In this study, rheological synergistic consequences of starch/carrageenan/milk proteins mixed systems were investigated by considering the role of each biopolymer type as well as their intrinsic characteristics. Through this study, we showed that starch endogenous proteins have no impact on the viscoelastic properties of starch/carrageenan mixed systems in presence or not of milk proteins. However, these properties were seen to be strongly dependent on the type of gelling carrageenans. Indeed, the ι-carrageenan-based systems led to lower viscoelastic properties compared to the κ-carrageenan-based systems, mainly due to the conformational ordering and the network characteristics of each gelling carrageenan-type. It was also demonstrated that whatever the type of gelling carrageenans, swollen starch granules filled in carrageenan gel network led to improved viscoelastic properties in comparison to pure carrageenan gels; the resulting filled composite gel strength becoming most pronounced in presence of milk proteins. This important reinforcement of the carrageenan gel networks in presence of both swollen starch granules and milk proteins appeared to be preferentially driven by the exclusion effect of swollen starch granules as well as their role as “filler”. In addition, the dominant role of the exclusion phenomenon of swollen starch granules became limited upon increasing carrageenan concentration in the mixed system; this effect being slightly less pronounced in presence of milk proteins. Finally, in terms of structural organization, the ternary mixed system can be regarded as a filled composite gel in which carrageenan chains and casein micelles form a three-dimensional hybrid network filled and strengthened by swollen starch granules.

In this study, physico-chemical interactions between carrageenan chains and starch granules were investigated through methylene blue spectrophotometric method by considering the chemical and macromolecular characteristics of each polysaccharide. For that purpose, various proportions of κ-, ι- or λ- carrageenan and different types of starches (native, chemically modified, “deproteinated” or not) were used. Blue dextran spectrophotometric method was also developed to evaluate the quantity of water absorbed by starch granules and hereby their swelling behavior allowing to determine the accurate amount of trapped carrageenan. It was demonstrated that the interactions between starch granules and carrageenan chains are not strongly impacted by starch chemical modifications. However, experiments with “deproteinated” starch showed that starch endogenous proteins can play a part in starch-carrageenan interactions. Three coexisting phenomena were found to occur when starch is pasted together with carrageenan. That consisted in: (i) partial penetration of carrageenan in starch granules, (ii) partial “exclusion” of carrageenan by starch granules and (iii) a predominant adsorption of carrageenan on starch granules due mainly to osmotic pressure effect. Moreover, the electrostatic interactions between starch endogenous proteins and carrageenan chains played a minor role. As expected, the interaction level appeared to be strongly depending from the carrageenan charge density: the lower the charge, the higher the interaction between starch and carrageenan, independently from starch chemical characteristics. Furthermore, it seemed that the rheological properties of the mixed systems are preferentially driven by the partial “exclusion” phenomenon as well as the role of starch granules as “filler”.

It is more and more realised that pectins are complex mixtures of many different molecules and research is directed towards the fractionation and characterisation of these pectic sub-populations. Since fractionation of pectins generally results in only low amounts of purified material, rapid characterisation methods using low amounts of samples are required. In this study, capillary electrophoresis was chosen to characterise pectins because only tiny amounts of sample are needed for the analysis. A new capillary electrophoresis (CE) protocol was developed to determine the degree of amidation, the degree of methyl-esterification (DM) and consequently the degree of substitution (DS) of pectins by analysing the pectins before and after removal of the methyl-esters. The CE results were compared with results obtained for the same pectins by using titration and Fourier transform infra-red (FTIR) spectroscopy methods. The CE method was found to be rather reliable resulting in small standard deviations for the DS and DAm. The CE method had the advantage of being rapid due to the limited sample preparation and automation of the analysis. In addition, CE was used successfully to determine the degree of blockiness of the free GalA residues over the pectic backbone.

