Reactivity in the hydrolysis of Acyl derivatives of bioorganic relevance

Abstract

The thesis entitled “Reactivity in the Hydrolysis of Acyl Derivatives of Bioorganic Relevance” is in two parts. Part A describes, in three chapters, studies on the possible stabilisation of transition states by hydrogen bonding. Part B describes, in two chapters, studies directed towards uncovering possible stereoelectronic effects in the mechanism of action of the lactam antibiotics.

Part A: Transition State Stabilisation by Hydrogen Bonding with Amide N-H The main objective was to study the kinetic effect of hydrogen bonding by the amide N-H moiety, and its possible amplification by relay along a chain of hydrogen bonds. The motivation for the study is the possibility that such effects may be employed by enzymes to disperse charge developed in a substrate at the transition state along the protein backbone. This could be the key to the Pauling theory of enzyme action. Specifically, the effects were sought in systems designed to undergo O or Nacyl cleavage, with an amide N-H moiety prepositioned near the carbonyl oxygen of the acyl group. A major complication is possible nucleophilic attack by the amide carbonyl, but this was avoided as described below. It should be noted that the effect sought is essentially a reversible hydrogen bond at the transition state. This is similar (but not identical) to general acid catalysis by an amide N-H, which is expected to be extremely small by the Brønsted law; thus, the effect is expected to be small but potentially substantial when amplified by relay. The effect also amounts to a form of internal solvation, and indeed secondary amides are excellent solvents for ionic processes.

Chapter I Offers an overview of organic and bioorganic reactivity with emphasis on recent conceptual developments. Chapter II Reviews the influence of hydrogen bonding on organic and bioorganic reactivity.

Chapter III Describes the attempts made to test the present hypothesis, divided into three sections: Section 1: The Anthranilate System Attempts were made to synthesize oligoanthranilamides (1) stepwise by reacting isatoic anhydride with methyl anthranilate in the neat state. These are potentially excellent models for the hydrogenbond relay described above. Higher oligomers (1b) and (1c) were obtained along with the expected 1a, suggesting the reactivity order: 1b > 1a > methyl anthranilate, attributed to increasing nucleophilicity of the terminal amino group in (1) with increasing length of the hydrogenbond relay. Parallel studies on methyl4aminobenzoate supported this conclusion. Further study of these systems was abandoned due to poor solubility of isatoic anhydride, though the results were encouraging.

Section 2: The Salicylate System Hydrolytic studies were carried out on novel 2carboxyphenyl [N(Nbenzoyl(alanyl))]alanate derivatives (2), in which charge relay via hydrogen bonds in sixmembered rings at the transition state is, in principle, possible. Importantly, the parent aspirin reacts via general base catalysis at neutral pH, at which nucleophilic attack by the amide carbonyl is precluded. Compounds (2) were prepared by DCCmediated coupling of benzyl salicylate with an alanine derivative, followed by benzyl deprotection. Rates of Oacyl cleavage were measured at pH 6.5 by monitoring salicylate release. At 39°C, the parent 2a was 1.9 times as reactive as aspirin, apparently demonstrating the soughtafter effect (buffer catalysis was negligible). Surprisingly, peptide 2b was only 0.7 times as reactive as 2a, attributed to preferential groundstate stabilisation in 2b by hydrogen bonding. Activation parameters support this:

Activation enthalpy: 18.0±0.1 vs 19.6±0.6kcal/mol Activation entropy: -22.3±0.1 vs -18.0±2.1e.u.

The entropy data in particular indicate greater groundstate preorganisation in 2b. As 2a may be similarly stabilised, the observed rate ratio of 1.9 may be an underestimate. Interestingly, Nmethyl derivative 2c, in which N-H hydrogen bonding is eliminated, is 1.2 times as reactive as 2a-likely due to loss of groundstate stabilisation as discussed above.

Section 3: The Phenanthridinone System Hydrolytic studies of the exocyclic acyl moiety in phenanthridinone derivatives (3) were undertaken to avoid the ambiguities encountered in the salicylate systems. The acyl aziridine moiety in 3a cannot be significantly stabilised in the ground state by hydrogen bonding (due to inhibited amide resonance) and nucleophilic participation by the endocyclic amide is ruled out sterically. 3a hydrolysed 5.3 times faster than Nbenzoylaziridine at pH10.9, as measured spectrophotometrically-clear evidence that amide N-H can stabilise a heterolytic transition state by hydrogen bonding. Again, given steric and electronic constraints, this may be an underestimate. Attempts to prepare the next higher analogue (bearing an additional N-H) to test the relay concept were unsuccessful. Preliminary results with ester 3b showed that it hydrolyses 0.7 times as fast as methyl4(Nbenzoylamino)benzoate, likely due to groundstate stabilisation as previously discussed.

Conclusion The evidence strongly indicates that amide N-H can stabilise heterolytic transition states by hydrogen bonding, although the relay concept remains unproven. The magnitudes observed are likely underestimates due to structural limitations of the model systems and the aqueous medium employed.