The processes affecting the distribution of bromide in blood, brain, and cerebrospinal fluid

The

Processes

Affecting

Bromide

the

in Blood,

Brain,

Cerebrospinal MICHAEL

Rfcfivfd

Distribution

of

and

Fluid I'OLLAY~

Septemhfr

10, 1966

The steady-state distribution of bromide (Br-) in the nervous system oi the rabbit was studied at various plasma concentration levels (05-20 rn>*). In addition, utilizing a ventriculo-cisternal perfusion system, the flus of Brbetween CSF and blood was measured under various experimental conditions. It was observed that the steady-state distribution of Brin brain and cerebrospinal fluid was concentration dependent and that brain served as a “sink” for the plasma and CSF. The results of the flux experiments revealed that the efllux of Brfrom the perfusion fluid was some 30% greater than the influx of Brfrom the blood. In addition, the movement of Brout of the perfusate was attenuated with increasing concentrations of Brand on death of the animal. The results suggest that, at low concentration levels, Bris rapidly cleared from the CSF (by an active transport system) and the brain (by either an active transport system or oxidation of Brto BrO,-), while at high concentration levels, the relative ineffectiveness of the proposed systems results in the accumulation of Brand the establishment of a slate of equilibrium between blood, brain and CSF. Introduction Recent studies have demonstrated the influence of active transport processes on the distribution of thiocyanate (CXS-) and iodine (J- ) between blood; brain and cerebrospinal fluid (13, 1.5). The available evidence from earlier studies on Br- distribution in the central nervous system seems to suggest that the apportionment of Br- between cerebrospinal fluid (CSF) and blood is similar to that of I- and CNS-, although in general the low plasma,/ CSF ratios have been ascribed to restricted diffusion between blood and CSF or the imposed loss of CSF Br- into brain tissue (3, 4, 9. 18. 19). It is therefore of some interest to re-examine the question of Br-- distribution in the nervous system with the intention of defining the processes which ’ This Xational technical

stud>- has been supported Institute of Neurological assistance of illan Stevens

in part by Public Health Service Grant XB Diseases and Blindness. This is to acknowled~r and Ray Counsellor. 74

0.;193, the

are primarily responsible for the movement of Kr cerebrospinal fluid.

between l~lood. brain, and

Methods New Zealand white rabbits (l.O-2.j 1;~) were utilized as the experimenta animals. The first group of experiments was concerned with the determination of the steady-state distribution of bromide at various plasma concentration levels ( < l.O-> 20.0 m>f) in brain and cerebrospinal Buid. In fourteen animals, solutions of sodium bromide ( 2-15 mmoles kg) containing trace amounts of BP were infused into the marginal ear vein over a period of 20 min. The specific activity of the infusion solution varied between 0.6 !IC Hr”” and 0.04 ftc RF; mmole of NaBr at low and high dose levels. respectively. In a manner to be subsequently described; the animals in this group were killed at 24 hours after infusion. In thirty animals receiving sodium bromide (NaBr”‘) at the very highest or lowest concentration levels, tissue samples were procured at 0.5 : 3.0, 6.0, 12.0, and 24.0 hours after injection. At the termination of the experiments, the animals were anesthetized with Nembutal (50 mg, ml) and blood samples obtained hy cardiac puncture. The CSF was collected from the cisterna magna by percutaneous puncture and the animals decapitated. The cranial cavity was then entered and three random samples of tissue from the cerebral hemispheres, containing both gray and white matter, were taken for analysis. The blood, brain, and CSF samples were prepared for countin g as described in a previous communication (13). All samples were prepared in triplicate and counted in a low-background proportional countin g system with a counting error of < 1..5c/o (20). The gross measurements of radioactivity were corrected fol background, decay, and self-absorption. The tissue concentrations were corrected for water content using values previously reported by l’ollay (13). In this study, the bromide space ( yOo) was defined as the ratio concentration Brp per Kg brain water to concentration Br- per Kg plasma watel x 100. The second group of experiments was concerned with the movement of Br between blood and CSF under various experimental conditions, The experimental arrangement. utilizing a ventriculo-cisternal perfusion system, as previously described by Pollay and Davson (14), consisted of an inflo\j cannula in one lateral cerebral ventricle and an outflow cannula from the cisterna magna. The ventricular perfusate contained in mmole,;l: KaCI 1.54. KC1 2.8, CaCI,, 1.1: RIgSo, OS. as well as trace quantities of inulin-Cll (0.01 !tc,‘mI). The rate of perfusion was maintained at 65 $‘min with a constant injection pump and the effluent from the cisterna magna collected under a controlled negative pressure of 15 cm of water. Outflow sample tubes mere changed at 30.min intervals o\:er a total experimental period of 2 hours. In

