Alpha Subunit of collagen prolyl-4-hydroxylase (PDB ID: 4BTA) from Homo sapiens

Created by: Ciara Hutchison

            Collagen prolyl 4-hydroxylase (PBD ID: 4BTA) is a protein that catalyzes proline hydroxylation of procollagen, and the alpha subunit of this protein is the section of the protein that has the catalytic domain for this activity (1). Collagen prolyl 4-hydroxylase-α is of interest to scientists because this function is necessary for collagens (which perform a vital structural role in many structural areas of the human body such as bone, cartilage, and skin) to grow and mature properly (1). Thus, under-activity of this enzyme can lead to weakened tissues in diseases such as scurvy. Treatment of these diseases has traditionally focused on consumption of a diet with sufficient iron and vitamin C, which are the ligands necessary for the enzyme to function; pharmacological treatments are not usually needed Additionally, over-formation of collagen can cause fibrotic diseases and play a role in breast cancer metastasis, so scientists are trying to develop therapeutic inhibitors of this protein to treat these diseases (1).

         The overall structure of the protein consists of two identical alpha and two identical beta subunits; however, the structures of interest for this analysis are the alpha subunit only (2). There are three domains of collagen prolyl 4-hydroxylase-α, the N domain (first 143 subunits), the binding domain for the substrate (144-244), and the domain that performs the catalytic reaction (245-517) (1). All of the crystallization studies that have been performed on this protein have been on the N domain and the substrate-binding domain, collectively called the N-terminal region, and the catalytic domain has yet to be crystallized. Thus, the complete Protein Data Bank crystal structure used for this analysis (4BTA) shows the two identical alpha subunits as a dimer, with the catalytic regions of both alpha subunits (amino acids 245-517) missing, and the beta subunits not shown. The region of the alpha subunit containing the substrate-binding domain of collagen prolyl 4-hydroxylase was crystallized using polyethylene glycols and polyethylene glycol monomethyl ethers, with the structural data obtained using X-ray diffraction (1). Other crystal structures that have been determined for this protein are not functionally different, but rather equivalent structures that have been obtained using different crystallization techniques and focusing on slightly different areas of the protein (1). 

The alpha subunit’s isoelectric point is at a pH of 5.36, and its molecular weight is 28.997971 kDa (3). This data was determined using calculators from the Expasy proteomics site, which is an online database for determining many different aspects of protein structure. The protein has a secondary structure of mainly alpha helices with some 3-10 helices and random coil, and has mixed polarity but is mostly polar, allowing it to be a soluble protein (1,2). The alpha helices in the N region are particularly important due to the coiled coil motif that forms between the 2 alpha subunits and allows dimerization (1). The alpha helices of the substrate binding domain create tetratricopeptide repeats (linear structures which encourage binding with other proteins) and allow the protein to be solvated, and the fold of helices in the catalytic domain also allow this domain to have the correct shape to bind to the ligands (1).

The overall tertiary structure that arises allows for the grooves and pockets that bind the substrate and catalytic ligands. Importantly, the random coil portions of the protein that are not part of secondary structure allow flexibility and movement for the protein’s function; particularly important are two flexible loops in the catalytic domain that allow it to reach the substrate (1).

Collagen prolyl 4-hydroxylase-α does not contain any prosthetic groups and is not usually attached to any substrates or products, except when it is performing the catalytic reaction (1). During the catalytic reaction (proline hydroxylation of procollagen), it associates with the metal ion Fe²⁺ as well as 2-oxoglutarate and ascorbic acid (vitamin C) (4). It also interacts with its collagen substrate through the substrate-binding domain. Although the detailed mechanism of the catalytic reaction is still unknown, the Fe²⁺ ion decarboxylates 2-oxoglutarate to create a ferryl ion which is known to be part of the carboxylation reaction of a procollagen proline (4).

Several drugs that inhibit this catalysis are known: they are oxalyl amino acid derivatives that act as competitive inhibitors by mimicking 2-oxoglutarate (5). The various inhibitors vary slightly in structure, but they all have -NH- instead of the –CH₂- that 2-oxoglutarate has at the number 2 position. The structure of the protein-drug complex is analogous to that of the protein-ligand complex during the catalytic reaction, with the drug binding at His-501 (6). However, these inhibitors are unable to react with the Fe²⁺ ion, meaning that the decarboxylation reaction of the 2-oxoglutarate analog and ferryl ion creation cannot occur, and they are therefore unable to be involved in carboxylating proline. Examples of these drugs are oxalylglycine and oxalylalanine (5).

