G protein coupled receptors (GPCRs) are a large class of seven transmembrane domain receptors that transduce signals from outside the cells when bound to an

appropriate ligand. The GPCRs have a myriad of functions, being involved in sensory perceptions, such as odor and vision, responding to pheromones, hormones and

neurotransmitters, where the ligands greatly vary in nature and size. The GPCRs can affect behavior and mood, the immune system, the sympathetic and parasympathetic nervous system, cell density sensing and there may be additional physiological activities that involve GPCRs in their pathway. The GPCRs are associated with a number of diseases and have been an active target of pharmaceutical companies .

As mentioned above, GPCRs are involved in a wide variety of physiological processes. Some examples of their physiological roles include:

1. The visual sense: the opsins use a

photoisomerization reaction to translate

electromagnetic radiation into cellular signals. Rhodopsin, for example, uses the conversion of 11- cis-retinal to all-trans-retinal for this purpose 2. The sense of smell: receptors of the olfactory epithelium bind odorants (olfactory receptors) and pheromones (vomeronasal receptors)

3. Behavioral and mood regulation: receptors in the

mammalian brain bind several different

neurotransmitters, including serotonin, dopamine, Gamma aminobutyric acid, and glutamate

4. Regulation of immune system activity and

inflammation: chemokine receptors bind ligands that mediate intercellular communication between cells of the immune system; receptors such as histamine receptors bind inflammatory mediators and engage target cell types in the inflammatory response

5. Autonomic nervous system transmission: both the

sympathetic and parasympathetic nervous systems are regulated by GPCR pathways, responsible for control of many automatic functions of the body such as blood pressure, heart rate, and digestive processes

6. Cell density sensing: A novel GPCR role in

regulating cell density sensing.

7. Homeostasis modulation (e.g., water balance)

8. Involved in growth and metastasis of some types of tumors . GPCRs are activated by an external signal resulting in a conformational change. It appears that once the receptor becomes bound it activates the G protein, which G protein is bound to ATP. The G protein is a trimer, which upon activation converts GTP (guanosine triphosphate) to GDP (guanosine diphosphate) . Active GPCRs are phosphorylated by protein-coupled receptor kinases. In many cases upon phosphorylation, the phosphorylated receptor becomes linked to arrestin. The binding to arrestin may result in translocation of the GPCR or other outcome.

In response to a stimulus, GPCRs activate heterotrimeric G proteins. In order to turn off this response, or adapt to a persistent stimulus, active receptors need to be desensitized. The first step is phosphorylation by a class of serine/threonin kinases called G protein coupled receptor kinases (GRKs) . GRK phosphorylation specifically prepares the activated receptor for arrestin binding. Arrestin binding to the receptor blocks further G

protein-mediated signaling and targets receptors for internalization, and redirects signaling to alternative G protein-independent pathways, such as β-arrestin

signaling. In addition to GPCRs, arrestins bind to other classes of cell surface receptors and a variety of other signaling proteins.

Arrestins are elongated molecules, in which several intra-molecular interactions hold the relative

orientation of the two domains. In unstimulated cell arrestins are localized in the cytoplasm in this basal "inactive" conformation. Active phosphorylated GPCRs recruit arrestin to the plasma membrane. Receptor binding induces a global conformational change that involves the movement of the two arrestin domains and the release of its C-terminal tail that contains clathrin and AP2 binding sites. Increased accessibility of these sites in receptor-bound arrestin targets the arrestin-receptor complex to the coated pit. Arrestins also bind

microtubules (part of the cellular "skeleton") , where they assume yet another conformation, different from both free and receptor-bound form. Microtubule-bound arrestins recruit certain proteins to the cytoskeleton, which affects their activity and/or redirects it to

microtubule-associated proteins. Arrestins shuttle between the cell nucleus and the cytoplasm. Their nuclear functions are currently under intense investigation and it was shown that all four mammalian arrestin subtypes remove some of their partners, such as protein kinase JNK3 or the ubiquitin ligase Mdm2, from the nucleus. Arrestins also modify gene expression by enhancing transcription of certain genes.

Mammals express four arrestin subtypes and each arrestin subtype is known by multiple aliases. The systematic arrestin name (1-4) plus the most widely used aliases for each arrestin subtype are listed in bold below:

• Arrestin-1 was originally identified as the S-antigen (SAG) causing uveitis (autoimmune eye disease) , then independently described as a 48 kDa protein that binds light-activated phosphorylated rhodopsin before it became clear that both are one and the same. It was later renamed visual arrestin, but when another cone- specific visual subtype was cloned the term rod

arrestin was coined. This also turned out to be a misnomer: arrestin-1 expresses at comparable very high levels in both rod and cone photoreceptor cells.

• Arrestin-2 was the first non-visual arrestin cloned. It was first named β-arrestin simply because between two GPCRs available in purified form at the time, rhodopsin and β2 ^~ 3άΓθηθ^ίο receptor, it showed preference for the latter .

• Arrestin 3: The second non-visual arrestin cloned was first termed (retroactively changing the name of β-arrestin into β-arrestin-l ) , even though by that time it was clear that non-visual arrestins interact with hundreds of different GPCRs, not just with β2 ^~ 3άΓθηθ^ίο receptor. Systematic names, arrestin- 2 and arrestin-3, respectively, were proposed soon after that.

• Arrestin-4 was cloned by two groups and termed cone

arrestin, after photoreceptor type that expresses it, and X-arrestin, after the chromosome where its gene resides. In the HUGO database its gene is called arrestin-3. Arrestins block GPCR coupling to G proteins via two mechanisms. First, arrestin binding to the cytoplasmic tip of the receptor occludes the binding site for the heterotrimeric G protein, preventing its activation

(desensitization) . Second, arrestins link the receptor to elements of the internalization machinery, clathrin and clathrin adaptor AP2 (C-terminal tail) , which promotes receptor internalization via the coated pits and

subsequent transport to internal compartments, called endosome. Subsequently, the receptor could be either directed to degradation compartments (lysosomes) or recycled back to the plasma membrane where it can once more act as a signal. The strength of arrestin-receptor interaction plays a role in this choice: tighter

complexes tend to increase the probability of receptor degradation, whereas more transient complexes favor recycling, although this "rule" is far from absolute.

Therefore, arrestins function as adapter proteins that facilitate desensitization, internalization and signaling of G protein-coupled receptors (GPCRs) . Quenching of G protein signaling via arrestins is best understood in the visual system where arrestin-1 quenches phototransduction via its ability to bind to the phosphorylated, light- activated form of the visual photoreceptor rhodopsin.

The binding ability of ligands to their respective GPCR, such as the arrestin-rhodopsin complex, opens a broad field of drug discovery and drug screening in order to diagnose and treat diseases in the field of the GPCR moderated exocytotic or endocytotic biochemical

processes. The international patent application WO

2008/114020 A2 discloses a method for selecting a GPCR with increased stability . Said method comprises a) providing one or more mutants of a parent GPCR, b) selecting a ligand, the ligand being one which binds to the parent GPCR when the GPCR is residing in a particular conformation, c) determining whether the or each mutant GPCR has increased stability with respect to binding the selected ligand compared to the stability of the parent GPCR with respect to binding that ligand, and d)

selecting those mutants that have an increased stability compared to the parent GPCR with respect to binding the selected ligand. Mutants of β-adrenergic receptor, adenosine receptor and neurotensin receptor are also disclosed in this application.

Unfortunately, the number of mutants of the GPCRs which represent the same biochemical behavior as the parent GPCR are limited . Further, mutant GPCRs a e rather difficult to be synthesized and therefore the binding assays are rather expensive due to the expensive mutant GPCRs .

It is therefore an obj ective of the present invention to provide a method for identifying the binding ability of GPCRs to ligand at lower cost and with a broader spectrum of ligand-GPCR pairs . Further, it is an ob ective of the present invention to stabilize the binding of the ligand and the GPCR by suitable ligands and/or suitable

ligand/GPCR pairs .

This ob ective is achieved according to the present invention by a method for determining the binding ability of a G-protein coupled receptor, hereinafter referred to as GPCR, and a mutateable ligand, said method comprising the steps of:

a) providing a well microtiter plate having the wells disposed in a matrix having a first number of rows and a second number of columns; b) providing a GPCR, such as rhodopsin, in each of said wells;

c) providing a number of mutants of the parent

ligand, wherein the parent ligand being one that binds to the GPCR when the GPCR is

residing in a particular conformation; d) bringing the mutants of the parent ligand in the wells into contact with the GPCR under conditions where the parent ligand would couple to the GPCR; and

e) determining for each mutant whether the mutant ligand has a weaker or stronger binding ability as compared to the standard binding ability of the parent ligand and the GPCR by determining the amount of coupled mutant-GPCR complex in said wells.

The present method therefore offers the opportunity to simultaneously scan the binding ability of the mutants of a parent ligand within the same set-up conditions of the assay for a number of mutants wherein the mutagenesis on the ligand can be executed at residue resolution due to the simpler organic structure of the mutateable ligand as compared to the structure of the GPCR that has been mutated for example in the WO 2008/114020 A2. Further, the combination of mutants allows the modification of binding affinity and stability of the GPCR-ligand complex for diagnostic purposes, pharmacological intervention or drug discovery.

