AMERICAN JOURNAL OF PHYSIOLOGY Vol. 228, No. 4, April 1975. Printed in U.S.A.

Distribution of H+ and HC03- between CSF and blood during metabolic alkalosis in dogs E. G. PAVLIN AND T. F. HORNBEIN (With Anesthesia Research Center, Department of Anesthesiology University of Washington, Seattle, Washington 98195 PAVLIN, E. G., AND T. F. HORNBEIN. Distribution of H+ and HCOabetween CSF and blood during metabolic alkalosis in dogs. Am. J. Physiol. 228(4) : 1141-l 144. 1975.-In anesthetized, paralyzed dogs ventilated to maintain a normal Pace,, metabolic alkalosis was induced and held constant over 6 h by infusion of sodium bicarbonate. Determination of pH, Pco~, [HCOa-1, and [lactate] in cisternal and lumbar cerebrospinal fluid (CSF) and in arterial plasma together with measurement of the CSF/ plasma DC potential differences permitted calculation of the electrochemical potential difference (II) for H+ and HCO3-; measurements were made prior to induction of metabolic alkalosis at pH, = 7.40, as soon after induction as stable arterial values were achieved and 3, 4.5, and 6 h thereafter. A steady state for ion distribution was reached by 4.5 h. Values of 1-1 for H+ and HC03returned to +O.l and +0.9 mV of control at 6 h for cisternal CSF and +0.6 and -0.4 mV for lumbar CSF. This return of pn+ and j&co3close to control in the steady state is compatible with passive distribution of these ions between brain extracellular fluid and blood. cisternal passive

and lumbar distribution;

CSF; acid-base CSF/plasma DC

balance; potential

ion

regulation;

PAPER REPORTS OBSERVATIONS on changes in the acidbase state of cisternal and lumbar cerebrospinal fluid (CSF) in dogs during sustained isocapnic metabolic alkalosis. The purpose of the study was to determine whether the distribution of H+ and HC03 between brain extracellular fluid and blood was regulated by a process of active ion transport or by passive distribution. The format of the study was like that used to assess the effects of metabolic acidosis on CSF acid-base balance (7). In that study, the electrochemical potential difference (p) between CSF and blood for H+ and HCO, returned to control by 6 h, suggesting that the distribution of these ions was passive (7). In the present study metabolic alkalosis was induced and maintained by intravenous infusion of sodium bicarbonate while arterial PCO~ was held constant. Values for cisternal and lumbar pH+ and pnco3- were determined periodically from concentrations of these ions and the measured DC potential difference (E) between CSF and blood. THIS

METHODS

Six studied

anesthetized, mechanically utilizing the methodology

ventilated and attention

dogs were to controls

the Technical and Department

Assistance of Carol Stowers) of Physiology; and ~~o~hysks,

detailed in the previous paper (7). At the end of a control period of 1.5 to 2 h at pH, = 7.40, ion concentrations in cisternal and lumbar CSF and blood were determined along with CSF/plasma E at the two sites. While Paooz was kept constant (Fig. 1), 0.9 N sodium bicarbonate was infused at rates that would quickly increase and then maintain pH, at about 7.52 over the next 6 h. Measurements were repeated approximately 1 h after starting bicarbonate infusion (time 0) and 3, 4.5, and 6 h thereafter. Electrochemical potential differences, pH+ and ~HeO~-, were calculated as previously described (7). RESULTS

By infusion of sodium bicarbonate, arterial pH was changed from 7.40 to 7.53 while PaeOz was held constant at 35 torr; plasma [HCO& changed from 23.3 to 32.4 meq/kg Hz0 (Table 1, Fig. 1). The CSF/plasma AE/ApH, of -41.1 mV for cisternal and -30.6 mV for lumbar locations (Fig. 2) at time 0 compares with values of -40.1 and -37.2 mV observed during metabolic acidosis (7) and with Held et al.‘s (5) cisternal value of -42.3 mV. With pH, held constant the values for cisternal and lumbar CSF/plasma E at 6 h were within 0.2 mV of the values observed at time 0. While arterial PCO2 remained constant, the PCO~ of both cisternal and lumbar CSF rose 4-5 torr over 6 h (Fig. 1). A g ra d ua 1 widening of the CSF-arterial PCO~ difference over time was noted during all four acid-base abnormalities as well as during control studies (6-8). In the present study, we estimate this change to represent a decrease in blood flow relative to metabolism of 37 % for the brainstem and 23 % for the lumbar spinal cord. These decreases in flow at 6 h are the largest observed during this series of studies. One wonders to what extent progressive alkalosis of CSF (Fig. 1) might have added to the tendency for flow to decrease with time (7). As discussed subsequently (6), the increase in CSF [lactate] (Table 1) may in part be explained by these changes in flow and CSF pH. The rise in cisternal and lumbar [HCOa-] to a new steady state by 4.5-6 h was associated with a rise in CSF pH (Fig. 1). The magnitude of rise in CSF pH was 21 % of the arterial change at the cistern and 33 70 for lumbar CSF. Changes were presumably attenuated by the rise in CSF Pco2 (Fig. 1). In spite of the slow, progressive rise of CSF Pco~, values of p for H+ and HC03 at both cisternal and lumbar

