The advent of stereochemical studies on phosphate esters and diesters with particular reference to their enzyme catalysed reactions, initially through the work of Knowles  and of Lowe , placed new demands on the nomenclature of the oxygen atoms of the transferring phosphoryl group, PO3−. 1 In early work employing thiophosphates made chiral by the specific introduction of oxygen-18 paired with oxygen-16, the direct application of Cahn-Ingold-Prelog (CIP) Rules for prochirality ,  resolved the problem by labelling the oxygen atoms (Rp) and (Sp), as appropriate , , . The more advanced use of 16O, 17O, and 18O bonded to the same phosphorus  led to the concept of pro-pro-pro-chirality at phosphorus, which was still capable of CIP identification , . However, such isotopic labelling is experimentally demanding and not necessarily applicable to stereochemical problems now more readily amenable to analysis through advances in protein crystallography. The increasing frequency of binary and tertiary structures of proteins in complex with phosphate ester substrates and/or analogues has enabled a rapidly expanding number of enzyme catalysed reactions to be investigated by structural and computational methods , , . Indeed, there are now over 1600 ligands in the PDB having a phosphoryl group component. These are associated with over 28 000 deposited structures. While many of these structures can be, and have been, labelled for their phosphorus and phosphoryl oxygen atoms through current practice, comparative studies of related structures easily identify multiple inconsistencies in labelling that arise from variable methods of naming N, O, and P atoms.
This situation has become increasingly complex as a result of the introduction and development of metal fluoride (MFx) analogues of the PO3− group in studies on transition state analogues (TSA) for phosphoryl transfer enzymes. Trifluoroberyllate (BeF3−)2 is a ground state analogue for phosphate, with characteristic tetrahedral geometry when ligated to anionic oxygen. Tetrafluoroaluminate (AlF4−)3 is a mimic for concerted phosphoryl transfer in multiple enzymes, though it has octahedral geometry. Aluminium trifluoride (AlF3) 4 forms trigonal bipyramidal (tbp) TSA complexes that have the correct 3-D geometry for concerted PO3− group transfer but lack its anionic charge. These two values converge in the relatively smaller number of trifluoromagnesate complexes (MgF3−), 5 which are both anionic and have tbp geometry. Indeed, some of the AlF3 complexes have been shown in reality to be MgF3− complexes in solution , , . The growth in use of these four types of MFx complexes is illustrated in Fig. 1. In addition, there are many significant structures of phosphoryl transfer enzyme complexes that include vanadium(V) or tungsten(VI) complexes, either as tetrahedral phosphate mimics or as tbp mimics of transition states. The relative growth in use of these six species is presented in Fig. 1. The double change from four coordinate tetrahedral PO4 to five coordinate tbp O–MF3–O and six coordinate octahedral O–MF4–O complexes adds a new dimension to the problem of the atomic description of these complexes. The need to solve this general problem provided the principal motivation for the development of these standardized naming conventions.
As the development of our protocols progressed, it became apparent to us that a rational, logical set of labels for the 5- and 6-coordinate systems described above could only be established on the basis of a clear definition of the systematic labelling of phosphorus atoms in standard multiple phosphate molecules, that already extends to eight in the case of hexaphosphoinositol bisphosphates . It needed to be followed by a comprehensive system for oxygen atom labelling to include both bridge and non-bridge atoms in linear chains of phosphates, as for the 13 oxygen atoms of 5′-adenosyl 5′″-guanosyl P1, P4-tetraphosphate  and the 3 non-isotopically identifiable oxygen atoms of the PO3− group of terminal phosphates. With those objectives accomplished, our recommendations could then be developed to incorporate both the fluorine ligands of MFx systems and the oxygen atoms of vanadate and tungstate analogues of phosphates and their TSAs.
The basic strategy of the recommendations is built on the recognition that a phosphate monoester comprises an alkoxy group and a phosphoryl group (ROH + PO3−), a monoalkyl diphosphate comprises a phosphate monoester and a second phosphoryl group (ROPO3− + PO3−), a monoalkyl triphosphate comprises a monoalkyl diphosphate and a third phosphoryl group, and so on. For simplicity, we have ignored anionic charges on phosphoryl oxygen atoms and we have treated P=O “double bonds” as P–O single bonds because there is no π-bonding in the phosphoryl group. While we do not seek to claim that our coverage has been exhaustive, we believe that the principles for naming atoms set out here will prove generally applicable to all cognate molecular species which share a geometrical relationship to phosphates, e.g. sulfates, perchlorates, etc.
Lastly, we provide an Appendix as a simple guide to the application of Cahn-Ingold-Prelog Rules to label prochiral, non-bridge oxygen atoms in molecules under inspection.
2 Existing recommendations
Phosphorus nomenclature and related IUPAC Recommendations
The nomenclature of phosphorus-containing compounds of biochemical importance, Recommendations 1976, was published in 1977 . It was concerned with the naming of compounds, but did not consider the identification of the individual atoms of the phosphate or polyphosphate groups, other than to label the phosphates of a nucleoside triphosphate α, β, and γ. It did cover the naming of polyphosphates where a bridging oxygen is replaced by a methylene or imino group. A variation on this was proposed in 1980 and revised in 1992 .
A document on the abbreviations and symbols for the description of conformation of polynucleotide chains, Recommendations 1982, was published in 1983 . In a related paper, it was proposed that the pro-S oxygen should be OP1 and the pro-R should be OP2 . This is the reverse of the system proposed here and it is also contrary to CIP nomenclature, where Rule 5 gives priority to R over S. We have chosen to adhere to CIP priority Rule 5.
IUPAC Recommendations for preferred names of derivatives of phosphoric acid are pertinent . They include the application of the CIP rules to chiral phosphates as well as CIP rules for a trigonal bipyramidal and octahedral systems. These are also described in IUPAC inorganic chemistry nomenclature systems for bipyramidal and octahedral structures .
3 Recommendations for labelling phosphorus atoms in phosphates
3.1 A. Labelling phosphorus atoms in polyphosphate species
3.1.1 A1. Species with one single polyphosphate chain
This requires a one-symbol code to describe the position of each phosphorus in a single chain of phosphates. Phosphorus descriptions use a capital letter that serves to discriminate sequential phosphorus atoms in the same chain (PDB usage).
Phosphorus atoms are named in progression from the RO- end as PA, PB, PG, PD, etc. 6 Hence, adenosine 5′-tetraphosphate (PDB ligand: AQP) has phosphorus atoms labelled as PA, PB, PG, PD starting from the ribose 5′-oxygen (Fig. A1a).
