Third ventricle choroid plexus function and its response to acute perturbations in plasma chemistry

Brain Research, 374 (1986) 137-146 Elsevier

137

BRE 11722

Third Ventricle Choroid Plexus Function and Its Response to Acute Perturbations in Plasma Chemistry RONALD E. HARBUT and CONRAD E. JOHANSON Department of Pharmacology, University of Utah School of Medicine, Salt Lake City, UT 84132 qU.S.A. ) (Accepted October 9th, 1985) Key words: third ventricle choroid plexus - - cerebrospinal fluid - - Na + - - K~ - - acidosis - - ion homeostasis - - tela choroidea

The homeostatic role of the third ventricle choroid plexus (3VCP) in the maintenance of CSF electrolytes was investigated by quantifying alterations in CP epithelial ion concentrations induced by chemical perturbations of plasma in adult Sprague-Dawley rats. Significant regional differences (third vs fourth (4VCP) and lateral ventricle CP (LVCP)) were found in epithelial content of Na ÷ and K÷, with respect to baseline levels as well as alterations caused by 5-60 min of systemic metabolic acidosis. 3VCP, which comprises ca. 10% of total choroidal tissue, has a water content, extracellular fluid volume and vascularity comparable to 4VCP and LVCP; yet 3VCP is characterized by relatively high and low values for cellular [Na÷] (68 mM) and [K+] (118 mM). Compared to time-matched controls, acute metabolic acidosis (i.p. NHaC1) effected a response, i.e. ]' [K÷] and ~ [Na+], in 3VCP that was less than in 4VCP, and substantially smaller than in LVCP. The onset and duration of induced electrolyte changes were qualitatively similar among the 3 plexus regions. Although systemic acidosis severely altered CP electrolyte concentrations, it did not compromise CSF homeostasis of [K+] and [Na+]. The function of 3VCP is discussed in terms of secretory capacity, embryological origin, and innervation. Overall, the findings indicate that transport/permeability phenomena which mediate transmembrane distribution of Na ÷ and K÷ in 3VCP differ quantitatively from other regions of the blood-CSF barrier. INTRODUCTION The choroid plexuses of the b l o o d - c e r e b r o s p i n a l fluid (CSF) barrier are collectively a substantial transport interface for bidirectionally translocating ions and hydrophilic solutes between plasma and CNS extracellular fluid 1,29. Although choroid plexus (CP) tissues in the lateral (LVCP) and fourth ventricles (4VCP) have been extensively analyzed for functional characteristics, the CP of the 3rd ventricle (3VCP) has not been nearly so rigorously investigated by physiologists28. This is unexpected since neuroendocrinologists have widely hypothesized that neurosecretory material in brain regions adjacent to the third ventricle can affect, and be modulated by, transport activity in the 3VCP. Such postulates are strengthened by Kozlowski's finding of neurosecretory innervation of the choroid plexus 9. Rodriguez's review of CSF as a pathway in n e u r o e n d o c r i n e integration stressed the need for delineation of transport/

absorption capacities of the respective CP epithelia22. The general lack of information for inorganic as well as organic ion transport by 3VCP prompted us to pursue the f u n d a m e n t a l problem concerning effects of systemic acid-base imbalance on electrolyte transm e m b r a n e distribution in this tissue. This model has been useful in analyzing the role of CP cation transport systems in CSF function 15,17-25. Numerous anatomical studies of the CP system have described significant variations in fine structure among 3VCP, 4VCP and L V C P 4.11,12,16,19. Innervation density of autonomic fibers, which modulate CSF secretion and Na +,K+-ATPase, also varies from one CP to anotherll.12 Such anatomical differences allow the possibility of functional differences among CP tissues, both in controls and in treated animals. Acute metabolic acidosis has proven to be expeditious for analyzing fluid turnover and ion homeostasis in the C P - C S F system 15,r.25. lntraperitoneal injection of NH4CI induces marked alterations in

Correspondence: C.E. Johanson, Department of Neurosurgery, Brown University and Rhode Island Hospital, 593 Eddy St., Providence, RI 02902, U.S.A. 0006-8993/86/$03.50 © 1986 Elsevier Science Publishers B.V. (Biomedical Division)

13~; [K-~]/[Na +] in the parenchyma (i.e. choroid epithelia) of the b l o o d - C S F barrierJL By comparing the onset, peak effect, and time course of acidosis-induced alterations in CP [K + ] and [Na+], we have provided evidence for similarities and differences in the transport function of 3VCP vs 4VCf ) and I.V('P. MATERIALS AND METHODS

Animals and anesthesia S p r a g u e - D a w l e y rats, 6 - 7 weeks old, were utilized; their average weight was 211 + 8.6 g (S.E.M.). Prior to experiments, the animals were housed in cages wherein they had access to food and water ad libitum, and were exposed to alternating 12-h periods of darkness or fluorescent lighting. Five min before termination of experiment, each rat was anesthetized i.p. with 80 mg/kg ketamine; subsequently, killing was done by exsanguination after blood and CSF sampling.

