doc1
Adenosine deaminase (PDB ID: 1WXY) from Bos taurus
Created by: Lillian Tan

Adenosine exerts multifaceted roles both inside and outside the cell. For example, the ribose group forms high-energy phosphodiester bonds in ATP, while its purine group acts as a generally inhibitory messenger received by ionotropic P2X and metabotropic P2Y receptors (figure 1, figure 2) (1). As with all messengers and metabolites, regulation is essential to maintaining homeostasis. The enzyme adenosine deaminase (PDB ID: 1WXY, abbreviation ADA) hydrolyzes adenosine to produce inosine, removing the amino group with the aid of water (figure 3) (2). An autosomal recessive disorder resulting in ADA deficiency leads to severe mental and physiological impediments (3). More generally, adenosine buildup is correlated with inflammation (2). Scientists and pharmacologists are interested in consulting the structure of ADA to develop drugs that augment or inhibit the enzyme’s activity through binding to active or allosteric sites.

ADA is classified as a hydrolase and has a molecular weight of 478.74 amu and an isolelectric point (pI) of 5.52 (4). There are 349 residues in the primary structure (5). The secondary structure includes ten α-helices and nine 310-helices, as well as a parallel β-sheet with eleven strands twisted into a barrel shape enclosing the active site. The right-handed crossovers of the parallel strands consist of βαβ motifs. ADA has a tertiary structure with helices wrapping the β-sheet, which in turn wraps around the core of the enzyme to produce an isolated environment for stabilizing the transition state of the deamination reaction. The helices on the exterior are likely amphiphilic, with the hydrophilic face exposed to the solvent and the hydrophobic face interacting with the β-sheet. The β-sheet, on the other hand, likely has a hydrophobic side facing the helices and a side with specialized residues suitable to recognizing and interacting with adenosine substrates.

The active site of ADA is named S0 (hydrophilic), and is surrounded by F0, F1 and F2 subsites (hydrophobic) that were analyzed during crystallography for the potential binding of inhibitors (2). These sites are organized such that the F sites are closer to the exterior of the protein, while S0 is larger and deeper within the β enclosure. The β-strand backbone of residues Leu-182 to Asp-185 provides a structural gate for the S0 site. The active site holds six water molecules (W0 – W6), and the heterocyclic purine and ribose groups of adenosine likely favor forming hydrogen bonds with these waters, which are held almost in one plane by several residues. Glu-186, Ser-265, and Thr-269 bind two off-site water molecules that form hydrogen bonds with W1 and W2 in the S0 site. W1 and W2 also form hydrogen bonds with Glu-217, which is likely in its negatively-charged carboxylate form since the S0 site is hydrophilic and the surrounding pH is greater than 5. Similarly, the basic form of Asp-296 interacts with W3 and W4. The most important water molecule is W4, as it both reacts with the substrate and dictates the open or closed conformation of ADA. W5 uniquely hydrogen-bonds with the -NH- backbone of Gly-184, and W6 is held by Asp-19 and His-17. The latter residue, His-17, is made electron-rich from the to coordination to zinc2+, and therefore finds stability in interaction with W6. In addition, the hydrophobic F0 site interacts with three water molecules (W7-W9), and subsites F1 and F2 weakly associate with two waters each (W10-W13). Zinc2+ is the critical ligand for this enzyme as the positive charge helps create a potent hydroxide nucleophile out of W4 to initiate the reaction. W0 coordinates to zinc from a distance of 2.1 Å. His-17 and two other histidine residues (unnumbered) also coordinate this cofactor to the active site.

The mechanism of deamination involves the zinc ion activating the W4 water, held by hydrogen bonds to Asp-296, by making the water more acidic since it acquires a partial negative charge (figure 4) (6). His-238 then extracts a proton with the imidazole group, creating an OH- nucleophile that subsequently attacks the 2-carbon on the purine ring. The aromaticity of adenosine breaks in a concerted step when the ring nitrogen closest to the aniline carbon receives a proton from the nearby Glu-217. The Asp-296 now forms hydrogen bonds to stabilize the OH group. This lowers the energy of the transition state, which features a partially formed C-O bond, a partially broken π-bond, and a partially formed N-H bond. In effect, ADA decreases the activation energy of the hydration slow step by stabilizing the transition state more than it stabilizes the substrate adenosine. Next, the returning aromaticity due to Glu-217 taking back its proton drives the NH2 group of the aniline to disassociate from the pyridine ring as a nucleophile and take off the His-238 proton, producing NH3. The deamination is predicted to be a faster step than hydration, and the product inosine is the inactive form of adenosine, with a hypoxanthine instead of a purine attached to ribose (figure 3).