For P1-(5′-adenosyl)-P5-(5′″-guanosyl) pentaphosphate (PDB Ligand: G5P), phosphorus atoms should be named PA, PB, PG, PD, and PE starting at the 5′-oxygen of the adenosine.
For the RO- group at the end of a phosphate chain, a nucleoside takes priority over a non-nucleoside. Thus, in uridine diphosphate glucose (PDB ligand: UPG), PA is bonded to uridine-O5′ and PB is bonded to O1″ of glucose (Fig. A1b). 7
A nucleic acid base takes priority over non-nucleic acid base (i.e. adenosine > nicotinamide riboside). Thus, in NAD+ (PDB ligand: NAD), PA is bonded to O5′ of adenosine, with PB bonded to O5′″ of the nicotinamide riboside (Fig. A1c).
Nucleosides take priority in their alphabetical order (A > C > G > dT > U). Thus, in P1-(5′-adenosyl) P4-(5′″-deoxythymidyl) tetraphosphate (Ap4dT) (PDB ligand: 4TA), the phosphorus atoms should be named PA, PB, PG, PD, starting at the 5′-oxygen of the adenosine (Fig. A1d).
Pentoses use CIP priorities: D-ribose > L-ribose > 2-deoxy-D-ribose > 2-deoxy-L-ribose. 8 Thus, a transition state for dAMP kinase should label the four phosphorus atoms PA, PB, PG, PD, starting from the adenosine 5′-oxygen (Fig. A1e).
In phosphonate and phosphoramidate analogues of polyphosphates, phosphorus atoms will be labelled in the same manner as for the parent polyphosphate molecule. Hence, for β,γ-methylene-GTP (PDB ligand: GCP), phosphorus atoms should be named PA, PB and PG from the 5′-oxygen (Fig. A1f1).
Likewise, for 2′-deoxyuridine 5′-α,β-imidotriphosphate (PDB ligand: DUP), phosphorus atoms should be named PA, PB and PG from the 5′-oxygen (Fig. A1f2).
3.1.2 A2. Species with multiple single phosphate chains
This requires a one-symbol code to describe the relationship of each phosphate chain to the parent molecule.
Inositol polyphosphates require a phosphorus label derived from the identity of the oxygen to which each single phosphate is attached. Thus, for myo-inositol 1,3,4,5,6-pentakis phosphate (InsP5) (PDB ligand: 5MY), the phosphorus atoms should be labelled P1, P3, P4, P5, and P6 (Fig. A2a1). 9 For fructose 1,6-bisphosphate (PDB label: FBP), the phosphorus atoms should be labelled P1 and P6 (Fig. A2a2).
3.1.3 A3. Species with multiple single phosphate and/or polyphosphate chains
This requires a two-symbol code to describe (i) the position of each phosphorus in a single chain of phosphates, and (ii) the relationship of that phosphate chain to the parent molecule.
Species with polyphosphates located on multiple oxygen atoms require a two-symbol code to designate their phosphorus atoms: a numerical code for the oxygen bridging to the parent molecule and an alphabetic code for the position of the phosphorus in the phosphate chain. Thus, in pppGpp (PDB ligand: 0O2), the 5′-phosphorus atoms should be named PA5, PB5, and PG5, and the 3′-phosphorus atoms named PA3 and PB3 (Fig. A3a).
Inositol polyphosphates having polyphosphate moieties require a two-symbol code to designate their phosphorus atoms. A numerical symbol designates the oxygen to which each single phosphate is attached and an alphabetic code designates the position of the phosphorus in the phosphate chain. In the case of monophosphates, the labels P1, P2, etc., should apply to single phosphorus entities, while PAn and PBn will apply to diphosphates, as in PP-InsP5 (PDB ligand: I7P) (Fig. A3b).4
4 Recommendations for labelling oxygen atoms in phosphates
4.1 B1. Non-terminal phosphates in molecules with one single phosphate chain
This requires a two-symbol code to describe (i) the identity of the oxygen relative to its congeners, and (ii) the identity of the parent phosphorus atom. Oxygen codes use a number first, to discriminate oxygen atoms bonded to the same phosphorus, followed by a letter to indicate the parent phosphorus.
The oxygen linking PA to the carbon moiety of the molecule will retain its regular label. Thus, in ATP, O5′ bonds PA to the ribose (Fig. B1a1). In Ap4G, O5′ bonds PA to adenosine, while O5′″ bonds PD to guanosine (Fig. B1a2) .
In each non-terminal phosphoryl group (PO3), the bridging oxygen bonding PX to P(X + 1) in the chain should be numbered O3X. Hence, in ATP, O3A joins PA to PB, and O3B joins PB to PG (Fig. B1b).
In each non-terminal phosphoryl group, the two non-bridging oxygen atoms will be labelled 1 and 2 according to their CIP pro-R- and pro-S-chiralities, respectively. 10 Hence, in ATP, PA will have non-bridging oxygen atoms O1A and O2A for the pro-R and pro-S oxygen atoms, respectively (Fig. B1b).
Bridging atoms in polyphosphate chains should be given priorities On′ < O3A < O3B < O3G < O3D < etc. (Section B3). This provides a relatively direct CIP route to the assignment of paired non-bridging oxygens on non-terminal PB, PG, etc., that can be illustrated here for O1B and O2B (Fig. B1d).
In chains containing a sulfur atom in a non-bridging, non-terminal position, the sulfur will take the name S1A (for substituent on PA), S1B (for substituent on PB), etc. The non-bridging oxygen is then named O2A, O2B, etc., and the bridging oxygen is O3A, O3B, etc., as above. This is shown for guanosine 5′-(Rp)-α-thio-triphosphate (PDB ligand: GAV) (Fig. B1e).
In modified polyphosphate chains having two-atom bridges replacing an O3N (where N = A, B, etc.) the bridging atoms X and Y will be labelled X3A and Y4A, progressively. Thus in β,γ-oxymethylene-ATP (AdoPOPOCH2P), the PB, PG-bridging atoms are O3B and C4B, respectively (Fig. B1f).
In polyphosphate chains with a bridging oxygen replaced by carbon or nitrogen, the prochirality 11 designations may change in consequence. Thus, in α,β-methylene adenosine 5′-triphosphate (PDB ligand: APC) (Fig. B1g), oxygen atoms O1A and O2A are reversed relative to their designation in ATP (Fig. B1b) because of the changed priority of C3A relative to O5′.
