Main

DS, a major cause of congenital heart disease and mental retardation, results in the majority of cases from the presence of an extra chromosome 21; however, other genotypes are responsible only in 5%(1). Recognized clinically by characteristic facial and limb features and hypotonia, DS also presents a high incidence of congenital duodenal stenosis, immune system abnormalities, and increased risk of leukemia and Alzheimer-like dementia(2). Various skeletal abnormalities also currently associated with DS include short stature and short neck, thin calvaria, delayed suture closure, and hypoplasia of cartilage-derived basilar, facial, and nasal bones(3).

DS patients have been reported to present postnatal growth retardation accentuated during early childhood and adolescence(4–8), microcephaly-associated brain growth retardation(9), and delayed bone age for chronologic age(10). Partial GH deficiency secondary to hypothalamic dysfunction has recently been suggested as a cause of growth retardation(11).

Detailed prenatal DS screening data have also revealed recognizable skeletal features such as long iliac angle(12), short humerus(13), short fifth digit(14), and femur/foot length ratio(15). A recent axial skeletal radiographic study(16) reported fetal malformations with variable distribution, the most frequent being agenesis and malformation of nasal bones and the cervical segment in the axial region. Although no growth retardation or other particular growth anomalies are documented in human DS fetuses, data from DS animal models have revealed early cartilage anomalies in prenatal life(17) and small size in 15-d fetuses(18).

Morphologic analyses of fetal rib, vertebra, and long bone growth plate cartilage are currently incorporated into pathologic practice and have shown extensive association of several skeletal malformations, mostly of genetic origin (for a review, see19), with fetal cartilage structure alterations. These changes reflect disturbances in chondrocyte proliferation and/or matrix mineralization processes. The aim of this study was to examine the morphologic appearance of skull, vertebra, rib, and epiphyseal growth plate in a group of human DS fetuses as a tool for detecting abnormalities in their chondrocytic phenotypes during the second trimester of fetal life that could affect pre- and/or postnatal skeletal growth and mineralization.

METHODS

Material. Twenty-nine human fetuses 17-22 wk old from therapeutic abortions were included in this study. Autopsy was performed with parental consent. Cartilage from femur, vertebra, skull, and rib was collected within the first 12 h postmortem and kept for pathologic examination.

The DS group consisted of 16 fetuses (19 ± 1 wk, 7 male and 9 female) presenting a complete 21 chromosome trisomy confirmed by prenatal karyotype screening (Table 1). The characteristic DS phenotype was found among individuals with some differences. Cardiopathy was demonstrated in only four cases. Thirteen fetuses (21 ± 1 wk old; 9 male and 4 female) who presented non-growth-related pathologies were selected as control subjects (Table 1). No significant differences were found in relation to weight, crownrump length, heel-rump length, foot length, and rib width, between the two groups, although gestational age was slightly lower in DS (p = 0.002).

Table 1 General data from human fetuses with DS and other pathologies not related to growth
Full size table

Tissue preparation and histologic analysis. Rib (4th and 5th), skull, femur, and vertebra were collected and processed according to Emergy and Kalpaktsoglou(20). Briefly, the material was fixed in 10% formaldehyde embedded in paraffin for 10 d. After decalcification in nitrous acid/EDTA (24-48 h, depending on tenderness), paraffin-embedded tissue was cut into 5-µm thick sections (five for each sample) and stained by hematoxylin/eosin, Masson trichrome, periodic acid Schiff, and Alcian blue. Femur and rib sections were cut longitudinally, and transverse sections made in vertebra and skull after being macroscopically scanned to discard sections with artifactual damage. Intact tissue sections were selected for histoquantitative analysis. Morphologic examination was performed with an Olympus BH2 microscope in cartilage and trabecular bone portions of rib and femur, cartilage structures of vertebra, and membranous structure of the skull. Measurements were taken by a Reichter Visopan projection microscope by a single observer, and the mean of three measurements was recorded. The ossification line was adopted as a reference to orientate rib and femur sections. The spatial arrangement of resting, proliferative, and hypertrophic chondrocytic zones was clearly identifiable. The resting zone containing small, round cells extended from the outer end of the cartilage to the the proliferative zone, which included proliferating cells with lacunae forming isogenic groups; the hypertrophic zone included the columns of mature chondrocytes up to the ossification line. Each zone was measured and its proportion quantified.