Experimental protocol A total of 54 animals was used to ascertain: (1) 3VCP compartmentation parameters (water content, volume of residual blood, extracellular fluid (ECF) volume; and (2) 3VCP tissue content of Na + and K +, in control rats and in those injected with NHaCI to induce metabolic acidosis. Due to the diminutive size of CP specimens (e.g. rat 3VCP wet wt. is less than 0.5 mg), it was necessary to carry out determinations of compartmental parameters and electrolyte concentrations in tissues from separate animals. Control and experimental animals refer to those injected with NaCI or NH4C1 salt solutions, respectively. Ten rats were used to quantify H20 content of plexus tissues by an electrobalance technique s. Residual blood volume (RBV) in tissues was estimated as the reduction in weight of tissues resulting from blood removal by perfusion of isotonic fluid (i.e. saline-sucrose solution containing albumin); RBV and blood composition data were used to correct for [Na +] and [K +] in the vascular compartment when CP cell ion concentrations were calculated (Fig. 2). Tissues were dried to constant weight at room temperature (2()-22 °C). E C F volume of CP 7.15 was determined by the steady-state 1-h volume of distribution (Vd) of [3H]raffinose. Controls received NaCI

(n -- 4) whereas treated animals were ~_~ivcn NH~({ (n = 4); both salt injectates (4.7 mmol'kg) contained [~H]raffinose (0.()15 mCi/ml, 0,3 m('i/kg), lhe d o w of NHaC1 was selected from dose-resp~mse analysis indicating that 4.7 mmol/kg elicited peak clevatio, m CP [K +] with negligible morbidity to the animals For analysis of Na+ and K + distribution, an additional 36 rats were divided into t~o equal-sized groups. Treated animals were administered Nlq~('l i.p.. 4.7 mmol/kg; time-matched controls received NaCl at the same osmolar dose. Five, I5, 30 or 60 rain after injection, a sample of blood was taken from the abdominal aorta; immediately thereafter, CSF was aspirated from the cisterna magna and the fourth, third and lateral ventricular choroid plexuses were rapidly removed in their entirety, The following measurements were done: hematocrit, osmolality, pH, pCO 2 and pO e in arterial blood; and K ~ and Na + in plasma, CSF and CP tissues. Detailed accounts of sampling techniques have been previously presented ~'-s,t7.25. Choroid plexus specimens were run over a glass slide to remove adhering CSF; since CSF is relatively high in [Na*] and in H~O content (99%), this wiping procedure helped to minimize artefactuat elevation in the content of Na and Hm() ( - 8()(~; ) in sampled tissues.

Analytical procedures Arterial gas tensions and pH were analyzed at 37 °C with a Radiometer A B L 2 Laboratory. Desiccated choroid plexus samples were weighed to the nearest 0.001 mg on a Cahn 4700 electrobalance, and were then extracted in Li-acid solution ( 15 mM LiCI and 0.02 N HNO3) prior to flame photometric analysis of Na + and K + concentration. CSF and plasma samples were weighed to the nearest 0.00l g, and were then diluted 1:200 with the Li-acid solution. Osmolality of samples was determined with a Wescor 5100B vapor pressure osmometer. Radioactivity analyses were performed with a Beckman LS 7500 liquid-scintillation counter. Micropipette tips for CSF sampling were manufactured with a David Kopf Instruments 700B vertical pipette puller

Calculations and statistical analyses Formulas and assumptions for the calculation of intracellular concentrations (mM or mEq/kg HmO ) have been thoroughly described 7-s,~7,~s>. Standard

139 A. INTERANIMAL ANALYSIS 040

ri-p

<0.05

,

MA).

Analytical reagent

grade NH4C1 and NaCI were purchased from Mallinckrodt, Inc. (Paris, KY), Bristol L a b o r a t o r i e s (Syracuse, NY) supplied Ketaject (ketamine HCI. 100 mg/ml). Disposable 0.01-ml micropipettes were acquired from D a d e Diagnostics (Miami, FL).

E

o.so x el

>'-

land Nuclear (Boston,

020

d3

RESULTS

0,10 -t-" (.9 uJ

Weights of CP tissues

0 B. INTRAANIMAL ANALYSIS 60

8

P <0.05 -i1

03

45 _.1 ¢'~

Oo

50 ~ i~i~i ~!: i~~

.,.J

3VCP

i~i,i,~ii!~iiiiil, 4VCP

I

LVCP

Fig. 1. A: Mean _+ S.E.M. weight (mg) of adult rat choroid plexus tissues dried to constant weight (n = 32 for each region). 3VCP and 4VCP refer to choroid plexus (CP) tissues in third and fourth ventricles, respectively. CP tissues from right and left lateral ventricles of each animal were pooled for the LVCP analysis. 4VCP weight is significantly greater than LVCP or 3VCP, P < (t.(15 by multiple range test. B: The CP weight in each region is expressed as % of total dry wt. of all CP tissues in the same animal. The two LVCP tissues in each rat were pooled. Intra-animal analysis was performed on each of 32 rats, for which mean + S.E.M. is presented. Statistical analysis and abbreviations are same as in A.

errors for derived variables were d e t e r m i n e d by equations from the H e w l e t t - P a c k a r d Statistical Library. Volume of distribution of radioisotopes was calculated by conventional formulasS,~5.25. Statistically significant differences among regions and between treatments were d e t e r m i n e d by multiple range test (Tukey version) and by Student's t-test. Coefficient of variation ( % ) , calculated as 100 x standard deviation + mean, was used to assess sampling and methodological variation associated with microtechniques for handling 3VCP.

