Main

Growth failure is a distinctive feature of many chronic diseases in children, such as CRF (1), inflammatory bowel diseases (2), rheumatoid arthritis (3), and cystic fibrosis (4). In mammals, poor statural growth results from disordered longitudinal bone growth, which in turn is primarily due to abnormal GP chondrogenesis (5). Chronic illnesses associated with growth retardation share a number of underlying mechanisms known to impair the function of the long bones' GP. These mechanisms are: 1) inflammation, 2) protein/calorie deprivation, 3) uremia/metabolic acidosis, 4) GH/IGF-I impaired action, and 5) GC (endogenous and exogenous).

This article reviews the experimental evidence supporting the association between these processes, impaired GP chondrogenesis, and longitudinal bone growth in chronic illnesses.

GP CHONDROGENESIS AND LONGITUDINAL BONE GROWTH

Longitudinal bone growth occurs at the GP (5). The cartilaginous GP is organized into three functionally and structurally distinct layers, the resting, the proliferative, and the hypertrophic zones (Fig. 1). In the resting zone, chondrocytes are irregularly arranged in a bed of cartilage matrix and rarely divide. Farther toward the metaphysis, in the proliferative zone, the chondrocytes show a flattened shape and are arranged in columns oriented parallel to the long axis of the bone. The proliferative chondrocytes farthest from the epiphysis stop replicating and instead enlarge to become hypertrophic chondrocytes. These terminally differentiated cells, which form a layer adjacent to the metaphysis termed the hypertrophic zone, eventually undergo apoptosis. Longitudinal bone growth occurs by endochondral ossification, a two-step process in which cartilage is first formed and then remodeled into bone. GP chondrocyte proliferation, hypertrophy, and extracellular matrix secretion lead to formation of new cartilage, chondrogenesis. After the terminally differentiated chondrocytes undergo apoptosis, the GP is invaded from the metaphysis by blood vessels and bone cell precursors that remodel the cartilage into bone. These two processes, chondrogenesis and ossification, are tightly coupled so that the width of the GP remains relatively constant while new bone is formed at the junction of the GP and the metaphyseal bone. The rates of GP chondrogenesis and, in turn, of longitudinal bone growth are regulated by multiple systemic (endocrine) factors (Fig. 2A). In addition, the underlying cellular processes of proliferation, differentiation, angiogenesis, and ossification appear to be regulated by a network of local (paracrine) factors, expressed in the GP (Fig. 2B).

Figure 1
figure 1

Histology of the GP.

Figure 2
figure 2

(A) Endocrine regulation of GP chondrogenesis. (B) Paracrine regulation of GP chondrogenesis. The arrow indicates on which GP zone systemic (A) or locally expressed (B) growth factors act directly, by stimulating (+) or by inhibiting (–) chondrocyte proliferation (in the GP resting and proliferative zones) or chondrocyte differentiation/hypertrophy (in the GP hypertrophic zone). IGF-I acts both as an endocrine and a paracrine growth factor. T3/T4, triiodothyronine//thyroxine; FGF, fibroblast growth factors; BMP, bone morphogenetic proteins; Ihh, Indian hedgehog; Sox9, sex determining region Y-box 9; Runx2, runt-related transcription factor 2; PTHrP, PTH-related protein.

