THE EFFECT OF TEMPERATURE ON THE ACID-BASE-PROTEIN EQUILIBRIUM AND ITS INFLUENCE ON THE

CO2 ABSORPTION CURVE OF WHOLE BLOOD,TRUE AND SEPARATED SERUM.

BY WILLIAM C. STADIE, J. HAROLD AUSTIN, AND HOWARD W.ROBINSON.

(From the John Herr Musser Department of Research Medicine, University ofPennsylvania, Philadelphia.)

(Received for publication, October 5, 1925.)

The influence of change of temperature on the carbonic acid-base-protein equilibrium in blood or serum involves three factors:(1) Change in solubility of CO2 in serum or blood. (2) Changein the equilibrium constants of the buffer systems in the blood.(3) Change in the equilibrium constant for the dissociation ofwater.

The fact that the solubility of CO2 changes with temperaturehas naturally been fully appreciated. On the other hand, thephysiological significance of the change in the dissociation constantof water is more or less obscure but has been discussed in a recentpaper by Austin and Cullen (1925). These topics will not beconsidered here.

The relations of temperature to the equilibrium constants ofthe buffer systems of the blood were discussed by Stadie andMartin (1924) whose work was confined to whole blood. Furtherwork from this laboratory by Cullen, Keeler, and Robinson (1925)together with that presented in this paper have enabled us toextend the equilibrium relations as affected by temperature soas to include true serum and separated serum as well as wholeblood. In addition the relations developed lead to a theoreticalexpression for one of the factors in the C correction of Cullenused to convert [pH] of serum at 200C. to pH at 38° which showsthe limitations of the C correction as a constant.

901

by guest on June 18, 2018http://w

ww

.jbc.org/D

ownloaded from

902 CO2 Absorption Curve

Equilibrium Relations of the Buffers of the Blood.

The equilibrium constants of the buffer systems of serum may bewritten as follows:

pH = PK'[I co- + log [BHC011112C0 + [H2CO4I

[BHPO 4I= pK [I2PO] + log [BH 2PO41

(1)[BPr]

= pI + [Pr[Prl

BX= K' + log HX

The portions of the above expression relating to carbonates andphosphates are logarithmic forms of the mass law equation, theformer being the familiar Henderson-Hasselbalch equation. Thelast portion of the expression is a similar expression for any addi-tional hypothetical weakly dissociated acid that may occur inserum under pathological conditions. There is evidence againstthe existence of any such acid in normal serum in significant quan-tity. The expression

[BPr]pH - pi + --P

B[Pr]

is a rearrangement of Van Slyke's buffer equation

[BPr] = [Prl (pH - pI)

(2)

(3)

where [BPrl = milli-equivalents of base bound by protein./ = the buffer constant for the particular serum or blood;

i.e., the change in milli-equivalents of base bound by1 unit of [Pr] for 1 unit ApH.

[Pr] = the concentration of protein for the serum or blood. Forserum protein the unit of [Pr] is 1 gm. For hemoglobinit is 1 millimol.

[pI] = an empirical constant that for a single protein is close tobut not necessarily identical with the pH at the isoelec-tric point.

by guest on June 18, 2018http://w

ww

.jbc.org/D

ownloaded from

Stadie, Austin, and Robinson

This equation (3) was first used by Van Slyke (1922) in his studyof buffers and in his application of his general formula for buffersto the special case of hemoglobin and serum protein. It is usedby Van Slyke, Wu, and McLean (1923) in their equations 54, 55,and 56. We rearrange it merely that it may be more readilycompared with the other expressions in equations (1). Thisbuffer equation of Van Slyke, applicable to any buffer system,happens to be especially convenient for the protein buffer of serumand whole blood over physiological pH range because of the factobserved by Van Slyke and his coworkers that over this range thebase bound by protein [BPr] in serum or in reduced blood isapproximately a linear function of pH and hence [Pr] is a constantover this pH range. The values for pI and 3 are different for cellsand for serum and hence for whole blood.

From the studies of Van Slyke, Hastings, and Neill (1922) andof Stadie and Martin (1924), it is clear that for oxygenatedhemoglobin there is a deviation from the linear relation of [BPr]to pH. However, one can calculate from Fig. 2 of Stadie andMartin that between pH of 7.0 and 8.0 for a blood of normalhemoglobin content the deviation from a mean straight line drawnthrough the true values does not exceed i 0.6 mM [BPr] at anypoint. For the purpose of this paper then the linear relationfor fully oxygenated as well as reduced blood is assumed as anapproximation.

