TSP-2 in the disc
TSP-2, a matricellular protein with recognized anti-angiogenic activity in vivo and in vitro, was found to be present in some, but not all, cells of the human and mouse annulus. Microarray analysis also verified expression of TSP-2 in the human annulus.
It is interesting that the response to a lack of TSP-2 expression in null mice was insufficient to produce vascular ingrowth into the disc. The disc may indeed be unusual in this regard, since surrounding soft tissues bordering the disc displayed the expected increased vascular bed previously seen in soft tissues of the TSP-2-null animal [
4]. The annulus contains TSP-2, but apparently does not rely solely upon it to inhibit blood vessel growth; TSP-2 may therefore not participate in regression of neonatal vessels in the disc.
Morphology of the annulus in the TSP-2-null mouse
Although there was no difference in the number of disc cells in the annulus of TSP-2-null mice versus WT animals, an important finding in the present study is the irregular, uneven collagen lamellar structure seen by polarized light microscopy in null mouse discs versus WT mouse discs. Appropriate annular morphology and integrity are essential to the function of the intervertebral disc, and cells in the outer annulus are polarized for directed secretion of ECM components [
24]. Ultimately it is the ECM that undergoes failure with disc degeneration; dehydration and matrix fraying culminate in tears within the annulus during biomechanical loading and torsion. Nucleus pulposus and annulus material rupture through these tears, and impinge on nerves, thus causing pain.
Kyriakides and colleagues found that dermal collagen fibers were disorganized in the TSP-2-null mouse, and that the skin of these mice displayed decreased tensile strength [
4]. These authors hypothesized that TSP-2 might function as a collagen fibril-associated protein that participates in the regulation of collagen fibril diameter and fibrillogenesis. This hypothesis suggests another avenue to be explored in disc cell biology. Other possibilities include the ability of TSP-2 to regulate the activity of matrix metalloproteinase-2 [
5] and tissue transglutaminase [
2]. Additional studies of the disc characteristics of the TSP-2-null mouse are needed to determine whether there are ultrastructural changes in collagen fibers similar to those seen in the tendons of TSP-2-null mice (that is, increased numbers of large-diameter fibrils) and to determine whether there is decreased tensile strength in TSP-2-null discs.
Vascular changes in soft tissue bordering the disc in the TSP-2-null mouse
Although there is a rich vascular supply to the developing and newborn disc, vascularity decreases with maturity, and the adult human disc is avascular. This condition is also present in many other species, including the sand rat, a small rodent model of spontaneous, age-related disc degeneration [
25,
26].
The small vascular beds along the dorsal and ventral annular surfaces, and vascularization of the vertebral endplate, constitute the main accesses to vasculature for the disc, and nutrients subsequently reach the cells of the disc via diffusion through the disc ECM. In humans and in some animals, including the sand rat, the endplate undergoes calcification with increasing age, and access to nutrients thus decreases further [
26‐
31].
Previous work has shown that the majority of cells in the outer annulus of the human disc and the sand rat disc contain TSP-1. Since TSP-2 has established anti-angiogenic properties, we hypothesized that this matricellular protein might contribute to the avascular state of the disc by its capacity to inhibit vascular ingrowth along the disc margin. One objective of this work was to examine mice with a targeted disruption of the TSP-2 gene to determine whether mice lacking TSP-2 would show enhanced vascularity of the adult annulus. Our studies show that, even in the absence of expression of the TSP-2 gene, vascular ingrowth into the body of the disc did not occur.
The results of the quantitative assessment of the vascular bed in soft tissue along the disc margins that we presented here are similar to those previously published by Kyriakides and colleagues, who counted blood vessels in adipose, dermis, and thymic tissues of WT mice and TSP-2-null mice [
4]. A similar increase in vasculature was therefore seen in the TSP-2-null mouse tissue in the margin of the disc. It is interesting to note that the previous investigators also showed that the differences seen in neonatal or embryonic dermis were not as great as those in adult tissue.
In the TSP-2-null specimens examined here, the expected increased vascular bed was present in the soft tissue margin adjacent to the disc. Although we have not measured any nutrient diffusion rates in these discs, we note that this increased adjacent vascularity may potentially result in an increased availability of nutrients to the disc.
It is also of interest to comment on the question of potential compensation between the two TSPs. There is no evidence to date for compensation. Although we have not tested whether upregulation of TSP1 might contribute to the lack of vascularization of the annulus in TSP-2-null mice, the evidence in the literature suggests that compensation by either of the two TSPs does not occur [
32].
Angiogenesis is seen in herniated disc tissue, and basic fibroblast growth factor has been noted in at least some of the blood vessels in the prolapsed disc. In contrast, no immunoreactivity for fibroblast growth factor was seen in intact, nonherniated discs [
33]. Vascular endothelial growth factor and platelet-derived growth factor have also been identified in herniated disc tissue [
34‐
37].
There is a complex biological relationship between the intact disc matrix and cells, and the nearby vasculature outside the disc. We now recognize a number of inhibitors of angiogenesis, as discussed in several reviews [
38‐
42]. There are angiogenesis inhibitors that function by inhibition of one, or more than one, angiogenic protein. Endogenous anti-angiogenic proteins have a number of interesting properties [
43]; some can specifically target newly-formed vasculature, but not older blood vessels. Relevant to the disc, tissue inhibitor of metalloproteinase-1 and ECM fragments, including those from collagen, merit further study of their anti-angiogenic potential. Also relevant to the disc are the findings that aggrecan may act in an anti-angiogenic factor [
44], as may other matrix proteoglycan components [
45].
Chondromodulin-I, which is present in the disc [
46], also acts as an endothelial cell growth inhibitor in fetal bovine cartilage, in growth plates, and in embryonic cartilaginous sites. In addition to TSPs, this matrix protein might exert an anti-angiogenic influence in the TSP-2-null mouse discs studied here. The work presented here points to the importance of additional studies of TSP-1-null mice and TSP-2-null mice. In addition, future studies of anti-angiogenic factors in the disc are needed to understand the change from the well-vascularized status of the fetal and young discs to the avascular adult human disc or small mammalian disc.