A short history of heme dioxygenases: rise, fall and rise again
- First Online:
- Cite this article as:
- Raven, E.L. J Biol Inorg Chem (2017) 22: 175. doi:10.1007/s00775-016-1412-5
- 818 Downloads
It is well established that there are two different classes of enzymes—tryptophan 2,3-dioxygenase (TDO) and indoleamine 2,3-dioxygenase (IDO)—that catalyse the O2-dependent oxidation of l-tryptophan to N-formylkynurenine. But it was not always so. This perspective presents a short history of the early TDO and IDO literature, the people that were involved in creating it, and the legacy that this left for the future.
Power to the people
There are fashions in science, just as there are in styles of trousers. Fashions in science are influenced by variables large and small: governments that can control the political climate; policy and funding streams; universities and other institutions that control scientific appointments; geography that can enhance or restrict access to ideas or technology; and the rate of development of technology itself which can either slow down or suddenly speed up scientific progress. But more often than not, fashions in science are also influenced to a greater or lesser extent by people, for it is the people who create the focus, the scientific stimulus, and the new ideas upon which future progress must be based.
In the case of the heme dioxygenase enzymes, a handful of people were highly influential and they laid the foundations for the development of the area over the next 60 years. This short perspective summarises these and other early contributions to the heme dioxygenase field.
In the beginning there were two
As often happens, two people drew more or less the same conclusions at more or less the same time. In 1955, Mason  and Hayaishi [2, 3] independently proposed that enzymatic incorporation of molecular oxygen into a substrate was possible. At the time, this was an almost unthinkable idea—probably because the prominent German chemist and Nobel Prize winner Heinrich Wieland (and naturally, therefore, almost everybody else) had ruled the possibility out—but this did not stop Mason and Hayaishi thinking about it quite a lot.
Where there’s muck there’s brass
At that time, the metabolism of tryptophan was just beginning to be clarified, and several people—including the distinguished A. Neuberger from Mill Hill in London1 [14, 15]—had come to the conclusion that NFK was part of the process. But the enzyme responsible for the activity had not been fully established, and it had been temporarily denominated as a “tryptophan peroxidase”. The early nomenclature, to put it mildly, would send shivers down the spine of an IUPAC committee. A list of terms as long as the Royal Mile appeared in print: tryptophan pyrrolase (which still pervades in the literature), tryptophan peroxidase, tryptophan oxidase, tryptophan peroxidase-oxidase, and tryptophan oxygenase were all used (see for example [14, 16, 17, 18, 19, 20, 21, 22]). Most authors evidently found the process of deciding between these terms to be an impossible task and so used them all at the same time. It was Hayaishi himself who brought some order to the confusion, by suggesting in 1970  that the enzyme would most sensibly be named tryptophan 2,3-dioxygenase (TDO), to distinguish its reactivity from any other enzymatic tryptophan activity (e.g. in the formation of tryptophan 5-monooxygenase). Even so, it took some years before the literature adjusted to this brave new world in which one enzyme had only one name.
It had been known at this time that there were other enzymes from different sources capable of catalysing the same reaction as TDO, but with much less substrate specificity than TDO. As far back as 1967, Hayaishi had identified one such enzyme from rabbit intestine  and it was initially identified as “tryptophan pyrrolase (tryptophan 2,3-dioxygenase)”. In view of the broad substrate specificity of these other enzymes, it was suggested , again by Hayaishi, that they be designated as indoleamine 2,3-dioxygenases (IDO), to differentiate them from the TDOs (which are specific for tryptophan) and to convey the message that other substituted indoles were also accessible by these enzymes. Although even as late as 1974 the community was still afflicted by chronic indecision on the names for their pet enzymes, as the early proposal  also suggested the very awkward and certainly confusing “indoleamine 2,3-dioxygenase (pyrrolase)” nomenclature. But by the end of the 1970s the literature was more consistent, with regular papers describing the properties of the now easily recognisable indoleamine 2,3-dioxygenase enzyme (see for example [25, 26, 27, 28, 29, 30, 31, 32, 33, 34]).
In the intervening years, a much clearer picture has emerged. It is now well known that the IDOs and the TDOs, whilst catalysing the same reaction, have slightly different properties. IDOs are monomeric, while the TDOs are tetrameric. IDOs have wide substrate specificity and will oxidise a range of indoleamine derivatives, while the TDOs are much more discriminating and typically oxidise only l-Trp at any respectable catalytic rate. Also, while IDO is widely distributed in all tissues but not the liver, TDO has most often been cited as being found only in the liver (although there is emerging evidence that it is also located in some cancer cells ).
