CENTRIFUGATION OF 

 PARTICULATES 



100 



CENTRIFUGATION OF 

 PARTICULATES 



fraction. With nuclei, at least, these ob- 

 stacles are in part overcome by the M. 

 Behrens technique (Ztschr. f. physiol. 

 Chem., 1932, 209, 59) by which tissue 

 is frozen and dehydrated at the outset, 

 fractionation being accomplished in non- 

 polar organic media. 



The factor of tonicity is poorly under- 

 stood. A solution isotonic to eryth- 

 rocytes may not be so to mitochondria 

 or nuclei. Mitochondria eventually 

 swell and burst in distilled water, but 

 appear well preserved in almost syrupy 

 sucrose solutions. The review of W. 

 C. Schneider and G. H. Hogeboom 

 (Cancer Res., 1951, 11, 1-22) should be 

 consulted in this respect. Nuclei, on 

 the other hand, do not appear to behave 

 as osmometers. 



Hydrogen ion concentration, of 

 course, is an important variable and 

 should be adjusted according to experi- 

 mental aims. In using citric acid 

 media, for example, we find a number of 



EH dependent effects. At pH 4 or 

 elow (2% citric acid) nuclei are readily 

 isolated in bulk, morphologically free 

 of cytoplasm, with no tendency to 

 clumping, and with an apparently com- 

 plete complement of desoxyribonucleic 

 acid (DNA) . At pH 4-6, agglutination 

 of cytoplasmic particles produces nu- 

 clear clumping. At pH 6, nuclei are 

 again dispersed, and contain, in addi- 

 tion to DNA, some enzymatic activity. 

 Above pH 6.5, they disintegrate com- 

 pletely, in contrast to cytoplasmic com- 

 ponents, which are better preserved in 

 nearly neutral media (Bounce, A. L., 

 Ann. N. Y. Acad. Sci., 1950, 50, 982- 

 999). 



It might appear that ideal media 

 would duplicate the tonicity, pH, and 

 electrolyte pattern of cytoplasm, and 

 some attempts in that direction have 

 been made (Wilbur, K. M. and Andre- 

 son, N. G., Exp. Cell Res., 1951, 2, 

 47-57). However, the most versatile 

 media employed to date have been simple 

 iso- or hypertonic sucrose solutions, 

 without buffers or electrolytes. These 

 media, especially if hypertonic, yield 

 excellent morphological preservation of 

 all components, including nuclei, which 

 resemble those of living cells in their 

 optical homogeneity (Schneider and 

 Hogeboom, Ibid.). One wonders if this 

 is really an advantage in the determina- 

 tion of nuclear nucleic acids, where 

 acid precipitation may be desirable, 

 but the question is not fully answered. 

 Bearing on this problem is the fact that 

 many such unbuffered homogenates are 

 slightly acid, due probably to glycoly- 

 sis, but that liver appears to be an 



exception in this respect (Wilbur and 

 Anderson, Ibid.). 



The separation of cell components 

 by centrifugation has relied more upon 

 difference in particle size, with resulting 

 difference in velocity of fall, than upon 

 any variation in specific gravity. This 

 would be expected from a consideration 

 of Stokes' Law, which states that the 

 velocity of particle fall is proportional 

 to the square of its radius, but only 

 directly proportional to the density 

 difference between particle and medium. 

 Moreover, we are dealing with semi- 

 permeable or frankly porous cell parti- 

 cles, whose density is altered with that 

 of the medium, so that very dense (and 

 correspondingly viscous) fluids must 

 be employed to effect separation by 

 flotation. In these cases we may be 

 roughly measuring 'dry weight densi- 

 ties', as in the Behrens procedure. An 

 obvious exception to these considera- 

 tions is that of lipid-rich constituents, 

 which migrate centripetally. 



As a result, whole cells are commonly 

 sedimented at very low speeds, nuclei 

 at approximately 5(X) x g mitochondria 

 at 2000-20,000 x g, and microsomes at 

 20,000 to over 100,000 x g. The "super- 

 nate" is the remaining non-sediment- 

 able fraction, and with the preceding 

 particulates, completes the list of 

 usually studied components. Others, 

 such as chromosomes and melanin gran- 

 ules, will not be considered here. The 

 time of centrifugation or field required 

 varies with the viscosity of the medium, 

 the time ranging from a few minutes to 

 several hours. 



A layering technique, as emphasized 

 by Wilbur and Anderson (Ibid.), is the 

 most efficient means of centrifugation, 

 as the mean distance of particle fall is 

 both increased and nearly equalized. 

 This naturally has the effect of both 

 isolating and washing the most rapidly 

 sedimenting fraction at a single step. 

 From this and the above considerations 

 it is obvious that a more efficient centri- 

 fuge would be one constructed to hold 

 longer tubes. With such an instru- 

 ment, employing tubes only double the 

 length of standard models, an even 

 further fractionation of particulates 

 might be achieved, with diminished loss 

 due to washing. Separation of nuclear 

 types suggests itself, and some en- 

 deavors in that direction are promising 

 (Marshak, A., Cancer Res., 1950, 10, 

 232). 



Below is a brief outline of some of 

 the more popular and useful techniques, 

 listed according to homogenizing fluids 

 employed. Centrifuging media are us- 

 ually similar to these, or made slightly 