INFLAMMATION

Juvenile rheumatoid arthritis (JRA) is a chronic inflammatory disease often associated with growth failure. Serum and synovial levels of IL-6, a major mediator of inflammation, are significantly elevated in children with JRA (6). In a recent study, it was found that transgenic mice overexpressing IL-6 experienced a stunted growth rate that led to a 50–70% smaller size when compared with age-matched littermates (7). Neutralization of IL-6 activity by a MAb produced a partial improvement of the animals' growth rate. IL-6 transgenic mice were also found with reduced circulating levels of IGF-I (which promotes GP chondrogenesis acting as an endocrine as well as a paracrine factor) and normal levels of GH. In the same study, similarly reduced IGF-I levels were found in 21 children with JRA, with serum IGF-I levels being negatively correlated with IL-6 levels. Other pro-inflammatory cytokines, like IL-1 and tumor necrosis factor-α (TNF-α), are also significantly elevated in chronic inflammatory illnesses such as JRA (8) or inflammatory bowel diseases (IBD) (9). In cultured whole rat metatarsal bones, high concentrations of both IL-1 and TNF-α impaired metatarsal longitudinal bone growth, decreased GP chondrocyte proliferation, and increased chondrocyte apoptosis (10). In contrast, IL-6 did not affect longitudinal bone growth. In another study, IL-1 affected proliferation of cultured rat costochondral chondrocytes, whereas IL-6 had no effect (11) All these studies suggest that the major inflammatory cytokines (IL-6, IL-1, and TNF-α) inhibit GP chondrogenesis and longitudinal bone growth either directly at the GP (IL-1 and TNF-α) or by reducing the systemic effects of IGF-I. Children with inflammatory bowel disease (more Crohn's disease than ulcerative colitis) often present with growth failure. At diagnosis, approximately 90% of children with Crohn's have short stature (12), which persists in 30–40% of these children. In an experimental model of Crohn's disease (trinitrobenzene sulfonic acid–induced colitis) in rats, it has been demonstrated that inflammation independently inhibits linear growth (13). In these animals, immunoneutralization of IL-6 normalized serum IGF-I levels and linear growth. On the other hand, immunoneutralization of another mediator of inflammation, TNF-α, induced a similarly stimulatory effect on the rats' growth without modifying their serum IGF-I levels. These findings would suggest that inflammatory cytokines may inhibit growth through multiple mechanisms (IGF-I and non-IGF-I mediated). In another study, experimental colitis induced in rats resulted in an abnormal GP morphology, with an increase of the resting zone's and reduction of both the proliferative and hypertrophic zone height (14).

Protein/Calorie Deprivation

Protein/calorie deprivation (either due to malnutrition or malabsorption) is known to inhibit longitudinal bone growth (15,16). Simmons et al. (17) first reported growth velocity in children treated with maintenance hemodialysis that was comparable to those with intact renal function when calories were supplemented to 70% of the recommended dietary allowance (RDA). Betts and Magrath (18) found a correlation between growth velocity and energy intake in children with CRF and concluded that diminished linear growth would occur when energy intake was <80% of the RDA.

Experimental evidence indicates that malnutrition affects the systemic GH-IGF-I system. In rats, fasting diminishes both GH production (19) and hepatic GH sensitivity. The latter is associated with reduced hepatic GH receptor mRNA, GH binding, IGF-I mRNA, and circulating IGF-I (2024). In humans, fasting induces increased circulating GH levels (25) and decreased IGF-I and GH binding protein levels (26). With respect to the effects of fasting on the GP, a 3-d fast in rats reduced the animals' growth rate to 30% of that of control animals, with a dramatic reduction of their overall GP height and with an apparent decrease of the chondrocyte number and volume in all the GP zones (27). In another study, rabbits fasted for 2 d exhibited decreased tibial growth velocity and GP width. In the fasted animals, serum IGF-I levels and hepatic IGF-I mRNA were decreased despite increased GH levels, suggesting hepatic resistance to GH (28).

