the Oxygen Transport through - Home - BiologiQ Transport... · other properties is necesary to provide a more detailed picture of the liquid. Measurements of the viscosity and thermal

other properties is necesary to providea more detailed picture of the liquid.Measurements of the viscosity andthermal conductivity at very low tem-peratures would provide useful infornia-tion. The predicted unusual behavior ofsound propagation in the liquid at verylow temperatures is of particular inter-est, and measurements of high-frequencysound propagation and attenuation atextremely low temperatures would be ofgreat value, together with possible ex-periments on scattering of light. Thefact that many of the critical experi-ments involve measurements at temper-atures reached only by magrnetic coolingprovides a technical challenge which isbeing taken up today in many institu-tions specializing in low-temperaturephysics.

References and Notes

1. L. W. Alvarez and R. Cornog, Phys. Rev.56, 379. 613 (1939).

2. S. G. Sydoriak, E. R. Grilly, E. F. Hammel,ibid. 75, 303 (1949).

3. J. G. Daunt, Advantces it Phys. 1, 209 (1952).4. See for example: J. Wilks, Nuiot'o cihtentto

10, suppl., 509 (1953); B. M. Abraham,D. W. Osborne, B. Weinstock, Sciettce 117,121 (1953); E. F. Hammel, in Progress intLow-Tentperature Physics, C. J. Gorter, Ed.(North Holland, Amsterdam, 1955), p. 78;R. A. Chentsov, Uspekhi Fiz. Nauk 55, 49(1955); J. G. Daunt, Ed., Symtposiumt onLiquid attd Solid He" (Ohio State Univ.Press. Columbus, 1957); V. P. Peshkov andK. N. Zinov'eva, Repts. Progr. Phys. 22,504 (1959).

5. F. London, Superfluids (Wiley, New York,1954), vol. 2.

6. A. A. Abrikosov and I. M. Khalatnikov,Repts. Progr. Phys. 22, 329 (1959).

7. K. R. Atkins, Liquid He/itiitt (CambridgeUniv. Press, New York, 1959).

8. H. L. Anderson, Phys. Rev. 76, 1460 (1949);

H. L. Anderson and A. Novick, ibid. 73,919 (1948).

9. F. London, Nature 163, 694 (1949); see alsoE. Pollard and W. L. Davidson, AppliedNuclear Ph/sics (Wiley, New York, 1942),p. 183; J. Franck, Phys. Ret. 70, 561 (1946).

10. J. de Boer, P/th'sica 14, 139 (1948);Repts. Progr. Phys. 12, 305 (1949);and B. S. Blaisse, Phy'sica 14, 149 (1948);J. de Boer and R. J. Lunbeck, ibid. 14,318, 510, 520 (1948).

11. W. H. Keesom, Heliuntt (Elsevier, Amsterdam,1942).

12. J. G. Daunt and R. S. Smith, Revs. ModerttPhys. 26, 172 (1954); K. Mendelssohn inHattdbttc/l der Phisik (Springer, Berlin, 1956),vol. 25. p. 370.

13. L. N. Cooper, R. L. Mills, A. M. Sesster,Phys. Ret'. 114, 1377 (1959). See also R. L.Mills. A. M. Sessler, S. A. Moszkowski,D. G. Shankland, Phvs. Ret'. Letters 3, 381(1959).

14. L. Landau, J. Phys. U.S.S.R. 5, 71 (1941);ibid. 8, 1 (1944).

15. R. P. Feynman, Phis. Ret'. 90, 116 (1953);ibid. 91, 1291, 1301 (1953); ibid. 94. 262(1954); - , in Progress int Lows-Temtpera-ture P/li'isics, C. J. Gorter, Ed. (North Hol-land, Am'nsterdam, 1955).

16. H. C. Kramers, J. D. Wasscher, C. J. Gorter,Physica 18, 329 (1952).

17. M. Cohen and R. P. Feynman, P/ys. Ret.101, 13 (1957).

18. Palevskv. Otnes, Larsson, Pauli, Stedman,ibid. 108, 1346, (1957); Palevsky, Otnes,Larson, ibid. 112, 11 (1958).

19. Yarnell, Arnold, Bendt, Kerr. Phvs. Rev.Letters 1, 9 (1958); Phys. Ret. 113. 1379(1959).

20. D. G. Henshaw, Phys. Ret. Letters 1, 127(1958).

21. P. L. Kapitza, Nature 141. 74 (1938).22. J. F. Allen and A. D. Misener, ibid. 141,

75 (1938).23. L. D. P. King and L. Goldstein, Phy's. Revt.

