Skip to main content
. Author manuscript; available in PMC: 2014 Jun 1.
Published in final edited form as: Diabetologia. 2013 Mar 19;56(6):1444–1453. doi: 10.1007/s00125-013-2884-2

Fig. 2.

Fig. 2

LPA was present in BVi and accounted for a major portion of the regression activity. (a) The amount of LPA in BVi was determined using a radioenzymatic LPA assay, which is described in the Methods. In this assay LPA is converted into radioactively labelled PA [30]. The reaction products were resolved by thin-layer chromatography. Compared with LPA standards BVi contained 0.8 ±0.13 µmol/l LPA. LPA present in BVi was cleared by treatment with isolated LPP. Mock treatment (EV; an immunoprecipitate prepared from cells that did not overproduce LPP1) had no effect on the concentration of vitreal LPA. (b,c) Retinal explants were isolated and cultured as described in Fig. 1. LPA (1 µmol/l) and BVi induced comparable regression. Clearing LPA from vitreous eliminated regression activity. Black and white arrows indicate stable and regressing vessels, respectively. Four independent experiments showed similar results. Data shown are means±SEM (n=4) *p<0.05 compared with buffer-treated (BT) explants; †p<0.05 compared with buffer-treated explants, no significant difference compared with LPA treatment; §p<0.05 compared with LPA, BVi and empty-vector (EV) treated BVi. (d,e) Primary BRECs were subjected to a collagen sandwich tube formation assay in low serum (0.5% [vol./vol.]); this assay is described in the Methods. Tube formation was photographed 12 h after the addition of VEGF-A (2.5 ng/ml), the indicated treatments were added and tubes were re-photographed after 4 h. Tubes regressed in response to either BVi or mock-treated BVi (white arrows in images), whereas they remained stable with LPP-treated BVi (black arrows). Magnification is ×100. Tubes were digitally outlined to facilitate their visualisation. The results when LPA (1 µmol/l) was added instead of BVi are also shown in (e). Data shown are means±SEM (n=4). Normalised tube length is the ratio of final and initial tube lengths. *p<0.05 compared with buffer-treated tubes; †p<0.05 compared with tubes treated with LPA or mock treated. ‡p<0.05 compared with buffer-treated (BT) tubes; §p<0.05 compared with BVi- and mock-treated tubes