The N-glycosylation defect in Lec5 and Lec9 CHO cells is caused by absence of the DHRSX gene.

Glycosylation-deficient Chinese hamster ovary (CHO) cell lines have been instrumental in the discovery of N-glycosylation machinery. Yet, the molecular causes of the glycosylation defects in the Lec5 and Lec9 mutants have been elusive, even though for both cell lines a defect in dolichol formation from polyprenol was previously established. We recently found that dolichol synthesis from polyprenol occurs in three steps consisting of the conversion of polyprenol to polyprenal by DHRSX, the reduction of polyprenal to dolichal by SRD5A3 and the reduction of dolichal to dolichol, again by DHRSX. This led us to investigate defective dolichol synthesis in Lec5 and Lec9 cells. Both cell lines showed increased levels of polyprenol and its derivatives, concomitant with decreased levels of dolichol and derivatives, but no change in polyprenal levels, suggesting DHRSX deficiency. Accordingly, N-glycan synthesis and changes in polyisoprenoid levels were corrected by complementation with human DHRSX but not with SRD5A3. Furthermore, the typical polyprenol dehydrogenase and dolichal reductase activities of DHRSX were absent in membrane preparations derived from Lec5 and Lec9 cells, while the reduction of polyprenal to dolichal, catalyzed by SRD5A3, was unaffected. Long-read whole genome sequencing of Lec5 and Lec9 cells did not reveal mutations in the ORF of SRD5A3, but the genomic region containing DHRSX was absent. Lastly, we established the sequence of Chinese hamster DHRSX and validated that this protein has similar kinetic properties to the human enzyme. Our work therefore identifies the basis of the dolichol synthesis defect in CHO Lec5 and Lec9 cells.

A pseudoautosomal glycosylation disorder prompts the revision of dolichol biosynthesis.

Dolichol is a lipid critical for N-glycosylation as a carrier for activated sugars and nascent oligosaccharides. It is commonly thought to be directly produced from polyprenol by the enzyme SRD5A3. Instead, we found that dolichol synthesis requires a three-step detour involving additional metabolites, where SRD5A3 catalyzes only the second reaction. The first and third steps are performed by DHRSX, whose gene resides on the pseudoautosomal regions of the X and Y chromosomes. Accordingly, we report a pseudoautosomal-recessive disease presenting as a congenital disorder of glycosylation in patients with missense variants in DHRSX (DHRSX-CDG). Of note, DHRSX has a unique dual substrate and cofactor specificity, allowing it to act as a NAD+-dependent dehydrogenase and as a NADPH-dependent reductase in two non-consecutive steps. Thus, our work reveals unexpected complexity in the terminal steps of dolichol biosynthesis. Furthermore, we provide insights into the mechanism by which dolichol metabolism defects contribute to disease.

SUZ domain-containing proteins have multiple effects on nonsense-mediated decay target transcripts.

Many transcripts are targeted by nonsense-mediated decay (NMD), leading to their degradation and the inhibition of their translation. We found that the protein SUZ domain-containing protein 1 (SZRD1) interacts with the key NMD factor up-frameshift 1. When recruited to NMD-sensitive reporter gene transcripts, SZRD1 increased protein production, at least in part, by relieving translational inhibition. The conserved SUZ domain in SZRD1 was required for this effect. The SUZ domain is present in only three other human proteins besides SZRD1: R3H domain-containing protein 1 and 2 (R3HDM1, R3HDM2) and cAMP-regulated phosphoprotein 21 (ARPP21). We found that ARPP21, similarly to SZRD1, can increase protein production from NMD-sensitive reporter transcripts in an SUZ domain-dependent manner. This indicated that the SUZ domain-containing proteins could prevent translational inhibition of transcripts targeted by NMD. Consistent with the idea that SZRD1 mainly prevents translational inhibition, we did not observe a systematic decrease in the abundance of NMD targets when we knocked down SZRD1. Surprisingly, knockdown of SZRD1 in two different cell lines led to reduced levels of the NMD component UPF3B, which was accompanied by increased levels in a subset of NMD targets. This suggests that SZRD1 is required to maintain normal UPF3B levels and indicates that the effect of SZRD1 on NMD targets is not limited to a relief from translational inhibition. Overall, our study reveals that human SUZ domain-containing proteins play a complex role in regulating protein output from transcripts targeted by NMD.

