Structure, Function, and Biosynthesis of the Coenzyme Methylofuran

Abstract

Methylotrophy is a metabolic trait that allows certain organisms to grow on carbon substrates without any C-C bonds, such as methanol or methane. The conversion and oxidation of these compounds usually proceeds via the toxic intermediate formaldehyde and is assisted by coenzymes that act as carriers for one-carbon units. Most methylotrophic bacteria rely on the tetrahydromethanopterin(H4MPT)-linked pathway for the oxidation of formaldehyde. This pathway involves not only H4MPT but also requires an analog of methanofuran (MFR), a coenzyme originally thought to be unique to methanogenic archaea. Previous biochemical and genetic evidence suggested that the structure of bacterial MFR must be similar to the one from methanogens; however, its purification and structural analysis remained challenging. In this thesis, the isolation and structural elucidation of MFR from the well-studied methylotroph Methylorubrum extorquens is described. Using preparative chromatography combined with high-resolution mass spectrometry and NMR, the bacterial analog of MFR—which was termed methylofuran (MYFR)—was identified and characterized. The core structure of MYFR was found to be identical to archaeal MFR, except for the presence of a tyrosine instead of a tyramine residue. Surprisingly, an unprecedented polyglutamate side chain consisting of up to 24 glutamate residues was connected to the tyrosine residue. NMR analysis further revealed that the glutamates in MYFR showed both α- and γ-linkages. In the H4MPT-linked pathway, MYFR is required by the formyltransferase/hydrolase complex (Fhc), where the coenzyme functions as a carrier of formyl units. To investigate whether the unusually large polyglutamate side chain of MYFR plays a role in the interaction with Fhc, the structure of the enzyme-coenzyme complex was solved at 3.1 Å. Interestingly, MYFR is bound as a non-covalent prosthetic group and the polyglutamate side chain tightly interacts with a large patch of positively charged residues of Fhc. This binding site is centrally located between the two active sites for formyl transfer and hydrolysis, thus suggesting that the polyglutamate chain functions as a flexible linker that allows the formyl-carrying aminomethylfuran moiety to reach both active sites of the bifunctional enzyme complex. Formyl units can thus be efficiently shuttled between the two active sites, without the need for MYFR to dissociate from Fhc. The electron density of Fhc-bound MYFR additionally revealed that the polyglutamate side chain of MYFR is branched, i.e. some glutamates are involved in isopeptide bonds with other glutamates. The branched polyglutamate structure might be required to support the strong interaction with Fhc and seems to be a unique feature of MYFR that is not present in archaeal MFRs. Since the H4MPT-linked pathway is widespread in Bacteria, MYFR is expected to be present in many strains. To determine whether there is structural diversity of MYFR, a survey comprising 12 proteobacterial strains was performed. Only in two strains, MYFR in the form present in M. extorquens was found. In six strains, a second type of MYFR was discovered which contained a tyramine instead of the tyrosine residue. For four strains, no MYFR could be identified. Interestingly, the number of glutamates in MYFR was not conserved across strains. While some had similar numbers as found in M. extorquens (around 16–20), two strains contained MYFR with 12 or fewer glutamates. The complex structure of the polyglutamate side chain of MYFR posed the question of its biosynthetic origin. In Proteobacteria, many genes essential for H4MPT-linked methylotrophy have previously been identified. For several of them, the function remained unknown. To identify genes involved in MYFR biosynthesis, strains with deletions in three of these genes (orf5, orfY, and orf17) were analyzed. All three mutants were unable to produce functional MYFR; however, the Dorf5 strain was accumulating MYFR-Glu2, a short intermediate of MYFR. Overexpression of orf5 in M. extorquens led to a significant increase in the number of glutamates attached to MYFR, as up to 40 glutamates were detected. The enzyme was thus renamed to MyfA, highlighting that this is the first enzyme discovered to be specifically involved in MYFR biosynthesis. In vitro assays with purified MyfA revealed de novo polyglutamate synthesis activity using L/D-glutamate and L-glutamine as substrates. Unexpectedly, L-glutamine was found to be an essential component of the assay. Assays with labeled glutamine showed that glutamine was serving as a source of glutamyl units for the incorporation into polyglutamates. The incorporation presumably took place after conversion to glutamate, as MyfA also showed glutaminase activity. Additionally, MyfA was able to cleave short glutamate containing peptides, thus also acting as a peptidase. These findings trigger the question of how the in vitro activities relate to MYFR biosynthesis in vivo. Taken together, the results obtained in this thesis shed light on various aspects of the structure, function, and biosynthesis of MYFR. They provide an in-depth understanding of the role MYFR plays in theH4MPTlinked pathway, thus expanding our knowledge about the biochemical basis of methylotrophy. The complex structure of MYFR that was revealed in this thesis, combined with the enzymes involved in its biosynthesis, will provide exciting opportunities for future research.