76

POILAY

certain experiments, blood samples were obtained by cardiac puncture at l-hour intervals. Tn order to measure the efilux of Brp from the ventricular system. KaBr (0.5-50 m&f) with trace quantities of Br”’ were added to the ventricular perfusate. The efflux of Bra” from the perfusion system was also studied when a constant level of unlabeled NaBr ( 10.0 mM) was maintained in the blood. In several experiments, after a steady state had been reached; the animals were killed with an overdose of Nembutal and the perfusion continued for an additional 2 hours. To measure the influx of Brp from blood to CSF, a constant blood level of NaBr (0.5 IXM) with trace amounts of Br”” was maintained in the blood. In these experiments, the ventricular perfusate contained only inulin-P4. The blood and perfusate samples were prepared for counting as previously described ( 13), but in the flux experiments. it was necessary to separate BP and Cl” activity with an IS mg /cm’) absorber. In those experiments in which unlabeled NaBr was used, the concentration of Br- was determined by the titrimetric method of Schales and Schales ( 16) after correcting for chloride content. Results

Steady-Stat? Ilistriblrtion oj RY. The steady-state values representing the distribution of Brm in blood, brain, and CSF at various intervals following the injection of the loading doseare shown in Figs. 1 and 2. .4 steady state

01 0

-

CSF

-

BRAIN/PLASMA

-

BRAIN/CSF

I 4

I 6

I PLASMA

I I.2

I 16

I 20

1 24

HOURS

FIG. 1. Ratio of concentrations ing inha\-ascular infusion of NaRr? prosimatcl~ 0.5mrbl Irvcl.

ui

Rrin blood, brain Plasma concentration

and CSF of Rr-

at interval% maintainv(l

i40~~9t ap-

is reached some I2 hours after iniection and maintained o\:er the ensuin: 12-hour period. In addition. at a plasma concentration level of approsimately 0.5 rmx (Fig. 1) the ratio of HI-~- CSF Br plasma is about 0.75. while at levels approsimatel~ 20 nm (Fig, 2) this ratio approaches unit!;. Similarly. the concentration of Br-- in brain tissue as compared to plasma and CSF reaches its maximum value only at high plasnla concentration levels. at which time the ratios brain plasma and brain, CSF are eyual. The same results were noted when steady-state ratios were determined at 24 hours after injection at various plasma concentration lel:els (Table I ). After

--

0

- BRAIN

I PLASMA

BRAIN/CSF

I

I

I

I

I

I

4

8

12

16

20

24

HOURS

the plasma concentration exceeded I ,i nm (Table l( line 3) the maximun~ distribution ratios at the steady state were reached. This is dramatically iIlustrated when the concentration of Hr-~ in (‘SF is plotted against the concentration of Br- in plasma (Fig. 3). The observed enlargement of the computed Br- space parallels the establishment of a state of equililxium between plasma and CSF (Fig. 3). At high plasma Br- levels, using either the CSF or plasma as a reference volume. a similar value is computed for the Br- space (Table 1~ columns (1 & 7). Br Fhrs Brtwcerl l’laswa alld C’SF, Thcorctic-al. The calculations for these experiments are based on equations previously derived by Bradbury and Davson ( I ) and I’aI~penheimer cl ul. ( 10) \Yhen the solute (nmole,~mI)

of

1~ Esperimentdl

b Ratio

values

concentrations.