The base in the blood is partly bound by strongly dissociatedacids which at all physiological pH values and temperatures existas salts and bind a constant amount of base. The remainder isdistributed between the various weakly dissociated acids of equa-tions (1). This may be expressed as follows:

[B] = [BA] + BHCOa] + [BPr] + ([BH2PO4] + [B2HPO4] + [BX]) (4)

where A includes the anion of all strongly dissociated acids,chiefly C1', and where the part in parenthesis is negligibly smallin normal blood and serum as compared with the [BHCOa] and[BPr]. Under these conditions base bound as [BHCO 3] +[BPr] may be considered a constant which we will call [b] for agiven sample of blood or separated serum regardless of changeswithin physiological limits of [02], [C02 ], pH, or t °. This isindicated diagramatically in Fjg. 1 where the distribution of base

903

by guest on June 18, 2018http://w

ww

.jbc.org/D

ownloaded from

CO2 Absorption Curve

between Pr' and HC0 3', as it is equilibrated with varying tensionsof CO2 with consequently varying pH, is shown for the same sep-arated serum at 200 and 38 ° .

Fig. 1 emphasizes the fact that the slope and location of theCO2 absorption curve of a serum (and the same is true of whole

Oq.Sq.

p,

FIG. 1. Showing the relation of CO2 absorption curve of a separatedserum expressed as IBHCO3 j against pH to the curve of [BPr] against pH.[BHCOsl3] + [BPr] = [b], a constant (approximately) for any given separatedserum.

blood) is determined essentially by the constant value for [BHCO3]+ [BPr] and by the value of [BPr] at any pH. The location andslope of the C02 absorption curve either of whole blood or of sep-arated serum is essentially determined by the location and slopeof the curve for [BPr]. In other words the carbon dioxide dis-sociation curve of separated serum is determined by the concen-

904

by guest on June 18, 2018http://w

ww

.jbc.org/D

ownloaded from

Stadie, Austin, and Robinson

tration of total CO2, the concentration of protein, and its constantspI and B. The relation, therefore,

A[BHCOs] = - A[BPrJ (5)

under the limitations stated is necessarily true.For true serum, subjected to changes in Pco,, pH, or tem-

perature while still in contact with the cells, that is as a phase ofwhole blood, the relationship between [BHCO3] and [BPr] isless simple. In true serum under these conditions with changesin [BPr] of cells and serum there is migration of all anions includingthe quantitatively important anion Cl' through the cell membranein amounts determined as shown by Van Slyke, Wu, and McLeanby the total amount of each anion in the blood and the requirementsof the Donnan equilibrium. Under these conditions [BA]TSis not constant and A(BHCOs)Ts does not equal - A [BPr]TsIf we neglect changes in the relative volume of cells and serum andchanges in the Donnan ratio of anions in cells and serum which areknown to be small over narrow pH ranges, the ratio of concentra-tion of [BHCO3]Ts to [BHCO3]c and therefore to [BHC03] Bwill be constant, and as an approximation we may write

A[BHCOS]T = [BHCO]B = -[BPr] B (6)

This question will be discussed at greater length below. There-fore, while changes in [BHC0 3] of separated serum are a measureof changes in [BPr] of separated serum, changes in [BHCO3 ]of true serum at 380 C. and of whole blood are both a measure ofchanges in [BPr] of whole blood. The quantitatively importantchange in [BPr] of whole blood is the change in base bound byhemoglobin. The slope and location of the CO2 absorption curveis determined therefore by equations (3) and (4) when the factorsin the parenthesis of (4) are quantitatively negligible so thatequation (5) holds. Since

pH = pI + [BPr (2)and[Pr

and if

[BHCO] + [BPr] = [b]

905

(7)

by guest on June 18, 2018http://w

ww

.jbc.org/D

ownloaded from

CO2 Absorption Curve

where [b] is a constant for the particular separated serum or bloodwe can write

pH = pI + [b]-[BHCO] (8)/[Pr]

Equation (8) defines the location of the C02 absorption curveexpressed as [BHCO3] against pH when the constants pI, [b] andO[Pr] are known. Conversely the curve of [BHCO3] against pHdetermines [Pr]; for

pH - pH' = pI [b] - [BHCO,3 (P + [b] - [BHCO ') (pH - pH' = pI + + ~[~-~ ] (#[Pr] 811r]

or

[BHCO,]' - [BHCO3 ] (10)/~[Pr] = (10)

pH - pH'

Further, if fB[Pr] is constant at all temperatures, then it willbe obvious from equation (2) that if we select points on the curvesat two temperatures having the same value for [BHCO3] it willbe true that:

pH t - pH t, = pI t - pI t, (11)

and we can evaluate the change of pI with temperature.The equation developed above may be used to evaluate experi-

mental data relating temperature and C02 absorption curves andin the present paper we present additional data on whole blood,true serum, and separated serum of dog and sheep at 20°C. and380 C.