4.2 B2. Non-terminal phosphates in molecules with multiple phosphate chains
This requires a three-symbol code, using (i) one symbol to describe the identity of the oxygen relative to its congeners, and (ii) two symbols for the identity of the parent phosphorus atom (see above)
In each non-terminal phosphoryl group, the two non-bridging oxygen atoms will be labelled 1 and 2 according to their CIP pro-R and pro-S chiralities, respectively. Hence, in ppGpp (PDB ligand: G4P), PA5 will have non-bridging oxygen atoms O1A5 and O2A5 for the pro-R and pro-S oxygen atoms, respectively, and PA3 will have non-bridging oxygen atoms O1A3 and O2A3 for the pro-R and pro-S oxygen atoms, respectively 12 (Fig. B2a).
In NAD+, the oxygen atoms on PA will be labelled O1A and O2A for the pro-R and pro-S oxygen atoms, respectively, and the oxygen atoms on PB will be labelled O1B and O2B for the pro-R and pro-S oxygen atoms, respectively (Fig. B2b). (NB Ade takes priority over Nicotinamide. Fig. A1d).
In ppIns5p, the oxygen atoms on PA5 will be labelled O1A5 and O2A5 for the pro-R and pro-S oxygen atoms, respectively (Fig. B2c).
In each non-terminal phosphoryl group (PO3), the bridging oxygen bonding PN to P(N + 1) (where N = A, B, etc.) in the chain should be numbered O3Nx, where x designates the parent oxygen of the polyphosphate chain. Hence, in ppGpp (PDB ligand: P4G), PA5 is joined to PB5 by O3A5, and PA3 is joined to PB3 by O3A3 (Fig. B2d).
4.3 B3. Non-terminal phosphates in molecules with a doubly-capped, single phosphate chain
This requires a two-symbol code to describe (i) the identity of the oxygen relative to its congeners and (ii) one symbol for the identity of the parent phosphorus atom (v.s.).
There are several examples of natural and non-natural dinucleosidyl polyphosphates with phosphate chain lengths from 2 to 6 phosphates. P1-(5′-adenosyl) P4-(5′″-deoxythymidyl) tetraphosphate (Ap4dT) (PDB ligand: 4TA) (Fig. A1d) and P1-(5′-adenosyl)-P4-(5′″-guanosyl) tetraphosphate (Ap4G) (Fig. B1a2) have been discussed above. While the accurate, and sometimes tricky, application of CIP rules can deliver an appropriate descriptor for paired non-bridging oxygens, for symmetrical molecules, including Ap3A and Ap5A, the oxygens on the central phosphate are stereochemically identical (being related by C2 symmetry). They become a diastereotopic pair only in a chiral environment, as when bound to a protein, for which an additional Rule is needed. The problem can easily be resolved by the application of an additional priority Rule for sequential phosphates in polyphosphate chains.
“Bridging atoms in polyphosphate chains are given priorities that increase with increasing separation from the priority nucleoside oxygen: On′ < O3A < O3B < O3G < O3D < etc.”
The application of this rule is illustrated for P1-(5′-adenosyl)-P3-(5′″-adenosyl) triphosphate, Ap3A (Fig. B3a). By extension, this rule also provides a relatively direct assignment for the non-terminal phosphates of ATP, p4A, etc. It is further illustrated in the Appendix (Section 9.5).
4.4 B4. Terminal phosphates in molecules with multiple phosphate chains
This requires a two-symbol code to describe (i) the identity of the oxygen relative to its congeners and (ii) the identity of the parent phosphorus atom (v.s.). The three oxygen atoms of a terminal phosphoryl group (PO3 ) are pro-pro-chiral. They can thus be labelled according to CIP rules in those (rare) cases where they are identified by isotopes 16O, 17 O, and 18 O.
In cases of a terminal phosphoryl oxygen being replaced by e.g. sulfur, fluorine, or nitrogen, the remaining two terminal oxygen atoms are prochiral and can be appropriately identified by CIP chirality rules. Thus, in GTPγS (PDB ligand: GSP), the sulfur has priority to be labelled S1G and the oxygen atoms are labelled O2G (pro-R) and O3G (pro-S), respectively (Fig. B4a).
Prochirality identification can be applied if one of the three oxygen atoms is promoted relative to the other two. In the context of enzyme-bound nucleotides, such promotion can often be identified by co-ordination of the terminal phosphate to a protein-bound metal ion, typically magnesium. Thus, for ATP bound in many kinases, the γ-phosphate is coordinated from one of its three oxygen atoms to magnesium. This oxygen is thus designated O1G. The remaining oxygen atoms are now prochiral and can be identified in the priority series: O3B > O1G > O2G > O3G. CIP rules then designate O2G as the pro-R oxygen and O3G as the pro-S oxygen, as illustrated for ATP bound in phosphoglycerate kinase (Fig. B4b; PDB entry: 1VJC).
In the absence of metal ion coordination to the terminal phosphate, H-bond donation from amino acids in the protein provides a means of priority identification for O1N. Hydrogen bonds are considered only if they have a length ≤3.0 Å. 13 Priority will be given according to donor atom XH priority with CIP rules (S > O > N). Hydrogen bonding to the amino acid of lowest primary sequence number will identify O1G in ATP, etc. If there is still ambiguity in the assignment, then backbone NH takes priority over sidechain NH. 14 This selection makes O2G and O3G prochiral, hence they can be assigned by the application of CIP rules. 15 Thus, in human bisphosphoglycerate mutase (PDB entry: 2A9J), the 3-phosphoglycerate has phosphoryl oxygen coordination from Arg100 and Arg116 to O1A, from Arg117 and Asn190 to O2A, and from Arg117 to O3A (Fig. B4c1). After O1A is promoted by amino acid linkage priority, O2A and O3A are assigned by prochirality rules (O3 > O1A > O2A > O3A).
In the case of human protein tyrosine phosphatase ptpn5 (C472S mutant), the tyrosine phosphate moiety is coordinated to residues in the loop Ala474-Arg478 (PDB entry: 2CJZ). Consideration of hydrogen bonds ≤3.0 Å shows oxygen O1P coordinated to Gly477 and Ile476 oxygen O2P coordinated to Ala474 and Arg478, and oxygen O3P coordinated to Arg478. Thus, we can now designate O1A as being coordinated to the lowest numbered amino acid, Ala474 (it is labelled as O2P in 2CJZ).10 The oxygen atom priority is O4′ > O1A > O2A > O3A, in which O2A and O3A are designated by CIP rules for prochirality10 as shown (O2A being pro-R and O3A is pro-S) (Fig. B4c2). (NB There are H-bonds from Ser472(OH) to O2P and O3P but both are longer than 3.0 Å and thus are ignored; distance from heavy atom to heavy atom).
