An enzyme in the test tube, and a transcription factor in the cell: Moonlighting proteins and cellular factors that affect their behavior

Abstract

In the cell, expression levels, allosteric modulators, post‐translational modifications, sequestration, and other factors can affect the level of protein function. For moonlighting proteins, cellular factors like these can also affect the kind of protein function. This minireview discusses examples of moonlighting proteins that illustrate how a single protein can have different functions in different cell types, in different intracellular locations, or under varying cellular conditions. This variability in the kind of protein activity, added to the variability in the amount of protein activity, contributes to the difficulty in predicting the behavior of proteins in the cell.

Introduction

Within a cell, many factors affect the level of a protein's activity. Varying rates of protein synthesis and degradation can result in different expression levels, and the level of intrinsic activity can be adjusted by the presence of allosteric modulators, post‐translational modifications (PTM), sequestration, and other means. For some proteins, not just the level of activity, but also the kind of activity can change. In moonlighting proteins, a single polypeptide chain performs two or more distinct and physiologically relevant biochemical or biophysical functions that are not due to gene fusions, multiple RNA splice variants, or pleiotropic effects.1 Which function a moonlighting protein exhibits at a given time is due to cellular factors that vary in different cell types, at different stages of the cell cycle or development, in different intracellular locations, and due to changes in the cell's environment (Fig. 1). The online MoonProt Database includes information about hundreds of moonlighting proteins for which biochemical or biophysical evidence supports the presence of at least two biochemical functions in one polypeptide chain (Ref. 2, moonlightingproteins.org). In this minireview, we will discuss some of the factors in the cell that enable proteins to exhibit different behaviors in different cellular contexts.

Figure 1.

Many factors can affect which function(s) is exhibited by a moonlighting protein in a cell. For example, an intracellular enzyme can also be a cell surface receptor, a part of a multiprotein complex, a secreted signaling protein, or a transcription factor. The expression of other proteins, concentrations of substrates and other ligands, and post‐translational modifications can all affect the role it plays in a given cell type and under different conditions.

Specialized Cell Types

Some of the first examples of moonlighting proteins to be identified were proteins that have a second function in specialized cell types. For example, β‐1,4‐galactosyltransferase transfers galactose to the N‐acetylglucosamine on the growing chains of glycoproteins. In mammary tissue, the binding of alpha‐lactalbumin, a noncatalytic homologue of lysozyme, causes the galactosyltransferase to switch to transferring galactose to glucose to make lactose, a major carbohydrate in milk.3, 4

About a dozen ubiquitous enzymes were adopted to play a second role in the specialized cells of the lens of the eye. These taxon‐specific crystallins are transparent water‐soluble proteins that are expressed at high concentrations to make up the structure of the lens. Interestingly, different enzymes were adopted to be crystallins in different branches of the evolutionary tree. For example, arginosuccinate lyase is an ubiquitous enzyme that catalyzes the fourth step of the urea cycle, the breakdown of argininosuccinate to produce arginine and fumarate. In birds and reptiles, it is also the delta‐2 crystallin.5, 6, 7 l‐gulonate 3‐dehydrogenase catalyzes the NAD‐linked dehydrogenation of l‐gulonate into dehydro‐l‐gulonate in the urinate cycle. In rabbits, it is also the lambda crystallin.8 While in the lens, the taxon‐specific crystallins do not exhibit their enzymatic activity (the lens lacks the enzymes' substrates), but they were adopted for this second function during evolution because their physical properties enable them to provide the correct refractive index while not blocking light transmission at high concentrations.

More recently, it was discovered that mammalian sperm use several intracellular proteins in a second role on the cell surface to bind to surface proteins on the egg during fertilization. Triosephosphate isomerase, pyruvate kinase, and fructose‐1,6‐bisphosphate aldolase in glycolysis, and glutathione S‐transferase, glutathione peroxidase 4, and peroxiredoxin 5 interact with zona pellucida proteins of the egg.9, 10, 11, 12 A transmembrane channel, the voltage‐dependent anion channel 2, which transports adenine nucleotides, Ca²⁺, and other metabolites, also binds to rhZP2 or rhZP3 on the surface of the egg.9

Cellular Location

The use of an ubiquitous cytoplasmic enzyme in a different place is not just for specialized cell types in animals. Throughout the evolutionary tree, intracellular enzymes, chaperones, and other proteins have been found to be secreted or displayed on the cell surface where they perform a different function. In multicellular organisms, intracellular/secreted proteins act as signaling molecules for coordinating the activities of different cells and tissues during growth of the organism, and some play a key role in signaling in the immune system (reviewed in Ref. 13). Some intracellular proteins are secreted by pathogens to bind to host proteins or tissues, where they can affect the activities of host cells and modulate the host immune responses to infection.

