THE OSMOTIC PRESSURE OF THE PROTEINS OF HUMAN SERUM AND PLASMA. BY E. B. VERNEY (Beit Memorial Research Fellow). (From the Laboratories of the Medical Unit, University College Hospital Medical School.) IN 1896 Starling(l) showed by direct measurement that the serum proteins exerted an osmotic pressure of 41 to 56 cm. H20 when put up in bell-shaped osmometers against 1F03 p.c. NaCl. In a later paper(2) this fact was confirmed, and by concentrating serum by filtration through a porous cell soaked in gelatin and allowing the concentrated serum to dialyse against the protein-free filtrate, it was shown that there was a rough proportionality between the protein content of the serum and the observed final pressure. For example, a serum containing 6-5 p.c. protein exerted a colloidal pressure of 38 cm. H20, and one containing 11 2 p.c. protein a pressure of 61 cm. H20, giving pressures of 36 and 34 cm. H20 respectively per grm. protein nitrogen p.c.1. In 1907 Moore and Roaf(3) measured the osmotic pressure of various colloidal solutions including serum. Their figures for pig's serum showed this to be 24 cm. H20 at 120 C. and 42 cm. H20 at 180 C., the protein content of the serum being 7-97 p.c. The magnitude of the pressure discovered by Starling was thus confirmed. Recently Govaerts(4) has described a method based on the apparatus of Moore and Roaf in which the pressure of the non-diffusible constituents is exerted on a bubble of air in a capillary tube, the final pressure of which is read by connecting it to a water manometer in the manner described by Krogh 5). Govaerts' apparatus can be used with as little as 1-5 c.c. serum or even less. The semi-permeable membrane used is "cellophane," which Govaerts states is permeable to NaCl and glucose and retains gum arabic and all proteins precipitable by trichloracetic acid. He finds normal human serum to exert an osmotic pressure of 35 to 40 cm. H20 or a mean of 29 cm. per grm. of protein nitrogen p.c. In clinical cases of severe cedema the pressure varies from 12 to 22 cm., giving a value of 13 to 19 cm. per grm. of protein nitrogen p.c. On the other hand, in pure hypertension 1 Corresponding to pressures of 5-8 and 5.5 cm. H20 respectively per grm. protein p.c.
320 B. B. VERNEY. cases the pressure was found to be raised (40-70 cm.), giving a value of 32 to 34 cm. H20 when calculated as the pressure exerted by 1 grm. protein nitrogen p.c. Govaerts states that the low values found in cases of cedema result from two factors: the dilution of the blood and the diminution of the pressure per grm. protein p.c., the reverse factors acting inversely in the group of hypertension cases. Govaerts gives no figures for the effect of simple dilution on the osmotic pressure of normal serum proteins, and it would seem desirable that this should be investigated before it is justifiable to invoke a second factor in the interpretation of the low values found in some of his clinical groups. This relationship has been determined and the results will be reported in this paper. Method and technique. An apparatus identical in its essential components with that used by Govaerts has been employed. I am indebted Fig. 1. Apparatus for testing permeability of "cellophane" membrane. Fig. 2. Equilibrium attainment curves between equivalent solutions of NaCl -*., CaCl2 --, CO(N 2L-.-*--, and distilled water. Ordinate =Milli-equivalents. Abscissa =Time in hours.
