Biomacromolecules 2003, 4, 1400-1409
1400
Effects of Amylopectin Structure and Molecular Weight on Microstructural and Rheological Properties of Mixed β-Lactoglobulin Gels Camilla Olsson,† Tuomo Friga˚rd,‡ Roger Andersson,‡ and Anne-Marie Hermansson*,† SIK-The Swedish Institute for Food and Biotechnology, P.O. Box 5401, SE-402 29 Go¨teborg, Sweden, and Department of Food Science, Swedish University of Agricultural Science, P.O. Box 7051, S-750 07 Uppsala, Sweden Received May 6, 2003
Nongelling amylopectin fractions from potato and barley have been used to form mixed β-lactoglobulin gels. The amylopectin fractions were produced by varying the time of R-amylase hydrolysis followed by sequential ethanol precipitation. The molecular weights, radius of gyration, chain length distribution, and viscosity of the fractions were established. The mixed gels were analyzed rheologically with dynamic mechanical analysis in shear and microstructurally with light microscopy, transmission electron microscopy, and nuclear magnetic resonance spectroscopy. The result of the gel studies clearly showed that small differences in the molecular weight of amylopectins have a significant influence on the kinetics of protein aggregation and thereby on the gel microstructure and the rheological behavior of the gel. Both an increase in the molecular weight and a higher concentration of amylopectins resulted in a more open protein network structure, with thicker strands of larger and more close-packed β-lactoglobulin clusters, which showed a larger storage modulus. The transmission electron micrographs revealed that degraded amylopectins were enclosed inside the protein clusters in the mixed gels, whereas nondegraded amylopectin was only found outside the protein clusters. The volume-weighted mean value of the molecular weight of the amylopectins was found to vary between 3.2 × 104 and 5.0 × 107 Da and the ratio of gyration between 14 and 61 nm. The maximum in chain length distribution was generally somewhat distributed toward longer chain lengths for potato compared to barley, but the differences in chain length distribution were minor compared to those seen in the molecular weight and ratio of gyration between the fractions. Introduction Mixed systems of biopolymers are often used to form products with specific textural properties based on their microstructure. Gel-forming biopolymers build up the microstructure, but nongelling components can also affect the structure formation by indirect means and thereby the final network structure and the texture. This has been shown for systems containing network-forming proteins by, for example, Bryant and McClements,1 Doublier et al.,2 Hermansson et al.,3 Olsson et al.,4 Tromp et al.,5 and Zasypkin et al.6 Even small alterations in the composition of nongelling components can result in dramatic changes in the protein network microstructure. In this paper, new information is presented on how small changes in the molecular weight of nongelling amylopectins affect the aggregation and gel formation of β-lactoglobulin (β-lg) protein. Adding nongelling polysaccharides to a gel-forming protein has been shown to strongly influence the kinetics of protein particle aggregation and thereby the final microstructure and texture. The stronger the tendency toward phase separation between the polysaccharide-rich phase and the * To whom correspondence should be addressed. Phone: +46-313351658. Fax: +46-31-833782. E-mail:
[email protected]. † SIK-The Swedish Institute for Food and Biotechnology. ‡ Swedish University of Agricultural Science.