4.5 B5. Terminal phosphates in molecules with multiple phosphate chains
This requires a three-symbol code, (i) one symbol to describe the identity of the oxygen relative to its congeners, and (ii) two symbols for the identity of the parent phosphorus atom (v.s.).
The rules described above (section B3) for single phosphate chains will apply with the addition of a descriptor symbol designating the point of attachment of that chain to the parent molecule. Thus, for human aldolase reductase (PDB entry: 2J8T), the bound NADP+ (PDB ligand: NAP) has the oxygen atoms of PA2 coordinating no metal and hydrogen bonded to Lys262, Ser263, Val264, Thr265, and Arg268. Thus, the oxygen coordinating Ser263 takes priority and is named O1A2. The oxygen atom priorities for PA2 are thus O2 > O1A2 > O2A2 > O3A, as shown (Fig. B5a).
4.6 B6. Isolated single phosphates
This requires prioritisation of two oxygen atoms by their coordination features thus allowing the third and fourth oxygen atoms to be assigned their prochirality by CIP rules.
Isolated phosphate with no metal ions. In a structure of the small G protein Rab-5c with GDP and Pi ligands in the catalytic site (PDB entry: 1Z0D), the isolated phosphate (PDB ligand: PO4) is not metal coordinated. Thus, the relative priorities of its 4 oxygen atoms are determined by H-bonds to amino acid residues. Ignoring H-bonds ≥ 3.0 Å, the PDB file assigns O1 coordinated to Ser30(OH), O2 coordinated to Leu80(NH), and O3 coordinated to Lys34(NH3+). O4 is only coordinated to ligands at distances ≥3.0 Å (oxygen atoms numbered as in 1Z0D) (Fig. B6a). (Hence, the PDB priority order is O1 > O3 > O2 > O4).
Assigning the top two oxygen priorities as O1P and O2P, respectively (Fig. B6b) converts the two remaining oxygen atoms into a prochiral pair. Promoting the ‘front’ oxygen to 18O gives phosphorus S chirality, thus identifying it as pro-S (O4P). By a similar analysis, the ‘rear’ oxygen (H-bonded to Leu80) is pro-R. Hence, the rear oxygen is designated O3P and the front oxygen is O4P (Fig. B6b) (NB The PDB file assigns PA and PB to the GDP ligand).
5 Recommendations for labelling fluorine and other atoms in phosphate transition state analogues
5.1 C1. Tetrahedral phosphate mimics – trifluoroberyllates
These use a two-symbol code that may be expanded to four when there are additional fluorine atoms in the species.
There are over 100 examples of trifluoroberyllates (BeF3−) in the PDB (PDB ligand: BEF). This phosphate mimic is invariably attached to a carboxylate or terminal phosphate oxyanion. Labelling the three fluorine atoms will follow the same rules as for the three oxygen atoms in a terminal tetrahedral phosphate. Prochirality10 identification can be applied if one of the three fluorine atoms is promoted relative to the other two. In the context of enzyme-bound trifluoroberyllates, such promotion can be generally identified by co-ordination of one of the fluorine atoms to a protein-bound metal ion, typically magnesium. For example, in β-phosphoglucose mutase (PDB entry: 2WF8), a BeF3 is coordinated to Asp8, while a catalytic magnesium bridges Asp8 and one fluorine. This fluorine is thus identified as F1Be. The prochiral fluorine atoms F2Be and F3Be are designated by CIP rules, as shown in the example (Fig. C1a). As there is no other fluorine in this structure, these labels can be abbreviated to F1, F2, and F3, respectively.
(Note that in PDB entry 2WF8, these fluorine atoms were labelled F3, F1, and F2, respectively).
There is one example of a BeF2 moiety bridging two anionic oxygen atoms. In this case, F1Be and F2Be will correspond to the (pro-R) and (pro-S) stereochemistry assigned by CIP rules. Thus, in UMPCMP kinase (PDB entry: 4UKD), BeF2 bonds to ADP O3B and to UDP O3G (Fig. C1b). Thus, the (pro-R) fluorine is F1Be and the (pro-S) fluorine is F2Be. 16 In this unique and rather complicated example, CIP rules give priority to O5′ over O5′″, since adenine (A) takes priority over uridine (U) (Section A1d).
There may be less common species where there is no metal ion coordinating the trifluoroberyllate. In these, the hydrogen bonding priorities set out in B3c can be applied.
In an example of multiple metal coordination where the distances of separation from both metals to fluorine are less than the sum of the two van der Waals radii, the coordinating metal with higher atomic number will take priority.
5.2 C2. Trigonal bipyramidal phosphate transition state analogues – trifluoromagnesates and aluminum trifluorides
This requires a two-symbol code to describe (i) the identity of the fluorine relative to its congeners, and (ii) the identity of the core metal ion.
For AlF3 (PDB code: AF3), MgF3 (PDB code: MGF), and ScF3 tbp transition state analogues (TSA), the three fluorine atoms are invariably equatorial with two axial oxygen ligands to the 5-coordinate metal. Priority identification can be applied when one of the three fluorine atoms is promoted relative to the other two and directional priority for the two axial ligands is established. In the context of enzyme-bound trifluoromagnesates and aluminates, such promotion is readily identified by closest proximity of one fluorine to a protein-bound metal ion, typically a catalytic magnesium. The direction of viewing is determined by CIP priority of one of the apical (oxygen) atoms over the second, viewing down the priority O-metal bond. Thus, in the small G protein, Ras (PDB entry: 1OW3), MgF3 is axially coordinated to GDP via O3B and to a water, and CIP priority gives O3B > OH2. Thus, the fluorine coordinated to the catalytic magnesium is designated F1Mg. F2Mg and F3Mg are then identified in a clockwise progression from F1Mg when viewed from O3B to Mg (Fig. C2a). 17
In case of multiple metal ion coordination where both distances of separation are less than the sum of the two van der Waals radii, the coordinating metal with highest atomic number will take priority. In cases where two fluorine atoms are coordinated to two equivalent metals, as for cAPK (PDB entry: 1L3R), in which the tbp complex of ADP·MgF3 is liganded to two catalytic magnesium atoms, F1Mg is prioritised as the fluorine coordinated to the magnesium of higher priority. Metal priority shall be determined by its amino acid coordination (see Section B3c). Viewing priority is determined by: O3B > O-Ser21′. In cAPK, one catalytic magnesium is coordinated to Asn171, to Asp184, and to a water; the second magnesium is coordinated to Asp184 and to two waters. Hence, the magnesium linked to Asn171 has priority and is coordinated to F1Mg; F2Mg, and F3Mg follow in clockwise progression (Fig. C2b).