Insights into energy balance dysregulation from a mouse model of methylmalonic aciduria.

Inherited disorders of mitochondrial metabolism, including isolated methylmalonic aciduria, present unique challenges to energetic homeostasis by disrupting energy-producing pathways. To better understand global responses to energy shortage, we investigated a hemizygous mouse model of methylmalonyl-CoA mutase (Mmut)-type methylmalonic aciduria. We found Mmut mutant mice to have reduced appetite, energy expenditure and body mass compared with littermate controls, along with a relative reduction in lean mass but increase in fat mass. Brown adipose tissue showed a process of whitening, in line with lower body surface temperature and lesser ability to cope with cold challenge. Mutant mice had dysregulated plasma glucose, delayed glucose clearance and a lesser ability to regulate energy sources when switching from the fed to fasted state, while liver investigations indicated metabolite accumulation and altered expression of peroxisome proliferator-activated receptor and Fgf21-controlled pathways. Together, these shed light on the mechanisms and adaptations behind energy imbalance in methylmalonic aciduria and provide insight into metabolic responses to chronic energy shortage, which may have important implications for disease understanding and patient management.

Duodenal CD8+ T resident memory cell apoptosis contributes to gut barrier dysfunction and microbial translocation in early alcohol-associated liver disease in humans.

BACKGROUND: Intestinal T cells are key in gut barrier function. Their role in early stages of alcohol-associated liver disease (ALD) remain unknown. AIM: To explore the links between intestinal T cells, microbial translocation and ALD METHODS: Patients with alcohol use disorder (AUD) following a rehabilitation programme were compared to subjects with non-alcoholic fatty liver disease (NAFLD) and healthy controls. Clinical and laboratory data (liver stiffness, controlled attenuation parameter, AST, ALT, K18-M65) served to identify AUD patients with isolated steatosis (minimal liver disease) or steatohepatitis/fibrosis (ALD). Serum microbial translocation markers were measured by ELISA, duodenal and plasma levels of sphingolipids by targeted LC-MS. T lymphocytes in duodenal biopsies were characterised by immunohistochemistry, flow cytometry and RNA sequencing on FACS-sorted cells. Mechanisms for T-cell alterations were assessed in vitro. RESULTS: Patients with ALD, but not those with minimal liver disease, showed reduced numbers of duodenal CD8+ T resident memory (TRM) cells compared to controls or patients with NAFLD. TRM transcriptomic analysis, in vitro analyses and pharmacological inhibition of cathepsin B confirmed TRM apoptosis driven by lysosomal membrane permeabilisation and cathepsin B release into the cytosol. Altered lipid metabolism and increased duodenal and plasma sphingolipids correlated with apoptosis. Dihydroceramide dose-dependently reduced viability of TRM. Duodenal TRM phenotypic changes, apoptosis and transcriptomic alterations correlated with increased levels of microbial translocation markers. Short-term abstinence did not reverse TRM cell death in patients with ALD. CONCLUSIONS: Duodenal CD8+ TRM apoptosis related to functional changes in lysosomes and lipid metabolism points to impaired gut adaptive immunity specifically in patients with AUD who developed early ALD.

Human cytosolic transaminases: side activities and patterns of discrimination towards physiologically available alternative substrates.

Transaminases play key roles in central metabolism, transferring the amino group from a donor substrate to an acceptor. These enzymes can often act, with low efficiency, on compounds different from the preferred substrates. To understand what might have shaped the substrate specificity of this class of enzymes, we examined the reactivity of six human cytosolic transaminases towards amino acids whose main degradative pathways do not include any transamination. We also tested whether sugars and sugar phosphates could serve as alternative amino group acceptors for these cytosolic enzymes. Each of the six aminotransferases reacted appreciably with at least three of the alternative amino acid substrates in vitro, albeit at usually feeble rates. Reactions with L-Thr, L-Arg, L-Lys and L-Asn were consistently very slow-a bias explained in part by the structural differences between these amino acids and the preferred substrates of the transaminases. On the other hand, L-His and L-Trp reacted more efficiently, particularly with GTK (glutamine transaminase K; also known as KYAT1). This points towards a role of GTK in the salvage of L-Trp (in cooperation with ω-amidase and possibly with the cytosolic malate dehydrogenase, MDH1, which efficiently reduced the product of L-Trp transamination). Finally, the transaminases were extremely ineffective at utilizing sugars and sugar derivatives, with the exception of the glycolytic intermediate dihydroxyacetone phosphate, which was slowly but appreciably transaminated by some of the enzymes to yield serinol phosphate. Evidence for the formation of this compound in a human cell line was also obtained. We discuss the biological and evolutionary implications of our results.