protein-rich phase, the coarser the final protein microstructure will be. This has been shown for systems containing the protein β-lg and different polysaccharides by using CLSM and SEM.4,7 The kinetics of β-lg particle aggregation are extremely sensitive to variations in process conditions such as pH and heating rate, as well as to the presence of other ingredients such as polysaccharides and emulsifiers.4,8-13 The particle aggregation can be manipulated to start already before heattreatment, i.e., prior to gel formation, as well to continue after gel formation, depending on the situation. This kinetically governed particle aggregation of the β-lg microstructure is very complex, because small changes in the additives/ process parameters can have a dramatic impact on the final result. The gelation of β-lg has been viewed as a two-stage sequential process including aggregation reactions and denaturation with unfolding reactions.14 At a pH close to the pI of the protein, the aggregation reactions start prior to the denaturation. Within this pH range, the final protein microstructure consists of protein-rich domains with a diameter of the order of micrometers, so-called particulate gels.13 These large aggregates form via a complex mechanism in which both electrostatic and hydrophobic interactions as well as hydrogen bonding and disulfide cross-linking are involved.15
10.1021/bm030038e CCC: $25.00 © 2003 American Chemical Society Published on Web 07/26/2003
Properties of Mixed β-Lactoglobulin Gels
Adding polysaccharides to suspensions of whey proteins may induce depletion interactions, which arises because a polymer molecule loses conformal entropy when confined between colloidal particle surfaces. This causes an osmotic situation where the remaining pure dispersion medium will tend to flow out from the gap between the protein colloids in order to dilute the bulk water. As a result, an effective attraction occurs between the protein colloids and accelerates phase separation effects.16-18 That there is strong incompatibility between a protein close to its pI and a nonionic polysaccharide is well documented.19 Tunier et al.18 have studied the depletion-induced phase separation of whey protein when mixed with an exopolysaccharide and noted that the phase separation was similar to that observed for spinodal decomposition of a single binary liquid. We have recently observed that, when amylopectin is added to a solution of β-lg, the aggregation of β-lg particles also shows behavior typical of spinodal decomposition, i.e., the sudden appearance of lighter and darker domains over the whole sample with a contrast that increases in time.4 When phase separation and aggregation occur simultaneously, though with different kinetics dependent on the situation, many different types of microstructures can be generated by using a protein whose microstructure is trapped by gelling. The rheological and microstructural behavior of β-lg gels when mixed with nondegraded (pap HV) and mechanically degraded potato amylopectin (pap LV) has recently been demonstrated by Olsson et al.4,20-21 It was found that increasing both the concentration and viscosity of pap led to a more open β-lg network with larger cluster volumes and larger pore volumes between the strands of clusters on an overall level of structure. The structure parameters were measured on light microscopy (LM) images by using a stereological image analysis approach. The shear modulus, G′, increased with the concentration of pap at a constant protein concentration. However, the gels containing pap HV increased in modulus only up to 0.25% pap HV, whereupon the modulus decreased, because of the degeneration of β-lg network connectivity at higher concentrations. In this study, we have very specific data on molecular weights as well as on the radius of gyration and chain length distribution of two different nongelling amylopectins, potato and barley. Both modern biotechnology and traditional plant breeding have resulted in new starches containing, for example, only amylopectin or higher amylose content. The chemical structure and molecular weight of starch can be modified under controlled conditions by R-amylase hydrolysis. In principle, R-amylase is an endo-specific enzyme commonly regarded as acting randomly on high-molecularweight substrates and catalyzing the hydrolysis of R-(1f4) linkages in amylose and amylopectin, leading to a decrease in molecular weight as well as unit chain length.22-23 Varying the time of R-amylase hydrolysis followed by sequential ethanol precipitation can produce fractions with a different chain length distribution and molecular weight at relatively high yields.24 Hydrolysis of potato and barley amylopectin resulted in fractions with both a decreased radius of gyration and molecular weight. The unit chain length distribution
Biomacromolecules, Vol. 4, No. 5, 2003 1401
