CB1 and cholesterol

A previous post postulated that some cannabinergic effects of terpenes on CB1 have to do with similarities of cholesterol and terpenes.

The CB1 structure diagram is from an excellent review by Dr. Patrica Reggio and others [1]. A good part of this review was devoted to modulation of CB1 by phosphorylation of the intracellular loops. These phosphorylation sites are pink. Amino acids with negatively charged side chains, which may mimic phosphorylation are colored blue. [1] Other residues discussed are highlighted in orange. Additional information from other sources has been added. The circles represent amino acids with single letter codes. The lay reader may look up the full names form these single letter codes as well as the amino acid properties on the proteinstructure.com site.

1. F3.36, is a phenylalanine in the 36^th amino acid in the 3^rd trans membrane helix that also extends into the cytoplasm the cell. The orange circles “YRC” that have been air brushed with red mark the start of the official intra cellular region. W6.48, and F3.36 are part of the switch that turns the receptor on and off.
2. W6.48 is part of the switch
3. D6.30, a glutamate forms a hydrogen bond with R3.50

More structural details are contained in the Reggio review.

Cholesterol binding domains

CARC is an acronym for a cholesterol amino acid recognition consensus sequence. Going from the N-terminus we have the following requirements [2]

1. a branched chain apolar amino acid like leucine or valine
2. a 1-5 amino acid spacer that can be anything
3. a tyrosine (Y)
4. another anything goes 1-5 amino acid spacer
5. a basic amino acid, i.e. lysine (K) or arinine ( R )

The CRAC domain is essentially an inverted CARC domain except that #3 can be a tyrosine or a phenylalanine. [1]. In this diagram the sequence runs vertically along the cartoon. The single amino acid codes of the 1-5 variable residues run horizontally. The 7^th trans membrane helix has both a CRAC and CARC domain. whereas the .4^th trans membrane helix has only a CRAC domain going from intracellular to extracellular: KAWAFCL.

N-acarchidonyl ethanolamide(AEA) revisited

AEA is thought to first anchor to branched aliphatic residues V351-I354. Molecular dynamic simulations in a phoshatidylcholine bilayer suggested that insertion initiated with a CWG motif located on M6. [3] This region is indicated by lime arrows in the CB1 receptor diagram. Note the presence of other branched aliphatic side chains as well as leucine. These residues are good for dissolving the iso octyl portion of cholesterol. Studies indicated that the agonist-induced receptor activation is associated with a modification of the respective orientations of TMH3 and TMH6 which results in the separation of two aromatic residues phenylalanine F200 and tryptophan W356. These side chains engage in aromatic π-π interactions in the inactive and antagonist bound states. [3] The DiScala review suggested that cholesterol be included in in silico molecular dynamic studies. [3] This review is mentioned in this post because it covers the complex dynamics of getting anandamide into the active site that includes being dissolved into a cholesterol rich micro domain in which CB1 resides and being carried to theat micro domain by albumin. [3]

Revisiting the CRAC and CARC domains

The concept of CARC/CRAC domains is brilliant in its simplicity, yet it does not seem to intuitively hold up in the three dimensional protein in which near by side chains need not necessarily come from adjacent regions in the linear, one dimensional structure. The method was the following

1. Search RSB.org for entries containing cholesterol and CB1
2. View the structures in NGL (Web GL) mode
3. Go to “ligand view” rather than the default of “structure view.”
4. screen capture images and record amino acids that are forming hydrogen bonds blue), hydrophobic interactions (gray), and Pi interactions (orange, green).,
5. Record amino acid residues involved with interaction with cholesterol.

AM11542

In the case of this structure histidine H154 fulfills the role of a K or an R. The ironic thing is that a carbonyl oxygen is forming an H-bond with the H on the -OH of cholesterol. The toggle switch W241 is forming hydrophobic contacts with the cholesterol ring system, A198 and I245 are interacting with the iso-octyl group.This structure also contains the agonist AM11542 and flavin mononucleotide (FMN).

AM841

Moving on to a similar structure, we have H152 instead of H154 in the general vicinity of the cholesterol -OH group. It is a carbonyl group again making the H-bond. W241 is again in the crystal structure interacting with the ring structure of cholesterol. I245 and L165 are making contacts with the iso ocytyl group.

Allosteric modulator Org27569

Just skimming PubMed abstracts Org27569 increases the affinity of normal orthosteric ligands while decreasing the Gi coupling that leads to the shut down of adenylyl cylase. One the other hand Org27569 increases alternate pathway ERK activation. According to RCSB 7V3Z these amino acids interact with cholesterol. allosteric modulator.

6N4B

This rcsb.org entry is of the human CB1 receptor docked to the heterotrimeric Gi complex, two cholesterol molecules and the highly potent agonist MDMB-Fubinaca (FUB). It is interesting to note that lipids other than cholesterol were not included in this structure.

The entire ribbon structure of heterotrimeric G1 docked to CB1 is shown for orientation. two different “ligand interaction” images are shown for what appears to be the same two stacked cholesterol molecules. The nitrogen of K232 does not appear to be forming a hydrogen bond with the-OH group of either cholesterol.

Is CARC/CRAC motif holding up?

This image was created with the Uniprot sequence for human CB1. A new line was started for each domain. Instead of dividing the domains into elements of secondary structure, UniProt divided the domains in relation to the cell membrane. The recorded interacting amino acids from the four CB1 crystal structures examined have been indicated.

The CRAC domain in H4 appears to be holding up, albeit not in the canonical motif. It is hard to say if these cholesterol binding sites would exist in a lipid bilayer with phospholipids that would likely influence the structures of the entire protein.

What this means for terpenes

In the previous post the possibility of terpenes interacting directly with CB1 was presented. Like cholesterol, terpenes are products of geranyl pyrophosphate. In examining canonical cholesterol binding motifs and actual crystal structures, there seems to be a certain amount of flexibility. The -OH of cholesterol does not have to be stabilized by hydrogen bonding to a lysine or arginine. Interestingly no pi bond interactions were indicated by the software on rcsb.org.

The 3D structures of these compounds are shown to illustrate that these compounds are not planar and that they have a certain degree of conformational flexibility. This post has not extensively covered the AEA review, but the same logic may apply to terpenes. If a terpene is highly soluble in the lipid microdomains containing CB1, It seems conceivable that these terpenes could interact with cholesterol or other sites.

References

1. Al-Zoubi, R., Morales, P., & Reggio, P. H. (2019). Structural Insights into CB1 Receptor Biased Signaling. International journal of molecular sciences, 20(8), 1837. PMC free article
2. Fantini, J., & Barrantes, F. J. (2013). How cholesterol interacts with membrane proteins: an exploration of cholesterol-binding sites including CRAC, CARC, and tilted domains. Frontiers in physiology, 4, 31. PMC free article
3. Di Scala, C., Fantini, J., Yahi, N., Barrantes, F. J., & Chahinian, H. (2018). Anandamide Revisited: How Cholesterol and Ceramides Control Receptor-Dependent and Receptor-Independent Signal Transmission Pathways of a Lipid Neurotransmitter. Biomolecules, 8(2), 31. PMC free article