0.12.; 0506

H,O

Br -

at 24 hours

rt 1.610 2 0.700

t I

0.777 5.640

16.220 21.300

kg

mmole

Plasma

after

injection

4.060 6..UO

0.144 1.160

mm&/kg

Brain

STEADY-STATE

H,O

expressed

-c 0.420 t 0.181

& 0.030 t 0.096

Br-

DISTRIBUTION

as means~

CSF

I-l.720 21.750

0.614 1.505

SD.

t r

k k

H,O

1.520 1.730

0.110 0.322

Br-

AT VARIOUS

1

mmole/kg

OF Br-

TABLE PLASM

0.91 1 .oo

o.i9 OS0

CSF/Plasma

LEVELS”

_~

I’

0.25 o..:o

0.19 0.21

Brain/Plasma”

0.28 0.29

02.3 0.26

BrainjCSFtj

is added to the fluid from the fluid is:

the ventricular

perfusing

,il, -z iri(c-(‘,,)

eiilus ui

systen~: the

solute

~ i7,.,,c,,,.

I'1

where, ii,, -: efiluv (nmwle, min ) C, = concentration in inflowing perfusate, C,, _ concentration of solute

of solute (nirnole ml ) (nmole nil ) in effluent

from

of CSF

the cisterna

I”i = rate at which

O’( 0

magna;

I’,.., = rate

perfusate

enters lateral

1 I5

IO

5 PLASMA

of secretion

br-1

(~(l,.‘miu).

and

ventricle.

20

-

20

-

I5

Er-

SPACE

W

JO

25

m moles/liter

The rate at which CSF is formed is calculated from the dilution of inulin which is added to the ventricular perfusate. The loss of inulin by diffusion across the lining ependyna or by Ml; absorption was considered negligible \vhen the pressure in the outflow s!;stem was less than --IO cm of water (6. 11). Therefore. I,.,,- = C“i--C“,,. I-7,> 121 where C’; and C”,, = the concentration of inulin-C’ ’ (cpni, ml ) in the inllo~z:in:: and outilowin~ fluids: respectively. At the steady state the iiet transventricular flus is the difference between the inRus and eftlus from the rsperiniental system. fi = kr,C, --k,c where

d

k,, and k, represent

ml, min. The average

intlus ventricular

and efilus

],I.

I .1 I

coefficients

concentration

(c,.)

wit!1 the dimetwions is taken

as ihe mean

80

POUAY

of the infow and outflow concentrations. The plasma (C,,,) and the average CSF concentration (c,;) are expressed in mmole,,‘ml. IIThen the concentration of solute in blood = 0, Eq. 13] is /i,, = k,,C,;. By substituting I3a] becomes

Eqs.

13aj

11 ( and 12 1 into ) 3a] and rearranging

k,, = 6’i( Ci-C,,)

-~~(C’i-Cro/

terms, Eq.

C’(J) C,b,‘C,-

141

When the solute is only added to the blood: the sign of the computed influx (/ii) will be negative, and

Ii, = i/i(-Cc,)-ti,.,fCc,. Therefore

151

the influx coefficient (hi) is kf = tiC’(, + ti( C’,-C’,,/C’,,)

II61

’ Co, C,,,.