varied with the time of hydrolysis, but the sequential ethanol precipitation process was also structure-dependent. The aim of this study was to elucidate the effects of amylopectins varying in molecular weight, ratio of gyration, and chain length distribution on β-lg network microstructure and texture. The microstructures were examined by means of light microscopy (LM) and transmission electron microscopy (TEM). The texture was analyzed using dynamic mechanical analysis in shear. Another aim was to locate the amylopectin phase in the gels, and both TEM and nuclear magnetic resonance spectroscopy (NMR) were used for this purpose. Experimental Section Materials. The materials used were naturally amylosefree amylopectin starch from potato (pap) developed by Lyckeby Sta¨rkelsen and Svalo¨f Weibull using genetic engineering and barley (ba) produced by traditional plant breeding from the University of Saskatchewan, Saskatoon, Canada. β-Lactoglobulin (β-lg), WPI PSDI-2400, was obtained from MD Food Ingredients, Denmark. A detailed description of the materials is given elsewhere.20,24 All calculations in this study were done on a dry matter basis, and all concentration values stated in % refer to wt %. Preparation of the Amylopectins. Pap and ba were mixed with distilled water, and the dispersion was heated to 120 °C, to fully dissolve the granule remnants, and then cooled at room temperature (20 °C). The viscosity of the pap and ba solutions in the mechanically degraded amylopectins and the enzymatically degraded amylopectins was reduced by shearing in a homogenizer, Ultra Turrax, T25, (IKAlabortechnik, JANKE & KUNKEL, Staufen, Germany), at a rate of 20 500 min-1 for 1 min. The amylopectins were enzymatically hydrolyzed using R-amylase from B. amyloliquifaciens [(1f4)-R-D-glucan glucanohydrolase; EC 3.2.1.1] with an activity of 10 mU/g starch (Boehringer-Mannheim, Mannheim, Germany) for 5, 15, and 60 min. The enzyme was inactivated, and aliquots of this solution were transferred to centrifuge tubes. Ethanol was then added sequentially to 20 and 40% (w/v) concentration to precipitate the partially degraded amylopectin. The fractions obtained are noted with enzyme degradation timeethanol concentration, i.e., 5-20, 15-20, and 60-40. Molecular Characterization of the Amylopectins. Static Light Scattering Measurements. Light-scattering measurements were done at room temperature with a DAWN DSP multiangle laser light-scattering detector (MALLS) in chromatographic mode using a He-Ne laser light source operating at 632.8 nm (Wyatt Technology Corp., Santa Barbara, CA). The solvent delivery system LC-10AD was equipped with an online degasser DGU-4A (Shimadtzu, Kyoto, Japan). A 0.1 µm online filter was added between the pump and the injector. The mobile phase was 0.1 M sodium nitrate with 0.02% sodium azide, the flow was set for 0.5 mL/min, and the injection volume was 75 µL. A thermostated (30 °C) automatic injector model Waters 717 (Millipore Corp., Marlborough, MA) was used and a column oven-set at a temperature of 30 °C (Microlab Universal thermostat,
1402
Biomacromolecules, Vol. 4, No. 5, 2003
Aarhus, Denmark). The columns used were in series: a guard column Aquagel-OH Guard (Polymer Laboratories, Amherst, MA) and size exclusion columns OH-Pak 803-HQ, 804-HQ, and 806-M-HQ (Shoko CO., Ltd, Tokyo, Japan). The online concentration and viscosity were measured with a Viscotek model 250 combined refractometer and viscometer operated at 30°. Toluene was used as the calibration standard at a 90° scattering angle. The light intensity of the diodes at the other angles was normalized to the 90° angle using the monodisperse isotropic scatterer, pullulan P-20 (MachereyNagel, Du¨ren, Germany). Samples were prepared by dissolving the amylopectin fractions (1-3 mg/mL) in 90% DMSO with 0.1 M NaNO3 in a water bath with boiling water for 15 min. The software used was ASTRA version 4.72.03 (Wyatt Technology Inc.) and TriSEC version 3 (Viscotek Inc.). All calculations in Astra were carried out on fitted data and using the Zimm formalism (K*c/Rθ). The model of angular dependence (Manual Viscotek) on unfitted data did calculations in TriSEC. The dn/dc value used for amylopectin in water solutions (0.146 mL/g) was obtained from the literature.25-26 Debranching and High Performance Anion Exchange Chromatography (HPAEC-PAD). The native as well as the partially degraded amylopectin was dissolved in 0.5 mL of 0.0625 M NaOAc buffer (pH 3.6) in a boiling water bath for 15 min (∼1.5 mg/mL). The mixture was cooled to room temperature and 5 µL of isoamylase from Pseudomonas amyloderamosa (glycogen-6- glucoanhydrolase; EC 3.2.1.68; 71 000 U/mg, Hayashibara Biochemical Lab., Okayama, Japan) was added. The samples were debranched overnight in a shaking water bath at 38 °C, after which the enzyme was inactivated by boiling for 10 min. The debranched amylopectin was analyzed with high performance anion exchange chromatography (HPAEC) consisting of a DX 500 chromatography instrument equipped with a pulsed amperometric detector (PAD) ED-40 (Dionex, Sunnyvale, CA) under conditions described in Koch et al.27 The detector response of the PAD is not quantitative with respect to carbohydrate concentration. The amount of each individual oligosaccharide chain of specific length was calculated using relative response factors. Viscosity Measurements. The enzymatically degraded amylopectins of pap and ba were dissolved in boiling water, whereas the β-lg was dissolved at room temperature. The pH of the β-lg solution was adjusted to 5.40 with 0.5 M HCl and was thereafter mixed with the cooled amylopectin solution and placed in LS 30 (Contraves AG, Zu¨rich, Switzerland). The viscosities were measured at shear rates 55-100 (s-1) at a temperature of 22 °C. Mixed Gels of β-Lactoglobulin and Amylopectin. For the rheological and the microstructural measurements, β-lg and amylopectin solutions were mixed, degassed, and poured into moulds of aluminum with cylindrical cross-sections, diameter 15 mm and length 60 mm. The ends were closed with rubber stoppers at the bottom and with heat-proof tape (scotch 425) at the top. The cylinders were greased to prevent the samples from sticking to the mould. The cylinders were then placed in a FP40-MS Julabo programmable silicon oil bath (Julabo Labortechnik GmbH, Seelbach, Germany). The
Olsson et al.
