TIMP-2 induces cAMP production in hMVECs, independent of MMP inhibitory activity. (A) Quiescent hMVECs were treated with TIMP-2 (100nM) for the indicated time points. The cAMP levels were measured as described in “cAMP determination.” Values represent the mean ± SD of 3 independent experiments. (B) Cells were pretreated with SQ22536 (100μM) for 20 minutes, followed by Ala + TIMP-2 for 1 minute, or treated with isoproterenol (10μM) and forskolin (20μM) for 1 minute. The results (mean ± SD) are expressed as the fold increase of cAMP levels in untreated cells. **P < .01, compared with Ala + TIMP-2 treatment alone. (C) Cells were pretreated with function-blocking anti-integrin antibodies (10 μg/mL) for 20 minutes, followed by Ala + TIMP-2 (100nM) for 1 minute. (D) Both GD25 and GD25-β1A MEFs were treated with Ala + TIMP-2 for the indicated time points. (E) A549 cells transfected with control or integrin α3 shRNA were treated with Ala + TIMP-2 as in panel D. The results from 3 independent experiments (mean ± SD) are presented as the percentage of maximally induced cAMP levels by Ala + TIMP-2 in the absence of integrin blocking antibodies or femtomoles of cAMP per microgram of protein in the samples to allow comparison between experiments. (F) Quiescent hMVECs transfected with control or integrin α3 siRNA were treated with Ala + TIMP-2 as in panel E. The results (mean ± SD) are expressed as the fold increase of cAMP levels in untreated cells. **P < .01, compared with Ala + TIMP-2 treatment alone. (G) Cells were pretreated with orthovanadate (2μM) for 20 minutes, followed by Ala + TIMP-2 for 1 minute. *P < .05, compared with Ala + TIMP-2 treatment alone. (H) Vector control and dn Shp-1 mutant-transduced hMVECs were treated with Ala + TIMP-2 for 1 minute. The results (mean ± SD) are presented as the percentage of maximal TIMP-2–induced cAMP levels in control vector cells (observed 1 minute after treatment). *P < .05, compared with Ala + TIMP-2 treatment alone in vector control cells.