Intracellular/secreted enzymes are also often used as cell surface receptors for small molecules or soluble proteins, interacting with the surface of other cells, or binding to extracellular matrix (ECM) as a means of attachment and colonization of the host (reviewed in Refs. 14 and 15). A few of these serve as cell surface receptors for soluble nutrients, signaling molecules, and other proteins. For example, GAPDH in several Staphylococcus species, as well as in mammals, has been shown to be a cell surface receptor for transferrin, a soluble iron carrier.16, 17 Iron is often a limiting factor in the growth of pathogenic bacteria, so the ability to scavenge host iron is important in infection. In pathogenic bacteria, several intracellular enzymes are also cell surface receptors for interacting with host cells. Streptococcus pyogenes GAPDH binds the uPAR/CD87 receptor on human cells,18 Streptococcus pneumoniae fructose 1,6‐bisphosphate aldolase binds to the Flamingo cadherin receptor on host cells.19 Listeria monocytogenes alcohol acetaldehyde dehydrogenase (ALDH) catalyzes two reactions in the metabolism of ethanol (ALDH and alcohol dehydrogenase) and also acts as an adhesion that binds to Hsp60 (another moonlighting protein) on the surface of host cells.20

Pathogenic bacteria, including Mycoplasma genitalium, and probiotic bacteria, including Lactobacillus acidophilus, use GAPDH and several other enzymes to bind to collagen, fibrinogen, and fibronectin in the ECM or to mucins contained in the mucus lining of the gastrointestinal tract and airway.21 In Streptococcus gordonii, the DNA‐directed RNA polymerase beta subunit, the glycolytic enzyme enolase, translation elongation factor Tu (EF‐Tu), and EF‐G have all been found to bind to the mucin MUC7.22 The probiotic Lactobacillus plantarum also uses enolase to bind to fibronectin.23, 24 Another glycolytic enzyme, glucose 6‐phosphate isomerase, is used by Lactobacillus crispatus to bind to laminin and collagen I.25

Some intracellular proteins are secreted to become insoluble materials outside the cell. The mouse SMC‐3 protein (Structural maintenance chromosome 3) works with SMC1, Scc3, and Scc1 (also called Rad21), in the cohesin complex to maintain pairwise alignment of chromosomes on the mitotic spindle and enable proper chromosome segregation during mitosis and meiosis. Outside the cell, SMC‐3 is the same protein as bamacan, a proteoglycan component of the basement membrane in the Engelbreth‐Holm‐Swarm tumor matrix, the renal mesangial matrix, and the basement membrane of other tissues. It plays a role in the control of cell growth and transformation.26, 27

Switching Functions in the Cytosol and Nucleus

Many cell types also make use of moonlighting proteins to perform two different functions within the cytosol of the same cell. These different functions often involve binding to another protein or multiprotein complex and regulating or coordinating signaling pathways, transcription, and/or translation. Several proteins that are part of the ribosome, a large protein/RNA complex, leave the ribosome and interact with different proteins. Saccharomyces cerevisiae RACK1 is also a scaffold protein in cytoplasmic signal transduction pathways.28 The L10 ribosomal protein from the plant Arabidopsis thaliana is a substrate and binding partner of NIK1 in the cytosol and is involved in a NSP‐interacting kinase (NIK) receptor‐mediated defense pathway to defend against geminivirus.29 In humans, the L11, L23, L5, and S7 ribosomal proteins bind to and inhibit HDM2, a ubiquitin ligase, which results in stabilization of the p53 tumor suppressor protein.30, 31, 32, 33 A. thaliana hexokinase 1 binds to a channel, the vacuolar H+‐ATPase B1 (VHA‐B1), and the 19S regulatory particle of proteasome subunit (RPT5B) to modulate transcription of specific target genes.34

In some cases, the moonlighting proteins interact with proteins in the cytoskeleton by affecting polymerization or by helping to attach other cellular components to cytoskeleton. Fructose‐bisphosphate aldolase A from Oryctolagus cuniculus (Rabbit) sequesters WASP (Wiskott–Aldrich Syndrome protein), which is involved in controlling actin dynamics and inhibits the WASP‐stimulated Arp2/3‐dependent actin polymerization reaction.35 The glycolytic enzyme fructose‐bisphosphate aldolase from Plasmodium berghei, the protozoan parasite that causes malaria, attaches actin filaments to TRAP proteins (transmembrane adhesive proteins of the thrombospondin‐related anonymous protein) and transduces the motor force across the surface of the plasmodium.36

Some cytosolic enzymes act as autologous activators or repressors of translation or affect RNA splicing or RNA half‐life. The binding sites in the RNA can be in the gene encoding regions, introns, or 5′ or 3′ untranslated regions (reviewed in Ref. 37. Human GAPDH and glutamyl‐prolyl tRNA synthetase, which catalyzes the attachment of amino acids glutamate and protein to their cognate tRNAs, are components of the GAIT complex (interferon‐gamma‐activated inhibitor of translation) that is involved in silencing ceruloplasmin mRNA translation (reviewed in Ref. 38). Human phosphoglycerate kinase 1 has a second function in binding mRNA to regulate expression of the urokinase receptor.39 Human methylglutaconyl‐CoA hydratase in leucine catabolism binds RNA and regulates mitochondrial protein synthesis.40 Dihydrofolate reductase binds DHFR mRNA to regulate DHFR synthesis, and the ligand methotrexate inhibits the interaction,41 and thymidylate synthase, which is important for de novo synthesis of thymidylate in pyrimidine metabolism, inhibits thymidylate synthase mRNA translation.42

