Evaluation of bovine colostrum powder quality by quantitation of native IgG : why to choose the Single Radial ImmunoDiffusion (SRID) Assay ?
The importance of colostrum to the health and growth of newborn mammals is well known. In the bovine, the maternal immunity is quite exclusively transferred to the agammaglobulinemic offspring via the colostrum. Thus, bovine colostrum is mostly characterized by its very high level of antibodies, particularly of the IgG1 subclass deriving from the circulating IgG1 pool [1]. The concentration of specific antibodies in colostrum can be raised by immunizing cows with viral or bacterial antigens and a great number of clinical studies have shown the effectivness of hyperimmune colostrum-based preparations for prevention or treatment of human enteric disease [2]. In addition, bovine colostrum is a rich source of other antimicrobial agents (complement, lactoferrin, lysozyme), growth factors, cytokines, vitamins and trace elements all acting synergistically. Studies have shown that colostrum is the only natural source of two major growth factors, namely transforming growth factors alpha and beta and insulin growth factors 1 and 2. These growth factors have significant muscle and cartilage repair characteristics, they promote wound healing with practical implication for trauma and surgical patients, and they have multiple regenerative effects that extend to all structural body cells [3]. Feeding bovine colostrum has been shown to increase bone-free lean body mass in healthy trained adults, and to increase the synthesis of myofibrillar protein. Thus, colostrum may have positive effects on muscle function, performance capacity and healthy status of physically active people [4].
As a consequence of these biological activities, the sales of colostrum derived products for the medical, food, and nutraceutical markets are dramatically rising. To ensure its microbiological quality, bovine colostrum must be pasteurised or sterilised. When performed at about 72°C for 15 s the heat treatment has no significant effect on the bioactive molecules present in the colostrum. However, the spray-drying process mostly used by colostrum-based products manufacturers have a more or less detrimental effect on colostrum quality. The final product is frequently discoloured and poorly soluble due to protein aggregation. Thus there is a need to control the quality of the bioactive components in the colostrum derived powders. IgG quantitation has become the accepted method to quantify the colostrum-based powders quality since it is 1) the predominant antibody in colostrum and 2) a particularly heat sensitive protein (e.g. 85% loss of activity by spray-drying at 130 °C[5]).
The generally accepted reference method for IgG quantitation in bovine milk and colostrum is the Single Radial Immunodiffusion [6], however IgG quantitation has been performed with physicochemical techniques such as polyacrylamide electrophoresis [7-9], cellulose acetate electrophoresis [10] capillary electrophoresis [11, 12] and HPLC [13, 14, 15]. These last methods cannot discriminate between native and heat-denatured IgG. Moreover, overestimated results have been reported when such methods were applied to heated milk since peaks of proteins oligomers can be confused with peaks of monomeric forms of higher molecular weight proteins [16]. Thus immunochemical techniques using antibodies directed against conformational epitopes must be preferred. They mostly include enzyme-linked immunosorbent assay (ELISA), immunonephelometry, SRID and Plasmon Surface Resonance. However their applicability to heat treated samples must be thoroughly evaluated.