Functionality of pectins as a food ingredient is strongly related to their chemical fine structure. Chemical characteristics of pectins are determined by many different parameters in their manufacture (choice of the raw material and extraction conditions). Pectin companies are thus in need of rapid methods to check the performance of extracted pectins. An important factor in the characterisation is the homogeneity of the pectin preparation, which is usually determined by laborious, time consuming, soft gel based chromatographic procedures. An HPLC method with a weak anion exchange column (WAX column) was found to discriminate between commercial pectins efficiently: pectins with similar DM (degree of methyl-esterification) or similar DS (degree of substitution: methyl-esters and amide groups) having different physical properties showed various populations. Furthermore the pectins with a blockwise or random distribution of the substituents showed different elution profiles. This indicates that the WAX column is sensitive not only to the amount of substituents but also to the distribution of these substituents.

Thickening and gelling properties of commercial amidated pectins depend on the degree of amidation and methyl-esterification, but also the distribution of these groups is of great importance. Methods have been developed during the last few years to determine the distribution of methyl esters over the pectic backbone. We applied the strategies developed for the analysis of high methyl-esterified pectins for studying the distribution of amide groups in amidated pectins. Low methyl-esterified amidated (LMA) pectins were digested before and after removal of methyl esters by an endo-polygalacturonase to determine the degree of blockiness of the substituents. The nature of the substituents (amide groups compared to methyl esters) did not modify the behavior of the enzyme. Oligomers released were separated by using high-performance anion exchange chromatography and pulsed amperometric detection (HPAEC-PAD) at pH 5. Fractions collected after on-line desalting were identified by using MALDI-TOF mass spectrometry. Oligomers were found to elute from the column as a function of their total charge. For the same overall charge and size, oligomers with methyl esters eluted before oligomers with amide groups. Both amide groups and methyl esters of the LMA pectins studied were found to be semirandomly distributed over the pectic backbone, but this may vary according to the amidation process used.

A protocol has been developed to fractionate sugar beet pectin using hydrophobic affinity chromatography. Three samples eluted from the column using 4 M NaCl as solvent (fractions 1A, 1B, and 1C), two fractions eluted using 2 M NaCl (fractions 2A and 2B), and one fraction eluted using water (fraction 3). The fractions were shown to be very polydisperse, and differences between the GPC refractive index and UV absorbance (214 nm) elution profiles demonstrated chemical heterogeneity. They were found to contain significantly different proportions of protein (1A, 2.79%; 1B, 0.97%; 1C, 0.77%; 2A, 1.41%; 2B, 5.09%; and 3, 5.89%) and ferulic acid (approximately 1A, 0.5%; 1B, 0.5%; 1C, 0.9%; 2B, 1.5%; and 3, 2%). The weight-average molecular mass, M(w), of the fractions also varied (1A, 153 kDa; 1B, 155 kDa; 1C, 306 kDa; 2A, 562 kDa; 2B, 470 kDa; 3, 282 kDa). Three fractions, that is, 1A, 1B, and 3, produced orange oil emulsions with a relatively small droplet size that were stable over a period of weeks. The other three fractions (1C, 2A, and 2B with higher M(w) values) produced emulsions with an initially larger droplet size, and the droplet size increased considerably over time. The increased droplet size may be influenced by the viscosity of the aqueous continuous phase. There was no simple relationship between protein or ferulic acid content and emulsification ability. For example, fraction 1B, which contained the lowest proportion of both protein and ferulic acid, produced stable emulsions of similar droplet size to fraction 3, which contained the largest proportion of protein and ferulic acid. The role of protein in the emulsification process was investigated by measuring the amount of protein in the aqueous phase before and after emulsification. It was clearly demonstrated that proteinaceous material adsorbed at the oil-water interface. It is evident that the emulsification properties of sugar beet pectin are influenced by the accessibility of the protein and ferulic acid groups to the surface of the oil droplets, the proportion of ester groups, and the molecular mass distribution of the fractions.