1141 Downloaded from www.physiology.org/journal/ajplegacy by ${individualUser.givenNames} ${individualUser.surname} (129.186.138.035) on January 12, 2019.

1142

E. G. PAVLIN

f

\

.. .

..

..

T. F. HORNBEIN

Cisternal Lumbar

AE / NH,

-f...

..

AND

= -41.8 = - 33.4

mV/pH mV/pH

Lumbar

‘. l

\

L

l . . . . . . l ******

‘I

..**.

+m.**

l

.

.

.

..I

.

.

.

.

.

.

.

.

.

.

.

.

3

\

\

\

\

\

rC’s+er

na ---

.-.e-----

---m

e I

I

5

.

I

I

I

I

Arter;oi

7.55

I

I

1

T

I

1

I

1

I

I

1

.. f

Lumbar

.I../!....*.I

-1 --- -I

CiS

I

I

1

I

I

I

I

1

I

I I I

1

I

I

I

Cistern0

I

I

I

1

I

1 6

,0 \

/’ ,0’ L umbor

2422I

I

I

I

C

0

I

2 lime

FIG.

lumbar thetized

1. Pco2, pH, CSF vs. time dogs. Vertical

I

3 (hours)

I

4

I

1

5

6

I

0

I

I

1

2 Time

and [HCOs-] of arterial plasma, cisternal and during isocapnic metabolic alkalosis in 6 anesbars indicate =t 1 SEM.

locations were essentially stable and within trol by 4.5 h (Tables 1 and 2, Fig. 2).

I

C

1 mV

of con-

DISCUSSION

As in our study (7) of metabolic acidosis, with stable isocapnic metabolic alkalosis pH+ and pHeo3- return very nearly to the initial control values (Fig. 2). Thus the changes in distribution of H+ and HC03between CSF and blood during metabolic alkalosis are consistent with the hypothesis that these ions distribute passively across the blood-brain barrier. The magnitude of the steady-state p for Hf and HC03 would, as already discussed (7), reflect the balance between production of metabolic H+ by brain tissue and permeability of the blood-brain barrier

I

3 (hours)

I

4

5

2. CSF/plasma DC potential difference (E) and electrochemical potential differences for H+ &+) and HC03(pHc@-) vs. time during isocapnic metabolic alkalosis in 6 anesthetized dogs. Values of AE/ApH, were calculated for change in time 0 sample from control. Values of pH+ and pHCO3are calculated for concentrations in CSF and mean capillary plasma water. Vertical bars indicate ztl SEM. FIG.

to these ions. Return of p is most easily interpreted as indicating that metabolic alkalosis does not alter brain H+ production or permeability. The alterations at the cisternal and lumbar locations are similar in magnitude and rate of change, implying a common process of regulation for brainstem and spinal cord. This finding adds further reassurance to our conclusion that cisternal CSF is in equilibrium with underlying medullary extracellular fluid rather than resembling choroid plexus secretion (7). The effect of metabolic alkalosis on CSF acid-base status has been studied by three other groups (Table 2). Only Adaro et al. (1) measured CSF/plasma E, but evi-

Downloaded from www.physiology.org/journal/ajplegacy by ${individualUser.givenNames} ${individualUser.surname} (129.186.138.035) on January 12, 2019.

