Biosynthesis and Insertion of Heme

Abstract

The red, iron containing tetrapyrrole heme is an essential cofactor of enzymes involved in the electron transport chain of energy generation and used for catalyzing chemically challenging reactions of the metabolism. It is also used for diatomic gas transport (O₂, CO, CO₂, NO, N₂O), catalysis, and detection. Multiple transcriptional regulators and transporters bind heme. This chapter focuses on the highly unusual pathways for heme biosynthesis and the integration of protoheme into target proteins. Today, three different biosynthetic routes for heme formation are known. The general precursor molecule of all tetrapyrroles 5-aminolevulinic acid is formed by two different pathways starting either with glutamyl-tRNA or succinyl-CoA and glycine. The conversion of 5-aminolevulinic acid to uroporphyrinogen III is common to all biosynthetic paths. Then the pathway branches to a classical route via protoporphyrin and two currently known alternative routes via coproporpyhrin III and siroheme. Various steps are catalyzed by up to three structurally unrelated enzymes. Finally, formed protoheme (heme b) gets actively inserted into proteins by the “Radical SAM” protein HemW. A detailed description of involved intermediates, enzymes, and their mechanisms are depicted below.

These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

1 Introduction

Hemes are red colored, iron containing porphyrins which belong to the class of tetrapyrroles. The word heme is derived from the Greek αἷμα haima which stands for “blood.” These molecules are all composed of four pyrrole rings connected by methine bridges. The centrally coordinated iron can be used as electron source or sink and usually switches between the Fe(II), Fe (III), and Fe (IV) state. In humans, defects of the enzymes involved in heme biosynthesis lead to severe diseases called porphyrias (Kaufholz et al. 2013b).

Overall, multiple different closed circular and open chain tetrapyrroles are known. Cyclic tetrapyrroles reveal characteristic reduction states of the ring system and typical metal irons chelated in the center (Dailey et al. 2017; Heinemann et al. 2008; Jahn and Jahn 2012; Layer et al. 2010). The pyrrole moieties are substituted by propionate, vinyl, acetate, and methyl groups.

Prominent members of the tetrapyrroles are the magnesium coordinating, green bacteriochlorophylls and chlorophylls (Bröcker et al. 2012; Wang and Grimm 2015), the light absorbing pigments of oxygenic and anoxygenic photosynthesis. The yellow nickel containing coenzyme F₄₃₀ plays an essential role in methanogenesis as part of the enzyme methyl coenzyme M (CoM) reductase (Moore et al. 2017; Zheng et al. 2016). Siroheme and heme d₁ are technically no hemes (Boss et al. 2017). Siroheme is an iron containing isobacteriochlorin and serves as prosthetic group of assimilatory sulfite and nitrite reductases. Heme d₁ is an iron containing dioxoisobacteriochlorin and the prosthetic group of the dissimilatory cd₁ nitrite reductase. The cobalt containing corrins of the cobalamin or vitamin B₁₂ class are part of methyltransferases, reductive dehalogenases, isomerases, and radical S-adensoyl-L-methionine (SAM) enzymes (Moore et al. 2017; Smith et al. 2018). However, the most ubiquitously found and versatile tetrapyrroles are hemes. They participate in various cellular processes including electron transfer, gas sensing and transport, signaling, catalysis, and transcriptional regulation.

