Ribonucleotide Reductase
Created by Jennifer Yi
Ribonucleotide reductase class Ib (pdb ID= 2BQ1) is an enzyme that reacts with ribonucleotides to produce deoxyribonucleotides, which are precursors that are used to repair and synthesize DNA in Salmonella typhimurium. Ribonucleotide reductase (RNR) has a molecular weight of 233585.08 daltons and an isoelectric point (pI) of 5.74. RNR uses a long electron radical transfer reaction to reduce the ribonucleotides to deoxyribonucleotides. This reaction requires an organic free radical that is generated and stored, and a pair of redox-active cysteine that provides the reducing power. RNR can be classified into three main classes called class I, class II, and class III based on how they generate the thiol radical from cysteine during the reaction (4). Class I RNR is subdivided into two groups called class Ia and class Ib. The major difference between class Ia and class Ib is that class Ib lacks the allosteric activity site. Without the allosteric activity site, different amounts of the four deoxyribonucleotides are produced, which can be toxic to the cell. Therefore, the RNR from Salmonella typhimurium is not biological active, and class Ib RNR is only found in prokaryotes. However, this enzyme is biologically active in Mycobacterium tuberculosis, which is responsible for the major global disease. Since the bacterium is becoming more resistant to the known drugs and treatments, RNR has become an anti-proliferating potential drug target (3).
RNR is a heterotetramer with four subunits that are made up with two different homodimers. The large dimer subunit called R1E and the small dimer subunit called R2F are encoded by the nrdE and nrdF genes respectively. In addition, the R1E subunits are made up of 44% alpha helices and 12% beta sheets, and the primary structure for each subunit is 714 amino acids long. The R2F subunits are made up of 64% alpha helices, and the primary structure for each subunit is 319 amino acids long (6). The R1E subunits reduce the ribonucleotide substrate at the active site, and it also contains the effector binding site. The R2F subunits generate the tyrosyl radical, which is required for catalysis, and it also contains the binding site for iron to stabilize the radical. The R1E subunits consist of three different domains: a ten stranded alpha-beta barrel, a small alpha-beta domain, and an N-terminal domain. When comparing the RNR structure to a hand, the subunit is the whole hand. The barrel is the palm of the hand, and the N-terminal domain is the finger. The barrel contains the active site in the center. It is made up of “two halves of parallel beta strands” that are anti-parallel to each other, and forms an elliptical barrel (3). In addition, each half-barrel contains four connecting helices and five parallel strands (5). The N-terminal domain that contains the catalytic residue Cys-388, Glu-390 and Asn-386 on its tip is inserted into the center of the barrel (9). In addition, the N-terminal is fifty amino acids shorter than the RNR class Ia N-terminal at the R1E subunit because the allosteric activity site is absent. The small alpha-beta domains are on the bottom side of the barrel, but it contains many conserved hydrophobic residues (8). Furthermore, the formation of the R1E subunits consists of two helices from each subunit, which forms a helix bundle (9).
The RNR R1E subunit contains loop 1, 2, and 3 in the catalytic site. These loops are involved in regulating allosteric specific reactions. The allosteric specific site is located at the R1E dimer interface and contains residues 185-190 and 215-230 from one subunit, and 201-203 and 244-248 from the other subunit (3). An allosteric regulation occurs when a nucleotide triphosphate binds to the allosteric specificity site. This allows for five different structures of the subunit, which explains how RNR can bind to four different ribonucleotides. Loop 1 is in both R1E subunits and is ordered. Loop 2 is partially flexible, and loop 3 is flexible in one subunit but ordered in the other subunit. In addition, loop 3 binds to the phosphate group of the effector, which is also positioned in the active site by loop 1 (6). Also, all three loops interact with the allosteric specific site and the effector, which causes a conformational change in the three loops. The base of the effector goes between the monomers and the hydrophobic pocket. In addition, the redox-active cysteine pair that is involved in the radical transfer reaction is within the last fifteen residues of the R1E C-terminal tail (3).
The R2F subunits are mostly made up of alpha helices, and the dimer is shaped like a heart without the tip because the subunits are almost positioned perpendicularly to each other (3). The helices are arranged in a helical bundle in which the di-iron binding site is buried in it. Two ferric ions are ligated by one aspartate, three glutamate, and two histidine residues. In addition, water binds to the iron atoms and makes an octahedral coordination. One histidine residue forms a hydrogen bond to the water molecule and makes a pocket that contains additional water molecules. The additional water molecules then form more hydrogen bond to Asn and Asp residues. Also, the carboxylate from two glutamate residues form the bridges of the two iron atoms with an oxygen ion. With this structure, the RNR is in its oxidized conformation (pdb ID= 2R2F). Therefore, when the oxygen bridge is formed, a radical goes onto Tyr-105, which is near the two iron ions. The tyrosine phenolic group is about 6.5-7.0Å away from the iron ions. To stabilize the radical, tyrosyl radical is stored deeply into the protein, and it is surrounded by a hydrophobic environment. In a reduced structure, the bridging of the oxygen is absent, but the iron ions are bridged by the carboxylates of two glutamate residues. When an oxygen molecule binds to two reduced ferrous ions, the protein becomes active, transforms ferrous to ferric, and creates a tyrosyl radical. RNR in an active state will have diferric ions and a tyrosyl radical. The shift between the reduced and oxidized forms occurs only with intramolecular changes, which explains why the RNR can change even when it is crystallized. In addition, Helix E, in the oxidized form, covers the metal center in R2F subunits. When RNR changes to the reduced form, the helix changes its conformation, which opens up the metal center to allow oxygen to go into the iron binding site. In addition, the last 30-33 residues in the R2F C-terminus are flexible to allow the formation of the holocomplex (10).