Inhibition of basal and glucagon-induced hepatic glucose production by 991 and other pharmacological AMPK activators.

Pharmacological AMPK activation represents an attractive approach for the treatment of type 2 diabetes (T2D). AMPK activation increases skeletal muscle glucose uptake, but there is controversy as to whether AMPK activation also inhibits hepatic glucose production (HGP) and pharmacological AMPK activators can have off-target effects that contribute to their anti-diabetic properties. The main aim was to investigate the effects of 991 and other direct AMPK activators on HGP and determine whether the observed effects were AMPK-dependent. In incubated hepatocytes, 991 substantially decreased gluconeogenesis from lactate, pyruvate and glycerol, but not from other substrates. Hepatocytes from AMPKß1-/- mice had substantially reduced liver AMPK activity, yet the inhibition of glucose production by 991 persisted. Also, the glucose-lowering effect of 991 was still seen in AMPKß1-/- mice subjected to an intraperitoneal pyruvate tolerance test. The AMPK-independent mechanism by which 991 treatment decreased gluconeogenesis could be explained by inhibition of mitochondrial pyruvate uptake and inhibition of mitochondrial sn-glycerol-3-phosphate dehydrogenase-2. However, 991 and new-generation direct small-molecule AMPK activators antagonized glucagon-induced gluconeogenesis in an AMPK-dependent manner. Our studies support the notion that direct pharmacological activation of hepatic AMPK as well as inhibition of pyruvate uptake could be an option for the treatment of T2D-linked hyperglycemia.

Urine metabolomics links dysregulation of the tryptophan-kynurenine pathway to inflammation and severity of COVID-19.

SARS-CoV-2 causes major disturbances in serum metabolite levels, associated with severity of the immune response. Despite the numerous advantages of urine for biomarker discovery, the potential association between urine metabolites and disease severity has not been investigated in coronavirus disease 2019 (COVID-19). In a proof-of-concept study, we performed quantitative urine metabolomics in patients hospitalized with COVID-19 and controls using LC-MS/MS. We assessed whether metabolites alterations were associated with COVID-19, disease severity, and inflammation. The study included 56 patients hospitalized with COVID-19 (26 non-critical and 30 critical disease); 16 healthy controls; and 3 controls with proximal tubule dysfunction unrelated to SARS-CoV-2. Metabolomic profiling revealed a major urinary increase of tryptophan metabolites kynurenine (P < 0.001), 3-hydroxykynurenine (P < 0.001) and 3-hydroxyanthranilate (P < 0.001) in SARS-CoV-2 infected patients. Urine levels of kynurenines were associated with disease severity and systemic inflammation (kynurenine, r 0.43, P = 0.001; 3-hydroxykynurenine, r 0.44, P < 0.001). Increased urinary levels of neutral amino acids and imino acid proline were also common in COVID-19, suggesting specific transport defects. Urine metabolomics identified major alterations in the tryptophan-kynurenine pathway, consistent with changes in host metabolism during SARS-CoV-2 infection. The association between increased urinary levels of kynurenines, inflammation and COVID-19 severity supports further evaluation of these easily available biomarkers.

Parkinson's disease protein PARK7 prevents metabolite and protein damage caused by a glycolytic metabolite.