2 Multiple Pathways for Heme Biosynthesis: An Overview

For many decades, a unique biosynthetic pathway for the biosynthesis of hemes in all organisms was proposed (Bogorad 1958; Bogorad and Granick 1953; Hoare and Heath 1958). First variations were observed in the late 1980s for the formation of the general precursor 5-aminolevulinc acid (ALA , (Jahn 1992). Originally, the presence of aminolevulinic acid synthase was assumed to be ubiquitous throughout all kingdoms of life (Figs. 1 and 2). But for plants and later on for most bacteria and archaea, a different so-called C₅-pathway, originating from the C₅ skeleton of glutamate and proceeding via a glutamyl-tRNA intermediate, was discovered (Czarnecki and Grimm 2013; Jahn et al. 1992). Later in the biosynthetic pathway, variation occurred at the enzymatic steps known to require molecular oxygen, the coproporphyrinogen III (HemF) and protoporphyrinogen IX oxidase (HemY) reactions (Fig. 2). For their anaerobic metabolism, bacteria had developed the “Radical SAM” enzyme coproporphyrinogen III dehydrogenase (HemN) and an electron chain coupled flavin protoporphyrinogen IX oxidase (HemG) (Boynton et al. 2009; Layer et al. 2004, 2005; Möbius et al. 2010). In the meantime, a third protoporphyrinogen IX oxidase (HemJ) was discovered in cyanobacteria (Kato et al. 2010; Skotnicova et al. 2018). With increasing access to multiple genomes, heme-synthesizing bacteria and archaea were discovered that lack genes for the enzymes of the late classical biosynthetic route to heme. Investigations from the late 1990s already proposed an alternative route via precorrin 2, an intermediate of the branch of tetrapyrrole biosynthesis toward cobalamin (B₁₂), F₄₃₀, siroheme, and heme d₁ (Ishida et al. 1998). Novel enzyme activities were measured in Desulfovibrio vulgaris; however, due to the isolation of the reaction products as oxidized ester, their exact nature was not determined with absolute certainty. It was demonstrated in 2006 for the archaeon Methanosarcina barkeri and in 2009 for D. vulgaris that heme can be synthesized from precorrin 2 (Buchenau et al. 2006; Lobo et al. 2009). A novel pathway via siroheme was proposed and demonstrated (Bali et al. 2011; Kuhner et al. 2016; Storbeck et al. 2009). Additionally, in many Gram-positive bacteria, genes for coproporphyrinogen III oxidases/dehydrogenases were missing. Recently, for these organisms a novel pathway with coproporphyrin III as intermediate was described (Dailey et al. 2015; Lobo et al. 2015). Here, a coproporphyrinogen III oxidase produces this tetrapyrrole which in turn is subjected to iron insertion and final decarboxylation to yield heme (Dailey et al. 2015; Hansson et al. 1997a; Hobbs et al. 2016, 2017; Lobo et al. 2014). Interestingly, the siroheme and coproporphyrin pathways to heme share the final step, the decarboxylation of Fe-coproporphyrin III (coproheme) to form heme (Lobo et al. 2014). In summary, today we know three different pathways for the formation of heme with typical intermediates protoporphyrin, siroheme and coproporphyrin (Fig. 2).

Fig. 1

Current knowledge about the biosynthesis of biologically active tetrapyrroles

Fig. 2

The threealternative pathways of heme biosynthesis. Typical intermediates and the end products are boxed

3 Two Routes for 5-Aminolevulinic Acid Formation

The C₅ compound ALA represents the general precursor of all known tetrapyrroles. Two different pathways for its formation are known. The most likely older pathway uses the C₅ skeleton of glutamate as precursor (Beale and Castelfranco 1973). Glutamate is loaded onto tRNA^Glu by glutamyl-tRNA synthetase, normally involved in protein biosynthesis (Schulze et al. 2006). An active site cysteine residue of glutamyl tRNA reductase (HemA, GtrR, GluTR) attacks the ester bond between the α-carbonyl of glutamate and tRNA^Glu with the formation of an enzyme-bound thioester intermediate (Fig. 3) and the release of free tRNA^Glu (Moser et al. 1999; Randau et al. 2004; Schauer et al. 2002). Hydride transfer from NADPH yields glutamate-1-semialdehyde (Lüer et al. 2007). Next, the pyridoxal-5′-phosphate/pyridoxamine-5′-phosphate-dependent glutamate-1-semialdehyde-2,1-aminomutase (HemL, GsaM) catalyzes the intramolecular transfer of an amino group using a modified aminotransferase mechanism (Grimm et al. 1992; Ilag and Jahn 1992). The product of both reactions is ALA. Currently, two crystal structures of GtrR and multiple structures of GsaM exist (Hennig et al. 1997; Li et al. 2018; Moser et al. 2001; Schulze et al. 2006; Zhao et al. 2014). Both enzymes form a stable channeling complex to protect the water-labile glutamate-1-semialdehyde intermediate (Lüer et al. 2005).

Fig. 3

Enzyme mechanisms of glutamyl-tRNA reductase (top) and 5-aminolevulinic acid synthase (bottom)

In a second, often termed Shemin pathway, 5-aminolevulinic acid synthase (HemA, AlaS) catalyzes the condensation of the C4 compound succinyl-CoA and the C2 amino acid glycine with elimination of CO₂ to form ALA (Fig. 2 (Gibson et al. 1958; Kikuchi et al. 1958; Shemin and Rittenberg 1945)). The pyridoxal-5′-phosphate-dependent enzyme proceeds after lysine binding through an internal aldimine, after pro-R-hydrogen abstraction a quinonoid I intermediate is formed prior to succinyl-CoA binding of the 2-amino-3-ketoadipate and after the release of coenzyme A the quinonoid II intermediates (Fig. 3). Decarboxylation and protonation yields ALA (Kaufholz et al. 2013a; Stojanovski et al. 2014). Two crystal structures of the enzyme are known (Astner et al. 2005; Brown et al. 2018).