There are several known functionally important residues in the substrate binding and catalytic sites that bind the ligands and therefore contribute to the protein’s catalytic activity. While the specific residues and binding mechanism for the substrate binding domain are not known, scientists predict that the binding is due to tyrosine residues exposed along the binding groove in the tetratricopeptide repeat section (1). Residues that bind to the other ligands are better known. His-412, Asp-414, and His-483 all bind to the Fe²⁺ ion. Lys-493, also known as subsite 1, binds to 2-oxoglutarate. His-501’s specific function has not yet been experimentally identified, but it is known to be part of the active site and thought to be involved in the decarboxylation reaction of 2-oxoglutarate (6). However, because a crystal structure for the catalytic domain of the protein has not yet been determined and does not exist on the Protein Data Bank,  ligand-binding and catalytic residues for this protein cannot be modeled.

Aside from these functionally important residues, there are also several structurally important residues in this protein, which, while not catalytic, are still vital for the protein. There are many hydrogen bonds and ionic interactions (salt bridges) in the protein, most of which contribute to the overall stability of the structure. This stability allows the protein to maintain the correct shape. First, the many helices in the protein are stabilized by hydrogen bonds. There are also specific irregular hydrogen bonds aside from those, including between Gln-110 and Glu-152, Lys-115 and Glu-160, and Gln-121 and Trp-169 (1). There are salt bridges (strong ionic interactions between a positively charged amino acid and a negatively charged amino acid) along the face between the two alpha subunits involving the residues Glu-33, Lys-40, Glu-82, and Arg-101, which allow dimerization (1). Additionally, Glu-18 in each subunit makes both an inter-chain salt bridge with Arg-73 and an inter-chain hydrogen bond with Trp-78 in the other subunit; and Glu-32 in each subunit makes both an inter-chain salt bridge with Lys-35 and an inter-chain hydrogen bond with Tyr-28 in the other subunit (1). All of these individual interactions are important: some of the hydrogen bonds help create the overall tertiary structure of the protein and keep it stable so that the catalytic site has the correct conformation to perform its function, and the salt bridges and other hydrogen bonds allow dimerization between the two alpha subunits, leading to quaternary structure (6).

A comparison protein (a protein with similar structure) was determined for collagen prolyl 4-hydroxylase-α using the Dali and PSI-BLAST servers. This comparison protein is the J chain of the human anaphase promoting complex (APC, PBD ID: 4UI9) (7,8). The PSI-BLAST server searches based on sequence and searches for proteins with a similar primary structure. It compares sequences for a protein of interest with other proteins and searches for gaps (regions that exist in the comparison protein, or “subject,” but not in the query) in the homology between the proteins. The Dali server searches based on overall shape and finds proteins with a similar tertiary structure by using a sum of squares method which compares intramolecular distances in a comparison protein versus a query. PSI-BLAST finds an E value for comparison proteins, and Dali finds a Z score; as measures of comparison, a high Z score (similar and a low E value (fewer gaps and therefore greater homology) mean a similar protein to the protein of interest. Cutoff scores for a protein considered significantly similar for a comparison are an E value of less than .05 and a Z score of greater than 2. APC-J has a Z score of 11.1 and an E value of 9e-5 (7,8).

While both APC-J and collagen prolyl 4-hydroxylase-α are based mostly on alpha helices, APC-J is a higher molecular weight and more complex, but also more compact structure (2,9). Collagen prolyl 4-hydroxylase-α also has proportionally more random coil than APC-J (2,9). APC-J is associated with Zn²⁺, unlike collagen prolyl 4-hydroxylase-α, which is only associated with a ligand when it is actively catalyzing a reaction (9). The main area of homology between these two proteins is in the tetratricopeptide repeat region (2,9).

The APC is a multimeric (19 subunits) ubiquitin ligase that is responsible for driving the cell toward anaphase of mitosis (separation of chromosomes). Due to the size and complexity of this molecule, detailed knowledge of the structure and function of every subunit of the APC (including J) is still forthcoming. However research shows that the catalytic domains are located in subunits 2 and 11, so the rest of the molecule is generally “static” (10). Thus, the main function of APC-J, as with the other non-catalytic subunits, is to provide structural stability (10). Unlike collagen prolyl 4-hydroxylase-α, APC-J does not have catalytic activity, so a more static, stable structure is necessary for APC-J. This is possible because it has relatively more alpha helix and less random coil than collagen prolyl 4-hydroxylase-α (9). Additionally, the more compact tertiary structure of APC-J allows it to fit with the many other subunits (9,10).

All in all, the structure of collagen prolyl 4-hydroxylase-α clearly contributes to its function. Individual amino acids in the primary structure contribute to the catalytic function of the enzyme as well as to non-covalent interactions for stability. The overall sequence also leads to the secondary and tertiary structures: the secondary structures (mainly alpha helices) give the protein stability and allow it to interact with substrates and other ligands, and the random coil in between give the protein flexibility, and the tertiary structures give the protein its overall stable structure and ability to catalyze the hydroxylation of procollagen.