Quantitative analysis. Data directly obtained from sections included: 1) skull and rib thickness; 2) vertebra, epiphyseal, and rib cartilage total length; and 3) percentages of resting cells, proliferative isogenic groups, and hypertrophic columns calculated in rib, femoral, and vertebral growth cartilages. Statistical analysis was performed using ANOVA/Mann-Whitney U test for nonparametric variables with the Statview software version 5.1 for Macintosh.

RESULTS

Skull cartilage study. Chondrocytes in the fetal skull are considered to be of neural crest origin and undergo a membranous ossification process that distinguishes them from other cartilages analyzed in this study. Although disturbance in the architecture of the craniofacial region is one of the most important consequences of DS, no irregular features were observed in transverse sections. All fetal skull sections presented a similar appearance, as expected at this stage of development, with spherical chondrocytes surrounded by a characteristic thick cartilaginous matrix (Fig. 1B). Thickness measurements showed no differences between groups (Table 2).

Figure 1
figure 1

Microscopic appearance of vertebra(A), skull (B), and femoral epiphyseal growth plate(C) in DS fetuses. Hematoxylin/eosin, 25×.

Full size image
Table 2 Histomorphometric analysis of the different zones of growth cartilage in different bones from human fetuses (17-22 wk) with DS or other fetal pathologies not affecting longitudinal skeletal growth (M± SD)
Full size table

Morphologic data of vertebra, rib, and epiphyseal growth plate cartilage. Endochondral ossification is a highly organized sequential transformation of cartilage into trabecular bone. The sequence at all endochondral ossification sites is similar, although in some areas of the skeleton, e.g. vertebrae, the sequence is foreshortened, or characteristically prolonged as at the costochondral junction. When total cartilage length of femoral epiphyseal growth plate, rib, and vertebra were compared, no significant differences were observed between DS and control subjects (Table 2).

Most cartilage consists of resting tissue containing a large amount of matrix within which single or paired cartilage cells appear to be randomly distributed, and much more extracellular matrix separates cells from each other than in other zones. Cartilage cells divide, near the ossification line, forming double and single rows of cells that display flattened morphology and are aligned parallel to the longitudinal axis. In the hypertrophic zone, the chondrocytes are enlarged and appear vacuolated with nuclear fragmentation. Length measurements of these zones in our study showed no significant differences between DS and non-DS fetuses for vertebra and femoral samples (Table 2). However, both hypertrophic and proliferative zones of the DS epiphyseal growth plate tended to increase in length at the expense of the resting zone. This tendency decreased when percentage values that normalize absolute data from the total extension of the growth plate were used (Table 2). Examination of trabecular bone adjacent to the DS growth plate provided no evidence of abnormal organization (Fig. 1C).

Several works have reported that rib, the most rapidly growing bone in a linear fashion(20,21), is the most useful for studying skeletal growth abnormalities. In addition, microscopic features of the costochondral junction are believed to be identical to those observed in growth plates of long tubular bones(22). In the present study, microscopic examination of DS costochondral junctions revealed significant previously unreported structural anomalies in contrast to the femoral growth plate (Fig. 2).

Figure 2
figure 2

Comparison of an identical length of DS(A) and control (B) costochondral junctions in the fetal rib. It includes hypertrophic (h) zone in A and B and proliferative (p) zone in B. Hematoxylin/eosin, 25×.

Full size image

Hypertrophic cells that formed a narrow line in the control subjects extended to form a thick stratum in DS. Hypertrophic DS chondrocytes not only demonstrated features characteristic of terminally differentiated cells, including enlarged nuclei and clear cytoplasm, but showed an altered organization of columns and adopted a network appearance in which distance between groups of cells tended to be shorter and cell-to-matrix ratio lower (Fig. 3). In addition to hypertrophic disorganization, histomorphometric data demonstrated a significant increase in the hypertrophic zone length accompanied by a decrease in that of the proliferative zone. Normalized data in percentages with respect to total length confirmed enlargement of the hypertrophic zone and reduction in proliferative and resting zones (Table 2 and Fig. 4).

Figure 3
figure 3

Detailed parts of the costochondral junction in DS (A-C) and control (D-F) subjects. The control includes resting (D), proliferative (E), and hypertrophic (F) zones. Equivalent portions in a DS are shown(A-C). Hematoxylin/eosin, 250×.