4 The Common Part of Heme Biosynthesis from 5-Aminolevulinic Acid to Uroporphyrinogen III

The monopyrrole porphobilinogen is formed via the asymmetric condensation of two ALA molecules (Dresel and Falk 1953; Granick 1954) by porphobilinogen synthase (Fig. 2, HemB, PbgS). For this purpose, the enzyme contains two ALA binding sites, termed A and P sites, referring to the contribution to the acetate or propionate moiety of porphobilinogen (Fig. 4). ALA 1 is bound via a Schiff base to a P site lysine, while ALA2 is also bound to a lysine residue of the sometimes metal containing A site (Spencer and Jordan 1995). This second Schiff base in the A site is converted to an enamine. In an aldol addition reaction, the C3 of ALA2 attacks the C4 of ALA1 with the formation of a C-C bond. Now the amino group of ALA 1 targets the Schiff base at the C4 atom of ALA2 yielding a C-N bond with the release from the lysine (Fig. 4). Subsequently, lysis of the remaining bond to the other lysine and aromatization of the formed ring system leads to porphobilinogen (Frere et al. 2002; Jaffe 2004). Overall three different binding sites for metal ions have been detected. These sites are filled in multiple combinations by zinc and magnesium (Jaffe 2016). Multiple crystal structures for the usually octameric enzymes have been elucidated (Erskine et al. 1997; Frankenberg et al. 1999; Frere et al. 2005; Jaffe et al. 2000). For the human enzyme, an equilibrium of functionally distinct octameric, hexameric, and two different dimers was described (Breinig et al. 2003; Jaffe and Lawrence 2012). Interestingly, the PbgS from Rhodobacter capsulatus is a hexamer without any metals (Bollivar et al. 2004).

Fig. 4

Enzymatic reaction for the formation of porphobilinogen from two molecules of 5-aminolevulinic acid by porphobilinogen synthase

During the next two enzymatic steps, the first circular closed tetrapyrrole uroporphyrinogen III is formed by hydroxymethylbilane synthase (previously porphobilinogen deaminase, HemC, HmbS) and uroporphyrinogen III synthase (HemD, UroS). Originally, these reactions were believed to be catalyzed by one enzyme. However, at end of the 1960s, both activities were separated (Stevens and Frydman 1968). During the 1970s, the new intermediate hydroxymethylbilane (pre-uroporphyrinogen) was identified via ¹³C NMR and shown to be the substrate for uroporphyrinogen III synthase (Burton et al. 1979; Jordan and Seehra 1979). Hydroxymethylbilane synthase catalyzes the polymerization of four porphobilinogen pyrroles starting with ring A followed by rings B,C, and D of the final tetrapyrrole (Battersby et al. 1979; Jordan and Seehra 1979). In 1987, the existence of covalently attached dipyrromethane cofactor, formed by the enzyme from two molecules of porphobilinogen, serving as primer for the polymerization reaction, was shown (Azim et al. 2014; Hart et al. 1987; Jordan et al. 1988; Warren and Jordan 1988). During catalysis first the dipyrromethane cofactor is assembled once after ribosomal enzyme formation. Incoming porphobilinogens are deaminated and polymerized at the cofactor. The roles of conserved aspartic acid and arginine amino acid residues have been described (Bung et al. 2018; Pluta et al. 2018; Woodcock and Jordan 1994). Finally, the bond between the cofactor and the tetrapyrrole is hydrolyzed and hydroxymethylbilane is released (Fig. 5). In water, it would cyclize to fully symmetric uroporphyrinogen I, an inhibitor of the next enzyme, uroporphyrinogen III synthase. Multiple crystal structures are available (Azim et al. 2014; Gill et al. 2009; Roberts et al. 2013). Recently, a regulation of hydroxymethylbilane synthase by heme was proposed (Uchida et al. 2018).

Fig. 5

Assembly of hydroxymethylbilane at a dipyrromethane cofactor by hydroxymethylbilane synthase (top) and the enzymatic conversion of hydroxymethylbilane to uroporphyrinogen III with the inversion of ring D catalyzed by uroporphyrinogen III synthase (bottom)

Uroporphyrinogen III synthase catalyzes the cyclization of linear hydroxymethylbilane with the inversion of ring D to form the cyclic but asymmetric uroporphyrinogen III. Uroporphyrinogen III as an intermediate of heme biosynthesis was proposed in the 1950s (Bogorad 1958). The enzyme was characterized and a mechanism proposed in the 1960s (Levin 1968; Mathewson and Corwin 1961). Catalysis starts with the loss of the hydroxyl group at ring A with the formation of the first azafulvene intermediate (Fig. 5). The reaction of the azafulvene with the substituted α-position of the D ring results in the formation of a spirocyclic pyrrolenine intermediate. A second azafulvene intermediate is formed on ring C by the breakage of the bond between rings C and D. This azafulvene reacts in the final steps with the free a-position before deprotonation, and rearrangement leads to the formation of uroporphyrinogen III (Hawker et al. 1998; Stark et al. 1985, 1986, 1993). Crystal structures for the enzyme have been published (Mathews et al. 2001; Peng et al. 2011; Schubert et al. 2008). Due to the low degree of amino acid conservation, the gene for the protein of a plant enzyme was discovered decades later (Tan et al. 2008).

