Original papers
Microstructural Observation of Dairy Products Using High-pressure Freezing in Combination with Cryo-scanning Electron Microscopy
2020 年 26 巻 6 号 p. 743-747
詳細
2020 年 26 巻 6 号 p. 743-747
Observation of bovine raw milk and cream was attempted by combining high-pressure freezing with cryo-scanning electron microscopy (SEM). In raw milk, casein micelles were observed to be distributed in the aqueous phase, with an average diameter of about 140 nm, which was in agreement with previous reports. In cream, the number of casein micelles adsorbed on the fat globule interface was observed to be larger compared with that observed under freeze replication with a cryoprotectant. Results showed that high-pressure freezing combined with cryo-SEM is a useful method for observing the native structures of raw milk and cream in their original states, which was difficult to achieve using conventional electron microscopy.
Dairy products are manufactured from bovine raw milk using manufacturing processes such as heating and homogenization, or lactic acid bacteria and enzymes. Therefore, product characteristics vary greatly depending on the processing conditions. Methods used to evaluate the characteristics of these products include sensory evaluation of taste by panelists, measurement of physical properties, such as hardness and stickiness, using a texture analyzer, and analysis of components, such as moisture and nutrients. Microstructural observation of the milk protein and fat contained in dairy products is also considered to be a useful evaluation method for determining milk composition and elucidating the effect of microstructure on texture and physical properties.
However, many dairy products have high moisture content, with milk fat containing saturated fatty acids at about 65% of total fatty acids (Tsuchiya, 1970). Chemical fixation is used as a pretreatment for general electron microscopy, whereby the sample is immersed in a fixing solution, thus making liquid samples such as milk difficult to process. Furthermore, osmic acid aqueous solution, which is the only fixative used to fix fat, cannot fix saturated fatty acids, making the fixation of milk fat difficult (Lee and Morr, 1992). Rapid freezing, whereby samples are immersed in a cryogen, such as liquid nitrogen, has generally been used for physical fixation. However, ice crystals that form during freezing disrupt the microstructure, and a higher water content results in larger ice crystals (Miyawaki, 2001). As an alternative method, cryoprotectants, such as glycerin, DMSO, and ethanol, are used in combination; however, changes caused by diluting or stirring the sample cannot be eliminated.
Since bovine milk is liquid, it is particularly difficult to fix using a fixative solution, and ice crystals are markedly generated during rapid freezing owing to its water content of about 88%. In previous studies, skim milk has been naturally dried and observed by the shadowing method (Kimura et al., 1979), deposited on carbon planchets, chemically fixed, dried at the critical point, and observed by scanning electron microscopy (SEM) (Dalgleish et al., 2004), observed by transmission electron microscopy (TEM) after dropping on a grid and freeze-drying (McMahon and Oommen, 2008), and observed by freeze replication after mixing with glycerin (Karlsson et al., 2007). In many cases, a drying process or cryoprotectant has been used, and observation of the original milk state has proven difficult. Cream is a liquid dairy product with a high milk fat content (18–50 g/100 g), making chemical fixation using osmic acid difficult. Even when using rapid freezing, the formation of ice crystals owing to the water content of the cream (44–56 g/100 g) is a major problem, disrupting the continuous water phase microstructure. In a previous study, cream was mixed with agar to form a gel, chemically fixed, and observed by SEM (Kalab, 1985). Also, cream has been freeze-fixed after mixing with cryoprotectants, such as ethanol or glycerin (Kimura, 2000). However, changes caused by diluting or stirring the sample are often an issue when mixing a gelling agent and cryoprotectant.
In this context, the present study focuses on high-pressure freezing, in which the sample is frozen while pressurizing at 210 MPa, to obtain a much deeper glassy freezing compared with conventional rapid freezing (Kamigaki et al., 2018; Jensen, 2013; Knomi et al., 2000). Furthermore, this technique was combined with cryo-SEM to observe the microstructure of raw milk and cream samples with fewer ice crystals. For comparison with conventional methods, raw milk was also observed by cryo-SEM using rapid freezing, while cream was also observed by TEM using freeze replication. The usefulness of the present method combining high-pressure freezing with cryo-SEM was evaluated.
