CXIX. NOTE ON THE FORMALDEHYDE TITRATION OF MILK-PROTEIN, AND ITS APPLICATION TO THE ESTIMATION OF CASEINOGEN.

BY GERALD THOMAS PYNE. From the Department of Dairy Chemistry, University College, Cork.

(Received May 8th, 1933.)

THE writer described in this Journal [1932] an improved technique for the determination of milk-proteins by formaldehyde titration. Very satisfactory results were obtained for mixed milks using a factor, 1 cc. N NaOH = 1-74 g. milk-protein, which however for milks of individual cows showed a somewhat wide variation, ranging from 1-63 to 1F87. It was tentatively suggested that this might arise from variations in the relative amounts of caseinogen and albumin from sample to sample and be susceptible perhaps of easy correction; and the further work now described has been carried out primarily to test this point. For this purpose the formaldehyde value, and protein content (N x 6.38) of a number of individual milks and of the corresponding caseinogen-free acid wheys were determined. Most of the samples were taken during two main periods-March to May 1932, and October to November, 1932-the entire herd of some 30 animals (Dairy Shorthorn breed) being tested on each occasion so as to obtain values representative of different stages of lactation. The aldehyde values of the milks were obtained by the method referred to above and those of the acid whey somewhat similarly. The latter was prepared by a modification of the isoelectric precipitation method for caseinogen, recommended by Moir [1931], the principal changes being that the milk was not diluted with more than about 80 % of its volume of water before adding the acetic acid and sodium acetate solutions, and that these were employed in somewhat higher concentrations, viz. 2-5 N and M respectively. In this way too great dilution was avoided, the final volume before filtration not being greater than twice that of the original milk. For the determination of its aldehyde value 20 cc. of the filtrate (corresponding approximately to 10 cc. milk) was roughly neutralised with 2N NaOH, 0-4 cc. saturated aqueous potassium oxalate and 1F0 cc. 0-5 % phenolphthalein were added, and the liquid was accurately neutralised with N/10 NaOH, the end-point being a faint pink which was matched with that of a similar quantity of whey and oxalate tinted with magenta. 2 cc. 40 % formaldehyde were then run in and the titration continued to the same tint as before. Allowance must of course be made for the acidity of the formaldehyde under similar conditions of dilution, etc. The results are summarised mainly as group averages in Table I. The numbers in the column headed "g. caseinogen calc." were obtained by dividing the difference in the protein contents of the milks and their corresponding quantities of acid whey, by the difference in their respective aldehyde titrations, and are seen to display a marked uniformity over the whole period, the extreme variation in even individual samples being largely attributable to
Biochem. 1933 xxKvii

58

916

G. T. PYNE
Table I.
1 cc. N NaOH=

Months after calving 1 2 3 4 5, 6 7 8 9 10-13 Mean

No. of samples 13 6 8 8 3 3
5

g. milkprotein

g. whey-

g.
caseinogen calculated 1-85 1-88 1-88 1-88 1-88 1-84 1-87

observed 1-69 1-73 1-75

1-75 1-76 1-76 1-77 1-78 1-80 1-74

3
5

(54)

protein observed 1-33 1-36 1-41 1-48 1-49 1-54 1-52 1-58 1-59 1-44
1-15-1-74

Caseinogen-N
Other N 3-3 3-3 3-2
2-8 2-9 3-1 2-7

Ratio:

1-85
1-89

3-1

2-8

1X87 1-77-1-97 5-3 %

3-4 %

Extreme values for individual samples Extreme variation from mean as % of mean Suggested % variation from mean attributable to experimental error

1-63-1-88

8-0 %
1-5 %

20-8 %

5-0 %

experimental error. Omission of the correction for the volume of precipitated protein and fat does not appreciably affect the value in this column. The last line in Table I is based on the assumption that the titration errors in an average aldehyde value determination for total protein, whey-protein, and caseinogen (say: 2-0, 0-6, and 1-4 cc. respectively), do not exceed about 0-03, 0-03, and 0-05 cc. respectively (the allowance for the last is larger as the difference of two titrations is involved), and is included merely as a guide to the degree of constancy to be expected from, or shown by, the various factors. From the data in Table I it may be seen that: (a) The total protein value of 1 cc. N NaOH gradually increases throughout the lactation, and this increase does not appear to depend to any marked extent on variations in the ratio of caseinogen to total protein', but arises from the more strongly marked increase in the whey-protein value of the alkali during this period. (b) The "caseinogen value" of 1 cc. N NaOH (mean 1-87 g.) is remarkably constant, making it evident that it is possible to estimate caseinogen by formaldehyde titration to at least the same degree of accuracy as total protein. If a 10 cc. pipette (delivering 10-25 g. milk), and N/10 alkali are used for the titration the formula becomes % caseinogen = f--x (difference in formaldehyde values of 10 cc. milk and corresponding quantity of acid whey). It seemed of interest to push the inquiry into variations in the formaldehyde factor of milk slightly further, and to look for the causes of the marked changes found in the whey-protein factor. The aldehyde titration of acid whey was already somewhat small for any detailed fractionation, and all that was attempted
1 For about 95 % of the milk samples the ratio Caseinogen- lay between 2-5 and 3-8. The 25ad38 h Other N variation in the mean factor, 1 cc. N NaOH = 1-74 g. protein, arising from this cause would not exceed 0-02, and is thus small compared with that found in practice for individual samples (see Table 1).

DETERMINATION OF MILK-PROTEINS

917

therefore was the removal of albumin by the usual heat-coagulation method and the determination of the nitrogen content and formaldehyde value of the filtrate. In six milk samples examined in this way, and chosen to represent various stages of lactation, the protein value of 1 cc. N NaOH was found to vary from 1-01 to 1-50 for the albumin-free filtrates, as against 1-22 to 1-47 for the corresponding acid wheys. While no great accuracy is claimed for these figures, on account of the smallness of the titrations on which they are based, they suggest that much of the variation in the formaldehyde factors of individual milks may be traced to variations in the composition of their non-protein-nitrogen fractions.

SUMMARY. 1. The factor for converting formaldehyde titration into milk-protein shows a progressive increase throughout the lactation period, which appears to arise immediately from a similar but more marked increase in the whey-protein factor and to be ultimately attributable largely to variations in the composition of the non-protein-nitrogen. 2. The formaldehyde titration method can be adapted to the estimation of caseinogen.
REFERENCES.

Moir (1931). Analyst, 56, 147. Pyne (1932). Biochem. J. 26, 1006.

58-2