OSMOTIC PRESSURE OF PROTEINS. 321 to him for the supply of "cellophane" with which these experiments have been carried out. The membrane has been tested for its permeability to various substances in the crystalloid state and to serum proteins and haemoglobin by making use of the apparatus shown in Fig. 1. It consists of two symmetrical and cylindrical glass cups A and B. The everted lips of these cups are ground flat, and between them is placed a circular disc of cellophane. The two vessels are clamped securely together, their respective cavities being then separated merely by the cellophane membrane. Four short tubulures, D, E, F, G, two from the top of each vessel, communicate with the exterior. Through two of them, D and E, pass thin glass tubes which on the one hand reach to the bottoms of the cylinders and on the other are connected to an air pressure pump, so that a slow stream of bubbles may keep the solutions on either side of the membrane efficiently mixed. The capacity of each vessel is 45 c.c. and the surface area of one side of the intervening membrane is 15 sq. cm., giving a diffusion area of 033 sq. cm. per c.c. of fluid as compared to 056 sq. cm. in the case of the osmometers used in the present investigation. When laked blood was placed in B and distilled water in A, no protein was detected in the latter at the end of 24 hours. The rate of attainment of H equilibrium between equivalent solutions of NaCl, CaCl2 and CO(NH2)2 and distilled water at room temperature, 19 to 200 C., is shown in Fig. 2. A was filled with distilled water and an equal volume of the solution in question placed in B. 2 c.c. were removed for analysis at intervals through Ml the tubulures F and G. It will be seen from the figure that equilibrium is estab- 0 lished in the case of these three bodies within 12 hours. The disposition of the osmometers has c been slightly altered from that adopted by Govaerts in order that the whole appa- N ratus may be submerged in a thermostat.. - A sectional drawing of one osmometer under tk 7 this arrangement is shown in Fig. 3. The D whole technique is carried out with strictly aseptic precautions. If these are not taken Fig. 3. Osmometer. organisms readily grow in the serum and in the dialysate, both of PH. LXI. 22
322 B. B. VERNEY. which become cloudy, and this infection is accompanied by a progressive fall in the osmotic pressure of the proteins. The rubber washers, cellophane discs and Ringer's solution are autoclaved for 45 minutes at 1 atmospheres and the osmometers and pipettes and centrifuge tubes dry sterilised at 1300 C. for 1 hour. The inside of the capsule, the perforated copper disc, and the rim around the periphery of the dialysate are coated with a layer of sterile paraffin wax, as recommended by Govaerts, so that in no place do the fluids come into contact with the metal of the osmometer. The capsule when completed holds 1 c.c., and before filling it with serum it is tested for leaks by placing a little Ringer on the surface of the membrane and raising the internal pressure. Serum or plasma is then pipetted into the capsule, the capillary D inserted to such an extent that the fluid passes round to the level shown in the figure at F, the tap G being open. The top of the capillary H is then connected to the mercury level (see Fig. 4) and the meniscus of the serum exposed to a pressure approximately equivalent to that which experience has shown to be the expected final osmotic pressure value. The capillary is then pushed further into the capsule, so as to bring the meniscus back to its original level, and sealed in position with paraffin wax. The tap G is closed, the osmometer cup placed over the capsule on to the thin rubber bung, and about 5 c.c. of Ringer placed on the surface of the bung by means of a curved pipette. The osmometer is now immersed in the thermostat and allowed to reach the bath temperature. This usually takes about half an hour and can be assumed to have occurred when re-exposure of the meniscus F to the same counter pressure as before causes no change in its position. 0 5 c.c. Ringer's fluid is now dropped on to the surface of the membrane by means of the special pipette shown at M, the point being guarded by a rubber tube N in order that it may not become contaminated by accidental contact with the inside of the tube 0 of the osmometer cup. A long tube is connected to the end of this, a piece of wool soaked in Ringer lightly plugs the junction, and the whole apparatus is more deeply immersed in the bath into the position shown in Fig. 4. Twentyfour hours later the pressure of the air confined between the meniscus and the tap is determined by observing the meniscus by means of a microscope fitted with a 3" objective and micrometer eyepiece, opening the tap and bringing the meniscus to its original level by adjusting the mercury cup (Fig. 4). The reading of the water manometer corrected for the difference in level between the meniscus and the surface of the dialysate, and for the capillarity of the tube D (Fig. 3), which in my