Materials [3H]Raffinose (7.8 mCi/mmol) and Biofluor liquid scintillation cocktail were o b t a i n e d from New Eng-

Since dry weight is less variable than wet, a comparison of regional differences in choroid plexus size is a p p r o p r i a t e l y based on dry weight 2°. A v e r a g e dry weight of rat ( 6 - 7 week) 3VCP was 0.078 _+ 0.003 mg (n = 32); see Fig. 1A. Thus, 3VCP is only about 1/4 the size of 4VCP, and 1/2 that of each LVCP. 3VCP comprises 10% of total weight of all choroid plexus tissues (Fig. 1B). Despite the relatively small size of 3VCP, the coefficient of variation (c.v.) associated with sampling size (17%) did not differ from that for 4VCP (16%) or L V C P (18%).

Baseline data for fluids and tissues 3VCP is distinct for relatively high amount of Na + (63 mEq/kg wet or 341 mEq/kg dry); thus, 3VCP contains about 30% more Na + than 4VCP or L V C P (Table I). Conversely, 3VCP [K +] is significantly lower than L V C P [K+]. H o w e v e r , with respect to c o m p a r t m e n t a t i o n p a r a m e t e r s (i.e. tissue H 2 0 content, E C F volume and residual blood), 3VCP values are c o m p a r a b l e to those for 4VCP and LVCP. NaClinjected controls (Table If) had arterial p a r a m e t e r s in the following ranges: p H (7.37-7.42); pCO~ (32-35 torr); and p O 2 (70-79 torr). Mean arterial hematocrit was 39%. Plasma [Na +] and [K +] were on average 154 _+ 0.5 and 3.86 _+ 0.08 m E q / k g H20, respectively, while CSF [Na +] and [K +] were 155 + 1.9 and 3.14 +_ 0.02 (n = 16 p o o l e d control values). The electrolyte and c o m p a r t m e n t a t i o n data were used to estimate parenchymal (i.e. epithelial) concentrations of ions (Fig. 2). Calculated values for cell Na + concentration, [Na+]i, were 68 mM for 3VCP, and 4 2 - 4 4 m M for 4VCP and LVCP. For [K+]i, mean values of 118 m M were o b t a i n e d for third and fourth plexuses vs 143 m M for the lateral plexus epithelium (P < 0.05). The volume of the cell water c o m p a r t m e n t , [H20]i, did not significantly differ among the 3 plexus tissues (Fig. 2).

t40 TABLE 1 Measurements ~)[electrolyte atld water concentrations in control rat~' a comparison o]third ventricle chorotd pte,tu~ ¢3 I ( [ ' ) ~**t/* ,~dwJ tissues Means _+ S.E.M. for n = 16 (electrolyte data) and n = 4-5 (compartmentation data). ECF, extracellular fluid volume c~,mnated b~ l-h V d of [3H]raffinose. In the electrolyte experiments, plasma [Na" ] and [K*] were 154 ± 0.5 and 3.86 :~_::(I.(t8 mEq/plasma H20; CSF [Na*] and [K -~] were 155 ± 1.9 and 3.14 ± (t.(12 mEq/CSF H20; and arterial pH, pCO z and pO2 were 7,40 ~: I).0t, 33.8 { i!7 tort, and 73.3 ± I. 1 torr, respectively. Residual blood determination is described in Materials and Methods, "¢1/( 7'

4 VCP

L 1/Ct '

63.4 Z 1.6 341 :g 8.6

48.(I ± 0.8" 274 ± 4.6*

48.6 ± I. 1 251 +_ 5 , 7

83.3 + 1.(! 448 ?_ 5.3

83.1 ± 0 . 9 ~* 475 + 5.1"*

9 6 . 6 2 1.2 408 ± 6.2 ;

81.4 ± 0.4

82.5 ± 0.3**

8(I.6 ± 0.4

12.3 ± 1.2

13.3 ± 1).4

i3.5 ± 11.6

1.4.4

15.0

17.3

Tissue [Na + ] mEq/kg wet wt. mEq/kg dry wt. Tissue [K +] mEq/kg wet wt. mEq/kg dry wt.

Total HzO content (%) 100 x kg H 2 0 / k g w e t wt.

Tissue ECF H20 (%) 10(I x kg HzO/kgwet wt.

Residual blood content (% of tissue weight)

* P < (/.05, 3VCP vs 4 VCP or LVCP, by multiple range test. ~* P < 0.05, 4VCP vs LVCP.

Compartmentation analysis allows CP ionic concentrations to be presented as raM, i.e. mEq/kg cell HsO. However, the establishment that measured amounts of ions in tissue reliably reflect calculated cellular levels precludes extensive compartmental analysis involving numerous additional experimental (i.e. acidotic) animals. Linear regression analysis of control data for 48 CP tissues from 16 animals revealed that for K + or Na +, there is a high degree of correlation (r2 > 0.99) between tissue and cellular concentrations. Ratio analysis of epithelial [K+]/

[Na +] can be advantageous in elucidating experimental effects on Na +- and K+-translocating systems ~7. Fig. 3 shows a high degree of correlation between cell [K+]/[Na +] and tissue [K+]/[Na+]. Thus, data for tissue [Na+], [K +] and [K+]/[Na +] (Figs. 5-7) should reliably reflect levels of ionic concentrations in the parenchymal cell compartment.

