The Use of IUPAC Names in Glossaries

What’s in a name?

To approach these questions, it might be helpful to look at some of the philosophical approaches to naming. Hilary Putnam, in The Meaning of “Meaning” [3], argues that objects referred to by what he calls ‘natural-kind terms’ are themselves core to the meaning of such terms. The terms are fixed by the expert community that deals with them, and thus water is fixed as H₂O by chemists. For the name ‘water’, the referent is individuated by its chemical formula, H₂O.

Putnam then uses a device common in contemporary analytical philosophy, that of parallel Worlds, to construct a thought experiment that has the naming of water as its subject. Earth and Twin Earth are identical in all respects, except that on Twin Earth ‘water’ has a chemical structure different from H₂O, say X_nY. Water and twin-water have similar physicochemical properties except as revealed through examination on the molecular level. Oscar and Twin Oscar have identical genetic makeup and histories in their respective worlds, and presumably have identical mental states. They both have cells that are sustained in solutions of what they call water, they both drink and bathe in water, and they both admire a sunset reflected on a lake of water. Is water ‘water’ on Twin Earth? Is ‘water’ water to Oscar? We will return to this idea below when asking what the consequences would be of naming an incorrect structure that has become embedded in chemical science and in broader human experience.

In his 1972 Princeton lectures [4], Saul Kripke elaborates upon the concept of naming, and poses a thought experiment that can be referred to as “Is Schmidt Gödel?”. Most of us may know little of Kurt Gödel, except perhaps that he discovered the incompleteness of arithmetic (I think I know what he looked like, because I have seen pictures, and I have read some biographical details, but I have no first hand knowledge of Gödel, and little opinion beyond a well-founded belief that he discovered the incompleteness theorem). Kripke, in his thought experiment, constructs the fictive scenario that Gödel had a friend Schmidt, whose remains and papers were later found in Vienna, proving he (Schmidt) actually discovered the theorem and Gödel only published it. Kripke asserts “So since the man who discovered the incompleteness of arithmetic is in fact Schmidt, we, when we talk about ‘Gödel’, are in fact always referring to Schmidt” but then concludes “… it seems to me we are not.” (Ref. [4] p. 84). I may justly be accused of oversimplification here, but I think this makes the point nicely that a finite set of attributes, perhaps consisting of only one element and perhaps even independent of the truth value of its appearance in that set, may help to identify the referent, but may not adequately name it. The title of Kripke’s published Princeton lectures, Naming and Necessity, raises the question to what extent assigning a name necessitates the properties, including existence, of the named entity. This should be borne in mind when we attempt to draw pairwise correspondences between a supposed structural reality, a formalized identifier, and an entity of which we believe we are talking. The glossarist is more concerned with intensional meaning and usage than with naming and extensional description, and so, presumably, are his or her readers.

The value of IUPAC names

It is indeed satisfying to be able to translate a structure into words in a systematic way. It lets us communicate through language rather than pictographically. It gives the student of chemistry a better grasp of the rules of chemical structure, and so the assigning of an IUPAC name to a structure has excellent pedagogical value. I believe that for many substances, and especially those of relatively complex structure, pedagogy becomes the dominant value of the IUPAC name. And so the question becomes: When is this formality of assigning an IUPAC name a reasonable requirement? I can think of no instance where the IUPAC name is more important than the structure it attempts to name, and indeed the name can impart no more information than the structure from which it is derived. Most substances, historically at least, have been given a ‘common’ name before their structure was known. We can only apply IUPAC rules of nomenclature to a substance later, when we know its chemical (molecular) structure—the ‘naming’ process then becomes one of translating structures into words. In Putnam’s terms, we might say that adrenaline is individuated as the correct three-dimensional connectivity of C₉H₁₃NO₃, and this knowledge comes to us from the realm of the experts in structure determination. But to the nomenclaturist, the exercise is then one of systematizing a way in which C₉H₁₃NO₃ is best rendered in words, as 4-[(1R)-1-hydroxy-2-(methylamino)ethyl]benzene-1,2-diol.

