This theory is the most widely held hypothesis. Hypertrophic chondrocytes of the growth plate undergo death by apoptosis, leaving behind a frame of cartilage matrix for osteoblasts that invade and lay down bone. Some supporting and opposing results for this hypothesis have been reported in several studies⁸⁾. Apoptosis-regulating proteins (such as caspases) were expressed in the growth plate and typical histologic changes in cell during apoptosis were shown in several studies in rat⁹⁾. However, another study found no signs of classical apoptosis in fusing human growth plates¹⁰⁾.

Autophagy is another method of programmed cell death that involves a catabolic process in which the cell degrades its own components through autophagosomes. Signs of autophagy (autophagosomes, double-membrane structures, and condensed chromatin) were observed in avian hypertrophic chondrocytes and chondrocytes of newborn mice^11,12). However, there were no autophagosomes or signs of autophagy in the growth plate in a study in humans⁷⁾.

Emons et al.¹⁰⁾ reported a dense border of thick bone surrounding growth plate remnants in a human growth plate tissue specimen undergoing epiphyseal fusion. They postulated that the dense border might act as a physical barrier preventing oxygen and nutrients from reaching the fusing growth plate, resulting in hypoxia and eventually cell death in a nonclassical apoptotic manner.

Stewart et al.¹³⁾ also observed increased expression of hypoxiainducible factor 2α mRNA during chick and murine chondrocyte differentiation in vitro.

This is the oldest hypothesis, in which terminal hypertrophic chondrocytes transdifferentiate into osteoblasts at the chondroosseous junction of the growth plate¹⁴⁾. This theory is mostly based on organ and cell culture models and there is a lack of direct evidence in human studies⁷⁾.

The conversion of progenitor cells in the mesenchymal condensation into chondrocyte lineage is controlled by the expression of the transcription factors Sox9, 5, and 6⁸⁾. Parathyroid hormone-related peptide (PTHrP) signaling, Indian hedgehog (IHH) pathway, and runt-related transcription factor 2 (Runx2) are required for further differentiation and hypertrophy of chondrocytes. In addition to those factors, fibroblast growth factor (FGF) pathway and bone morphogenic proteins (BMP) are necessary for the development of the perichondrium, periosteum, chondrocytes, and osteoblasts^7,15).

Of these paracrine factors, the best studied one is PTHrP that is known as secreted by periarticular chondrocytes of long bones^16,17,18). Hirai et al.¹⁹⁾ reported that PTHrP diffuses across the growth plate cartilage maintaining chondrocytes in the proliferative state.

IHH is secreted by prehypertrophic and hypertrophic chondrocytes and positively regulates PTHrP production. And it also has independent effects on chondrocyte differentiation¹⁸⁾. BMP signaling across the growth plate is considered to contribute to the progressive differentiation of resting to proliferative to hypertrophic chondrocytes^20,21,22). FGF and its receptor (FGFR) system are also important for growth plate development. Results from various in vivo studies indicate that FGFR1 and FGFR3 are growth-inhibiting, while FGFR-2 is growth-promoting^23,24).

Vascular invasion is a prerequisite for the replacement of avascular cartilage by vascular bone and marrow. The most important stimulating factor for vessel invasion is hypoxia. Vessel formation is mediated by transcription factors, including hypoxia-inducible factor-1 and vascular endothelial growth factor (VEGF)^25,26).

Runx2, FGFs, BMPs, transforming growth factor (TGF), insulin-like growth factor (IGF), and platelet-derived growth factor (PDGF) are also required for VEGF expression and vessel formation²⁷⁾.

There are five steps in differentiation of the osteoblastic lineage. These steps are the preosteoblast, mature osteoblast, osteoid osteocyte, early osteocyte, and mature osteocyte²⁸⁾. Wingless-type mouse mammary tumor virus integration site (Wnts)/β-catenin signaling, Runx2, Osterix, Ihh, BMPs, and IGFs are required for these processes. During these steps, bone can deposit minerals from the extracellular matrix rich in type I collagen, completing ossification⁴⁾.

