224

G. M. WARNOCK AND J. DUCKWVORH

1944

are not used as a source of minerals to sustain the mammary supply during lactation. In other words, an osteoporotic skeleton is not an obligatory consequence of lactation in this species at least. Considerable difficulties exist in the determination of these requirements by the metabolic balance technique, since this method does not indicate when mobilization of minerals from the spongiosa is accompanied by mobilization from the shaft. Further, the balance over a whole lactation is often small in relation to the total quantities of materials analyzed and may be close to the analytical errors of the methods when the diet given is rich in minerals. The present observations may find some application in the determination of certain mineral requirements of lactation. Because the losses of Ca and P are unavoidable, involving the gradual liberation of reserves laid down at an earlier peripd, it has been difficult to estimate the requirements for lactation on the bafsis ofthe negative values obtained by the balance technique. Since the present study shows that, in the rat, it is possible to restrict the withdrawal of bone salt to the spongiosa, leaving the shaft unaffected, it seems reasonable to suggest that lactation requirements could be defined as those intakes which, for a given output of minerals in milk, will restrict the withdrawals to the spongiosa. Thus, using a similar technique, the problem ofthe mineral requirements for lactation in domestic aniimals might become more amenable to experimental attack. This preservation of the diaphyses may be an unattainable ideal in the high-producing
Bauer, W., Aub, J. C. & Albright, F. (1929). J. exp. Med. 49, 145. Bloom, W. & Bloom, M. A. (1940). Anat. Bec. 78, 497. & McLean, F. C. (1941). Anat. Bec. 81, 443. Coons, C. M., Schiefelbuwch, A. T., Marshall, G. B. & Coons, R. R. (1935). BuU. Okla. agric. Exp. Sta. no. 223. Duckworth, J. & Warnock, G. M. (1942-3). Nutr. Abetr. Rev. 12, 167. Forbes, E. B., Black, A., Braman, W. W., Frear, D. E. H., Kahlenberg, 0. J., McClure, F. J., Swift, R. W. &

dairy cow, but in sheep and swine it may be possible as in the rat. Also since Theiler (1934) pointed out that, in cattle, P deficiency had a more marked effect on-the skeleton than Ca deficiency, the method migit be more suitable for the determination- of P than Ca requirement in the cow. While no evidence has yet been produced to show that a mild degree of osteoporosis is seriously detrimental to the individual, its avoidance would preserve the value of the skeleton from a struictural standpoint.

SUMMARY 1. A technique for studying changes in the ends and shafts of the femur and tibia of the rat is described and used to follow the effect of gestation and lactation upon bone. 2. The percentage ash content and weight of dry, fat-free bones of well-fed rats were determined. During lactation and pregnancy the changes which occurred were confined to the bone ends. 3. The resorptio4 of the spongioea whicll occurred in lactating rats was not observed in nonlactating aniinals. The rate of replenishment of the depleted spongiosa at the end of lactation was the same, irrespective of whether the animals were then pregnant or non-pregnant. 4. The shafts of the bones of rats were little affected during lactation. The use of this finding in the determination of the Ca and P requirements of domestic animals is discussed. The' authors are indebted to Mr A. C. Dalgarno for assistance in the care of the animals.
Voris, L. (1935). Tech. BuU. Pa agric. Exp. Sta. no. 319. Goss, H. & Schmidt, C. L. A. (1930). J. biol. CIem. 6, 417. Kyes, P. & Potter, T. S. (1934). Anat. Bec. 60, 377. McLean, F. C. & Bloom, W. (1940). Anat. Bec. 78, 333. - - (1941). Arch. Path. 32, 315. Theiler, A. (1934). Vet. J. 90, 143. &; Green, H. H. (1931-2). Nutr. Abetr. Rev. 1, 359. Camb. 36, 24. Thomson, W. (1936). J. Hyg.,J

REFERENCES

Qualitative Analysis of Proteins: a Partition Chromatographic Method Using Paper
BY R. CONSDEN, A. H. GORDON AND A. J. P. MARTIN, Wool Industries Research A88ociation, Totridon, Headingley, Leeds, 6

(Received 13 May 1944)'

Gordon, Martin & Synge (1943b) attempted to separate amino-acids on a silica gel partition chromatogram, but found it impracticable owing to. adsorption by the silica of various amino-acids.