TJle E&Y of Br- From CSF. The rate of outflow from the ventriculocisternal perfusion system became constant after the first 30 min as did the observed loss of Br- from the perfusate. The experiments lasted for approximately 2 hours. The average rate of production of CSF in thirty animals. as calculated from Eq. 121, was 12.1 b&/rnin sn i I .2. The data and calculations of one experiment in which the efflux of Br- at a low concentration (0.5 mM) was measured are shown in Table 2. The rate of efklux (ri,,) at various CSF concentration levels (0.5.50 m&l) revealed the nonlinearity of this flux at CSF concentration levels above 15 mM (Fig. 4). The clearance of Br- from the ventricular perfusate (ko) demonstrated a similar attenuation at high (50 mM) perfusate concentration levels (Table 3). In four esperiments, the egress of NaB?“. at low CSF concentrations (0.5 mivr). was TABLE

fi

C’i

(Ld/min) ~~-____ 6.5 i7,,,, ,j,,

13 79.1

65 (1.3 79.1-11 = 6,; x IO-” (0.50 = 8.05 Ikmole,‘min

k,, =

C’,,

(cpm,rnl)

__.

=

8.08

x

(1 Steady-state b Equations

741111

lo-“/‘0.41

at for

z

X lo--:‘--0..12

X lo-:’

values used

ill)

~__~-

Il..3

0.0197

c,,

(mmolc

x low:~/ml)

pl!min

ml. min

t = 60 min.

cakulation-see

c, 050

11 741

X IO-::)-11.3

z

2

test.

X IO-“,0..11

X 10pCF

O..Z2

c,; 0.11

measured with plasma Hr concentraticm~ of 10 mar. It IUS olxer\;e(l that the nxkmum rate of mo\:ement of Hr-- out of the C‘SF is unaffectetl hi- an adverse concentration gradient between (‘ST; and l~lood (,Table 3). In two experiments. after reaching the steads state. the animals were killed \vith an overdose of Nenibutal and the ~~entriculo-cist~rlial perfusion continued for

6r

ol It

0

IO

20

30

40

50

i-0

about 2 hours. The results of these experiments (Fig. 5 ) reveal a marked decrease in the amount of III cleared from the perfusate in its Ifow thrctrtgh the ventricular system. .Is might be expected. the production of C‘SF ceased after the animals were killed. Although the loss of the diluting effect of the newly formed C‘SF would tend to increase the measured Bt--- in the cfi?rrent,

82

POLLAY

it does not account for the marked increase observed following the death of the animal. The Movctttmt oj Hr- Front Blood to CSF. In these experiments. a constant plasma concentration of IYaBP was maintained at a 0.5-m, level over a 2-hour period, during which time the ventricular system was perfused with a fluid devoid of Br-. The influx coefficient (k?) computed from this data was some 3Oc/ less than that found for the efflux coefficient (k,,) at the

42

2

0 t 0

1

30

60

-0

90

MINUTES

FK. 5. Effect of death of the animal on the stead--state loss of Hrirom the ventricular perfusate and the rate of secretion of CSF. Loss of Rrp from pel-fusion system expressed as the ratio of outflow concentration (mmole,‘ml) to inflow concentration

(mmole~ml).

same concentration Rrp as represented CSF.

Rate

of CSF

production

(i.,,Y,)

calculated

from

dilution

of inulin.

level (Table 3). Note that the clearance of plasma by kj is very nearly the calculated rate of secretion

of of

Discussion

The steady-state distribution ratio for Brp in plasma and CSF has been shown to be significantly less than unity. The reported values in man. dog and rabbit are 0.33, 0.77 and 0.72: respectively (3. 4, 7> 8, lS, 19, 22). In common to all species studied is the observation that the distribution ratio increases as the concentration in the blood is elevated, although the level at which this occurs varies considerably between species (3: 5, 20, 21). In the present study, it was also noted that the distribution of Br- between the brain and CSF compartments was remarkably dependent on the plasma concentration level (Table 1. Figs. 1 and 2). In fact, the computed distribu-

