Fibrillin-1 and -2 differentially modulate endogenous TGF-β and BMP bioavailability during bone formation
Harikiran Nistala, Sui Lee-Arteaga, Silvia Smaldone, Gabriella Siciliano, Luca Carta, Robert N. Ono, Gerhard Sengle, Emilio Arteaga-Solis, Regis Levasseur, Patricia Ducy, Lynn Y. Sakai, Gerard Karsenty, and Francesco Ramirez
INTRODUCTION
The extra-cellular matrix (ECM) has been shown to be an important component of growth signaling and development. Amongst the ECM proteins, fibrillins and their associated proteins have been found to be important for cell maturation and growth and are typically bound by TGF-β and BMP. This paper looks at the effects of mutated proteins/genes for fibrillin-1 and -2 individually, focusing on TGF-β and BMP and their effects on osteoblast growth in the context of Marfan Syndrom (MFS) and congenital contractural arachnodactyly (CCA).
Figure 9. Model of microfibril-mediated control of osteoblast maturation. The scheme summarizes the distinct contributions of osteoblast-produced fibrillin-1 and fibrillin-2 microfibrils to osteogenic differentiation through the differential regulation of endogenous TGF-β and BMP signals that together calibrate the rate of bone formation.
RESULTS
Figure 2. Impaired maturation of Fbn2-null osteoblasts. (A) Illustrative images of von Kossa staining of neonatal cOb (left) and adult MSCs (right) along with magnified van Geison–counterstained images (middle) of cOb differentiated for 21 d after OS administration. Histograms summarize the number of mineralized nodules in WT and Fbn2-null (Fbn2-/-) cOb (n = 15) and MSC (n = 5) cultures. Bars, 200 μm. (B) AP activity of WT and Fbn2-/- cOb measured 3 d before (day -3) and 4 d after (day 4) OS administration (day 0) and normalized to total protein levels (n = 5). (C) qPCR estimates of indicated transcripts in total RNA isolated from day 4 cOb cultures (n = 4; top) or P4 calvariae of WT and Fbn2-/- mice (n = 4; bottom). (D and E) MTT and BrdU incorporation assays (D) and C-myc and Ccnd1 mRNA levels (E) at the indicated days of cOb differentiation (n = 6). (F and G) Cell survival evaluated by Trypan blue exclusion (F; n = 4) and cell apoptosis (G) measured as the fraction of cleaved caspase-3 over full-length protein (with histograms representing densitometric analyses) in WT and mutant cObs. Error bars indicate mean ± SD, and asterisks indicate statistically significant differences (P <>
Figure 3. Osterix and collagen down-regulation in Fbn2-null osteoblasts. (A and C) Illustrative images of pOBCol2.3GFP (A) or Osx-GFP::Cre transgene expression (C) in Fbn2-null and WT cOb at day 17 or 4 of differentiation, respectively, with histograms summarizing the number of GFP-positive cells (n = 3); nuclei are DAPI stained. (B) Illustrative images of tibiae of Fbn2-/- and WT mice harboring the pOBCol2.3GFP transgene that include (from top to bottom) hematoxylin/eosin staining, GFP expression, and magnified views of toluidine- and GFP-positive cells; histograms summarize cell counts in the last images (n = 3). (D) von Kossa staining after 21 d of differentiation of cOb treated with or without 100 ng/ml rhBMP2 with histograms (right) summarizing numbers of mineralized nodules and levels of Osx and Col1a2 transcripts in mutant and WT samples (n = 3). Error bars indicate mean ± SD, and asterisks indicate statistically significant differences (P <>
Figure 4. Fibrillin-2 controls TGF-β bioavailability. (A) Surface plasmon resonance sensograms of binding of immobilized L1K (left) and L4K (right) to rF86 at concentrations between 0 (baseline tracing) and 200 nM (top tracing). Binding was recorded as resonance units (RU), and nonspecific binding to control surface was subtracted at a molar ratio of 1:1. The table summarizes the affinity data (expressed as Kd) for interactions between the LTBPs and fibrillin-2 peptide. (B) Immunodetection of nuclear pSmad2 accumulation in day 4 cOb cultured in low serum with or without SB431542; nuclei are DAPI stained. Histograms summarize the percentage of pSmad2-positive nuclei in WT (black) and mutant (gray) cells (n = 3). (C) Transcriptional activity of p3TP-lux reporter plasmid transfected in WT or Fbn2-null cOb cultured in low serum (n = 3). (D) TMLC bioassays (n = 5) measuring active TGF-β in WT or Fbn2-null cOb cultures (left) or total TGF-β in heat-activated conditioned media of the same cultures (right). (E) qPCR estimates of Tgf-β transcripts in WT and Fbn2-null cOb (n = 3). Error bars indicate mean ± SD, and asterisks indicate statistically significant differences (P <>
Figure 5. Elevated TGF-β signaling limits Fbn2-null osteoblast maturation. (A) Illustrative images of immunoreactive material corresponding to the indicated proteins deposited in the ECM of overconfluent WT, Fbn1-null, and Fbn2-null cOb cultures after 4 d of differentiation; nuclei are DAPI stained. Bars, 50 μm. (B) Maturation of WT and Fbn2-null cOb cultures treated with 1 μM SB431542, 300 ng/ml neutralizing pan–TGF-β antibody, or 50 μM Alk5 siRNA with histograms summarizing the number of mineralized nodules in each treatment (n = 3). (C) qPCR estimates of the indicated mRNA levels in the Alk5 silencing experiments. Error bars indicate mean ± SD, and asterisks indicate statistically significant differences (P <>