Methods.

Blood was drawn from heart of the dog, or jugular vein of thesheep under oil or from the rat by instant decapitation. Theblood was defibrinated by stirring under oil with a glass rod.

Equilibration of blood or serum was performed by the first orsecond saturation method of Austin, Cullen, Hastings, McLean,Peters, and Van Slyke (1922) and the technique there describedfor separation of true serum was employed.

Electrometric pH was done by the method and with the stand-ardization described by Cullen, Keeler, and Robinson (1925).

906

by guest on June 18, 2018http://w

ww

.jbc.org/D

ownloaded from

Stadle, Austin, and Robinson

Colorimetric pH at 20° was read by the technique of Cullen(1922) and at 38° by that of Hastings and Sendroy (1924) usingfor both the latters' bicolor standards. Colorimetric readingsare expressed with bracketed figures, [pH]20 indicating the colorgiven at 20° by the phosphate buffer of corresponding pH at 200and [pH]38 indicating the color given at 380 by the phosphatebuffer of corresponding pH at 380. Unbracketed pH figuresindicate gasometric or electrometric pH. [CO2] analyses wereperformed by Van Slyke's method. Gasometric pH was cal-culated by the Henderson-Hasselbalch equation from data onseparated serum or true serum using Bohr's a (0.975 a water)and pK'38 = 6.10, pK'20 = 6.19 (Cullen, Keeler, and Robinson)

DISCUSSION.

Our experimental data are shown in Table I. From these data[BHCO3] was plotted against pH. The resulting slopes are given

A[BHCO,]in Table I in the columns headed zH for each true serum,

ApHwhole blood, and separated serum.

Effect of Temperature on fl[Pr].

The effect of temperature on the slope of the CO2 absorptioncurve of whole blood, i.e. upon the value of [Pr]B, is not com-pletely settled by the data of Table I. Two of the four experi-ments show a greater slope at 38°C. than at 200 C., one the reverseoutside the limits of experimental error. One gives within ex-perimental limits the same slope. Such evidence neither provesnor disproves the assumption of Stadie and Martin that [Pr]does not vary with temperature and furnishes no evidence forabandoning for the present the assumption. We tentativelyassume, therefore, that at both 200 and 380

[BHCOa] B = - A[BPr]B (12)

total available base [b]B of blood being constant.It will be seen from Table I that both at 20°C. and 38°C. the

slope [BHCO] is constantly steeper for true serum than for thecorresponding blo e ratio of the slopes is 1.13 0.04 withcorresponding blood. The ratio of the slopes is 1.13 ± 0.04 with

907

by guest on June 18, 2018http://w

ww

.jbc.org/D

ownloaded from

CO2 Absorption Curve

Nq Co Cl t- co Co 00- -! 1 0 0 0 '-4

- - -4 4 -4 -

_~C _ lO O O

_ _ __ _ _

cni 00 ll4 -i ci 0 -4 k-to 00 01 m 0 01 xf) C'-01 01 01 0 014 01 -4

Ni t4 0 C 0 X 000 C I- ka - C 4 0N N N N N I Cl

5-4 d LC *o_ -- -z

~00*0t'-t-C0000O0OI4 4 C O C o o O k Co la C o

cl Cl c N N N _- - - _ _

..

0 Co 0 00 O N- 0 C0 C 0 CD 0~ COCO~oCIOC-0Coto1C! t--~ll Clt.-

to m X0 c' _ 0 j CD _ CD X N co t O- o

C Co, O O 4C to CO cO to oCO C 6 C O0

0 0 0 0 o C0 - 0 ) N to ) I- N 00 0S t 0 N t 0 N ot X o t000' 1004 O E oo - _ O N O~ N O _ 'Id! N ONO

> to N _ CO O d *, l c N 0 0 cl N

ao3h'jj OO tCdddouo cir-cq

t-

S f I_ _ 10 0 bco CD l 10 o C0 C CO 1O12N N N N CO N 14_N_ N N N _ N Nq N N~ N~

00 N 00 O0 c O _ t t o C N 1 N - c 'lv ~ c 0co 0 0 "d 0 2 Co o '0t 00 00 C6

mO CO L- m O O - c O O N N O-O OO

C 00 00 000 00 000 0000 00 0 o000_ m C4 C 4 m N N mC CI CO N CID Cm N N co

0 _ _ 0 _ __ O_ 0 C3 ___

'04)

a a C4 a Z)o o o o ~3--O N O +o

C 0

4..4Ca N N411=

Co CO CoNl C'D CO

b0 0 0O O O

c~l

0 0CoA

908

I -;;cii

o

a"

