The Journal of Physical Chemistry B Article /-SPD-DIR d DIR /-SPD-DI E2 86420 Time(ns) Time(ns) b 802086420 e D2R SPD- D2R D2R 3 E2一 8。z Time(ns) Time(ns) /-SPD-D3R SPD-D3R e2E 0864 f38号9z Figure 3. Distance fluctuations in TM3-TM6 and n-o during the MD simulations. for both DIR and D2R were found: cluster 1 is surrounded by DiR shortened gradually from ca. 20 A to ca. 12 A at the first TM3, TM6, and TM7; cluster 2 is in TM2 and TM4; and 30 ns and remained stable for the rest 20 ns in its unliganded cluster 3 is in TM3 and TM5. However, differences are state simulation. It indicated that a transition from active state observed at the cytoplasmic face of DiR, with an outward to inactive state is through a continuum of intermediate movement of TMS and TM6 and an inward movement of TM3 conformations; this kind of behavior is in consistency with the and TM7 in the present model. Such movement of trans- postulate proposed by Brian Kobilka et al. that GPCRs are membrane segments is a typical feature of activated B,AR."As more like molecular rheostats that are able to sample the present DIR model was constructed based on the active- continuum of conformations. Upon binding with 1-SPD, the state B2AR, it is presumably in an active state and is more eparation in TM3-TM6 was kept at ca 15 A(above 14 A) uitable for investigating the agonist effect of 1-SPD during the entire simulation(Figure 3a), implying that 1-SPD Cytoplasmic Helical Movement in Agonistic DIR and has the capacity of stabilizing DiR in its active state. In contrast ntagonistic D2R/D3R. The three molecula to the conformational changes induced by 1-SPD in DIR, 1- assembled by embedding DIR, D2R, and D3R in a rectangular SPD-induced conformational changes at the cytoplasmic faces ox comprising a POPC lipid bilayer surrounded by explicit of both D2R and D3R were not remarkable. The separation in water were simulated for 50 ns each. The MD simulations TM3-TM6 in unliganded D2R is relatively stable and is kept at showed that, for all systems, temperature, mass density, and ca. 12 a during the entire simulation. While in I-SPD bound volume were relatively stable after 2 ns. After that, the D2R, the separation in TM3-TM6 fluctuated slightly at around fluctuation scale became much smaller for the rms deviations of 11 a during the first 8 ns, and then it jumped to ca. 12 A. As for he backbone atoms of all simulation systems( Figure 2), D3R, the overall landscape of distance in TM3-TM6 is indicating that the molecular systems behaved well thereafter. stabilized at 10-ll A in both liganded and unliganded D3Rs The most prominent differences in conformation between ( Figure 3b, c). active and inactive GPCRs are resided at the cytoplasmic face, lonic Lock Formation/Breakage in Agonistic D1R and with the outward movement of TM3 and TM6, as exemplified Antagonistic D2R/D3R. It is also established that the basal by the active crystal structures of B2AR-T4L-Nb80 and B2 AR- conformation of GPCRs are maintained by interhelix loops and lexes" and by the active cry noncovalent interactions between side chains of a set of crucial receptor.A systematic examination of currently available residues in TM regions. One of the best characterized GPCR crystal structures revealed that the movement of TM6 examples of noncovalent interaction is a salt bridge formed apart from TM3 is the key event in the process of GPCR between R3. 50, part of the highly conserved(D/E)RY motif in activation(Table S2), as such conformational change opens the transmembrane helix 3, and the conserved E6.30 in trans- docking pocket for G protein. The distances between TM3 and membrane helix 6. This salt bridge network was nicknamed TM6 in the cytoplasmic ends( distance between centroid of the "ionic lock". Though it was originally proposed to account for Ca atoms of four residues located at the tip of each TM)were the activation of B,AR, accumulated studies have demonstrated onitored for both liganded and unliganded DiR, D2R, and that this interaction was crucial in stabilizing the inactive state D3R as shown in Figure 3. The separation in TM3-TM6 for of the rhodopsin family GPCRs. In order to get a more 8124 dx. dolora/o.021/p30492351 