In the absence of fluorine coordination to a metal, hydrogen bonding to amino acids can be used to determine fluorine priority (see section B4c). 18
A significant number of structures in the PDB ( > 24) have a trigonal bipyramidal complex assigned as tetrafluoromagnesate(2−) (PDB ligand: MF4). The best resolved of these (PDB entry: 1WPG, 2.30 Å resolution) has electron density and bond lengths that can be equally well assigned as a regular Asp351-CO2−·MgF3−·OH2 complex. This can be labelled as for C2a (above) using coordination to a catalytic Mg to give priority to F1. 19
5.3 C3. Octahedral phosphate transition state mimics
This requires a two-symbol code to describe (i) the identity of the fluorine relative to its congeners and (ii) the identity of the core metal ion.
For tetrafluoroaluminate, AlF4− octahedral TSA analogues (PDB ligand: ALF), the four fluorine atoms are invariably equatorial with two trans-oxygen ligands to the 6-coordinate aluminium. Priority identification can be applied by promoting one of the four fluorine atoms relative to the other three. In the context of enzyme-bound tetrafluoroaluminates, such promotion is invariably identified by closest proximity of one fluorine to a protein-bound metal ion, usually magnesium. The direction of viewing is determined by CIP priority of one of the apical oxygen atoms over the second and viewing down the priority O-metal bond. Thus, in the structure of βPGM (PDB entry: 4C4R), the fluorine coordinated to the catalytic magnesium is identified as F1Al, while F2Al, F3Al, and F4Al follow in clockwise progression viewed from Asp8–Od1, which has priority over glucose-O6. (NB The corresponding PDB designations are F2, F1, F3, and F4, respectively) 20 (Fig. C3a).
There are at least 3 examples of octahedral trifluoroaluminate complexes (in PDB to 2015) having three fluorines in equatorial positions with the fourth equatorial ligand identified as oxygen. An example of this is the transition state analogue for enzymatic hydrolysis of dUTP (PDB entry: 4DL8). Axial priority for viewing is established by the CIP precedence of O3A > OWat401 (Fig. C3b). One fluorine is coordinated to two catalytic magnesium atoms and so is designated F1B. A progression viewed in the priority direction then identifies the bridging oxygen as the second priority ligand, O1B, with F2B and F3B completing the clockwise equatorial sequence.
In case of multiple metal coordination, the coordinating metal with highest atomic number will take priority, where the distance of separation is less than the sum of the two van der Waals radii (as for Fig. C2b above).
In the absence of fluorine coordination to a metal, hydrogen bonding to amino acids will be used to determine fluorine priority. Thus, in the fructose 2,6-bisphosphatase reaction of the enzyme PFKFB3, an AlF4− complex with His253Ne2 has been described (PDB entry: 3QPW. Fig. C3c). This octahedral complex is completed by water coordination trans to the histidine nitrogen. The four fluorine atoms are coordinated F1 to water, F2 to Arg252 and Gln388, F3 to His387 and water, and F4 to Arg252 and Asn259 (fluorines numbered as in PDB 3QPW). As F2 is coordinated to Ne of Arg252 and F4 is coordinated to Arg252-NH1, F2 takes priority since its H-bonding is to the nitrogen nearer to Cα of the lowest numbered coordinating amino acid. Hence, F1Al is coordinated to Arg252 and Gln388 and the progression to F2Al, F3Al, and F4Al proceeds clockwise as viewed from the water apex of the octahedral complex (CIP priority is O > N, magenta arrow) (Fig. C3c). 21
6 Recommendations for labelling vanadate and tungstate analogues of phosphates
6.1 C4. Vanadates
Orthovanadate, VO4 3– , is encountered as an analogue of phosphate in a variety of forms. They are invariably trigonal bipyramidal and thus mimic a five-coordinate phosphoryl transfer process.
Monosubstituted vanadate (V). In isolation, vanadate (PDB ligand: VO4) can mimic the transition state for phosphoryl group transfer as a trigonal bipyramidal complex substituted by either one or two axial oxygen ligands that represent nucleophile and leaving group. A typical example is the Xac nucleotide pyrophosphatase/phosphodiesterase structure (PDB entry: 2GSO) where the vanadate is axially coordinated to Thr90. The three equatorial oxygen atoms are numbered O1V, O2V, and O3V, with the axial oxygen O4V being trans to the hydroxylic oxygen of Thr90 (Fig. C4a). The equatorial oxygen coordinated to two zinc ions takes priority and is O1V. The direction of viewing is determined by the priority Thr90 oxygen > O4V (magenta arrow). Thus, a clockwise progression identifies O2V at the front and O3V at the rear of the trigonal planar array.
Disubstituted vanadate. A transition state analogue complex for phosphorylation of glucose 1-phosphate on O6 by α-phosphoglucomutase has vanadate linearly coordinated by oxygen-3 of Ser116 and by oxygen-6 of glucose 1-phosphate (PDB entry: 1C4G). CIP priority analysis gives O6G > O3S. The three equatorial oxygen atoms take priority from O1V by its coordination to cobalt, which substitutes for a native catalytic magnesium. Assignment of O2V and O3V follows a clockwise progression when viewed from O6G (magenta arrow) (Fig. C4b).
For the nucleoside-diphosphate kinase from B. burgdorferi, a vanadate transition state complex links ADP and His134 as axial ligands (PDB entry 4DZ6). There is no catalytic metal to coordinate the three equatorial oxygen atoms. Thus, oxygen H-bonded to Lys13 takes priority as O1V over oxygen O2V H-bonded to Arg94, while O3V is not H-bonded to any amino acid. These assignments are in accord with those in the PDB entry.
Trisubstituted vanadate. Tyrosyl-DNA phosphodiesterase (Tdp1) is a DNA repair enzyme that catalyzes the hydrolysis of a phosphodiester bond linking a tyrosine residue to a DNA 3′-phosphate. Orthovanadate is central in a transition state analogue structure in which vanadium is linked to the tyrosine oxygen, to the 3′-oxygen of the scissile nucleotide, and to His262 of the enzyme (PDB entry: 1RFF). Axial ligand priority is Tyr-O > HisNε2. Equatorial ligand priority is assigned to Thd-O3′. Hence, O2V and O3V follow in a clockwise progression when viewed from the Tyr-oxygen (Fig. C4c).