-

00

m

De

1v

*0

m

*00

COd q;3

n:*0

A "

(6

. d

Aaw

0

U20CO

PqS

`-----

. o

by guest on June 18, 2018http://w

ww

.jbc.org/D

ownloaded from

Stadie, Austin, and Robinson 909

no striking difference in the dog between the slopes at the twotemperatures. Equation (12) as an approximation holds fortrue serum but could be more accurately written

A[BHCOslTS = k lBHCOs]B = - k A[BPr]B (13)

We may also write at constant pH but changing temperature

t - t [BHCO,]TS = k (t-t' A[BHCO31B) = -k (t- _ t' A[BPr]B)(14)

At present we may put k = 1.1.That the slope of the true serum curve is greater than that for

whole blood is to be expected from the following considerations.For whole blood, if v is the volume of the cells we have

(1 - v) [BHCO]Ts + v[BHCO3]C = [BHCO3]B (15)

Assuming v constant a similar expression could be written for ahigher CO2 tension so that for a small increase we may write

(I - v) A[BHCOsITs + vA[BHCOa]C = A[BHCOs] B (16)

or

A[BHCO3]TS = A[BHCO]IB + v(A[BHCOs]Ts - A[BHCO] C) (17)

If the concentrations of BHCO3 be referred to the water contentof cells and serum denoted by w and w' respectively, the Donnanratio can be written

[BHCO,]c w'· -- = r (18)

[BHCO3]TS w

or for a small change at constant r, w and w'

A[BHCO31s = wr a[BHCOs] c (19)

The data on pp. 844 and 845 of the paper by Van Slyke, Wu,WI

and McLean (1923) show for horse blood that -- is always greaterwr

than 1.0 between pH 7.0 and 7.7. It follows then that(A[BHCO8]Ts - A [BHCO 3]c) is finite and positive and equation(17) becomes

A[BHCOO]TS = A[BHCOs]B + x[BHCO3] where x is positive; (20)

by guest on June 18, 2018http://w

ww

.jbc.org/D

ownloaded from

CO2 Absorption Curve

or as an approximation we may write

A[BHCOa]TS = k A[BHCO,B = -k A[BPr]B (21)

where k = 1.1 approximately.It would, of course, be possible to pursue this matter further

and elaborate a definite exact theoretical expression relating[BHCOS]Ts to [BHCO3]B in terms of protein concentration,buffer slope, water content, pH, r, and total anion concentrationof serum and cells respectively. This would involve merely arearrangement of the equations derived by Van Slyke, Wu, andMcLean (1923). However, the establishment of the validity ofthis relation would be nothing more than an experimental con-firmation of their hypotheses. Theoretically this would be noadvance and practically for each blood the relation would bedifferent because of variation in the constants. Peters, Bulger,and Eisenman's empirical solution of this problem is eminentlysatisfactory for practically all purposes.

In twenty-one determinations of the CO2 curve of whole bloodand true plasma on human blood at 38°C., Peters, Bulger, andEisenman (1924) found a greater slope for plasma in nineteencases. The ratio of the slopes in these nineteen cases was 1.08 inclose agreement with the value found here.

The approximate position of the C02 absorption curve of trueserum relative to the whole blood curve may be located if thevalue of [BHCO 3]Ts at one point is known by drawing a linethrough this point on a [BHCO3 ] - pH plot having a slope 1.1of the slope of the whole blood. Lacking the one point on thetrue serum curve the location of this line is more hazardous butmay be approximated by the method outlined by Peters, Bulger,and Eisenman (1924). Employing the principles outlined herethe true serum curve located at any one temperature may be ap-proximately located at other temperatures.

For separated serum the one experiment shows within the limitsof experimental error the same slope at 20° and 380C.; i.e., [Pr]ss= constant, so that we may write at any temperature

A[BHCOs]ss = - A[BPr]s s (22)

total available base [b]ss of separated serum being constant.

910

by guest on June 18, 2018http://w

ww

.jbc.org/D

ownloaded from

Stadie, Austin, and Robinson

From the data of Table I certain approximations as to the effectof temperature upon the acid-base equilibrium may be justifiablydrawn. Assuming as an approximation a linear relationship of{BHCO3] against pH at both temperatures, we calculate from thedata of Table I the changes with temperature of [BHCOJ atconstant pH, at constant [H 2CO 3], and at constant Pco, and alsothe changes of pH at constant [BHCOs], [H2C0 31, and Pco2-These relations are tabulated in Table II.

