About Me

Originally published as JCI: Bones can also affect your metabolism and appetite | intensive reading of the literature Foreword Recent studies have shown that bone, as an endocrine organ, secretes at least three hormones: fibroblast growth factor 23 (FGF23), apolipoprotein 2 (LCN2), and osteocalcin (OCN). Osteoblasts are cells that synthesize OCN, a hormone involved in the regulation of glucose and energy metabolism, among other functions. On a normal diet, Ocn-deficient mice (Ocn-/-) exhibit reduced glucose tolerance, insulin sensitivity, and circulating insulin levels, as well as reduced energy expenditure and increased fat content. Mechanistic studies in cell culture and mice have shown that OCN regulates glucose better by promoting insulin secretion in pancreatic beta cells, promoting glucose uptake in muscle fibers, and increasing energy expenditure. Mouse experiments indicate that the function of OCN in beta cells and myofibers is mediated by group C 6 member A of the G protein-coupled receptor family (GPRC6A). This pathway appears to be preserved in humans because human OCN can bind to and activate human GPRC6A, and mutations or genetic polymorphisms in human GPRC6A are associated with insulin resistance. OCN is a small protein (46 amino acids in mice and 49 amino acids in humans) that is synthesized by osteoblasts and whose three glutamic acid residues (Glu) are gamma-carboxylated before being secreted. γ-Carboxylation of OCN occurs in the endoplasmic reticulum and is mediated by γ-glutamyl carboxylase (GGCX), a process that requires reduced levels of vitamin K as an essential cofactor. This post-transcriptional modification increases the affinity of OCN for hydroxyapatite, a mineral component of the extracellular matrix (ECM) of bone cells. Therefore, most of the OCN secreted by osteoblasts are deposited in the ECM of bone matrix, constituting a large number of non-collagen polypeptides. Although both γ-carboxylated (Gla) and partially carboxylated OCN (ucOCN) can be detected in serum, most in vivo and in vitro studies have shown that the endocrine function of OCN in humans and mice is performed by ucOCN. Recent studies have shown that by specifically inactivating Ggcx in osteoblasts, circulating levels of ucOCN in mice can be increased and glucose tolerance can be improved. Other studies have shown that osteoclasts are responsible for the partial decarboxylation and activation of OCN in bone ECM. Taken together, these findings suggest that inactivated γ-carboxylated OCN is synthesized by osteoblasts and deposited in the bone ECM, and is activated by osteoclasts and released as ucOCN during bone resorption. Like many other peptide hormones, the sequence of Ocn cDNA suggests that it was first synthesized as a preprohormone consisting of a signal peptide, a propeptide, and a mature hormone. Propeptides of such preprohormones, such as prothrombin and cofactor IX, are characterized by a signal recognized by gamma-glutamyl carboxylase, the propeptide of which is gamma-carboxylated. However, the importance of γ-carboxylation of the propeptide and OCN secretion has not been studied in vivo. In addition, it is not clear whether the removal of OCN propeptide depends on the γ-carboxylation process. OCN is unique among all known γ-carboxylated proteins because it can be efficiently γ-carboxylated in the absence of the propeptide. Finally, and most importantly, who are the endopeptidases that process OCN prohormones (pro-OCN), and the importance of OCN in bone endocrine function is unclear. Therefore, we searched for pro-OCN converting enzymes. For this purpose, we focused on proprotein convertases (PCs). PCs are a class of serine proteases that target specific motifs commonly found in prehormonal sequences consisting of basic residues such as arginine or lysine. These enzymes act in the secretory pathway or outside the cell to cleave other proteins, activate or inactivate them, and then participate in the regulation of many biological processes.
