Multiple Sequence Alignment

6.2.2.1 Prediction of Transmembrane Helices

Even though ADRB1 and ADRB2 seem to be the best structural templates to build the backbone of the 5-HT2C receptor from their 3D coordinates, the four crystal

Table 6.1 Calculated RMSD between the backbone of the four crystal structures of GPCRs

Bovine rhodopsin

Human pi adrenergic receptor

Meleagris p2

adrenergic receptor

Human A2A

adenosine receptor

Bovine rhodopsin Human b1 adrenergic receptor Meleagris b2 adrenergic receptor Human A2A adenosine receptor

5.14

o o o structures are highly interesting in order to carry out a 3D prediction of TM segments in the 5-HT2C receptor. Indeed a profile alignment between a multiple sequence alignment, inherited from the structural superimposition of the four 7-TM bundles crystal structures, and the sequence of the human 5-HT2C receptor has been automatically produced (Thompson et al. 1994) and manually adjusted to delete gaps within the TM regions (Fig. 6.2). This permits one to emphasize the consensual delimiting positions of helices in the 5-HT2C sequence. In a parallel manner, bioinformatics methods have been used to predict the TM regions (Bray and Goddard 2008) on the assumptions that the outward facing sections of the TM helices must be hydrophobic because they are in contact with the hydrocarbon tails of the lipid bilayer and that the hydrophobic center of each helix should be at the center of the membrane (Donnelly et al. 1994).

The Table 6.2 summarizes the different TM predictions resulting of previous works based on RHO crystal structure as the only template or on the previously evoked bioinformatics method in comparison with the 3D-based prediction from the closest structural templates, ADRB1 and ADRB2. Sequence homol-ogy percentages within TM regions of RHO, ADRB1, ADRB2, and A2AAR, deduced from the TM prediction, are 44, 62, 62, and 53, respectively, and support the idea that TM regions of ADRB1 and ADRB2 are the best templates. In a general manner, the length of TM helices from the RHO-based prediction by Zuo is shorter than those of the other predictions. It appears that RHO-based prediction by Farce et al., Rashid et al., and the ADRB-based method apply the full-length 3D information of helices from crystal structures contrary to the other RHO-based method whereas all keep the same TM center. The most specific trait of ADRB-based prediction is the fourty percent longer C-terminal region of helix 5 compared with other TM5 of other predictions. Otherwise, even though 2D-based method shows a shift of the center in TM 4 and 6 toward their C-terminal region, all predictions seem to be homogeneous and independently provide crucial insights on ligand binding and activation mechanisms within the 5-HT2C receptor. If the seven TM helices are widely discussed for ligand binding and receptor activation, a small eighth helix exists in the C-terminal region of all crystal templates with a strictly conserved arginine residue, which is consequently labeled as Arg8.50. These regions present a good sequence alignment with the sequence of 5-HT2C and belong to segments further modeled as conserved regions.

6.2.2.2 Prediction of Common Structural Features

All reported discussions should be read while inspecting the multiple sequence alignment (Fig. 6.2). Important conserved features between the crystal structure and the 5-HT2C sequences are reported here using the Ballesteros and Weinstein numbering system to designate the more conserved residue in each helix as X.50, where X is the TM helix number, and the 5-HT2C numbering for loop regions.

Fig. 6.2 Alignment of human 5-HT2C receptor and crystal template sequences with the Jalview sequence editor tools. The predicted TM helices of 5-HT2C (human_5-HT2C_TM_pred) are illustrated as red bars below the alignment and result from the delimiting positions of TM segments in crystal templates. Homology rates are rendered for residues overlined with graduated blue levels, whereas significant residues discussed in the text are labeled according to the Ballesteros and Weinstein numbering system to designate the most conserved residue in each helix as X.50, where X is the helix number (For interpretation of the colors in this figure, the reader is referred to the web version of this chapter)

Fig. 6.2 Alignment of human 5-HT2C receptor and crystal template sequences with the Jalview sequence editor tools. The predicted TM helices of 5-HT2C (human_5-HT2C_TM_pred) are illustrated as red bars below the alignment and result from the delimiting positions of TM segments in crystal templates. Homology rates are rendered for residues overlined with graduated blue levels, whereas significant residues discussed in the text are labeled according to the Ballesteros and Weinstein numbering system to designate the most conserved residue in each helix as X.50, where X is the helix number (For interpretation of the colors in this figure, the reader is referred to the web version of this chapter)

