Metaphosphate anion

For comparison one could look at the well known case of the monoester monoanion, which is known to have a transition state close to the metaphosphate anion. 

It has to be assumed that these processes are occurring on the boundary between SN1(P) and SN2(P) mechanisms in whose transition states considerable P—0(—Ar) bond cleavage takes place. The lifetime of the resulting, more or less free metaphosphate anion 102 then depends upon the nucleophilicity of the surrounding solvent. With pyridine, for example, a very fast reaction occurs so that the overall process approaches an SN2 reaction. Acceleration of the reaction by amines such as 2,6-lutidine, which are disqualified from acting as nucleophiles by steric hindrance, or by solvents such as dioxane, whiche are presumably too... 

When monomeric metaphosphate anion POf (102) is generated form the phos-phonate dianion 170 in the presence of the hindered base 2,2,6,6-tetramethylpiper-idine, it undergoes reaction with added carbonyl compounds147), Thus, it phosphoryl-ates acetophenone to yield the enol phosphate, whereas in the presence of acetophenone and aniline the Schiff base is formed from both compounds, probably by way of the intermediate C6H5—C(CH3) (OPO e) ( NH2C6HS). This reactivity pattern closely resembles that of monomeric methyl metaphosphate 151 (see Sect. 4.4.2). 

This phosphotransferase [EC 2.7.2.1] catalyzes the thermodynamically favored phosphorylation of ADP to form ATP Aeq = [ATP][acetate]/ [acetyl phosphate] [ADP] = 3000). GDP is also an effective phosphoryl group acceptor. This enzyme is easily cold-denatured, and one must use glycerol to maintain full catalytic activity. Initial kinetic evidence, as well as borohydride reduction experiments, suggested the formation of an enzyme-bound acyl-phosphate intermediate, but later kinetic and stereochemicaT data indicate that the kinetic mechanism is sequential and that there is direct in-line phosphoryl transfer. Incidental generation of a metaphosphate anion during catalysis may explain the formation of an enzyme-bound acyl-phosphate. Acetate kinase is ideally suited for the regeneration of ATP or GTP from ADP or GDP, respectively. 

Acyl-phosphates can participate in so-called in-line phosphoryl transfer reactions, and depending on reaction conditions, acyl-phosphate compounds can form a metaphosphate anion ... 

The parallelism in reactions with HMPT and mixtures of phosphorus pentoxide and amines can best be realized by looking at the mechanism of HMPT reactions [1 5 ]. In the reaction of p-methoxy-benzylalcohol with HMPT which produce the corresponding N,N-dimethylbenzyl amine--p-methoxybenzyl phosphate was identified in the XP NMR spectrum of the reaction mixture. Since pyrophosphate was also observed during the reaction of the benzyl alcohol with HMPT, the intermediate formation of the metaphosphate anion can be assumed. Addition of the benzyl alcohol to that anion then produce the benzyl phosphate. Phosphate ions are known to be good leaving groups so that a benzylamine can easily be produced from the phosphate by reaction with di-methylamine released from the dimethylammonium ion. The phosphoric acid produced during the reaction is believed to react with HMPT, so that more dihydro-gen-bis(dimethylammonium) pyrophosphate is formed. 

However, structures containing two metaphosphate anions are very rare, an example is Pb2Cs3(P40i2)(P03)3 in which both cyclotetraphosphate and catenaphosphate gronps are present. 

Phosphoryl and nucleotidyl transfer reactions are nucleophilic displacements on phosphorus 3—6), and like analogous displacements on carbon (7) or on metal ions (5), have been found to take place by mechanisms varying between two extreme or limiting cases 1. In the dissociative or SnI mechanism the initial departure of the leaving nucleophile, yields the planar triply coordinate metaphosphate anion as a reactive chemical intermediate (3, 8), which then combines with the entering ligand on either face of the metaphosphate plane 2. In the associative... 

The second mechanism for a hydrolysis reaction at phosphorus involves the initial dissociative loss of the leaving group to generate the metaphosphate anion... 