Proprotein convertases subtilisin/Kexin include proprotein convertase 1 (PC1), proprotein convertase 2 (PC2), furin, proprotein invertase 4 (PC4), and proprotein invertases 5A and 5B (PC5A and PC5B). Paired basic amino acid lyase 4 (PACE4) and PC7, which play a key role in regulating the maturation and secretion of a variety of peptide hormones such as insulin, glucagon, adrenaline (ACTH), glucagon-like peptide 1 (GLP-1), and parathyroid hormone (PTH). In this paper, we adopted cytological and genetic based parameters and found that furin functions as an endopeptidase during pro-OCN processing in osteoblasts. We also found that OCN γ-carboxylation and OCN disposal in osteoblasts are two independent processes in cell culture and in vivo. In furin-deficient mouse osteoblasts, proteolysis of pro-OCN is critical for activation of the hormone. We also found that furin may affect appetite by regulating energy metabolism through pathways unrelated to OCN. Fig. 1. PC cleaves pro-OCN at the RXRR motif of osteoblasts. Expand the full text A. Schematic representation of the precursor protein of the precursor hormone pro-OCN, including the approximate location of Gla residues and amino acid sequences of OCN propeptide sequences from different vertebrates: the conserved RX (R/K) R motif is colored yellow. The conformance symbols are contained in the following sequence. Single asterisks represent fully conserved residues, colons represent highly conserved residues, and periods represent moderately or weakly conserved residues. B. WB analysis of cell supernatants from primary osteoblasts transfected with OCN-V5 or the R46a/R48a/R49a OCN mutation (OCNAAA-V5), all labeled at the C-terminus with a V5 epitope. C. WB analysis of endogenous OCN in cell supernatants from differentiated calvarial osteoblasts from mice treated with or without 50.m u.M Dec-RVKR-CMK (RVKR). D. WB analysis of supernatants and cell extracts from CHO-IdID cells transfected with OCN-V5, treated with or without 50 μM Dec-RVKR-CMK (RVKR). E. WB analysis of supernatants from primary osteoblasts transfected with OCN-V5 or OCNAAA-V5 with or without treatment with 50 μM Dec-RVKR-CMK or 20 μM D6R. The C-terminal motif of the OCN propeptide consists of 3 basic amino acid residues: Arg-Leu-Arg-Arg (RLRR49) (R, Arg, arginine; L, Leu, leucine), which is highly conserved in vertebrates (Fig. 1A). This basic residue motif has been found to be a recognition site for PCs in a variety of secreted proteins, so we hypothesized that pro-OCN may be cleaved by one or more PCs in osteoblasts. When the RLRR sequence was mutated to ALAA, the molecular weight of OCN secreted in the culture medium of primary cultured osteoblasts changed, which was consistent with the molecular weight when the propeptide in the OCN mutant protein was retained (Fig. 1B), supporting our hypothesis that the RLRR sequence is the cleavage site of PCs. Similarly, primary cultured osteoblasts were treated with Dec-RVKR-CMK, a cell-permeable subtilisin/kexin-like PCs inhibitor, Resulting in changes in the molecular weight of endogenous OCN secreted by primary cultured osteoblasts or V5-labeled osteoblasts or CHO-ldlD cells (fig. 1 C-E). Dec-RVKR-CML-treated and ALAA mutations induced the same molecular weight change, suggesting that RLRR is at least the major, if not the only, sequence recognized and cleaved by PC (Figure 1E).