Table 6.2 Predictions of TM segments (length within brackets) of the human 5-HT receptor

TM Predicted TM region

Prediction method

Authors

1 DGVQNWPALSIVIIIIMTIGGNILVIMAVSM (31) RHO

based RHO based RHO based Hydrophobic profile

PDGVQNWPALSIVIIIIMTIGGNILVIMAVSM (32) ADRB based

GVQNWPALSIVIIIIMTIGGNILVIMAVSM (30) NWPALSIVIIIIMTIGGNILVIMAV (27) GVQNWPALSIVIIIIMTIGGNILVIMAVSME (31)

ATNYFLMSLAIADMLVGLLVMPLSLLAI (28) ATNYFLMSLAIADMLVGLLVMPLSLLAILY (30) FLMSLAIADMLVGLLVMPLS (20) TNYFLMSLAIADMLVGLLVMPLSLLAILYD (30)

RHO based RHO based RHO based Hydrophobic profile

NATNYFLMSLAIADMLVGLLVMPLSLLAILY (31) ADRB based

3 RYLCPVWISLDVLFSTASIMHLCAISLDRY (30) LPRYLCPVWISLDVLFSTASIMHLCAISLDRYVAI

VWISLDVLFSTASIMHLCAISLDRY (25) RYLCPVWISLDVLFSTASIMHLCAISLDR (29)

LPRYLCPVWISLDVLFSTASIMHLCAISLDRYVAIR

4 AIMKIAIVWAISIGVSVPIP (20) RTKAIMKIAIVWAISIGVSVPIP (23) TKAIMKIAIVWAISIGVSVPIPVI (24) AIMKIAIVWAISIGVSVPIPVIGL (24)

SRTKAIMKIAIVWAISIGVSVPIPV (25)

5 DPNFVLIGSFVAFFIPLTIMVITYC (25) DPNFVLIGSFVAFFIPLTIMVITYC (25) NFVLIGSFVAFFIPLTIMVIT (21) FVLIGSFVAFFIPLTIMVITYCLTIY (26)

RHO based RHO based

RHO based Hydrophobic profile ADRB based

RHO based RHO based RHO based Hydrophobic profile ADRB based

RHO based RHO based RHO based Hydrophobic profile

DPNFVLIGSFVAFFIPLTIMVITYCLTIYVLRRQ (34) ADRB based

ERKASKVLGIVFFVFLIMWCPFFITNILSVL (32) RHO based INNERKASKVLGIVFFVFLIMWCPFFITNILSV (33) RHO based NERKASKVLGIVFFVFLIMWCPFFITNI (28) RHO based

LGIVFFVFLIMWCPFFITNILSVLCE (26) Hydrophobic profile

NERKASKVLGIVFFVFLIMWCPFFITNILSVL (32) ADRB based

(Rashid et al.

2003) (Farce et al. 2006) (Zuo et al. 2007) (Bray and Goddard 2008) (Renault et al. 2010) (Rashid et al. 2003) (Farce et al. 2006) (Zuo et al. 2007) (Bray and Goddard 2008) (Renault et al. 2010) (Rashid et al. 2003) (Farce et al. 2006)

(Zuo et al. 2007) (Bray and Goddard 2008) (Renault et al. 2010) (Rashid et al. 2003) (Farce et al. 2006) (Zuo et al. 2007) (Bray and Goddard 2008) (Renault et al. 2010) (Rashid et al. 2003) (Farce et al. 2006) (Zuo et al. 2007) (Bray and Goddard 2008) (Renault et al. 2010) (Rashid et al. 2003) (Farce et al. 2006) (Zuo et al. 2007) (Bray and Goddard 2008) (Renault et al. 2010)

(continued)

Table 6.2 (continued)

Prediction

TM Predicted TM region

method

Authors

7 EKLLNVFVWIGYVCSGINPLVY (22)

RHO based

(Rashid et al. 2003)

KLLNVFVWIGYVCSGINPLVYTLF (24)

RHO based

(Farce et al. 2006)

KLLNVFVWIGYVCSGINPLVYTLPN (25)

RHO based

(Zuo et al. 2007)

EKLLNVFVWIGYVCSGINPLVYT (23)

Hydrophobic

(Bray and Goddard

profile

2008)

MEKLLNVFVWIGYVCSGINPLVYT (24)

ADRB

(Renault et al.

based

2010)