Two commercially extracted pectins having different physical properties but similar chemical characteristics were fractionated into sub-populations using ion exchange chromatography. Individual sub-populations were characterised using established strategies (galacturonic acid and neutral sugar content, degree of methyl-esterification) including the use of enzymes (endo- and exo-polygalacturonases) as analytical tool. Some purified populations showed similar degree of methyl-esterification whereas they were eluting at different ionic strength. It was shown that these populations mainly differed in the number of galacturonic acid moieties in ‘endo-polygalacturonase degradable blocks’ and in the location of these blocks within the molecule. The size of the blocks present at the non-reducing end of the pectin was also different within the molecules. The separation of pectins on anion exchanger combined with the use of enzymes allowed us to differentiate pectic sub-populations. Commercial pectins appear to be a mixture of several polymers differing in total charge as well as in the distribution of the charges.

Carrageenans, sulphated polysaccharides extracted from red seaweeds, are widely added to milk systems because of the gelation they promote and for their ability to recover a structure at rest after shearing. The gelation of iota-carrageenan is closely related to the helix-coil transition they undergo at approximately 48°C in milk. Carrageenan gels (e.g. iota with and without nu precursors) were formed in milk or permeate by decreasing the temperature from 60 to 10°C following various mechanical treatments, e.g. gelation and ageing of the gel at rest, gelation under shearing and ageing at rest. Gels were characterized rheologically during cooling and ageing, performed in the dynamic mode, to follow the mechanical recovery of the mixed systems at rest.In many cases, a strain-hardening behavior was observed using strain sweep measurements: above the upper limit of the linear domain, stress increased more than proportionally to the strain. Two types of iota-carrageenan were compared: iota with and without nu precursors. The content of the precursor affects the ability to form helical structures. Samples without precursor gave the highest level of helical structures. The carrageenan concentration effect was tested in the range 0.1–0.5wt%. Differences in sheared and unsheared gel properties, particularly in large strain conditions, are discussed in terms of gelation process, phase separation, carrageenan/milk proportion and carrageenan type. The industrial interest in using either type of carrageenan is discussed.

Weak gels of biopolymers are frequently used in food applications. Dynamic rheological measurements, performed at low strain in the linear domain are useful to characterize the network properties, e.g. gelation, ageing and mechanical recovery after shearing. However, the rupture properties of these gels are also of great interest to complete their characterization.Several weak gels containing carrageenan, xanthan–locust bean gum, pectin or alginate–pectin were investigated. The behaviour of these gels under large strain was studied either in dynamic mode or under constant low shear rate conditions. In many cases, a strain-hardening behaviour was observed: above the upper limit of the linear domain, the stress increased more than proportionally to the strain. A good agreement in stress–strain variations was observed between the results from dynamic and constant shear rate modes.This hardening effect, generally associated with a large deformability, gave to these gels a large resistance to rupture, with regards to their modulus. This can be related to the functionality of the biopolymer systems in terms of texture or suspension properties. As an example, in fluid gels, it allows the suspension of particles with a very weak network (typically a modulus of 1 Pa).

We have studied the influence of the calcium ion concentration, [Ca2+], and the pH on the storage (G′) and loss (G″) shear modulus at 1Hz of low methoxyl pectin solutions and gels. Upon lowering the temperature in the presence of Ca2+, G′ and G″ increase immediately followed by a further slow logarithmic increase with time. The immediate response increases with increasing [Ca2+] and decreasing temperature. The ‘gel’ temperature (Tg) where G′ and G″ cross increases with increasing [Ca2+]. However, G′ and G″ have a universal temperature dependence at all [Ca2+] if plotted as a function of T−Tg.The pH was reduced slowly by addition of GDL. Reduction of the pH to below 3 leads to the formation of strong gels for amidated pectin and much weaker gels for non-amidated pectin. For non-amidated pectin lowering the pH below 3 leads to a weakening of the gel formed by Ca2+, while for amidated pectin the gel is reinforced.