Finally, treatment of osteoblasts with hexa-D-Darginine (D6R), a non-permeable inhibitor that blocks extracellular but not intracellular PCs, was found to have no effect (fig. 1E),wiped film evaporator, indicating that lysis occurs intracellularly. These data suggest that intracellular PC contributes to the conversion of pro-OCN to OCN in osteoblasts. Fig. 2. Furin cleaves pro-OCN in vitro. A. Relative mRNA expression levels of undifferentiated and differentiated osteoblasts in kexin-like PCs in mice. The copy number was calculated from the standard curve of mouse DNA, and Actb was used as an internal reference. B. Analysis of GST-pro-OCN treated in vitro. Each group of furin, PC5A, PC7, and PACE4 contained the same number of enzyme units at 0 or 15 min of incubation. Released OCN was measured by WB and OCN/GST-pro-OCN relative values. C. Evaluation of the time course of furin treatment of GST-pro-OCN by WB. D. The effect of increasing furin dose on GST-pro-OCN treatment in vitro was evaluated by WB, and the action time was 60 min. The effect of 30 min (E) or various incubation times (F) of furin on GST-pro-OCN and R48a/R49a OCN mutations (GST-RR/AA-pro-OCN) in vitro was analyzed by E and F, WB. G, WB analysis of furin protein expression in specific tissues and cell lines. To determine which PC is responsible for pro-OCN cleavage in osteoblasts, we first evaluated the expression of eight subtilisin/kexin-like PCs in this cell. As shown in Fig. 2A, mRNAs encoding furin (Furin) and PACE4 (PCSK6) were highly expressed in this cell, while mRNAs encoding PC7 (PCSK7) and PC5A (PCSK5A) were lowly expressed. The mRNAs of Pcsk1, Pcsk2, Pcsk5b and Pcsk 4 were undetectable or expressed at very low levels in osteoblasts. Notably, the expression of Furin and Pcsk5a is induced during osteoblast differentiation. Based on these results, we then tested the ability of furin, PC5A, PC7, and PACE4 to cleave pro-OCN in vitro. Recombinant GST-pro-OCN proteins produced in bacteria were incubated in the conditioned medium of HEK293 cells and transfected either with empty vectors or with vectors expressing furin, PC5A or PC7 of the soluble extracellular enzyme domain, or with vectors expressing and purifying recombinant soluble PACE4 in insect cells. Furin, PC5A, and PACE4, but not PC7, were detected to cleave GST-pro-OCN, releasing mature OCN (Figure 2B). We also noted that furin was more effective at cleaving pro-OCN than PC5A and PACE4 when incubated for a shorter time,wiped film evaporator, e.g., 15 min (fig. 2B). In addition, further tests showed that furin was able to cleave more than 80% of pro-OCN within 30 min in a dose-related manner (fig. 2, C and D). Since PC5A and PACE4 are known to act primarily on extracellular or plasma membrane-bound substrates, whereas furin functions primarily in the intracellular secretory pathway, these results, together with our observation that pro-OCN lysis in osteoblasts is inhibited by cell-permeable Dec-RVKE-CMV, But not by the non-cell-permeable D6R (fig. 1 E), suggesting that pro-OCN is primarily a substrate for furin. Importantly, the mutation of the RLRR motif to RLAA completely abolished the cleavage ability of furin for pro-OCN in vitro (fig. 2, E and F), thus indicating that furin specifically cleaves pro-OCN at this dibasic group position.
Consistent with the possibility that furin may play an important role in osteoblasts, furin was found to be expressed at comparable or higher levels in osteoblasts than it was in liver cells, thymus, MIN β cells, and no other redundant function for this enzyme is currently known (Figure 2G). The above data indicate that furin specifically cleaves pro-OCN in vitro. Figure 3. Furin but not PC5 is required for pro-OCN processing by osteoblasts A. Furin mRNA expression levels were assessed by PCR after transfection of Ad-GFP (control osteoblasts) or Ad-Cre (Furin-/-osteoblasts) in Cre-mediated Furin inactivated Furinfl/fl osteoblasts. B. WB analysis of OCN secreted from osteoblasts derived from differentiated Furinfl/fl after transfection with Ad-GFP or Ad-Cre. C, Pcsk5fl/fl osteoblasts inactivated by Cre-mediated Pcsk5 were transfected with Ad-GFP or Ad-Cre, and the mRNA expression level of Pcsk5 was assessed by PCR. D. WB analysis of OCN secreted from osteoblasts derived from differentiated Pcsk5fl/fl after transfection with