Monopyrroles as natural products

Heterocycles have played an important role in the evolution of life. Most coenzymes and vitamines contain heterocylces. The heterocyclic ring is often the central constituent necessary to fulfill the biological function of the coenzyme (see Figure 1).

Figure 1 Important coenzymes containing heterocycles.

This observation has led to the idea that coenzymes were formed early in the history of life (they may have been present during the prebiotic period of evolution). In this context it is remarkable that the pyridine ring as a typical representative of the electron poor heterocycles is present in cofactors like NAD, NADH. In contrast to the electron poor heterocycles which are widely present in nature the electron rich heterocycles like pyrroles, furans and thiophenes are found less often. Despite the fact that pyrrole belongs to the structurally simple five-membered heterocycles, relatively few natural products containing only one pyrrole ring are known (see Figure 2)1-4.

Figure 2 Monopyrrolic natural products.

Their biological functions are as varied as are their structures 1-65-7. Some natural pyrroles are pheromones 2 8,9, plant hormones 4 10 or act as antibiotics 6 11.

An important class of pyrrole derived natural products containing more than one pyrrole ring are neotropsin 7 and distamycin 8 which bind to the minor groove of DNA (see Figure 3) 12. They contain a series of pyrrole rings which are linked by amide bonds.

Figure 3 DNA-binders neotropsin 7 and distamycin 8.

In spite of the differences in the structure of these natural products most of them have one point in common: they are stabilised by an electron-withdrawing substituent or by another aromatic ring. Without these stabilising substituents the electron rich pyrrole ring is easily attacked. Under the influence of small amounts of acids unstabilised pyrroles are polymerised and autooxidised to give so called pyrrole black.

Porphobilinogen (= PBG 9) is a remarkable exception to this rule (see Figure 4). PBG 9 is a trisubstituted pyrrole which contains only alkyl substituents.

Figure 4 Porphobilinogen

The lack of substituents which can stabilise the pyrrole ring via conjugation is attributed to the high reactivity of PBG. The reactivity of PBG is used in the biosynthesis of the tetrapyrrolic pigments. More than 1010 tons of chlorophyll are synthesised each year 13,14. In order to avoid undesired side reactions due to the high reactivity of PBG only small concentrations of the pyrrole 9 are present in living organisms. Only if the normal biosynthetic pathway is severely disturbed do large quantities of PBG accumulate. In humans this accumulation leads to acute intermittent porphyria, a rare but severe illness 15,16.

PBG is a dedicated intermediate in the biosynthesis of tetrapyrroles 17-19. The tetrapyrrolic pigments 10 - 12 play an important role for central processes of live (see Figure 5). They are universally distributed and have therefore been named the "pigments of life" 20,21.

Figure 5 Some "pigments of life"

They function as indispensable cofactors for the transport of oxygen or of electrons (10), they harvest light and transform the energy of the photons into redox energy (11), the reduction of nitrate and sulfite (12)22,23.

The synthesis of porphobilinogen

The synthesis of PBG 9 has attracted the attention of chemists for different reasons: Initial synthetic efforts were undertaken to prove the structure 24,25; subsequently interest was mainly to synthesise PBG 9 labelled at specific positions for biosynthetic studies 26-28. Labelled and unlabelled PBG 9 or precursors thereof were used in the synthesis of pyrromethanes, tripyrrenes, bilanes and porphyrinogens 29-36. Despite the exorbitant price of PBG which is almost a thousand times that of gold 37 there have been only a limited number of fundamentally different approaches to PBG 9 or to analogues of porphobilinogen reported in the literature. Especially surprising is that since 1979 when the synthesis of PBG 9 was last reviewed, by Frydman, 38 very few new results have appeared.

Five synthetic strategies have been used for the synthesis of PBG 9 (see Figure 6).

Figure 6 Synthetic strategies for the synthesis of porphobilinogen

The first and historically the oldest strategy uses a classic Knorr synthesis to obtain a suitable precursor. To obtain the correct substitution pattern the MacDonald group has invested a considerable amount of work into the modification of the side chains obtained directly from the Knorr synthesis.

In the second strategy developed by Plieninger 39 and Evans 40 the pyrrole ring is formed by condensation of a C3-unit with a C-N-unit. In one case the variant of Kleinspehn of the Knorr synthesis is used, whereas in the second case the ring closure is achieved in a stepwise fashion. In this strategy both the acetic acid and the propionic acid side chains are in place right from the beginning.

The third strategy is due to Frydman and Rapoport 41. They started with a pyridine derivative, which they successfully transformed into a suitably substituted azaindole. Hydrogenation of the azaindole led to the porphobilinogen lactam, which could be hydrolysed to PBG 9.

The fourth strategy stems from Anderson and collaborators, who started with the unsubstituted pyrrole 42,43. Introducing step by step the acetic acid side chain in the b-position, the nitrile group as precursor of the methylamino group in the a-position and the propionic acid side chain in the b'-position finally gave PBG 9.

The fifth strategy was developed by the group of Adamckzyk and was published recently 44. The condensation of an a-acetoxynitro compound with benzyl isocyanoacetate is used to construct the scaffold of PBG.

Biosynthesis of porphobilinogen

All tetrapyrroles are derived from a common tetrapyrrolic precursor uroporphyrinogen III (14)45-47. Uroporphyrinogen III (14) itself is biosynthesised in three steps from eight molecules of 5-aminolevulinic acid (13) (see Figure 7).