ELISA cannot be used with heat denatured proteins since their oligomerisation upon heating greatly overestimates their quantification [17]. Moreover, positive results in ELISA can be obtained with only one remaining native epitope while at least 3 native epitopes are required for precipitation in immunonephelometry or SRID. Surface Plasmon Resonance quantifies in real time the mass of protein captured by the antibody coated on the sensor surface. IgG oligomers produced upon heat treatment will greatly surestimate this quantification. Moreover, this technique requires a very expensive equipment which is not justified for routine IgG quantitation. Nephelometric immunoassays have been developed for bovine IgG quantification in milk [18,19] and similar results were obtained when bovine IgG were quantified in whey by kinetic immunonephelometry or SRID using the same antisera and standards [20]. Major drawbacks of immunonephelometry are the high antibody consumption and the need of very clear samples. This latter point requires labour intensive preparation such as centrifugation for defatting followed by sampling of the lower non-fat milk component by piercing a hole in the bottom of the centrifuge tube [21] or by 0.2 µ filtration [22]. Single Radial Immunodiffusion has proven to be useful for IgG quantification in milk products [23-29]. Due to the filtering effect of the agar, SRID allows direct analysis of whole colostrum or milk, as recommanded by Fleenor and Stott [23]. Thus the hard 0.45 µ filtration of colostrum powders solutions can be avoided. Moreover, highly aggregated IgG do not diffuse in the agar and oligomers interference in IgG quantitation is limited since they diffuse more slowly than the native monomeric IgG. However, unacceptable discrepancies in IgG concentrations and values exceeding the level of total whey protein have been reported by users of commercially available SRID plates [30, 31]. These discrepancies can be mostly explained by the use of antisera and/or standards obtained from unadequate IgG subclass [32, 33]. In the bovine, the two major subclasses IgG1 and IgG2 are approximately 90/10 in bovine colostrum or milk and 55/45 in serum [1, 34]. Commercially available bovine IgG are generally purified from bovine serum using salt fractionation or ion-exchange chromatography and are thus mostly of the IgG2 subclass. Antisera raised against these IgG partially crossreact with IgG1 and will give overestimated results in SRID due to larger, while less contrasted, precipitates [32]. Consequently, antisera and standards used for SRID quantification of IgG in colostrum or milk must be preferably obtained from purified colostral IgG1 [20, 24, 33, 35]. In such conditions the species in which the antibodies are produced is of secondary importance since similar results were obtained in the present work using goat or horse antibodies and since hen yolk has been reported as a feasible alternative source of antibodies for immunoassay of IgG in bovine milk [21, 33]. Lastly, discrepancies could result from incorrect fitting of the standard curve. As an example, Collin et al [21] have reported a better relationship between their SRID results and those of other assays when the SRID standard curve was constructed using a log-log regression rather than the linear regression recommended by the manufacturer of their SRID plates (ColtestTM, ICP Ltd).
The repeatability obtained with GTP Immuno IDRing ® RID kits and the IDRing® Meter software for the diameters measurement is routinely ≤ 4%, in accordance with previous reports [5, 35, 36]. When using less precise diameter measurement tools such as the McSwiney ruler, higher within-run CV% (6.3-9.5) have been reported [21]. Protein G Affinity Chromtography (PGAC) C has proven to be a very useful method for the isolation of mammalian IgG and fast on-line flow injection analyses have been developed for IgG monitoring in bioprocesses and applied to cell culture supernatants [37, 38]. In cattle, protein G binds strongly to IgG1 and IgG2 [39] while protein A of Staphylococcus aureus binds mainly to IgG2 [40]. Thus, protein G is particularly suited for IgG quantification in bovine dairy products since IgG1 is the predominant IgG subclass, and a PGAC method has been developed for IgG quantitation in bovine whey [41]. However, highly overestimated results when PGAC was compared to Single Radial Immunodiffusion for IgG quantitation in colostrum-based powders, have been observed. This could be partly explained by protein G binding of non-IgG compounds such as alpha 2-macroglobulin [42], cholesterol, phospholipids [43], and albumin [44] while this last binding can be avoided using recombinant protein G. Recently, it has been shown that