Ad-GFP or Ad-Cre. E. WB analysis of OCN secreted after treatment of differentiated Furinfl/fl calvarial osteoblasts, transfected or not with Ad-GFP or Ad-Cre, with or without 50 μM Dec-RVKR-CMK or 20 μM D6R. We therefore assessed the necessity of furin for osteoblast pro-OCN maturation. To this end, we transfected Furinfl/fl primary osteoblasts with adenovirus-expressing GFP (Ad-GFP) or with adenovirus-expressing Cre-GFP (Ad-Cre) as a control group against furin-deficient osteoblasts (Furin-/-osteoblasts). The expression of Furin was significantly decreased in osteoblasts transfected with Ad-Cre compared with controls (Fig. 3A). It is clear that the inactivation of Furin in primary osteoblasts is sufficient to induce molecular weight changes in secreted OCN (fig. 3 B). In contrast, deletion of Pcsk5, encoding PC5A and PC5B, had no significant effect on the molecular weight of OCN in primary osteoblasts (fig. 3, C and D). In addition, neither Dec-RVKR-CMK nor D6R had a significant effect on the molecular weight of OCN in Furin-deficient osteoblasts (Figure 3 E), suggesting that secreted OCN remained in the propeptide state. These data suggest that Furin is required for pro-OCN processing by osteoblasts by studying the defective expression of Furin. Figure 4. Localization of Furin and Pro-OCN in osteoblasts. A, Typical immunofluorescence picture showing primary osteoblasts transfected with FLAG-pro-OCN or R46A/R48A/R49A FLAG-pro-OCN mutations from p3xFLAG-Myc-CMV-23 vector (FLAG pro-OCNAAA). Furin is green, FLAG (e.g. Pro-OCN) is red, and DNA is blue. B. Quantitative analysis of the red signal region of primary osteoblasts transfected with or without FLAG-pro-OCN or FLAG-pro-OCNAAA. C. WB analysis of cell supernatant and cell extracts from primary osteoblast transfected with FLAG-pro-OCN or FLAG-pro-OCNAAA. To further investigate the interaction between furin and pro-OCN, mouse osteoblasts were transfected with the pro-OCN construct, the propeptide was N-terminally labeled with a 3 X FLAG epitope, and the intracellular localization of pro-OCN and furin was assessed by immunofluorescence.
As shown in fig. 4A, both the FLAG-pro-OCN and non-lytic mutant (FLAG-pro-OCNAAA) signals overlap with the furin signal (middle and bottom panels), suggesting colocalization of pro-OCN and furin. However, FLAG-pro-OCNAAA was more abundant intracellularly than FLAG-pro-OCN (fig. 4, middle panel and bottom panel). Quantitative analysis of FLAG-pro-OCN immunofluorescence signal intensity and WB confirmed that intracellular pro-OCN levels were increased when the furin cleavage site was mutated (fig. 4, B and C). Consistent with this, when furin-deficient osteoblasts were used (fig. 3 B), FLAG-pro-OCNAAA was still secreted (fig. 4 C), suggesting that retention of the propeptide did not prevent secretion of this protein (OCN). Taken together, these in vitro and cell biology assays support furin as the primary, if not the only, PC responsible for processing pro-OCN in osteoblasts. Fig. 5. Pro-OCN treatment and γ-carboxylation processes are independent in osteoblasts A and B, WB Cell supernatants and cell extracts from differentiated mouse osteoblasts were analyzed for endogenous total OCN and γ-carboxylated OCN (Gla OCN) with or without treatment with 50 μM warfarin (A) or 50 μM Dec-RVKR-CMK (B). C. WB analysis of supernatant from osteoblasts transfect with OCN-V5 or E13D/E17D/E20D OCN-V5 mutations (OCNDDD-V5), treated with or without 50. Mu. M Dec-RVKR-CMK. D. WB analysis of OCN immunoprecipitation in sera from control mice (Ggcxfl/fl) and from mice deficient in osteoblastic gamma-hydroxylation (Cgcxosb-/-). Total OCN and r-hydroxylated OCN were assessed by WB. E. LC-MS/MS analysis of cell supernatants from mouse differentiated osteoblasts with or without 50 μM Dec-RVKR-CMK or 50 μM warfarin. The OCN propeptide area was quantitatively compared to the total peptide area. Table 1. LC-MS/MS analysis of mouse differentiated osteoblast supernatants, treated with or without 50 μM Dec-RVKR-CMK or 50 μM warfarin In liver cells, γ-carboxylation, decarboxylation after extraction ,jacketed glass reactor, a widely studied protein post-transcriptional modification, occurs primarily in the endoplasmic reticulum (ER). To test whether the two processes of pro-OCN treatment and γ-carboxylation are interrelated in osteoblasts, differentiated