Several other cytosolic enzymes respond to changes in the cell by moving to the nucleus to act as transcription factors. For example, rat pterin‐4‐alpha‐carbinolamine dehydratase converts 4alpha‐hydroxy tetrahydropterin to quinonoid dihydrobiopterin in a phenylalanine hydroxylation reaction. The enzyme also binds to and causes dimerization of hepatocyte nuclear factor 1alpha, a homeodomain transcription factor, resulting in transcription activation.43, 44 Escherichia coli transketolase from the pentose phosphate pathway derepresses the marRAB multiple antibiotic resistance operon by binding to the MarR repressor.45 Instead of activating transcription, Listeria monocytogenes GmaR is an O‐GlcNAc transferase (glycosyltransferase) that modifies flagellin and acts as a transcriptional anti‐repressor by binding to the MogR transcriptional repressor and preventing it from binding to DNA.46

Palmitoylation, the attachment of a fatty acid, of the mouse estrogen receptor causes it to move in the opposite direction—from the nucleus to the plasma membrane to participate in membrane initiated steroid pathways that causes dilatation of the vasculature and increased endothelial repair. A mutation at the palmitoylation site (C451A‐ERα) inhibits targeting to the plasma membrane and signaling, but the mutant protein retains the nuclear functions.47

Post‐translational Modifications

Other PTMs also cause moonlighting proteins to switch between functions, and because they can be dynamic and reversible, PTMs on moonlighting proteins can contribute significantly to a cell's ability to adapt quickly to changes in the cell's needs, and the functions in different cell types can depend on the presence of those modifying enzymes. Phosphorylation of several ribosomal proteins results in them leaving the ribosome and move into the nucleus. In the nucleus, human ribosomal protein S3 joins a multiprotein complex that binds DNA and is involved in NF‐kappaB‐mediated transcription.48 The trigger for A. thaliana ribosomal protein L10a to leave the 40S subunit and participate in the NIK receptor‐mediated defense pathway against geminivirus reproduction, mentioned above, is phosphorylation by its binding partner, NIK1.49 Human ribosomal protein L13a is part of the 60S subunit,50 but upon phosphorylation is released from the ribosome and becomes a component of the GAIT complex described above.51 Another protein involved in protein synthesis, the human glutamyl‐prolyl tRNA synthetase, which catalyzes the attachment of amino acids to cognate tRNAs, also becomes part of GAIT complex upon phosphorylation.52

Small Molecules

In addition to responding to factors such as the presence of other proteins or PTMs, the behaviors of many moonlighting proteins depend directly on cellular concentrations of substrates, products, cofactors, or other ligands. Aconitase in mammals and in the bacterium Mycobacterium tuberculosis comprise a well‐known example of a protein that switches functions due to binding or release of a ligand. As an enzyme in the citric acid cycle, aconitase uses an iron–sulfur cluster in its active site. When cellular iron concentrations are low, aconitase loses its 4Fe‐4S cluster, undergoes a large conformational change and is then able to bind to iron‐responsive elements in mRNA that encode proteins that are involved in iron uptake and use.53, 54, 55

Conclusions

The growing number of known moonlighting proteins, their wide occurrence in the evolutionary tree, and the vast variety of their functions and combinations of functions suggests that many more proteins may also have multiple functions. Because there does not seem to be a clear pattern or specific protein characteristics that can inform us as to which proteins have multiple functions, knowing the protein complement encoded by a genome or even which proteins are expressed in a given cell type can still leave ambiguity about which cellular processes are active there. In addition, the ability of many factors to affect a moonlighting protein's function adds to the difficulty of predicting protein behavior within the complex environment of a cell.

In order to improve predictions of protein behavior in the cell, there is still much more that needs to be elucidated about the biochemical or biophysical functions of moonlighting proteins (and other proteins) and the cellular factors involved in their regulation. Much of what we know comes from studies of individual proteins for which one or more functions were sometimes only partly characterized, for example, there is little or no information available about the mechanisms of regulation of the activities for some of the known moonlighting proteins, so many additional studies to characterize enzyme activity, binding to other molecules, allosteric regulation, and protein structures, as well as other biochemical and biophysical studies are still needed. In addition, because many protein functions have been studied using only one or a few representative proteins from a protein family, accurate extrapolation of information about function and regulation to the millions of protein sequences becoming available from sequencing projects will require more extensive information about how similar homologous proteins need to be in order to share both, one, or neither function(s) and regulatory mechanisms. Accessibility, clarity, and accuracy of all this experimental information will also be important for predicting and understanding the functions of proteins within the cell, so it will also be valuable to continue manual annotation of the experimental results providing information about protein function, along with clear information about which version of a protein (which paralogue, from which species, etc.) performs that function, in databases such as the UniProt Knowledgebase (UniProtKB)56 and the MoonProt Database.