ZMP: A Master Regulator of One-Carbon Metabolism

Abstract

In this issue, Kim et al. (2015) show that ZMP (5-aminoimidazole-4-carboxamide ribonucleotide) binds to and activates a conserved riboswitch to regulate expression of one-carbon metabolism genes.

Cell growth requires biosynthesis of balanced quantities of protein, RNA, DNA, lipids, and their associated metabolic precursors. An important family of building blocks is one-carbon units, which are carried by the cofactor tetrahydrofolate (THF). An essential vitamin in humans, folate is required for growth and development, and inadequate dietary intake remains a leading cause of birth defects globally (World Health Organization, 2008). Conversely, pharmacological targeting of folate metabolism led to the development of some of the most important antibiotics and chemotherapeutics. It has long been observed that defects in folate metabolism can manifest biochemically as an accumulation of a late intermediate in purine biosynthesis, 5-aminoimidazole-4-carboxamide ribonucleotide (ZMP or AICAR; note that AICAR is also sometimes used to refer to the analogous ribonucleoside, which is commonly used as a pharmacological agonist of AMPK in eukaryotic cells) (Gots and Chu, 1952). However, despite intense interest over several decades, a ZMP-selective receptor had not been identified, and the role of this metabolic signal remained unclear. In this issue, Kim and colleagues identify a conserved mechanism whereby ZMP, and the analogous nucleotide triphosphate ZTP, binds to and activates a riboswitch present in multiple bacterial lineages to directly regulate the expression of one-carbon metabolism genes (Kim et al., 2015). This finding presents the first mechanism, independent of the production or import of THF itself, to systematically regulate the flux of one-carbon units through the folate pathway. Given the conservation of one-carbon metabolism throughout all domains of life, it is attractive to hypothesize that ZMP may also be a master regulator of this pathway in eukaryotes.

Riboswitches are conserved regulatory elements in the 5′ UTR of many bacterial, fungal, and plant mRNAs. They consist of a small molecule or ion binding RNA aptamer coupled with a RNA sequence, known as the variable expression platform, that can positively or negatively control expression of the subsequent transcript. This mechanism allows a metabolite to directly suppress (or enhance) the expression of multiple metabolic genes simultaneously, bypassing the need for transcription factors. Riboswitches are now known to be the regulatory elements in many metabolic control circuits, including those for the synthesis of important co-factors such as folate, thiamine, and riboflavin (Serganov and Patel, 2012). The authors had previously identified a conserved RNA motif termed pfl as a candidate riboswitch that was associated with one-carbon metabolism genes, but the ligand remained a mystery (Weinberg et al., 2010).

Here, through a series of biochemical and cellular experiments, the authors demonstrate persuasively that the pfl motif is a functional riboswitch whose ligand is the purine intermediate ZMP. The pfl riboswitch binds ZMP with nanomolar efficiency and is greater than 1,000-fold selective over other purine monophosphates including AMP, IMP, and XMP. Specific elements of ZMP recognized by the pfl motif include the primary amine and terminal amide groups and the riboside itself. By using a pfl-lacZ reporter construct, the authors confirm the in vivo functionality of this ZMP riboswitch in response to folate stress. Impressively, they also identify the pfl motif in over 2,000 bacterial gene sequences (Kim et al., 2015).

Interestingly, unlike other known phosphate binding aptamers, this riboswitch does not discriminate by the number of phosphates. Thus, while pfl is the first confirmed receptor target for ZTP, ZMP is equally active. This raises the possibility that ZTP, which was originally identified as an “alarmone” in S. Typhimurium cultures subjected to various folate stresses (Bochner and Ames, 1982), is an accidental metabolite produced due to the promiscuous activity of PRPP synthetase (Sabina et al., 1984). Alternatively, ZTP may play a functional role in modulating signaling dynamics. For example, if ZTP were to be turned over more slowly than ZMP, ZTP might produce more sustained one-carbon gene activation.

Why is ZMP such an attractive intermediate metabolite for regulating one-carbon metabolism? ZMP is the penultimate intermediate in purine biosynthesis, lacking only the carbon from 10-formyl-THF to form IMP, and can be produced by two separate pathways, the THF-dependent de novo purine biosynthetic pathway and the histidine biosynthetic pathway (Figure 1). The existence of an alternative salvage pathway to ZMP can explain why an upstream purine biosynthetic intermediate, glycineamide ribonucleotide (GAR), is not the signal for one-carbon stress. High flux through purine synthesis in rapidly dividing cells ensures that deficiency in 10-formyl-THF will be quickly reflected in ZMP levels, leading to a positive signal of the metabolic defect. As activated THF species are both unstable and structurally complex (with variable polyglutamate tails of 3–10 residues), a positive signal indicated by increased ZMP levels enables response to a single species likely providing advantages over detection of 10-formyl-THF depletion directly.

Figure 1. ZMP Regulates Expression of Tetrahydrofolate Genes.

The purine intermediate ZMP (highlighted in pink) is a positive regulator of expression of tetrahydrofolate genes involved in one-carbon metabolism. ZMP can also be produced from ATP via the histidine biosynthesis pathway. A conserved riboswitch is activated by ZMP binding, leading to expression of metabolic genes highlighted in pink. GlyA (glycine hydroxymethyltransferase), FolD (5,10-methylene-tetrahydrofolate dehydrogenase/5,10-methylene-tetrahydrofolate cyclohydrolase), Fhs (formate tetrahydrofolate ligase), PurH (AICAR formyltransferase/IMP cyclohydrolase).

Given the apparent conservation of the ZMP sensor riboswitch as a regulator of one-carbon metabolism genes in bacteria, it is tempting to speculate whether a similar circuit exists in higher organisms. In mammalian cells, ZMP has been extensively studied in its capacity as an AMP mimetic and activator for various AMP-utilizing enzymes including the master energy sensing kinase, AMPK. Direct activation of AMPK, a heterotrimeric complex that exists in at least 12 different forms, either by the native substrate AMP or by ZMP, downregulates mTORC1 activity and upregulates cellular catabolic processes including autophagy and fatty acid catabolism. Administration of the cell-permeable ZMP precursor AICA-riboside induces large effects in systemic metabolism in an AMPK-dependent fashion (Narkar et al., 2008). What is not known, however, is whether any of these identified AMP-mimetic mechanisms of ZMP activity are engaged by physiological concentrations of ZMP, as opposed to the pharmacological treatments used to date. Notably, AMPK activity does not regulate one-carbon metabolism genes, and a preference for ZMP over AMP by any AMPK-family member nucleotide binding site has not been identified.

Thus, it remains simply an attractive hypothesis that ZMP-specific processes exist in mammalian cells to regulate one-carbon metabolism. Like in bacteria, one-carbon stress induced either by nutrient deprivation or by antifolates causes large increases in ZMP in humans as evidenced by the urinary excretion of AICA-riboside (Luhby and Cooperman, 1962). In S. cerevisiae, ZMP can bind to and activate transcription factors that control the expression of tetrahydrofolate and purine synthesis genes (Pinson et al., 2009), yet similar regulatory elements have yet to be identified in animals. Given the importance of anti-folate therapies in the treatment of many human cancers, there is strong motivation to further explore the regulatory networks that control these processes.