osteoblasts were treated with warfarin, which inhibits γ-carboxylation by blocking the decrease in vitamin K, or Dec-RVKR-CMK, and then pro-OCN cleavage and γ-carboxylation were assessed. As shown in fig. 5A, warfarin effectively blocked the γ-carboxylation of OCN but did not affect its cleavage. In contrast, Dec-RVKR-CMK effectively blocked the cleavage of OCN but had no effect on its γ-carboxylation (fig. 5B). Similarly, mutation of the three OCN glutamate (E) residues to aspartate (D) for γ-carboxylation, which is known to prevent post-transcriptional modification of prothrombin, did not affect pro-OCN to OCN maturation in primary osteoblasts (fig. 5C). We also assessed the effect of γ-carboxylation of pro-OCN on its cleavage process by studying in vivo Ggcx deficient mice (Ggcxfl/fl OCN-Cre mice, Ggcxosb-/-mice) in osteoblasts. Because loss of γ-carboxylation prevented the accumulation of OCN in the bone ECM of these mice, immunoprecipitation and WB were used to assess the processing of OCN in serum.
Although circulating OCN was not γ-carboxylated in Ggcxosb-/-mice, its translocation was similar to that of OCN in control sera, suggesting that OCN was processed to a mature state in vitro, even in the absence of γ-carboxylation (fig. 5D). Finally, we tried to determine the more precise structure of OCN in osteoblasts under normal conditions or in the presence of inhibition of furin-dependent lysis or γ-carboxylation. For this purpose, OCN secreted by differentiated osteoblasts treated with vehicle, Dec-RVKR-CMV or warfarin was analyzed by liquid chromatography-tandem mass spectrometry (LC-MS/MS). The secreted protein is collected, centrifuged, and digested with Arg-C, which cleaves any arginine residues. Thus, if pro-OCN is present in the medium, Arg-C can release an OCN propeptide lacking three residues, such as KPSGPESDKAFMSKQEGNKVVNR. Interestingly, we did not find this peptide in the supernatants of osteoblasts treated with vehicle or warfarin, but only in the supernatants of osteoblasts treated with PC inhibitors, accompanied by some truncated fragments of the N-terminal OCN propeptide (Table 1 and Fig. 5e). These data further suggest that pro-OCN cleavage is not related to γ-carboxylation of OCN and that the pro-OCN treatment process is fully effective in osteoblasts because pro-OCN is completely lost in vehicle-treated osteoblasts. Finally, since the OCN propeptide is not detectable in the supernatant of osteoblasts under normal conditions, it is suggested that the cleaved propeptide (containing the three RRRs that are degraded) is degraded rather than secreted intracellularly. This is consistent with earlier reports that the OCN propeptide was not detectable by radioimmunoassay in human serum or human osteosarcoma supernatants. The above data suggest that pro-OCN treatment and γ-carboxylation are independent processes in osteoblasts. Fig. 6. Impaired pro-OCN processing in Furinosb-/-mice. A. Genomic DNA was extracted from different tissues, and PCR was used to detect the knockout of Furin alleles (PCR), and Flox PCR was used as the loading control. WAT, white adipose tissue; BAT, brown adipose tissue. B. WB analysis of Furin expression in bone marrow-derived osteoblasts in Furinfl/fl or Furinosb-/-mice. C. Quantitative analysis of Furin protein expression in Furinfl/fl mice. D. WB analysis of total OCN and gamma-carboxylated OCN expression in bone extracts in Furinfl/fl or Furinosb-/-mice at 9 months of age. E, WB analysis of OCN immunoprecipitation in serum from Furinfl/fl or Furinosb-/-mice. F, ELSA determination of total and γ-carboxylated OCN after homogenization of 9-month-old Furinfl/fl or Furinosb-/-mouse bone. G, ELSA determination of total and ucOCN in the serum of 9-month-old Furinfl/fl or Furinosb-/-mice. Our cell culture data show that furin is responsible for pro-OCN processing (i.e., lysis) in osteoblasts, and we wondered whether furin is also required for pro-OCN processing in vivo. To address this issue, we generated mice with conditional furin knockout in osteoblasts (Furinfl/flOCN-Cre mice, Furinosb-/-mice), because mice lacking furin in all cells die in the womb before bone formation. For this purpose, Furinfl/fl mice bred with OCN-Cre transgenic mice can generate Cre recombination to control the promoter of human OCN in differentiated osteoblasts.
