The specific activity of the crystalline phosphoserine so obtained was not altered significantly by three recrystallizations from aqueous ethanol. The melting point of the product was 168-170° with decomposition.

Other Phosphorylated Amino Acids-In experiments similar to that described for phosphoserine, synthetic unlabeled phospho-l-hydroxyproline, phospho-l-tyrosine, and phospho-l-threonine were added to portions of crude hydrolysate of "phosphoprotein" and recovered by chromatographic methods. In contrast to the results with phospho-l-serine, none of these phosphorylated amino acids showed significant amounts (more than 1 per cent) of the radioactivity of the hydrolysate after persistent rechromatog-raphy. These experiments do not prove, however, that the phosphorylated amino acids tested are not present in the hydrolysate in non-radioactive form, since it is only by their radioactivity that they could have been detected.

Although no significant amounts of the other radioactive phosphorylated amino acids could be found, only about 30 per cent of the activity of the crude hydrolysate could be accounted for as phosphoserine. Inspection of the chromatogram in Fig. 1 reveals several minor peaks eluted more rapidly than phosphoserine. The material in one of these peaks (Tubes 53 to 57 of the chromatogram shown) was concentrated and chromatographed on filter paper. The radioactivity was found to be associated with a rather diffuse ninhydrin spot with an Rr value of 0.40. The specific activity of this radioactive ninhydrin-reacting substance was fully as high as that of phosphoserine. No material of high specific activity was found in any experiment except in ninhydrin-positive fractions.

Since the conditions of hydrolysis of the "phosphoprotein" were deliberately chosen to be rather mild so that total breakdown of phosphate esters could be avoided, it is fairly certain that the amounts of free phosphoserine found are only a fraction of the total phosphorylated serine present. It seems reasonable to assume that the faster running material represents peptides of phosphoserine. No further effort to identify this material has been made.

"Phosphoprotein" of Mitochondria of Normal Liver Tissue-A few experiments were carried out with the "phosphoprotein" fraction of normal rat liver mitochondria, made radioactive by incubation in vitro with inorganic phosphate labeled with P32 (5, 6). Results essentially similar to those obtained with the Ehrlich ascites tumor "phosphoprotein" were obtained, although the radioactivity of the mitochondrial "phosphoprotein" was considerably less than that derived from the tumor tissue. As in the case of the tumor "phosphoprotein," the principal radioactive component of the mitochondria "phosphoprotein" appeared to have chromatographic properties identical with that of phosphoserine.

Discussion

The evidence presented in this paper demonstrates that the phosphate moiety of phosphoserine in the phosphoprotein of the Ehrlich ascites tumor undergoes a truly remarkable rate of turnover. Results presented in Table I show that the specific activity of the "phosphoprotein" fraction obtained by the Schneider technique is about 30 times that of the nucleic acid fraction. The specific activity of the purified isolated phosphoserine is about 4 times that of the crude "phosphoprotein" fraction, or about 120 times that of the nucleic acid fraction.

What is the physiological significance of this high rate of renewal? Is it a reflection of a very rapid rate of turnover of the whole protein molecule, or is it the result of rapid dephosphorylation and rephosphorylation of the otherwise intact protein? These questions are of obvious importance in protein metabolism and function, but the answers await further fundamental study.

It is by no means certain that the high metabolic activity indicated by the rapid rate of turnover of the phosphoserine linkage is a characteristic of all types of "phosphoprotein" molecules in the Ehrlich ascites tumor cells. The possibility must be considered that this rate of turnover is a result of enzymatic activity on the part of enzymes which function by cyclic phosphorylation and dephosphorylation of the serine hydroxyl group, perhaps as a stage in the utilization of the phosphate bond energy of adeno-sinetriphosphate. Mechanisms of enzyme action involving intermediary phosphorylation of the enzyme itself have been put forward by several authors. Recently, Lipmann and his collaborators (19) have presented evidence for the formation of an enzyme-phosphate-adenosine compound which may be formed during the course of formation of acetyl coenzyme A from adenosinetriphosphate and acetate. It is also of considerable interest in this connection that Schaffer et al. (20) have isolated phosphoserine from hydrolysates of chymotrypsin after treatment with diisopropyl fluorophosphate, suggesting that a serine residue is the active center of this enzyme.