OSMOTIC PRESSURE OF PROTEINS. 323 apparatus amounted to about 1 cm. H20, gives the osmotic pressure of the non-diffusible constituents of the serum. The osmometer is then dismantled, the capillary tube taken out without opening the tap G, and 4L the serum removed for the determination of the total and non-protein nitrogen by means of a Kjeldahl apparatus. Six similar osmometers have been used and they have been set up 22-2
324 E. B. VERNEY. in parallel and connected to the mercury level by means of a capillary tree as shown in Fig. 4. Experimental results. The degree of accuracy of the method is shown in Table I. TABLE I. Cat's serum. 1 c.c. placed in each osmometer. Dialysate=0-5 c.c. Ringer without bicarbonate. Set up 29. xi. 24 at 4 p.m. under gross counter pressure of 30 cm. H20. Temperature = 17-40 C. Max. Max. percentage Gross Mean deviation error No. of Temp. pressure Level Capil- Osmotic osmotic osmo- of reading diff. larity pres- prescm. cm. H20 sure sure mean from from meter bath cm. H2O mean Time 1. xii. 24 10.30 a.m. 1 2 3 4 5 17-4 9 9 9 42-2 400 41-0 43-3 41-0 -0-9 -003-0-9-1-0-0-5-1-2-1-3-1-3-1-6-1-3 40-1 38-4 38-8 40-7 39-2 39.3 1-4 43-6 1. xii. 24 5 p.m. 1 18-0 42-3 -0-9 -1-2 40-2 2 9 40-6 -003-1-3 39-0 3 3 40-3 -009-1-3 38-1 9 4 9 41-4 - 1-0 -1-6 38-8 9 5 Y 41-5 -0.5-1-3 39.7 TABLE II. Human serum. 10 c.c. venous blood drawn 9. xii. 24. Repeated 1 hours later. Serum A and B set up under initial pressure of 36 cm. HBO against 0-5 c.c. Ringer (RSB) at 6.15 p.m. in duplicate. Temp. = 18.40 C. Osmotic Protein Osmotic pressure pressure Mean nitrogen per grm. protein Time Temp. 'C. cm. H20 value mgrm. p.c. nitrogen p.c. A B 10. xii. 24 1.30 p.m. 4.30 p.m. 6p.m. 10. xii. 24 1.30 p.m. 4.30 p.m. 6 p.m. The Ringer's composition: 18-4 18-4 34-4 35-8 34-2 35-2 35-1 35.7 36-2 36-9 36-0 37.4 36-3 37.4 35-0 1150 36-7 1150 30-5 31-8 fluids used in these experiments had the following NaCl KCI CaC12 NaHCO3 0-85 p.c. 0-042 = RCB 0-024 0-02
OSMOTIC PRESSURE OF PROTEINS. 325 on the one hand, and the same fluid without the bicarbonate (= RSB) on the other. The osmometers were set up invariably in triplicate or duplicate. Two control experiments on human serum are given in Table II. No appreciable difference between serum and plasma proteins could be detected in the osmotic pressure values, calculated per grm. protein nitrogen p.c., for the blood of the same subject drawn at different times, as the following table shows. TABLE III. A =Human serum readings. 29. v. 25, 12 c.c. blood drawn from vein. Serum set up against 0*5 c.c. RCB against initial pressure of 30 cm. HEO at temperature of 150 C. in triplicate, at 6 p.m. B=Human plasma readings. 3. vi. 25, 11 c.c. blood drawn from vein into 0 1 c.c. saturated neutral potassium oxalate. Plasma set up as serum A against initial pressure of 35 cm. H20 at 150C. in duplicate at 7 p.m. Osmotic Protein Osmotic pressure Temp. pressure Mean nitrogen per grm. protein Time 0 C. cm. H120 value mgrm. p.c. nitrogen p.c. A 30.v.25 1.30 p.m. 15.0 36-9 - 36-4 35-5 36-8 1200 30-1 2. vi. 25 11 a.m. 15-5 9 37.5 91 t 37-0 B 4. vi. 25 7 p.m. 17-7 33-1 pi, ll 33.7 5. vi. 25 10 a.m. 17-9 32*9 33-6 1120 30 0 33.5 2 p.m. 1851 34.4 9.9 34-0 When the plasma was diluted with Ringer a fall was encountered in the osmotic pressure greater than could be attributed to the proportional fall in protein concentration (see Table IV). TABLE IV. 8 c.c. blood drawn from arm vein into 0-1 c.c. saturated potassium oxalate. Plasma diluted 1: 1 with Ringer (RSB). Osmometers set up as usual and exposed to initial counter pressure of 16 cm. H2O in triplicate at 4 p.m., 8. vi. 25. Osmotic Protein Osmotic pressure Temp. pressure Mean nitrogen per grm. protein Time C. cm. H4O value mgrm. p.c. nitrogen p.c. 9. vi. 25 10 a.m. 19-8 12-8 ~~~~11*1 9 9 10.7 3.15 p.m. 200 13-0,oil ils, 11*0 pi, 9 10-6 11*5 550 20 9