The R1E and R2F dimer interaction is asymmetrical. The hydrophobic cleft of the R1E binds with the R2F subunit C-terminus, which forms a helix turn. The hydrophobic cleft in the R1E subunits is between alpha-I and alpha-D from the ten stranded alpha-beta barrel and alpha-10 helix from the small alpha-beta domain. There are three aromatic residues at the bottom of the cleft, and they are Phe-351 on alpha-D, Trp-684 on alpha-I and Phe-297 on alpha-10. In addition, Arg-685 binds to the R2F C-terminal. The dimer interaction is asymmetric because only one R2F C-terminal binds to one of the R1E hydrophobic cleft. Therefore, a R1E dimer will bind to two R2F dimers. The flexible C-terminus of the R1E does not bind to the hydrophobic cleft because the R2F C-terminus has a higher affinity to it, and the R1E C-terminus contains a cysteine pair to provide electrons to the active site. Furthermore, Val-9 and Met-10 on the N-terminal R1E subunits bind to the hydrophobic cleft on the other R1E subunit (3).
RNR catalyzes each of the four ribonucleotides into its corresponding deoxyribonucleotides by reduction. Three cysteine, one glutamate, and one asparagine residues are essential in the active site. In addition, four electrons from the external source, the tyrosyl radical, and the two ferrous ions are required to activate the reaction. When the substrate goes into the active site in the R1E subunits, Glu-390 and Asn-386, which are located at the finger loop, form hydrogen bonds to the oxygen atoms on the 2’ and 3’ carbons of the substrate. Therefore, the diphosphate group of the substrate is bound deep into the active site, and the ribose ring is located between a generated thiol radical that is at the tip of the finger in the barrel and the redox-active cysteine pair in the active site. The R2F subunits respond to a signal and release the stored organic free radical. The free organic radical or the tyrosyl radical goes up from the R2F subunits through a hydrogen bonded pathway between residues from each subunit and onto a cysteine residue in the active site of R1E subunits. This produces a thiol radical from Cys-388, which is located at the center of the barrel. The thiol then attacks the substrate to make a substrate radical by abstracting the hydrogen atom on the 3’ carbon. This allows a protonated hydroxyl group at 2’ carbon to be a good leaving group. The substrate radical is reduced by the reducing power from the redox-active cysteine pair, and a hydrogen atom is given to the 3’ carbon from the cysteine to regenerate the thiol radical. In addition, the disulfide bond between Cys-178 and Cys-415 is formed in an oxidized state, and steric hindrance prevents the cysteine residues from being accessible to the catalytic site. These cysteine residues are redox-active, and they are positioned on adjacent strands of the barrel. In a reduced conformation, Cys-415 moves away from the active site. Once the substrate is reduced and released, the redox-active cysteine pair has to be regenerated. Therefore, the cysteine pair is re-reduced by another cysteine pair in the C-terminus of R1E subunits when the C-terminus tail swings to the active site. Then an external electron donor re-reduces the redox-active cysteine pair in the C-terminus. The external electron donor is NrdH-redoxin, which acts like thioredoxin and has a primary structure like glutaredoxin, is coded by an additional operon. When the radical transfer occurs, the subunits are rearranged transiently so that it forms a tight symmetric complex (3).
The key to a successful reaction is controlling the free radical when it is transferred from the R2F subunits to the R1E subunits. RNR has a specific pathway to protect the radical from undergoing unwanted side reactions. The radical transfer pathway requires Tyr-304, and iron ligands His-101, Asp-67, Trp-31, and Glu-191 at the C-terminal tail of the R2F subunit. Tyr-304 is the last residue on the R2F subunit, and it is between Trp-31 of R2F subunit and Tyr-693 in the R1E holocomplex. Therefore, it is crucial for the pathway, and it acts like a bridge to transfer the radical between the two dimers. Every time Tyr-304 transfers the free organic radical, it reduces and re-oxidizes, and the distance between Tyr-304 and a radical transferring residue is 15Å. In the R1E subunit, Tyr-693 is closest to the R2F subunit so it accepts the radical from Tyr-304. In addition, it has another tyrosine residue stacked on top of it in the same beta strand, which causes the formation of hydrogen bonds of the phenolic oxygens to Cys-388 (3).