Although phosphate monoesters chiral by virtue of oxygen isotope substitutions cannot be used in stereochemical studies of phosphate monoester hydrolysis (since there are only three stable isotopes of oxygen), they have been used profitably in studies of phosphoryl transfer reactions relevant to the question of the intermediacy of monomeric metaphosphate anion in phosphoryl transfer reactions (see Section III,A). The laboratories of Knowles and Lowe have reported general methods for the synthesis of phosphate monoesters chiral by virtue of oxygen isotope substitution, and these syntheses are summarized in this section. 

As noted previously, studies of the mechanisms of phosphate monoester solvolysis have been extended to the mechanisms of the analogous phosphorothioate ester solvolysis because the thiometaphosphate anion is believed to be more stable than the metaphosphate anion. Thus, a general method based upon P NMR spectroscopy for the configurational analysis of chiral thiophosphate monoesters (see Fig. 10) was described recently by Cullis and co-workers (38). 

In summary, Cullis and Lowe have demonstrated that the generation of stable thiometaphosphate in solution is possible. The next section describes analogous studies on probing for the existence of free metaphosphate anion, since it is this intermediate that has potential significance in enzyme-catalyzed phosphoryl transfer reactions. 

Knowles has cogently summarized evidence obtained from a wide range of studies of solvolyses of phosphate monoesters that supports the participation of metaphosphate anion (55). The most recent evidence will be briefly summarized in this section before the results of stereochemical studies are discussed. 

A very clever three-phase test for the detection of metaphosphate intermediates in phosphoryl transfer reactions has been described by Rebek and coworkers (44). The basis of this test is the use of two polymers suspended in solution. The donor polymer contains a potential precursor to metaphosphate anion, e.g., an acyl phosphate or a phosphoramidate, and the recipient polymer contains an acceptor nucleophile, e.g., an amine. After reaction and physical separation of the polymers, the recipient polymer is analyzed for covalently bound phosphate. Since very few of the phosphoryl groups to be transferred will be on the surface of the donor polymer, detection of significant transfer to the recipient polymer provides evidence for a diffusible intermediate, i.e., free metaphosphate anion. Significant transfer did occur in dioxane or acetonitrile suspensions of the polymers, thereby providing evidence for an intermediate. However, this test for diffusible and, therefore, relatively stable metaphosphate anion is compromised by the choice of solvent. Both dioxane and acetonitrile can provide unshared electron pairs for the highly electrophilic metaphosphate anion such that the actual species that migrates from the donor polymer to the recipient polymer may be a complex between metaphosphate anion and the solvent. Such a role for solvent has been investigated stereochemically, the results of which will be described later in this section. 

Ramirez and Marecek have investigated the solvolyses of 2,4-dinitrophenyl phosphate in aprotic and protic solvents and described reaction conditions that are consistent with the involvement of free metaphosphate anion 45, 46). In particular, the dianion of the reactive phosphate monoester is capable of transferring its phosphoryl group to /err-butanol whereas the monoanion of the same ester is essentially unreactive in the same reaction. Since rert-butanol is unreac-tive as a nucleophile for steric reasons, the phosphorylation of this alcohol is considered to be a diagnostic test for the involvement of the highly reactive metaphosphate anion. 

Skoog and Jencks initially investigated the rates of transfer of the phosphoryl group from phosphorylated 3-methoxypyridine to a variety of substituted pyri-dines and amines that differ in pK and concluded that free metaphosphate anion is not involved as an intermediate (50). The basis for this experiment is the expectation that if metaphosphate anion is a true intermediate, a change in the rate-determining step should occur as the basicity of the acceptor pyridine or amine is varied. When the acceptor is less basic (nucleophilic) than 3-... 

All of these studies are consistent with the solvolyses of phosphate monoesters involving some form of metaphosphate anion however, on the basis of these experiments, the question of whether the anion exists symmetrically solvated and, therefore, can be considered to be a free intermediate in solution remains uncertain. Knowles, Cullis, and co-workers have used stereochemical techniques to examine the intermediacy of free metaphosphate anion in hydrolyses and alcoholyses of chiral phosphate monoesters. These experiments are important since they place significant constraints on the lifetime of the reactive intermediate and, therefore, clarify the large number of observations made in other laboratories on the question of whether metaphosphate anion is sufficiently stable that it can be considered a mechanistically significant intermediate. 