5 The Classical Pathway via Protoporphyrin IX

Here, uroporphyrinogen III is first subject to a stepwise decarboxylation of each of the four pyrrole ring acyl side chains at the C2, C7, C12, and C18 to four methyl groups (Mauzerall and Granick 1958). The reaction is catalyzed by uroporphyrinogen III decarboxylase (HemE, UroD), which starts at ring D and proceeds clockwise via ring A and B to ring C (Jackson et al. 1976). The enzyme does not require any associated cofactor. The enzyme is a homodimer with juxtaposed and facing active sites. Two major different models were proposed for enzyme activity. In the first model, the homodimer shuttles a single substrate molecule forth and back between both active sites without solvent contact (Phillips et al. 2003; Whitby et al. 1998). In the second model, solely one active site of one subunit is used, the substrate is rotated by 90° following each decarboxylation. Interestingly, a dimer of an active and an inactive uroporphyrinogen III decarboxylase subunit is still able to perform the complete reaction pointing toward model 2 (Phillips et al. 2009). Several crystal structures have been reported (Fan et al. 2007; Martins et al. 2001; Phillips et al. 2003; Whitby et al. 1998). A uroporphyrinogen III decarboxylase structure with the product coproporphyrinogen III revealed that the substrate/product binds in a dome-shaped structure with the four NH groups facing a 2x hydrogen bond distance to a conserved aspartate residue. Furthermore, three conserved arginines, one conserved histidine, and one tyrosine residues were proposed to be involved in catalysis. One of the arginine residues was identified as the general acid catalyst. Upon its protonation the decarboxylation reaction becomes rate-limiting (Silva et al. 2010). Regardless of the exact mechanism, the enzyme was suggested as a “benchmark” for catalytic proficiency among enzymes without cofactors due to a calculated enzyme enhancement value of the various decarboxylation reaction of around 10¹⁷ (Lewis and Wolfenden 2008).

During the next biosynthetic step toward the formation of heme, the propionate side chain of coproporphyrinogen III ring A and B undergoes an oxidative decarboxylation to the corresponding vinyl groups by two different, structurally not related enzymes. The oxygen-dependent cofactor-free coproporphyrinogen III oxidase (HemF, CpgC) and the radical SAM enzyme coproporphyrinogen III dehydrogenase (HemN, CgdH) are catalyzing the conversion of coproporphyrinogen III into protoporphyrinogen IX. The general reaction was discovered in the 1950s (Granick and Mauzerall 1958) and the corresponding oxidase enzyme was described shortly thereafter (Sano and Granick 1961). First, it was shown that coproporphyrinogen III oxidase catalyzes the decarboxylation of ring A prior to that of ring B with the intermediate haderoporphyrinogen (Cavaleiro et al. 1974; Elder and Evans 1978). A mechanism for the oxygen-dependent reaction by the cofactor-free enzyme was proposed (Lash 2005; Silva and Ramos 2008). During the first step a base catalyzed deprotonation of the pyrrole NH-group yields an azacyclopentadienyl anion, which reacts with molecular oxygen at the α-position to form a pyrrole peroxide anion. Next, a proton at the β-positon of the substrate gets abstracted by the peroxide via a six-membered ring transition state with the formation of an exocyclic double bond (Fig. 6). Elimination of CO₂ and H₂O₂ with the following bond rearrangements results in the formation of the vinyl group of the product protoporphyrinogen IX (Breckau et al. 2003). Solved crystal structures were without substrate (Lee et al. 2005; Phillips et al. 2004). Nevertheless, an aspartate and two conserved arginine residues were proposed to be in catalysis and substrate binding (Stephenson et al. 2007).