CSF-BLOOD

H+

AND

HCO3-

WITH

METABOLIC

1143

ALKALOSIS

TABLE 1. Values for acid-base parameters in arterial plasma, cisternal and lumbar CSF, CSFlplasma E, and calculated electrochemical potential dzjferences (p) for H+ and HCOa- in six dogs during constant isocapnic metabolic alkalosis

Time, h

zt * *

0.010 0.015 0.011

23.3 24.6 25.0

rt 0.8 dz 0.5 zt 0.6

35.1 47.9 49.3

A It =t

1.1 1.4 1.4

1.5 1.7 1.6

+ zt +

0.3 0.1 0.1

3.5 6.2

Arterial CSF-cistern CSF-lumbar

7.522 7.340 7.340

+ * zt

0.015 0.017 0.017

33.7 26.3 26.3

zk 0.8 zt 0.6 & 0.5

37.7 51.1 51.8

* * =t

0.9 1.3 0.9

1.9 1.7 1.6

* 0.2 Zt 0.1 & 0.2

-1.6 2.4

=t =t

Arterial CSF-cistern CSF-lumbar

7.524 7.366 7.367

zk 0.016 =t 0.014 + 0.014

32.4 28.3 29.1

zt 0.9 A 1.3 Zt 1.5

34.6 52.0 53.2

rk =t zk

0.9 1.8 1.6

1.4 2.3 2.5

* A zt

0.2 0.4 0.4

-1.4 2.7

Arterial CSF-cistern CSF-lumbar

7.528 7.370 7.381

+ zt zt

0.015 0.008 0.015

32.3 28.8 30.1

zt zt zt

0.8 1.2 1.4

35.6 52.5 54.0

h & &

0.7 2.1 1.6

1.4 2.6 2.7

zk 0.2 zt 0.3 zk 0.4

Arterial CSF-cistern CSF-lumbar

7.528 7.371 7.384

A * h

0.015 0.010 0.018

32.4 29.3 30.3

& zt =t

0.8 1.2 1.5

35.1 53.3 53.1

zt 0.8 =f= 2.1 =t 1.4

1.3 2.7 3.1

It zt zt

3

6

Values are means =t SE. in METHODS of ref. 7.

The

p for

H+

and

HC03

2. Cisternal plasma p for H+ and HCOa-:

are

calculated

dzyerence

(AI() between sustained metabolic alkalosis and normal

Chazan Fencl

Present

&a+#

et al.

(1)

et al. et al.

study

(3)*

(2)*

mv

4~~03-)

Comments

mv

+1.4

-1.7

Dogs, about 5 h. Combined with respiratory acidosis. E measured simultaneously

+2.0

-3.0

Dogs,

-0.1

-0.8

Unanesthetized days

+0.1

+0.9

Dogs, 6 h E simultaneously

6-9

from

0.2 0.4 0.5

CSF-mean

PH+,

pHCO1-,

mv

3.6 6.5

zk 0.8 zk 0.9

-3.5 -5.6

* +

0.9 1.1

0.4 0.4

6.3 10.5

=t 0.7 z#z 0.8

-6.0 -10.0

=t •t

0.9 0.9

=t zt

0.3 0.4

4.4 8.2

+ zt

0.6 0.8

-3.2 -6.8

zt 0.9 zt 0.9

-1.4 2.6

rt +

0.3 0.3

3.8 7.4

zt 0.5 zk 0.5

-2.4 -5.8

zt zt

0.8 0.8

-1.5 2.6

=t A

0.3 0.3

3.7 7.1

* 0.4 zk 0.8

-2.6 -6.0

zt zt

0.9 1.0

capillary

zt 0.3 =t 0.5

mV

concentration

differences

as described

pH+ for their goats behaved similarly; possible reasons why remained -3.0 mV from normal will be discussed subsequently (6). Fencl et al. (3) offered the hypothesis that regulation of CSF acid-base status was by active transport, the pump for H+ or HCOabeing located at the blood-brain barrier. Their data appear to support quite as comfortably a process of passive distribution. The active transport proposed by Fencl et al. (3) afforded explanation for the small changes in CSF pH from normal, about 0.022 pH for an increase in plasma [HCOa-] from 28 to 40 meq/kg HPO. For a lesser degree of metabolic alkalosis, we observed a greater change in cisternal CSF pH (Table 1). The closer regulation of CSF pH observed by Fencl et al. (3) is best explained by compensatory hypoventilation in their goats as contrasted to the iatrogenic isocapnia of the present study. With respiratory compensation for a primary metabolic alkalosis, CSF pH would be expected to return toward normal. Indeed because of the large value for AE/ApH8. occurring with metabolic rather than respiratory acid-base changes, one would predict from the Nerst equation that with complete compensation i.e., an arterial pH of 7.40, the pH of CSF would be more acid than normal (see Fig. 3 of ref. 6). /-cHCO3-