It will be seen that a rather constant temperature effect is the20o - 38o ApH at constant [BHC03]. This we have pointed outabove, if the slope of [BHC03] against pH be the same at the twotemperatures, is a measure of 20 - 38. ApI of whole blood when theApH measured is that of true serum and of 200 - 38 ApI of separatedserum when the ApH measured is that of separated serum. Thedata suggest that the 200 - 38oApI is for whole blood and separatedserum approximately:

A pI ApH [BHCO3 constant-A p = --0.017 to -0.023 (23)At, A t

If it is desired to approximate the position of a [BHCO3] curveof a serum or blood against pH at some other temperature (t')when the curve at t is known, it can be done conveniently by meansof equation (23) and the assumption that the slope of the [BHCO 3]curve against pH at t' is the same as at t.

In order to aid in visualizing the nature of these changes nomo-grams (Figs. 2 and 3) have been prepared of the relations in thetrue serum of Dog 33, June 4, and in the separated serum of Dog27, January 14. It will be seen that while the interval betweenthe lines is much greater in separated serum due to its smaller,B[Pr] value and consequently flatter C02 absorption curve, theslope of the lines in physiological ranges is almost the same fortrue serum and for separated serum. These nomograms alsobring out the fact, pointed out by Austin and Cullen (1925) thatwith change in temperature from 38° to 20 ° at constant Pco 2 thepH of serum varies very little. They pointed out, however, thatthere is a marked associated change in pOH.

911

by guest on June 18, 2018http://w

ww

.jbc.org/D

ownloaded from

CO2 Absorption Curve

'aW V

~0-0

Z 2m 0.11

'3D. ad|HdV-I

'0o .ad['ODH]V--

wo 'D.o Jeda HdV-

'I0-

000o

0

000~

0

0

E 5 c

-

- 00_~ 1>00 00 0

04

C 00 CO

CO 0

1-4=

0 ~ E 0o o 0 0o

o o o o

= 00 0

5 a0 0 0l

tO t°3dl S' X 00 O C-m 0. I - "" I aQ 30 id 0 04-1

o.9 [eo3Hg]V- 0 5 0 02 2 Hd | t_ to | co i

CO CO COma8~ H C- C- C-r:r;~

43 e 'aOlaA-

00a lad HdV-

Hd

CEi

I0

- - _

- 0 04 - 0 04 00o o 0 0C 0 0 0

0 0 0 0

.O .0 ": . 0 .40

kO 00 co v - 00 00o Co c o

44,s.- ~ ~ ~ ~ ~ C __

-4I-eDQ

Cl04

c0cq

4-

bCO

0a)

912

'