temperature was increased to 90 °C at a heating rate of 2.5 °C/min and held at 90 °C for 1 h. Thereafter, the temperature was decreased to 20 °C at a cooling rate of 2.5 °C/min. The moulds were stored overnight, 15-18 h, at a temperature of 20 °C. For the NMR experiments, the gels were formed in 10 mm glass tubes with plastic stoppers. Rheological Characterization. Gels were cut into 5 mm high cylindrical test pieces for oscillatory measurements in shear with a parallel plate system. The measurements were done with a Bohlin VOR rheometer (Bohlin Rheology, Chichester, UK) equipped with a 15 mm plate, at a frequency of 1 Hz and a strain of 8 × 10-4. A thin filter paper with a diameter of 15 mm was placed on both plates to prevent slipping during measurements. The gel pieces were compressed by 5%, and the moduli and the phase angle were measured after a time sweep of 180 s when the moduli were constant. The amount of enzymatically degraded amylopectin was limited, but it was nevertheless enough to produce gels on at least 2 occasions with 2-3 gel pieces on each occasion. The error bars in the figures showing modulus behavior show the standard deviation. Light Microscopy. Small cubes (1.5 × 1.5 × 1.5 mm) were cut from gels prepared in the same way as for the oscillatory measurements. The preparation procedure followed that described by Langton and Hermansson.28 The procedure started with fixation with 2 vol % glutaraldehyde and 0.1% (w/v) ruthenium red in 0.1 M phosphate buffer of pH 5.9 for 3 h and then washing twice in buffer for 10 min. The gel cubes were then fixed a second time, with 1% (w/v) OsO4 for 2 h, followed by rinsing in the phosphate buffer, before being dehydrated in a graded ethanol series, 50, 70, 90, 95, and 99.5 vol %. The gels were then embedded in polybed, (TAAB 812). Semithin sections (1.0 µm) were cut out with glass knives, stained with toluidine blue, and examined under a LM, Microphot-FXA (Nikon Corp., Tokyo, Japan). At least gels from 2 different occasions were prepared. Transmission Electron Microscopy. The gels embedded for LM were also used for TEM. Thin sections of about 70 nm were cut with a diamond knife and double-stained with 5% uranyl acetate and 0.3% lead citrate, to visualize the protein phase. To visualize the amylopectin, the thin sections were double-stained with 1% periodic acid and thiosemicarbazide. The sections were examined in a TEM, LEO 906e, (LEO Electron Microscopy Ltd, Oberkochen, Germany) at an accelerating voltage of 100 kV. Image Analysis. A contextvision microGOP S2000 system (Contextvision AB, Linko¨ping, Sweden) was used for measuring the average pore size as the star volume, which estimates the volume-weighted mean volume. The star volume is defined as “the mean volume of all parts of an object which can be seen unobscured in all directions from a particular point”.21,29 Nuclear Magnetic Resonance (NMR) Spectroscopy. NMR experiments were done on amylopectin and β-lg mixed gels to measure the proton transverse relaxation times (T2). The instrument used was a Bruker DRX spectrometer operating on a standard bore (L52 mm) 400 MHz magnet (Bruker Spectrospin Canada, Milton, Ont., Canada), equipped with a microimaging accessory with a coil diameter of 10 mm
Biomacromolecules, Vol. 4, No. 5, 2003 1403
Properties of Mixed β-Lactoglobulin Gels Table 1. Weight Average Molecular Weights (M h w) and Radius of Gyration of (〈sj〉w) of the Different Amylopectin Fractions Used sample/fraction
M h w (Da)
〈sj〉w
Pap HV Pap LV Pap 5-20 Pap 15-20 Ba HV Ba 5-20 Ba 15-20 Ba 60-40
5.01 × 107 2.66 × 107 6.40 × 106 6.45 × 105 4.87 × 107 1.58 × 107 2.38 × 106 3.19 × 105
60 55 35 15 61 54 27 14
