We have described a TM-inducible Col2-Cre™
line that can be used to selectively activate Cre in chondrocytes, overcoming a limitation of current Col2-Cre
transgenic lines that express Cre in both chondrocytes and the perichondrium (Long et al., 2004
; Sakai et al., 2001
). Serendipitously, we found that the new line also targets a subset of cells associated with the trabecular bone surface, likely reflecting the endogenous activity of the Colα1(II)
promoter that directs expression of the prechondrogenic Colα1(II)A
isoform in these cells. Finally, utilization of the Col2-Cre™
line in localized removal of Smo
has provided evidence that direct Ihh input is likely required for both PTHrP
expression in the periarticular chondrocytes, and the proper morphology of columnar chondrocytes.
The versatility of the Col2-Cre™ line lies in the fact that both cell type-selectivity (chondrocytes vs. perichondrial cells) and localized activity (subsets of chondrocytes) can be achieved by adjusting the timing or dosage of TM administration. It should be noted however, that although a higher TM dosage generally correlated with more cells expressing Cre activity, certain variability has been observed among embryos from pregnant females receiving the same amount of TM. In addition, the location of Cre-active cells within the growth plate also varied among embryos. These observations reflect a stochastic aspect of TM-mediated Cre activation, likely resulted from the variable availability of TM to each cell especially when a lower dosage was used.
In addition to its utility in embryonic studies, the Col2-Cre™ line also promises to be a useful tool for postnatal genetic studies. In particular, our results with young pups showed that nearly all of columnar chondrocytes activated Cre after receiving two doses of TM. Similarly, a high percentage of the articular chondrocytes exhibited TM-induced Cre activity. Thus the Col2-Cre™ line may be useful for postnatal studies involving either the growth plate or the joints.
Several lines of evidence from the present study support that direct Ihh input is likely required for periarticular chondrocytes to express PTHrP. First, loss of Ihh response in subsets of periarticular chondrocytes correlated with the loss of PTHrP in those cells, whereas cells outside the affected domains maintained their normal PTHrP expression profile. Second, loss of Ihh response in chondrocytes outside the normal PTHrP-expressing periarticular domain did not diminish PTHrP expression. Third, Ihh response in the perichondrium did not correlate with periarticular PTHrP expression. However, the current data does not exclude the possibility that other signals may emanate from within the PTHrP-expressing group of cells and function in either an autocrine or a paracrine manner.
The immediate regulator of PTHrP downstream of Ihh is presently unknown. In an effort to examine potential direct regulation of the PTHrP promoter by Ihh, we identified several putative Gli binding sites approximately 4 kb upstream of the first exon of the murine PTHrP gene. However, studies of promoter fragments containing these potential sites by transient transfections in C3H10T1/2 cells have not revealed a consistent response to exogenous Hh. The most trivial explanation for the negative results is that the Hh responsive elements were not included in the tested fragments. However, since PTHrP promoter activity in the embryonic growth plate generally correlates with low- but not high-level Ihh signaling, as indicated by the low level of Ptch1 in the PTHrP-expressing cells, it is possible that a “proper” level of Hh signaling required for activating PTHrP expression was not achieved in the cell culture model. Finally, Hh may not directly control the promoter activity of PTHrP. Future studies are required to distinguish these possibilities.
The present study indicates that direct Ihh input may be required for the proper morphology of columnar chondrocytes. Local removal of Ihh responsiveness resulted in abnormal cell morphology and disorganization of chondrocytes within the affected regions. Specifically, the number of flat cells was greatly reduced and the normal “stacks” of cells were no longer evident. The morphological defect was reminiscent to that observed in Col2-Cre; Smon/c
embryos where the entire columnar region was absent (Long et al., 2006
), but the underlying mechanism remains unknown. Ihh could directly control the cell morphology, or, as previously suggested, regulate a “differentiation” process that encompasses the morphological changes (Kobayashi et al., 2005
). Alternatively, the loss of flat cells could be secondary to the marked decrease in cell number within the region in the absence of Ihh signaling (Long et al., 2001
). In any case, the regulation appears to be mediated by antagonizing Gli3 repressor activity, as removal of Gli3
restored the columnar chondrocytes in the absence of Ihh
(Hilton et al., 2005
Finally, the current study has identified a Col2-Cre™-positive cell population primarily associated with the trabecular, but not the cortical, bone surface. These cells appear to represent only a subset of the bone surface cells, and are morphologically heterogeneous as appearing on tissue sections. The role of these cells in vivo remains unknown at present, but they are capable of producing mineralized bone nodules in vitro. Molecularly, they do not express high levels of Colα1(I) or Bsp. However, it should be noted that in the “co-localization” experiment, tissue sections were first subjected to Lac Z staining prior to in situ hybridization for Colα1(I) or Bsp mRNA. Because we have noticed that the staining procedure notably reduces the in situ hybridization signals, it is possible that the Col2-Cre™-positive cells express a lower level of Colα1(I) or Bsp that is below the current sensitivity threshold. Finally, since the endogenous Colα1(II) promoter is known to be active in skeletal precursors during embryogenesis, the Col2-Cre™-expressing cells may represent a progenitor population uniquely associated with the trabecular bone both in the embryo and in postnatal life. Future studies are necessary to determine whether this is indeed the case.