CAPILLARY ELECTROPHORESIS AS A TOOL FOR THE CHARACTERISATION OF PECTIN FINE STRUCTURE

M.A.K. Williams, G.M.C. Buffet and T.J. Foster

Unilever Research Colworth, Colworth House, Shambrook, MK44 1LQ, Bedford, UK

1. INTRODUCTION

1.1 Pectin Fine Structure

While pectin is essentially a linear co-polymer of galacturonic acid and its methylesterified counterpart, it is arguably one of the most complex of the plant cell polysaccharides. This complexity, manifest in a host of possible fine structure variants, gives pectin its utility of function and, combined with its horde of accompanying enzymes in-vivo, its ability to respond to a myriad of environmentally triggered stimuli. More generally, any process capable of modifying pectin's molecular level polymeric characteristics, such as the molar mass, the degree and distribution of methylesterification, amidation and acetylation, and the content, length and distribution of neutral sugar side-chains, may yield a tangible change in its manifest macroscopic properties. The determination of such polymeric structure attributes forms an extensive and exciting field of research of which only a small part can be covered in this article. In particular, the work reported here is focused on the characterisation of homogalactan regions and, more specifically, on the elucidation of the distribution of methylesterified residues.

1.2 Conventional Characterisation of Methylesterification

1.2.1 Average Degree of Methylesterification. At present the classical titration method'is still the official method used for the determination of the average degree of methylesterification (DE) of a pectin sample. Traditional alternatives involve the liberation of methanol by de-esterifying enzymes or by acid or alkali treatments, and its subsequent quantification by chromatography, while more recently methods have been described using infra-red spectroscopy.

1.2.2 Intermolecular Distribution of Methylesterification. Previously reported chromatographic methods for obtaining the intermolecular DE distribution (the variation in DE between different molecules in the same sample) involve binding pectin onto an ion exchange column under conditions of low ionic strength and subsequent elution of the pectin with a buffer, the ionic strength of which is gradually increased.' Fractions are collected and classically the pectin mass in the column eluates has been determined by a colorimetric assay with meta-hydroxy diphenyl (MHDP) while the DE has been inferred from the elution time." However, it should be noted that the ion-exchange chromatography (IEC) method is not without its problems. Pectins with blockwise intramolecular DE distributions are found to elute from the column in an irregular manner, which is potentially a serious problem, as the intramolecular charge distribution will be polydisperse to some degree even in a sample in which all chains possess the same average charge per residue. Furthermore, the binding of pectin to the column will be a function of its degree of polymerisation since the equilibrium constant for the binding of a polyanionic molecule to a polycationic IEC stationary phase depends on electrostatic interactions summed over all groups.

In an attempt to correct for such problems a method has been described in which eluates from IEC have been analysed by size exclusion chromatography (SEC) using a dual detection system monitoring refractive index and conductivity. In this method the DE of the eluates is actually measured, using a calibration linking the refractive index / conductivity ratio to DE, developed using pectin standards, rather than inferred from the IEC elution time. However, the eluates from IEC may still be polydisperse with respect to DE and it should be noted therefore that the DE distribution derived using these average values is still an approximation to some degree.

1.2.3 Intramolecular Distribution of Methylesterification. In general there are two approaches to the determination of intramolecular DE distribution (the spatial distribution of methylesterified residues along the pectic backbone). These are i) to attempt to measure the distribution directly by recording a property of the residues that is sensitive to the local residue type environment, and ii) to fragment the chain according to rules that depend upon the residue type environment, and analyse the fragments generated. Broadly speaking NMR methods have been applied that follow the first methodology,' while many enzymatic and chemical methods' have been described that utilise the second approach. Recent advances in this area include the separation of methylesterified enzyme digest fragments using high performance anion exchange chromatography (HPAEC) at pH 5, and the use of tandem mass spectrometry, which is able to locate the position of methylesterified residues within the surviving enzyme-resistant fragments.

The question of how to exploit the rich vein of experimental data that is now being obtained, in order to derive fine structure descriptors that efficiently encode the maximum structural information and offer useful substrate discrimination, drives a further interesting area of the work. Recently the "degree of blockiness" of a pectin substrate has been defined using the results from endo-polygalacturonase (endo-PG) digests, based on the liberation of unesterified mono-, di-, tri- and tetra-mers of galacturonic acid. Furthermore, using the proportion of unesterified mono-, di- and tri-galacturonic acid released in combination with the percentage of the total substrate galacturonic acid that is liberated in an unesterified form, and the ratio of the sums of peak areas obtained for methylesterified and unesterified fragments, sequence similarities have been represented by distance trees using computational techniques commonly exploited in the analysis of DNA and proteins. These approaches have been used to compare pectins with various average DE values and differing intramolecular distributions (generated by random or blockwise de-esterification) and clearly demonstrate a useful discriminatory ability.

Complementary approaches focus on minimising the set of pectin fine structures that are consistent with available digest information, in order to gain the most realistic representation of the polymeric backbone. This philosophy is essentially one of chain reconstruction. Practically, this type of work involves the computer simulation of test substrates and their digestion, with subsequent comparison of digest characteristics with those measured experimentally. To date this has been attempted simply with the aim of the reconstruction of the measured molecular weight distribution. However, recent attempts have been made at building more specific models of enzyme action that could, in principle, be used in such a fashion. An obvious problem here is the practical issue of computational time. To construct, and subsequently digest according to specific enzyme rules, all (statistically) possible fine structures that exist for a pectin molecule of some 500 residues in length is simply impractical. However, as more is understood about pectin biosynthesis it is possible that by limiting the set of trial pectin substrates to those that are biosynthetically likely such an approach may become more realistic.

1.3 Capillary Electrophoresis

While capillary electrophoresis...