A. Relative expression of BMP and BMP inhibitor genes by cultured gingival fibroblasts Expression normalized to averaged Eif4a2 and Atp5b expression. Data shown as mean ± SD (n = 5 animals). Assays carried out in triplicate. *P < 0.05—Grem1 expression significantly higher than all others tested. Significance tested by one‐way ANOVA with Bonferroni post‐tests. B, In situ hybridization of BMP inhibitor mRNA expression in molar periodontal tissue of 8 wk post‐natal WT mice. A, H&E staining, B, Grem1 expression restricted to the inner half of the gingival lamina propria closest to the alveolar bone and periodontal ligament. C‐E, Expression of Grem2 (D), Nbl1 (E) and Noggin (F) undetectable. OE, oral epithelium; SE, sulcular epithelium; S, sulcus; D, dentine; PL, periodontal ligament; AB, alveolar bone; *, processing artefact. Scale bar in F = 100 μm for A‐E (n = 2 mice). Results confirmed in two independent experiments. C, Western blotting of gingival fibroblast conditioned media confirming the presence of Gremlin1 and removal with immunoprecipitation beads. Lanes: No IP—Concentrated gingival fibroblast conditioned media containing proteins >5 kDa, IP‐Grem1—Concentrated GFCM treated with IP beads with Gremlin1 antibody adsorbed, IP‐control—Concentrated GFCM treated with IP beads with isotype control antibody adsorbed, IP‐beads—GFCM treated with IP beads without antibody adsorbed. 10 μg total protein/well loaded. D, Inhibition of ROS cell ALP activity of Gremlin1‐depleted gingival fibroblast conditioned medium is rescued with addition of Gremlin1. GFCM that underwent IP were analysed in the ROS cell ALP activity assay. SM—standard media. GFCM—concentrated GFCM. IP‐Grem1—Concentrated GFCM treated with IP beads with Gremlin1 antibody adsorbed. IP‐beads—GFCM treated with IP beads without antibody adsorbed. IP‐control—Concentrated GFCM treated with IP beads with isotype control antibody adsorbed, Grem1—100 ng/ml rmGremlin1 added immediately prior to ROS assay. No Grem1—no Gremlin1 added. Results shown as mean ± SD. *—significantly different from standard media with no Gremlin1, †—significantly different from matched sample without Gremlin1 added. P < 0.05. Each treatment tested using five replicates from 1 cell line. Representative results from experiments testing three independent primary cell lines. Significance tested using one‐way ANOVA with Bonferroni post‐tests