-^ -^ -

�

- -

by guest on June 18, 2018http://w

ww

.jbc.org/D

ownloaded from

Stadie, Austin, and Robinson

00 0 o 0

oI °C! CD 9 C O l

- 0 C

o 0 1 O C

00 00 I 00~ ~ ~~~~~

. * S

CO GO 0o 00

Co OC Ci

o o o o

Co Co co

Co Io Co o

o o 0 0

oo 0 0 0 0

- ho

- ~~~~~~~C

o U o 0o 0

C o Co ho 0) C- 0 Co00 C- 00 NC - ) = 00

- t- - -

_1 ~~~~~t0

o m 0 0 0 O 0 ) 0 0 0o

o 0 0O0 0 0 5 5

O O

O O, - ' e

OCI COO CI Co COC Co CZ CoO

CII00 00 hoohh 00 hOh ho

O O O O O O O

O~ C O$ O\ O O) O O O

_s M_ t_ Ut O 0 0 D CQ0 lN D C

A C O O OO C oo h 0 O ) C o 1o

i-1 ee -- N Ce Cqe Ceq - CIl

1-0 1_0 1-0 C-O IO C O eqO eOc0 o oo o 000 00 0o o 00 0 00

iCO _ICO 'Co Co Cl CO N CI CICO CICO

* w

) Oo- o 7~ 0~

P? ic ~~~~~ t-e~C

0)~~~

co

o

C;I a)GOm/

913

c o

CO

C!

c0o4

0 0

Oho COO

CN

0) K

00 oo

cq

0q

by guest on June 18, 2018http://w

ww

.jbc.org/D

ownloaded from

914 CO2 Absorption Curve

FIG. 2. Nomogram of true serum Dog 33 (June 4) showing relations of[BHC031s, PCO2, [H2 C03]6 , pH, and t°C.

pri

FIG. 3. Nomogram of separated serum Dog 27 (Jan. 14) showing relationof [BHCOs], Po2, [H2C03]s, pH, and t°C.

by guest on June 18, 2018http://w

ww

.jbc.org/D

ownloaded from

Stadie, Austin, and Robinson 915

Change of pH of Serum with Temperature at Constant Total C02 andConstant Total Available Base: Calculation of t - tvApH.

Separated serum equilibrated under known conditions at agiven temperature, t, or true serum separated from cells underequilibrium conditions and kept so that no loss or gain of CO2may occur (under oil or in mercury receivers) will on change oftemperature assume a new equilibrium with consequent changeof pH. The same thing is necessarily true for diluted serum.This change of pH enters into the C correction of Cullen's (1922)colorimetric method for pH of serum and plasma since the con-ditions defined are exactly those employed by Cullen in readingat 20°C. the [pH] of a serum equilibrated at 38°C. Austin, Stadie,and Robinson (1925) have shown elsewhere, however, that thispH change is not necessarily the most important cause of variationin the C correction. Disregarding this for the moment we willproceed to show the factors involved in the change of pH of serumwith change of temperature from 380 C. to 20°C.; i.e., t - tApHand its possible variations. If we accept the evidence that withchange in temperature the equilibrium of serum protein-carbonicacid systems is determined by change in pI of equation (3) and ofpK' and a of the Henderson-Hasselbalch equation, it is possibleto calculate the t - tApH under the above conditions as follows.

The amount of base in equivalents bound by serum proteinat t is given by the linear relation, equation (3)

[BPr] = B[Pr] (pHt - p t ) (3)

where [Pr] = protein concentration gm. per literB = protein buffer change in equivalents of base bound per gm.

of protein per 1 unit ApH.PIt = pH close to but not necessarily identical with the isoelectric

point at t°C.

The equivalents of base bound by HCO, at t are given by thefamiliar equation

[BHco] [CO +21 (24)k' + [H+]t

At equilibrium the sum of the base [BP] + [BHCO3] is [b]

Ib] = #[Pr] (pHt - pIt) + k- [CO2] (25)kt + [H+]t

by guest on June 18, 2018http://w

ww

.jbc.org/D

ownloaded from

CO2 Absorption Curve

At t'°C. if no loss of CO2 or change of protein concentrationoccurs

[b] = /[Pr] (pHt , - Pi t') + k +, [CO2] (26)kt , [H+]t,

Equating, and assuming the total available base [BP] - [BHCO8 ]which we call [b] unchanged, we have

k:. k C(- ' ApH - t p I ) [Pr] = (k. + [H+]t, k[ + [H+]t) [CO] (27)

Equation (27) shows that t tpH in serum (and likewise indiluted serum) at constant [CO2] and [b] is dependent on therelation of the total [CO 2] to [Pr], the magnitude of ApI withtemperature, the change in k' (the dissociation of carbonic acid)with temperature and also upon the initial value of [H+] or pH.

The relation of k't and k't, of equation (27) to the Ato for the

Henderson-Hasselbalch equation is as follows:

kt'log - = PKt.- pIK = 0.09 (28)

between 38°C. and 20°C.Evidence that 3 for hemoglobin is independent of temperature