Cells are continuously exposed to potentially dangerous compounds. Progressive accumulation of damage is suspected to contribute to neurodegenerative diseases and aging, but the molecular identity of the damage remains largely unknown. Here we report that PARK7, an enzyme mutated in hereditary Parkinson's disease, prevents damage of proteins and metabolites caused by a metabolite of glycolysis. We found that the glycolytic metabolite 1,3-bisphosphoglycerate (1,3-BPG) spontaneously forms a novel reactive intermediate that avidly reacts with amino groups. PARK7 acts by destroying this intermediate, thereby preventing the formation of proteins and metabolites with glycerate and phosphoglycerate modifications on amino groups. As a consequence, inactivation of PARK7 (or its orthologs) in human cell lines, mouse brain, and Drosophila melanogaster leads to the accumulation of these damaged compounds, most of which have not been described before. Our work demonstrates that PARK7 function represents a highly conserved strategy to prevent damage in cells that metabolize carbohydrates. This represents a fundamental link between metabolism and a type of cellular damage that might contribute to the development of Parkinson's disease.

ECHDC1 knockout mice accumulate ethyl-branched lipids and excrete abnormal intermediates of branched-chain fatty acid metabolism.

The cytosolic enzyme ethylmalonyl-CoA decarboxylase (ECHDC1) decarboxylates ethyl- or methyl-malonyl-CoA, two side products of acetyl-CoA carboxylase. These CoA derivatives can be used to synthesize a subset of branched-chain fatty acids (FAs). We previously found that ECHDC1 limits the synthesis of these abnormal FAs in cell lines, but its effects in vivo are unknown. To further evaluate the effects of ECHDC1 deficiency, we generated knockout mice. These mice were viable, fertile, showed normal postnatal growth, and lacked obvious macroscopic and histologic changes. Surprisingly, tissues from wild-type mice already contained methyl-branched FAs due to methylmalonyl-CoA incorporation, but these FAs were only increased in the intraorbital glands of ECHDC1 knockout mice. In contrast, ECHDC1 knockout mice accumulated 16-20-carbon FAs carrying ethyl-branches in all tissues, which were undetectable in wild-type mice. Ethyl-branched FAs were incorporated into different lipids, including acylcarnitines, phosphatidylcholines, plasmanylcholines, and triglycerides. Interestingly, we found a variety of unusual glycine-conjugates in the urine of knockout mice, which included adducts of ethyl-branched compounds in different stages of oxidation. This suggests that the excretion of potentially toxic intermediates of branched-chain FA metabolism might prevent a more dramatic phenotype in these mice. Curiously, ECHDC1 knockout mice also accumulated 2,2-dimethylmalonyl-CoA. This indicates that the broad specificity of ECHDC1 might help eliminate a variety of potentially dangerous branched-chain dicarboxylyl-CoAs. We conclude that ECHDC1 prevents the formation of ethyl-branched FAs and that urinary excretion of glycine-conjugates allows mice to eliminate potentially deleterious intermediates of branched-chain FA metabolism.

Impaired glucose-1,6-biphosphate production due to bi-allelic PGM2L1 mutations is associated with a neurodevelopmental disorder.

We describe a genetic syndrome due to PGM2L1 deficiency. PGM2 and PGM2L1 make hexose-bisphosphates, like glucose-1,6-bisphosphate, which are indispensable cofactors for sugar phosphomutases. These enzymes form the hexose-1-phosphates crucial for NDP-sugars synthesis and ensuing glycosylation reactions. While PGM2 has a wide tissue distribution, PGM2L1 is highly expressed in the brain, accounting for the elevated concentrations of glucose-1,6-bisphosphate found there. Four individuals (three females and one male aged between 2 and 7.5 years) with bi-allelic inactivating mutations of PGM2L1 were identified by exome sequencing. All four had severe developmental and speech delay, dysmorphic facial features, ear anomalies, high arched palate, strabismus, hypotonia, and keratosis pilaris. Early obesity and seizures were present in three individuals. Analysis of the children's fibroblasts showed that glucose-1,6-bisphosphate and other sugar bisphosphates were markedly reduced but still present at concentrations able to stimulate phosphomutases maximally. Hence, the concentrations of NDP-sugars and glycosylation of the heavily glycosylated protein LAMP2 were normal. Consistent with this, serum transferrin was normally glycosylated in affected individuals. PGM2L1 deficiency does not appear to be a glycosylation defect, but the clinical features observed in this neurodevelopmental disorder point toward an important but still unknown role of glucose-1,6-bisphosphate or other sugar bisphosphates in brain metabolism.