Fig. 6

Enzymatic conversion ofcoproporphyrinogen III in coproporphyrinogen III by the coproporphyrinogen III oxidase (top) and the coproporphyrinogen III dehydrogenase (bottom)

Many bacteria possess anaerobic heme containing respiratory chains for energy generation. Consequently, an oxygen-independent coproporphyrinogen III conversion was needed. At the end of the 1960s Tait described such enzyme activity for Rhodobacter sphaeroides (Tait 1969, 1972). Around 10 years later the identical stereochemistry of an initial pro-S-hydrogen abstraction at the β-carbon as observed for the oxygen-dependent catalysis was described for the coproporphyrinogen III dehydrogenase (Seehra et al. 1983). Similarly, haderoporphyrinogen was identified as reaction intermediate (Rand et al. 2010). Genetic approaches led to the isolation of the corresponding genes in the 1990s (Lieb et al. 1998; Troup et al. 1995; Xu and Elliott 1994). Intensive biochemical and structural analysis of recombinant coproporphyrinogen III dehydrogenase from Escherichia coli identified the protein as “Radical SAM” enzyme. It carries a [4Fe-4S] cluster coordinated by three cysteine residues and one S-adenosyl-L-methionine (SAM) molecule (Layer et al. 2002, 2005; Lieb et al. 1998; Troup et al. 1995; Xu and Elliott 1994). The reaction starts with the reduction of the [4Fe-4S] cluster by an unknown electron donor. The electron is subsequently transferred to SAM, which in turn undergoes a homolytic cleavage with the generation of methionine and a 5' deoxyadenosyl radical. The highly reactive radical abstracts stereo-specifically the pro-S-hydrogen at the β-carbon (Fig. 6). Finally, elimination of CO₂, transfer of the remaining electron to a yet unknown electron acceptor, and structural rearrangements with formation of the vinyl group finalize the reaction which has to occur twice (Layer et al. 2004, 2006). The crystal structure revealed the presence of two SAM molecules, which would allow a complete reaction cycle without release of a reaction intermediate (Layer et al. 2003).

The six electron oxidation of protoporphyrinogen IX to the red colored protoporphyrin IX is catalyzed by three distinct enzymes all named protoporphyrinogen IX oxidase (HemY or PgoX, HemG or PgdH1, HemJ or PgdH2). Under aerobic conditions the reaction can occur auto-catalytically. Already in the early 1960s the corresponding enzyme was discovered (Porra and Falk 1961, 1964; Sano and Granick 1961). The FAD-containing, oxygen-dependent PgoX is found in all heme-synthesizing eukaryotes and a few Gram-negative bacteria. Currently, the crystal structure of three enzymes are known (Corradi et al. 2006; Koch et al. 2004; Qi et al. 2002), however, without bound substrate or product. A model for substrate binding was verified via kinetic studies (Heinemann et al. 2007). Due to the FAD cofactor one can assume that the reaction proceeds via three two-electron steps from porphyrinogen via tetrahydro and dihydro intermediates to the fully oxidized porphyrin. Kinetic studies revealed that three meso-carbon hydride ions are removed in a sequential fashion with the concomitant removal of the NH proton (Akhtar 2003). Alternatively, it was proposed that all hydride abstractions occur at the C-20 meso-carbon including total ring hydrogen rearrangement via enamine-imine tautomerizations (Koch et al. 2004). Based on the PgoX structure, a complex with the following enzyme ferrochelatase was proposed and finally demonstrated (Koch et al. 2004; Masoumi et al. 2008).

Again, PgoX is oxygen-dependent, and bacteria require an alternative system for anaerobic heme biosynthesis. First description of an oxygen-independent system channeling the six abstracted electrons into respiratory chains was reported in the 1970s from Jacobs and Jacobs (Ishihara et al. 1995; Jacobs and Jacobs 1978; Jacobs et al. 1970, 1971). In E. coli, the corresponding gene hemG was mapped and the corresponding mutant used for cloning of the gene (Nishimura et al. 1995; Sasarman et al. 1993). A detailed biochemical characterization of PgdH1 followed at the beginning of this decade (Boynton et al. 2009; Möbius et al. 2010). The FMN-containing protein belongs to the class of long chain flavodoxins. It was shown that the abstracted six electrons are transferred via ubiquinone to terminal oxidases (Cyo, Cyd) under aerobic conditions and via menaquinone to nitrate and fumarate reductase under anaerobic conditions (Möbius et al. 2010). A similar mechanism as proposed for PgoX is also possible for PgdH1 (Fig. 7). Interestingly, the only eukaryotic organism utilizing PgdH1 is the Leishmania major (Zwerschke et al. 2014).

Fig. 7

Respiratory chain coupled, six electron oxidation of protoporphyrinogen IX to protoporphyrin IX by the FMN containing protoporphyrinogen IX oxidase (HemG, PgdH1)

For the third enzyme PgdH2 (HemJ), little is known. The corresponding hemJ gene was discovered in the cyanobacterium Synechocystis 6803 (Boynton et al. 2011; Kato et al. 2010). The corresponding membrane protein contained heme and was proposed to interact with coproporphyrinogen III oxidase (Skotnicova et al. 2018).