RNR’s associated ligands are ferrous ion, 2’deoxyguanosine-5’triphosphate, and magnesium ion. Iron is a cofactor that stabilizes the tyrosyl radical in the R2F subunits. In the R2F subunit, it becomes a bridged with oxygen and also helps generate the radical. 2’deoxyguanosine-5’triphosphate (dGTP) and magnesium are used to induce crystallization, but they can also bind or interact with the protein. The deoxyguanosine triphosphate ligand binds to the allosteric specific site and causes the RNR to reduce a specific ribonucleotide. When dGTP binds to the specific site, the base forms a hydrogen bond from the second nitrogen atom to the carboxyl group of Tyr-244, which is close to the 2’ carbon. Also, the three phosphate groups are surrounded by positively charged residues. In addition, dGTP causes ADP to be reduced in the substrate binding site. Each nucleoside triphosphate forms hydrogen bonds with different residues. Therefore they have different effects on the substrate preference. However, all nucleoside triphosphates form hydrogen bonds from the 3’ hydroxyl group to Asp-185. The allosteric specific site has a higher affinity to deoxyribonucleotides since the 2’ hydroxyl group of the ribonucleotides cause steric hindrance. In addition, magnesium ion can coordinate with the phosphate groups of the effector (3).
The human ribonucleotide reductase R1 subunit (pdb ID= 2WGH) from Homo sapiens is a homodimer, and it is functionally similar to the ribonucleotide reductase from Salmonella typhimurium (7). The human RNR’s primary structure is about 42% similar to the Salmonella typhimurium RNR. The two enzymes are also similar in the primary and tertiary structures because of the results from Blast in which the E value is 1e-45 and from DALI in which the Z score is 37.6 and the rmsd is 2.3 (1). A large Z score value indicates that the tertiary structures of the two enzymes are similar, which explains why the functions are also similar. A low E value indicates that there is a low chance that another enzyme with similar primary structure alignment will be found, which shows that the primary structures of the two proteins have a lot of similarities. The human ribonucleotide reductase can form an alpha-beta multi-subunit protein, and it has to form at least a heterotetramer in order to be active. Therefore, the human RNR can form a tetramer, but it is unstable and can serve as an intermediate for the formation of the hexameric state. Therefore, the human RNR is mostly in a hexameric state with six alpha and two beta subunits. As the concentration of dATP increase, the concentration of hexameric holocomplex increases also. However, the conformational change due to binding the dATP causes the radical transfer reaction to be disrupted in the hexameric holocomplex. In addition, the function of the subunits is similar to the ribonucleotide reductase. The smaller subunits generate and store the tyrosyl free radical while the larger subunits contain the catalytic site. However, one thing that is different between the two enzymes is that the human RNR has two allosteric sites, while RNR has only one allosteric site. This difference is shown because the human RNR is a class Ia enzyme that is usually in eukaryotes while RNR is a class Ib enzyme that is only found in prokaryotes. The extra allosteric active site in class Ia acts like an on and off switch for the enzyme. This allows the enzyme to produce equal amounts of dNTP and to control the rate of catalysis. Therefore, the dNTP also serves as an allosteric inhibitor in human RNR. The allosteric active site is significant because an unbalanced amount of dNTP is toxic to the cell. This site also contains a beta hairpin, which shifts to allow dATP and ATP to bind. Furthermore, when the human RNR forms a hexamer, it is active with ATP and inactive with dATP. Also, the large subunit has an ATP-binding cone with a beta-cap (7). RNR from Salmonella typhimurium only forms a tetramer, which is biologically inactive, and it has three domains in the large subunits. However, the Salmonella typhimurium RNR and the human RNR have similarities in their primary sequence. Some conserved amino acids in all class I RNR sequences are the three cysteine residues in the active site, Cys-439, Glu-441, Leu-438, and Asn-437 in the loop that goes into the center of the barrel, the two tyrosine residues for the radical transfer, and the hydrophobic and polar residues that reside in the active site cavity that is in between two domains (8).
Since RNR is also involved with the rate of cell proliferation, it is a drug target for a variety of diseases. Targeting the RNR would inhibit cellular division and also inhibits growth, and there are several ways to do that. Adding a chelating compound would remove the iron ion, which is also a growth factor, and stop cell proliferation. Also, peptidomimetic inhibitors, which are based on peptides from the C-terminus of R2F dimer, prevent the holoenzyme complex from forming. Inhibiting the radical transfer reaction and the binding of the effectors to the allosteric specificity site, and destroying the active site with substrate analogues would also inhibit cellular growth. An example of a drug that is a substrate analogue is gemcitabine. Gemcitabine is used to treat lung cancer and pancreatic cancer. Substrate analogues bind to the active site and cause side reactions that would inhibit RNR from functioning. In addition, radical scavengers, like hydroxyurea or hydroxylamine, can convert the tyrosyl radical into a tyrosine residue. Even though hydroxyurea is used in cancer therapy, it is nonspecific and takes a long time (2).