tion volume2 for 131 in brain tissue was some .50’/, greater after the C‘SF and blood compartments reached equilibrium (Table I. Fig. 3 ). It is apparent that the movement of Hr- within the nervous system cannot be explained on a simple kinetic basis as proposed in the past (3. 12 ). 1t was following the demonstration by T’ollay and Davson ( 14) that certain anions (I- and CiXS-) are transported out of the CSF that it became possible to theorize about the effect of a CSF pump on the movement and distribution of anions in the nervous system (4) Reed et al. ( 15) and Pollay (13) proposed and supported with experimental data the concept that the rapid egress (by transport) of I- and CS!? from CSF to blood is responsiblefor the CSF/‘plasma ratio of less than unity. They also demonstrated that, by altering the effectivenessof this pump (by saturation. competitive inhibition and derangement of cellular respiration). the blood. brain, and CSF compartments approached eyuilibrium. It was on the basis of these studies that the movement of Rr- between CSF and blood was reexamined. The demonstration. in the flux experiments, that the maximum clearance of Brp from the CSF was concentration dependent, occurred against a concentration gradient. and required cellular energy. certainly suggeststhat the movement. of Br- from CSF to blood is similar to that of T- and UK-. There was also evidence that Br- enters the ventricles at a rate ,greater than that found for T-- or CSS- (3. 13. Ii). It seems probable that Rr- enters the ventricu!ar system in the newly formed Auid at a concentration similar to that found in the plasma, since the rate Brmwas cleared from the blood (12.7 ul/‘min) was almost identical to the rate of CSF production ( 12.1 bll/min) The demonstration that the steady-state ratio CSF!‘plasma is 0.S suggeststhat the nascent fluid is modified by CSF transport of Brp into blood and diffusion into adjacent brain tissue during its passage through the ventricular and subarachnoid channels. Presumably. on elevating the concentration of Br- in plasma, the capacity of the pump is reached with a subsequentaccumulation of Br- in the CSF. Lynder these conditions, the ratio CSF /plasma approachesunity. Tt has been shown that the distribution of Br- in brain tissue is dependent on both the plasma and CSF concentration levels (Table 1). This relationship is similar to that noted in the I-- and CNS- studies (13. 15). except that in the case of Hr. the brain acts as a “sink” for the CSF. This situation implies that, at low plasma concentrations. Br-- enters the brain from both the CSF and plasma. in spite of the fact that CSF is being

84

POLIAk

rapidly cleared of Br-. If the movement of Br- into and out of the brain was due only to diffusion across the blood-brain and brain-CSF barriers: it would be expected that at low concentration levels? the concentration of Br- in brain (based on a maximum distribution volume of some 303~ of total brain water) would eventually be greater than the concentration in CSF. This is the situation reported for CKS- but is not that observed for Br- (13). It is therefore necessary to propose a system which explains the rapid epress of Br- from brain which is overcome at high plasma and CSF concentration levels. One possible explanation is that Br- is oxidized to BrO,<- by brain tissue and eliminated by diffusion into brain capillaries (3). Presunably, at high concentration levels. the brain tissue would be unable to oxidize a significant proportion of the increasing quantities of Br- entering the brain compartment from CSF and blood and, hence, Br- would accumulate in the brain tissue with the resulting establishment of equilibrium with the (‘SF and blood compartments. It could also be proposed that a Krpump is interposed between the brain and blood compartments. Tn the event of small Br- loads. the pm7~p could effectively maintain the low concentration of Br- in brain tissue with the resulting drain of Br- from plasma and CSF. On the other hand, at high Br- loads (which would occur at high plasma and CSF concentration levels), the pump would be relatively ineffective and Br- would accumulate in brain tissue. The data available do not support one proposal over the other, but only indicate that a significant flus must occur by a specialized system operating across the brain-blood and. perhaps. the brain-C% barriers. References

I.