Overall Structure

Inspection of ADRB2 crystal (Cherezov et al. 2007; Rosenbaum et al. 2007) enables us to perceive the main interactions responsible for the maintaining its architecture and occurring in helical regions or some connecting loop regions conserved in the 5-HT2C sequence. Starting with helix 1 and 2, the side chain of the more conserved residue, Asn1.50, is included in a hydrogen bond with the carbonyl group of Ser7.46 whereas Asn2.40 hydrogen bonds with Tyr7.53. At the top of helix 3, Cys3.25, strictly conserved within the GPCR family, forms a disulfide bridge with a conserved cysteine residue (Cys127 in 5-HT2C) of the extracellular loop 2 (EL2). Near this cysteine, Trp3.28, conserved within the GPCR-type monoamine receptors, 5-HT2C and ADRB, forms a p-p stacking interaction with the conserved tryptophan residue (respectively, numbered 99 and 120 in ADRB2 and 5-HT2C) of the adjacent EL1. This interaction would restrain EL1 from pulling down toward the TM bundle and making the top of helices 2 and 3 closer. At the bottom of TM3, all of the mammalian monoamine GPCRs have a highly conserved DRY (ERY in rhodopsin) pattern that is involved in the receptor activation (Ballesteros et al. 2001; Jensen et al. 2001). Indeed, in RHO crystal, Arg3.50 is trapped by two salt bridges between Asp3.49 and Glu6.30, which would maintain the receptor in an inactive state. Previous studies on the G-coupled lutropin receptor (LHR) have shown that breaking of the charge-reinforced H-bonding interaction between Arg3.50 and Asp6.30 would increase the solvent accessibility of the cytosolic extensions of helices 3 and 6, which probably induce rotation of helices 3 and 6 to get drastic changes in the receptor-G protein interface (Angelova et al. 2002; Greasley et al. 2002; Zhang et al. 2005). In ADRB2 crystal, the salt bridges are disrupted by the presence of a sulfate ion, which is a stronger counterion than glutamate. The more conserved residue in helix 4, Trp4.50, also seems to be stabilizing for the receptor architecture since it is implied in an interhelical interaction by hydrogen bonding with Ser2.45 (Asn in RHO) of TM2, itself binding with the polar residue in position 3.42 (respectively, Ser, Thr, and His in RHO, ADRB2, and 5-HT2C), whereas the reference residue in helix 5, Pro5.50, introduces a kink whose the angle could be decreased during the receptor activation (Crozier et al. 2007). Other conserved features in TM 3, 5, 6, and 7 are either hydrophobic clusters that maintain the stacking of the interior core of the protein and cover the outward section in order to interact with the lipid tails of the membrane or amino acids directly involved in the ligand binding, which will be discussed in Sect. 6.1.2.2.2.

The more recent rhodopsin crystal structures, available in the Protein Data Bank (Berman et al. 2000) under 1GZM (Li et al. 2004), 1L9H (Okada et al. 2002), and 1U19 (Okada et al. 2004) entries, have also revealed structural waters participating in an extensive hydrogen-bond network (Pardo et al. 2007) between conserved residues of helices 1, 2, 6, and 7, which are amenable to play a critical role in activation mechanism in a similar manner to that described above for the suggested impact of the DRY sequence in TM3 and resulting in a disrupted network of hydrogen bonds. This phenomenon was also analyzed in ADRB2 crystal structure taking care to decompose the examination of the extensive hydrogen bond network in order to reveal substantial differences with RHO-focused inspection (Fig. 6.3). The first such difference concerns the conformation of Tyr7.53, which is oriented toward one additional water included in the network of hydrogen bonds between Asn1.50 and Asn7.49 rather than toward the bottom of the helix bundle. The displacement of Tyr7.53 disrupts the p-p stacking interaction between side chains of Tyr7.53 and Phe7.60 (Tyr in 5-HT2C) as well as a hydrogen bond with Asn2.40 and could suggest that the receptor is not in complete inactive state anymore, if we take into account that Tyr7.53 is strictly conserved in GPCRs and that ADRB2 is cocrystallized with carazolol, an inverse agonist. The second variation is a gain of one water-relayed hydrogen bond between helices 1 (Gly1.61) and 7 (Trp7.40). What is interesting in modeling GPCR-type monoamine receptors is the conserved network between 7.38, 6.47, and 6.51, which is strictly conserved within monoamine receptors and directly interacting with the ligand in ADRB2. On the assumption that rotation of TM6 could lead to the activation of the receptor, specific interactions of the ligand with this aromatic residue would be one of the starting events to switch off the hydrogen-bond signaling along the TM6.