a good correlation (r = 0.962) could be obtained between IgG quantitation by PGAC and Single Radial Immunodiffusion on colostrum powder solutions precipitated at pH 4.6 [45]. The authors suggested that the major factor contributing to the overestimation in PGAC is the presence of casein, although it was not possible to deduce whether the casein was bound directly to the column matrix or to the bound IgG. Most of the discrepancies between SRID and PGAC can be explained by protein G binding of light scattering heat-aggregated IgG which are overestimated when subsequently quantified by absorption at 280 nm. Tthese IgG aggregates are probably maintained by strong hydrophobic interactions and/or covalent links such as disulfide bridges since their size cannot be reduced by ultrasonic treatments in the presence of triton X100 (Levieux D, unpublished results) The protein G binding of highly polymeric IgG populations has been previously reported [46] and can be explained by the high protein G affinity for the IgG Fc region [47]. This region is less sensitive to heat denaturation than the Fab region involved in the antibody activity and it has been reported that Fc of heat-denatured and aggregated IgG maintains a native conformation since able to regulate mannose receptor expression by macrophages [48]. During repeated use of a PGAC column, a decline in the IgG peak and in the slope of the calibration curve has been reported [38]. This could be explained by loss of binding capacity of protein G due to its denaturation upon repeated elutions [38] or to progressive irreversible adsorption of IgG [49]. Using acidic elution (pH 2.0-2.5) the column could be theorically used hundreds of cycles [38]. However, the lifespan of the costly affinity protein G columns was greatly reduced when analyzing colostrum-based powders probably due to retention of protein aggregates and lipids which cannot be easily eliminated by cleaning procedures without deleterious effect on the binding capacity of the column [50]. To conclude, heat-aggregated IgGs in colostrum powder greatly overestimate native IgG quantitation in protein G analysis by an increase in optical absorbance at 280 nm due to light scattering. Founded on IgG concentration, the price of a colostrum-based powders would thus paradoxally increase with the percentage of IgG denaturation. The same applies to methods using an optical absorbance measurement such as capillary electrophoresis or chromatography must be also avoided. Among antibody based methods, ELISA and Surface Plasmon Resonance are highly sensitive to oligomerized or polymerized proteins and are thus also unsuitable. Consequently, Single radial Immunodiffusion using proper antisera and standards remains the best choice for the quantitation of native IgG in colostrum-based products. Moreover, the improvements proposed by GTP Immuno, such as manual or automatic diameters measurement on numeric pictures now fulfill the quality assurance criteria for traceability and repeatability required by pharmaceutical industry.
References
[4] A. Mero, J. Kakhkönen, T. Nykanen, T. Parviainen, I. Jokinen, T. Takala, T. Nikula, S. Rasi, J. Leppaluoto, J. Appl. Physiol. 93 (2002) 732-739.
[5] C.C. Chen, Y.Y. Tu, H.M. Chang, Food Sci. 26 (1999) 487-495.
[6] G. Mancini , A.O. Carbonara, J.F. Heremans, Immunochemistry 2 (1965) 235-254.
[7] H.E. Randolph, R.E. Erwin, R.L. Richter, J. Dairy Sci. 57 (1974) 15-23.
[8] K.F. Ng-Kwai-Hang, J.F. Hayes ; J.E. Moxley, H.G. Monardes, J. Dairy Sci. 70 (1987) 563-570.
[9] G.O. Regester, W.S. Geoffrey, J. Dairy Sci. 74 (1991) 796-802.
[10] A.R. Deshmukh, J.D. Donker, P.B. Addis, R. Jenness, J. Dairy Sci. 72 (1989) 12-17.
[11] U. Bilitewski Analytical Chem. 72 (2000) 692A-701A.
[12] L.E. Bennett, W.N. Charman, D.B. Williams, S.A. Charman, J. Pharm. Biomed. Anal. 12 (1994) 1103-1108.
[13] G.O. Regester, W.S. Geoffrey, J. Dairy Sci. 74 (1991) 796-802.
[14] S. Ostersen, J. Foldager, J.E. Hermansen J. Dairy Res. 64 (1997) 207-219.
[15] D.F. Elgar, C.S. Norris, J.S. Ayers, M. Pritchard, D.E. Otter, K.P. Palman, J.
[16] D. Levieux Ann. Rech. Vet. 11 (1980) 89-97.
[17] M. Rumbo, F.G. Chirdo, C.A. Fossati, M.C. Anon J. Agric. Food Chem. 44 (1996) 3793-3798.
[18] J.P. Lebreton, F. Joisel, S. Boutleux, B. Lannuzel, F. Sauger, Lait 61 (1981) 465-480
[19] R. Collin, C. Prosser, R. McLaren, M. Thomson, D. Malcom, J. Dairy Res. 69 (2002) 27-35
[20] D. Levieux, Lait 71 (1991) 327-328
[21] R. Collin, C. Prosser, R. McLaren, M. Thomson, D. Malcom, J. Dairy Res. 69 (2002) 27-35
[22] J.P. Lebreton, F. Joisel, S. Boutleux, B. Lannuzel, F. Sauger, Lait 61 (1981) 465-480
[23] W.A. Fleenor, G.H. Stott, J. Dairy Sci. 64 (1981) 740-747
[24] D. Levieux, Ann. Rech. Vet. 5 (1974) 329-342.