The sol–gel transition of iota carrageenan and pectin is followed by measuring the evolution of ultrasonic velocity of a compressional wave at a frequency of 500 kHz. Measurements are performed as a function of temperature, from about 90 to 20 °C. Results are compared to those obtained in rheology in oscillatory conditions. It is shown that the ultrasonic velocity is sensitive to the sol–gel transition when going from a viscous to a “solid-like” state with a large elastic component. However, the method fails when the transition occurs gradually to give a weak elastic gel.

Samples of calcium sensitive (CS) and non-calcium sensitive (NCS) pectin were compared in acid dairy drinks (ADD) with different protein concentrations. The CS pectin is more efficient in terms of stabilization than the NCS pectin and produces a network at all MSNF values. The gels formed are weak but present thixotropic properties. They are also very distortable and can be broken. The gels therefore present a yield point. The NCS pectin is quite good for stabilizing ADD showing that the calcium sensitivity is not the only parameter governing stability.

The negatively charged polysaccharide, iota-carrageenan has been mixed with two types of gelatin (PS and LH) which differ by their pI, about 9 and 4.5, respectively. At pH 7, gelatin PS chains have a net positive charge and gelatin LH chains have a net negative charge. Phase diagrams in non-gelling conditions (60°C) and pH 7 have been established in both 25 and 100mM NaCl. The iota-carrageenan/gelatin PS phase diagrams are typical of systems involving attractive interactions leading to the formation of complexes: both polymers are mostly in the lower phase and increasing salt concentration leads to a reduction of the phase separated domain. The iota-carrageenan/gelatin LH mixture shows a very different behavior. No macroscopic bulk phase separation is observed. However, an increase in the turbidity measured by spectrophotometry is observed for certain compositions which would indicate microscopic phase separation. From the limit between the turbid and transparent concentration domains, the binodal may be drawn. Considering the negative net charge of gelatin LH chains at pH 7, repulsive interactions were expected to govern the phase separation. However, in this case too, by increasing salt concentration the phase-separated domain is reduced, where an increase would have been expected for thermodynamic incompatibility. By following the evolution of the absorption spectrum of a methylene blue/iota-carrageenan/gelatin LH mixture as a function of the salt concentration, it has been shown that gelatin LH chains associate with carrageenan ones in low salt conditions up to 200mM NaCl.

Acylgalactosylceramides (AGC) from forebrains of normal and dysmyelinating (quaking and shiverer) mice were purified by Florisil column chromatography and preparative TLC. These procedures resolved the AGC on the basis of their Rf values into two main fractions which co‐nigrate with their homologs from rat forebrains. In control animals, AGC were detectable in mouse forebrains from the eighth postnatal day and reached maximal values within 20 days. The same developmental pattern was obtained in dysmyelinating shiverer mice but the AGC content was reduced to approximately 30% of control values. In quaking mutants, the AGC were hardly detected. They were also present in sciatic nerve of normal mice and to a lesser extent in trembler mice. Gas chromatography‐mass spectrometry analysis of both ester‐ and amide‐linked fatty acids isolated from AGC of normal and shiverer mice shows that the shiverer mutant AGC display a chemical structure similar to that of normal AGC. AGC constituents of control myelin are reduced by approximately 70% in shiverer myelin, indicating that these molecules can be considered as early markers of oligodendrocyte differentiation. The early arrest of myelinogenesis in the quaking animals and the near absence of AGC are in good agreement with this proposal. Moreover, the reduced amount of AGC in the trembler PNS indicates that AGC could also be early markers for differentiation of the Schwann cell.

1. The results obtained in this study confirm the previous ones related to the presence in Nereis (both in control and contaminated animals) of a low molecular weight protein having the capacity to bind Cd (Dennaï et al., 1986). This component probably also binds Zn and Cu.