They obtained, however, good separations by using cellulose in the formn of strips of filter paper. Following further work along these lines, the present paper describes a qualitative micro-analytical tech-

Vol. 38

PARTITION CHROMATOGRAPHY WITH PAPER

225

nique for proteins. Using only 200,ug. of wool, it is possible by this method to demonstrate the presence of all the amino-acids which have been shown to be there by other methods. The method is rather similar to the 'capillary analysis' method of Schonbein and Goppelsroeder (reviewed by Rheinboldt, 1925) except that the separation depends on the differences in partition coefficient between the mobile phase and watersaturated cellulose, instead of differences in adsorption by the cellulose. That adsorption of the aminoacids by the cellulose plays no significant part is seen from Table 1, where the partition coefficient calculated from the rates of movement of the bands are compared with those found directly by England & Cohn (1935). Too much stress should not be laid tipon the agreement of these figures, which are based upon an assumed water content of the saturated cellulose and the assumption that the ratio of the weight of n-butanol to paper is constant in all parts of the strip. This assumption does not hold accurately. Nevertheless, the conclusion seems justified that the cellulose is playing the role of an inert support. The most satisfactory solvents are those which are partially miscible with water. Within a homologous series of solvents the corresponding rates of movement of the amino-acids change in the same direction as the water solubility of the solvent: solvents completely miscible with water can be employed provided that the water content is not too high. In this case, presumably, the cfallulose, by a 'salting out' effect, allows the system to function as a partition chromatogram. However, the aminoacid bands obtained are much broader than is the case with immiscible solvents. This is no doubt due to a variation in the composition of the phases caused by the presence of the amino-acids. This effect has been noticed in the n-butanol-water system by England & Cohn (1935) and is the limiting factor in the amount of amino-acid that can be employed in a given chromatogram. It is reasonable to suppose that the phases will be more easily disturbed by employing a miscible rather than an imrmiscible solvent. The main effect of temperature on the rates of movement of amino-acids is also explicable in terms of the change in composition of the phases. Thus, in the phenol-water system, increase of temperature increases the miscibility and the rate of movement of the bands. In the collidinewater system the reverse is true. Further, the greater the difference between the working temperature and the critical solution temperature, the less sensitive will the rates be to change of temperature. However, though the absolute partition coefficients may be greatly changed, the ratios of the partition coefficients of the respective acids are almost unaltered.

There is an obvious advantage in working with unsubstituted amaino-acids in that any substituent group, even though small, renders the physical properties of the derivatives more similar and hence increases the difficulties of separation. Martin & Synge (1941) and Gordon, Martin & Synge (1943a) failed to separate the slower moving acids when acetylated, whereas all the acids are separable by the present technique. Moreover, selective losses associated with acetylation and extraction are obviated. A considerable number of solvents has been tried. The relative positions of the amino-acids in the developed chromatogram depend upon the solvent used. Hence by development first in one direction with one solvent followed by development in a direction at right angles with another solvent, amino-acids (e.g. a drop of protein hydrolysate) placed near the corer of a sheet of paper become distributed in a pattern across the sheet to give a two-dimensional chromatogram characteristic of the pair of solvents used. Advantage is taken of the colour reaction with ninhydrin (see Copley, 1941, for bibliography) to reveal the positions of the amino-acids.

Relation of partition coefficient to rate of movement of bands
The relation of the partition coefficient to the rate of movement of the band may be calculated by the method of Martin & Synge (1941). A = cross-sectional area of paper + water + solvent, AL = cross-sectional area of solvent phase, As = cross-sectional area of water phase, a =partition coefficient concentration in water phase cpncentration in solvent phase' A R =

AL +cAS'

However, R is not conveniently measurable in paper chromatograms, so a new symbol, RF is introduced.*
R

movement of band movement of advancing front of liquid

RAL
A
or

AL
AL+ OLAS
)

1 RFAS AS AS (RF
A
A

z

ALIAS is equal to the ratio of the volumes of solvent and water phase in the chromatogram. * The symbol RF is equivalent to the symbol R used by LeRosen (1942) who apparently received the paper of Martin & Synge (1941) too late to avoid using the symbol R in a different sense.