75, 1366 (1949).24. G. de Vries and J. G. Daunt, ibid. 92, 1572

(1953); ibid. 93, 631 (1954).25. T. R. Roberts and S. G. Sydoriak, ibid. 93,

1418 (1954); ibid. 98. 1672 (1955).26. D. W. Osborne. B. M. Abraham, B. Wein-

stock, ibid. 94, 202 (1954); B. M. Abraham,D. W. Osborne, B. Weinstock, ibid. 98,551 (1958).

27. D. W. Osborne. B. Weinstock, B. M. Abra-ham, ibid. 75, 988 (1949).

28. J. G. Daunt and C. V. Heer, ibid. 79, 46(19511). For more recent evaluations of thelambda temperature as a function of Hes

Oxygen Transport throughHenog-lobin Solutions

How does the presence of hemoglobin in a wet membranemediate an eightfold increase in oxygen passage?

P. F. Scholander

Evolution from single cells to or-ganisms is linked intimately with thedevelopment of a circulatory system.Without this, both size and activitywould be severely limited by the slow-ness of diffusion. But even with a cir-

26 FEBRUARY 1960

culatory system, oxygen transportwould be hampered by still another"unfitness of the environment"-name-ly, the very low solubility of oxygen inwater. This difficulty was overcome bythe evolution of oxygen-carrying pig-

composition, see, for example, T. R. Robertsand S. G. Sydoriaik, Loitw Temiperatuire Plhysic.anid Chtenisrt'y (Uniiv. of Wisconsini Press,Madison, 1958), p. 170; S. D. Elliott, Jr.,and H. A. Fairbank, ibidl. p. 180.

29. D. F. Brewer, J. G. Daunt, and A. K.Sreedhar [Phys. Rev. 115, 836 (1959)] giverefer-ences to and discussions of this questioni.

30. D. F. Brewer, A. K. Sreedhar, H. C. Kramers,J. G. Daunt, Phys. Rev'. 110, 282 (1958).

31. K. A. BruLeckner and J. L. Gammel, inSyjnposium oni Liqid anid Solid He- (OhioState Univ. Press, Columbus, 1957), p. 186;Phys. Ret'. 109, 1040 (1958).

32. L. Landau, Zhutr, Eksptl. i Teoret. Fiz. 30,1()58 (1956) [tranislation, So- iet Phyis. JETP3, 920 (1957)]; Zltiir. Eksptl. i Teoret. Fiz. 32,59 (1957) [translation, Soviet Phys. JETP 5,101 (1957)]; Zlttir Eksptl. i Teoret Fiz. 35,97 (1958) [translation, Soviet Pitys. JETP 35,70 (1959)].

33. A. A. Abrikosov and 1. M. Khalatnikov,Zliur. Eksptl. i Teoret. Fiz. 32, 915 (1957)[translation, Soviet Pitys. JETP 5, 745 (1957)].

34. W. M. Fairbank, W. B. Ard, H. G. Dehmelt.W. Gordy, S. R. Wllliams, Phys. Ret. 92,208 (1953); W. M. Fairbank, W. B. Ard,G. K. Walters, ibid. 95. 566 (1954); G. K.Walters and W. M. Fairbank, ibid. 103, 263(1956); W. M. Fairbank anid G. K. Walters,in Symtposium ott Liquid and Solid Hes(Ohio State Univ. Press, Columbus, 1957),p. 20)5.

35. A. A. Abrikosov anid 1. M. Khalatnikov,Zhur. Eksptl. i Teoret. Fiz. 32, 1084 (1957)[translation, Soviet Phys. JETP 5, 887 (1957)].

36. Similar conclusions were arrived at earlieralso by R. A. Buckingham and H. N. V.Temperley [P/tys. Rev. 78, 482 (1950)] andby I. Pomerancliuk, [Zhir. Eksptl. i Teoret.Fiz. 20, 919 (1950)].

37. K. N. Zinov'eva, Z/ltr. Eksptl. i Teoret. Fiz.34, 609 (1958) [translation, Soviet P/hys.JETP 7, 421 (1958)].

38. H. A. Fairbank and D. M. Lee, in Symposiumon Liquid attd Solid He" (Ohio State Univ.Press, Columbus, 1957), p. 26.

39. A. A. Abrikosov and 1. M. Khalatnikov,Zhur. Eksptl. i Teoret. Fiz. 33, 110 (1957)[translation, Soviet Phys. JETP 6, 84 (1958)].