The stereochemical consequences of the methanolyses of the monoanion of phenyl [ 0, 0, 0]phosphate, the dianion of 4-nitrophenyl [" 0, 0, 0] phosphate, and [ 0, 0, 0]phosphocreatine have been determined by Knowles and co-workers (35). The monoanion of phenyl phosphate behaves as a typical phosphate monoester in that its rate of hydrolysis is maximal at pH 4, where an intramolecular proton transfer is possible. The dianion of 4-nitrophenyl phosphate is highly reactive since protonation of the leaving group is not necessary. Finally, A-phosphoguanidines have been reported to be the most reactive phosphoryl compound (the chiral phosphocreatine can be enzymically synthesized from [ y- 0, 0, 0]ATP). Thus, the solvolyses of all three of these compounds are believed to involve the participation of metaphosphate anion. The meth-anolysis of each of these compounds proceeds with quantitative inversion of configuration. 

Given the inability to implicate free metaphosphate anion in these studies, Cullis and Rous examined the stereochemical consequence of an alcoholysis reaction (55). When [ 0, 0, 0]P -0-ethyl-P -thiopyrophosphate is treated with methyl iodide in ethanol, the S-methyl derivative rapidly decomposes to form 0-ethyl-S-methyl thiophosphate and ethyl [ 0, 0, 0]phosphate this reaction is thought to proceed via a mechanism involving metaphosphate anion, but, as would be expected on the basis of the observations by Knowles, the stereochemical consequence is nearly quantitative inversion of configuration. In dichloromethane solution, the decomposition of the oxygen chiral substrate yields ethyl [ 0, 0, 0]phosphate that has suffered extensive racemization (approximately 70%). This stereochemical result can be rationalized by the intermediacy of free metaphosphate anion in the absence of nucleophile acceptors. 

The laboratories of both Knowles and Cullis have described solvolysis conditions in which free metaphosphate anion can exist. Initially both laboratories investigated the possible stabilization of metaphosphate anion by acetonitrile, since this solvent was reported by Rebek et al. to allow a successful application of the three-phase text for free metaphosphate anion (44). The Harvard laboratory studied the reaction of phenyl (, 0, 0]phosphate with rert-butanol in acetonitrile (56), and the Leicester laboratory studied the reaction of [)8- 0, 0, 0]ADP with 2-0-benzyl-(5)-1,2-propanediol in acetonitrile (57). In both cases, complete racemization was observed, and this can be explained by the complexation of metaphosphate anion by the acetonitrile solvent. Thus, the success of the three-phase test of Rebek et al. for metaphosphate presumably can be attributed to diffusion of an acetonitrile-metaphosphate anion complex rather than free metaphosphate anion. 

The criterion given by Ramirez for the participation of metaphosphate in a solvolysis reaction was that the sterically hindered (and, therefore, nonnucleo-philic) alcohol ter -butanol could be phosphorylated. Accordingly, the logical extension of this hypothesis to stereochemical studies searching for conditions favoring free metaphosphate anion was to conduct solvolysis reactions in neat tm-butanol. The solvolysis of 4-nitrophenyl [ 0, 0, 0]phosphate in tert-butanol was studied in the laboratory of Knowles and the tm-butyl [ 0, 0, 0]phosphate product was found to be completely racemic (56). This result can be explained only by the generation of a free metaphosphate anion in the sterically hindered solvent, which, once generated and symmetrically solvent, can react with the solvent to form the sterically hindered product. This conclusion is supported by complementary P NMR experiments performed in the... 

Thus, even though free metaphosphate anion cannot be considered a mechanistically significant intermediate in enzyme-catalyzed phosphate monoester ester hydrolysis, since it is unlikely that an acceptor nucleophile will not be present to participate in a preassociative mechanism, it can exist in solution under appropriate solvent conditions when an acceptor nucleophile is unavailable. 

The pH dependent chemical hydrolyses of the phosphorothioate analogues 2 -, 3 - and 5 -UMPS have been investigated. Between pH 2 and 5 dethio-phosphorylation is between 200 and 300 times faster than dephosphorylation due to the higher stability of the intermediate thiometaphosphate compared to the metaphosphate anion. At pH values less than 1, no migration of the thio-phosphoryl group takes place but rather, desulfurisation occurs. 

---