Two more serious arguments have been put forward for the use of IUPAC names for all substances in a glossary. It is suggested that they remove ambiguity for the chemist, and that they meet a legal requirement for which IUPAC’s representatives might be called to account in legal proceedings. Importantly for both objectives, the rules of IUPAC nomenclature should provide clarity in referencing a particular chemical compound.

Take again the example of adrenaline, known in some parts of the English-speaking world as epinephrine. Here, the IUPAC name can serve a useful role in disambiguation, and confirm that indeed the names adrenaline and epinephrine refer to the same substance. But in practice there can be no reasonable doubt what adrenaline ‘is’, any more than we doubt that adrenaline is not tryptophan or that Gödel is not Schmidt. Adrenaline was isolated, identified, and synthesized more than a century ago; for the better part of a century it has been a household name to biochemists, it has permeated endocrinology and neurophysiology, medical practice has been imbued with it, and ‘adrenaline rush’ is used in common speech, indicating general awareness of its physiological effects. I find it difficult to imagine a scenario where the legal obligations of IUPAC would not be fulfilled by mentioning adrenaline in a recommendation without referring to its IUPAC name.

With further regard to the legal requirement, it is a structure, and not a name, that is patented. Drugs such as sertraline, fentanyl, and haloperidol—even if they go by different common or generic names in different jurisdictions—are defined by the manufacturers that hold their patents (and by the physicians that prescribe them) as the structures they are believed to represent, and not by IUPAC names that may be written correctly (or incorrectly on occasion) according to more than one convention (see Preferred IUPAC Name, below). As Hellwich has put it in these pages, “the thing named by the word remains the same, and many of its other properties remain uncertain, even if another designation is selected”. [5]

Confirming that acetaminophen and paracetamol are one and the same substance by providing an IUPAC name can be helpful in discourse, but does not supersede the legal status of the structure.

The practical use of glossaries

Inclusion of the relatively simple IUPAC name for noradrenaline (see sketch below) is non-intrusive in the text of a glossary, and immediately conveys the information to the chemist that it is a catechol structure with a primary amine functionality, and that there is an optically active center that probably matters. Thus, while not an essential part of a strict glossary definition (“catecholamine hormone acting as a postganglionic adrenergic mediator at α- and β-adrenergic receptors” [2] seems sufficient to differentiate noradrenaline from other substances and so inform the reader), this added piece of information—the IUPAC name—is interesting and non-distracting. It also serves to reassure the reader that noradrenaline is identical to the substance also called norepinephrine, and furthermore to distinguish it from the levorotatory isomer, 4-[(1S)-2-amino-1-hydroxyethyl]benzene-1,2-diol, a synthetic compound, “having greater pressor activity than the natural dextrorotatory isomer” [2]. Note, however, that while to define ‘noradrenaline’ simply as ‘4-[(1R)-2-amino-1-hydroxyethyl]benzene-1,2-diol’, or more accurately as ‘substance whose structure is denoted by 4-[(1R)-2-amino-1-hydroxyethyl]benzene-1,2-diol’ is, strictly speaking, correct, it would be sure to disappoint a user of the glossary if the definition were restricted to this; the reader is looking for more than an extensional definition of a natural-kind term (e.g., saying that by tiger we mean any member of the set of all things that are tigers). On the other hand, a glossary’s work is generally done when a common name of a substance is entered and its intensional meaning explained—the IUPAC name may be helpful in conveying intensional meaning when it does not detract by its complexity.

Noradrenaline has been known by this name, or alternatively as norepinephrine, throughout the English-speaking world for more than a century. Its IUPAC name, 4-[(1R)-2-amino-1-hydroxyethyl]benzene-1,2-diol, immediately informs the reader of its basic structure, hints that its stereochemistry may be important for function - indeed, distinguishing it from its also biologically active levorotatory 1S isomer, levaterenol—and clearly differentiates it from the related substance adrenaline (epinephrine; (R)-4-(1-hydroxy-2-(methylamino)ethyl)benzene-1,2-diol).

Domoic acid, or (2S,3S,4S)-4-[(2Z,4E,6R)-6-carboxyhepta-2,4-dien-2-yl]-3-(carboxymethyl)pyrrolidine-2-carboxylic acid, is a good example of the latter case.