In order provide conditions that deliver a certain gradient of the reaction conditions in the wells, a preferred embodiment of the present inventions is present where for the wells of each row or each column the same mutant is used and wherein the reaction condition in the well, such as salt concentration or solvent agent, changes from well to well in said row or in said column. For providing the advantageous and comparable reaction conditions in the wells, the reaction condition can be kept constant in those wells belonging to the same row or to the same column.

In order to provide a reference value of the binding of the parent ligand and the GPCR to the assay, the parent ligand may be added to all wells belonging to the same row or the same column of the well microtiter plate. With other words, the conditions under which the mutants are tested are also applied to the parent ligand-GPCR pair which enhances the scalability of the results for the mutants simultaneously tested. Any impact on the reaction conditions that might differ from one well microtiter plate assay to the other, can therefore be eliminated due to the reference value identified for the parent ligand- GPCR-pair .

Advantageously, the mutants are provided in solubilized form that enables rather simple assay conditions for the investigation on the binding ability of a specific mutant ligand to the GPCR.

In a further preferred embodiment of the present

invention, the parent ligand can be from the agonist class of ligands and the particular conformation is an agonist conformation, or the parent ligand is from the antagonist class of ligands and the particular

conformation is an antagonist conformation. Preferably, the parent ligand is any one of a full agonist, a partial agonist, an inverse agonist, an antagonist, or the parent ligand is from the inverse agonist class of ligands and the particular conformation is an inverse agonist

conformation. Therefore, these ligands guarantee the application of the assay trials with respect to the binding ability on the interesting spectrum of in-vivo biochemical reactions that are controlled/influenced by the GPCR and its binding stabilization to the mutant ligands in question. In a further preferred embodiment of the present

invention, the parent ligand is a polypeptide which binds to the GPCR. Preferably, the polypeptide is any of an antibody, an ankyrin, a G protein, an RGS protein, an arrestin, a GPCR kinase, a receptor tyrosine kinase, a RAMP, a NSF, a GPCR, an NMDA receptor subunit NR1 or NR2a, a calcyon, a fibronectin domain framework, or a fragment or derivative thereof that binds to the GPCR. In a further preferred embodiment of the present

invention the mutants of the parent ligand are provided in a form that one of more mutated amino acid residues are replacing the parent amino acid residue. Preferably, a parent amino acid residue can be mutated individually to alanine/glycine.

Excellent results on the binding ability can be achieved when the parent ligand is one of arrestin 1, arrestin 2, arrestin 3 and arrestin 4. Preferably, mutants of the parent ligand having a higher binding affinity to the GPCR than the parent mutant are candidates for drug discovery .

With respect to the pair of a mutant ligand and a GPCR, a solution for the objective mentioned above is achieved by a pair of a mutant ligand and a GPCR wherein the binding stability of said pair is higher as compared to the pair consisting of the parent ligand and the GPCR, preferably identified according to the method of any of the claims 1 to 10. Accordingly, an alternative solution for a pair of mutant ligand and a GPCR is achieved by a pair of a mutant ligand and a GPCR wherein the binding stability of said pair is lower as compared to the pair consisting of the parent ligand and the GPCR, preferably identified according to the method of any of the claims 1 to 10.

Further, the objective is achieved according to the present invention with respect to the ligand by a mutant ligand of the parent ligand of the arrestin type having a higher binding stability with a GPCR than the pair of the parent ligand and the GPCR, preferably identified

according to the method of any of the claims 1 to 10. Alternatively, this objective is achieved according to the present invention by a mutant ligand of the parent ligand of the arrestin type having a lower binding stability with a GPCR than the pair of the parent ligand and the GPCR, preferably identified according to the method of any of the claims 1 to 10.

Further aspect of the present invention is achieved by the use of the pair according to claim 11 or 12 or the mutant ligand of claim 13 or 14 in a drug screening using complementation assay with mutant ligands optimized for either higher or lower affinity to the GPCR as compared to the binding affinity of the parent ligand and the GPCR.

Preferred embodiments of the present invention are hereinafter described in more detail with reference to the following drawings which depict in :

Figure 1 an overview of the scanning mutagenesis on

arrestin;

Figure 2 schematically an binding assay using arrestin- mCherry and rhodopsin in native rod outer segment (ROS ) membranes ;

Figure 3 an overview on the course of the half maximal inhibitory concentration IC ₅₀ of the mutant ligands of arrestin;

Figure 4 an overview on the combination of mutations to create an arrestin "Super Binder" ; Figure 5 schematically a systematic construction of an arrestin "Super Binder" ;

Figure 6 schematically the arrestin binding mechanism at single amino acid resolution;

Figure 7 schematically the model of the arrestin- rhodopsing complex ;

Figure 8 schematically the C-tail exchange mechanism at single amino acid resolution by which the C- tail of arrestin is replaced by the

phosphorylated C-tail of a GPCR; and

Figure 9 schematically the transfer of mutations to

other arrestins .

Table 1 (at the end of the specification) shows the relative binding affinities for mutants covering the complete sequence of the GPCR ligand arrestin-1. Herein, each amino acid position in the arrestin has been mutated to alanine (A) or glycine (G) .

Figure 1 shows the amino acid sequence of arrestin . The scanning mutagenesis on arrestin was executed by known sequencing technologies in order to mutate each residue individually to alanine (A) or glycine (G) . During the first sequencing round 74% of the 403 mutant ligands have been successfully generated . Complete coverage has been achieved after further mutagenesis and sequencing rounds .

As schematically illustrated in Figure 2, the mutants of the arrestin-1 that have been obtained by scanning mutagenesis are transformed into Escherichia coli (E . coli) . The relative expression level of each mutant with respect to the wild type (parent ligand) was determined using fluorescence of a C-terminal mCherry fusion to arrestin. For complex formation arrestin is combined with rhodopsin in rod outer segment (ROS) membranes that had been phosphorylated with native rhodopsin kinase (GRK1 ) . To minimize the effect of variations in the expression of different arrestin mutants the assay contained 1.25 μΜ rhodopsin, far in excess of the 5-50 nM apparent binding affinity of arrestin-1. Consequently, no correlation has been observed between the amounts of functionally

expressed arrestin-1 and the ability of a mutant to bind rhodopsin under increasing ionic strength measured as described below .

Although, the present examples will now discuss in more detail the complex (pair) of arrestin as parent ligand and rhodopsin as GPCR, the method according to the present invention is open to scan any arbitrary pair of parent ligand and GPCR . In particular, the use of the well microtiter plate offers a brought range of scanning experiments which can be established simultaneously under reaction conditions that are equal for all mutated ligands .

In more detail, Figure 2 shows the direct binding assay of arrestin-mCherry and rhodopsin in native ROS

membranes. Plasmids containing arrestin mutated by scanning mutagenesis were transformed into E. coli. The relative expression level of each mutant with respect to the wild type was determined using fluorescence of a C- terminal mCherry fusion to arrestin. For complex

formation arrestin is combined with rhodopsin in ROS membranes that had been phosphorylated with native rhodopsin kinase (GRK1) . For comparison of relative binding in the present example, each time 11 arrestin mutants and wild-type arrestin as control were combined in a 96-well microtiter plate and probed for binding to to light-activated rhodopsin in eight different salt concentrations. For dark-adapted rhodopsin, only single point measurements were conducted for each mutant. After centrifugation and washing steps, the amount of bound arrestin was quantified using fluorescence of the mCherry fusion protein. The resulting data were fitted to

sigmoidal dose-response curves with variable slope to extract the IC50 values and 95% confidence intervals as listed in Table 1. A library of 403 arrestin mutants has been screened for their IC ₅₀ values. All measurements were performed in the frame of fluorescence quantification of arrestin-mCherry fusion proteins . To find mutant combinations that would increase binding, 23 of 24 mutants with the highest IC ₅₀ values measured were selected, ranging from 1.14 M for G297A till 0.56 M for R291A, from the arrestin-1 mutant library . They were combined with the strongest binding mutation F375A ( IC ₅₀ = 1.32 ± 0.31 M) . Further, 15 mutants were selected with significantly higher IC ₅₀ values than WT (wild type) . As control 10 mutants were chosen with IC ₅₀ values similar to WT (within the standard deviation of WT measurements ) and 2 mutants with significantly lower IC ₅₀ values than WT as well as 3 mutants that showed a weak signal ( I24A, V57A, and I149A) due to low functio- nal expression levels . Altogether, 53 double mutants have been constructed . The screening procedure for IC ₅₀ values was then extended additionally to the one previously employed with a second range of assayed sodium chloride concentrations to be able to fit binding data of mutants with high IC ₅₀ values with adequate accuracy . Each

combined mutant was sub ected to both sodium chloride screening ranges to derive IC ₅₀ values (Fig . 4 ) .