40. H. L. Laquer, S. G. Sydoriak, T. R. Roberts.Phys. Re'. 113, 417 (1959).

41. K. R. Atkins and H. Flicker, ibid. 113, 959(1959).

42. A. A. Abrikosov and I. M. Khalatnikov,Zhur. Eksptl. i Teoret Fiz. 34, 198 (1958)[translation, Soviet Phys. JETP 7, 135 (1958)].

43. D. F. Brewer and J. G. DatLnt, Phys. Re'.115, 843 (1959).

44. L. Goldstein, Ann. phys. 8, 390 (1959).

ments, which when circulated to thetissues could carry many limes moreoxygen than can water alone.

Oxygen-carrying pigments appearednot only in blood, however, but alsoin a vast area where visible means fortransport of the pigment is lacking-namely, as myoglobin in the musclesystem. Here it is found within themuscle cells, providing, so to speak,the last leg of the supply line to theoxygen-needy contractile machinery.But how could this myoglobin enhanceoxygen transport unless it were circu-lated within the cell? Simlple diffusioncould hardly be aided by the pigment.It is true enough that the increasedoxygen capacity could help to smoothout a fluctuating demand, as demon-strated by Millikan (1), but could itpossibly also be that the pigment might

-L

The author is professor of physiology atScripps Institution of Oceanograpphy, Universityof California, La Jolla.

585

on

Nov

embe

r 22

, 201

0w

ww

.sci

ence

mag

.org

Dow

nloa

ded

from

serve as a specific conveyor belt foroxygen, enhancing its rate of delivery?

If we turn from a normally aeratedenvironment to habitats of low oxygentension we find that many animalsdevelop high concentrations of oxygen-carrying pigments. One may mentionmud-dwelling worms or insect larvae(chironomids), certain crustaceans instagnant pools (Daphnia), intestinalnematodes and maggots (Ascaris andGastrophilus), and many mammals ex-posed naturally or experimentally tohigh altitudes. And again we are drawnto the fundamental question: Is it pos-sible that an oxygen-carrying pigmentcan enhance oxygen transfer througha stationary solution?A great deal of penetrating experi-

mental and theoretical work has beendone to elucidate the kinetics of situa-tions where oxygen, carbon monoxide,and other gases load and unload hemo-globin solutions (2). Unfortunately,however, this approach is not so suit-able if one wishes specifically to studywhat happens in a pure steady-statesituation, since the capacity factor caneasily obscure the steady-state events.The primary aim of the investiga-

tion under discussion has been to findout what happens when air diffusesthrough human hemoglobin solutions.Complications from oxygen capacityor storage have been eliminated by ad-hering strictly to steady-state condi-

D'

tions. It is shown below that the rateof oxygen transport may indeed be en-hanced by the presence of hemoglobin,and a few preliminary experiments onmyoglobin and other pigments indicatethat the phenomenon may have widerapplications (3).

Methods

Principle. Moist air at various pres-sures is diffused through a Milliporemembrane charged with a hemoglobinsolution; on the other side of the mem-brane a moist vacuum is maintained.When a steady state is reached, theamounts of oxygen and nitrogen gascoming through the membrane in ameasured time interval are analyzed.With the nitrogen values as a base line,the oxygen transport can be calculated.

Instrumentation. The diffusion ap-paratus (Fig. 1) consists of a diffusionchamber which may be divided intotwo compartments by a wet Milliporefilter. The membrane rests on a stain-less-steel grid. The upper compartmentis connected to a vacuum gauge and aglass-stoppered vent tube (D); thelower compartment, to the vacuumchamber by means of a taper joint.Each compartment is kept moist by apiece of wet filter paper.

The vacuum chamber is a 300-mlmercury reservoir connected through a

Hg LEVELBULB

VACUUM

Fig. 1. Apparatus for steady-state diffusion.