was presented by Stadie and Martin and as already pointed outour data furnish no evidence for change of [Pr] of serum proteinor whole blood with temperature. A graphical solution of equation(27) yields the graphs shown in Fig. 4, where abscissae are initialpH3 8, ordinates are t - tApH = pH2 0 - pH38. It is at onceevident that t - tApH is not a constant. Variations in initialpH from 7.0 to 8.0 cause variations as large as 0.16 in t - t,ApH.Changes in protein buffer of serum well within the limits of varia-tions observed in different species cause considerable change in thevalues of t - tApH as well as in the importance of change in initialpH. Thus for a serum with [CO2] = 20 m, initial pH = 7.4and /[Pr] = 6, we find 20 - 8ApH = 0.34. A change in total[CO2] of 10 mM causes change in 20 - 38ApH (at same initial pH)of 0.04 to 0.05 pH.

It should be pointed out that here over narrow physiological

916

by guest on June 18, 2018http://w

ww

.jbc.org/D

ownloaded from

Stadie, Austin, and Robinson

ranges, 7.30 to 7.50, the changes in 20 - 38ApH are small (0.04 to0.05 pH). However, the use of a constant value for Cullen's Ccorrection (which involves 20 - sApH) in experiments on separatedserum saturated at different CO2 tensions is associated with greatervariations of t - tApH because lower initial pH is associated withhigher [CO2 ]. From Fig. 4 it will be seen that lowering initialpHs8 and increasing [CO2 ] both tend to diminish 20 _ssapH; wehave here, therefore, a summation of effects. For true serum theA [CO2] at different initial pHt values is still greater with greateralteration in 20 - sApH at different initial pH values.

The use of a constant Cullen's C correction therefore withinphysiological ranges introduces the possibility of errors due to

20a-35 s 1-' = 0.4 20-3' - -

1SEPr= 2.0 13 [Pr]= 4.0

FIG. 4. Graphs showing 2o0-3APH when separated serum undergoeschange of temperature from 38 ° to 200C. at constant total [CO2]s fromvarious initial pH38 values. Graphs are drawn for three different [CO2]svalues and for two values for #[Pr].

variation in t - tApH of from 0.06 to 0.09 pH units between 200and 38°C. In experimental work on blood or'serum over widerpH ranges still larger errors would be introduced as pointed outabove.

Equation (27) must be valid, as already pointed out, if equation(3) is valid and if change in temperature affects only the constantspK' and a of the Henderson-Hasselbalch equation and pI whileO[Pr] remains constant independent of temperature.

However, since the special interest of this aspect of the paperlies in the 20 _ S8ApH of serum under conditions of constant total[CO2] and [b], experiments were done to test equation (27) underthese special conditions. The data are given in Table III.

The experiments are on both separated and true serum and

917

by guest on June 18, 2018http://w

ww

.jbc.org/D

ownloaded from

918 CO2 Absorption Curve

divide themselves into two groups, Nos. 1 to 12 and Nos. 13 to19. The sample of serum was equilibrated at 380 and its gaso-metric pH38 calculated from the Henderson-Hasselbalch equation.

TABLE III.Effect of Temperature on pH of Serum at Constant [C02].

Source.

Dog 27.

Sheep 1.

Same with addedNaHCO3

Dog 33, true serum.

Sheep 1, true serum.

Sheep 1.

Rat pooled serum.

Dog 27.

* Assumed.

0

00

123

456

78

910

11

1314

1516

171819

ICO]I

mM

27.024.522.4

23.221.619.4

46.240.6

25.519.9

25.3I'd.

3124

2624

251917

#[Pr]

mM

4.04.04.0

2.32.32.3

2.32.3

4.0*4.0*

2.3*" o

pH

20°

7.327.657.80

7.177.577.90

7.227.94

7.287.54

7.307.57

[pH]

[7.22]> [7.95]

[7.37][7.67]

[7.22][7.61][7.87]

380

7.087.407.58

6.987.307.60

7.087.73

7.087.27

7.067.26[pH]

[7.101[7.75]

[7.21][7.47]

[7.05][7.40][7.63]

2o-4SApH

Ob-served.

0.240.250.22

0.190.270.30

0.140.21

0.200.27

0.240.31

0.12>0.20

0.1 0.20

0.170.210.24

Calcu-lated.

0.220.290.33

0.150.230.32

0.100.21

0.220.29

0.160.24

Differ-ence ob-served-

calcu-lated.

+0.02-0.05-0.11

+0.04+0.04+0.02

+0.040.00

-0.02-0.02

+0.08+0.07

The color reading [pH]2o at 20°C. of each sample was determinedcolorimetrically. The gasometric pH20 and the correction([pH]2o - pH2o) were determined independently on samples of thesame serum by equilibration at 20°C. over a range of pH. These

ll� l- - --- �-

Y 2.;j1Y

by guest on June 18, 2018http://w

ww

.jbc.org/D

ownloaded from

Stadie, Austin, and Robinson 919

([pH]2o - pH20) corrections were used to correct the other 20 °

colorimetric readings on the same serum equilibrated at 38 sothat the values given as pH2o in Table III are essentially gaso-metric pH2o values. [Pr] was determined from the slope of theserum CO2 absorption curves. The "calculated 2o _ 3 8ApH" valuesare obtained from Fig. 4.