Furinosb-/-mice were born without significant developmental abnormalities as calculated by the Mendelian ratio. We further confirmed by PCR that furin site recombination occurred exclusively in skeletal tissue (fig. 6a) and that furin mRNA and protein expression levels were significantly reduced in Furinosb-/-mice compared to control osteocyte cultures (fig. 6, B and C). As shown in fig. 6d, inactivation of Furin result in accumulation of pro-OCN and loss of mature OCN in bone. Correspondingly, OCN was only present in Furinosb-/-mouse serum as a precursor (i.e., pro-OCN) (fig. 6E). Importantly, furin deficiency did not affect γ-carboxylation of pro-OCN in bone ECM as demonstrated by in vitro assays with osteoblast cultures (WB and ELISA) (fig. 6, D and F). Interestingly, although total bone OCN content was not reduced in Furinosb-/-mice, circulating total ucOCN and ucOCN levels were approximately halved as assessed by ELISAs, which were used to assess OCN and pro-OCN (fig. 6G). We therefore hypothesized that in Furinosb-/-mice, the decarboxylation and release of pro-OCN during bone resorption is reduced in effectiveness compared to mature OCN. The above data indicate that the processing of pro-OCN is impaired in Furinosb-/-mice. Fig. 7. Pro-OCN is not effectively decarboxylated by osteoclasts. A and B In vitro resorption analysis of inactivated mouse calvaria derived from Furinfl/fl or Furinosb-/-in the presence of osteoblast-like RAW246.7 cells with or without treatment with 10 ng/ml RANKL. (A) ELISA determination of total OCN and ucOCN of cell culture supernatants. (B) Typical TRAP staining map of skull. C. Percentage of TRAP positive area. D. WB analysis of mouse bone extracts from Furinfl/fl or Furinosb-/-after 14 days of incubation in 37 degree PBS at pH 7.5 or 4.3. To test the hypothesis that pro-OCN cannot be effectively decarboxylated by osteoclasts, we used RAW 264.7 osteoclast precursor cells obtained from inactivated calvarial cultures of control or Furinosb-/-mice to induce osteoclast differentiation with RANKL. When osteoclasts of the control bone differentiated, we could detect the release of total OCN and ucOCN, whereas when the osteoclasts of Furinosb-/-mice differentiated, the total OCN released by the osteoclasts was reduced and ucOCN was not detected (Figure 7 A). Importantly, in TRAP staining, osteoclast differentiation did not appear to be impaired in Furinosb-/-mice compared to controls (fig. 7, B and C), suggesting that furin loss in osteoblasts does not affect osteoclast differentiation. Previous studies have found that the acidic pH generated during bone resorption is responsible for the decarboxylation of OCN. Therefore, we compared the decarboxylation of mature OCN and pro-OCN when incubated under pH conditions similar to the bone resorption lacunae environment of osteoclasts (e.g., pH 4.3). When OCN was extracted from the bone of control mice, it was effectively decarboxylated at pH 4.3, whereas pro-OCN extracted from the bone of Furinosb-/-mice was not decarboxylated (Figure 7D). Taken together, these results suggest that in the absence of furin, osteoblasts secrete γ-carboxylated pro-OCN, which, like mature γ-carboxylated OCN, is deposited in the bone ECM. However, the decarboxylation and release of pro-OCN during bone resorption is weak, resulting in a decrease in circulating ucOCN levels.