-----

Study Adaro

E, mV

7.398 7.344 7.341

0

4.5

[Lactate-], meq/kg H20

Pco2, torr

Arterial CSF-cistern CSF-lumbar

Control

TABLE

[HCOa-I, meq/kg SO

PH

days goats,

measured

* CSF/plasma E was calculated assuming a AE/ApHa of -42.3 mV/pH for the metabolic component of pH, change and -30.5 mV,‘pH, for respiratory alterations of PH.. Separation of metabolic and respiratory contributions to the change in pH, and E was made assuming an in vivo buffer value of 10 slykes.

dence that a steady state had been attained is lacking. Their dogs experienced a combined respiratory acidosis and metabolic alkalosis. Utilizing reported values for AE/ApH, (5), cisternal p for H+ and HCO,can be calculated for the other two studies. These values are at best estimates, in part because CSF/plasma E was not measured and documentation of the stability of the control values is lacking. Chazan et al. (2) acutely anesthetized their dogs for sampling of CSF, creating the possibility of an unstable acid-base state. From the studies by Fencl et al. (3) of awake goats, our estimated values for j&t and &C03during metabolic alkalosis are close to those of their normal animals. During metabolic acidosis (Table 3 of ref. 7)

5

We Watson This

thank Ms. for patience work was

15991-05 National T. Award Received

Patricia Reynolds, Mica1 Middaugh, and help. supported by Public Health Service

from the National Institutes of Health.

Institute

F. Hornbein is supported by Grant 5 K03 HE0961 7-09. for

publication

3 June

of

General

Research

Medical Career

and

Mary

Grant

GM

Sciences,

Development

1974.

Downloaded from www.physiology.org/journal/ajplegacy by ${individualUser.givenNames} ${individualUser.surname} (129.186.138.035) on January 12, 2019.

1144

E.

G,

PAVLIN

AND

T. F. HORNBEIN

REFERENCES 1. ADARO, F. V. M., E. E. ROEHR, A. R. VIOLA, AND C. WYMERSZBERG DE OBRUTZKY. Acid base equilibrium between blood and cerebrospinal fluid in acute hypercapnia. J. A#. Physiol. 27: 271-275, 1969. 2. CHAZAN, J. A., F. M. APPLETON, A. M. LONDON, AND W. B. SCHWARTZ. Effects of chronic metabolic acid-base disturbances on the composition of cerebrospinal fluid in the dog. Clin. Ski. 36: 345-358, 1969. 3. FENCL, V., T. B. MILLER, AND J. R. PAPPENHEIMER. Studies on the respiratory response to disturbances of acid-base balance, with. deductions concerning the ionic composition of cerebral interstitial fluid. Am. J. Physiol. 210: 459-472, 1966. 4. FENCL, V., J. R. VALE, AND J. A. BROCH. Respiration and cerebral

5. 6.

7.

8.

blood flow in metabolic acidosis and alkalosis in humans. J. Ap@. Physiol. 27: 67-76, 1969. HELD, D., V. FENCL, AND J. R. PAPPENHEIMER. Electrical potential of cerebrospinal fluid. J. Neurophysiol. 27 : 942-959, 1964. HORNBEIN, T. F., AND E. G. PAVLIN. Distribution of H+ and HC03between CSF and blood during respiratory alkalosis in dogs. Am. J. Physiol. 228: 1149-l 154, 1975. PAVLIN, E. G., AND T. F. HORNBEIN. Distribution of Hf and HC03between CSF and blood during metabolic acidosis in dogs. Am. J. Physiol. 228: 1134-I MO, 1975. PAVLIN, E. G., AND T. F. HORNBEIN. Distribution of Hf and HCO3between CSF and blood during respiratory acidosis in dogs. Am. J. Physiol. 228: 1145-l 148, 1975.

Downloaded from www.physiology.org/journal/ajplegacy by ${individualUser.givenNames} ${individualUser.surname} (129.186.138.035) on January 12, 2019.