Cyclic trisubstituted vanadate. Trisubstituted vanadate provides a transition state analogue structure for hairpin ribozyme cleavage of a phosphodiester (PDB entry: 1M5O). The axial O2′ has CIP priority over the axial O5′. Priority in the three equatorial oxygen atoms is taken by the ribose-O3′ leading to assignment of O1V followed by O2V in a clockwise progression (Fig. C4d).
6.2 C5. Tungstates
Tungstate(VI) ion, WO4 = (PDB ligand code: WO4) is a mimic of tetrahedral phosphate in a small but significant range of structures in the PDB. In such systems, two oxygen atoms need to be assigned priority to enable the remaining two to be assigned by prochirality rules.
Isolated tungstate(VI) with two metal ions. In a structure of purple acid phosphatase (PDB entry: 3KBP), an isolated tungstate(VI) ion mimics phosphate. It is coordinated both to zinc and to iron. Zinc, with atomic number 30, takes CIP priority over iron (atomic number 26) and so the two tungstate oxygen atoms coordinated to these metal ions are labelled O1W and O2W, respectively (Fig. C5a). The remaining two tungstate oxygen atoms are now prochiral and can be labelled O3W and O4W by CIP rules described above.
Isolated tungstate(VI) with one metal ion. In a structure of a tungstate complex of CheYN59D/E89R, the isolated tungstate(VI) ion is coordinated to manganese and several amino acids (PDB entry: 3RVS). Thus, O1W is identified by its coordination to manganese. Coordination to oxygen gives precedence over coordination to nitrogen. Coordination to oxygen is only considered if the distance of the heavy atoms ≤3.0 Å (see Section B3c). Hence, O2W is coordinated to Asp59 and takes precedence over the third oxygen that is coordinated to Thr87. The remaining two tungstate oxygen atoms are now prochiral and can be labelled O3W and O4W by CIP rules described above (Fig. C5b).
Isolated tungstate(VI) with no metal ion. For an isolated WO4− species, a similar procedure of prioritisation by amino acid coordination can be used to identify O1W and O2W. O3W and O4W can then be assigned by the prochirality procedure. Thus, in a structure of Yersinia enterocolitica PTPase complexed with tungstate (PDB entry 3F9A), an isolated tungstate is encircled by a loop of amino acids 404–409, with three of its oxygen atoms coordinated to NHs. As isolated water coordination is ignored, 22 priority is given to coordination from Arg404 to O1W followed by coordination from Val407 to O2W. Hence, O3W and O4W are now prochiral and can be assigned using CIP rules (Fig. C5c).
Tungstate(VI) coordinated to a substrate ligand. A compound example of tungstate as a dual analogue of phosphate is found in the structure of a protein of the histidine triad family in which adenosyl 5′-ditungstate (PDB ligand: ADW), an analogue of ADP, is coordinated to His112 (PDB entry: 1KPE). This situation calls for labelling both tungstens and seven oxygen atoms, since the first tungstate is a trigonal bipyramidal TSA of PA and the second tungstate is a tetrahedral analogue of PB of ADP. Tungsten WA is equatorially linked to the adenosyl 5′-oxygen and axially linked to His112-Nτ. As in the case of polyphosphates (Section B1c), the bridging oxygen to WB is designated O3A. This enables the assignment of the two prochiral equatorial oxygen atoms as O1A and O2A (when viewed in the axial direction O3A to Nτ). For WB, oxygen O3A has highest CIP priority because it is coordinated to WA. The oxygen coordinated to Gly105 takes precedence over the oxygen coordinated to Ser107 and is therefore identified as O1B. This enables the prochiral pair of oxygen atoms to be assigned as O2B and O3B as shown (Fig. C5d).
The recommendations presented here have been developed to describe molecules derived from orthophosphoric acid and its derivatives, analogues, and transition state analogues. We have built on the existing usage of nomenclature wherever possible, especially on Cahn-Ingold-Prelog Rules. We have introduced only one Rule to override CIP analysis, and that for the purpose of simplicity of use by non-experts in stereochemistry. In our experience, the recommendations have worked well for the most demanding species we have examined, e.g. C3c and C5d. However, we recognise that the recommendations may be equally relevant to other species with tetrahedral geometry, such as sulfates and sulfonamides, or with tbp or octahedral geometries. We also recognise that there may be existing, or as yet non-existent, structures that could require an extension of these recommendations, and we remain receptive for advice on such problems.
Appendix – A guide for the use of Cahn-Ingold-Prelog rules for prochirality
Short procedure for identification of paired non-bridging oxygen atoms (or paired fluorine atoms) using Cahn-Ingold-Prelog Rules for prochirality (enantiotopicity)
Two non-bridging oxygen atoms bonded to the same phosphorus are enantiotopic if promoting one of them from isotope-16 to isotope-18 generates the opposite enantiomer compared to promotion of the other. 23 This is illustrated for methyl phenylphosphonate (Fig. X1). Promoting the ‘front’ oxygen (Step a) gives molecule (A) where the phosphorus is a stereogenic centre and is labelled R in Cahn-Ingold-Prelog nomenclature. Promoting the ‘rear’ oxygen (Step b) gives molecule (B) where the stereogenic phosphorus is labelled S. This analysis is based on the CIP priority rule O(Me) > 18O > 16O > C; on viewing the face of the P-centered tetrahedron with the lowest priority ligand (C) at the rear (magenta arrow), a clockwise progression from high to low priority is designated R (as shown) and an anticlockwise progression is S. Because A and B are enantiomers, the two non-bridge oxygen atoms are enantiotopic. In extension, the paired, non-bridge oxygen atoms can be labelled (pro-R) for the front one (clockwise progression) and (pro-S) for the rear one (C, right).