Effects' of acidosis on fluids and tissues Rats were rendered acidotic and analyzed 5, 15, 30 or 60 min later. The effects described below refer to

T A B L E II

Time-course analysis of the effects o f NaCl- or NH4Cl-injection on arterial p H and blood gases Means + S.E.M. for 4 adult male rats. Samples were taken from abdominal aorta 5, 15, 30 or 60 rain after i.p. injection (4.67 mmol/kg) of either NaC1 (control) or NH4Ct (treatment). The significance of the induced differences from time-matched controls was determined by Student's t-test. 5 rain

15 rain

30 rain

O0 min

7.37 ± (I.(12 34.7 ± 2.1 71,3 ± 2.3

7.41 ± I1.(/1 33.9 ± 1.7 70.1 ± 3.1

7.42 k 0.01 32.0 ± 0.49 73.0 __+ 1.8

7.39 ___+0.0l 34.4 + 0.77 78.7 ± 0.57

7.35 ± (t.01 30.0 + 1.1 79.6 ± 1.6"

7.32 _+ (1.01"* 3 1 . 0 ± 1.8 80.0 ± 2.5*

7.29 + (I.(11"* 3 1 . 8 ± 1.3 77.3 + 3.2

7.31 ± 0.02* 30.1 L 1.6 80.5 ±. 2.1)

Control pH pCO 2 (torr) pO 2 (torr)

Treatment pH pCO 2 (torr) pO 2 (torr)

* P < 0.05. ** P < 0.01.

141 mEq Ion kg Cell HzO

4.0

kg Cell HzO x l o o kg Wet Tissue

rz : 0.993 p < 0.05 3.5

[Na]i

[K]i

~/

["20]i

200

3.0

75 • T..

m

;.~:~,f.~::: . .-.- -.-. . . ...........

/M,

80

-:.-?-~: : - -

~';;f;

.;~.~ :

,,,.,,.,. ::::::::.

40

,'{~,/,']_, :: i ! -

;iii;

45

"

30 15

"-'.,','4::: :: :: I/l/X:

~////I i "////]i:: :: - 3V

4V

LV

3V

4V

LV

2.5

.....

120

o

/

z

6O "r

o 3V

4V

LV

Fig. 2. Means + S.E.M. for concentrations of cellular Na +, K + and H~O in choroid plexuses from 3 different regions of the ventricular system in adult rats. LV, lateral ventricle. Cell ionic concentrations were calculated from tissue electrolyte data (n = 16 pooled controls) and means for compartmentation values in Table I. Equations, assumptions, and statistical analyses (i.e., S.E.M. for derived data) are discussed in Materials and Methods. * P < 0.05, 3VCP vs 4VCP or LVCP, by multiple range test.

NH4Cl-induced increases or decreases from timematched control data: thus, 'baseline' and 'treatment' refer to effects o b s e r v e d after injection of NaCI and NH4C1, respectively. In t r e a t e d animals, arterial [H] increased maximally from 38 to 51 nM at 30 min (Fig, 4); this corresponds to a decrease in pH from 7.42 to 7.29 (P < 0.05) (Table II). NaC1 injection caused a small reduction in baseline plasma [H] (Fig. 4); this may relate to plasma dilution due to volume of injected solution (2 ml/100 g b.wt.). Plasma [ H C Q ] was reduced below baseline at least 4 m M during metabolic acidosis: maximal reduction from 21 to 14 mM occurred at 30 rain (P < 0.05; Fig. 4). T r e a t m e n t t e n d e d to lower arterial p C O 2 by 2 - 4 torr, but the reductions were generally not statistically significant (Table I1); also as the result of hyperventilatory c o m p e n s a t i o n p O 2 was increased by up to 10 torr (P < 0.05, at 5 and 15 rain). In metabolic acidosis plasma [K +] increased maximally from 3.6 to 5.3 m M at 5 rain (Fig. 4); b e y o n d 15 min, [K +] continuously declined but was always greater than corresponding baseline values. A treatment-induced maximal decrease of 7 m M in p l a s m a [Na +] occurred at 30 min (Fig. 4). Systemic acidosis

/ LvcP

ua

2.0

4VCP

3VCP 1.5

I.O I.O

I i.25

I 1.50

I i.75

I 2.00

TISSUE[q/[No]

2 25

Fig. 3. Relationship between cellular and tissue concentrations of cations. Linear regression analysis of cell [K*]/[Na +] vs tissue [K+]/[Na *] yielded a slope significantly different from zero and a correlation coefficient squared > 0.99. The least-squares regression represents control electrolyte data obtained for 48 tissues, i.e. 16 for each CP region. Limits are S,E.M. The high degree of correlation between cell and tissue concentration ratios is a reflection of the relatively large size of the parenchymal cell (epithelial) compartment (see H20 i data in Fig. 2) and the relatively small and similar volume of the extracellular fluid compartment in all 3 CP regions. Thus, in experiments in which ECF volume is not significantly altered (e.g. metabolic acidosis), changes in tissue [K+]/[Na +] should accurately reflect alterations in cell [K+]/[Na+].

over the 60-min period did not significantly alter CSF [K +] and [Na+]. A u g m e n t e d [K +] in CP occurred as early as 5 - 1 5 rain after NH4CI injection (Fig. 5). 3VCP, 4VCP and L V C P r e s p o n d e d to acidosis with peak increases at 30 min. These consistent increases (A) in mean [K +] were in the o r d e r 3VCP [d 48] < 4VCP [A 81] < L V C P [d 103]; values in brackets are concentration changes in m E q / k g dry tissue. N o n e of the increases in tissue [K+] returned to baseline by 60 rain (Fig. 5). [Na +] in CP was generally r e d u c e d by NH4CI (Fig. 6), even at 5 min in LVCP. Increases in [K+]/[Na +] in CP tissues caused by acidosis were proportional to baseline ratios of plexus [K+]/[Na+]; thus, induced changes in [K+]/[Na +] were greatest in L V C P , and least in 3VCP (Fig. 7). C o m p a r t m e n t a t i o n p a r a m e t e r s were analyzed in the 60-min treated animals. Tissue water content (n = 5) increased by 1 - 2 % in the 3 CP tissues, but the differences were generally not statistically significant from NaCl controls. Tissue E C F volume was not significantly altered by NH4C1 injection. The % values

142 6.0

55 J.

,.rl

~: 5,0

.J

--

"-

n ...................