having a 180° pulse of about 35 ms. Spectroscopy was done at an ambient temperature (22 °C), and the 1H NMR 1D profiles were acquired using the CPMG spin-echo sequence: 90° x - (τ - 180° y - 2τ - 180° y - τ - Acq)n′. The delay time was 0.5 ms, and the number of echoes was 32 in order to record the decay of the signal. The T2 values of the mobile fraction, as opposed to rigid solids, in the samples were derived from the monoexponential decay curve according to the function described by Colquhoun et al.30 No attempt was made to investigate the contributions of crystalline regions (rigid solids) to the decay of the transverse magnetization or the compartmentalization of water and hence different T2 populations. A two-component model was attempted but it did not improve the degree of fit. Results and Discussion Characterization of the Amylopectins. Molecular Properties. The pap and ba samples were mechanically degraded and incubated with R-amylase for different periods of time and sequentially precipitated in different ethanol concentrations in order to obtain fractions of different molecular weights. (The yield of pap was slightly larger than the yield of ba.24) Light scattering data showed a decrease in both the weight average molecular weight, M h w, and the radius of gyration, 〈sj〉w, with increased lengths of R-amylase treatment. This is shown in Table 1 for different pap and ba fractions. The pap HV and ba HV have a M h w of 5.0 × 107 and 4.9 × 7 10 Da, and a 〈sj〉w of 60 and 61 nm, respectively. However, the analysis values of the native pap HV and ba HV starches are to be considered just as crude estimates because of the difficulty in dissolving and passing them through the chromatography system. For the pap samples subjected to mechanical and enzymatic treatment, the M h w ranged from 6.8 × 105 to 2.6 × 107 Da and 〈sj〉w from 15 to 55 nm and for the ba samples M h w ranged from 3.2 × 105 to 1.6 × 107 Da and 〈sj〉w from 14 to 54 nm. We can conclude that the enzymatic degradation of amylopectins has a somewhat larger effect on pap compared to ba. Although the pap HV has M h w and 〈sj〉w of the same size as of ba HV, the degraded fractions pap 5-20 and pap 15-20 have lower values than the corresponding fractions of ba. Debranching of the amylopectin and analysis of unit chains with HPAEC-PAD reveal information about the unit chain distribution. Figure 1 shows the distribution for pap HV and ba HV and the fractions isolated after different treatments with R-amylase and ethanol. It appears that pap HV, pap LV, and pap 5-20 has bimodal distribution consisting of short
Figure 1. Chain length distributions of pap HV, ba HV, pap LV, pap 5-20, ba 5-20, pap 15-20, ba 15-20, and ba 60-40.
chains (DP ∼ 14-15) and long chains (DP ∼ 40-50). By contrast, ba HV, ba 5-20, and ba 15-20 shows a trimodal chain distribution consisting of short chains (DP ∼ 10-12), medium chains (DP ∼ 19-21), and long chains (DP ∼ 4050). The more extensively amylase-hydrolyzed pap 15-20 and ba 60-40 exhibits polymodal chain distribution. Generally, for the fractions compared, the maximum in the distribution is somewhat displaced toward longer chain lengths for the pap compared to ba. However, the differences in chain length distribution were minor compared to the differences seen in M h w and 〈sj〉w between the fractions. Viscosity Properties. Figure 2 shows viscosity versus shear rate for the enzymatically degraded amylopectins mixed with 6% β-lg. The figure shows that the viscosity of all solutions behaves in a Newtonian manner. We can also observe that the result follow the expected pattern, which means, the higher the M h w the higher the viscosity, irrespective of type of amylopectin. The viscosity versus shear rate of pap HV and LV solutions has been studied previously.20 The results showed a shear-thinning behavior of 0.25% pap HV and a Newtonian behavior of the 2.0% mechanically degraded pap LV. The viscosity of both the amylopectins was around 0.014 Pas at a shear rate of 30 1/s. Thus considerably higher than the enzymatically degraded amylopectins. Rheological Behavior of Mixed Gels. Amylopectins from Different Sources. Figure 3 shows the shear modulus, G′, for gels containing 3.6% ba 5-20/6% β-lg and 3.6% pap 5-20/6% β-lg. The gel containing ba 5-20 with a higher
1404
Biomacromolecules, Vol. 4, No. 5, 2003
Olsson et al.
