mom

Momordin (PDB ID: 1MOM) is a type I ribosome inactivating protein (Type 1 RIP) extracted and purified from seeds of the bitter melon Momordica charantia (1). The bitter melon M. charantia has been routinely used as an alternative medicine for anti-cancerous, anti-diabetic, and anti-tumorous treatments (2). Due to M. charantia’s success in alternative treatments, many biochemists have directed their research toward the individual proteins responsible for the success, focusing primarily upon ribosome inactivating proteins. Ribosome inactivating proteins are of particular interest due to their ability to irreversibly damage and inhibit ribosomes (3). Whereas Type II RIPs are composed of both an A and B-chain and demonstrate high cytotoxicity, Type I RIPs such as momordin are comprised solely of an A-chain component and exhibit much lower cytotoxicity. Interest arose over momordin and other Type I RIPs based upon their similarity to either the A or B-chain of Type II RIPs as well as their inhibiting capabilities. Momordin is known for its ability to inhibit activator protein-1 (AP-1), and has even been observed in past studies to display an inhibitory effect against tumorous cells, one such example being Ehrlich ascites-tumour cells (4, 5). Current interest lies in the selective toxicity of momordin toward T-cells and HIV infected macrophages for use as a potential specialized treatment. The specificity for either cell type would ultimately be accomplished through conjugation to antibodies and hormones (1).

When allowed to equilibrate against 20-40% w/v polyethylene glycol 4000 at pH 7.4, momordin was crystallized completely as rhombohedral crystals. However, with lactose included as a crystallization additive, isomorphous crystals could be obtained. Despite altered morphology and improved diffraction characteristics, electron density imaging provides reliable evidence of the similar amino acid sequences. The only notable difference between the two crystal types was a substitution of Leu-64 for Val-64 in the isomorphous crystals (1).

Upon comparing the structure of momordin to that of a Type II RIP such as Ricin (PDB ID: 2AAI), additional structural differences are observed beyond the absence of a B-chain. In Ricin, a length of ten hydrophobic residues follows Cys-259, which are capable of being buried within its A-chain or by its associated B-chain. These respective residues in momordin can be buried through folding of the A-chain. Those hydrophobic residues that would have been incapable of correct and/or sufficient folding appear to have been substituted by amino acids containing hydrophilic or uncharged side chains (1). Momordin is comprised of a single A-chain unit with a respective isoelectric point of 9.32 and a molecular weight of 27382.38 Da (6). The primary structure of momordin consists of 246 amino acid residues interacting to compose a secondary structure of 40% helices and 24% beta sheets. The beta sheet portion of the protein comprises 60 residues, producing a total of 15 strands throughout. The helical portion of the protein comprises 100 residues, producing twelve helical structures overall (7). When comparing the published sequence of momordin to the electron density determined sequence, there is an overall difference of 17 residues. The published sequence reports a total of 263 residues. The missing residues are reported to be missing from the C-terminus of the protein, most likely the result of proteolytic cleavage (1).

The structure of momordin is what ultimately determines its inhibiting activity. The mechanism through which momordin acts was determined by analyzing the momordin protein in the Japanese Ampelopsis Root, or Ampelopsis radix. In the mechanism, momordin continues to inhibit the de novo synthesis of AP-1. Upon analysis, it was presumed that the inhibiting ability of momordin originates from its interaction with the nucleic acid in the AP-1 site of DNA (8). Out of the twelve helical structures comprising momordin, eight correspond to alpha-helices enclosing the active site cleft. The active site of momordin appears to be defined by the four amino acid residues Tyr-70, Tyr-111, Glu-160, Arg-163, and Trp-192. Both Glu-160 and Arg-163 are proposed to be the central catalytic residues. Tyr-70 is primarily responsible for shielding Trp-192 in the active site from external solvent whereas it is normally exposed in other similar unliganded structures. In addition, it is suspected that Trp-192 may adopt two different conformations. This allows for hydrogen bonding to be selective to either Leu-239 or Tyr-70. In addition, momordin is also glycosylated at the residues Ile-208, Thr-226, Asn-227, Thr-229, and Ser-230 with N-Acetyl-D-Glucosamine (NAD), which seems to be utilized primarily for binding sites and signaling with cells (1).

Momordin was compared to similar protein structures using the Dali Server and the Position-Specific Iterated Basic Local Assignment Search tool (PSI-BLAST). The purpose of the Dali Server is to find proteins with similar tertiary structures to that of the protein query. This is accomplished through a sum-of-squares approach by which a measure of similarity is calculated by comparison of intermolecular distances. Upon completion of the calculation, a protein is assigned a Z-score with a value greater than 2 indicating high similarity between tertiary structures of the selected protein and the query (9). The purpose of PSI-Blast, on the other hand, is to find proteins with a similar primary structure to that of the protein query. Similar proteins are evaluated on the basis of homology and assigned a corresponding E-value. The smaller the value—below 0.05 for best results—indicates a strong primary sequence homology between the query and a selected protein (10). Using both the Dali Server and PSI-BLAST, the Type II RIP Luffaculin-1 (PDB ID: 2OQA) was determined to be a comparable structure. Luffaculin-1’s low E value of 7e-110 and a high Z score of 40.1 indicates high similarities with the primary and tertiary structures of momordin. Momordin was compared to a Type II RIP due to the fact the Type I RIP match results corresponded primarily to the exact momordin structure in other plant species (9, 10). As a result of its classification and similarities, the biological significance of luffaculin is extremely similar to that of momordin as well as other Type I and II RIPS in that it functions as an inhibitor of protein synthesis, leaving potential for pharmaceutical and disease treatments.

The largest differences noted between luffaculin and momordin is that luffaculin is a Type II RIP composed of both an A-chain and a B-chain. Momordin’s structure appears to correlate closely with that of luffaculin’s A-chain. When comparing the two A-chains, it is obvious that luffaculin’s A-chain is shorter and varies in secondary structure from that of momordin. Luffaculin is comprised of 241 residues, 39% of which correspond to eleven helical structures and another 26% corresponding to 15 beta strands. Of the eleven helical structures, nine of them correspond directly to alpha helical structures. The remaining two helical structures correspond to 3/10-helices. The majority of the amino acids in either of the protein sequences are highly conserved, with only small strips of protein diverging from homology. When analyzing those diverging sections, it is noticed that there is no pattern or rule for amino acid substitutions. In some cases, similar amino acids correspond to the same position between the proteins, such as a hydrophobic amino acid in luffaculin corresponding to a hydrophobic amino acid in the corresponding location of momordin. In other cases, there are opposing amino acids, such as a hydrophilic in luffaculin corresponding to a hydrophobic in momordin. The noticeable difference between the lengths of the two structures is due to the absence of momordin’s four terminating residues as well as possible deletion of two residues within the structure of luffaculin during its evolutionary history. Upon comparing the sequences, it appears the deletion corresponds to somewhere around residues 97 to 103 as well as Leu-240 of momordin. Residues 97-103 of momordin and the corresponding residues of luffaculin are not homologous, and therefore it is difficult to determine which amino acid position present in momordin is not present. Another principle difference between the momordin and luffaculin revolves around the number of binding sites and associated prosthetic groups. Unlike momordin, which contains only NAD and its corresponding binding site, luffaculin contains NAD, tetraethylene glycol (PG4), di(hydroxyethyl)ether (PEG) (solely on the B-chain), as well as their corresponding binding sites. When looking at the active site of luffaculin, it is noticeably similar to momordin, containing eight helical structures defining the cleft. This is partially due to active sites among Type I and Type II RIPs being highly conserved (7, 11).