Figure 4 shows the half-maximal inhibitory concentration ( IC ₅₀ ) values of sodium chloride for formation of com ^¬ plexes composed of mutant arrestin-1 and phosphorylated activated rhodopsin . Double mutants (circles) are sorted by decreasing IC ₅ o value (from left to right) and are composed of F375A+X, where the single mutant X is shown underneath and named on the x-axis. Triple mutants

(triangles) are composed of either F375A+A307G+X or

F375A+T304A+X, while quadruple mutants are combined like F375A+T304A+E341A+X or like F375A+T304A+F380A+X. The size of the shape encodes the functional expression level of regarding mutant and is scaled in relation to the arres- tin-1 wild type functional expression level (legend) . The functional expression level tells how much of arrestin protein, functional in terms of rhodopsin binding, was expressed and pulled down for regarding mutant in

relation to wild type arrestin-mCherry construct at 100 mM NaCl . In case of combined mutants with elevated IC50, the reference point was mutant F375A at 492 mM NaCl .

The IC ₅₀ values could be derived for 49 out of 53 double mutants, 4 double mutants exposed signal intensities below detection limit. About two thirds, exactly 33 of 49 mutants, exhibited IC ₅₀ values similar to F375A (within the standard deviation of the IC ₅₀ value derived in 23 independent measurements for F375A, see above) . A sum of 12 mutants had significantly higher IC ₅₀ values than F375A and 4 double mutants significantly lower values . Another series of mutations has been added on top of double mu ^¬ tant A307G+F375A, which was leading the screen with the highest IC ₅₀ value observed (2.83 M) , or on top of mutant T304A+F375A, which was with an IC ₅₀ value of 1.51 M under the 12 best-binding mutants. Of 38 constructed triple mutants IC ₅₀ values for 35 mutants could be determined and those were ranging from 3.52 to 1.01 M. Although triple mutants containing A307G bound in very high salt

concentrations to light-activated phosphorylated

rhodopsin (R*-P) , quantitatively the amounts of formed complexes were low. Thus triple mutants E341A+T304+F375A and F380A+T304A+F375A were chosen with IC ₅₀ values of 2.75 M and 2.09 M, respectively, to design 15 quadruple mutants. For quadruple mutants, IC ₅₀ values from 2.95 M to 1.37 M have been obtained. The two quadruple mutants R171A+E341A+T304A+F375A and D303A+E341A+T304A+F375A reached IC ₅₀ values amounting to 720% and 710% of the WT value (0.41 ± 0.05 M) (Fig. 4 and Table 2) . Conclusively, it was possible to increase complex stability more than 7 times under the pressure of high ionic strength by selection and combination of single mutants to quadruple mutants .

As a short example, for comparison of relative binding, 11 arrestin mutants and WT arrestin as control were combined in a 96 well microtiter plate and probed for binding to dark and light activated rhodopsin in 8 different salt concentrations (100 mM to 2.4 M) . After centrifugation and washing steps the amount of bound arrestin has been quantified using fluorescence of the mCherry fusion protein . The resulting data were fitted to sigmoidal dose-response curves with variable slope to extract half maximal inhibitory concentration ( IC ₅₀ ) values and 95% confidence intervals listed in table 1 (at the end of the specification) . A selection of strong and weak binding mutations discussed in the main text have been measured several times to increase accuracy of the determined IC ₅₀ values . The variation of IC ₅₀ values for WT arrestin was 0.41 ± 0.04 M from 59 independent

experiments in agreement with previous reports using radiolabeled arrestin-1. Among the 25 best binding mutations , 13 affected polar residues including 10 residues that are charged under physiological conditions . Similarly, 10 of the worst 25 binding mutations affected polar residues including 4 charged residues . This even distribution between polar and hydrophobic residues demonstrates that the assay is not biased even though increasing ionic strength predominately affects

hydrophilic interactions . This is in agreement with the idea that arrestin binding to rhodopsin involves a multitude of hydrophilic and hydrophobic interactions, as well as specific conformational changes and is not dominated by a few charged interactions. Figure 3 provides an overview on the course of the half maximal inhibitory concentrations ( IC _. 50 values ) of all single mutant ligands of arrestin. The investigation involved the binding ability of all 403 arrestin mutants listed in table 1. Using the pull downs of an arrestin- mCherry fusion protein, the analysis comprised the comparison of the rhodopsin binding of all 403 mutants covering the complete arrestin sequence . This information provides a functional 4 ^th dimension to the crystal structures of inactive, preactivated and active

arrestins . The resulting single amino acid resolution functional maps reveal a series of critical interactions in the polar core and along the C-tail of arrestin that are interrupted during arrestin activation . The data further reveals several patches of amino acids that strongly reduce binding and act as direct binding interfaces to rhodopsin . This information in combination with computational molecular docking of active arrestin and light-activated rhodopsin allow to develop a model of the arrestin-rhodopsin complex as shown in Figure 7.

Figure 4 shows the combination of mutations of the arrestin- 1 ligand in terms of their binding affinity to rhodopsin . Rectangular symbols represent the mutant ligands which have just one mutated amino acid position; the circular symbols represent the assays with mutant ligands having two mutated amino acid positions , and the triangular symbol represent assays with mutant ligands comprising tree mutated amino acid positions . Rhombic symbols represent quadruple mutant ligands of arrestin-1.

Therefore, the triangular image in Figure 5 represents the systematic construction of a mutated arrestin having excellent binding properties to the GPCR rhodopsin. These "super binders" are identified by an iterative

combination of the mutations of arrestin. For example, those mutant arrestins with a single mutated amino acid position showing in this first class of the single mutated amino acid positions relatively high binding ability are a candidate for the following assay using this single mutated arrestin having now a second mutated amino acid position and so on .

The triangle in Figure 5 therefore shows in the

triangular radar among the binding affinity, the mutant arrestin stability and the functional expression of the mutant arrestin significantly higher values for the IC ₅₀ especially for those mutant ligands which have three or four mutated amino acid positions as explained above in more detail with reference to Figure 4.

Spoken more generally, the method according to the present invention therefore can apply an iterative approach, too . In this sense, the starting point is the scan of a ligand having a single mutated amino acid position and observing the respective response in the binding ability . Those of the single mutated ligands having a relatively high binding ability are then the subj ect for the second assay wherein an additional second amino acid position is mutated . Accordingly, those double mutated ligands showing superior binding ability are the subj ect for the third assay wherein an additional third amino acid position is mutated, and so on . This iterative approach can therefore be executed until a desired level of biochemical reactivity/binding ability/ functional potential is achieved . It has to be noted that the iterative approach can be also executed in the opposite sense searching for multiple mutated ligand that has particularly low biochemical reactivity/binding ability/functional potential as compared to the parent ligand in relationship to the GPCR it binds to.

The model in Figures 6 and 7 show how the arrestin finger loop and beta-strands XV and XVI interact with TM5/TM6 of rhodopsin and act as a sensor for the active receptor conformation . Phosphosensing is achieved by a series of amino acids that anchor the C-tail of arrestin in a position that blocks binding of the receptor (see Figure 8) . In a C-tail exchange mechanism, the C-tail of

arrestin is released and subsequently replaced by the phosphorylated C-terminus of the receptor .

Figure 7 shows the conceptual model of an arrestin-GPCR complex derived from molecular docking of active arrestin and light-activated rhodopsin . The phosphorylated C- terminus of rhodopsin binds along the arrestin N-domain and interacts with several charged residues exposed during release of the arrestin C-terminus . The finger and lariat loops (upper and middle inset) fit into the crevice opening during rhodopsin activation . In this position Gln69 or Asp73 in the arrestin finger loop can interact with Leu72 and Asn73, two residues in TM2 of rhodopsin that are critical for the binding of arrestin- 1. The lariat loop mediates contacts to the cytoplasmic ends of TM6 and TM7/H8, two regions whose relative position is involved in biased signaling of β-adrenergic receptors and arrestin binding to rhodopsin . The edge of the C-domain (lower inset) contains a set of amino acids that could interact with the phospholipid membrane or form a secondary binding site for GPCR dimers .

Figure 8 shows at single amino acid resolution functional maps of arrestin-1. Binding of 403 arrestin mutants (IC ₅₀ ) values shown as increasing ribbon width and as spectrum ranging from red over white to blue) plotted on the crystal structures of inactive ² , preactivated p44 arrestin and an arrestin-1 model based on the crystal structure of arrestin-2 with bound receptor

phosphopetide ⁴ . Several residues including three

hydrophobic phenylalanines (F375, F377, F380) and the charged R382 anchor the C-tail of arrestin-1 into the 3- element interaction and polar core regulatory sites.

Mutation of these key residues and their interaction partners leads to strongly increased binding to

phosphorylated, light activated rhodopsin. Truncation of the C-tail in p44 arrestin leads to a conformational change that releases R29 from the polar core and exposes several other charged residues (K14, K15, Q69, K300) whose mutation leads to strongly reduced binding to phosphorylated rhodopsin. In active arrestin these residues directly interact with phosphorylated serines and threonines in the receptor peptide (green, V2Rpp) . Together these data suggest a phosphosensing mechanism in which the C-tail of arrestin-1 is exchanged for the phosphorylated C-terminus of rhodopsin during formation of the desensitizing arrestin complex.