Growth hormone (GH) and insulin-like growth factor-I (IGF-I) are the main stimulators of longitudinal bone growth. They are also important for the acquisition of bone mass during the prepubertal period and maintenance of bone homeostasis throughout life. GH stimulates the synthesis and secretion of IGF-I in the liver and growth plate. Longitudinal bone growth is mediated by GH, circulating IGF-I, and more importantly, local IGF-I in the growth plate. For differentiation, proliferation, and hypertrophy of chondrocytes; the production of extracellular matrix; and ossification in the growth plate, IGF-I produced from chondrocytes of the epiphyseal plate is important²⁹⁾. There is a close interplay between estrogen and GH in the regulation of growth and development in puberty. During puberty, there can be a 1.5- to 3-fold increase in the pulsatile secretion of GH and a more than 3-fold increase in the concentration of serum IGF-I. However, the interactions between estrogen and GH at the growth plate remain unclear because there is a lack of evidence of the independent roles of these two hormones at the cellular level in the growth plate^30,31).

During puberty, estrogen induces the stimulation of the GH-IGF-I axis and a pubertal growth spurt. In the classical pathway, estrogen acts after binding with its receptor (estrogen receptor, ER). There are two subtypes of ER, ERα, and ERβ. Both subtypes seem to be involved in the augmentation of GH secretion and are expressed in the resting, proliferative, and hypertrophic zones of the growth plate. Indirect evidence suggests that epiphyseal fusion occurs when the proliferative capacity of growth plate chondrocytes is exhausted and estrogen acts by advancing growth plate senescence. Therefore, the binding of estrogen with each subtype of ER is thought to be related to the pubertal growth spurt and epiphyseal fusion^32,33,34,35).

In mice studies, ERα looks like the dominant mediator of estrogen actions in bone. ERβ has some repressive functions in bone of females. Evidence for a role of ERβ in males is lacking. A definite role for ERβ in humans remains unclear³⁶⁾.

Borjesson et al.³⁷⁾ studied the mechanisms of the pubertal growth spurt and epiphyseal fusion with cartilage-specific ERα knockout mice and reported that the effects of estrogen and ERα on growth plate thickness were unremarkable in early puberty, but obvious in late puberty. They suspected that activation of the GH-IGF-I axis with low doses of estrogen is important for the pubertal growth spurt in early puberty, while high doses of estrogen binding to its receptors in growth plate cartilage are essential for epiphyseal fusion in late puberty.

The effects of estrogen on the growth plate have been studied in many rodent models. However, there are some differences in growth plate physiology between rodents and humans. In rats or mice, the pubertal growth spurt is unremarkable and longitudinal bone growth continues even after sexual maturation, as epiphyseal fusion does not occur at the time of sexual maturation. Therefore, there are difficulties in the direct application of growth plate studies in rodents to humans⁵⁾.

Further molecular studies are needed to elucidate the consequences of estrogen and the ER in growth plates. Estrogen is known to promote bone formation by stimulating osteoblastogenesis and inhibiting apoptosis of mature osteoblasts. It is also known to decrease the osteoclast production by inhibiting the reaction with receptor activator of nuclear factor-κB ligand or production of interleukin-1, -6, -7, and tumor necrosis factor-α. Estrogen may also increase osteoclast apoptosis by stimulating the Fas/FasL pathway⁴⁾.

Androgen itself also contributes to bone formation and the pubertal growth spurt, perhaps through a direct interaction with growth plate chondrocytes³⁸⁾. Dihydrotestosterone can stimulate proliferation and proteoglycan synthesis in growth plate chondrocytes in vitro and testosterone stimulates chondrocyte proliferation in an organ culture model study with increased local IGF-I production⁵⁾.

In epiphyseal fusion, the activity of androgen is due to the aromatization of androgens to estrogens in various peripheral tissues, including growth plate cartilage³⁶⁾.