Figure 6. Abnormally high TGF-β activity in differentiating Fbn1-null osteoblasts. (A) Illustrative images of von Kossa–stained WT and Fbn1-null (Fbn1-/-) cOb after 21 d of differentiation with histograms summarizing the number of mineralized nodules in each sample (n = 5). (B) Cell proliferation of WT and mutant cOb at day -3 of cell culture as assessed by BrdU incorporation and qPCR quantification of C-myc and Ccnd1 transcripts (n = 3). (C and D) qPCR estimates of indicated transcripts in total RNA isolated from day 4 differentiating WT and mutant cOb cultures (C; n = 4) and from P4 WT and Fbn1-/- calvarial bones (D; n = 3). (E) TMLC bioassays (n = 5) measuring active TGF-β in WT or Fbn1-null cOb cultures (left) or total TGF-β in heat-activated conditioned media of the same cultures (right). (F) Transcriptional activity of p3TP-lux reporter plasmid transfected in WT or Fbn1-null cOb cultured in low serum with or without 1 μg/ml of noggin (n = 3). (G) qPCR estimates of TGF-β transcripts in WT and Fbn2-null cOb (n = 3). Error bars indicate mean ± SD, and asterisks indicate statistically significant differences (P <>
Figure 7. Fibrillin-1 regulates BMP signaling in cultured osteoblasts. (A) Immunodetection of nuclear pSmad1/5/8 accumulation in day 4 differentiating WT and Fbn1-null (Fbn1-/-) and Fbn2-null (Fbn2-/-) cOb cultured in low serum; nuclei are DAPI stained, and histograms summarize the percentage of pSmad1/5/8-positive nuclei in WT (black), Fbn1-null (white), or Fbn2-null (gray) cells (n = 3). Bars, 25 μm. (B) Transcriptional activity of pBRE-lux reporter plasmid in day 4 differentiating WT, Fbn1-null, and Fbn2-null cOb (n = 5). (C) C2Cl2BRA bioassay measuring BMP signaling in conditioned media from WT, Fbn1-null, and Fbn2-null cOb and from the first two cell cultures treated with either 1 μg/ml noggin or 1 μM SB431542 (n = 3). (D and E) qPCR estimates of Bmp transcripts in day 4 Fbn1-/- and Fbn2-/- cOb cultures, respectively (n = 3). Error bars indicate mean ± SD, and asterisks indicate statistically significant differences (P < id="p-3">Figure 8. Microfibrils are not a structural substrate for matrix mineralization. (A) Illustrative von Kossa–stained WT cOb cultures in which Fbn1 or Fbn2 expression was silenced by RNAi with histograms on the immediate and far right showing the number of mineralized nodules and levels of indicated transcripts (n = 3), respectively. (B) Illustrative von Kossa–stained Fbn2-/- cOb cultures in which Fbn1 was silenced with histograms on the immediate and far right, showing the number of mineralized nodules and levels of indicated transcripts, respectively (n = 3). Error bars indicate mean ± SD, and asterisks indicate statistically significant differences (P < href="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEjDeAk42Bq4q3eP7jMCQcWz9gL_SOkNd5F9Eh800aXmrwE5oFq1TxpUGjF1w-qC8uEf9hC7aIaaTo1y2EIT_HN_rjSIoPKhUaX4_yY-4IX49neG3JiBi60SY7U0hgobqKvAN4Eq/s1600/fig9.bmp">DISCUSSION
In short, fibrillin-1 which is associated with MFS is shown to impair bone growth, as the absence of the related gene results in faster maturation than typical cells. On the other hand, the lack of fibrillin-2 is associated with CCA results in impaired "bone formation and osteoblast maturation." In the end, it can be said that the combination of fibrillins as well as their associated and binding proteins are responsible for both "anabolic and catabolic phases of bone homeostasis."
The TGF-β and BMP proteins are important in the specificity of the cellular signaling and resulting cascades. The combination of these two proteins is important only in the combined number of the two. The relative amounts in that sum has been shown to be unimportant. All of these effects, however, are said to be bone-specific after observing the effects of MFS and CCA. The details of this specificity is still unknown and needs to be further explored for a more comprehensive understanding of these protein effects and diseases.
METHODS
Critique
The conclusions drawn from the data of cultured cells were for endogenous activity; however, the experiments used exogenous stimulators/inhibitors. These in vitro conclusions are probably reasonably reliable however, as it is probably the activity levels rather than the activators themselves that matter. It would still have been nice if there was some way to implement an endogenous set of stimulators/inhibitors, so that the conclusions could be definite. On the other hand, the experiments on the modified mice did prove to be amongst the first to unambiguously show the importance of the fibrillin bindings and associations in the ECM for osteogenic differentiation. The combination of the findings from the cultured cells and modified mice also serve to reinforce the conclusions drawn from the in vitro experiments.
A couple of issues were not dealt with in these set of experiments. Firstly, the other cascades that could have resulted from high TGF-β activity were not accounted for. Any other resulting cascade could have influenced the observed effects or even produced some nuances. There is no reason to state that the observations and results are the effect of this single pathway. Another issue left untouched was the interactions between signaling events. Events occur in both the space and time dimensions and the signals are bound to have some sort of effect on each other. Spatiotemporal signaling events and consequences were not accounted for, and this something that should be looked into.