586

GASTRANSFER

Hg CUP

gas trap and stopcock to the levelingbulb. The chamber has one wide-bore(5-mm) stopcock (A) to the diffusionchamber and a regular one of narrowerbore (B) to the vacuum line. The up-per end of the vacuum chamber ter-minates in a 1-mm-bore capillary andcup which can be closed under mer-cury by inserting into it a polyethylene-tipped wire plug. The capillary carriesa movable millimeter scale to meas-ure the total amount of gas.Gas is transferred to the micro gas

analyzer by means of a micro syringe.This is made from a glass tube of 2 mminside diameter fitted with a fine poly-ethylene tip and a wire plunger. Thelatter is made airtight by means of apiece of polyethylene tubing which isflared at the end by heating slightly.This micro syringe is filled with mer-cury, and the gas is always protectedby mercury during handling. From thispipette the gas is introduced into themicro analyzer (4).The Millipore membrane (HA

grade) holding the hemoglobin solutionhas a porosity fine enough so that airpressure of 1 atm will not break thecapillarity. This filter, which is 0.15mm thick, has about 80 percent spacefor the liquid.

Procedure. The basic data were ob-tained from heparinized human blood.The cells were washed twice in 0.9percent saline and hemolyzed by re-peated freezing and thawing. The ox-ygen capacity of the hemoglobin so-lutions was determined by the syringemethod (5).To saturate the dry Millipore mem-

brane with the hemolyzed cell solution,it is placed on top of the solution untilit is soaked through, whereuponit is submerged. It is then blotted onboth sides and placed on the grid ofthe lower half of the diffusion cham-ber, which is kept moist by a ring ofwet filter paper. The upper half, alsocarrying a wet filter paper, is screwedvacuum-tight onto the lower half withthe wet membrane in between. A widerubber band prevents evaporation fromthe crack. The chamber is put on thetaper joint of the vacuum chamber,which is stoppered by the plug, ani(evacuated through B. B is closed, Ais opened momentarily, and the cham-ber is again evacuated through B. Thisprocess is repeated twice, and the airpressure on the membrane is adjustedto the desired level through screwclamp C and vent D.

SCIENCE, VOL. 131

on

Nov

embe

r 22

, 201

0w

ww

.sci

ence

mag

.org

Dow

nloa

ded

from

Diffusion of gases now proceedsfrom the upper air compartmentthrough the membrane and into thewide-bore channel to -the vacuum cham-ber. The time of diffusion is clockedby means of a stop watch, and theamount of gas which has movedthrough the membrane is periodicallychecked by closing off the diffusionchamber 'at stopcock A, letting themercury rise, and measuring thevolume of the gas bubble in the capil-lary. Twenty cubic millimeters of gassuffice for an accurate determinationof the quantity as well as the compo-sition.The gas transfer is executed with the

10002 P 02 IN MmHgN2

2/3 /3 t /12 °AIR PRESSURE IN ATM

Fig. 2. Diffusion of air through hemoglo-bin solutions of three concentrations.Curve 1/1 represents oxygen capacity of22.9 volumes percent; the other curvesrepresent oxygen capacity of 1/2 and /4this amount, respectively. Horizontal line,02/N2 ratio through plasma; lower diag-onal line, rate of oxygen diffusion throughplasma; three dashed parallel lines, rela-tive rates of oxygen transport calculatedfrom the corresponding 02/N2 data;shaded area, water-vapor tension.26 FEBRUARY 1960

gas pressure in the capillary slightlybelow atmospheric pressure. The plugis removed and quickly replaced by themercury-filled transfer pipette. The gasis drawn up into this and is followedby mercury, whereupon the pipette isremoved and replaced by the plug. Thegas sample is now transferred to thecup of the analyzer containing alkalinecitrate, and the oxygen and nitrogenare determined by absorption in theconventional way (4).

Successive samples are analyzed un-til the 02/ N2 ratio becomes constant--that is, until the oxygen content of thefilter is constant. Under the least favor-able conditions (high pressure and highhemoglobin concentration) the mem-brane contains initially, at most, some14 to 20 mm3 of oxygen, but diffusionthen proceeds so quickly that some 6to 8 times this amount goes throughbefore final sampling. At low pressuresthe membrane holds less than 1/20 ofthe amount analyzed.

Results

Diffusion of air through a membranecharged with water. The rate of simplegas diffusion through a membrane isproportional to the pressure differenceand solubility of the gas and is in-versely proportional to the square rootof the molecular weight. When air dif-fuses through a layer of water, thesteady-state ratio of oxygen to nitro-gen should accordingly be 49.0 per-cent at 240C and 48.7 percent at 26°C.In actual fact, our figures for air dif-fusing ithrough water held in a Milli-pore filter or a dialyzing membranecome out as 56 ± 1 percent at all pres-sures. The reason for this quite sub-stantial deviation from the theoreticalvalue is not known, but it may be thata slight diffusion through the membranesubstance itself and its 30 percent orso of bound water would favor theoxygen. The oxygen-nitrogen ratio wasunchanged by charging the filter witha lampblack suspension or by using theVF grade Millipore filter; both of theseprocedures give a higher filter-to-waterratio. The same figure was also obtainedby diffusing air through blood plasmaand methemoglobin solutions (see Figs.2-4).