Figure 9 shows an overview of a transfer of mutations to other arrestins. In bold letters, the 20 best binders are identified. Light grey letters represent the 20 worst binders to the rhodopsin GPCR.

Table 1 lists the binding parameters for 403

alanine/glycine mutants in arrestin-1. These finding according to the existing data now can be used to modify the binding of arrestin-2+3 to

pharmacological interesting GPCRs . Combination of mutants allows modification of binding affinity and stability of GPCR-arrestin complexes for diagnostic purposes,

pharmacological intervention or drug discovery (e.g.

beta-arrestin recruitment assays , structure-based drug discovery, silencing of hyperactive GPCRs etc . ) . Table 1 below shows a list of I C ₅₀ values of NaCl on phosphorylated and light-activated rod outer segment (ROS*-P) membrane binding for 403 arrestin mutations covering the complete arrestin-1 sequence. I C ₅₀ values for each single mutant have been obtained from 8 measurements in increasing concentrations of NaCl (100 mM to 2.4 M) with the quality of fit indicated as R ² and the variation as 95% confidence interval. A selection of functionally important residues has been measured multiple times and the values averaged. 16 mutations listed under remarks have been removed from analysis either because they did not express (as indicated by in gel-fluorescence of the mCherry fluorescence marker) or expression was too low to obtain a reliable signal.