An occupied W4 site supports the open conformation since the nearby Phe-65 residue must move away from W4 due to the hydrophobic nature of the phenyl group. Phe-65 is attached to the α-helix of residues Thr-57 to Ala-73, and this helix hovers over the core of the β-barrel such that the movement of phenyalanine towards W4 would induce the entire helix to block the active site, resulting in the closed conformation. The Phe-65 helix is thus the structural gate of the enzyme, and several inhibitory drugs target the prevention of water from occupying W4 (2).

A study of the adenosine mimic 6-hydroxyl-1,6-dihydroxypurine riboside (HDPR) found that the compound completely occupies the S0 site such that all water molecules W0-W6 would be excluded (figure 5) (2). Adenosine substrates can no longer bind to the active site. Moreover, HDPR shuts the structural gate as Leu-58 and Leu-62 on the helix move into the respective F1 and F2 hydrophobic sites. A variety of other purines have been synthesized and tested in vitro and in vivo, including erythro-9-(2-hydroxy-3-nonyl) adenine hydrochloride (figure 6, EHNA) (7). These competitive inhibitors have generally been validated to reduce ADA activity. However, purine compounds that mimic adenosine may have undesired side effects, and a variety of novel amphiphilic molecules were synthesized to allosterically inhibit ADA. Research has shown that the phenyl groups of complexes like FR104783 interact at the F0 site (figure 7) (2). A carbon or nitrogen atom on such inhibitors also replaces the W4 water, although it seems that the structural gate may not close due to steric encounters with the large inhibitor. Additional pharmacological considerations include the KI (inhibitor equilibrium constant) and stability of these amphiphilic molecules. Their aromatic hydrocarbon groups increase the energy of the hydrophilic S0 site, but nonpolar groups are required for gaining access to the F0, F1, and F2 sites. The design of inhibitors is thus further complicated by the necessity to control the kinetics and thermodynamics of inhibition, and the tertiary structure of ADA poses many challenges for researchers seeking the perfect inhibitor.

Adenosine deaminase is evolutionarily conserved, and many studies arise from ADA extracted from Bos taurus and Mus musculus. To study the sequence homology, PSI-BLAST compares a protein’s primary structure with that of other proteins. A “gap” on BLAST indicates a mismatch between subject and query amino acids, and the greater the number of gaps, the higher the E value. An E value closer to zero indicates strong similarities in amino acid sequences. Therefore, when choosing comparison proteins, the upper boundary for consideration is E < 0.05. Another useful tool is the Dali Server, which compares tertiary structures by analyzing intramolecular distances and geometries. When a Z-score greater than 2 is obtained for a comparison of two structures, this would indicate that the proteins have significant resemblances in 3D conformation. 

Adenosine deaminase from Mus musculus (PDB ID: 3mvt) yielded an E value of 0.0 < 0.05 and a Z score of 60.8 > 2 (8, 9). A superposition of the mouse and cow ADA suggests that most residues of this hydrolase are conserved during the evolution of mammals as the tertiary structures are nearly identical. In fact, the sequence of mouse ADA almost exactly matches that of cow ADA, and thus the very same secondary structures arise with the helices and sheets wrapping the active site. Nonetheless, the Mus musculus ADA was crystallized with two identical subunits joined together in the quaternary structure and therefore catalyzes two adenosine deaminations at once (10).

Cytosine deaminase (PDB ID: 3O7U, abbreviation CDA) from E. coli also had a low E value (2 x 10-77 < 0.5) and a high Z value (22.4 > 2) when compared with ADA from Bos taurus, but this enzyme has a slightly different function: hydrolyzing cytosine (figure 8) (8, 9). The primary structure of CDA is noticeably different from that of ADA due to the misalignments when the two enzymes are superimposed. The components of secondary structure of CDA is similar to ADA, consisting mostly of α-helices and β-sheets, with the β-sheets wrapping around the active site. However, there is a greater presence of random coils in CDA. The active site at the center of CDA likely has a different heterocyclic ring recognition environment compared with that of the ADA active site since cytosine does not have a ribose group like adenosine does (cytosine is a base while cytidine is the nucleoside counterpart). Nonetheless, crystallographers reported a similar mechanism of hydrolysis involving a histidine, a glutamate, and an aspartate as the three catalytic residues (11). These key hydrolase residues are likely conserved from an early point in evolution (before eukaryotes) due to their optimal efficiency in the deamination reaction. While the structure of ADA has been well elucidated, the best synthetic enzyme regulators and the role of this indispensible hydrolase throughout evolutionary history have yet to be extensively elucidated.