Fig. 8. Impaired glucose tolerance in Furinosb-/-mice. A, Fasting and postprandial blood glucose levels in 9-month-old Furinfl/fl or Furinosb-/-mice. B. GTT results in 9-month-old Furinfl/fl or Furinosb-/-mice. Mice were fasted for 16 H and then injected with 2 G/kg glucose. C. Fasting and postprandial serum insulin levels in 9-month-old Furinfl/fl or Furinosb-/-mice. D. Pancreatic insulin content of 9-month-old Furinfl/fl or Furinosb-/-mice. E, GTT results of 4-month-old Furinfl/fl or Furinosb-/-mice after 10 weeks on a high-fat, high-sugar diet. Mice were fasted for 6 hours and then injected with 1 G/kg glucose for metabolic assessment in male mice fed either A normal feed diet (A-D) or A high-fat, high-sugar diet (F). To determine the effect of uncleaved pro-OCN on the endocrine function of OCN, we then studied glucose metabolism in Furinosb-/-mice. A significant increase in postprandial blood glucose levels was observed in 9-month-old Furinosb-/-mice compared to littermate controls (fig. 8A). In addition, 9-month-old Furinosb-/-mice showed impaired glucose tolerance when subjected to glucose loading (fig. 8B). Furthermore, consistent with ucOCN normally supporting insulin secretion by beta cells, serum insulin levels were significantly reduced in the postprandial state in Furinosb-/-mice, but fasting serum insulin levels were not affected (fig. 8C). These results are consistent with the fact that in Furinosb-/-mice, the blood glucose level is elevated only in the postprandial blood glucose or in the glucose tolerance test (GTT) (fig. 8, a and B). In addition, in agreement with previous reports, ucOCN had a positive effect on insulin synthesis in beta cells, and insulin content was also reduced in Furinosb-/-mice compared to normal mice (fig. 8D). Finally, 4-month-old Furinosb-/-mice fed with high fat and high glucose for 10 weeks showed reduced glucose tolerance compared to young mice (fig. 8 E). Fig. 9. Decreased energy expenditure and food intake in Furinosb-/-mice. Metabolic parameters, (A) oxygen consumption, (B) carbon dioxide release, (C) cardiac production (energy expenditure), X-axis of activity (D) and Y-axis of activity (E) of Furinfl/fl and Furinosb-/-mice aged A-E, 6 months. F, epididymal fat pad weight after normal feeding in Furinfl/f- and Furinosb-/-mice at 12 months of age. G-J, 6-month-old Furinfl/fl, Furinosb-/-, OCN +/+, OCN-/-mice. (G and I) Day and night food intake. (H and I) Total food intake over a 3-day period Because OCN also increases energy expenditure, we assessed the energy balance of control and Furinosb-/-mice by indirect calorimetry. Diet-fed Furinosb-/-mice showed an increase in oxygen consumption and carbon dioxide production during the dark phase at 3 and 6 months of age. For example, when the mice were more active (fig. 9, a and B), this resulted in a significant reduction in the overall level of energy expenditure (fig. 9 C). We observed that the activity of Furinosb-/-mice was not reduced in the dark and light phases (fig. 9, D and E), suggesting that the reduction in energy expenditure in Furinosb-/-mice was not due to the reduced activity of these animals. Decreased energy expenditure in Furinosb-/-mice at 12 months of age was associated with an increase in epididymal fat pad weight and an increase in the percentage of total body adipose tissue, although they did not show a significant increase in body weight compared to control mice with either normal or high-fat feeding (fig. 9 F).