In all cases, the total (and partial)amount of the gases diffusing throughwas proportional to the pressure (Fig.5).

Oxygen transport through hemoglobinsolutions. Typical curves for oxygentransport through hemoglobin solutionsare given in Fig. 2. The basic curve( 1/1) was obtained from a hemoglobinsolution with an oxygen capacity of22.9 volumes percent. It will be seenthat the 02/N2 ratio at air pressure of1 atm is 95 percent instead of 56 per-cent as it is in water or plasma. Theratio increases rapidly with loweringof the air pressure, reaching over 400percent at 1/12 atm. But when the rateof nitrogen diffusion is calculated from

10002 P02 IN MM H,N2 --

AIR PRESSURE IN ATM

Fig. 3. Diffusion of air through two he-moglobin solutions of the same concentra-tions (02 capacity, 15.8 volumes percent)but with different pH's. Horizontal line,02/N2 ratio of diffusion through a met-hemoglobin solution; lower diagonal line,rate of oxygen diffusion through water;two parallel curves, relative rates of oxy-gen transport calculated from the 02/N2values; shaded area, water-vapor tension.

587

on

Nov

embe

r 22

, 201

0w

ww

.sci

ence

mag

.org

Dow

nloa

ded

from

the total gas volume and the 02/N2 ratio,we find that it is always proportional tothe pressure (Fig. 5). Two things fol-low: (i) the increased 02/N2 ratio isexclusively caused by an increased rateof oxygen transport, and (ii) the oxy-gen transport is proportional to theproduct of 02/N2 and Ap, where Ap isthe diffusion pressure.

This enables us to calculate the rela-tive rate curves for oxygen transportin Fig. 2. It will be seen that the trans-port rates through hemoglobin in allthree dilutions are parallel to the dif-fusion rate through water-that is,numerically the oxygen transport is theresult of two additive processes, theone, plain diffusion through the water,the other, a specific transport mediatedby the hemoglobin. The latter rate re-mains constant over a large gradientof oxygen pressures. This simple re-lation holds quite precisely for all casesinvestigated, with various concentra-

2.2. P 0 IN MM HgN2

AIR PRESSURE IN ATM

Fig. 4. Diffusion of air through (i) wholeblood (02 capacity, 21 volumes percent)hemolyzed by HA membrane, and (ii)washed cells smeared on the under sideof a dry VF Millipore filter. Horizontalline, 02/N2 ratio through water; lowerdiagonal curve, rate of oxygen diffusionthrough water; parallel curves above this,relative rates of oxygen transport cal-culated from the O2/N2 data; shaded area,water-vapor tension.

588

tions of hemoglobin or diffusingthrough intact red cells (Figs. 2-4).

Figure 3 shows the effect of pH onthe same hemoglobin solution. The onlysignificant difference appears at thevery lowest pressure-namely, 1/12atm where the transport drops offin the acid sample but keeps on un-diminished in the alkaline sample.As would be expected, the hemo-

globin concentration has a marked ef-fect on the oxygen transport. In Fig. 2the decrease in specific transport wasproportionately less than the dilutionof the hemoglobin. At half strengththe rate of transport was reduced to62 percent and at quarter strength, to39 percent. When the concentrationwas doubled there was a marked in-crease in viscosity and a decrease inthe rate of oxygen transport. The rateof nitrogen diffusion through this so-lution also dropped, and this findingimplicates the high viscosity as thecause (2, 6).

The retarding effect of increased vis-cosity was demonstrated by adding 10percent of gelatine to the hemoglobinsolution, which was enough to make itsolidify. This almost halved the specificrate of oxygen transport by the hemo-globin but hardly affected the nitrogendiffusion.Oxygen transport through red cells.

In these experiments the washed redcells were smeared on either the upperor the lower side of the finest Milliporefilter (VF grade). It will be seen fromFig. 4 that the oxygen transport wasalmost doubled at pressure of ½13 atm.When this procedure was tried on aregular HA filter, the capillary forcesin the filter ruptured the blood cellsand the hemoglobin solution camethrough on the other side. This didnot happen on a VF filter, and whenthe filter was soaked afterward in iso-tonic saline there was little, if any,evidence that hemolysis had takenplace. It is therefore indicated that thetransport effect displayed by hemo-globin solutions is also operative inthe intact red cell.