Phys. Chem. B2012116,8121-813for both D1R and D2R were found: cluster 1 is surrounded by TM3, TM6, and TM7; cluster 2 is in TM2 and TM4; and cluster 3 is in TM3 and TM5.20 However, differences are observed at the cytoplasmic face of D1R, with an outward movement of TM5 and TM6 and an inward movement of TM3 and TM7 in the present model. Such movement of trans￾membrane segments is a typical feature of activated β2AR.4 As the present D1R model was constructed based on the active￾state β2AR, it is presumably in an active state and is more suitable for investigating the agonist effect of l-SPD. Cytoplasmic Helical Movement in Agonistic D1R and Antagonistic D2R/D3R. The three molecular systems assembled by embedding D1R, D2R, and D3R in a rectangular box comprising a POPC lipid bilayer surrounded by explicit water were simulated for 50 ns each. The MD simulations showed that, for all systems, temperature, mass density, and volume were relatively stable after 2 ns. After that, the fluctuation scale became much smaller for the rms deviations of the backbone atoms of all simulation systems (Figure 2), indicating that the molecular systems behaved well thereafter. The most prominent differences in conformation between active and inactive GPCRs are resided at the cytoplasmic face, with the outward movement of TM3 and TM6, as exemplified by the active crystal structures of β2AR-T4L-Nb80 and β2AR￾Gs complexes4 and by the active crystal structures of opsin receptor.42 A systematic examination of currently available GPCR crystal structures revealed that the movement of TM6 apart from TM3 is the key event in the process of GPCR activation (Table S2), as such conformational change opens the docking pocket for G protein. The distances between TM3 and TM6 in the cytoplasmic ends (distance between centroid of the Cα atoms of four residues located at the tip of each TM) were monitored for both liganded and unliganded D1R, D2R, and D3R as shown in Figure 3. The separation in TM3-TM6 for D1R shortened gradually from ca. 20 Å to ca. 12 Å at the first 30 ns and remained stable for the rest 20 ns in its unliganded state simulation. It indicated that a transition from active state to inactive state is through a continuum of intermediate conformations; this kind of behavior is in consistency with the postulate proposed by Brian Kobilka et al. that GPCRs are more like molecular rheostats that are able to sample a continuum of conformations.25 Upon binding with l-SPD, the separation in TM3−TM6 was kept at ca. 15 Å (above 14 Å) during the entire simulation (Figure 3a), implying that l-SPD has the capacity of stabilizing D1R in its active state. In contrast to the conformational changes induced by l-SPD in D1R, l￾SPD-induced conformational changes at the cytoplasmic faces of both D2R and D3R were not remarkable. The separation in TM3−TM6 in unliganded D2R is relatively stable and is kept at ca. 12 Å during the entire simulation. While in l-SPD bound D2R, the separation in TM3−TM6 fluctuated slightly at around 11 Å during the first 8 ns, and then it jumped to ca. 12 Å. As for D3R, the overall landscape of distance in TM3−TM6 is stabilized at 10−11 Å in both liganded and unliganded D3Rs (Figure 3b,c). Ionic Lock Formation/Breakage in Agonistic D1R and Antagonistic D2R/D3R. It is also established that the basal conformation of GPCRs are maintained by interhelix loops and noncovalent interactions between side chains of a set of crucial residues in TM regions.24 One of the best characterized examples of noncovalent interaction is a salt bridge formed between R3.50, part of the highly conserved (D/E) RY motif in transmembrane helix 3, and the conserved E6.30 in trans￾membrane helix 6. This salt bridge network was nicknamed “ionic lock”. 43 Though it was originally proposed to account for the activation of β2AR, accumulated studies have demonstrated that this interaction was crucial in stabilizing the inactive state of the rhodopsin family GPCRs.43,44 In order to get a more Figure 3. Distance fluctuations in TM3−TM6 and N−O during the MD simulations. The Journal of Physical Chemistry B Article 8124 dx.doi.org/10.1021/jp3049235 | J. Phys. Chem. B 2012, 116, 8121−8130