On the synthesis of N -maleoyl amino acids in aqueous media: cautionary tales for the unwary traveller

The high cost or protracted syntheses associated with N -maleoyl amino acid derivatives coupled with a considerable demand from both the life and physical sciences has spurred the search for facile, low-cost routes to these compounds. Herein, we demonstrate that a recently published method purporting to deliver N -maleoyl amino acids by cyclization of maleamic acids in water, instead results in a hydrolysis product, namely the corresponding hemi-maleate salt.


Introduction
There is a considerable demand for low-cost, high purity N-maleoyl amino acids (NMAAs) for use in such diverse pursuits as the preparation of polymers, 1 the elaboration of metal surfaces 2 and the synthesis of bioconjugates. 3 The widespread use of NMAA active esters as heterobifunctional crosslinkers has resulted in the development of a number of methods 4 for their rapid preparation; however, for applications requiring the free acids the synthetic options are less appealing.
The traditional methods used to prepare NMAAs are limited in their utility by either a lack of scalability, 5 that they are cumbersome from a practical perspective, 6 or that they require a protracted synthesis. 7It is, therefore, of particular interest when a new methodology appears in the literature offering a facile synthesis 8 of these highly desirable compounds.Two such procedures have now been published that employ water as the reaction solvent.
The first method to describe the preparation of NMAAs in aqueous media was that of Ondruš and coworkers 9 (Scheme 1).This involves the reaction of the desired ω-amino acid, for example -alanine 1, with the commercially available exo-3,6-epoxy-1,2,3,6-tetrahydrophthalic anhydride 2, to furnish the corresponding NMAA, such as 3-maleimido propionic acid 3.A detailed Scheme 1. Attempted preparation of maleimides using the Ondruš methodology.
In a recent paper, Song and coworkers 11 describe the preparation of NMAAs by heat-induced cyclization of maleamic acids.This method is akin to that of Rich and coworkers, 6 but in stark contrast to carrying out the reaction in toluene with azeotropic removal of water, as Rich did, Song and coworkers employed water as a solvent.This method is extremely attractive as it avoids all the problems associated with the traditional methods of preparing these compounds and even removes the need for harsh organic solvents, as used in a contemporary methodology. 8

Scheme 2. Preparation and attempted cyclization using the Song methodology i) Method A acetic acid 8 h ii) Method B water 4 h.
Unfortunately, analysis the products from the attempted aqueous cyclization of maleamic acids reveals that they are hydrolyzed.
The cyclization of N-(3-carboxy-acryloyl)--alanine 5, which was prepared, as described by Song and coworkers (Method A), by reacting maleic anhydride 6 with -alanine 1 (Scheme 2) in acetic acid, resulted in a product with an NMR spectrum identical to that described by the authors; however, this material has different physical and chemical properties to that of an authentic sample of 3-maleimido propionic acid 3 (Bachem GmbH).In order to establish if contamination by residual acid was impeding the cyclization in our hands, a quantity of the maleamic acid 5 was prepared in water, using a known methodology (Method B), 12 and then subjected to the same cyclization conditions.When the proton and carbon-13 NMR spectra of the products from both attempted cyclizations were compared (Figures 1 and 2) with the spectrum of an authentic sample of 3-maleimido propionic acid 3 as well as one of -alanine hemi-maleate 4 it was clear that the products of the attempted cyclizations were, as with the Ondruš methodology, unfortunately, the corresponding hemi-maleate 4.
The NMR shifts and coupling constants (all taken in DMSO-d6) for the hemi-maleate and the putatively cyclized materials were essentially identical and contrasted sharply with that of the authentic maleimide 3.As noted by Corrie, 10 the olefinic protons can be particularly diagnostic when comparing maleates to maleimides; in this instance, having a shift of 6.06 ppm in the hemimaleate 4 and one of 7.02 ppm in the authentic maleimide 3. The two methylene triplets have shifts of 3.01 and 2.58 ppm with a coupling constant of 6.8 Hz in the hemi-maleate and 3.63 and 2.51 ppm with a constant of 7.3 Hz in the maleimide species (Figure 1).The carbon-13 spectra (Figure 2) also contain marked differences.In particular, one of the carbonyl peaks which resonates at 167.29 ppm in the hemi-maleate 4 is at 170.69 ppm in the maleimide 3; the olefinic carbons are at 135.94 ppm in the hemi-maleate 4 and upfield at 134.56 ppm in the maleimide 3 and the methylene carbons are at 34.82 and 31.45ppm, in the hemimaleate 4, and at 33.29 and 32.39 ppm in the maleimide 3.In addition to the NMR data, the melting points of the products from the attempted cyclizations were in good accord with that of the hemi-maleate 4 (121-124 °C for Method A, 122-123.5 °C for Method B and 123-124 °C for the hemi-maleate).The melting point of the authentic maleimide 3 is significantly lower at 110-112 °C (lit.108-109 °C). 8In conclusion, contrary to the published method, maleamic acids are hydrolyzed on refluxing in water.

Experimental Section
General.NMR spectra were recorded on a Bruker Avance II 500 MHz instrument using DMSO-d6 as a solvent.Proton and carbon-13 spectra were recorded at 500 MHz and 126 MHz, respectively.Chemical shifts are reported in ppm and are calibrated against the residual solvent signal.Infrared spectra were recorded in the research laboratories of Dr. B. D. Alexander, University of Greenwich at Medway, UK, on a Perkin-Elmer Spectrum RX1 FT-IR instrument, and are reported in reciprocal centimetres.Melting points were determined on a Reichert hotplate apparatus and are uncorrected.Elemental analysis was carried out by Medac Ltd, Egham, Surrey, UK.