Oxygen transport through myoglobinand other pigments. A preliminarystudy was made on oxygen transportthrough a myoglobin solution. Thiswas prepared from the very dark pec-torals of a California sea lion. Thinslices cut across the fibers were re-peatedly washed in isotonic saline so-lution and wrung in a dry towel. Theamount of blood left after this pro-

cedure is very small compared to thehigh concentration of myoglobin. Theslices were passed through a meatgrinder and repeatedly frozen andthawed, and the macerate was groundunder a few milliliters of water ina mortar and centrifuged. The fluidwas passed through an HA Milliporefilter.

This solution at an air pressure of1/3 atm gave initial 02/N2 ratios of100 percent and 87 percent-that is,there was a substantial oxygen enhance-ment. After a few hours, and contraryto the finding for hemoglobin solutions,the ratio dropped to that of water. Al-though this is an unstable preparation,it demonstrates that myoglobin in vitrois capable of implementing a potentsteady-state oxygen transport.A marine sand-dwelling worm, Thor-

acophelia, when cut into pieces andcentrifuged, yielded a deep red solu-tion with an oxygen capacity of 5.3volumes percent. This mixture of bloodand body liquid yielded large amountsof C02, together with oxygen andnitrogen, but nevertheless showed some12- to 1 5-percent enhancement of theoxygen transport when measured atair pressures of 1/6 or 1/ 12 atm. Inview of the loss of oxygen in the mem-brane, these are clearly minimumfigures.

Hemolyzed red cells from fish blood,-mackerel, with an 02 capacity of 12.1percent and yellowtail (Seriola dor-salis), with a higher 02 capacity-showed an enhancement of 100 per-cent at air pressure of 1/3 atm in themackerel and 110 percent in the yel-lowtail.

Discussion

We have observed a steady-state rateof oxygen transport which can be at-tributed to the added effects of twosimultaneous processes: one, a diffu-sion through the solvent; the other, a

transport specifically dependent uponhemoglobin. The latter proceeds at a

constant rate over a wide range ofpressures and depends upon the kinet-ics and oxygen-binding properties ofthe hemoglobin molecule.It therefore seems that when a tension

gradient of oxygen is imposed upon a

hemoglobin solution, oxygen moleculesare handed down from one hemoglobinmolecule to the next in a chain or"bucket-brigade" fashion. Provided the

SCIENCE, VOL. 131

on

Nov

embe

r 22

, 201

0w

ww

.sci

ence

mag

.org

Dow

nloa

ded

from

"buckets" are emptied at one end andfilled at the other, a steady-state systenmis set up which results in facilitation ofoxygen transport through the chain (seeFig. 6).The steady-state situation requires

that the oxygen tension at any levelbetween the upper and lower surfacesof the membrane remain constant.Above, it is near the tension of the gasphase; below, it is near zero. The oxygensaturation at various levels is deter-mined by the oxygen dissociation curve.but the exact relations are here imma-terial as long as conditions remain con-stant. One may tentatively assume thata linear tension gradient exists and thatthe hemoglobin loses is first oxygen atair pressure of approximately 1/3 atm.its second at 1/6 atm, and its third at1/12 atm. Near the lower surface onlythe fourth molecule remains attachedto the hemoglobin (Fig. 6). This is theone which jumps off into the vacuumand which is then replenished from thehemoglobin chain above. The maximunmrate at which the chain can keep theoxygen moving evidently depends uponthe constant kinetic motion of thehemoglobin molecules, for the rate ofoxygen transport via this route is con-stant and independent of the oxygengradient.

That a rate-limiting factor is locatedat the vacuum end of the chain is indi-cated by the pH effect. It was foundthat an increased affinity for oxygen-that is, a high pH-is able to maintainthe transport unimipaired at very lowpressures, whereas lowering of the pHslowed it down. A glance at the lastcolumn of Fig. 6 will show that at lowpH and a pressure head of 1/6 atm, thelower half of the chain gets increasing-ly disrupted by completely reducedhemoglobin molecules, and we may as-sume that as a consequence of this thedelivery slows down.

If the head of pressure is only I/1 2atm to begin with, one would expect agreat number of void (unoxygenated)molecules in the lower part of the chain,even at high pH; nevertheless, oxygenwas delivered at full rate. This seenis toweaken the explanation of the pH effectjust mentioned, unless possibly anotherfactor entered-namely, back pressureof oxygen. From the lower surface ofthe membrane into the vacuum cham-ber is a long way for oxygen to go,and this simple diffusion depends, ofcourse, upon a concentration gradient.It is possible that the oxygen back pres-26 FEBRUARY 1960

MM N210 MIN.