~5

<[

'~ 45

nuJ I-£K

o3

4.0

<

-'

13.

/

35

1

I

I

I

1

,,,1

3.0

25

l

l

I

. 1

1

__t...__-J

160

O

2O

L

,55

,

...J

E

t.ld 1-117

15

,50 d. ^

I(

1

I

I

20

0

I

t

40

POST-INJECTION

I

60

TIME(min)

"

~:

--J (3_

145 0

I

-k

I

20

POST-INJECTION

o,.

I

~

I

40 TIME

t,

60 (min)

Fig. 4. Time course of change in arterial electrolyte concentrations in metabolic acidosis. Means _+ S.E.M. for 4 adult rats. Open squares represent NH4Cl-injected (acidotic) animals; filled squares are NaCl-injected time-matched controls. [K~], INa +] and [HCO3] are in millimolar concentrations (mEq/kg plasma H20), whereas [H] is in nanomolar units. * P < 0.[]5, acidosis vs timematched control, by Student's t-test.

3VCP 6 0 0 t A

, ....................

500

400 t

t

~

6001 B 4VCP

i

i

* i

,',-

J

~-

]- ~.... ~. . . . . .5 . . . . . . . . . .

a

[K] 400 L

,

~

for V d of [3H]raffinose (n = 4) in acidotic animals were 14.8 _+ 2.4 (3VCP); 15.6 _+ 0.8 (4VCP); and 14.3 + 1.1 (LVCP); these are not significantly different from corresponding control values in Table I. The 60-min V d of [3H]raffinose in CSF was significantly reduced by NH4C1treatment, from 1,50 _+ 0.1 to 1.18 + 0.1 (P < 0.05); this suggests an acidosis-induced decrease in the effective permeability of the b l o o d - C S F barrier to the non-electrolyte. DISCUSSION

,

,

~

,

Size and function of 3 VCP 600

*

LVC P

T h e 3 V C P is c o n s p i c u o u s for its c e n t r a l l o c a t i o n ; i.e. in t h e m i d l i n e o f t h e C S F axis a n d also p o s i t i o n e d

[K]

4001 J L I. I I I 0 20 40 60 POST-INJECTION TIME (rain) Fig. 5. Time-course analysis of the effect of acidosis on [K +] in various regions of the choroid plexus system. Lateral (LVCP), third (3VCP) and fourth (4VCP) ventricular choroid plexuses

were sampled 5, 15, 30 or 60 min after i.p. injection of contro! (NaCI) or acidifying salt (NH4C1). Values are m e a n (_+ S.E.M.) amounts of K + (mEq/kg dry weight) i n C P tissues of control (filled symbols) or treated (open symbols) animals. The choroid plexus content of residual erythrocytes was not significantly altered by NH4CI treatment (15). * P < 0.(]5. control vs acidosis, by Student's t-test.

143

t

ventricular fluid and neurosecretory fibersg, 22. In view of the 'strategic' location of 3VCP, it is surprising there have been so few studies characterizing the

A

+100

~K

,

.,.

o, ~No _1-~---; . . . .

in 3VCP

-~ . . . .

OJ

-t00

io 4vcP

f ~---..

..... • i

"

J

IC

+IOOF

i

i

.~

0~-v'\

-I00|

6KnOr ANo LVCP

---~ j.

+loo~B AK or ANo

,

L

*

I ~-'"~'""

0~-

-'~

/-t/? t

I

i

t

* -~ .....

*

. . . . . . "~ £ ,

0 20 40 60 POST-INJECTION TIME (rain)

Fig. 6. Time-course analysis of magnitude of induced changes by NH4CI treatment, of choroid plexus ionic concentrations (mEq/kg dry wt.). A, experimental value minus corresponding time-matched control. Abbreviations, n values, and statistical analysis are the same as in Fig. 5. Squares and circles symbolize K+ and Na +, respectively. * Significant difference from timematched mean control value, P < 0.05.

between lateral and fourth ventricles. The proximity of the 3 V C P - C S F system to neuroendocrine systems in adjacent brain tissue has prompted numerous hypotheses stating functional interaction between third

3.5 T

[.]

i

2.5 ,

5

T

0.5 3VCP

4VCP

LVCP

REGIONS OF THE BLOOD-CSF BARRIER

Fig. 7. Effect of acute metabolic acidosis (NH4CI injection) on [K+]/[Na+] in choroid plexus tissues from the third, fourth and lateral ventricles. Tissues were sampled at 30 min after i.p. injection of either 4.7 mmol/kg NaCl (control; unfilled bars) or NH4CI (acidosis; filled bars). [K+]/[Na+] refers to the ratio of concentration of K+ to Na + in dried specimens of choroid plexus. * P < 0.05, control vs acidosis, by Student's t-test.

