Advanced glycation end products stimulate osteoblast apoptosis via the MAP kinase and cytosolic apoptotic pathways

Abstract

We have previously shown that diabetes significantly enhances apoptosis of osteoblastic cells in vivo and that the enhanced apoptosis contributes to diabetes impaired new bone formation. A potential mechanism is enhanced apoptosis stimulated by advanced glycation end products (AGEs). To investigate this further, an advanced glycation product, carboxymethyl lysine modified collagen (CML-collagen), was injected in vivo and stimulated a 5-fold increase in calvarial periosteal cell apoptosis compared to unmodified collagen. It also induced apoptosis in primary cultures of human or neonatal rat osteoblastic cells or MC3T3-E1 cells in vitro. Moreover, the apoptotic effect was largely mediated through RAGE receptor. CML-collagen increased p38 and JNK activity 3.2- and 4.4-fold, respectively. Inhibition of p38 and JNK reduced CML-collagen stimulated apoptosis by 45% and 59% and by 90% when used together (P<0.05). The predominant apoptotic pathway induced by CML-collagen involved caspase-8 activation of caspase-3 and was independent of NF-kappaB activation. When osteoblastic cells were exposed to a long-term low dose incubation with CML-collagen, there was a higher degree of apoptosis compared to short-term incubation. In more differentiated osteoblastic cultures, apoptosis was enhanced even further. These results indicate that advanced glycation end products, which accumulate in diabetic and aged individuals, may promote apoptosis of osteoblastic cells and contribute to deficient bone formation.

Figure 1. CML-collagen induces apoptosis in bone-lining cells in vivo and osteoblastic cells in vitro

A. CML-collagen (100 μg) or control unmodified collagen (100 μg) was injected into the scalp as described in Materials and Methods and mice were euthanized 24 hours later. Apoptotic cells were identified in paraffin sections by the TUNEL assay. The number of TUNEL-positive bone-lining periosteal cells was counted at 1000x magnification between the occipital and coronal sutures. Each value represents the mean of 6 specimens ± SEM. B. CML-collagen (200 μg/ml) or an equal amount of unmodified control collagen were incubated with human, MC3T3-E1 or neonatal rat calvarial osteoblastic cells for 24 hours and compared to controls which were incubated in assay medium alone. Each value represents the mean of 4 replicates ± SEM and each experiment was performed three times with similar results.

Figure 2. CML-collagen stimulates a dose and time dependent increase in apoptosis of osteoblastic cells

A: Human osteoblastic cells were incubated with different concentrations of CML- collagen or unmodified collagen for 24 hours. The extent of apoptosis was determined by ELISA. B. Human osteoblasts were incubated with AGE-collagen or unmodified collagen (200μg/ml) as indicated. Apoptosis was determined by ELISA. Each value represents the mean of 4 replicates ± SEM and each experiment was performed three times with similar results. C. Primary human osteoblasts were incubated with CML-collagen (50 or 200 μg/ml) or unmodified collagen for 24 hours or 1 week. As an additional control cells were incubated with medium alone. At the end of each incubation period apoptosis was determined by ELISA. The ELISA values were normalized by cell number. Each value represents the mean of 4 replicates ± SEM. The experiment was performed three times with similar results.

Figure 3. CML-collagen inducedCollagen stimulates apoptosis is mediated through RAGE

Primary cultures of human adult osteoblastic cells were incubated with CML-collagen (200μg/ml) in presence or absence of antibody specific for the extra-cellular domain of RAGE (10 μg/ul) or non-immune serum (10 μg/ul) for 24 hours. In some cases cells were incubated with antiserum or non-immune serum alone. The extent of apoptosis was determined by ELISA. Each value represents the mean of 4 replicates ± SEM. The experiment was performed three times with similar results.

Figure 4. CML-Collagen stimulates apoptosis through p38 and JNK MAP kinase activity

A Primary human osteoblasts were incubated for 0–30 mins with 200 μg/ml of CML-collagen or unmodified collagen. Cytoplasmic extracts were tested for activated (A) p38 and (B) JNK activity using a phosphosite-specific ELISA. (C) Cells were pre-incubated for 2 hours with a specific inhibitors for p38, SB203580 or JNK inhibitory peptide, H-GRKKRRQRRRPPRPKRPTTLNLFPQVPRSQDT-NH2. The inhibitors were also present in assay media. At the end of the incubation period the number of cells was counted and apoptosis was determined by ELISA. The ELISA values were normalized by cell number. For Panels A and B each value represents the mean of 3 replicates ± SEM and experiments were performed three times with similar results. For Panel C each value represents the mean of 4 replicates ± SEM and the experiment was performed four times with similar results.

Figure 5. CML-collagen stimulates caspase activity in osteoblasts

A. Human osteoblasts were incubated with CML- collagen with or without different caspase inhibitors for 24 hours. A. Caspase-3, -8 and -9 activity was measured by fluorimetric assay in lysates from cell cultures obtained after 24 hours stimulation with CML-collagen (200 μg/ml). B. Caspase-3 activity was measured using a fluorimetric assay in lysates from cell cultures obtained after 24 hours stimulation with CML-collagen (200 μg/ml) with or without caspase-8 or -9 inhibitors. (C) Cells were incubated with control collagen (200μg/ml) or CML- collagen (200 μg/ml) with or without inhibitors to caspase-3, -8 or -9. Apoptosis was measured 24 hours later by ELISA. For each panel the value represents the mean of 4 replicates ± SEM and experiments were performed three times with similar results.

Figure 6. Apoptotic effect of CML-Collagen increased in differentiated osteoblasts

MC3T3-E1 cells (A) or primary neonatal calvarial osteoblastic cells (B) were cultured in the presence or absence of ascorbic acid (50 μg/ml) for one or two weeks. Cells were then incubated with CML-collagen or unmodified collagen (200 μg/ml) for 24h. Control cells were incubated in assay medium alone. Apoptosis was determined by ELISA and the values were normalized by cell number. Each value represents the mean of 4 replicates ± SEM. The experiment was performed three times with similar results.

Figure 7. AGEs stimulate osteoblast apoptosis through an NF-κB-independent mechanism

A. Primary human osteoblasts were incubated for 24 hours with 200 μg/ml of CML-collagen or unmodified collagen in presence or absence of NF-κB inhibitor SN50 (100μg/ml). Nuclear extracts were isolated and incubated with biotinylated NF-κB probe or probe plus excess unlabelled probe. EMSA was carried out as described in Materials and Methods. The arrow points to the NFκB band in lane 2. The experiment was performed three times with similar results. B. Primary human osteoblasts were incubated with CML-collagen (200 μg/ml) with or without NF-κB inhibitor SN50 (100μg/ml) or unmodified collagen (200 μg/ml). Apoptosis was measured 24 hrs later by ELISA. Each value represents the mean of 5 replicates ± SEM. The experiment was performed three times with similar results.

Figure 8. CML-Collagen compared to TNF-induced apoptosis in primary human osteoblasts

A. Cells were incubated with different concentration of TNF or CML-collagen for 24h. The extent of apoptosis was determined by ELISA. B. Cells were incubated with a suboptimal dose of CML-collagen (100 μg/ml) and increasing concentrations of TNF-α (10–20ng/ml). The extent of apoptosis was determined by ELISA. The experiments were performed two (B) or three times (A) with similar results. Each value represents the mean of 3 or 4 replicates ± SEM.