Table 1

E39A 0.37 0.35 to 0.39 0.9988 2

P40A 0.37 0.36 to 0.40 0.9984 2

V41A 0.40 0.37 to 0.43 0.9981 1

D42A 0.40 0.36 to 0.45 0.9964 2

G43A 0.50 0.45 to 0.55 0.9948 3 signal weak

V44A 0.51 0.49 to 0.53 0.9994 2

V45A 0.33 0.30 to 0.36 0.9971 1

L46A 0.38 0.34 to 0.42 0.9967 1

V47A 0.44 0.42 to 0.46 0.9993 1

D48A 0.38 0.30 to 0.51 0.9730 2

P49A 0.47 0.41 to 0.54 0.9885 1

E50A 0.32 0.28 to 0.36 0.9941 1

L51A 0.30 0.27 to 0.34 0.9948 1

V52A 0.31 0.28 to 0.35 0.9963 1

K53A 0.31 0.28 to 0.35 0.9953 1

G54A 0.42 0.38 to 0.46 0.9975 1

K55A 0.43 0.41 to 0.46 0.9989 1

R56A 0.42 0.40 to 0.45 0.9986 1

V57A 0.53 0.46 to 0.62 0.9870 2 signal weak

Y58A 0.35 0.33 to 0.38 0.9980 1

V59A 0.42 0.46 to 0.50 0.9978 2

S60A 0.48 0.46 to 0.50 0.9993 1

L61A 0.35 0.29 to 0.43 0.9822 1

T62A 0.31 0.27 to 0.36 0.9928 1

C63A 0.41 0.36 to 0.76 0.9959 3

A64G 0.45 0.43 to 0.47 0.9990 1

F65A 0.47 0.45 to 0.50 0.9984 2

R66A 0.64 0.59 to 0.69 0.9978 3

Y67A 0.41 0.39 to 0.42 0.9997 1

G68A 0.45 0.43 to 0.47 0.9990 1

Q69A 0.24 0.20 to 0.28 0.9873 2

E70A 0.45 0.39 to 0.52 0.9852 1

D71A 0.45 0.43 to 0.48 0.9979 1

I72A 0.45 0.42 to 0.49 0.9963 1

D73A 0.31 0.29 to 0.32 0.9992 1

V74A 0.32 0.27 to 0.38 0.9875 2

M75A 0.26 0.24 to 0.28 0.9980 2

G76A 0.41 0.35 to 0.47 0.9947 1

L77A 0.33 0.31 to 0.35 0.9980 2

S78A 0.46 0.32 to 0.66 0.9495 1

F79A 0.43 0.36 to 0.51 0.9903 1

R80A 0.58 0.54 to 0.63 0.9984 2

R81A 0.47 0.45 to 0.49 0.9989 1

D82A 0.46 0.41 to 0.51 0.9959 1

L83A 0.44 0.43 to 0.45 0.9997 1

Y84A 0.51 0.48 to 0.54 0.9986 2

F85A 0.48 0.43 to 0.53 0.9961 2

S86A 0.47 0.44 to 0.51 0.9995 1

Q87A 0.54 0.50 to 0.59 0.9974 2

V88A 0.48 0.45 to 0.52 0.9975 1

Q89A 0.37 0.36 to 0.38 0.9998 1

V90A 0.44 0.41 to 0.47 0.9979 2

F91A 0.41 0.40 to 0.43 0.9994 1

P92A 0.32 0.28 to 0.36 0.9931 2

P93A 0.44 0.40 to 0.48 0.9978 1

V94A 0.40 0.35 to 0.46 0.9927 2

G95A 0.37 0.35 to 0.39 0.9990 1

A96G 0.37 0.35 to 0.40 0.9984 1

S97A 0.34 0.30 to 0.39 0.9947 1

G98A 0.39 0.34 to 0.45 0.9944 1

A99G 0.38 0.37 to 0.40 0.9993 1

T100A 0.35 0.32 to 0.39 0.9962 1

T101A 0.42 0.36 to 0.49 0.9918 2 signal weak

R102A 0.35 0.30 to 0.39 0.9949 1

L103A 0.48 0.42 to 0.55 0.9953 1

Q104A 0.30 0.29 to 0.32 0.9981 1

E105A 0.35 0.32 to 0.39 0.9969 1

S106A 0.32 0.29 to 0.35 0.9964 1

L107A 0.44 0.43 to 0.45 0.9996 1

I108A 0.44 0.41 to 0.47 0.9986 1

K109A 0.42 0.38 to 0.46 0.9968 1

110A 0.37 0.34 to 0.41 0.9975 1

L111A 0.70 0.64 to 0.76 0.9980 2

G112A 0.45 0.42 to 0.48 0.9977 1 A113G 0.45 0.41 to 0.48 0.9977 1

N114A 0.48 0.45 to 0.50 0.9988 1

T115A 0.45 0.41 to 0.49 0.9973 1

Y116A 0.51 0.48 to 0.54 0.9975

P117A 0.44 0.42 to 0.46 0.9992 1

F118A 0.33 0.32 to 0.35 0.9990 1

L119A 0.39 0.37 to 0.41 0.9989 1

L120A 0.46 0.41 to 0.52 0.9910 1

T121A 0.41 0.39 to 0.44 0.9988 1

F122A 0.41 0.40 to 0.42 0.9999 1

P123A 0.40 0.35 to 0.45 0.9947 1

D124A 0.38 0.33 to 0.42 0.9956 1

Y125A 0.37 0.35 to 0.41 0.9979 1

L126A 0.41 0.38 to 0.44 0.9981 1

P127A 0.33 0.30 to 0.38 0.9953 1

C128A 0.40 0.33 to 0.49 0.9905 1

S129A 0.53 0.47 to 0.60 0.9970

V130A 0.41 0.37 to 0.45 0.9974

M131A 0.39 0.37 to 0.41 0.9989 1

L132A 0.47 0.45 to 0.49 0.9991

Q133A 0.35 0.32 to 0.39 0.9953 1

P134A 0.45 0.42 to 0.48 0.9986 1

A135G 0.49 0.45 to 0.53 0.9961

P136A 0.46 0.45 to 0.48 0.9995 1

Q137A 0.45 0.42 to 0.49 0.9952

D138A 0.41 0.39 to 0.42 0.9995 1

V139A 0.39 0.37 to 0.41 0.9988 1

G140A 0.38 0.36 to 0.39 0.9991 1

K141A 0.49 0.48 to 0.50 0.9998 1

S142A 0.39 0.36 to 0.42 0.9976 2

C143A 0.50 0.48 to 0.51 0.9997 2

G144A 0.57 0.48 to 0.68 0.9921 2

V145A 0.49 0.47 to 0.51 0.9994 1

D146A 0.57 0.51 to 0.64 0.9947 2

F147A 0.43 0.46 to 0.56 0.9945 3

E148A 0.42 0.40 to 0.44 0.9988 2

I149A 0.74 0.64 to 0.85 0.9901 3 signal weak

K150A 0.42 0.40 to 0.44 0.9993 1

A151G 0.35 0.34 to 0.36 0.9994 1

F152A 0.42 0.37 to 0.46 0.9952 1

A153G 0.32 0.31 to 0.34 0.9993

T154A 0.44 0.35 to 0.55 0.9809 1

H155A 0.35 0.32 to 0.38 0.9970 1

S156A 0.33 0.31 to 0.36 0.9977 1

T157A 0.62 0.55 to 0.70 0.9965 signal weak

D158A 0.39 0.32 to 0.38 0.9991 1

V159A 0.37 0.35 to 0.38 0.9995 1

E160A 0.44 0.42 to 0.46 0.9991 1

E161A 0.37 0.35 to 0.40 0.9985 1

D162A 0.40 0.38 to 0.42 0.9992 1

K163A 0.43 0.40 to 0.47 0.9968

I164A 0.41 0.38 to 0.44 0.9979

P165A 0.42 0.38 to 0.46 0.9970 1

K166A 0.38 0.33 to 0.42 0.9960 1

K167A 0.36 0.34 to 0.39 0.9979 1

S168A 0.48 0.46 to 0.51 0.9986 1

S169A 0.46 0.42 to 0.50 0.9973 1

V170A 0.58 0.51 to 0.66 0.9975 2

R171A 0.54 0.47 to 0.61 0.9962 2

L172A 0.28 0.26 to 0.30 0.9976 2

L173A 0.51 0.47 to 0.55 0.9983 2

I174A 0.38 0.36 to 0.40 0.9990 2

R175A 0.91 0.77 to 1.08 0.9948 3

K176A 0.47 0.43 to 0.51 0.9971 1

V177A 0.33 0.29 to 0.37 0.9962 2

Q178A 0.40 0.34 to 0.45 0.9941 1

H179A 0.32 0.29 to 0.35 0.9963 2

A180G 0.37 0.31 to 0.43 0.9913 1

P181A 0.38 0.35 to 0.40 0.9984 1

R182A 0.41 0.34 to 0.49 0.9916 1

D183A 0.40 0.37 to 0.43 0.9985 1

Ml 84 A 0.46 0.43 to 0.49 0.9987 1

G185A 0.33 0.29 to 0.37 0.9951 1

P186A 0.37 0.35 to 0.40 0.9984 1 Q187A 0.40 0.35 to 0.45 0.9957 1

P188A 0.27 0.21 to 0.56 0.9973 3

R189A 0.40 0.38 to 0.43 0.9988 1

A190G 0.41 0.37 to 0.45 0.9974 1

E191A 0.32 0.29 to 0.36 0.9961 1

A192G 0.36 0.33 to 0.38 0.9982 1

S193A 0.43 0.38 to 0.48 0.9962 1

W194A 0.24 0.22 to 0.28 0.9899

Q195A 0.33 0.31 to 0.36 0.9976 1

F196A 0.30 0.27 to 0.34 0.9949 1

F197A 0.31 0.28 to 0.34 0.9963

M198A 0.32 0.28 to 0.36 0.9952 1

S199A 0.31 0.28 to 0.33 0.9975 1

D200A 0.42 0.38 to 0.46 0.9964

K201A 0.40 0.37 to 0.43 0.9983 1

P202A 0.35 0.31 to 0.39 0.9948

L203A 0.39 0.35 to 0.43 0.9957 1

R204A 0.44 0.41 to 0.46 0.9989 1

L205A 0.40 0.38 to 0.42 0.9991 1

A206G 0.36 0.34 to 0.39 0.9986 1

V207A 0.42 0.40 to 0.44 0.9990 1

S208A 0.41 0.39 to 0.42 0.9994 1

L209A 0.47 0.42 to 0.51 0.9966

S210A 0.35 0.33 to 0.37 0.9988 1

K211A 0.40 0.38 to 0.43 0.9986 1

E212A 0.46 0.42 to 0.50 0.9963 1

I213A 0.36 0.34 to 0.38 0.9985 1

Y214A 0.37 0.35 to 0.39 0.9988 1

Y215A 0.42 0.34 to 0.50 0.9918 1

H216A 0.40 0.38 to 0.42 0.9988 1

G217A 0.36 0.34 to 0.38 0.9986 1

E218A 0.42 0.41 to 0.44 0.9995 1

P219A 0.44 0.41 to 0.46 0.9985 1

I220A 0.53 0.46 to 0.61 0.9895 signal weak

P221A 0.43 0.40 to 0.46 0.9981 1

V222A 0.51 0.48 to 0.53 0.9989

T223A 0.41 0.38 to 0.43 0.9985 1

V224A 0.51 0.39 to 0.67 0.9416 signal weak

A225G 0.44 0.41 to 0.47 0.9983 1

V226A 0.54 0.45 to 0.66 0.9856 signal weak

T227A 0.38 0.37 to 0.40 0.9991 1

N228A 0.34 0.32 to 0.37 0.9975 1

S229A 0.31 0.29 to 0.34 0.9971 1

T230A 0.42 0.41 to 0.43 0.9996 1

E231A 0.48 0.46 to 0.50 0.9994 1

K232A 0.31 0.27 to 0.53 0.9970

T233A 0.36 0.33 to 0.38 0.9978 1

V234A 0.31 0.29 to 0.33 0.9987 1

K235A 0.34 0.32 to 0.37 0.9976 1

K236A 0.40 0.37 to 0.44 0.9979 1

I237A 0.40 0.37 to 0.43 0.9974 1

K238A 0.40 0.35 to 0.46 0.9946 1

V239A 0.38 0.36 to 0.40 0.9988 1

L240A 0.44 0.40 to 0.48 0.9961 1

V241A - - - no expression

E242A 0.36 0.34 to 0.37 0.9992 1

Q243A 0.40 0.38 to 0.41 0.9992 1

V244A 0.39 0.37 to 0.41 0.9990 1