Materials Bovine raw milk was obtained from the Cheese Research Laboratory (MEGMILK SNOW BRAND Co., Ltd., Tokyo, Japan). Before observation, the milk was stored at 5 °C for 1 day and then incubated at 30 °C for 1 h. Cream with a fat content of 47% was obtained by centrifuging raw milk at 47 °C, followed by heat sterilization at 125 °C for 3 s, cooling to 5 °C in ice water, and then storing at 5 °C for 2 days.
Experimental procedure The specimen carriers (dome-shaped Cu-Au, 16770132, Leica Microsystems, Vienna, Austria) were filled with 0.2 µL each of raw milk or cream samples, using a pipette to prevent air bubbles from entering, and then frozen using a high-pressure freezing device (EM HPM100, Leica Microsystems, Vienna, Austria). The user cannot alter the device parameters, and the samples were frozen under the conditions (pressure 210 MPa, freezing rate −12 000 K/s) set by the manufacturer. Cryofixed samples were transferred to the cryo-stage (pp3010T, Quorum, Laughton, UK) and fractured at −130 °C. Sublimation was performed at −70 °C for 5 min, and platinum coating was performed at −130 °C. The treated samples were inserted into the cryo-SEM (S4300, Hitachi High-Technologies, Tokyo, Japan) and observed at −130 °C. The images of raw milk obtained by combining high-pressure freezing with cryo-SEM were analyzed using image analysis software (Azokun, Asahi Kasei Engineering Corp., Kawasaki, Japan). These images were binarized, all casein micelles were assumed to be spherical, and the average size of casein micelles was calculated.
The present observation method combining high-pressure freezing with cryo-SEM was compared with conventional methods. Raw milk was observed using cryo-SEM after immersion in slush nitrogen. Cream was observed using freeze replication after mixing with an equal amount of cryoprotectant (30% glycerin solution), immersing in slush nitrogen, and coating with platinum carbon after fracturing at −130 °C (JFD2, JEOL, Tokyo, Japan). The obtained replica film was observed using a transmission electron microscope (JEM-2000FXII, JEOL, Tokyo, Japan).
Figure 1a and b show cryo-SEM images taken after rapid freezing in slush nitrogen. Fat globules were confirmed in the image, and the aqueous phase had a honeycomb structure appearance. Therefore, the continuous aqueous phase was considered to be completely destroyed by the formation of ice crystals, and the presence of casein micelles could not be confirmed. Figure 2a and b show low- and high-magnification images of raw milk observed by combining high-pressure freezing with cryo-SEM. Almost no structural damage due to ice crystals, i.e., the honeycomb structures, was observed in the image of raw milk when using high-pressure freezing. Some large fat globules (2–4 µm in diameter) and spherical structures thought to be casein micelles (40–600 nm in diameter) were observed in the aqueous phase. Table 1 shows the average diameter of casein micelles calculated by analyzing the images obtained by high-pressure freezing and cryo-SEM. The average diameter of casein micelles in raw milk was 140 nm.
(a,b) Microstructures of raw milk observed by cryo-scanning electron microscopy combined with rapid freezing using slush nitrogen.
(a,b) Microstructures of raw milk observed by combining high-pressure freezing and cryo-scanning electron microscopy.
| Average diameter (nm) |
Standard deviation (nm) |
n | |
|---|---|---|---|
| Raw milk | 140 | 6 | 429 |
Figure 3a and b show low- and high-magnification images of cream observed using cryo-SEM after high-pressure freezing. Figure 3b shows an enlarged image of the outlined area of Figure 3a. These images show many spherical structures of about 100–200 nm in size, which appear to be casein micelles adsorbed on the interface of fat globules. Figure 4a and b show low- and high-magnification images of cream observed by freeze replication. In these images, casein micelles were observed as aggregates of multiple particles, with almost no casein micelles confirmed to be adsorbed on the fat globule interface in Figure 4a. In Figure 4b, only one or two casein micelles were confirmed to be adsorbed on the fat globule interface.
