nt12dreuar3esd

300 | Nature | Vol 579 | 12 March 2020

Article

and G protein^14. In M2R–Go, opening of the cytoplasmic portion of
TM6 accommodates the α5-helix of Go. Similarly, displacement of
TM6 accommodates the finger loop of βarr1, which occupies a simi-
lar position to the α5-helix of the Ras domain of Go, albeit at a slightly
lower depth and involving a smaller interaction surface (Fig. 2d). This
difference probably accounts for the lower affinity and inherent con-
formational variability of the βarr1–7TM interaction compared with
the M2R–Go interaction.
In the cryo-EM structure of M2Rpp–βarr1 and the crystal structure
of rhodopsin–visual arrestin (denoted Rho–Arr1); PDB: 5W0P^8 , the
arrestins show notable differences despite their similar orientations
relative to the 7TM bundles (Fig. 2e). βarr1 is tilted approximately 7°
further towards the membrane (Extended Data Fig. 4a)—this is probably
due to the interaction of the C domain with the nanodisc phospholip-
ids, although crystal packing interactions in the structure of Rho–Arr1
might also contribute to these differences. In addition, the finger loops
display differences in both positioning and structure. Whereas the
finger loop of visual arrestin is modelled with an α-helical segment
(residues E71–G77), the finger loop of βarr1 adopts an extended loop
configuration. The other arrestin interdomain loops adopt similar
conformations in these two GPCR–arrestin structures (Extended Data
Fig. 4b).

Interaction of βarr1 with M2R

βarr1 interacts with M2Rpp through interfaces with the phosphoryl-
ated C terminus, the 7TM core, and ICL2 of the receptor (Fig.  3 ). The
phosphopeptide–βarr1 interaction is essentially identical to that in
our crystal structure of V2Rpp–βarr1–Fab30^7 , and six phosphoryl-
ated residues are well resolved in the 3.6 Å map. The peptide binds to
a positively charged crevice on the N domain by displacing the C ter-
minus of βarr1, which destabilizes the arrestin polar core and enables
the gate loop to flip towards the N domain. A critical phosphorylated
residue on the C terminus of the receptor, T491 (T360 in V2Rpp), estab-
lishes interactions with R25 (N domain) and K294 (gate loop) of βarr1,

stabilizing an activated conformation characterized by interdomain

twisting^21 (Fig. 3a). The fused V2Rpp is 36 residues long and provides

ample flexibility so as not to limit the orientation of arrestin; this is also

evident by the lack of observed density for the 23 residues that do not

engage arrestin. Similarly, although we cannot rule out the possibil-

ity that natively phosphorylated ICL3 residues influence the relative

orientation of βarr1, this seems unlikely given the length of this loop

(152 residues) and its lack of order in our structure.

The second interface between the finger loop of βarr1 and the 7TM

bundle of the receptor involves both hydrophobic and electrostatic

interactions (Fig. 3b). The C-terminal finger loop region—including

residues L68, V70, L71 and F75—is hydrophobically packed against the

side of the receptor pocket formed by TM3, TM5, TM6 and ICL2. On the

opposite side, the map of the N-terminal part of the finger loop displays

partial side-chain densities of ionic residues R65, E66 and D67, reveal-

ing both their positions and their relative mobilities. A potential salt

bridge between R65 (finger loop) and D135 (middle loop) of βarr1 could

stabilize the conformation of the finger loop, whereas E66—although

resolved only through Cβ—points towards R62 of the finger loop and

K138 of the middle loop (Extended Data Fig. 5a). At the tip of the fin-

ger loop, D69 seems to be positioned to form hydrogen bonds or salt

bridges with several receptor residues, including N582.39 and R1213.50

of the highly conserved DRY motif (superscripts denote Ballesteros–

Weinstein numbering for GPCRs^22 ). These interactions formed consist-

ently across multiple, independent molecular dynamics simulations,

suggesting that this interaction network stabilizes the M2Rpp–βarr1

complex (Extended Data Fig. 5b, c). Considering that R3.50–E6.30 salt

bridges stabilize the inactive receptor state of M2R and other Class A

GPCRs, βarr1 may stabilize an active receptor partly by engaging R3.50.

The mutant βarr1(D69A) displays significantly reduced coupling to

M2Rpp, which supports this hypothesis (Extended Data Fig. 5d).

The finger-loop fold, the interdomain twist, and the binding mode

through the C terminus of V2Rpp that we observe here for βarr1

are very similar to those in the crystal structure of the active βarr1–

V2Rpp–Fab30^7 (PDB: 4JQI, Extended Data Fig. 4c). However, the crystal

a

b TM3

TM1

TM3

TM2

TM4

H8

Finger

loop

ICL2

TM7

V2Rpp

C terminus

TM4 TM3

TM5

F75

L71 D69

L68 D67

R65

V70

T74

L73

R1213.50

H295

K294

R25 L293

R7

K10

N582.39

N592.40

TM6

TM5

pS488

C489 R165

pT490

pT491

pS494 pS493

pS495

c

N domain

C domain

C-edge

Lipid bilayer

L12L129 9 34.5134.51

Fig. 3 | Regions of interaction between M 2 R and βarr1. The M2Rpp–βarr1
complex, with dashed boxes indicating the main interaction sites between
βarr1 (teal) and receptor (orange). a, Enlarged view of the finger loop of βarr1
inserting into the TM bundle of M2R. The model includes side-chain positions
from PHENIX refinement with OPLS3e electrostatics. b, Enlarged view of the N
domain of βarr1 bound to the C terminus of phosphorylated M2Rpp.

c, Expanded view of the interaction between ICL2 of M2R (ribbon) and a

hydrophobic cleft in βarr1 (rendered as electrostatic surface; red, blue and

white graduations indicate negative, positive and neutral surface potential,

respectively). The mesh in all panels depicts the 3.6 Å cryo-EM map contoured

at σ = 3.0 (where σ is the root mean square of the electron density in the unit

cell) with a masked 2.0 Å zone around the atoms depicted.