The isolation of radioactive phosphoserine from the acid hydrolysates of the "phosphoprotein" fraction is, of course, no conclusive evidence that the phosphate of the intact protein is present as a simple monoester of serine. If, for example, the actual structural linkage is that of a pyrophosphate or phosphorus diester bridge, it is still possible that acid hydrolysis would yield phosphoserine. The nature of phosphorus linkages in "phosphoproteins" has been discussed by Perlmann (21) who has also presented evidence indicating the possible existence of nitrogen-phosphorus linkages in "phosphoproteins."

We are indebted to Mr. William Bradford for growth and transfer of the Ehrlich ascites tumor in mice, and to Dr. H. G. Williams-Ashman for valuable advice and for performing the experiment shown in Table I.

Summary

After incubation of washed, intact Ehrlich ascites tumor cells in the presence of inorganic phosphate labeled with P8, the specific activity of the "phosphoprotein" residue has been found to be many times higher than that of any other trichloroacetic acid-insoluble fraction. Upon partial acid hydrolysis of the crude "phosphoprotein" phosphoserine of very high specific activity could be isolated by chromatographic techniques. The possible significance of the high metabolic activity of the phosphoserine residues of "phosphoproteins" is discussed.

Preliminary experiments with the "phosphoprotein" of mitochondria from normal liver tissue yielded essentially similar results.

Bibliography

1. Schneider, W. C, J. Biol. Chem., 164, 241 (1946).

2. Araki, M., Yonezawa, T., Chin, S., Kuga, M., Shimada, N., Imagi, S., and Ochiai, Y., Oann, 43, 69 (1952).

3. Davidson, J. N., Frazer, S. C, and Hutchison, W. C, Biochem. J., 49, 311 (1951).

4. Johnson, R. M., and Albert, S., J. Biol. Chem., 200, 335 (1953).

5. Friedkin, M., and Lehninger, A. L., J. Biol. Chem., 177, 775 (1949).

6. Williams-Ashman, H. G., and Kennedy, E. P., Cancer Res., 12, 415 (1952).

7. Jeener, R., Bull. Soc. chim. biol., 31, 731 (1949).

8. von Euler, H., Hevesy, G., and Solodkowska, S., Ark. Kemi, Mineral, o. Qeol.,26 a, 1 (1949).

9. 8chmidt, G., and Thannhauser, S. J., J. Biol. Chem., 161, 83 (1945).

10. Lipmann, F., and Levene, P. A., J. Biol. Chem., 98, 109 (1932).

11. Lipmann P., Biochem. Z., 262, 3 (1933).

12. de Verdier, C.-H., Acta chem. Scand., 7, 196 (1953).

13. Kun, E., Talalay, P., and Williams-Ashman, H. G., Cancer Res., 11, 855 (1951).

14. Kennedy, E. P., J. Biol. Chem., 201, 399 (1953).

15. Plimmer, R. H. A., Biochem. J., 36, 462 (1941).

16. Moore, S., and Stein, W. H., J. Biol. Chem., 176, 367 (1948).

17. Gomori, G., J. Lab. and Clin. Med., 27, 955 (1942).

18. Busch, H., Hurlbert, R. B., and Potter, V. R., /. Biol. Chem., 196, 717 (1952).

19. Jones, M. E., Lipmann, F., Hilz, H., and Lynen, F., J. Am. Chem. Soc., 76, 3285

(1953).

20. Schaffer, N. K., May, S. C, Jr., and Summerson, W. H., J. Biol. Chem., 202, 67

(1953).

21. Perlmann, G., in McElroy, W. D., and Glass, B., Phosphorus metabolism, Baltimore, 2 (1952).