Domoic Acid

A cause of amnesic shellfish poisoning, it occurs in some algal blooms, and is neurotoxic by virtue of its acting as an analog of kainic acid at a defined class of glutamate receptors [Kainic acid, or (2S,3S,4S)-3-(carboxymethyl)-4-(prop-1-en-2-yl)pyrrolidine-2-carboxylic acid, is itself a natural product produced by some seaweeds that is used as an experimental agent to induce seizures in animals in an experimental model of epilepsy]. Although domoic acid was known historically by its effects, an outbreak of shellfish poisoning in Prince Edward Island, Canada, in 1987 that led to three deaths prompted the effort of structure determination. Inclusion of the IUPAC name in the Glossary [2], together with that of kainic acid, gives the reader an idea of general structure, the relative degree of complexity, perhaps an insight into the approach that would have been taken to structure determination, and a quick comparison with the prototypic kainic acid, and thus at least a flavor of intension, all without the need to punctuate the glossary with structural drawings.

When we take a substance like cocaine, with a broadly contextualized societal significance, we are closer to Oscar and Twin Oscar discussing water (Of course, living in parallel Worlds they couldn’t, but we can still imagine the conversation). Most chemists will have a general idea of at least the structure of cocaine; and, if they consult a Glossary of Terms in Neurotoxicology, it is unlikely to be in order to look for, or to confirm, the structure, but rather to learn more about its bioactive properties. True, the designation “methyl(1R,2R,3S,5S)-3-(benzoyloxy)-8-methyl-8-azabicyclo[3.2.1]octane-2-carboxylate”, in principle, immediately lets the savvy chemist visualize the structure of cocaine.

Cocaine

But, if it turned out that this structure were wrong, that an error in stereochemical assignment, say of the methyl carboxylate at C2, had been made initially and propagated through the literature for decades, this would have profound consequences for medicinal chemists and receptor physiologists; but it wouldn’t matter much to the reader of a glossary who wanted to know what that recreational white powder ‘is’ in a neurochemical context. The intensional component of the glossary entry wouldn’t be changed, just as if Twin-Earth ‘water’ turned out really to be X_nZ instead of X_nY, it wouldn’t much affect the conversation between the Oscars that we imagined above. If we wish to be enlightened on the incompleteness of arithmetic, do we first ask “Is Schmidt ‘Gödel’”?

So, the IUPAC name for noradrenaline is helpful and informative, for cocaine it is interesting, and for domoic acid it may be both. On the other hand, when a chemist looks up sertraline in a Glossary of Terms in Neurotoxicology, (s)he is interested in the biological properties and mechanisms of action of an antidepressant drug, possibly in its structure, and probably not in its IUPAC name. While it is important to have a specification of the exact structure of the substance under discussion, sertraline is a generic drug name, and its precise chemical structure is adequately specified in international patents, independent of the IUPAC name assigned to that structure.

Sertaline

Nominare ad absurdum

A strict mandate for IUPAC names can quickly get out of hand, especially when biologicals are involved. When is the formal IUPAC name of the structure desirable in a glossary? A good example for mulling over this question is tetrodotoxin. There may be some for whom the IUPAC name and structure of tetrodotoxin (see sketch below) immediately impart the same information; I am not one of them. For most chemists, I believe, the reconstruction of the structure from the IUPAC name, and vice versa, are rather laborious (even error-prone), though not intractable, tasks.

