Sarkar, Gobinda

Abstract Increasing evidence suggests a potential role for activin in bone formation. However, the cognate receptors through which activins function with respect to skeletal tissues have not yet been identified. Identification and regulation of expression of these receptors are necessary prerequisites to understanding the role of activins in bone metabolism. We detected mRNAs for three activin receptors, type I (ActRI), type II (ActRII), and type IIB (ActRIIB), in multiple skeletal tissues in rat, including tibia and costochondral growth plate, and also in cultured osteoblasts. To gain information about the relationship between receptor expression and different skeletal cell functions, we evaluated expression of the three receptors in a semiquantitative manner during the early stages of fracture healing, a model for rapid bone formation. Relatively high levels of ActRI and ActRII expression were detected in the callus at 7, 10, and 14 days after fracture, times that correlate with the interval of rapid intramembranous bone formation and the initiation of endochondral bone formation. Expression of the ActRIIB in the fracture callus was strikingly lower than either ActRI or ActRII. Immunostaining of the fracture callus and the newborn rat femur with an anti-ActRII antibody localized the receptor to osteoblasts at regions of membranous and endochondral bone formation. No staining of osteoblasts in fracture callus or bone was seen with an anti-ActRIIB antibody. These results provide strong evidence of the identification of the principal receptors through which activins could function in the skeletal system and further shed light on activin's mechanism of action in bone formation.

Abstract Knowledge of the cis-acting elements is required for identifying trans-acting splicing factors underlying cartilage-specific alternative splicing of Col2 pre-mRNA. By performing desired deletions in the mouse Col2 pre-mRNA, location of the intronic cis-acting elements was narrowed down to be at or near splice-junction sequences flanking exon 2 of the gene. Introduction: Type II collagen (Col2) pre-mRNA undergoes cartilage-specific alternative splicing involving exon 2 during chondrocyte differentiation. Thus, the trans-acting protein factors that regulate the splicing are associated with the differentiation of chondrocytes. Knowledge of the cognate cis-acting elements is necessary to eventually identify the trans-acting factors. Materials and Methods: To localize the cis-acting sequences, we created several deletions within a minigene containing exon 1 to exon 4 of mouse Col 2 gene and evaluated alternative splicing of the resulting pre-mRNAs in ATDC5 cells, a model of insulin-stimulated chondrocyte differentiation. The first deletion reduced intron 1 from 3799 to 259 bp, the second reduced intron 2 from 1108 to 94 bp, the third combined the above two deletions, and the fourth was derived from the third by removing intron 3 and exon 4. ATDC5 cells harboring these constructs were cultured for up to 21 days with or without insulin. Alternative splicing was evaluated by determining the ratio of Col2B (lacks exon 2) to Col2A (has exon 2) RNAs by reverse transcription-polymerase chain reaction. Results: The deletion in intron 1 had no effect on the alternative splicing while other deletions affected splicing (demonstrated by the presence of splicing intermediates) in cells cultured without insulin or with insulin for 1 week. The splicing intermediates were not seen from any construct when cells were cultured longer (14–21 days) with insulin. Conclusion: These results show that the 259-bp intron 1, the 94-bp intron 2, and exon 2 sequences retained in the fourth construct provide cis-acting signal sufficient for insulin-induced cartilage-specific alternative splicing of Col2 pre-mRNA.