The promiscuous binding pocket of SLC35A1 ensures redundant transport of CDP-ribitol to the Golgi.

The glycoprotein α-dystroglycan helps to link the intracellular cytoskeleton to the extracellular matrix. A unique glycan structure attached to this protein is required for its interaction with extracellular matrix proteins such as laminin. Up to now, this is the only mammalian glycan known to contain ribitol phosphate groups. Enzymes in the Golgi apparatus use CDP-ribitol to incorporate ribitol phosphate into the glycan chain of α-dystroglycan. Since CDP-ribitol is synthesized in the cytoplasm, we hypothesized that an unknown transporter must be required for its import into the Golgi apparatus. We discovered that CDP-ribitol transport relies on the CMP-sialic acid transporter SLC35A1 and the transporter SLC35A4 in a redundant manner. These two transporters are closely related, but bulky residues in the predicted binding pocket of SLC35A4 limit its size. We hypothesized that the large binding pocket SLC35A1 might accommodate the bulky CMP-sialic acid and the smaller CDP-ribitol, whereas SLC35A4 might only accept CDP-ribitol. To test this, we expressed SLC35A1 with mutations in its binding pocket in SLC35A1 KO cell lines. When we restricted the binding site of SLC35A1 by introducing the bulky residues present in SLC35A4, the mutant transporter was unable to support sialylation of proteins in cells but still supported ribitol phosphorylation. This demonstrates that the size of the binding pocket determines the substrate specificity of SLC35A1, allowing a variety of cytosine nucleotide conjugates to be transported. The redundancy with SLC35A4 also explains why patients with SLC35A1 mutations do not show symptoms of α-dystroglycan deficiency.

The microRNA miR-33 is a pleiotropic regulator of metabolic and developmental processes in Drosophila melanogaster.

BACKGROUND: miR-33 family members are well characterized regulators of cellular lipid levels in mammals. Previous studies have shown that overexpression of miR-33 in Drosophila melanogaster leads to elevated triacylglycerol (TAG) levels in certain contexts. Although loss of miR-33 in flies causes subtle defects in larval and adult ovaries, the effects of miR-33 deficiency on lipid metabolism and other phenotypes impacted by metabolic state have not yet been characterized. RESULTS: We found that loss of miR-33 predisposes flies to elevated TAG levels, and we identified genes involved in TAG synthesis as direct targets of miR-33, including atpcl, midway, and Akt1. miR-33 mutants survived longer upon starvation but showed greater sensitivity to an oxidative stressor. We also found evidence that miR-33 is a negative regulator of cuticle pigmentation and that miR-33 mutants show a reduction in interfollicular stalk cells during oogenesis. CONCLUSION: Our data suggest that miR-33 is a conserved regulator of lipid homeostasis, and its targets are involved in both degradation and synthesis of fatty acids and TAG. The constellation of phenotypes involving tissues that are highly sensitive to metabolic state suggests that miR-33 serves to prevent extreme fluctuations in metabolically sensitive tissues.

The metalloprotein YhcH is an anomerase providing N-acetylneuraminate aldolase with the open form of its substrate.

N-acetylneuraminate (Neu5Ac), an abundant sugar present in glycans in vertebrates and some bacteria, can be used as an energy source by several prokaryotes, including Escherichia coli. In solution, more than 99% of Neu5Ac is in cyclic form (≈92% beta-anomer and ≈7% alpha-anomer), whereas <0.5% is in the open form. The aldolase that initiates Neu5Ac metabolism in E. coli, NanA, has been reported to act on the alpha-anomer. Surprisingly, when we performed this reaction at pH 6 to minimize spontaneous anomerization, we found NanA and its human homolog NPL preferentially metabolize the open form of this substrate. We tested whether the E. coli Neu5Ac anomerase NanM could promote turnover, finding it stimulated the utilization of both beta and alpha-anomers by NanA in vitro. However, NanM is localized in the periplasmic space and cannot facilitate Neu5Ac metabolism by NanA in the cytoplasm in vivo. We discovered that YhcH, a cytoplasmic protein encoded by many Neu5Ac catabolic operons and belonging to a protein family of unknown function (DUF386), also facilitated Neu5Ac utilization by NanA and NPL and displayed Neu5Ac anomerase activity in vitro. YhcH contains Zn, and its accelerating effect on the aldolase reaction was inhibited by metal chelators. Remarkably, several transition metals accelerated Neu5Ac anomerization in the absence of enzyme. Experiments with E. coli mutants indicated that YhcH expression provides a selective advantage for growth on Neu5Ac. In conclusion, YhcH plays the unprecedented role of providing an aldolase with the preferred unstable open form of its substrate.