Protoporphyrin ferrochelatase (HemH, PpfC) catalyzes the final step of heme biosynthesis, the insertion of ferrous iron into protoporphyrin IX with the formation of protoheme (heme b). The first description of ferrochelatase from avian erythrocytes dates back to the 1950s (Ashenbrucker et al. 1956). Eukaryotic enzymes are usually membrane associated and contain with the exception of the plant enzymes one [2Fe-2S] cluster of unknown function per subunit (Shepherd et al. 2006). Bacterial enzymes are found with and without the cluster. The tetrapyrrole binds to the enzyme in an open conformation. Binding of the substrate triggers a rearrangement of a hydrogen bond network among conserved active site amino acid residues. Possibly, due to an abstraction of one pyrrole hydrogen a closed conformation is induced. Now the macrocycle is engulfed which causes an approximately 12 degree distortion of the bound tetrapyrrole. This distortion obviously facilitates metal chelation by the porphyrin with the simultaneous displacement of the second pyrrole hydrogen to a conserved histidine residue. The imidazole ring of the histidine moves, which causes structural rearrangements of the enzyme to adapt to the release conformation. Finally, heme b is released (Medlock et al. 2007, 2009; Sigfridsson and Ryde 2003; Wang et al. 2009, 2013).

6 The Alternative Pathway via Coproporphyrin III

Gram-positive bacteria utilize an alternative, only recently discovered pathway for heme biosynthesis. Coproporpyhrinogen III synthesized via the classical pathway gets oxidized to coproporphyrin III followed by metal insertion with the formation of Fe-coproheme III. Finally, the propionate side chains at ring A and B of Fe-coproheme III get decarboxylated to form heme b (Dailey et al. 2015; Hansson et al. 1997b; Hobbs et al. 2016, 2017; Lobo et al. 2015). Obviously, oxidation of the ring system of the porphyrinogen to form a porphyrin occurs at the level of coproporphyrinogen III instead of protoporphyrinogen IX as seen in the classical pathway. Consequently, iron insertion follows and the decarboxylation of ring A and B, as performed at the level of coproporphyrinogen by two different enzymes (CpgC, CgdH) in the classical pathway, utilizes the iron containing Fe-coproheme III in the novel pathway. The first committed step of the novel pathway, the oxidation of coproporpyhrinogen III to coproporphyrin III is performed by a coproporpyhrinogen III oxidase (HemY, CgoX). It was originally described for Bacillus subtilis as protoporphyrinogen oxidase (Corrigall et al. 1998; Hansson and Hederstedt 1994; Qin et al. 2010). However, already then it was reported that the enzyme catalyzes the oxidation of coproporpyhrinogen III to coproporphyrin III at a higher rate as the proposed protoporpyhrinogen oxidation (Han et al. 2013; Hansson et al. 1997b). With the discovery of the novel pathway this observation makes sense. The mechanism of the oxygen-dependent six electron oxidation performed by the FAD enzyme is highly similar to the oxygen-dependent protoporphyrinogen oxidase (FgoX) mechanism with the abstraction of six protons from the porphyrinogen and the formation of three molecules H₂O₂. However, the active site pocket of CgoXs is with 1173 ų is much larger than those of FgoX protoporphyrinogen IX oxidases with 527 to 440 ų and contains more positively charged surface areas (Qin et al. 2010). One explanation is the accommodation of the remaining propionate side chain containing coproporpyhrinogen III by CgoX. The oxygen-independent enzyme is currently unknown.

Next, coproporphyrinogen ferrochelatase (HemH, CpfC) catalyzes the insertion of ferrous iron into coproporpyhrin III to generate coproheme III. The best characterized enzyme is again from B. subtilis. It is water soluble and possesses a [2Fe-2S] cluster of unknown function. The mechanism is most likely highly similar to those of the ferrochelatase PpfC (Hansson et al. 2007; Karlberg et al. 2002; Lecerof et al. 2000, 2003; Olsson et al. 2002).