T245A 0.38 0.35 to 0.40 0.9978 1

N246A 0.36 0.33 to 0.40 0.9960 2

V247A 0.59 0.55 to 0.62 0.9989 2

V248A 0.40 0.38 to 0.42 0.9991 1

L249A 0.31 0.28 to 0.33 0.9968 3

Y250A 0.31 0.27 to 0.48 0.9938 3

S251A 0.33 0.31 to 0.35 0.9975 3

S252A 0.32 0.31 to 0.33 0.9995 1

D253A 0.50 0.45 to 0.56 0.9912 1

Y254A 0.31 0.26 to 0.36 0.9913 1

Y255A 0.44 0.42 to 0.46 0.9988 1

I256A 0.44 0.39 to 0.48 0.9957 2

K257A 0.50 0.46 to 0.55 0.9968 2

T258A - - - 5 no expression

V259A 0.40 0.37 to 0.42 0.9983 1

A260G 0.41 0.37 to 0.44 0.9975 1 A261G 0.46 0.43 to 0.49 0.9986 2

E262A 0.33 0.31 to 0.35 0.9986 1

E263A 0.47 0.43 to 0.51 0.9963 1

A264G 0.37 0.36 to 0.39 0.9992 1

Q265A 0.42 0.33 to 0.53 0.9822 1

E266A 0.35 0.30 to 0.41 0.9919 1

K267A 0.36 0.29 to 0.45 0.9863 1

V268A 0.40 0.38 to 0.43 0.9984 1

P269A 0.42 0.39 to 0.46 0.9972 1

P270A 0.32 0.21 to 0.48 0.9631 1

N271A 0.39 0.37 to 0.42 0.9979 1

S272A 0.36 0.32 to 0.40 0.9964 1

S273A 0.38 0.35 to 0.43 0.9966 1

L274A 0.40 0.37 to 0.43 0.9984 1

T275A 0.42 0.39 to 0.45 0.9982 1

K276A 0.39 0.36 to 0.42 0.9972 1

T277A 0.53 0.48 to 0.69 0.9990 1

L278A 0.46 0.43 to 0.49 0.9985 1

T279A 0.39 0.36 to 0.43 0.9974 1

L280A 0.32 0.27 to 0.37 0.9866 1

V281A 0.41 0.38 to 0.45 0.9975 1

P282A 0.30 0.28 to 0.32 0.9975

L283A 0.35 0.33 to 0.38 0.9977 1

L284A 0.56 0.51 to 0.60 0.9986

A285G 0.35 0.32 to 0.38 0.9971 1

N286A 0.40 0.37 to 0.42 0.9985 1

N287A 0.33 0.31 to 0.35 0.9982 1

R288A 0.39 0.37 to 0.42 0.9989 1

E289A 0.41 0.37 to 0.45 0.9973 1

R290A 0.45 0.41 to 0.49 0.9979 1

R291A 0.56 0.51 to 0.62 0.9967

G292A 0.40 0.35 to 0.47 0.9939 1

I293A 0.34 0.26 to 0.45 0.9786 1

A294G 0.37 0.33 to 0.43 0.9951 1

L295A 0.44 0.41 to 0.48 0.9979 2

D296A 0.92 0.71 to 1.21 0.9860 3

G297A 1.14 0.82 to 1.75 0.9879 3

K298A 0.52 0.41 to 0.66 0.9906 2

I299A 0.32 0.29 to 0.36 0.9956 1

K300A 0.33 0.31 to 0.35 0.9983 1

H301A 0.35 0.32 to 0.39 0.9968 1

E302A 0.41 0.36 to 0.47 0.9940 2

D303A 0.87 0.76 to 0.99 0.9964 3

T304A 0.84 0.74 to 0.96 0.9967 3

N305A 0.34 0.33 to 0.35 0.9997 1

L306A 0.36 0.35 to 0.37 0.9998 1

A307G 0.64 0.57 to 0.72 0.9967

S308A 0.31 0.31 to 0.32 0.9999 1

S309A 0.39 0.37 to 0.41 0.9986 1

T310A 0.46 0.41 to 0.53 0.9893 1

I311A 0.37 0.35 to 0.38 0.9991 1

I312A 0.41 0.19 to 0.88 0.9593 1

K313A 0.42 0.41 to 0.43 0.9996 1

E314A 0.35 0.34 to 0.37 0.9993 1

G315A 0.42 0.40 to 0.43 0.9995 1

I316A 0.40 0.39 to 0.41 0.9996 1

D317A 0.49 0.46 to 0.52 0.9984 1

K318A 0.36 0.35 to 0.37 0.9996 1

T319A 0.43 0.39 to 0.47 0.9976 1

V320A 0.25 0.24 to 0.27 0.9979

M321A 0.40 0.34 to 0.47 0.9930 1

G322A 0.20 0.04 to 0.73 0.8993 signal weak

I323A 0.32 0.31 to 0.33 0.9993 1

L324A 0.44 0.42 to 0.45 0.9994 1

V325A - - - 7 no expression

S326A 0.38 0.35 to 0.41 0.9980 2

Y327A - - - 7 no expression

Q328A 0.45 0.41 to 0.50 0.9969 1

I329A 0.45 0.44 to 0.46 0.9998 1

K330A 0.41 0.38 to 0.43 0.9985 1

V331A 0.41 0.40 to 0.43 0.9994 1

K332A 0.46 0.44 to 0.48 0.9992 1

L333A 0.39 0.36 to 0.43 0.9969 1

T334A 0.40 0.39 to 0.41 0.9997 1 V335A 0.40 0.38 to 0.42 0.9991 1

S336A 0.36 0.35 to 0.37 0.9993 1

G337A 0.29 0.28 to 0.31 0.9988 2

L338A 0.38 0.34 to 0.42 0.9955 2

L339A 0.30 0.27 to 0.34 0.9954 2

G340A 0.39 0.36 to 0.42 0.9976 1

E341A 0.57 0.52 to 0.63 0.9961 2

L342A 0.40 0.30 to 0.34 0.9969 2

T343A 0.32 0.31 to 0.33 0.9992 1

S344A 0.40 0.39 to 0.42 0.9996 1

S345A 0.34 0.33 to 0.36 0.9988 1

E346A 0.46 0.42 to 0.50 0.9946 1

V347A 0.36 0.35 to 0.38 0.9994 1

A348G 0.41 0.39 to 0.43 0.9990 1

T349A 0.30 0.29 to 0.32 0.9990 1

E350A 0.40 0.39 to 0.42 0.9996 1

V351A 0.43 0.40 to 0.45 0.9985 1

P352A 0.39 0.37 to 0.40 0.9993 1

F353A 0.52 0.47 to 0.58 0.9942 1

R354A 0.34 0.28 to 0.42 0.9859

L355A 0.45 0.42 to 0.48 0.9978 1

M356A 0.37 0.36 to 0.38 0.9997 1

H357A 0.48 0.45 to 0.51 0.9975 1

P358A 0.35 0.34 to 0.36 0.9997 1

Q359A 0.44 0.43 to 0.45 0.9997 1

P360A 0.38 0.35 to 0.42 0.9968 1

E361A 0.44 0.43 to 0.44 0.9999 1

D362A 0.45 0.41 to 0.50 0.9946 1

P363A 0.44 0.43 to 0.46 0.9994 1

D364A 0.44 0.43 to 0.45 0.9996 1

T365A 0.41 0.38 to 0.44 0.9979 1

A366G 0.43 0.41 to 0.46 0.9985 1

K367A 0.41 0.39 to 0.43 0.9993 1

E368A 0.44 0.41 to 0.47 0.9976 1

S369A 0.43 0.41 to 0.47 0.9992 1

F370A 0.46 0.44 to 0.49 0.9984 1

Q371A 0.36 0.35 to 0.37 0.9994 1

D372A 0.42 0.41 to 0.42 0.9998 1

E373A 0.44 0.43 to 0.45 0.9998 1

N374A 0.34 0.30 to 0.39 0.9932 2

F375A 1.28 0.93 to 1.88 0.9916 4

V376A 0.66 0.58 to 0.75 0.9957 3

F377A 1.08 0.82 to 1.43 0.9949 3

E378A 0.47 0.45 to 0.49 0.9992 1

E379A 0.43 0.42 to 0.44 0.9998 1

F380A 1.00 0.70 to 1.45 0.9903 3

A381G 0.46 0.37 to 0.59 0.9592 2 signal weak

R382A 0.97 0.76 to 1.24 0.9944 4

Q383A 0.39 0.37 to 0.42 0.9987 2

N384A 0.35 0.33 to 0.37 0.9988 1

L385A 0.44 0.41 to 0.48 0.9966 1

K386A 0.40 0.39 to 0.41 0.9996 1

D387A 0.46 0.43 to 0.49 0.9978 1

A388G 0.40 0.37 to 0.45 0.9971 1

G389A 0.40 0.38 to 0.42 0.9987 1

E390A 0.41 0.36 to 0.48 0.9929 2

Y391A 0.41 0.39 to 0.44 0.9983 1

K392A 0.37 0.35 to 0.39 0.9989 1

E393A 0.42 0.40 to 0.44 0.9991 1

E394A 0.44 0.41 to 0.49 0.9973 2

K395A 0.44 0.29 to 0.70 0.9495 3 signal weak

T396A 0.45 0.43 to 0.46 0.9996 1

D397A 0.41 0.38 to 0.45 0.9981 2

Q398A 0.34 0.30 to 0.38 0.9958 1

E399A 0.35 0.32 to 0.40 0.9959 1

A400G 0.36 0.32 to 0.41 0.9957 1

A401G 0.34 0.30 to 0.39 0.9947 1

M402A 0.40 0.36 to 0.44 0.9962 2

D403A 0.43 0.40 to 0.46 0.9984 1

E404A 0.41 0.36 to 0.46 0.9962 1 The following mutants were constructed earlier and belong to prior art: K2A ³ , I12A ⁴ , K14A ⁴ , K15A ⁴ , R18A ⁵ , Y25A ⁶ , D30A ³ , V44A ⁶ , L46A ⁶ , F65A ⁵ , D72A ⁷ , R102A ⁶ , L103A ⁶ , Q104A ⁶ , E105A ⁶ , S106A ⁶ , L107A ⁶ , I108A ⁶ , K109A ⁶ , K110A ⁶ , L111A ⁶ , D138A ⁵ , K142A ⁷ , D162A ⁵ , K166A ⁵ , V170A ¹ , L172A ¹ , L173A ¹ , I174A ¹ , R175A ¹ , V177A ¹ , Q178A ¹ , K235A ⁵ , Y250A ⁵ , E346A ⁵ , D296A ³ , D303A ³ , F375A ² , V376A ² , F377A ² , F380A ² , R382A ³ The earlier described mutants were analyzed in one-point measurements for binding to different states of

rhodopsin. IC50 values of sodium chloride were not derived earlier for those mutants.

¹ Gurevich & Benovic, 1996: Mechanism of phosphorylation- recognition by visual arrestin and the transition of arrestin into a high affinity binding state.

² Gurevich, 1998: The selectivity of visual arrestin for light-activated phosphorhodopsin is controlled by multiple nonredundant mechanisms.

³ Vishnivetskiy, Paz, Schubert, Hirsch & Gurevich, 1999: How does arrestin respond to the phosphorylated state of rhodopsin?

⁴ Vishnivetskiy, Schubert, Climaco, Gurevich, Velez,

Gurevich, 2000: An additional phosphate-binding element in arrestin molecule: Implications for the mechanism of arrestin activation.

⁵ Hanson & Gurevich, 2005: The differential engagement of arrestin surface charges by the various functional forms of the receptor.

⁶ Vishnivetskiy, Francis, Eps, Kim, Hanson, Klug, Hubbell, Gurevich, 2010: The role of arrestin alpha-helix I in receptor binding.

