups. No differences were found for the ratio matrix
volumeper chondrocyte or for the mean number of chondrocytes per column.
Proliferative activity of chondrocytes did not differ significantlyamong
groups (Table 4). Regardless of
the group, the distributionof BrdU-labeled chondrocytes was fairly
heterogeneous (Figure 3),although labeled cells were confined to the proliferativezone. As for other
kinetic parameters, the growth fraction showedno differences among the three
groups of rats. However, thecell turnover per column per day clearly was
reduced in AC ratscompared with either of the nonacidotic groups. Estimation
ofdaily rates of cellular turnover showed that cell columns eliminate
approximately8 chondrocytes per day (1 chondrocyte every 3 h) in C rats and
6.2or 4.4 cells per day (1 chondrocyte every 3.9 or 5.5 h, respectively)in
PF and AC rats. Changes in cell turnover were not accompaniedby modifications
in the duration of the hypertrophic activityphase, but the linear velocity of
advance of chondrocytes inthe vertical columns decreased significantly in AC
rats whencompared with either PF or C rats.
Figure 3. 5-Bromo-2'-deoxyuridine—labeled cells in the proximal tibial
growth plate of C (a) and AC (b) rats. Sections counterstained with
hematoxylin. Magnification, x350.
In the study presented here, growth impairment induced by acidosislikely
was the result of the combined effect of reduced foodintake and the acidotic
state itself because AC rats grew lessthan PF rats, and, in turn, this group
with a normal acid-basebalance gained less weight and length than C rats. As
also shownin Table 1,
differences in body growth paralleled those foundin tibial length,
longitudinal bone growth rate, and volumeof recently formed bone.
More interesting, chronic metabolic acidosis caused significantalterations
in the growth plate that were not present in thetwo groups of nonacidotic
rats. As shown in Table 2, the
tibialgrowth plate of AC rats was much smaller, as confirmed by amarkedly
reduced height and volume, than that of PF and C rats.Growth plates of PF and
C rats had a similar size. Elongationof long bones is the result of the
interplay of two coupledprocesses: continual and vectorial production of
cartilage bythe growth plate and replacement of cartilage by bone tissueat
the epiphysial/metaphysial interface. A strict coordinationamong the
processes of cartilage enlargement, cartilage resorption,and osseous tissue
formation at the metaphysial end is requiredfor normal bone growth. Thus, the
association of stunted longitudinalgrowth and reduced growth plate height
found in the AC ratsindicates a disruption of this equilibrium because
cartilageproduction and bone formation both were decreased, althoughat
different intensity. In these rats, the production of cartilagewas slowed
significantly through the replacement of cartilageby bone, resulting in a
shortened growth plate. Our study showsthat the chondrocyte advance and its
turnover (Table 4), aswell as
the formation of osseous tissue (Table
1), were lowerin the AC rats, demonstrating the inhibitory effect
of metabolicacidosis on cartilage cell progression and endochondral bone
formation.Thus, the decreased height of the growth plate induced by acidosis
wasthe result of alterations at both sides of the calcificationfront. In
vitro studies
(19,20)
showed that acidosis exertsa direct effect on bone cells, suppressing
synthesis of collagenand alkaline phosphatase activity in osteoblasts and
stimulatingosteoclastic bone resorption. Therefore, the reduction of daily
boneformation observed in our AC rats may be linked to a decreasedcollagen
synthesis.
Our study also indicates that the processes of chondrocyte proliferation,
asassessed by the BrdU labeling index, and maturation, as assessedby
stereologic estimators of size and shape of chondrocytesand the volume of
matrix per cell, were not impaired by acidosis(Tables
2 and
3). It has been shown that
growth may be disturbedseverely in the presence of normal cartilage cell
proliferationat the growth plate
(12,21),
and, in agreement with the presentresults, growth plate chondrocytes of
growth-retarded uremicrats have been shown to have normal proliferative
activity (10).However, the
preserved morphology of chondrocytes at the differentstages of their
progression through the growth plate was a findingessentially dissimilar to
that found in other metabolic andrenal disorders associated with an altered
process of chondrocytematuration, such as vitamin D—deficient rickets
(22) andsevere chronic renal
failure (10). In rachitic
rats, the hypertrophicchondrocytes become much bigger than normal, whereas
the hypertrophicchondrocytes of uremic rats are abnormally small. Like the
acidoticrats, knockout of the PTHrP gene causes very short growth
plates,but in these transgenic mice, the chondrocytes do not proliferate
normallyand the process of chondrocyte differentiation is acceleratedso that
their extracellular matrix is mineralized prematurelyand the cells undergo
early apoptosis
(23,24).
Although markersof chondrocyte differentiation, e.g., type X
collagen synthesis,were not analyzed, the stereologic analysis performed in
ourstudy suggests that the process of chondrocyte maturation wasunaltered in
the AC rats. Therefore, the mechanism responsiblefor the slowed transit of
chondrocytes along the growth plateremains to be established. Because the
movement of chondrocytesis not passive but requires vectorial degradation of
matrixat the metaphysial cell pole and resynthesis at the epiphysialcell
pole, it is tempting to speculate that acidosis impairsthis process of matrix
remodeling at the hypertrophic zone ofthe growth plate. Evidence supporting
this hypothesis clearlydeserves further investigation.
In conclusion, our study characterizes, for the first time,the
modifications caused by sustained metabolic acidosis inthe young rat's growth
plate. In addition, our findings suggestthat acidosis is not the main
pathogenic factor responsiblefor the marked impairment of chondrocyte
maturation found inthe growth plate of severely uremic animals
(10).