Tetrodotoxin

Most will know that eating the puffer fish as a delicacy is fraught; the fish produces tetrodotoxin, and there is a risk of consuming the toxin if the fish is not properly prepared. But tetrodotoxin has also been a very useful reagent in studying neuromuscular function, as it selectively blocks Na⁺ conduction across some classes of ion channels, and as such it has taken on a life of its own in the scientific literature, where it graces textbooks and is frequently used as an adjective describing certain ion channels, as in tetrodotoxin-insensitive Na⁺ channels. Suppose the only mention of tetrodotoxin in a glossary, thus precluding cross-referencing, was in the definition of that class of Na⁺ channels as “pores in cell membranes selectively permeable to Na⁺, independent of the presence of a specific fish toxin”, for example. Would obligatory use of the IUPAC name then condemn the entry term to be “(4R,4aR,5R,6S,7S,8S,8aR,10S,12S)-2-azaniumylidene-4,6,8,12-tetrahydroxy-6-(hydroxymethyl)octahydro-1H-8a,10-methano-5,7(epoxymethanooxy)quinazolin-10-olate-insensitive Na⁺ channel”? This example may seem contrived, but because tetrodotoxin is commonly used as an adjective, I see no syntactic difference, in principle, in using its IUPAC name interchangeably as the same part of speech; would it be desirable to require a separate entry for tetrodotoxin (the noun) just to state the IUPAC name, and thus allow the necessary entry for the common term “tetrodotoxin-insensitive Na⁺ channel” (italics implying cross-reference)?

To complicate matters further, brevetoxins illustrate the problem of multiple ways of systematizing IUPAC names. These represent several naturally-occurring structural variants of substances falling into the classes brevetoxin A and B; the prototype for brevetoxin A is brevetoxin-1, and for brevetoxin B is brevetoxin-2. These are shown, designated with different IUPAC naming conventions, next page. The occurrence of these different conventions (for example, naming brevetoxin A according to rules for fused-ring systems and brevetoxin B as a bridged compound) has led to the concept of a “preferred IUPAC name” (PIN), and can contribute to confusion when a relatively complex IUPAC name intervenes between a substance’s entry in a glossary, and its chemical structure. The convention for naming the brevetoxins may be a matter of taste, and I realize that the preference would be to treat them both as the fused ring system - I include the bridged compound name for brevetoxin B only to illustrate an alternative “correct” IUPAC name. But, the inclusion of either name is of little interest to the reader who consults a glossary to learn that these polyether compounds are produced by dinoflagellates and cause a type of neurotoxic shellfish poisoning by binding to neuronal voltage-gated Na⁺ channels.

In some cases, such as with the brevetoxins, dinophysis toxins, and pectenotoxins, IUPAC names were avoided in the Glossary of Terms in Neurotoxicology because the common names refer not to unique substances but to collections of related structures produced through variations in biosynthetic pathways. Yessotoxin, however, is a unique compound and was designated, according to one IUPAC convention and no doubt to the enlightenment of very few readers, by 2-[(2S,4aS,5aR,6R,6aS,7aR,8S,10aS,11aR,13aS,14aR,15aS,16aR,18S,19R,20aS,21aR,22aS,23aR,24aS,25aR,26aS,27aR,28aS,29aR)-6-hydroxy-2-[(2R,3E)-2-hydroxy-5-methylidene-octa-3,7-dien-2-yl]-5a,8,10a,11a,19-pentamethyl-3-methylidene-18-(sulfooxy)octatriacontahydropyrano[2””’,3””’:5””,6””]pyrano[2”,3”:5’,6’]pyrano[2’,3’:5,6]pyrano[3,2-b]pyrano[2””’,3””’:5””,6”’]pyrano[2”’,3””:5”’,6”’]pyrano[2”’,3”’:5”,6”]pyrano[2”,3”:6’,7’]oxepino[2’,3’:5,6]pyrano[2,3-g]oxocin-19-yl]ethyl hydrogen sulfate.

InChIKey to the rescue?

Some hope for resolving the issue of mandatory IUPAC designation, while maintaining perspective on the intended use of various IUPAC documents, may be offered by coding strategies such as the InChIKey, or hashed InChI. This allows a uniformly 27-character designator to be assigned to each structure, and these designators could be embedded in a glossary text rather unobtrusively, perhaps in a reduced font size, or perhaps collected in a table of all substances referred to in the document, placed at the end. The InChI is generated unambiguously by an algorithm and is non-proprietary. It permits database searching for structures. Currently, an InChIKey has not been generated and assigned for all substances to my knowledge, and for those where it has been, it is not readily available to every user. And, with present coding, there is a finite (though vanishingly small) chance that two structures, each with unique InChI, could have the same InChIKey. However, it is clear that the ongoing development of InChIKey technology will solve these problems. There may still be a limited role for conventional IUPAC names in IUPAC glossaries, but when the user is likely to want to access a structure, or when legal requirements become of concern, they will be superseded by less unwieldy coding. Back to adrenaline, UCTWMZQNUQWSLP-VIFPVBQESA-N may not embed more unobtrusively in a glossary than its IUPAC name, 4-[(1R)-1-hydroxy-2-(methylamino)ethyl]benzene-1,2-diol, though I doubt this could be said for HCYDZFJGUKMTQB-AVHIVUAZSA-N (yessotoxin—compare IUPAC name above).