The putative Escherichia coli dehydrogenase YjhC metabolises two dehydrated forms of N-acetylneuraminate produced by some sialidases.

Homologues of the putative dehydrogenase YjhC are found in operons involved in the metabolism of N-acetylneuraminate (Neu5Ac) or related compounds. We observed that purified recombinant YjhC forms Neu5Ac from two dehydrated forms of this compound, 2,7-anhydro-N-acetylneuraminate (2,7-AN) and 2-deoxy-2,3-didehydro-N-acetylneuraminate (2,3-EN) that are produced during the degradation of sialoconjugates by some sialidases. The conversion of 2,7-AN into Neu5Ac is reversible and reaches its equilibrium when the ratio of 2,7-AN to Neu5Ac is ≈1/6. The conversion of 2,3-EN is irreversible, leading to a mixture of Neu5Ac and 2,7-AN. NMR analysis of the reaction catalysed by YjhC on 2,3-EN indicated that Neu5Ac was produced as the α-anomer. All conversions require NAD+ as a cofactor, which is regenerated in the reaction. They appear to involve the formation of keto (presumably 4-keto) intermediates of 2,7-AN, 2,3-EN and Neu5Ac, which were detected by liquid chromatography-mass spectrometry (LC-MS). The proposed reaction mechanism is reminiscent of the one catalysed by family 4 ß-glycosidases, which also use NAD+ as a cofactor. Both 2,7-AN and 2,3-EN support the growth of Escherichia coli provided the repressor NanR, which negatively controls the expression of the yjhBC operons, has been inactivated. Inactivation of either YjhC or YjhB in NanR-deficient cells prevents the growth on 2,7-AN and 2,3-EN. This confirms the role of YjhC in 2,7-AN and 2,3-EN metabolism and indicates that transport of 2,7-AN and 2,3-EN is carried out by YjhB, which is homologous to the Neu5Ac transporter NanT.

Brief O2 uploading during continuous hypothermic machine perfusion is simple yet effective oxygenation method to improve initial kidney function in a porcine autotransplant model.

With oxygenation proposed as a resuscitative measure during hypothermic models of preservation, the aim of this study was to evaluate the optimal start time of oxygenation during continuous hypothermic machine perfusion (HMP). In this porcine ischemia-reperfusion autotransplant model, the left kidney of a ±40 kg pig was exposed to 30 minutes of warm ischemia prior to 22 hours of HMP and autotransplantation. Kidneys were randomized to receive 2 hours of oxygenation during HMP either at the start (n = 6), or end of the perfusion (n = 5) and outcomes were compared to standard, nonoxygenated HMP (n = 6) and continuous oxygenated HMP (n = 8). The brief initial and continuous oxygenated HMP groups were associated with superior graft recovery compared to either standard, nonoxygenated HMP or kidneys oxygenated at the end of HMP. This correlated with significant metabolic differences in perfusate (eg, lactate, succinate, flavin mononucleotide) and tissues (eg, succinate, adenosine triphosphate, hypoxia-inducible factor-1α, nuclear factor erythroid 2-related factor 2) suggesting superior mitochondrial preservation with initial oxygenation. Brief initial O2 uploading during HMP at procurement site might be an easy and effective preservation strategy to maintain aerobic metabolism, protect mitochondria, and achieve an improved early renal graft function compared with standard HMP or oxygen supply shortly at the end of HMP preservation.

Metabolite Repair Enzymes Control Metabolic Damage in Glycolysis.