Figure 4. Shear modulus, G′, for mixed gels containing 3.6% ba 5-20/6.0% β-lactoglobulin and 3.6% ba 60-40/6.0% β-lactoglobulin. The horizontal line in the background indicates the G′ value of the pure 6% β-lg gel. Figure 2. Viscosity as a function of shear rate for 3.6% ba 5-20/ 6.0% β-lg (-9-) 3.6% pap 5-20/6.0% β-lg (-2-), 1.8% ba 5-20/ 6.0% β-lg (-4-), and 3.6% ba 60-40/6.0% β-lg (-(-).
Figure 5. Shear modulus, G′, versus M h w of amylopectin for gels containing 6% β-lg mixed with 3.6% ba 5-20 and 3.6% pap 5-20. Figure 3. Shear modulus, G′, for mixed gels containing 3.6% ba 5-20/6.0% β-lactoglobulin and 3.6% pap 5-20/6.0% β-lactoglobulin. The horizontal line in the background indicates the G′ value of the pure 6% β-lg gel.
Μ h w and thereby higher viscosity has a G′ of almost 6kPa, whereas the gel containing pap 5-20 with a lower M h w and thus a lower viscosity has a G′ of less than 5 kPa. The horizontal line in the background indicates the G′ value of the pure 6% β-lg gel. The G′ values of the mixed gels are more than double compared to those of pure β-lg gel. The contribution from amylopectin is likely to be negligible, because amylopectin only forms weak gels at low temperatures after a long storage time and at high concentrations. We therefore consider amylopectin to be nongelling and only influencing the formation of the aggregated β-lg network. For example, it has been reported that a 20% amylopectin gel formed at 2 °C after 60 days of storage had a G′ of around 13 mPa.31 Amylopectins with Different Enzymatic Treatment. Two different enzymatic treatments of ba were compared, in which ba 60-40 had a much lower M h w compared to ba 5-20, see Table 1. The consequences of the differences in M h w of ba in mixed gels with 6% β-lg are shown in Figure 4. The gel containing 3.6% ba 60-40/6% β-lg has a G′ of only 2.8 kPa, whereas the gel containing 3.6% ba 5-20/6% β-lg has a G′ of almost 6 kPa. Comparison of Mixed Gels Containing Low and High Molecular Weights of Amylopectin. Figure 5 shows G′ as a function of the amylopectin M h w for mixed gels containing
6% β-lg and 3.6% amylopectin. Filled circles show the effect of pap (fractions 5-20 and 15-20) and empty circles show the effect of ba (fractions 5-20, 15-20 and 60-40) on the gel strength. As expected from previous studies of gels containing pap LV and HV, an increase in G′ is observed with increase in the M h w of the amylopectins. In earlier studies the viscosity of the pap LV and HV was an indirect measure of differences in M h w, but now the results confirm that the M h w has a significant influence on the gel strength. The gels containing ba seem to increase more in G′ with increasing M h w compared to pap; that is, the curve for gels containing ba in Figure 5 has a somewhat steeper slope. A comparison implies that ba has a greater effect on the β-lg network at high M h w, whereas pap has a greater effect on the protein network at low M h w. There are further known differences in the fine structure between pap and ba with regard to chain length distribution and polydispersity.24 The chain length has a bearing on how a polysaccharide affects the compatibility with a protein in solution. Incompatibility is reduced in branched polysaccharides compared to linear polysaccharides according to Grindberg and Tolstoguzov.19 Increased polydispersity weakens the depletion effect according to an investigation of a colloidal dispersion containing small and large hard spheres.32 However, whether these properties have influenced the behavior of the mixed gels or not in the present study has not been elucidated. The polydispersity is the same for ba 5-20 as for pap 5-20, whereas ba 15-20 has a somewhat larger polydispersity than pap 15-20.24 Concerning chain length distribution, ba contains proportionately
Properties of Mixed β-Lactoglobulin Gels
Biomacromolecules, Vol. 4, No. 5, 2003 1405
Figure 6. LM micrographs, visualizing the protein phase, of mixed gels containing 6% β-lactoglobulin, (a) pure or mixed with (b) 3.6% ba 60-40, (c) 1.8% ba5-20, (d) 3.6% pap 5-20, (e) 3.6% ba 5-20, and (f) 2.0% pap LV.
more short chains (10 000×. The particles thereafter form particle aggregates with a diameter of around 0.5 µm, and these structures are possible to resolve by using the 100× objective in the LM. The particle aggregates in turn form larger clusters, which are connected in a network. These cluster networks are studied on an overall level of structure, where structures larger than 5 µm are resolved when using objectives in LM with a magnification