In summary (fig. 8 + fig. 9A-F), the phenotypic characterization of FurinoSB-/-mice shows that furin, which is expressed by osteoblasts, plays an important role in regulating glucose and energy metabolism. Our findings also support the assertion that furin cleavage is required for the full activation of pro-OCN. Fig. 10. Paired feeding reveals a more severe metabolic phenotype in Furinosb-/-mice. Metabolic phenotype of 4-month-old, pair-fed Furinfl/fl or Furinosb-/-mice. A. Body weight of Furinfl/fl or Furinosb-/-mice before and 2 and 4 weeks after pair-feeding. B. Percentage of total body fat to body weight after 4 weeks of paired feeding. C. Postprandial blood glucose levels after 3 weeks of paired feeding. D. GTT results after 4 weeks of paired feeding. E. ITT results after 4 weeks of paired feeding. Glucose tolerance was normal in 3-month-old and 6-month-old Furinosb-/-mice on normal feeding (see Supplementary Figures, not shown). In addition, insulin sensitivity was assessed using insulin tolerance tests (ITTs) in Furinosb-/-mice of all ages, and the results were normal (see Supplementary Figures, not shown). This is in contrast to Ocn inactivation resulting in glucose tolerance and insulin resistance in 3 month old normally fed mice. Similarly, under normal feeding, Furinosb-/-mice had less fat accumulation than Ocn-/-mice (fig. 9F). Thus, the metabolic phenotype of Furinosb-/-mice is delayed and more moderate than that of Ocn-/-mice. Although there are several possibilities to explain these differences, one possibility is that furin negatively regulates other aspects of energy metabolism and is not related to OCN. Phenomena supporting this hypothesis include an approximately 30% reduction in dark-phase food intake in 6-month-old Furinosb-/-mice (fig. 9 G). Correspondingly, total food intake was reduced by approximately 20% over a 3-day period (fig. 9H). In contrast, Ocn-/-mice of the same age (6 months) and mice of the same genetic background (e.g., C57BL/6J) had normal food intake compared to wild-type mice (fig. 9, I and J). Taken together, these findings suggest that Furinosb-/-mice do not develop as strong insulin resistance as Ocn-/-mice, possibly in part because of reduced caloric intake, a condition previously thought to improve insulin sensitivity in mice. To directly confirm this possibility, we paired 4-month-old control and Furinosb-/-mice for 2 to 4 weeks to maintain the same calorie intake before assessing glucose tolerance and insulin sensitivity. As shown in fig. 10a, the Furinosb-/-mice were heavier than the pair-cultured control mice, which correlated with an increase in the specific gravity of the total adipose tissue content (fig. 10B). Furinosb-/-mice showed high postprandial blood glucose levels and a glucose intolerance phenotype when paired fed (fig. 10, C and D). Finally, under these conditions, Furinosb-/-mice were found to be insulin resistant as assessed by ITT compared to control mice (fig. 10 E). Taken together, these results support that the inactivation of furin in osteoblasts produces two distinct effects of energy metabolism regulation: on the one hand, it reduces energy expenditure, glucose tolerance, and insulin sensitivity; on the other hand, it reduces appetite. The former effect may be OCN-dependent, while the latter effect may be OCN-independent. In this study, it was found that furin, as a PC, was responsible for the conversion of pro-OCN to mature OCN in osteoblasts. The results show that the process of cleavage of pro-OCN to mature OCN can regulate the level of active ucONC in the blood circulation, because pro-OCN deposited in bone ECM is difficult to be effectively released and decarboxylated during bone resorption.
Correspondingly, the osteoblasts of furin gene inactivated mice have metabolic abnormalities similar to those reported in Ocn-/-mice, but there are still some phenotypic differences. This suggests that furin may regulate energy metabolism in two ways, one related to OCN and the other unrelated to OCN. In conclusion, our work has uncovered a previously undescribed regulatory mechanism of bone endocrine; similar to other peptide hormones, it is associated with a specific PC activity. References (public number dialog box reply "JC70" to download the original text): Al Rifai Omar,Chow Jacqueline,Lacombe Julie et al. Proprotein convertase furin regulates osteocalcin and bone endocrine function. [J] .J. Clin. Invest., 2017,50l rotovap, 127(11): 4104-4117. >>>> For more intensive reading of literature, please click on the general catalogue of intensive reading of literature to return to Sohu to see more. Dear readers, Responsible Editor:. toptiontech.com