The last step of the novel pathway is the decarboxylation of coproheme III to heme b and was named coproheme decarboxylase or heme synthase (ChdC, HemQ) . An enzyme (AhbD) of identical catalysis is also part of the novel siroheme pathway for heme biosynthesis described below, but structurally not related to ChdC. While AhbD belongs to the family of “Radical SAM” enzymes, ChdC is a member of the chlorite dismutase family (Celis et al. 2015; Dailey et al. 2015; Pfanzagl et al. 2018). The structures of several incorrectly annotated as potential chlorite dismutase ChdCs were solved (PDB accession numbers 1T0T, 3DZT, 1VDH, 4WWS, SLOQ). The Listeria monocytogenes enzyme was co-crystallized with the product heme (Hofbauer et al. 2016). Two molecules of H₂O₂ are needed for catalysis (Celis et al. 2015; Hofbauer et al. 2014, 2016). During catalysis coproheme acts as substrate and cofactor. The coproheme ferric iron is coordinated by a conserved histidine residue. Under aerobic conditions O₂ gets to the coproheme iron bound and oxidizes it to the ferric state. The subsequent second-order reaction between the ferric complex and H₂O₂ is slow and pH-dependent. First evidence for ferryl porphyrin cation radical was obtained (Streit et al. 2018). A tyrosine, hydrogen bonding to the propionate at ring A, is essential for decarboxylation. It is proposed that an oxidizing equivalent from the most likely radical allows for the formation of a tyrosine radical which abstracts the hydrogen from the propionate side chain. Migration of the unpaired propionyl electron back to the coproheme would yield ferric haderoheme and CO₂. The propionate at ring B forms salt bridges to a lysine residue. Now, a similar pathway is proposed with this lysine as the essential proton shuttle for the second decarboxylation reaction (Celis et al. 2017). An alternative version of the ChdC protein called PitA, representing a fusion of ChdC with a monooxygenase domain, was described for the aerobic growth of Halferax volcanii (Kosugi et al. 2017).

7 The Second Alternative Pathway via Siroheme

The second alternative pathway starts directly at uroporphyrinogen III and uses the initial steps of cobalamin, F₄₃₀, heme d₁, and the complete siroheme biosynthesis (Bali et al. 2014; Kuhner et al. 2014). The pathway was originally proposed for archaea but subsequently found in multiple bacteria (Storbeck et al. 2010). The three reactions of siroheme formation are the SAM-dependent methylation of uroporphyrinogen II at position C-2 and C-7 to form precorrin-2 via precorrin-1, the following NAD-dependent ring dehydrogenation of precorrin-2 to form sirohydrochlorin and NADH, and the final iron insertion into sirohydrochlorin to yield siroheme. These three catalytic challenging reactions can be performed by one multifunctional enzyme called siroheme synthase (CysG) (Spencer et al. 1993; Warren et al. 1990). The enzyme contains two independent enzymatic modules, one for the methyltransferase and the other for the combined dehydrogenase-ferrochelatase function (Anderson et al. 2001; Lobo et al. 2009; Strey et al. 1999). Alternatively, the identical reactions are performed by three different enzymes, uroporphyrinogen-III C-methyltransferase (SUMT = NirE, CobA, SirA, Met1p) for the SAM-dependent methylation of uroporphyrinogen II at position C-2 and C-7 to form precorrin-2 via precorrin, the precorrin 2-dehydrogenase for the NAD-dependent ring dehydrogenation of precorrin-2 to form sirohydrochlorin, and the sirohydrochlorin ferrochelatase (SirB) for iron insertion. Interestingly, some bacteria carry a protein fusion of uroporphyrinogen III synthase and SUMT (Anderson et al. 2001; Lobo et al. 2009), most likely allowing direct channeling of the uroporphyrinogen III intermediate into the precorrin-2 pathway.

Uroporphyrinogen-III C-methyltransferase (SUMT = NirE, CobA, SirA, Met1p) was first described for cobalamin biosynthesis in Pseudomonas denitrificans (Blanche et al. 1989). The crystal structures of SUMT involved in cobalamin and heme d₁ biosynthesis were solved (Rehse et al. 2005; Storbeck et al. 2011; Vevodova et al. 2004). The NirE structure revealed the coordination of the tetrapyrrole by three arginines, a histidine, and a methionine residue. A mechanism induced by the arginine-mediated proton abstraction from the C-20 position was proposed (Storbeck et al. 2011). The subsequent movement of electrons facilitates the nucleophilic attack of C-2 at the methyl group of SAM. Upon proton abstraction from C-20 a new double bond between C-20 and C-1 is formed. After methyl transfer, rearrangements of double bonds within the macrocycle occur to form presorrin-1. After the first round of methylation the intermediate and side product SAH are released from the active site. Precorrin-1 and new SAM have to bind again to initiate a novel round of methylation at C-7 (Storbeck et al. 2011).

Precorrin-2 dehydrogenase (SirC, Met8p) catalyzes the NAD⁺-dependent oxidation of precorrin-2 (dipyrrocorphin) to form sirohydrochlorin (Raux et al. 2003; Schubert et al. 2008). The identical reaction is catalyzed by domains of the multifunctional CysG and Met8p. The crystal structures of SirC and Met8p were solved. Both enzymes were found to bind metals including Co(II) and Cu(II). It was proposed that SirC evolved from a Met8p-type protein via loosing its chelatase domain (Raux et al. 2003; Schubert et al. 2008).