⁷ Vishnivetskiy, Baameur, Findley and Gurevich, 2013:

Critical role of the central 139-loop in stability and binding selectivity of arrestin-1

Detailed information on the experimental work Mutants were expressed and cells disrupted in buffer C [10 mM Hepes (pH 7.0), 100 mM NaCl, 1 mM DTT, 1 mM MgCl ₂ , and 0.1 mM EDTA] or buffer D (containing 1.842 M NaCl) , both supplemented with 0.2 mg/ mL lysozyme, 20 yg/mL DNase, 1.5 mM PMSF, and protease inhibitor mixture Roche Complete. Procedure C using buffer C and plate C was applied to wild type, single and combined mutants of the arrestin-mCherry construct. Procedure D using buffer D and plate D was applied to single mutant F375A and to combined mutants of the arrestin-mCherry construct. In detail, 1.024 mL cleared cell lysate containing wild type or mutant arrestin-mCherry construct in buffer C was mixed with 76 yL ROS-P*, while 637.1 yL cleared cell ly ^¬ sate in buffer D was mixed with 82.9 yL of the same ROS- P* (1.45 mg/mL) stock, obtaining master mixes C or D, respectively. Master mix C was distributed in 100-yL portions to 8 wells of a 96-well plate (in the following called plate C) with each well containing 100 yL buffer C with increasing amounts of sodium chloride, finally yielding 100, 247, 492, 737, and 982 mM and 1.472, 1.962, and 2.403 M NaCl in the 8 reaction mixes. Each plate C contained wild type arrestin-mCherry for reference and 11 different arrestin-mCherry mutants. Master mix D was portioned in 60-yL fractions and transferred to 8 wells of a 96-well plate (below called plate D) with each well con ^¬ taining 140 yL of the same buffer with different amounts of sodium chloride, resulting in 492, 737, and 982 mM and 1.472, 1.962, 2.403, 3.176, and 3.949 M NaCl. 11 dif ^¬ ferent arrestin-mCherry mutants were assayed with each plate C, containing arrestin-mCherry wild type construct (as reference) , and the same amount of mutants with each plate D, containing arrestin-mCherry mutant F375A (as reference) . Samples in each well were mixed, at 37 °C for 5 min incubated and for 6 min light activated. Separate 96-well plates were filled with the following samples and processed in parallel in the dark: one 100-yL fraction of each master mix C was combined with 100 yL buffer C or one 60-yL portion of each master mix D with 140 yL buffer which was supplemented with NaCl to yield 492 mM NaCl . All plates were centrifuged and supernatants removed and pellets washed by carefully adding 100 yL buffer C to plates C or 100 yL buffer with 492 mM NaCl to plates D. Dark controls were treated accordingly. Pellets in plates C were resuspended with buffer C and pellets in plates D with buffer containing 492 mM NaCl. Quantification of pulled-down arrestin-mCherry was conducted. Table 2 lists the constructed mutants and includes the number of measurements and thereof derived IC ₅₀ and R ² values as well as 95% confidence intervals.

In-gel fluorescence thermo-stability assay.

Arrestin-mCherry fusion proteins were expressed,

harvested and lysed. The cell lysate from a 50-mL cell- culture fraction was cleared by centrifugation

(Centrifuge 5424R; Eppendorf) at 21,100 x g for 20 min at 4 °C. The lysate was distributed in 100-yL portions to eleven 1.5-mL tubes (Sarstedt) , which were put into a heating block (Dri-Block; Techne) that was equilibrated at 30 °C. The temperature was ramped up to 80 °C manually in 5 °C-steps each 2.5 min. Samples were removed

successively all 2.5 min and cooled down on ice.

Precipitant was removed by centrifugation for 1 h. 12 yL supernatant of each sample were mixed with 3 yL 5x SDS- loading dye. Full-length arrestin-mCherry construct was separated from degraded protein by SDS-PAGE for 1 h at 200 V and 80 mA in MOPS buffer using an 8-12% Bis-Tris gradient gel (Novex NuPAGE; Life Technologies) in sup ^¬ plied chamber (Novex NuPAGE SDS-PAGE gel system; Life Technologies) . Fluorescence-emission of mCherry or mCherry-protein fusions was detected by exciting the protein at 312 or 365 nm using a 605 nm-filter

(ImageQuant RT ECL; GE Healthcare) . Fluorescence

intensities were quantified by ImageJ (NIH) and plotted in Prism. Boltzmann sigmoidal fitting allowed to determine melting temperatures (T _M ) and R values and standard errors. T _M values of wild type, V170A, L173A and R175A derived by described novel in-gel fluorescence thermostability assay were compared with T _M values derived by a standard thermo-shift assay utilizing the

fluorescent dye CPM, N- [4- (7-diethylamino-4-methyl-3- coumarinyl ) phenyl ] malemeide . The in-gel fluorescence thermostability assay is superior to the CPM assay in terms of simplicity: It does not require protein

purification and uses cheaper instrumentation also available in low-budget laboratories.

The following table 2 shows a list of constructed mutants that were screened for half-maximal inhibitory concentra- tion (IC ₅₀ ) of NaCl to disrupt formation of complexes with P-R* . Binding of each mutant to P-R* in its natural environment, the rod outer segment (ROS) membranes, was quantified in 8 different sodium chloride concentrations, ranging from 100 to 2403 mM. The measurement was repeated for the range from 492 to 3949 mM salt if the fitted sigmoidal dose-response curve could not reach the bottom plateau. The number of test sets is indicated from which IC ₅₀ , R ² and 95% confidence interval were derived. It is remarked if expression of functional arrestin protein was too low to determine IC ₅₀ values reliably. The melting temperature of arrestin mutants (T _M ) was determined by above described in-gel fluorescence assay.

Arrestin-l mutants

95% confidence Quality of fit Number Func. expresT _M (arrestin) [°C]

Mutant IC ₅₀ [M]

interval (R ² ) of curves sion level [%] ( ° of measurements)

WT 0.41 0.37 to 0.46 0.9937 74 100 ± 53 64 ± 1 (10)

VI 1 A 0.61 0.56 to 0.67 0.9982 2 50-150 62 ± 1 (2)

F13A 0.84 0.74 to 0.95 0.9974 3 50-150 57 ± 2 (1)

Y25A 0.59 0.51 to 0.71 0.9949 3 50-150 61 ± 0 (1)

V44A 0.51 0.49 to 0.53 0.9994 2 50-150 64 ± 1 (1)

R66A 0.64 0.59 to 0.69 0.9978 3 50-150 59 ± 1 (1)

Q87A 0.54 0.50 to 0.59 0.9974 2 50-150 weak signal

L111A 0.70 0.64 to 0.76 0.9980 2 50-150 61 ± 1 (1)

C143A 0.50 0.48 to 0.51 0.9997 2 50-150 weak signal

G144A 0.57 0.48 to 0.68 0.9921 2 50-150 60 ± 0 (1)

I149A 0.74 0.64 to 0.85 0.9901 3 <20 weak signal

V170A 0.58 0.51 to 0.66 0.9975 2 50-150 60 ± 1 (2)

R171A 0.54 0.47 to 0.61 0.9962 2 50-150 60 ± 1 (2)

L173A 0.51 0.47 to 0.55 0.9983 2 50-150 62 ± 1 (1)

R175A 0.91 0.77 to 1.08 0.9948 3 50-150 53 ± 1 (2)

V247A 0.59 0.55 to 0.62 0.9989 2 50-150 54 ± 1 (1)

R291A 0.56 0.51 to 0.62 0.9967 2 50-150 63 ± 0 (1)

D296A 0.92 0.71 to 1.21 0.9860 3 50-150 59 ± 1 (2)

G297A 1.14 0.82 to 1.75 0.9879 3 50-150 53 ± 1 (1)

K298A 0.52 0.41 to 0.66 0.9906 2 50-150 64 ± 0 (1)

D303A 0.87 0.76 to 0.99 0.9964 3 50-150 55 ± 1 (1)

T304A 0.84 0.74 to 0.96 0.9967 3 50-150 60 ± 1 (2)

A307G 0.64 0.57 to 0.72 0.9967 2 50-150 57 ± 0 (3)

E341A 0.57 0.52 to 0.63 0.9961 2 50-150 63 ± 1 (2)

F375A 1.32 0.94 to 1.95 0.9849 23 107 ± 26 63 ± 1 (3)

V376A 0.66 0.58 to 0.75 0.9957 3 50-150 62 ± 1 (1)

F377A 1.08 0.82 to 1.43 0.9949 3 >150 66 ± 1 (2)

F380A 1.00 0.70 to 1.45 0.9903 3 >150 55 ± 0 (1)

R382A 0.97 0.76 to 1.24 0.9944 4 50-150 62 ± 1 (1)

H10A+F375A 1.54 1.40 to 1.69 0.9967 2 50-150

V11A+F375A 0.99 0.64 to 1.67 0.9805 2 50-150 60 ± 1 (1)

F13A+F375A 1.02 0.91 to 1.16 0.9968 2 50-150

I24A+F375A 1.39 0.95 to 2.02 0.9884 2 <30

Y25A+F375A 1.10 0.92 to 1.31 0.9958 3 50-150

D30A+F375A 1.41 0.87 to 2.34 0.9893 3 50-150

V44A+F375A 1.69 1.07 to 2.80 0.9871 3 <30

P49A+F375A 1.75 0.81 to 4.18 0.9660 3 <30

R56A+F375A 1.45 1.14 to 1.85 0.9968 3 50-150

V57A+F375A 2.17 1.25 to 3.83 0.9743 3 <30

V59A+F375A 1.92 1.29 to 2.96 0.9806 3 <30

A64G+F375A 1.16 0.78 to 1.75 0.9907 3 50-150

R66A+F375A 2.04 1.08 to 3.89 0.9674 3 50-150 59 ± 1 (2)

