Data concerning the fractionation of individual whey proteins and casein micelles by microfiltration with ceramic gradient membranes

Data are related to the research article “Fractionation of casein micelles and immunoglobulins by microfiltration in diafiltration mode Study of the transmission and yield of IgG, IgA and IgM” [1]. The data show the transmission and yield of the individual whey proteins α-Lactalbumin (α-La), β -Lactoglobulin (β -Lg), blood serum albumin (BSA), lactoferrin (LF), lactoperoxidase (LPO) and the immunoglobulins IgG, IgA, IgM during microfiltration (0.14 μm) performed in diafiltration mode at 50 °C with different applied transmembrane pressures (0.6-3 bar). The data provide information on the decrease of the respective proteins in the microfiltration retentate and their increase in the UF retentate. The relevant analytical methods for the individual protein detection were performed by reversed phase high performance liquid chromatography and ELISA. The isoelectric point of IgG and IgM was measured with the Zetasizer Nano ZS.


Data
The dataset contains information on the transmission and yield of individual whey proteins during milk protein fractionation by microfiltration. Furthermore, data on the comparison of analytical methods for the determination of bovine IgG and b-Lg are as well as the influence of temperature during filtration on the respective whey proteins are presented.
Specifications Table   Subject

Experimental design, materials, and methods
The preparation of the samples, the equipment and analytical methods for analysis are described in detail by Heidebrecht and Kulozik (2019) [1]. The only method not described was the measurement of the zeta potential, however this methods has been described by Dombrowski et al. (2016) [2].

Comparison of IgG measured with ELISA and RP-HPLC
IgG of different filtration samples were measured with reversed phase high performance liquid chromatography (RP-HPLC) and enzyme-linked immunosorbent assay (ELISA). The binding mechanism to the RP column is based on hydrophobic interactions, while ELISA is used to measure the binding capacity to an antigen.  For the evaluation of the analytical equivalence, the correlation coefficient according to Lin (1989) was calculated in addition to the stability index according to Pearson [3].
The concordance correlation coefficient was 0.962 with a confident interval of 95% of 0.946e0.974. A coefficient between 0.95 and 0.99 is considered to be an essential correspondence of a set of pairs from two measurements [4]. It should be noted that this correlation may not be valid at high denaturation levels. Native Ig is soluble at pH 4.6 and begins to precipitate at pH below pH 3.5 [5]. Proteins which are not in their native state are precipitated by the adjustment to pH 4.6 and thus not recognized by the RP-HPLC measurement. However, they may have a certain binding capacity and be detected by the ELISA measurement.

Analytical equivalence of two independent RP-HPLC methods for b-Lg quantification
The advantage of the method 1 according to Dumpler et al. (2017) [6] is, that caseins and whey proteins elute one after the other and can therefore be measured in one run. However, IgG cannot be measured with this method, which was the primary goal of the related study. Since caseins and whey proteins do not lie on top of each other, it is possible to determine the total b-Lg content as well as the soluble content at pH 4.6 (also referred to as native b-Lg) and thus to determine the degree of denaturation.
When using the PLRP-S column of method 2 according to Ref. [7], caseins and whey proteins are partly over each other, so that caseins must be precipitated by adjusting the pH to 4.6 during sample preparation. Thus only the native b -Lg content can be measured with this method. The native b-Lg concentration measured with both methods is shown in Fig. 2. The correlation coefficient according to Lin and was 0.995 (confident interval 95% 0.9931e0.9964).  (2) C A is sum of the respective soluble and insoluble whey protein concentration and C N is only the soluble concentration.
It should be noted that the time scale for Figs. 10 and 11 were not normalized to 1 m 2 membrane area, as it was the case with Figs. 6 and 8. The DD of b-Lg increased exponentially from 0 to 60% in the    Fig. 11 increases disproportionally. After about 4 DF-steps, the curves deviate from each other, which is better illustrated by the percentage difference in Fig. 11. As the concentration of native proteins decreases, the ratio of native to aggregated proteins in the MF-retentate shifts in the direction of the aggregated (see Fig. 10). However, the 60% denaturation at the end of the DF process is misleading in terms of the nativity of the proteins. Therefore, we also calculated the absolute degree of denaturation (ADD) (Fig. 10), which has the same validity as the DD during batch filtration. The ADD expresses the total amount of denatured protein in the MF-retentate at a given time relative to the total amount of native protein in the milk before filtration. The ADD is defined by Eq. (3) where V t is the volume at a specific time, V 0 is the initial volume of skimmed milk, C A is the absolute concentration of b-Lg and C N is the native concentration of b-Lg.  2)) and absolute degree of denaturation (Eq. (3)) in the MF-retentate as well as degree of denaturation in the UF feed tank. Fig. 11. Progress of native and total b-Lg (native plus denatured) as well as the percentage difference of the two curves (right y-axis) as function of the filtration time.