43 k3 l'6 1/12 0AIR PRESSURE IN ATM

Fig. 5. Rate of nitrogen diffusion throughvarious concentrations of hemoglobin.Curve I / I represents oxygen capacity of22.9 volumes percent; the other curvesrepresent fractions of this amotunt. Ordi-nate = vol/min.

sure built up sufficiently to keep thechain saturated to the very end. Thisraises a fundamental question: If thelow-pressure side were raised deliber-ately, would the transport system stillwork? The answer can be obtainedonly by the use of a different technicalapproach.What then, one may wonder, is the

possible biological significance of thisoxygen-specific, steady-state, enhancedrate of oxygen transport?

It will be seen from the various ex-

ATM. AIR PRESS.

:1,:

2/3

j

ID

pH 81/3

U

rr

l/6

n1)

71/6

0

000

MOIST VACUUM

Fig. 6. Schematic presentation of possiblemode of oxygen transport via hemoglobinmolecules. Solid circles, oxygen mole-cules: open circles, hemoglobin molecules.

periments in vitro that the oxygen trans-port mediated by human hemoglobinmay dominate straight diffusion by afactor of 8 or more at low pressuresand can at full atmospheric pressurealmost double the rate. The effect hasalso been demonstrated in intact redcells, and it is possible, although thishas not been demonstrated, that it playsa role in their normal gas exchange,even when the low-pressure side is notzero.More relevant, however, is the phys-

iological implication of the role ofenhanced rate of oxygen transport inthe case of myoglobin. One may visual-ize a system which in addition to itsstorage function would grease, so tospeak, the oxygen transfer from theblood in the capillaries into the meta-bolic machinery of the muscle fiber.Such a transport would be enhancedeven more if the niyoglobin were mo-bile within the muscle cells, somewhatlike water in a sponge, rather than heldin a viscous medium. But whatever isthe case, muscle contractions wouldundoubtedly favor the transport.

It was found that sea lion myoglobin,indeed, enhances oxygen transport invitro. The increase in hemoglobin andsometimes in myoglobin observed inmany high-altitude mammals fits wellinto this same line of argument, andthe same holds true also for thoselarvae, worms, intestinal parasites, andother animals which live under lowoxygen tension and harbor red hemo-globin-like pigments. A steady-statetransport effect of the blood pigmentfrom one such worm has indeed beendemonstrated.

Summary

A study has been made of steady-state diffusion of air at various pres-sures through hemoglobin solutions.Whereas nitrogen diffused in proportionto the pressure, the rate of oxygentransport was greatly enhanced andseemingly proceeded by means of twoprocesses which are additive. One is aregular diffusion through the solvent(water), which is proportional to thepressure; the other is a specific trans-port mediated by the hemoglobin mole-cules. The rate of the latter is constantover a wide pressure range, and theprocess may at low tensions transportover eight times more oxygen than doesstraight diffusion. Preliniinary studies

589

on

Nov

embe

r 22

, 201

0w

ww

.sci

ence

mag

.org

Dow

nloa

ded

from

have established that myoglobin anda few other pigments in vitro have thesame property.

References and Notes

1. G. A. Millikan, Proc. Roy. Soc. (London)B123, 218 (1937); Physiol. Revs. 19, 503 (1939).

2. I. S. Longmuir and F. J. W. Roughton,J. Physiol. (London) 118, 264 (1952); F. Kreu-zer, Helv. Physiol. et Pharmacol. Acta. 9, 379(1951); - , Ibid. 11, suppl. 9 (1953).

3. This investigation was supported by a researchgrant (No. RG-5979) from the U. S. Depart-

ment of Health, Education, and Welfare, Pub-lic Health Service, and by the Norwegian Re-search Council. I wish to thank A. BairdHastings, Denis L. Fox, and Y. Zotterman fortheir stimulating interest and advice, and Igratefully acknowledge the excellent laboratoryfacilities extended to me by Edmund L. Keeneyand Hastings at the Scripps Clinic and Re-search Foundation. The effect described herewas found in 1956 at the Institute of Zoo-physiology, University of Olso, where valuableassistance was rendered by G. Sundnes and H.Leivestad.