(a,b) Low- and high-magnification images of cream observed by combining high pressure freezing with cryo-scanning electron microscopy.
Arrow, casein micelles adsorbed at the interface of a fat globule.
(a,b) Low- and high-magnification images of cream observed by freeze replication.
Enclosed portion, aggregate of multiple particles appearing to be casein micelles.
High-pressure freezing at 210 MPa delays the formation of ice crystals during freezing, resulting in amorphous ice at the surface region of the sample at a depth of 200–600 µm (Kamigaki et al., 2018; Jensen, 2013). Although the influence of high-pressure freezing on the sample is not well understood, the size of casein micelles quantified in this study was in agreement with previously reported values of about 140 nm from cryo-electron microscopy of vitreous sections (CEMOVIS) (Kamigaki et al., 2018), and 100–150 nm obtained using a replica method (Schmidt, 1982; Kimura and Taneya, 1989) and light-scattering method (Ono, 2005). Gebhardt et al. (2006) reported that casein micelles disintegrate into small fragments after pressure treatment at 50–250 MPa. Knudsen et al. (2010) reported that some large micelles and many small micelles with diameters of 20–50 nm coexist under pressure treatment at 150–200 MPa. Considering these previous reports of high-pressure treatment altering the size of casein micelles, the present study indicates that the adverse effects of high-pressure treatment were small, as the size of casein micelles was not reduced.
Cream is an oil-in-water emulsion, in which milk proteins at the interface between the fat globules and aqueous phase are considered to greatly influence emulsion stability (Relkin and Sourdet, 2005; Goff, 1997; Sourdet et al., 2003). Although many reports have stated that casein micelles or casein exist at the interface of fat globules (Darling and Butcher, 1978; Luo et al., 2014; Cano-Ruiz and Richter, 1997), this was visualized using high-pressure freezing in this study. After high-pressure freezing, cream was observed to be frozen without dilution, while freeze replication involved mixing the cream with an equal amount of glycerin. As the surface concentration of protein adsorbed on the fat globule interface and shear stability depend on the protein concentration in the aqueous phase (Segall and Goff, 2002), their dilution or agitation might have caused the number of casein micelles adsorbed on the fat globule interface to decrease.
Therefore, high-pressure freezing was considered to be a useful fixation method for electron microscopy, capable of reducing ice crystal formation without cryoprotectants and maintaining the original microstructure of the sample. High-pressure freezing can be applied to observation not only using cryo-SEM, but also cryo-TEM using ultrathin frozen sections (CEMOVIS) or TEM using freeze substitution. For samples that are difficult to observe using conventional fixation methods, high-pressure freezing is a useful pretreatment that preserves the original microstructure of a sample without dilution by cryoprotectants or immersion in a fixative. The microstructures of food with a high moisture or fat content might be more accurately observed using this method.
The observation of bovine raw milk and cream was attempted using a high-pressure freezing and cryo-SEM method. For raw milk, almost no structural destruction due to ice crystals was observed, with casein micelles observed in the aqueous phase and their sizes quantified. For cream, structures that appeared to be casein micelles adsorbed on the fat globule interface were observed. As the number of casein micelles was larger than that observed in samples obtained by freeze replication with a cryoprotectant, high-pressure freezing was considered to preserve the native structure of the sample. From these results, the observation method combining high-pressure freezing with cryo-SEM allowed observation of the native structure of raw milk and cream, which was difficult to achieve using conventional chemical fixation or rapid freezing through immersion in slush nitrogen.