Sirohydrochlorin ferrochelatase (SirB, Met8p) catalyzes the insertion of iron into sirohydrochlorin to form siroheme (Schubert et al. 2002). SirB solely catalyzes the iron chelation reaction. The multifunctional Met8p carries one active site with a catalytically important aspartate residue for both reactions, the dehydrogenase and chelatase reaction (Schubert et al. 2002).

The heterodimeric Siroheme decarboxylase (NirDLGH, AhbAB) catalyzes the conversion of siroheme to didecarboxysiroheme (Palmer et al. 2014). The two acyl side chains attached to C-12 and C-18 are decarboxylated to the corresponding methyl groups. Surprisingly, the crystal structure of the Desulfovibrio desulfuricans protein was solved and revealed structural similarity to proteins of the Asn/Lrp transcriptional regulator family proteins. A enzyme mechanism was proposed (Palmer et al. 2014). The coproheme (Fe-coproporphyrin) synthase (AhbC) converts didecarboxysiroheme into coproheme (Fe-coproporphyrin). Recombinant AhbC protein from Methanosarcina barkeri was shown to transform in the presence of SAM and the reducing agent dithionite 12,18-didecarboxysiroheme into Fe-coproporphyrin III (coproheme). Thus, the enzyme catalyzes the loss of the acyl side chains at C2 and C7 (Bali et al. 2011). The exact enzymatic mechanism remains to be determined. Heme synthase (AhbD) catalyzes the oxidative decarboxylation of coproheme into heme b (Kuhner et al. 2016). The protein belongs to the “Radical SAM” family of enzymes. A close mechanistic relationship to the coproporphyrinogen III dehydrogenase reaction can be assumed. Interestingly, the protein contains two [4Fe-4S] clusters. Besides the “classical” [4Fe-4S] cluster I involved in coproporphyrinogen III dehydrogenase type catalysis, the second auxiliary [4Fe-4S] cluster was identified to be involved in the electron transfer to the final electron acceptor of the reaction (Kuhner et al. 2016).

8 Heme b Insertion by the Heme Chaperone HemW

The “radical SAM” protein HemW revealed a high degree of amino acid sequence homology to coproporphyrinogen III dehydrogenases. However, the corresponding enzyme activity was never demonstrated for these proteins. A genetic investigation in Lactobacillus lactis indicated HemW’s participation in the generation of cytochromes. Subsequently, stable heme binding of the protein was demonstrated. Very recently the E. coli HemW was shown to be a heme chaperone involved in the active integration of heme b into the usually heme containing respiratory nitrate NarGHI which is involved in the anaerobic energy generation of the bacterium. The human counterpart RSAD1 was also shown to stably bind heme indicating its heme chaperone function (Abicht et al. 2012; Haskamp et al. 2018).

Like other radical SAM proteins, HemW contains three cysteines and one SAM coordinating a [4Fe-4S] cluster. The intact iron-sulfur cluster is required for HemW dimerization, which in turn causes membrane localization. The intact iron-sulfur cluster is not required for stable covalent heme binding. Bacterioferritins and the heme-containing subunit NarI of the respiratory nitrate reductase NarGHI were shown to interact directly with HemW. Bacterioferritins might serve as heme donors for HemW, while the cytochrome subunit NarI of nitrate reductase represents the target of HemW. During contact heme covalently bound to HemW gets actively transferred to a heme-depleted, catalytically inactive nitrate reductase, restoring its nitrate-reducing enzyme activity (Fig. 8). For the transfer process, an intact a [4Fe-4S] cluster is required. The exact mechanism of heme transfer remains to be determined (Abicht et al. 2012; Haskamp et al. 2018).

Fig. 8

Model of E. coli HemW activity. Heme from heme biosynthesis is most likely transferred via bacterioferritin (Bfr, light green) to the heme chaperone HemW (blue). The [4Fe-4S] cluster-containing HemW dimerizes and localizes to the membrane, where it interacts with its target protein NarI (yellow), a subunit of the respiratory nitrate reductase NarGHI. After heme incorporation into apo-NarI, the holo-NarGHI catalyzes the reduction of nitrate to nitrite

9 Research Needs

Some of the enzyme activities of the various alternative pathways, including the oxygen-independent coproporphyrinogen oxidase for the formation of coproporphyrin III, are unknown. Moreover, the exact nature of various enzymatic mechanisms and enzyme structures require further investigations. Possibly, additional alternative pathways shall be discovered.