[25] J.L. Burton, B.W. Kennedy, E.B. Burnside, B.N. Wilkie, J.H. Burton, J. Dairy Sci. 72 (1989) 135-149.
[26] P.J. Koning, J. Jong, J. Kaper, Voedingsmiddelentechnologie 22 (1989) 24-26.
[27] L.C. Pritchett, C.C. Gay, T.E. Besser, D.D. Handcock, J. Dairy Sci. 74 (1991) 2336-2341
[28] J.D. Quigkley, K.R. Martin, L.B. Dowlen, L.B. Wallis, J. Kamar, J. Dairy Sci., 77 (1994) 264-269.
[29] G. Mainer, M.D. Perez, L. Sanchez, P. Puyol, M.A. Millan, J.M. Ena, E. Dominguez, M. Calvo, Milchwissenschaft 55 (2000) 613-617.
[30] NorthField Laboratories, http://www.intact.com.au/case_example.htm (2006).
[31] Institute of Colostrum Research, http://www.colostrumresearch.org /About_ICR /qualityScale_full.html#Immunoglobulins (2006)
[32] B.A. Bockkout, J. Immunol Methods 7 (1975) 187-210.
[33] E.C. Li-Chan, A. Kummer, J. Dairy Sci. 80 (1997) 1038-1046.
[34] N.L. Norcross, J. Amer. Vet. Med. Assoc. 181 (1982) 1057-1060.
[35] H. Fey, J. Pfister, J. Messerli, N. Sturzenegger, F. Grolimund, Zbl. Vet. Med B, 23 (1976) 269-300.
[36] D. Levieux, A. Ollier, J. Dairy Res. 66 (1999) 421-430.
[37] G.S. Blank, Vetterlein D. Analytical Biochem.190 (1990) 317-320.
[38] M. Reinecke, T. Scheper, J. Biotechnol 59 (1997) 145-153.
[39] J.J. Langone, Adv. Immunol. 32 (1982) 157-252.
[40] J. Goudswaard, J.A. Van Der Donk, A. Noordz, R.H. Van Dan, J.P. Vaerman Sand. J. Immunol. 8 (1978) 21-28.
[41] N.M. Kinghorn, C.S. Norris, G.F. Paterson, D.E. Otter, J. Chromatogr. 700 (1995) 111-123.
[42] U. Sjöbring, J. Trojnar, A. Grubb, B. Ackerström, L. Björk, J. Immunol. 143 (1989) 2948-2954.
[43] E. Kitsiouli, M.E. Lekka, G. Nakos, C. Cassagne, L. Maneta-Peyret, J. Immunol Methods 271 (2002) 107-110.
[44] U. Sjöbring, L. Björk, W. Kastern, J. Biol. Chem. 266 (1991) 399-405.
[45] D.E.J. Copestake, H.E. Indyk, D.E. Otter, J. AOAC Int.89 (2006) 1249-1256. Chromatogr. A278 (2000) 183-196.
[46] Y. Kanamura, S. Nagaoka, Y. Kuzuya, Anim. Sci. Technol. 63 (1992) 383-393.
[47] A.E. Sauer-Eriksson, G.T. Kleywegt, M. Uhlen, T.A. Jones, Structure 3 (1995) 265-278.
[48] S. Schreiber, W.F. Stenson, R.P. MacDermott, J.C. Chappel, S.L. Teitelbaum, S.L. Perkin, J. Immunol. 147 (1991) 1377-1382.
[49] Z. Yan, J. Huang, J. Chromatogr. 738 (2000) 149-154.
[50] Z. Yan, J. Huang, J. Immunol. Methods 237 (2000) 203-205.