Full size image
Figure 4
figure 4

Percentage of total growth cartilage length occupied by each zone (resting, proliferative, and hypertrophic) (M± SD) in ribs from human fetuses with DS (▪) and other pathologies not affecting longitudinal skeletal growth (control,□).

Full size image

DISCUSSION

In epiphyseal and rib cartilage, several zones resulting from chondrocyte differentiation and maturation associated with quantitative and qualitative changes in cell morphology and matrix composition can be distinguished. These topographic differences in cell population morphology and cell/matrix ratio reflect chondrocyte maturation and stage of development, and underline the importance of histologic study of cartilage cell functions when skeletal growth alterations are analyzed.

In the present work, differences in growth cartilage of two groups of second trimester fetuses were studied in an attempt to ascertain whether special traits in cartilage development exist in DS. Statistically significant alterations were found in the hypertrophic portion of costochondral junctions. On the basis of data obtained, we suggest that the chondrodysplastic phenotype in DS-affected fetuses is partly due to a reduction in the proliferation process of chondrocytes, which seem to present an increased rate of differentiation. The fact that this anomaly appears to be restricted to the rib leads to the suspicion that it is a local early manifestation, as occurs in other pathologies(22), and could form part of a more general phenomenon.

Commitment to differentiation stages leads chondrocytes to assume changes in their proliferative capacity, acquisition, or loss of expression of specific genes and changes in their response to several growth factors and hormones. It is not feasible to recognize a priori the altered step which discretely supports DS chondrocyte hypertrophy, although genes that could act as a signal to hypertrophy have recently been described(23–25); however, their real function remains unknown. On the other hand, many studies support the important role of IGF-I, transforming growth factor-β, parathormone/parathormone-related peptide, triiodothyronine, 1,25-(OH)-vitamin D, and gonadal steroids(26–30) as main regulators of progression to hypertrophy and mineralization, a process which begins with the halt in proliferation and when type X collagen, alkaline phosphatase, osteocalcin, osteopontin, and bone sialoprotein are expressed(31).

In general, constitutional bone development disorders (a heterogeneous group which includes osteochondrodysplasia, dysostoses, idiopathic osteolysis, chromosomal aberrations, primary metabolic abnormalities, and miscellaneous disorders with osseous involvement) present structural growth cartilage abnormalities whose well known origin could provide some insights into the hypertrophy process. This is the case for hypophosphatasia, an autosomal recessive disorder caused by defective regulation of the alkaline phosphatase gene, which presents a similar appearance of broad columns of unmineralized hypercellular cartilage in the epiphyseal growth plate persisting deep into the metaphysis, or also vitamin D-deficient rickets, which presents enlargement of the hypertrophic zone but without the coincident characteristics observed in DS. In both cases, other authors suggest that the increase in the proportion of the hypertrophic zone results from the impossibility of progression to bone formation(32).

The abnormal chondrogenesis in DS could be due to a genetic imbalance, trisomy 21, that may affect chondrogenesis more or less directly. Although no results that might explain the skeletal anomalies found in human DS patients are currently available, studies in transgenic mice with altered Ets2 expression, which mimics the human DS phenotype, have been published. Expression of this proto-oncogene and transcription factor(17) located in chromosome 21 occurs in a variety of cell types. During murine development, Ets2 is highly expressed in newly forming cartilage, including skull precursor cells and vertebral primordia, and its overexpression in mice has been implicated in the genesis of some skeletal abnormalities. Complete phenotypic mapping of human DS, based on molecular analysis of chromosome 21 duplications, may permit establishment of the link between doses and effects of a particular gene.

Other alterations found in DS individuals, such as cellular aging(33), apoptosis, and increased generation of reactive oxygen species(34), although not specifically associated with skeletal development, may play a role in the cartilage anomaly described herein. However, using morphologic criteria (presence of pycnotic nuclei and apoptotic bodies) we observed no increase in apoptotic/necrotic cells in DS versus non-DS sections (data not shown).

In summary, to our knowledge, the data obtained in this study are the first concerning an abnormal process of cartilage development in fetuses with DS. The reduction in resting and proliferative zones suggests that chondrogenesis at the chondrocostal junctions may be accelerated, providing an example of a more general process that manifests postnatally in long bones, leading to diminished growth rates. Further studies are required to define the precise mechanisms by which systemic or local factors influence this increase in growth cartilage hypertrophy.