226

R. CONSDEN, A. H. GORDON AND A. J. P. MARTIN
Table 1. Partition coefficient8 clulated from Rp vaiu es 2 4 3 I 1-No. of run 22-6 18-0 17-7 % water in paper' ... 28-7 4-56 2-93 3-70 ... 3-25 AL fAa Partition coefficients: 70-4 70-4 70-4 Glycine 70-4
Alanine Valine Norvaline Leucine
35-9 12-2 8-7

1944
Direct measurements England & Cohn (1935)
70-4 42-3 13-8
5-5 3-2

Norleucine

4.5 3.5

39.9 14-1 10-8

43-7 14-8

4-2

5.4

10-5
5-6 4-4

36-6 12-5 9-2 6-0 4-6

9.5

Assuming a given water content of the paper,

AL/AS may be deduced from the ratio of weight of dry paper to that of the developed chromatogram. The water content of the paper is apt to vary from experiment to experiment and is difficult to measure in the presence of n-butanol or other solvent. Table 1 shows the partition coefficient calculated from the Rp and AL/As values for four separate runs under slightly different conditions. The water content has been so chosen that the partition coefficient for glycine is equal to tha't given by England & Cohn (1935). The last column gives the direct measurements of England & Cohn. The water content of saturated cellulose, according to the Intertional Citicl Tables, is 22 % on a dry-weight basis.

strips of sheet glass which replace the bars. The cross-section is shown in Fig. 4. Paper. Whatman no. 1 filter paper has been used almost exclusively. Whatman no. 42, while giving satisfactory separations of amino-acids, was too dense to permit a convenient rate of flow of solvent. In an attempt to separate larger quantities of amino-acids, thicker papers have been tried. Of these, Whatman 'Accelerator' paper (^, in. thick) gave satisfactory separations but was too fragile to withstand normal handling. Whatman 'Seed Testing' paper ( in. thick) was unsatisfactory as uniform flow of solvent could not be obtained. With most solvents the advancing front of liquid is coloured yellowish brown by material extracted from the paper. This contaminant can be partially removed by Soxhlet extraction with ethanol. Chromatographic washing is more effective. However, as the contaminant is so fast-running, relative to the amino-acids in most solvents (except phenol), it therefore does not interfere, and unextraoted paper has usually been used.

EXPERIMENTAL The essentials of the apparatus consist of a filterpaper strip, the upper end of which is immersed in a trough containing the water-saturated solvent. The strip hangs in an airtight chamber in which is maintained an atmosphere saturated with water and solvent. The trough is provided with a bar over which the paper passes to prevent capillary siphoning between the outside of the trough and the paper. The apparatus is shown in Figs. 1-4. Chamber. For one-dimensional experiments, stoneware drain-pipes (6 or 9 in. in diameter and 2 ft. and 2 ft. 6 in. long respectively) have been found convenient. The pipe stands in a close-fitting tray hammered from a sheet of lead. The top of the pipe is grould flat and covered by a sheet of glass (Fig. 3). The glaze on these pipes is imperfect and a pipe ahould be reserved for each solvent and atmosphere. For two-dimensional experiments, the chamber consists of a glass-sided lead box, 30 x 30 x 5 in., made airtight with a well-fitting lead lid. A removable lead tray rests on the floor of the box. A completely air-tight tinplate box, using the tray as a water seal, was used when an atmosphere of coal gas was required. Troughs. The troughs used in the one-dimensional experiments are constructed from i in. bore glass tubing, opened along its length by grinding (Figs. 1, 2). For two-dimensional work, the larger troughs required would be too fragile if of the same pattern. Therefore an iron frame supports the trough (of j in. bore thick-walled tubing) and carries