physiology of this secretory and reabsorptive epithelium. Relative inaccessibility and diminutive size of 3VCP are presumably factors previously discouraging in vivo transport/permeability analyses of this b i o o d - C S F barrier region. Assessment of the role of 3VCP in CSF dynamics must take into account tissue size. 3VCP in Rattus norvegicus (mixed strains) attains maximum weight of 80-85/~g (dry weight) at 5 - 7 weeks postpartum 20, the same age of our S p r a g u e - D a w l e y rats in which 3VCP averaged 80 #g. Thus, 3VCP comprises about 10% of total plexus weight in rats (Fig. 1; ref. 20) and in dogs 4. Intra-animal analysis reveals that the proportion of this regional tissue is strikingly invariant (Fig. 1B). Since isolated preparations of L V C P as well as ventriculocisternal perfusion systems in various mammals have shown that CSF formation is proportional to CP weight 1, it is tempting to postulate that 3VCP elaborates 1/10 of total choroidal fluid production. However, Deck et al. z found evidence that 3VCP and L V C P produced far less fluid per g of tissue than 4VCP; their results are complicated, though, by ventricular injection of a contrast agent which could inhibit CSF formation 5. In situ isolation of 3VCP, although technically difficult, would help to clarify secretory capacity.

Fluid compartments in 3 V C P Quantitative analysis of vascular, interstitial and intracellular compartments in choroid plexus furnishes insight on function. 3VCP vascularity is comparable to 4VCP and LVCP, i.e. about 15% of wet tissue weight is taken up by residual blood as determined indirectly by blood removal (Table I), or directly from nCr-tagged erythrocyte V d data in conjunction with hematocrit values 7,8,L7. By micrographic analysis, Quay reported vessel luminal volumes of 13.6, 13.1 and 13.3% of total tissue volume for adult rat 3VCP, 4VCP and LVCP, respectively 2j. Thus, various methodologies for assessing blood content and vessel capacity indicate that 3VCP is extensively vascularized, comparably in magnitude to other regions of CP. The interstitium, or stroma, is the region between blood vessels and epithelial cells; the size and compo-

144 sition of this region may importantly affect fluid movement 4. The V d of [3H]raffinose, an estimate of ECF volume (interstitial fluid plus plasma), ranged narrowly from 12.3% in 3VCP to 13.5% in LVCP (P > 0.05). Similarly, the relative volume of stroma analyzed by planimetry of micrographs was within a narrow range, i.e. 15-18% of tissue volume 21, Stromal volume overestimates interstitial fluid volume since the former includes connective tissue solids and endothelia. Overall, the data indicate that ECF volume between blood and parenchymal cell fluid is close to 15% both in 3VCP and in tissues of fourth and lateral ventricles. 3VCP water content (81.4%) is intermediate between values for LVCP (80.6) and 4VCP (82.5). Comparable absolute values and statistically significant regional relationships in total H20 content of plexus tissues have also been found in another strain of rats 2°. Small differences in H20 content in the various compartments composing the plexus could combine to explain the small, but significant, variation in total H20 content; e.g. Gomez and Potts reported more interstitial connective tissue in LVCP than in 4VCP 4. Volume of intracellular H20, i.e. t-120i, as calculated from tissue compartmentation data, was not significantly different in 3VCP compared to LVCP and 4VCP (Fig. 2). Quay has reported that relative volume of the choroid epithelial (ependymal) cells is the same for the 3 choroid plexus regions 2~. Collectively, the compartmentation data obtained by different laboratories using micrographic, radioisotopic, and other techniques indicate more similarities than differences among ventricular tissues. This is in line with findings by scanning electron microscopy that the epithelia of the different plexuses are macroscopically homogeneous z3. Thus, the gross structural (compartmental) uniformity among plexuses occurs even though there are differences in embryological origin and time course of development20.

Sodium and potassium in 3VCP Although structurally similar, the various choroid plexuses show metabolic, enzymatic and transport differences 3,17,19,26. An unexpected finding, though, was the markedly higher level of Na + in 3VCP tissue, i.e., about 30% greater than in 4VCP and LVCP; this difference was not attributable to a larger proportion

of Na ~ in ECF, since raffinose V d w a s least m 3~,( t ' (Table 1). Conversely, [K+I was substantially less m 3VCP than in LVCP. The striking 100% difference m [K+]i/[Na+]i between 3VCP and LVCP {Fig. 13} is probably not due to differences in Na t,K~-ATPase activity 14, or to an ischemic artefact associated with time for specimen removal (LVCP was sampled last but it maintained highest [K ÷1 indicating sufficient energy substrate for the Na+-K e pump under the protocol). As in a previous investigationtL [K+]i/[Na~], was significantly less in 4VCP than in LVCf'. Electrophysiological measurements are needed to elucidate if regional variation in [K + L/[Na ~ ]i is due to differences in ion pump-leak ratio~L The relatively high [Na']i in 3VCP could affect transport of organic solutes (e.g. hormones, neurotransmitters, etc.} dependent on the transmembrane Na + gradient26. Choroid plexuses, particularly 3VCP, clear hormones (e.g. melatonin) from CSF>; this could be an aspect of the circadian rhythm of the CSF hormone levels. Other consequences of high [Na+]i include effects on rate of apical Na+-K ~ exchange 7,24,29 and basolateral Na+-H antiport 15,25,29 systems.