Figure 7 Biosynthesis of uroporphyrinogen III

There are two major biosynthetic pathways to synthesise 5-aminolevulinic acid (13). The final structural complexity of the tetrapyrrolic pigments is produced modifying uroporphyrinogen III (14) by a series of chemical transformations. These transformations allow one to adapt the properties of the tetrapyrrole structure to the specific task it has to fulfil. The size of the coordination hole modulates the chelation properties of the ligand and most remarkably the redox potential of the metal complexes 48-50. Modification of the functional groups attached to the side chains of the disk-like macrocycle attributes hydrophilic or hydrophobic properties to the metal complexes. Esterifying one of the carboxylic acids with phytol (11) allows one to incorporate the complexes into membranes. The vinyl groups can also be used to covalently link the chelates to certain proteins (10).

The first step in the biosynthesis differs between plants and animals (see Figure 7). The final steps of the biosynthesis starting from uroporphyrinogen III (14) vary according to the pigment to be synthesised. In contrast, the central part of the biosynthesis is common to all organisms and for all biosynthetic pathways studied. The physiological importance of the tetrapyrroles and the fact that tetrapyrroles posses a dedicated biosynthetic pathway has stimulated interest in PBG (9).

The two central biosynthetic steps are highly attractive to the synthetic chemist. Using only one molecule as starting material a complex ligand is synthesised in a convergent manner. A mechanistic analysis shows that both processes, the dimeroidization of 5-aminolevulinic acid (13) to PBG and the tetrameroidization of PBG (9) to uroporphyrinogen III (14), belong to the general class of oligomeroidization. Both reactions are exothermic.

From a mechanistic point of view there is one big difference between the two processes. The tetrameroidization can easily be achieved in the absence of an enzyme, whereas it is difficult to realise the dimeroidization in a reaction vessel without enzyme.

Early studies of the reactivity showed that tetrameroidzation of PBG 9 could be achieved easily even without the help of an enzyme (see Figure 8) 55-54.

Figure 8 Tetrameroidization of PBG (9)

These observations immediatly raise the question of the mechanism of the transformation catalysed by porphobilinogen synthase (=PBGS) and of the comparison of the enzyme catalysed mechanism with its chemical analogue the Knorr pyrrole synthesis (see Figure 9)

Figure 9 Comparision between Knorr pyrrole synthesis and porphobilinogen biosynthesis

No high resolution X-ray structure of the enzyme PBGS is known so far. Despite the knowledge of almost twenty gene derived proteine sequences for PBGS from different sources, the sequence of events on the enzyme and the mechanistic details of the transformation are stilllargely unknown . The following experimental findings are relevant to the mechanism of PBGS (see Figure 10).

Figure 10 Experiments relevant to the mechanism of PBGS

The substrate forming the propionic acid side chain interacts first. At least one of the substrates forms a covalent bond with the enzyme via a Schiff base. Most of the enzymes isolated so far need Zn2+ as an essential cofactor. To bind the Zn2+ the cysteines have to be in their reduced form. The enzyme is usually a homooctamer. The deprotonation leading from the pyrrolenine tautomer to the aromatic pyrrole is enantioselective and therefore occurs on the enzyme.

An argument, which was used in the early studies to postulate a mechanism, was the formation of "mixed pyrroles" 18 and 19 by PBGS from R. spheroides (see Figure 11).

Figure 11 Enzymatic formation of "mixed pyrroles"

Using the enzyme isolated from R. spheroides, Shemin observed the formation of "mixed" or heterologous pyrroles (see Figure 14)55-57. Shemin had only indirect indications for the structures of the two "mixed pyrroles" 18 and 19 (TLC, radioactive labelling). Assuming that the postulated structures of the "mixed pyrroles" are correct, Shemin drew the following conclusion: A pyrrole analogous to porphobilinogen can only be formed between levulinic acid and 5-aminolevulinic acid if the levulinic acid is using the "A-site". If the levulinic acid is bound to the "P-site" the formation of a "mixed" pyrrole is impossible, he therefore deduced, that the sequence of binding is "A-site" first "P-site" second (see Figure 12). Using this postulate and drawing a close analogy between PBGS and class I aldolases, Shemin proposed a mechanism for PBGS.

Figure 12 Shemin's mechanism for PBGS

To study the order in which the two substrate molecules bind to the enzyme, Jordan performed highly elegant single-turnover experiments 58-60. Stochiometric equivalents of labelled substrate and porphobilinogen synthase were rapidly mixed and after ca.100 ms added to a large excess of unlabelled substrate. The position of the radioactive label was determined by degradation. The pulse labelling could also be done using [5-13C] 5-aminolevulinic acid. The 13C NMR spectrum of the product allowed to identify the position of the label directly 60.

Starting from his observations Jordan postulated an alternative mechanism for the formation of porphobilinogen 58-60 (see Figure 13).

Figure 13 Jordan's mechanism for PBGS

Jordan postulates that after the formation of the Schiff base between the enzyme and the first substrate molecule, the second substrate molecule forms a new Schiff base to the enzyme bound 5-aminolevulinate. Only after this step can the aldol reaction and the elimination which leads, after deprotonation to the product, occur.

Despite the efforts of several research groups, the mechanism of the enzymatic synthesis of porphobilinogen is not yetestablished . Clear is that the sequence of recognition of the two substrate molecules is "P-site" first, "A-site" second, at least for the bovine liver and the human erythrocyte enzymes. The substrate at the "P-site" forms a Schiff's base to a lysine of the active site. The second substrate may be bound non-covalently to the enzyme. One Zn2+ probably complexed to cysteines helps in the catalytic step. There is good circumstantial evidence, that the enzyme shows half-the-site reactivity. Assuming that the active site is formed at the interface of a dimer would be an attractive interpretation of these observations. Finally, the product forms a relatively stable complex with the enzyme. This observation is in agreement with the fact that the last chemical step, the deprotonation, still occurs at the enzyme.