Two non-bridging oxygen atoms bonded to the same phosphorus are diastereotopic if promoting one of them from isotope-16 to isotope-18 generates a different diastereoisomer compared to promotion of the other. 24 In the case of adenosine 5′-diphosphate (ADP), the two non-bridging paired oxygen atoms on PA are diastereotopic. Promoting the ‘front’ oxygen to 18O generates a new stereogenic centre at PA (D; CIP label S), while promoting the ‘rear’ oxygen to 18O generates a stereogenic centre with the opposite sense at PA (E; CIP label R) (Fig. X2). NB the D and E stereoisomers are not mirror images because the stereochemistry of the D-ribose is unchanged. Since they are not enantiomers, they are termed diastereoisomers. 25
We can now apply the priority rules described in Section X1 to label the non-bridging oxygen atoms on PA in ADP. This analysis begins with the CIP priority rule that ranks di-coordinate oxygen above mono-coordinate oxygen. Thus, O5′ and O3A rank above the two non-bridging oxygen atoms. For these bridging oxygen atoms, relative priority is determined by the next atom in the chain: priority is given to the atom with the higher atomic number. In the case of ADP, the sub-adjacent atoms along the chain are PB and C5′. Hence, O3A has priority over O5′ as P has a higher atomic number than C. The CIP priority ranking is thus O3A > O5′ > 18O > 16O (Fig. X3). 26
Viewing the P-centered tetrahedron in stereoisomer (D) from the face with 16O at the rear gives an anticlockwise progression from high to low priority ligands (Fig. X3 left) and so PA in D has S-configuration. Hence, the two paired-oxygen atoms in ADP can be labelled pro-S for the front one (as its promotion to 18 O makes PA an S chiral centre) and pro-R for the rear one (as its promotion to 18 O makes PA an R chiral centre) (Fig. X3 center). We can now apply CIP Rule 5 that gives R priority over S. Thus, the pro-R oxygen is labelled O1A and the pro-S oxygen is labelled O2A (Fig. X3 right).
The rigorous application of the CIP rules almost inevitably means that there are unexpected outcomes. For example, the stereochemistry of the non-bridging oxygen atoms at PB in guanosine 5′-triphosphate (GTP) and in γ-thioGTP (GSP) have opposite assignments. For GTP, the rules for the in-chain atoms flanking PB identify O5′ bonded to C5′, thereby taking priority over all oxygen atoms bonded to PG (O1G, O2G, O3G). Hence, the priority sequence for the four GTP oxygen ligands at PB is O3A > O3B and thus the front oxygen is pro-R and the rear oxygen is pro-S (Fig. X4a). Hence, the pro-R oxygen is labelled O1B and the pro-S oxygen oxygen is O2B (Fig. X4a).
By contrast, for GTPγS, the sulfur atom on PG takes CIP priority over O5′, with the consequence that O3A takes priority over O3B (Fig. X4b). The result is that in GTPγS (as presented) the rear oxygen is pro-R (and thus O1B) and the front oxygen is pro-S (and thus O2B) (Fig. X4b). This is the opposite 3D spatial outcome compared to GTP. We note that a like situation holds for GTPγF, but not for γ-amino-GTP.
The above problem is readily resolved by a user-friendly solution.
The practical use of these interlinked analogues of nucleoside triphosphates, especially for ATP and GTP, calls for a simplified nomenclature system, designed to deliver comparable labels for diastereotopic oxygens on non-terminal phosphates across the series of phosphoramidate, phosphonate, phosphofluoridate, and phosphorothioate analogs of NTPs. That can be achieved by the use of a new nomenclature rule that gives priority to bridging oxygens in an extended polyphosphate chain based simply on their position in the chain, rather than on their often convoluted substituent atom priorities:
“Bridging atoms in polyphosphate chains should be given priorities that increase with separation from the priority nucleoside oxygen: On′ < O3A < O3B < O3G < O3D < etc.”
This rule, already described and used in Section 4B1d, is applied here to resolve the nomenclature discrepancy for O1B and O2B encountered in structures X4a,b (above). In the case of GTP analogs, having O3G replaced by another heavy atom, the rule assigns priorities as: O3B > O3A > O1B > O2B (Fig. X5a). This now assigns the same descriptors to the prochiral oxygens at PB as that in Fig. X4b (and the contrary to that for Fig. X4a).
The simplicity of use of this rule is also illustrated in naming the diastereotopic pairs of oxygens on PA, PB and PG for adenosyl-5′ tetraphosphate, p4A (Fig. X5b). The priority order for the bridging oxygens, On′ < O3A < O3B < O3G, results in a consistent naming for the non-bridging oxygens relative to their spatial location as O1A, O1B, and O1G (all frontal) and O2A, O2B, and O2G (all rearwards) (Fig. X5b).
Membership of the sponsoring body
Membership of the IUPAC Organic and Biomolecular Chemistry Division Committee for the period 2014–2015 is as follows:
President: M. J. Garson (Australia); Vice President: M. A. Brimble (New Zealand); Secretary: A. Griesbeck (Germany); Past President: K. N. Ganesh (India); Titular Members: G. Blackburn (UK); A. Brandi (Italy); T. Carell (Germany); B. Han (China); F. Nicotra (Italy) N. E. Nifantiev (Russia); Associate Members: A. Al-Aboudi (Jordan); V. Dimitrov (Bulgaria); J. F. Honek (Canada); P. Mátyus (Hungary); A. P. A. E. Knight: Total internal reflection fluorescence microscopy 1319 Rauter (Portugal); Z. Xi (China/Beijing); National Representatives: Y.-M. Choo (Malaysia); O. M. Demchuk (Poland); M. Ludwig (Czech Republic); V. Milata (Slovakia); M. Olire Edema (Nigeria); P. M. Pihko (Finland); N. Sultana (Bangladesh); H. Vanik (Croatia); B.-J. Uang (China/Taipei); T. Vilaivan (Thailand).
Membership of the IUPAC Organic and Biomolecular Chemistry Division Committee for the period 2016–2017 is as follows:
President: M.A. Brimble (New Zealand); Vice President: F. Nicotra (Italy); Secretary: A. Rauter (Portugal); Past President: M. Garson (Australia); Titular Members: P. Andersson (Sweden); J. Clardy (USA); N. Nifantiev (Russia); G. Pandey (India); J. Scott (UK); Chen Xi (China): Associate Members: V. Bolzani (Brazil): T. Carell (Germany); E. Uggerud (Norway); O. Demchuk (Poland); N. M. Carballeira (Puerto Rico); B. Sener (Turkey); National Representatives: E. Naydenova (Bulgaria); Ken-Tsung Wong (China/Taipei); M. Ludwig (Czech Republic); S. Marque (France); A. Al-Aboudi (Jordan); I. Shin (Korea); K. Lammertsma (Netherlands); M. Ali Munawar (Pakistan); V. Milata (Slovakia); L. Mammino (South Africa).