R80A+F375A 3 <detection

F85A+F375A 1.02 0.65 to 1.71 0.9883 2 50-150

Q87A+F375A 2.01 1.33 to 3.10 0.9890 2 <30

L107A+F375A 1.27 0.81 to 2.00 0.9834 2 <30

L111A+F375A 1.53 0.74 to 3.55 0.9796 2 50-150

D124A+F375A 1.08 0.87 to 1.37 0.9942 2 50-150

C128A+F375A 1.19 1.00 to 1.42 0.9975 2 50-150

C143A+F375A 1.03 1.00 to 1.18 0.9969 3 50-150

G144A+F375A 1.48 0.85 to 2.78 0.9888 3 50-150

V145A+F375A 3 <detection

D146A+F375A 1.53 1.30 to 1.80 0.9916 2 50-150

F147A+F375A 3 <detection

I149A+F375A 2.24 1.62 to 3.09 0.9844 2 <30 50 ± 2 (2)

K163A+F375A 1.19 0.65 to 2.16 0.9866 3 50-150

V170A+F375A 1.06 0.70 to 1.70 0.9866 4 50-150 weak signal (1)

R171A+F375A 1.73 1.29 to 2.34 0.9844 2 50-150

L173A+F375A 1.58 1.34 to 1.87 0.9921 2 50-150

R175A+F375A 2.48 1.55 to 4.21 0.9723 3 50-150

D183A+F375A 0.98 0.67 to 1.43 0.9927 1 50-150

V222A+F375A 0.44 0.39 to 0.49 0.9949 1 <30

A225G+F375A 1.42 0.73 to 3.51 0.9738 2 50-150

V247A+F375A 3 <30

D253A+F375A 1.22 1.00 to 1.50 0.9968 3 50-150

I256A+F375A 1.23 0.83 to 1.84 0.9866 2 50-150

K257A+F375A 1.45 1.11 to 1.91 0.9947 3 50-150

A261G+F375A 1.46 0.84 to 2.74 0.9897 3 50-150

S272A+F375A 1.57 1.35 to 1.84 0.9931 2 50-150

R291A+F375A 1.42 0.91 to 2.40 0.9850 3 50-150 63 ± 1 (2)

D296A+F375A 2.51 1.39 to 4.62 0.9604 2 <30 56 ± 2 (1)

G297A+F375A 1.34 0.84 to 2.20 0.9420 2 <30

K298A+F375A 1.60 0.87 to 3.07 0.9829 4 50-150 61 ± 0 (2)

D303A+F375A 1.46 0.85 to 2.89 0.9841 4 50-150 62 ± 1 (1)

T304A+F375A 1.51 0.69 to 3.34 0.9848 2 <30 58 ± 0 (2)

A307G+F375A 2.83 1.38 to 5.78 0.9569 1 50-150 51 ± 1 (2)

E341A+F375A 2.06 1.53 to 2.80 0.9904 3 50-150 58 ± 1 (3)

H357A+F375A 0.86 0.50 to 1.47 0.9855 2 50-150

N374A+F375A 1.09 0.87 to 1.36 0.9968 2 50-150 58 ± 1 (2)

V376A+F375A 1.21 0.93 to 1.59 0.9899 3 50-150

F380A+F375A 1.39 0.97 to 2.11 0.9874 2 50-150 57 ± 1 (2)

R382A+F375A 1.44 0.93 to 2.27 0.9957 2 50-150 61 ± 1 (1)

H10A+T304A+F375A 1.58 1.13 to 2.38 0.9895 2 50-150

V11A+T304A+F375A 1.29 0.97 to 1.74 0.9888 3 50-150

D30A+T304A+F375A 1.53 1.12 to 2.11 0.9845 2 50-150

R80A+T304A+F375A 1.50 0.81 to 2.75 0.9363 2 50-150

D82A+T304A+F375A 1.32 0.85 to 2.21 0.9811 3 50-150

V90A+T304A+F375A 2.07 1.60 to 2.68 0.9691 2 50-150

^ n L111A+T304A+F375A 1.96 1.33 to 2.96 0.9770 2 50-150

P123A+T304A+F375A 2.02 1.46 to 2.79 0.9887 2 50-150

C143A+T304A+F375A 1.39 1.16 to 1.67 0.9956 2 50-150

G144A+T304A+F375A 1.93 1.09 to 4.68 0.9837 2 50-150

D146A+T304A+F375A 1.94 1.57 to 2.40 0.9907 3 50-150

I149A+T304A+F375A 3 <detection

V170A+T304A+F375A 1.22 1.00 to 1.48 0.9988 2 50-150

L173A+T304A+F375A 1.53 1.07 to 2.22 0.9969 2 50-150

R175A+T304A+F375A 1.51 1.13 to 2.02 0.9784 1 <30

V247+T304A+F375A 1.01 0.57 to 1.89 0.9936 2 <30

D296A+T304A+F375A 1.39 1.06 to 1.85 0.9941 2 50-150

G297A+T304A+F375A 1.25 1.02 to 1.53 0.9975 1 50-150

K298A+T304A+F375A 2.01 1.42 to 2.85 0.9832 3 50-150

D303A+T304A+F375A 2.45 1.50 to 4.26 0.9698 2 50-150

A307G+T304A+F375A 2.89 1.10 to 9.37 0.9446 2 <30

E341A+T304A+F375A 2.75 1.19 to 6.37 0.9481 1 50-150 61 ± 1 (2)

V376A+T304A+F375A 2.14 1.25 to 4.39 0.9822 2 50-150

F380A+T304A+F375A 2.10 1.00 to 5.95 0.9664 2 50-150 59 ± 1 (4)

R382A+T304A+F375A 2.71 1.64 to 4.74 0.9493 2 <30

V11A+A307G+F375A 2.40 1.37 to 4.23 0.9591 1 50-150

R66A+A307G+F375A 3.37 1.20 to 9.48 0.9544 1 <30

L111A+A307G+F375A 2.16 0.76 to 9.30 0.9842 2 <30

P123A+A307G+F375A 2.67 1.65 to 4.32 0.9700 1 <30

I149A+A307G+F375A 1 <detection

R175A+A307G+F375A 1 <detection

V247A+A307G+F375A 3.41 1.48 to 7.88 0.9421 1 <30

R291A+A307G+F375A 2.69 2.01 to 3.61 0.9883 1 <30

D296A+A307G+F375A 2.42 1.96 to 2.99 0.8012 1 <30

G297A+A307G+F375A 3.52 0.71 to 17.37 0.9228 1 <30

V376A+A307G+F375A 2.33 1.25 to 4.51 0.9836 2 50-150

F380A+A307G+F375A 1.98 1.04 to 3.88 0.9062 2 50-150

R382A+A307G+F375A 3.11 1.29 to 7.52 0.9290 1 50-150

VI 1 A+E341A+T304A+F375A 2.10 0.87 to 5.62 0.9660 3 50-150 60 ± 1 (1)

R66A+E341A+T304A+F375A 2.59 1.07 to 6.74 0.9507 2 50-150 56 ± 1 (1)

R171A+E341A+T304A+F375A 2.95 1.23 to 7.85 0.9405 2 50-150 59 ± 0 (1)

R291A+E341A+T304A+F375A 2.10 1.05 to 4.31 0.9591 3 50-150 61 ± 1 (1)

K298A+E341A+T304A+F375A 2.24 1.14 to 4.47 0.9741 3 50-150 61 ± 1 (1)

D303A+E341A+T304A+F375A 2.91 0.85 to 9.99 0.9105 1 50-150 59 ± 1 (1)

R382A+E341A+T304A+F375A 2.48 1.26 to 4.92 0.9714 3 50-150 56 ± 1 (1)

VI 1 A+F380A+T304A+F375A 1.63 1.14 to 2.33 0.9852 3 50-150

R66A+F380A+T304A+F375A 2.64 0.80 to 8.84 0.9603 2 50-150 58 ± 0 (1)

F85A+F380A+T304A+F375A 1.78 0.96 to 3.62 0.9726 3 50-150 61 ± 1 (1)

R171A+F380A+T304A+F375A 2.07 0.75 to 6.59 0.9408 3 50-150 56 ± 1 (1)

R291 A+F380A+T304A+F375A 1.92 0.90 to 4.35 0.9714 2 50-150 61 ± 0 (1)

K298A+F380A+T304A+F375A 1.37 1.02 to 1.87 0.9868 2 50-150 61 ± 0 (1)

D303+F380A+T304A+F375A 1.94 0.94 to 4.03 0.9678 2 50-150 60 ± 1 (1)

E341A+F380A+T304A+F375A 2.00 0.91 to 5.39 0.9598 2 50-150 57 ± 0 (1)

Table 2: List of constructed mutants that were screened for half-maximal inhibitory concentration ( I C ₅₀ ) as also shown in Figure 4.

In the attached sequence protocol, the following

relationships apply: arrestin-1 = SEQ ID No:l, arrestin-2 = SEQ ID No:2, arrestin-3 = SEQ ID No:3, and arrestin-4 = SEQ ID No : 4.