The 1970s: the emergence of heavy metal
The 1980s onwards
A new dawn from 2000: arise again
The Dawson review was very timely, because it included a focused but detailed summary of all of the previous IDO and TDO work. With expression systems emerging soon afterwards (see above), the review set the scene for a resurgence in interest in these enzymes over the next two decades, Fig. 5. Mauk has referred to this as a “renaissance” . Much of the new work in the last few years has been motivated by the search for IDO inhibitors relevant to therapeutic application in cancer [71, 72, 73].
In terms of functional analyses, there have been some substantial developments since 2000 (see also previous reviews [74, 75, 76]). Of special note is the landmark human IDO structure from Sugimoto and Shiro , which gave the first glimpse of the highly hydrophobic IDO active site in complex with the inhibitor 4-phenylimidazole bound to the heme; other structures in complex with related inhibitors have recently appeared [77, 78] and form an important structural framework for structure-based drug design in the future.
The structure of the X. campestris TDO in complex with tryptophan , and other TDO structures have also been important [62, 66]. The structure of human TDO in the apo form (i.e. without heme bound) has also been reported . There are no structures for inhibitor-bound TDOs, with structure-based virtual screening providing the best information so far . It has been suggested from spectroscopic work that the heme sites in (tetrameric) TDO may not be equivalent . The recent structure of human TDO in complex with both O2 and l-Trp  is another step forward, and allows the first reliable visualisation of the binding orientation in the ternary complex.
There is evidence, at least in IDO, that the active site and other regions of protein structure that are not visible in the X-ray maps are conformationally mobile and that this might affect reactivity ; similar flexibility is known to be important in the P450cam system (see for example [84, 85, 86]).
Techniques other than crystallography have been needed to make progress on mechanism, and there is much work to do yet before the mechanism is fully clarified. Early proposals for the mechanism of NFK formation  have been substantially revised in recent years. The generational echoes have resonated loudly, as some of the newer ideas on mechanism  were derived from mass spectrometry experiments (as in the early days ).
Early proposals  for tryptophan oxidation suggested a base-catalysed abstraction mechanism and no change in oxidation state of the metal, but several groups had independently reported [42, 88, 94] that the 1-Me-l-Trp analogue was also reactive, and it was noted  that this is not consistent with a base-catalysed abstraction mechanism. Mutational data where the presumed active site base (histidine) had been removed were also not consistent with base-catalysed abstraction . Two other mechanisms, Fig. 6, have been put forward [88, 90, 91, 97], but there is little in the way of firm evidence for either. Electrophilic addition from the ferrous oxy species, Fig. 6, is one possibility: recent evidence in TDO  (using modified hemes that were first used more than 30 years ago ) supports this. We have noted [74, 75] that oxygen may not be an especially good electrophile if it is bound to the heme as a ferric superoxide species, and there is spectroscopic evidence for a ferric superoxide species  from Raman’s work. An alternative suggestion  is radical addition from the ferric superoxide, Fig. 6 (bottom). Both pathways lead to formation of a ferryl (FeIV) species. There is mass spectrometry evidence for epoxide formation , but later intermediates in the mechanism are not clarified. Addition of oxygen across either the C2 or the C3 position of the substrate is possible for both the radical and electrophilic mechanisms, and at present this is a moot point. Both possibilities have been suggested [82, 88, 90, 91, 93, 97].
Substrate binding and catalysis
It had been noted from very early on [17, 104] that the rate of tryptophan turnover in IDO decreases at high concentrations of substrate. This was originally proposed  to be a consequence of substrate binding to the ferric form of the enzyme, but this is not consistent with the known [51, 105] increase in reduction potential on substrate binding and has therefore been questioned . Some evidence suggests that the sequence of binding of O2 and the substrate at high and low substrate concentrations is important [106, 107, 108], possibly linked to changes in the reduction potential on substrate binding . Others have suggested  that there is a second (inhibitory) binding site in IDO and that this is the origin of the inhibition—this is also plausible and there is evidence for more than one binding site (or at least multiple binding conformations) [61, 109, 110, 111, 112], including in a recent structure for human TDO where a second l-Trp binding site (referred to as an exo site) has been clearly observed at >40 Å from the active site .
What goes around comes around: the lasting contribution of Osamu Hayaishi
Heme dioxygenases have floated into fashion, out of it, and back in again. The early contributions that Hayaishi made to the dioxygenase field are a lasting legacy that form a framework of reference to this day and will serve us all well as the field moves to the future.
Fred Sanger was Neuberger’s first Ph.D. student.
ER acknowledges Dr. J. Basran (University of Leicester), Dr. I. Rowlands (University of Leicester Library), and Prof. Almira Correia for helpful discussions.
Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.