J. R. Knowles. Annu. Rev. Biochem. 49, 877 (1980). Google Scholar
G. Lowe. Acc. Chem. Res. 16, 244 (1983). Google Scholar
R. S. Cahn, C. Ingold, V. Prelog. Angew. Chem. 78, 413 (1966). Google Scholar
R. S. Cahn, C. Ingold, V. Prelog. Angew. Chem. Int. Ed. Engl. 5, 385 (1966). Google Scholar
G. A. Orr, J. Simon, S. R. Jones, G. J. Chin, J. R. Knowles. Proc. Natl. Acad. Sci. USA 75, 2230 (1978). Google Scholar
K.-F. Sheu, P. A. Frey. J. Biol. Chem. 253, 378 (1978). Google Scholar
R. L. Jarvest, G. Lowe. J. Chem. Soc. Chem. Commun. 364 (1979). Google Scholar
S. R. Jones, L. A. Kindman, J. R. Knowles. Nature 275, (5680), 564 (1978). Google Scholar
M. W. Bowler, M. J. Cliff, J. P. Waltho, G. M. Blackburn. New J. Chem. 34, 784 (2010). Google Scholar
S. C. Kamerlin, P. K. Sharma, R. B. Prasad, A. Warshel. Quart. Revs. Biophys. 46, 1 (2013). Google Scholar
N. J. Baxter, L. F. Olguin, M. Goličnik, G. Feng, A. M. Hounslow, W. Bermel, G. M. Blackburn, F. Hollfelder, J. P. Waltho, N. H. Williams. Proc. Natl. Acad. Sci. USA 103, 14732 (2006). Google Scholar
N. J. Baxter, G. M. Blackburn, J. P. Marston, A. M. Hounslow, M. J. Cliff, W. Bermel, N. H. Williams, F. Hollfelder, D. E. Wemmer, J. P. Waltho. J. Am. Chem. Soc. 130, 3952 (2008). Google Scholar
Y. Jin, M. J. Cliff, N. J. Baxter, H. R. W. Dannatt, A. M. Hounslow, M. W. Bowler, G. M. Blackburn, J. P. Waltho. Angew. Chem. Int. Ed. 51, 12242 (2012). Google Scholar
J. Mishra, U. S. Bhalla. Biophys. J. 83, 1298 (2002). Google Scholar
M. A. G. Sillero, A. de Diego, E. Silles, F. Pérez-Zúñiga, A. Sillero. FEBS Lett. 580, 5723 (2006). Google Scholar
Nomenclature of Phosphorus-Containing Compounds of Biochemical Importance (Recommendations 1976) IUPAC-IUB Commission on Biochemical Nomenclature. Proc. Natl. Acad. Sci. USA 74, 2222 (1977). Google Scholar
International Union of Biochemistry and Molecular Biology, Biochemical Nomenclature and related documents, 2nd edition, Portland Press 1992, 109–114, 256–264, and 335 [ISBN 1-85578-005-4]. (see also IUPAC-IUB Joint Commission on Biochemical Nomenclature, Pure Appl. Chem. 55, 1273–1280 (1983)). Google Scholar
Abbreviations and symbols for the description of conformations of polynucleotide chains. Recommendations 1982. Eur. J. Biochem. 131, 9 (1983). Google Scholar
Newsletter of the Nomenclature Committees of the International Union of Biochemistry and Molecular Biology. Eur. J. Biochem. 122, 437 (1982). [also ref 12, p. 265]. Google Scholar
IUPAC Nomenclature of Organic Chemistry, IUPAC Recommendations and Preferred Names 2013 (H. A. Favre, W. H. Powell, Eds.), Royal Society of Chemistry, Cambridge, UK (2014). (a) P-93.2.4 p. 1215, (b) P-18.104.22.168 p. 1222–1223, (c) P-22.214.171.124 p. 1225–1226, H. A. Favre, W. H. Powell Royal Society of Chemistry (2013). Corrections http://www.chem.qmul.ac.uk/iupac/bibliog/BBerrors.html; p. 432ff P-42.3-4 Phosphorus acids etc.; p. 915ff P-67 Phosphorus acids etc.; p. 992ff P-68.3 Phosphorus compounds; p. 1215ff P-93.2.4 Stereochemistry of phosphates etc.; p. 1420ff P-105 Nucleosides; p. 1425ff P-106 Nucleotides.
Nomenclature of Inorganic Chemistry: IUPAC Recommendations 2005, Royal Society of Chemistry (2005), Corrections http://www.chem.qmul.ac.uk/iupac/bibliog/RBcorrect.html which has links to later corrections. PDF:http://www.iupac.org/nc/home/publications/iupac-books/books-db/book-details.html?tx_wfqbe_pi1[bookid] = 5; p.180ff, IR 126.96.36.199, octahedral coordination; p184ff, IR 188.8.131.52, bipyramidal coordination; p. 189ff, IR 184.108.40.206, chiral octahedral; p. 190ff, IR 220.127.116.11, chiral bipyramid.
The description of the trigonal planar monoanionic PO3 species as a phosphoryl group is in line with long-established usage in biochemistry, biomolecular chemistry, and molecular biology. We note that the IUPAC recommendation is that the species ‘phosphoryl group’ relates to the neutral, diatomic trivalent species ≡PO.
The Ångstrom as the unit of distance in protein structures is standard throughout biochemistry, bioorganic chemistry, and molecular biology. The IUPAC use of nm as the IS unit of sub-micrometer distance is readily accommodated by the conversion factor 1 nm = 10 Å.
In determining priorities, coordination to an isolated water is ignored, because the presence or absence of a particular isolated water in a crystal structure can be a function of the structural resolution achieved, which makes water a variable object. However, waters coordinated to metal ions can be used.
For CIP Rules see the IUPAC Blue Book p1162 . For the use of pro-R and pro-S see “Basic Terminology of Stereochemistry (IUPAC Recommendations 1996)” Pure Appl. Chem. 68, 2193–2222 (1996).
Coordination to an oxygen of an isolated water is ignored. This is because the presence or absence of water in a PDB structure may be a function of the resolution of the structure, and therefore may vary from one structure to another of the same protein-ligand complex. Also note the use of PDB style numbering for atoms in amino acids (which avoids the use of Greek symbols).
About the article
Published Online: 2017-04-29
Published in Print: 2017-05-01
Funding: International Union of Pure and Applied Chemistry, (Grant/Award Number: ‘USD 10,000’).
Citation Information: Pure and Applied Chemistry, Volume 89, Issue 5, Pages 653–675, ISSN (Online) 1365-3075, ISSN (Print) 0033-4545, DOI: https://doi.org/10.1515/pac-2016-0202.http://creativecommons.org/licenses/by-nc-nd/4.0/.