BRA~BTJRY, M. LV. B., and H. DAYSON. 1965. Thr transport of urea, creatinine, and certain monosaccharides between blood and fluid perfusing the cerebral wntricular system 0i rabbits. .I. Pkysiol. Lofzdon 170: 395-211. 2 RRATTG.ARD, S. Il., and T. Lmxn~urs~. 10.54. Demonstration of Brqz in nerve cells, .r. ‘?-Pwol. ~wfwnwr~. Pcy-h?bfr. 17: 1 l-1.3. .i DAVSON, H. 1956. ‘LPhysiolocy of the Ocular and Ccl-ebrosl~inal Fluids.” Little, Brown. Boston, Massachuselts. 4. Damoh-, H. 1964. “.\ Textbook oi General Physiology.” Little, Brown. RoEton, Massachusetts. 5 GRIXKRERC., D. M.. R. B. AIRI), M. D. D. BOEIXLK, L2’. W. CA>IJW.I.I,. n:. E. COIIS. and M. M. MURAYAIL~.A. 1914. A study with radioactive isotopes of the prrmeability of the blood-cerehrospinal fluid barrier to ions. ANZ. J. Physiol. 140: 47-64. 6. HEJSW. S. R., D. Hero, and J. R. P.APPLKIIEIMIX. 1962. Bulk ilow and diffusion in the cerebrospinal Ruid spstem of the goat. din J. Pkysiol. 203: 7;s.781. 7. HUX~LR, G., H. 17. SWTII, and L. M. TAXOR. 1954. On the bromide test of permeahilily of the barrirr lwtwren blood and cerehrospinal fluid-an assrssmrnt. Riorlt~rs. J. 56: 588-.j9i. S. LWIXL. H. 1X:., and J. R. BROWK. 19.34. Results of repeated determinations oi the I)lood-ccl-ebrospinal fluid barl-ier. Pror. SW. Erpt2. Ri,l. dfcd. 32: 216.S20.

10.

11.

11. 1.i. 11. 15.

I (I Ii. IS.

19

20.

21. 12.

I’oI.I.A\-. M. 1966. Ccrehrospinal fluid Iransport and the thiocyanatr space II~ the brain. Am. J. Pl~~~siol. 210: 27;.279. PVILAY, >I., and H. I~~~vson;. 196.1. The pawyx of ccl-tain substances out oi the eel-ebwspinal fluid. Brni~ 86: ldi-1.50. RI’I:D. D. J ., I>. M. Woonncw, L. JAWRS, and Ii. SQ~-nws. 196.5. Factors affecting! distribution of iodide in brain and cerchrospinal Huid. AIII. .I. Pky~iol. 209: ;.ifiG4. SCRALIS~ O., and S. S. SCI+.U.I:S. 1911. ;Z simple and accurate method for the determination of chloride in biological fluids. J. Hid. Chow. 140: Si9-S8.1. \;AN HAKRIW~LD, A. 1966. Extracellular space in the central nervous rystcm. Kouinkl. A-cd. ilknil. U’dcnsrhn,b. Pror. 69: I i-11. ~~XIACX. G. B.. and B. B. Bonn:. 1939. The distribution of iodide, thioqanatr. bromide, and chloride in the central nervous system and spinal fluid. J. Pknrmnrol. Exptl. Tlwra~~. 65: 220.226. CV.AI.L.WI:. G. B., and B. B. HRWII-. 1940. The passage of bromide, iodide and thioqxnate into and nut of the cerrhrnspinal fluid. 1. Phavr~ncol. E.xpfl. ‘I‘hwap. 68: SO-.5.T. \VI:IR. E. G. 1932. The intluence of the wrum bromide concentration upon the distribution oi bromide ion hetwccn serum and qlinal fluid. A ~7. J. Physiol. 137: 109-11.3. L!‘FAK: E. G. 1915. The effects of intracisternal injection of sodium bromide upon the blond-spinal fuid hart-ier. .4r~. J. Ph~siol. 143: s.~-Ys. \VJz!: E. G.. and A. B. H~wrscs. 19.19. The distribution nf I)romidc and chloride in tissue? and body fluids. .I. Bj,l. C~ICIU. 129: .5ii-SSS.