Journal: The Journal of Biological Chemistry

Article Title: Structural and Functional Characterization of the Kindlin-1 Pleckstrin Homology Domain

doi: 10.1074/jbc.M112.422089

Figure Lengend Snippet: MD simulations of the kindlin-1 PH domain. A , left , shown is separation between the charged groups of Arg 380 and Glu 416 , which together form a salt bridge in the crystal structure that appears to partially occlude the putative lipid binding cavity. Running averages taken every 100 ps show the transient breaking of the salt bridge over the course of both the GROMOS96 43a1 simulation ( green ) and the OPLS-AA/L simulation ( blue ). The dashed line at 4 Å indicates the separation at which the salt bridge is considered to be formed/broken following the published definition . Center , shown is a simulation snapshot of the occluded conformation used for docking, with the Arg 380 –Glu 416 salt bridge intact. Right , shown is a simulation snapshot of the open conformation used for docking, with the Arg 380 –Glu 416 salt bridge now broken and replaced by an alternative Arg 380 -Glu 450 salt bridge, leaving the binding pocket exposed. B , shown are conformational dynamics of the kindlin-1 PH domain during the MD simulations. Left , Cα r.m.s.d. was calculated as a function of time over the MD simulations using the GROMOS96 43a1 forcefield ( green ) and the OPLS-AA/L forcefield ( blue ). Right , shown is Cα root mean square fluctuation calculated for each residue over the course of the MD simulations. C , molecular docking of Ins(1,3,4,5)P 4 (equivalent to the headgroup of PtdIns(3,4,5)P 3 ) using AutoDock. Left , centers of mass of the ligand clusters when docking to the occluded conformation are shown. Based on a cluster tolerance of 1.5 Å r.m.s.d. from 100 docking runs, there were a total of 48 distinct conformational clusters, of which 22 were multimember clusters. Right , shown are centers of mass of the ligand clusters when docking to the open conformation. In this case, from 100 docking runs there were now 41 distinct conformational clusters, of which only 14 had multiple members. D , shown are estimated free energies of binding and number of clusters using AutoDock. Left , shown is mean estimated free energy of binding for all 100 docking runs using AutoDock for each of the six PtdIns species. Results are shown for docking to the wild type ( closed ) PH domain ( black squares ), wild type PH domain ( blue triangles ), and E416A mutant PH domain ( red diamonds ). Error bars are ±1 S.D. Lines are guides for the eye. A more negative value is indicative of more favorable binding. Center , mean estimated free energy of binding for the members of the largest cluster after docking is based on a cluster tolerance of 1.5 Å r.m.s.d. Similar trends to those in the left hand panel are observed. Right , number of clusters after 100 docking runs for each case is shown. A lower number of clusters suggests a greater degree of consensus as to the location of the binding site.

Article Snippet: Glo-PIPs were purchased from Echelon Biosciences as follows: PtdIns-(3)P (C-03B6a), PtdIns(5)P (C-05B6a), PtdIns(3,5)P 2 (C-35B6a), and PtdIns(3,4,5)P 3 (C-39B6a).

Techniques: Binding Assay, Mutagenesis