Procedure' To run a one-dimensional chromatogram a strip of paper, 1-5 cm. or more in width and 20-56 cm. in length, is used. A pencil line is drawn across the strip about 5 cm. from one end. The solution, 2-4,ul. containing 5-15I&g. of each amino-acid to be analyzed, is applied along the centre portion of this line from the tip of a capillary tube. The end of the paper is fixed in the trough wi'th a microscope slide. The trough and paper are now transferred to the chamber, which has been previously prepared by covering the bottom of the tray with a two-phase mixture of water and solvent to provide an atmosphere saturated with both components. The trough is filled with the water-saturated solvent and the lid put on the chamber. When the solvent has run a convenient, distance (15-25 cm. in 6 hr.; 30-50 cm. in 24 hr., depending on solvent and temperature), the paper is removed and the position of the solvent front is marked. The strip is dried, either in ar oven at 1 10' or by hanging in a drying cupboard through which hot air is sucked by a fan exhausting to the outside. After drying, the paper is sprayed with a solution of ninhydrin (0-1 % in n-butanol) and again dried. Finally, the paper is heated at 800 for 5 min. The bands are outlined in pencil, as fading of the colotur takes place after a few days. When it is

BIOCHEMICAL JOURNAL, VOL. 38, NO. 3

PLATE

A

B

C
Two-dimensional chromatogram of a wool hydrolysate (lSOpg.) on Whatman no. 1 sheet. Hydrolysate applied at.circkf Run with collidmne for 3 days in direction AB, then in direction AC with phenol for 27 hr. mn an atmosphere of cos gas and NH,, (produced from a 083% NH3 solution). The filter employed in photographing renders the yellow prolin spot scarcely visible. (Photography by J. Manby, photographer to the University of Leeds.)

LATE 2

BIOCHEMICAL JOURNAL, VOL. 38, NO. 3

A

B

C of each) on Whatman *wo-dimensional chromatogram of a mixture of the hydrochlorides of 22 amino-acids no. 1 sheet, carried out as described under P1. 1. As before, the yellow proline and orange. hydroxyproline spots are not visible in the photograph. (Photography by J. Manby, photographer to the University of Leeds.)

(6-12jsg.

VoI. 38

PARTITION CHROMATOGRAPHY WITH PAPER

desired to run a number of chromatograms simultaneously, the individual solutions may conveniently be placed side by side on a wide strip. It is seldom necessary to leave more than an interval of 1 cm. between the spots, but it is undesirable for the amino-acids to be too near the edge of the paper, as irregularities of flow are usually more pronounced there.

227 2 of development, again for 24-48 hr., now proceeds, the chamber, tray and trough having been prepared
Glass sheet

Glass

Trough

Drain-

pipe

Fig. 1. Cross-section of small trough.

Paper
= -=
-

. :
--

Lead

Water and solvent

tray

Fig. 3. Elevation of smaU trough in drain-pipe.

Fig. 2. Plan of small trough in pipe.

For two-dimensional analyses, a standard sheet 18 x 22i in. is used (Pls. 1 and 2). The solution to be analyzed (6-12 A., representing 200-400,ug. of protein) is placed near the corner, 6 cm. from either edge. The paper is held with pne edge slightly overlapping the opening of the trough and pressed into it with a strip of sheet glass somewhat longer than the paper. After transfer to the chamber, prepared as above, the chromatogram is allowed to develop for 24-72 hr. The paper is then removed and dried in the drying cupboard, turned through a right angle, and returned to the trough. The next stage

Fig. 4. Cross-section of large trough.

for the second solvent during the drying of the sheets. Subsequent treatment is the same as for the strips.

228
E $

R. CONSDEN, A. H. GORDON AND A. J. P. MARTIN o o o o o o o o o o

'944

.0M
Q

g cZ o

0

> o 0

0° t o o o o

0 o o,

o

o

o

0Zo 66 66666666666 ,6 2 0~~~~~~~~~~~~P~ 0 ~ ~ ~ ~ ~

o

2,_§ o _ v O =oZ O

o6 6 6

.o
O

c

e

c

F

-

oC
O

o0I'
O O

0.
O O O O

O

CO

O

2
, o 0

6 6 6 v 0

6

t_ O U: e4 aq0 co eqt- &o cqecc 6 6 6 6 6 6 6 6 6 6 6 6 6 ~66 6
0

ooooooo *

W

6 65

6 C ac< -~~~~~C
9

2
0

~~~~~~~~~~~~~~o

-6e _

c

OO o

CO

(a

C

O

C

°-.

.. 0

CO

o °C>
41 i.
:5
0

6