Fig. 6.3 Hydrogen bond network relayed by structural waters in ADRB2. The extensive hydrogen bond network is decomposed into three pictures. Interacting amino acids are represented in the transparent TM bundle as sticks depicted by atom type, whereas structural waters are illustrated as spheres and hydrogen bonds, as dashed lines

Fig. 6.3 Hydrogen bond network relayed by structural waters in ADRB2. The extensive hydrogen bond network is decomposed into three pictures. Interacting amino acids are represented in the transparent TM bundle as sticks depicted by atom type, whereas structural waters are illustrated as spheres and hydrogen bonds, as dashed lines

Ligand-Binding Site

In crystal structure of ADRB2, the cocrystallized ligand carazolol is caged in the top of helix bundle between TM 3, 5, 6, and 7 and EL2. The amino acids directly involved in the binding of ligand or the pocket plasticity and conserved in 5-HT2C receptor enable a prediction of a coarse fingerprint of interaction between serotonin and this receptor. This is supported by the neighbouring chemical structure of the two compounds since both include a benzyl group, itself in aromatic system (respectively, indol and carbazol groups in serotonin and carazolol), connected to a protonated nitrogen through a three-atom (serotonin) or four-atom (carazolol) spacer. The protonated amine of the ligand is bound by an electrostatic interaction with Asp3.32 and one hydrogen bond with Tyr7.43, two specific residues for ligand binding among all mammalian biogenic amine receptors (Bywater RP 2005). Moreover, Asp3.32 was found to anchor the terminal amine moiety of serotonin in 5-HT2A (Kristiansen et al. 2000). Conserved amino acids of ADRB2 TM5 in the vicinity of ligand position are Ser5.43 (first suggested to interact with the NH of the indole in serotonin (Ebersole et al. 2003) and then strongly suspected to bind to the 5-OH of serotonin (Braden and Nichols 2007)) and Phe5.47 (found to have significant interactions with serotonin in 5-HT2A (Shapiro et al. 2000) and more generally with agonist ligands (Salom et al. 2006)). However, inspection of ADRB2 crystal structure shows that there is no contact between Phe5.47 and the inverse agonist ligand but that it forms a p-p stacking interaction with Phe6.52, itself binding the ligand by an edge-to-face p-p interaction as well as Phe6.51, another strictly conserved residue within the monoamine receptors found to affect the binding of serotonin while mutated to a leucine residue (Choudhary et al. 1993). This suggests a different binding mode between the 5-HT receptor subtypes or between agonist and antagonist ligands.

Otherwise, the strictly conserved Trp6.48 in the GPCR family, previously described in the extended hydrogen-bond network going across the ADRB2 receptor, was found to cause an almost 1,000-fold decrease in serotonin binding in 5-HT2A (Roth et al. 1997). The ligand binding specificity associated with these conserved structural features within GPCR-type biogenic amine receptors must be distinguished from the ligand binding selectivity for receptor subtype of this GPCR subfamily. In this way the extracellular loop 2 (EL2) connecting helices 4 and 5 is suspected to play a crucial role from inspection of crystal structure and site-directed mutagenesis (Zhao et al. 1996; Wurch and Pauwels 2000). This is supported by the idea that EL2 has a very variable length and a low sequence homology, suggesting that it could partially determine the strict selectivity of a GPCR-type receptor. In addition to the idea that EL2 contributes to a hydrogen-bonding network that is thought to maintain rhodopsin in an inactive state (Klco et al. 2005), the examination of the two crystal structures of RHO and ADRB2 shows the propensity of EL2 to act as a gate for ligand accessibility. Moreover, chimeras of a1A and a1B adrenoceptors, built from the reciprocal exchange of three consecutive residues in EL2 (Zhao et al. 1996) as well as a single substitution in the canine 5-HT1D sequence shifting toward human 5-HT1D sequence (Wurch and Pauwels 2000), have induced the permutation of ligand selectivity within the two couples of engineered receptors. The lack of sequence homology and the variable length of EL2 render this loop very challenging for homology modeling of every GPCR.

Defeat Drugs and Live Free

Defeat Drugs and Live Free

Being addicted to drugs is a complicated matter condition that's been specified as a disorder that evidences in the obsessional thinking about and utilization of drugs. It's a matter that might continue to get worse and become disastrous and deadly if left untreated.

Get My Free Ebook


Post a comment