326 E. B. VERNEY. It was of interest to investigate this further and consequently a series of experiments was carried out, the oxalated plasma of the same subject being diluted to varying degrees with Ringer both with and without bicarbonate, and the resultant protein osmotic pressures determined in the manner already described. The results are summarised in Table V. TABLE V. Protein Osmotic Osmotic nitrogen pressure pressure Temp. mgrm. p.c. cm. H20 per grm. V Nature of Nature of 0C. =PN =p PN p.c. 1/p i.e. 1/PN diluent dialysate 19.0 1504 49-8 33-2 67-5 74.5 RSB 17-9 1120 33-6 30-0 100 100 - RSB 25-5 1047 32-4 29-6 103 5 107 - RSB 16-3 1025 29-5 28-7 114 109 RSB 18-8 1010 28-2 28-0 119 111 - RCB 25-3 1005 27-5 27-4 122 112 - RSB 16-7 992 29-2 29-4 115 113 RSB 17-5 940 20-0 21-3 168 119 RCB RCB 18-8 740 17-1 23-1 197 151 RCB RCB 16-7 725 15-8 21-8 213 154 RSB RSB 18-8 700 14-9 21-3 226 160 RCB RCB 25-3 675 13-6 20-2 248 166 RCB RSB 25-5 670 13-4 20-0 250 167 RSB RSB 16-3 618 12-9 20-8 261 181 RSB RSB 25-5 606 13-3 21-6 257 185 RCB RSB 17-5 570 11-7 20-6 287 197 RCB RCB 19-9 550 11-5 20-8 292 204 RSB RSB 16-7 538 10-8 20-1 311 208 RSB RSB 17-5 500 7.4 14-8 454 224 RCB RCB 16-3 442 6-6 14-9 510 253 RSB RSB RCB =Ringer's fluid with 0-02 p.c. NaHCOO. RSB =Ringer's fluid without NaHCO3. The figures in the first row are those obtained from plasma concentrated by ultrafiltration. Those in the second row are taken as the normal plasma values, with which the remainder are compared in columns5 and 6. It will be seen that the osmotic pressure per grm. protein nitrogen p.c. gradually falls as the plasma is diluted. Ada ir(6) has observed this phenomenon in the case of haemoglobin dissolved in either N/10 NaCl, or(7) distilled water. The possible cause for this which first occurred to the mind was that the molecular volume of the colloidal particles was comparatively large, and that one was therefore dealing with asolution in a state analogous to that exhibited by a gas when highly compressed. The reciprocal of the osmotic pressure and the reciprocal of the protein nitrogen were therefore calculated, the value of each for normal plasma being taken arbitrarily as 100, the values for the diluted and concentrated plasmas being interpreted in figures relative to this. The figures are given in columns 5 and 6 and are plotted in Fig. 5. It will be seen that for dilutions ranging up to 50 p.c. of the original concentration, the points
OSMOTIC PRESSURE OF PROTEINS. 327 lie fairly accurately on a straight line'. The figures in the first line of Table V were obtained from a plasma which had been concentrated by 500 + 400~~~~~~~~~~~~~~ 400 -t 200 p(,-b) =k 100 jt B A Fig. 5. b*e 50 100 200 A, original plasma; B, plasma concentrated by ultrafiltration. Ordinate =Osmotic Pressure Original Plasma Osmotic Pressure Diluted Plasma Abscissa =Protein Nitrogen Original Plasma x 100 Protein Nitrogen Diluted Plasma +represents readings when bicarbonate-free Ringer, and 0 readings when bicarbonate Ringer, was used as the diluent. ultrafiltration through cellophane, and it will be observed that the plotted point for this concentrated plasma lies in the neighbourhood of the same straight line. It seemed to make no appreciable difference to 1 Dr G. S. Adair informs me that he has obtained a similar form of curve in the case of sheep and of horse plasma with protein concentrations ranging from 1 p.c. to 14 p.c. 100
328 E. B. VERNEY. the values obtained whether the plasma were diluted with Ringer containing bicarbonate or no bicarbonate. Dilutions in the neighbourhood of 40 p.c. of the original concentration give rise to a relatively smaller pressure and fall distinctly above the line to which the figures of the more concentrated solutions adhere. If this line be produced it will be seen that it cuts the v axis at the value of 50. In other words, the osmotic pressure of the plasma proteins reacts to their dilution, within limits, in a manner such as would be expected of a non-ionised colloidal solution in which the colloidal molecules occupied an effective volume as large as 50 p.c. of the original. SUMMARY AND CONCLUSIONS. 1. The osmotic pressure of the proteins of human serum and plasma has been determined. 2. Dilution of the plasma with Ringer's fluid gives rise to a relatively larger fall in osmotic pressure than the concomitant fall in the protein concentration. 3. If p = the osmotic pressure of the proteins, v= the reciprocal of the protein nitrogen, and b= a constant, the relation p (v - b) = k is shown to hold for dilutions of the plasma up to 50 p.c. of the original concentration. 4. The constant b has a value of 50 p.c. of the original volume of the plasma. The expenses of this research were defrayed in part out of a grant from the Govemment Grant Committee of the Royal Society. REFERENCES. 1. Starling. This Journ. 19. p. 312. 1896. 2. Starling. Ibid. 24. p. 317. 1899. 3. Moore and Roaf. Biochem. Journ. 2. p. 34. 1907. 4. Govaerts. Bull. de l'acad. Roy. de Med. de Belg. 4. p. 161. 1924. 5. Krogh. The Anatomy and Physiology of the Capillaries. Yale University Press, 1924. 6. Adair. Proc. Camb. Phil. Soc. (Biol. Sciences), 1. p. 75. 1924. 7. Adair. Proc. Roy. Soc. A, 109. p. 292. 1925.