UREMIA/METABOLIC ACIDOSIS

Growth retardation is a major manifestation of CRF in children. The uremic rat model has been extensively used to investigate growth impairment in CRF. In uremic rats, the height of the GP has been found to be greater (29,30), equal (31,32), or smaller (33,34) than control animals. Some evidence indicates that GP morphology may vary according to the degree of secondary hyperparathyroidism. Sanchez et al. (34) have reported that, compared with control animals, the height of the GP cartilage of the proximal tibia of uremic rats remained unchanged in animals with mild hyperparathyroidism, but was markedly decreased in those with severe hyperparathyroidism. In contrast, increased growth cartilage and hypertrophic zone heights were observed in uremic rats with hypercalcemia, hypophosphatemia, and depressed serum PTH levels induced by calcium supplementation (32). Another study suggests that the height of the GP may depend on the severity and duration of renal failure, with a positive correlation between the degree of renal failure and the height of GP cartilage (30). Thus, only the animals with severe renal insufficiency exhibit an increased size of their GP. In rats with a milder degree of renal insufficiency, the size of the GP as well as the longitudinal bone growth rate may not be different from control. A consistent finding is that the increased GP height in uremic rats mainly results from an expansion of the GP hypertrophic zone (29), likely due to a prolonged duration of the hypertrophic phase. In addition to the expansion of the GP hypertrophic zone, the disorganized chondrocyte columns in the proliferative zone also suggests an overall disturbed GP chondrogenesis in uremic rats (29). A decreased expression of GH receptor protein (32) and IGF-I mRNA (35) in the proliferative zone of uremic rat GP are supportive of reduced chondrocyte proliferation. In nephrectomized rats fed a high calcium diet (which induces biochemical changes consistent with adynamic osteodystrophy), linear growth and tibial length were reduced, with the heights of the whole GP and of the GP hypertrophic zone significantly increased (36). Such morphologic changes were associated with a diminished chondroclastic/osteoclastic activity (decreased TRAP staining and MMP-9 mRNA expression), likely responsible for reduced cartilage degradation and resorption. Chronic metabolic acidosis, which is typically seen in renal diseases such as renal failure and renal tubular acidosis, is associated with growth retardation. Experimental studies have shown that metabolic acidosis decreases pulsatile GH secretion and reduces serum IGF-I levels (37). In addition, it has been shown that animals with severe metabolic acidosis exhibit reduced expression of chondrocyte IGF-I mRNA in the epiphyseal GP (38) and the lack of response to exogenous GH administration (39). Rats made acidotic by administration of ammonium chloride for 14 d and compared with pair-fed control rats exhibited a reduced longitudinal growth rate and a thinner GP (40). The decreased height of the GP was mainly due to a reduced height of the hypertrophic zone and, to a lesser degree, of the proliferative zone. In an in vitro study, murine mandibular condyles cultured for 3 d in acidic conditions (pH 7.15) grew significantly less than control bones cultured in neutral conditions (41). In addition, acidosis down-regulated cartilage matrix proteoglycan and collagen II synthesis, and the expression of IGF-I, IGF-I receptor, and PTH receptor in the GP.

GLUCOCORTICOIDS

GC are widely used as anti-inflammatory and immunosuppressive drugs in children with chronic diseases. Long-term, high-dose GC treatment often leads to growth failure, which in mammals reflects impaired longitudinal bone growth (42).

To understand the mechanisms underlying GC-mediated growth failure, several investigators have focused their attention on the systemic and local effects of GC on the GP, the site where longitudinal bone growth takes place.

Systemic administration of GC in mice causes reduced whole GP width and tibial length, and growth retardation (43,44). Similar results have been observed in rats (45) and in rabbits (46). Such effects on the GP width are likely due to both a decreased number of proliferative chondrocytes (43,47) as well as increased apoptosis of hypertrophic chondrocytes (43,48,49). The apoptotic effect of GC in the GP is also confirmed by the observed increased expression of apoptotic proteins, caspase-3 (49) and Bax (50), and decreased expression of Bcl-2 and Bcl-x (49,50) anti-apoptotic proteins. Regarding the molecular mechanisms underlying chondrocyte decreased proliferation and increased apoptosis, short-term systemic administration of GC in rodents decreases IGF-I expression in the GP (43,51), whereas long-term treatment (1 mo) increases it (52). Growth-suppressive doses of dexamethasone given to rabbits increase GH receptor mRNA expression in the GP.

The suppressive effects of GC on GP chondrogenesis and longitudinal bone growth may be indirect (mediated by other systemic growth factors) and/or direct (mediated by the activation of the GC receptor in the GP) (45,52,53). A direct effect of GC in the GP has been shown by a study in which local infusion of dexamethasone into a rabbit tibial GP caused a decreased growth rate of the treated tibia compared with the contralateral untreated one (54). To support this finding, observations in cultured chondrogenic cell line (55) and primary GP chondrocytes (56) have shown a suppressive effect of dexamethasone on cell proliferation. In addition, GC suppress GH receptor expression in cultured rat GP chondrocytes, whereas type-1 IGF receptor expression is not affected (56). In contrast, GC induces type-1 IGF receptor expression in porcine chondrocytes (57).