Hundreds of metabolic enzymes work together smoothly in a cell. These enzymes are highly specific. Nevertheless, under physiological conditions, many perform side-reactions at low rates, producing potentially toxic side-products. An increasing number of metabolite repair enzymes are being discovered that serve to eliminate these noncanonical metabolites. Some of these enzymes are extraordinarily conserved, and their deficiency can lead to diseases in humans or embryonic lethality in mice, indicating their central role in cellular metabolism. We discuss how metabolite repair enzymes eliminate glycolytic side-products and prevent negative interference within and beyond this core metabolic pathway. Extrapolating from the number of metabolite repair enzymes involved in glycolysis, hundreds more likely remain to be discovered that protect a wide range of metabolic pathways.

Inborn errors of metabolite repair.

It is traditionally assumed that enzymes of intermediary metabolism are extremely specific and that this is sufficient to prevent the production of useless and/or toxic side-products. Recent work indicates that this statement is not entirely correct. In reality, enzymes are not strictly specific, they often display weak side activities on intracellular metabolites (substrate promiscuity) that resemble their physiological substrate or slowly catalyse abnormal reactions on their physiological substrate (catalytic promiscuity). They thereby produce non-classical metabolites that are not efficiently metabolised by conventional enzymes. In an increasing number of cases, metabolite repair enzymes are being discovered that serve to eliminate these non-classical metabolites and prevent their accumulation. Metabolite repair enzymes also eliminate non-classical metabolites that are formed through spontaneous (ie, not enzyme-catalysed) reactions. Importantly, genetic deficiencies in several metabolite repair enzymes lead to 'inborn errors of metabolite repair', such as L-2-hydroxyglutaric aciduria, D-2-hydroxyglutaric aciduria, 'ubiquitous glucose-6-phosphatase' (G6PC3) deficiency, the neutropenia present in Glycogen Storage Disease type Ib or defects in the enzymes that repair the hydrated forms of NADH or NADPH. Metabolite repair defects may be difficult to identify as such, because the mutated enzymes are non-classical enzymes that act on non-classical metabolites, which in some cases accumulate only inside the cells, and at rather low, yet toxic, concentrations. It is therefore likely that many additional metabolite repair enzymes remain to be discovered and that many diseases of metabolite repair still await elucidation.

Pyridoxamine-phosphate oxidases and pyridoxamine-phosphate oxidase-related proteins catalyze the oxidation of 6-NAD(P)H to NAD(P).

6-NADH and 6-NADPH are strong inhibitors of several dehydrogenases that may form spontaneously from NAD(P)H. They are known to be oxidized to NAD(P)+ by mammalian renalase, an FAD-linked enzyme mainly present in heart and kidney, and by related bacterial enzymes. We partially purified an enzyme oxidizing 6-NADPH from rat liver, and, surprisingly, identified it as pyridoxamine-phosphate oxidase (PNPO). This was confirmed by the finding that recombinant mouse PNPO oxidized 6-NADH and 6-NADPH with catalytic efficiencies comparable to those observed with pyridoxine- and pyridoxamine-5'-phosphate. PNPOs from Escherichia coli, Saccharomyces cerevisiae and Arabidopsis thaliana also displayed 6-NAD(P)H oxidase activity, indicating that this 'side-activity' is conserved. Remarkably, 'pyridoxamine-phosphate oxidase-related proteins' (PNPO-RP) from Nostoc punctiforme, A. thaliana and the yeast S. cerevisiae (Ygr017w) were not detectably active on pyridox(am)ine-5'-P, but oxidized 6-NADH, 6-NADPH and 2-NADH suggesting that this may be their main catalytic function. Their specificity profiles were therefore similar to that of renalase. Inactivation of renalase and of PNPO in mammalian cells and of Ygr017w in yeasts led to the accumulation of a reduced form of 6-NADH, tentatively identified as 4,5,6-NADH3, which can also be produced in vitro by reduction of 6-NADH by glyceraldehyde-3-phosphate dehydrogenase or glucose-6-phosphate dehydrogenase. As 4,5,6-NADH3 is not a substrate for renalase, PNPO or PNPO-RP, its accumulation presumably reflects the block in the oxidation of 6-NADH. These findings indicate that two different classes of enzymes using either FAD (renalase) or FMN (PNPOs and PNPO-RPs) as a cofactor play an as yet unsuspected role in removing damaged forms of NAD(P).