Brevetoxins. These natural products are a cause of neurotoxic shellfish poisoning when ingested with contaminated shellfish. Several structural variants are produced by the dinoflagellate Karenia brevis. The prototypical structures are brevetoxin A and B.

A) Brevetoxin A. When R = -CH₂C(=CH₂)CHO the compound is designated brevetoxin-1 and can be named as a fused-ring structure, 2-{[(3aR,4aS,5aR,6aS,8Z,10aR,11aS,14aR,15aS,16aR,18R,20S,20aS,21aR,22aS,23aR,24Z,26aS,27aR,28S,29aS,30aR,32R,33aS)-20-hydroxy-22a,28,30a,32-tetramethyl-2-oxo-3,3a,4a,5,5a,6a,7,10,10a,11a,12,13,14,14a,15a,16,16a,18,19,20,20a,21a,22,22a,23a,26,26a,27a,28,29,29a,30a,31,32,33,33a-hexatriacontahydro-2H-furo[2’’’,3’’’:7’’,8’’]oxocino[2’’,3’’:5’,6’]pyrano[2’,3’:6,7]oxepino[3,2-b]pyrano[2’’’’,3’’’’:5’’’,6’’’]pyrano[2’’’,3’’’:5’’,6’’]pyrano[2’’,3’’:7’,8’]oxocino[2’,3’:7,8]oxocino[2,3-h]oxonin-18-yl]methyl}prop-2-enal.

B) Brevetoxin B was synthesized by K.C. Nicolaou and coworkers [6] in 1995, in 123 steps with 91% average yield (final yield ~9·10⁻⁶). When R = -CH₂C(=CH₂)CHO, the structure is designated brevetoxin-2 and can be named correctly according to rules for a bridged cyclic compound as 2-{[(1R,3S,5R,7S,9R,11S,12S,14R,16R,18S,20R,21Z,24S,26R,28S,30R,31R,33S,35R,37S,42R,44S,46R,48S)-12-hydroxy-1,3,11,24,31,41,44-heptamethyl-39-oxo-2,6,10,15,19,25,29,34,38,43,47-undecaoxaundecacyclo[26.22.0.0³,²⁶.0⁵,²⁴.0⁷,²⁰.0⁹,¹⁸.0¹¹,¹⁶.0³⁰,⁴⁸.0³³,⁴⁶.0³⁵,⁴⁴.0³⁷,⁴²]pentaconta-21,40-dien-14-yl]methyl}prop-2-enal, although a fused-ring convention as given here for brevetoxin-1 is preferred.

Summary

1. Assigning a systematic IUPAC name to a proposed chemical structure does not necessitate the physical existence of the structure, nor does an error in naming supersede the primacy of the structure.

2. Although IUPAC names should be used whenever possible in the primary research literature, the reader of a glossary is generally seeking intensional meaning, not necessarily extensional specification, and may or may not be concerned with structure in a particular case.

3. IUPAC names may serve a useful pedagogical role in a glossary when they are unobtrusive and more or less transparent to the educated reader, but otherwise may be an unhelpful distraction.

4. Very complex names and choice of a Preferred IUPAC Name (PIN) among various naming conventions may not advance the needs of a glossary user. If exact specification of a structure is needed, an endorsed coding system such as the InChIKey may be a more welcome addition.

I realize these comments will be controversial. I share with others in IUPAC the belief that using IUPAC names in scientific documents is highly desirable. But I submit that a good case can be made for allowing authors to exercise judgment in their inclusion when compiling glossaries that include a large number of complex natural products and pharmaceuticals.