Effects of systemic acidosis on 3VCP Induced augmentation of [K+]/[Na +] in LVCP and 4VCP has been reported for a 1-h systemic acidosis modeW. However, even 5-15 min after NH4CI injection there are significant alterations in plexus electrolytes, i.e. K + retention and Na + depletion (Fig. 6). The rapidity of ion redistribution in CP makes it improbable that enhanced [K+]/[Na +] is due to synthesis of new Na+,K+-ATPase, which would accelerate Na + extrusion and K ÷ accumulation; although elevated arterial pO 2 occurs (Table I1), it is unlikely that the 5-10 torr increase is sufficient to increase activity of apical membrane Na+,K+-ATPase. Although K + transport into choroidal epithelium is mainly in the apical membrane7, 24,29, there is basolateral transport of K + into CP 18,29, and this system could be affected by acidemia; however, acidosis, if anything, inhibits K+-uptake systems, and we found that extracellular acidosis does not alter uptake of S6Rb (an analog of K ÷) by in vitro CP (unpublished data). Although apical Na+-K ÷ pump stimulation could theoretically be mediated indirectly through catecholamines released into plasma in acidosis 17, current evidence indicates

145 this is not the prominent effect. An alternative, more likely, explanation is that elevated [K+]/[Na +] in CP during acidosis results primarily from reductions in Na + uptake and K + effiux across limiting membranes. Following NHgCI injection, the lowered plasma [Na +] together with augmented arterial [H] cause diminution in the basolateral transmembrane gradients for Na + and H. Consequently, this lowers the driving force for interstitial Na + uptake into the cell via Na+-H exchange6.15, 25, thereby resulting in a reduction in cell [Na+]15, iv. In vivo kinetic studies of 22Na+ penetration into the choroid plexus-CSF system support this model 15. Diffusive effiux of K + from CP has been described by Zeuthen and Wright30; in acidosis, a reduction in K + diffusion from the cell, secondary to decreased membrane permeability, would promote K + build-up and thus contribute to maintenance of cellular electro-osmotic balance in the face of Na + depletion. However, since variations in concentrations of ions do not necessarily reflect alterations in their activities, there is the need for ion-specific electrode measurements to determine if Na + and K + activities in the cytoplasm are altered by acidosis. With respect to K + and Na + redistribution, the 3VCP showed least reaction to systemic acidosis. On the other hand, the greater increases in [K+]/[Na +] after NH4C1 injection occurred in those CP tissues (i.e. LVCP and 4VCP) with the larger ratios of [K +] to [Na +] in the control state (Fig. 7). Thus, the uniquely low reactivity of 3VCP may be attributable, in part, to the apparent atypically low baseline transmembrane Na gradient. Previous pharmacological 27 and toxicological l0 studies also revealed that 3VCP is less reactive than LVCP or 4VCP to chemical challenges. Further investigations are needed to assess whether the generally more diminutive responses of 3VCP are related to observations that this tissue has the lowest 02 uptake capacity ~9 and the greatest degree of innervation 11.12.16of all CP regions. CSF ion homeostasis Despite ionic redistribution between plasma and

REFERENCES l Cserr, H.F., Physiology of the choroid plexus, Physiol. Rev., 51 (197l)273-311.

CP, [Na +] and [K +] in CSF remained stable even after 60 rain of acidemia. One-h treatment with either NH4C1 (systemic acidosis inducer) or acetazolamide (a carbonic anhydrase inhibitor) causes a decreased penetration rate of 22[Na+] from plasma into both the CP epithelium and CSF 15. Since CP secretory rate is proportional to Na turnover from plasma to CSF, the acute treatment of either NH~C1 or acetazolamide leads to inhibition of flowlS-2L The stability of CSF [K +] and [Na +] after NHaCI or acetazolamide indicates that CSF formation per se is integrally linked to CSF ion homeostasis; thus, the stoichiometric relationship between Na +, K + and HeO fluxes across the apical membrane apparently remains relatively constant even when fluid turnover is curtailed by 40-50% (ref. 15, 25). CSF homeostasis of [K +] and [Na+], in the face of substantial plasma perturbations of these ions, is probably effected largely via maintained, or even decreased, blood-CSF barrier permeability. Further evidence for this hypothesis is that the 1-h V d of raffinose (a non-electrolyte with molecular weight of 504) in CSF decreased in acidosis by 20%. Since non-electrolytes penetrate the choroid plexus mainly paracellularly, i.e. through tight junctions, the results point to a decrease in permeability of the entire choroid epithelial interface in acidotic animals. This study has demonstrated that parenchymal elements of CP in the various ventricles, when stressed with perturbations in plasma chemistry, undergo quantitatively-different alterations in ion content. It is of homeostatic significance that the ionic concentration changes in choroid plexuses are not readily transmitted across the apical membranes of the epithelia into CSF. ACKNOWLEDGEMENTS This work was supported by National Institutes of Health Grants NS 13988 and GM 07579, and a Research Career Development Award to C.E.J. We "thank Dr. W.B. Quay for critical reading of the manuscript and Dr. V.A. Murphy for constructive contributions to the project.

2 Deck, M.D.F., Deonarine, V. and Potts, D.G., The rate of cerebrospinal fluid formation proximal and distal to aqueductal obstruction in the dog, Radiologist, 108 (1973) 607-611.