4. P. F. Scholander, L. van Dam, C. L. Claff,J. W. Kanwisher, Biol. Bull. 109, 328 (1955).

5. F. J. W. Roughton and P. F. Scholander,J. Biol. Chem. 148, 541 (1943).

6. A. Klug, F. Kreuzer, and F. J. W. RoughtonLHelv. Physiol. et Pharmacol. Acta 14, 121(1956)] exposed thin films (0.1 by 45 mm) ofreduced hemoglobin solutions to oxygen pres-sure of 0.9 atm and studied optically theoxygenation rate. On comparing their findingswith earlier determinations of nitrogen diffu-sion (diffusion into a system much less sensi-tive to errors from surface disturbance andconvection), they estimated an oxygen en-

hancement of 100 percent at low viscosityvanishing at high hemogIobin concentrations.This effect, attributed to diffusion of oxyhemo-globin, would at a pressure of 0.2 atm amountto 450 percent. This is totally out of the rangeof our findings and the effect is seemingly,therefore, of a different nature.

When an ecologist studies the rela-tions between the species of animals andplants in a given type of environment,he tries to measure, as far as he can,the various external factors, but at thesame time he is aware of the fact thatthe degree or the speed of reaction tothese factors varies individually withinmore or less specific limits for eachspecies. Obviously, the more responsiveindividuals will be either helped orharmed, by any external change, to agreater degree than the less responsiveones, but, by and large, the ecologistcomes to assume a mean response foreach species and then proceeds to treatall the individuals of that species asessentially alike. Similarly, in this at-tempt to survey the present environmentin which we work, I assume an averagezoologist, if there be such a person, andpay attention to the environment inwhich he goes through the motions andactivities inherent in being a zoologist.The old, traditional concept of the

The author is head curator of the departmentof zoology, Smithsonian Institution, Washington,D.C. This article is adapted from his vice presi-dential address, delivered before Section F(Zoological Sciences), during the Chicago meetingof the AAAS, 26-31 Dec. 1959.

590

scholar or the research worker is thatof a relatively solitary individual,plodding along with his studies andonly too glad to be left alone to pursuethem. For many years, for centurieseven, in the early history of science, thescholar was considered to be possiblyquite interesting, but neither a directcontributor to the welfare of mankind,nor actually harmful. It was only whenit began to be realized generally that noknowledge was without potential valuethat society began to pay an increasingamount of attention to the individualsengaged in extending the limits ofknowledge. There is no need to tracethe details of this historical change ofattitude from the time of the Renais-sance, when scholars were still theproteges of powerful and enlightenedprinces and dukes and even of lesser,local VIP's, so we may jump to thesituation as it was in our own experi-ence a generation or two ago, limitingthe picture to the zoological portion ofthe scientific panorama, and comparingthat situation with the present one.The situation as it was then was far

more favorable for the prosecution ofzoological work than it had been at

any previous time. There were literallyhundreds of laboratories in our collegesand universities, not all equally wellequipped or of equal coverage, but stillcapable of training future zoologists andof providing, within very variable limits,the opportunity for research. Most ofour major research museums were inexistence, although their collections werefar less complete than they are now,and library facilities, which have sinceincreased greatly in scope and in com-pleteness, were generally adequate.There were some, but not many, or-ganizations from which research grantsmight be obtained, although most ofthese organizations had established fairlydefinite perimeters to their interests.

Within this setting the hypotheticalaverage zoologist went about his work,and the results, seen in the perspectiveof 30 years, were good. There waslittle need to do anything other thanenlarge the facilities and provide oppor-tunities for an increasing number ofpeople in the zoological field. To judgeby present conditions there was a moreleisurely atmosphere then than now.

Urgency and Research

At the present time we are facedwith a tension in the political atmos-phere that has unfortunately invadedthe scientific environment to a danger-ous degree. We are expected to doresearch, but with an urgency that hasno place in research; we must do thingswithout delay, lest the Russians antici-pate us; we are told that our survivaldepends on our intensity and applicationin scientific work. WVhile this unfor-tunate trend may or may not have somevalidity in matters of missiles and otherdefense developments, it has no placein many areas of science, includingzoology. The preachers of urgency for-get that research is one of the mostennobling forms of human endeavor and

SCIENCE, VOL. 131

Changing Environment ofZoological Research

The era of abundance in funds and transportationopens new vistas for research and exploration.

Herbert Friedmann

on

Nov

embe

r 22

, 201

0w

ww

.sci

ence

mag

.org

Dow

nloa

ded

from