In conclusion, a large body of data consistently indicates that GC suppress statural growth by inhibiting GP chondrogenesis. In contrast, conflicting experimental evidence exists on the molecular mechanisms underlying the GC-mediated suppression of GP function.

GH/IGF-I AXIS

As discussed so far, most of the mechanisms responsible for growth retardation associated with chronic illnesses disturb the GH/IGF-I action on GP chondrocytes.

The effects of GH on the GP are in part mediated by the systemic effects of IGF-I. On the other hand, local GH injection in a tibial GP stimulates the ipsilateral tibial growth rate, without affecting the contralateral one (58). In cultured GP chondrocytes, GH induces resting zone chondrocyte proliferation and IGF-I secretion (59), which induces expansion of proliferating chondrocytes acting as a autocrine/paracrine factor. In support of a direct effect of GH on the GP, GH receptor has been demonstrated on chondrocytes in rabbit (60), rat (61), and human GP (62).

The possibility that GH has IGF-I-independent effects on longitudinal bone growth has long been debated. A recent study has shown that, in mice with targeted deletion of IGF-I, tibial growth rate was reduced by 35%, whereas it was reduced by 65% in GH receptor null mice (63). The IGF-I null mice showed a GP with a significantly enlarged resting zone, a normal GP proliferative zone, and a reduced hypertrophic zones. In contrast, GH receptor null mice exhibit all GP zone significantly narrowed.

GH receptor null mice demonstrate attenuation of both numbers and size of GP chondrocytes, whereas IGF-I null mice exhibit only reduced chondrocyte size. The fact that the resting zone is enlarged and chondrocyte proliferation normal in the GP of IGF-I null mice supports the view that GH enhances chondrocyte generation and proliferation independent of IGF-I. Further support of an IGF-I-independent effect of GH on longitudinal bone growth derives from the increased width of the GP resting zone in IGF-I null mice injected with GH (64).

Additional experimental evidence suggests that GH acts in the GP by inducing resting cells to enter the proliferative cycle (65), whereas IGF-I stimulates GP chondrocyte proliferation and hypertrophy (64).

Abnormalities of the GH–IGF-I axis are thought to be one of the main mechanisms underlying growth failure in children with chronic diseases. Evidence indicates that a state of GH resistance, rather than GH deficiency, typically occurs in chronic illness. Children with cystic fibrosis (66,67), chronic bowel inflammatory diseases (68,69), or JRA (70) exhibit normal GH secretory pattern and low IGF-I and IGFBP-3 levels. In children with CRF, serum GH levels are normal or elevated (71,72). Tonshoff et al. (73) have shown down-regulation of hepatic GH receptor gene expression in uremic animals. In addition, circulating GH binding protein levels were found to be decreased in children and adults with CRF (74). In concordance with these findings, a reduced (nutrition-independent) GH receptor mRNA level has been described in uremic rats (73,75). With respect to circulating IGF-I levels, they tend to be normal in children with preterminal CRF (76), whereas they are slightly decreased in end-stage renal disease (77). To explain the latter finding, it has been hypothesized that the inhibitory effect on IGF activity may be due to an excess of high-affinity IGFBP, especially IGFBP-1, -2, -4, and -6 (78,79).

CONCLUSIONS

Growth failure in chronic illness is primarily due to impaired longitudinal bone growth. The rate of growth elongation depends on the rate of GP chondrogenesis, which is regulated by the interaction of multiple endocrine and paracrine signals. Processes such as malnutrition, acidosis, and uremia, and molecules like inflammatory cytokines and GC impaired the cellular events of chondrogenesis, acting systemically and/or locally at the GP (Table 1). Such detrimental effect on GP chondrogenesis is often, but not always, due to a state of relative GH resistance and decreased IGF-I bioactivity, characterized by reduced GH receptor and IGF-I expression and altered expression of IGFBPs. Further studies are needed to elucidate the causative role of other growth factors expressed in the GP in the growth failure associated with chronic diseases.

Table 1 Underlying mechanisms of growth retardation in chronic illnesses and their effects on the cellular processes characterizing GP chondrogenesis and ossification