3 Decker. J.F. and Quay, W.B., Stimulatory eflects oi melatonin on ependymal epithelium of choroid plexuses in gel& en hamsters, J. Neural Transm. 55 (19821 53- ~7 4 Gomez, D.G. and Potts, D.G., l'he lateral, third and fourth ventricle choroid plexus of the dog: a structural and ultrastructural study, Ann. Neurol.. 1(! (1981) 333 -341) 5 Harnish, P.P. and DiStefano. V , I)ecreased cerebrospinal fluid production by intravenou~ sodium diatriz~mle, Invest Radiol., 19 (19841 318-323. 6 Johanson, C.E., Differential effects el acetazolamide, bcnzolamide, and systemic acidosis on hydrogen and bicarbonate gradients across the apical and basolateral membranes of the choroid plexus. J. Pharmacol. E.~p 17ter., 231 119841 502-511, 7 Johanson, C.E., Reed, D.J. and Woodbury, 1).M., Active transport of sodium and potassium by the choroid plexus ~f the rat, J. Physiol. (London), 241 11974) 359-372. 8 Johanson. C.E., Reed, I).J. and Woodbur,,, D.M., Developmental studies of the compartmentalization of water and electrolytes m the choroid plexus of the neonatal rat brain, Brain Research, 116 (1976) 35- 48. 9 Kozlowski, G.P., Neurosecretory supply to extrahypothalamic structures: choroid plexus, circumventrieular organs, and limbic system. In E.M. Rodriguez and T.B. van Wimersma Greidanus (Eds.), Frontiers of Hormone Research ( Cerebrospinal fluid (CSF) and Peptide Hormones ~. Vol. 9, Karger, Basel, 1982, pp. 105-118. 111 Levine, S.. Degeneration of choroid plexus epithelium induced by some tertiary amines. In L. Roizin. H. Shiraki and N. Crcevic (Eds.), Neurotoxicology, Raven Press, New York, 1977, pp. 419-425. 11 Lindvall, M., Fluorescence histochemical study on regional differences in the sympathetic nerve supply of the choroid plexus from various laboratory animals, (Tell Tissue Res.. 198 (1979) 261-267, 12 Lindvall, M,, Edvinsson. L. and Owman, C., hlistochemical study on regional differences in the cholinergic nerve supply el the choroid plexus from various laboratory animals, Exp. Neurol., 55 (1977) 152-t59. 13 Masuzawa, T.. Saito, T. and Sate, F., Cytochemical study on enzyme activity associated with cerebrospinal fluid secretion in the choroid plexus and ventricular ependyma. Brain Research, 222 (198l) 3(19-322, 14 Miwa, S.. Inagaki, C., Fujiwara, M. and Takaori, S., Na-, K-, Mg- and HCO3-adenosine triphosphatases in the rabbit brain choroid plexus, Jap, ,l Pharmacol, 3(! (1980) 337-345. 15 Murphy, V., Sodium-Hydrogen Exchange in the Rat Choreid Plexus, Ph.D. thesis, University of Utah, 1984

16 Napoleone. P.. Sancesafio, G. and Amenia. t ' lndoleammergic innervation of rat choroid plexus: a lluorescenec his tochemical study, NeuroscL Lelt.+ 34 <19821 14';- t47. 17 Pershing. L,K and Johansom ('.[!., ,\cid~si+, induced e n hanced activity of the Na-K L,xchangc pump m the in rive choroid plexus: an untogcnctic analysis el 0~ssiblc r+)le in cerebrospinal fluid pH homcosla~is, ! ~'c',,.,hcm ?,~ 119821 322-3?2. 18 Pollav. M,, Transport mechanisms in the ~:hort~id plexus. t;?,d. Prec. fi'd. Am. Soc. Exp [?,~o1.,33 (19-:41 2(11~4=20(~9. 19 Quay, W,B., Regional differences in metabolism and composition of choroid plexuses. Brain Rcs'ea;~t~, ? ([tJ¢~0~ 378-389. 211 Qua}', W.B., Regional and quantitative dlflcrcnccs in lhc postweaning development el choroid plexuses in the ~at brain, Brain Research, 36 119721 37--45. 21 Quay, W,B., Regional differences in postwcaning growth by choroid plexuses as affected by dietary sal! deficiencies, Am. J. Anat.. 134 (19721 59-7!t 22 Rodriguez, E.M.. The cerebrospinal fluid as a pathway in neuroendocrine integration. J. Endear. "71 (19761 4~17-443 23 Scott, D.E., Paull, W.K. and Dudley, ( ; K , , A comparative scanning electron microscopic analysis of the human cerebral ventricular system. I. The third ventricle. Z. Zellfi)rsch.. 132 (19721 203-215. 24 Smith, Q.R. and Johanson, ( . E . , Effect of ouabain and po. tassium on ion concentrations in the choroidal epithelium, Ant. J. Physiol., 238 (1980) F399-F406. 25 Smith, Q,R. and Johanson. C.E., Eftiect of carbonic anhydrase inhibitors and acidosis on choroid plexus epithelial cell sodium and potassium, J Pharrnacol. Exp Ther., 215 (19811) 673-6811. 26 Smulders, A.P. and Wright, E.M., Role o1 choroid plexus m transport of melatonin between blood and brain. Brain Regearch. 191 11980) 555-558. 27 Suzuki, Y.and Ito, T., Reserpine-induced glycogen accumulation in the epithelial cells of the mouse choroid plexus. Brain Research, 70 119741 113- 122. 28 Wright, E.M., Accumulation and transport of amino acids by the frog choroid plexus, Brain Research. 44 /19721 207 -219. 29 Wright, E.M., Transport processes in the formation of the cerebrospinal fluid, Rev. Physiol. Biochem. Pharmacol.. 83 (1978) 1-34. 311 Zeuthen, T. and Wright, E.M., Epithelial potassium transport: tracer and electrophysiological studies in choroid plexus, J. Mernbr. Biol., 60 ( 198 t ) 105-128