In the second group of observations (experiments 13 to 19) no([pH] 20 - pH2o) corrections were determined so that' only thecolorimetric readings of the diluted serum at 200 and 38°C. aregiven. Direct comparison therefore cannot be made with thecalculated 20 - 38ApH (see Austin, Stadie, and Robinson, 1925).The experimental data with the exception of No. 3 are in completequalitative as well as quantitative agreement with the calculated

.values. t - t,ApH as observed is not constant but regularly(except No. 3) increases with pH. Increasing total CO2 (Nos.4 to 8) decreases 20 _ 3sApH roughly about 0.04 per 10 mM increaseof CO2 per liter. The calculated t - tApH is in agreement with theobserved for most of the observations. In Nos. 11 and 12 thedifference is outside the experimental error but calculated andobserved - tApIH vary with change in [CO2] and with initialpH38 in the same direction and to the same extent. No reasonsfor the discrepancies in Nos. 3, 11, and 12 are known.

SUMMARY.

1. The dependence of the CO2 absorption curve of serum orblood upon the base bound by protein at varying pH is discussedand the relation stated in equations (8) and (10).

2. The change in CO2 absorption curve between 380 and 20°C.is determined experimentally for whole blood, true serum, andseparated serum in dog and sheep.

3. Our data furnish no evidence that the value of fB[Pr] is alteredconsistently by this change of temperature.

4. On the assumption that [Pr] is the same at both tempera-tures one can evaluate the change of pI of equations (3) or (8)with temperature.

5. It is found that approximately for whole blood, true serum,and separated serum

A pI A pH [BHCO] constant - -0.017 to -0.023.At 0 At 0

by guest on June 18, 2018http://w

ww

.jbc.org/D

ownloaded from

920 CO2 Absorption Curve

6. When it is desired to approximate the position of a [BHCOs]curve against pH for either separated serum, true serum, or wholeblood at another temperature, t', when the curve at t is known, itcan best be done by means of the equation in conclusion 5 andthe assumption that the slope of the [BHCO3 ] curve against pHat t' is the same as at t.

7. The slope of the true serum curve is found to be higher thanthat of the corresponding blood both at 20°C. and 38°C. Thereason for this is discussed and an empirical relation

A [BHCO]TS = 1.1 A [BHCOS]B

given relating the two slopes at any temperature.8. The change in pH of a separated serum or whole blood under-

going change in temperature at constant [CO2] is calculated and-the influence of initial pH, initial [CO2], and initial value forB[Pr] is shown.

9. This change in pH (which enters into the C correction ofCullen's colorimetric method) may vary 0.09 pH units withinphysiological ranges with change in temperature from 20 ° to 38° ,and is a factor therefore in the variation of the C correction.

BIBLIOGRAPHY.

Austin, J. H., Cullen, G. E., Hastings, A. B., McLean, F. C., Peters, J. P.,and Van Slyke, D. D., J. Biol. Chem., 1922, liv, 121.

Austin, J. H., and Cullen, G. E., Medicine, 1925, iv, 275.Austin, J. H., Stadie, W. C., and Robinson, H. W., J. Biol. Chem., 1925,

lxvi, 505.Cullen, G. E., J. Biol. Chem., 1922, lii, 501.Cullen, G. E., Keeler, H. R., and Robinson, H. W., J. Biol. Chem., 1925,

lxvi, 301.Hastings, A. B., and Sendroy, J., Jr., J. Biol. Chem., 1924, lxi, 695.Peters, J. P., Bulger, H. A., and Eisenman, A. J., J. Biol. Chem., 1923-

24, lviii, 773.Stadie, W. C., and Martin, K. A., J. Biol. Chem., 1924, Ix, 191.Van Slyke, D. D., J. Biol. Chem., 1922, lii, 525.Van Slyke, D. D., Hastings, A. B., and Neill, J. M., J. Biol. Chem., 1922,

liv, 507.Van Slyke, D. D., Wu, H., and McLean, F. C., J. Biol. Chem., 1923, lvi, 765.

by guest on June 18, 2018http://w

ww

.jbc.org/D

ownloaded from

RobinsonWilliam C. Stadie, J. Harold Austin and Howard W.

SEPARATED SERUMCURVE OF WHOLE BLOOD, TRUE AND

ABSORPTION2INFLUENCE ON THE COITSACID-BASE-PROTEIN EQUILIBRIUM AND

THE EFFECT OF TEMPERATURE ON THE

1925, 66:901-920.J. Biol. Chem. 

  http://www.jbc.org/content/66/2/901.citation

Access the most updated version of this article at

 Alerts:

  When a correction for this article is posted• 

When this article is cited• 

to choose from all of JBC's e-mail alertsClick here

  t-1

http://www.jbc.org/content/66/2/901.citation.full.html#ref-lisfree atThis article cites 0 references, 0 of which can be accessed

by guest on June 18, 2018http://w

ww

.jbc.org/D

ownloaded from