Acknowledgments
This research was supported by grant PM96-0110 from the
DirecciónGeneral de
Investigación
Científica y
Técnicadel Ministerio de
Educación y Cultura.
Cooke RE, Boyden DG, Haller E: The relationship of acidosis and
growth retardation. J Pediatr57
: 326-337,1960[Medline]
Santos F, Chan JCM: Renal tubular acidosis in children.
Am J Nephrol 6:289
-295, 1996
Maniar S, Caldas A, Laouari D, Kleinkecht C: Severity of chronic
metabolic acidosis and growth of rats with chronic uremia. Miner
Electrolyte Metab 18:241
-244, 1992[Medline]
McSherry E: Acidosis and growth in no uremic renal disease.
Kidney Int 14:349
-354, 1978[Medline]
McSherry E, Morris RC Jr: Attainment and maintenance of normal
stature with alkali therapy in infants and children with classic renal tubular
acidosis. J Clin Invest 61:509
-527, 1978
Potter DE, Greifer I: Statural growth of children with renal
disease. Kidney Int 14:334
-339, 1978[Medline]
Challa A, Chan W, Krieg RJ Jr, Thabet MA, Liu F, Hintz RL, Chan
JCM: Effect of metabolic acidosis on the expression of insulin-like growth
factor and growth hormone receptor. Kidney Int44
: 1224-1227,1993[Medline]
McSherry E: Renal tubular acidosis in childhood. Kidney
Int 20: 799-809,1981[Medline]
Cobo A, Carbajo E, Santos F, García
E, López JM: Morphometry of uremic rat growth
plate. Miner Electrolyte Metab22
: 192-195,1996[Medline]
Cobo A, López JM, Carbajo E, Santos
F, Alvarez J, Fernández M, Weruaga A: Growth
plate cartilage formation and resorption are differentially depressed in
growth retarded uremic rats. J Am Soc Nephrol10
: 971-979,1999[Abstract/Free Full Text]
Sanchez CP, Salusky IB, Kuizon BD, Abdella P, Juppner H, Goodman
WG: Growth of long bones in renal failure: Roles of hyperparathyroidism,
growth hormone and calcitriol. Kidney Int54
: 1879-1887,1998[Medline]
Breur GJ, Farnum CE, Padgett GA, Wilsman NJ: Cellular basis of
decreased rate of longitudinal growth of bone in pseudoachon-droplastic dogs.
J Bone Joint Surg Am 74:516
-528, 1992[Abstract/Free Full Text]
Hunziker EB: Mechanism of longitudinal bone growth and its
regulation by growth plate chondrocytes. Microsc Res
Tech 28: 505-519,1994[Medline]
Mehls O, Ritz E, Hunziker EB, Eggli P, Heinrich U, Zapf J:
Improvement of growth and food utilization by human recombinant growth hormone
in uremia. Kidney Int 33:45
-52, 1988[Medline]
Cruz-Orive LM, Hunziker EB: Stereology for anisotropic cells:
Application to growth cartilage. J Microsc143
: 47-80,1986[Medline]
Hansson LI, Menander-Sellman K, Stenstrom A, Thorngren KG: Rate of
normal longitudinal bone growth in the rat. Calcif Tissue
Res 10: 238-251,1972[Medline]
Breur GJ, Turgai J, Vanenkevort BA, Farnum CE, Wilsman NJ:
Stereological and serial section analysis of chondrocytic enlargement in the
proximal tibial growth plate of the rat. Anat Rec239
: 255-268,1994[Medline]
Hunziker EB, Schenk RK: Physiological mechanisms adopted by
chondrocytes in regulating longitudinal bone growth in rats. J
Physiol 414:55
-71, 1989[Abstract/Free Full Text]
Krieger NS, Sessler NE, Bushinsky DA: Acidosis inhibits
osteoblastic and stimulates osteoclastic activity in vitro. Am J
Physiol 262:F442
-F448, 1992[Abstract/Free Full Text]
Bushinsky DA: Stimulated osteoclastic and suppressed osteoblastic
activity in metabolic but not respiratory acidosis. Am J
Physiol 268:C80
-C88, 1995[Abstract/Free Full Text]
Hunziker EB, Wagner J, Zapf J: Differential effects of IGF-I and
hGH on the various developmental stages of growth plate chondrocytes in vitro.
J Clin Invest 93:1078
-1086, 1994
Dean DD, Muñiz OE, Berman I, Pita
JC, Carreño MR, Woessner JF, Howell DS:
Localization of collagenase in the growth plate of rachitic rats. J
Clin Invest 76:716
-722, 1984
Karaplis AC, Luz A, Glowacki J, Bronson RT, Tybulewicz VL,
Kronenberg HM, Mulligan RC: Lethal skeletal dysplasia from targeted disruption
of the parathyroid hormone-related peptide gene. Genes
Dev 8: 277-289,1994[Abstract/Free Full Text]
Amizuka N, Henderson JE, Hoshi K, Warshawsky H, Ozawa H, Goltzman
D, Karaplis AC: Programmed cell death of chondrocytes and aberrant
chondrogenesis in mice homozygous for parathyroid hormone-related peptide gene
deletion. Endocrinology 137:5055
-5067, 1996[Abstract]
Received for publication May 22, 2000.
Accepted for publication November 21, 2000.
Top
Abstract
Introduction
