Metabolic Alkalosis in the - JCI

Metabolic Alkalosis in the Rat

EVIDENCE THAT REDUCEDGLOMERULARFILTRATION

RATHERTHAN ENHANCEDTUBULAR BICARBONATEREABSORPTION

IS RESPONSIBLE FOR MAINTAINING THE ALKALOTIC STATE

MARTIN G. COGANand FU-YING Liu, Department of Medicine, University ofCalifornia, San Francisco, California 94143

A B S T R A C T Maintenance of chronic metabolic al-kalosis might occur by a reduction in glomerular fil-tration rate (GFR) without increased bicarbonate reab-sorption or, alternatively, by augmentation of bicar-bonate reabsorption with a normal GFR. Todifferentiate these possibilities, free-flow micropunc-ture was performed in alkalotic Munich-Wistar ratswith a glomerular ultrafiltrate total CO2concentrationof 46.5±0.9 mM(vs. 27.7±0.9 mMin controls). Al-kalotic animals had a markedly reduced single neph-ron GFRcompared with controls (27.4±1.5 vs. 51.6±1.6nl/min) and consequently unchanged filtered load ofbicarbonate. Absolute proximal bicarbonate reabsorp-tion in alkalotic animals was similar to controls (981±49vs. 1,081±57 pmol/min), despite a higher luminal bi-carbonate concentration, contracted extracellular vol-ume, and potassium depletion. When single nephronGFRduring alkalosis was increased toward normal byisohydric volume expansion or in another group byisotonic bicarbonate loading, absolute proximal bicar-bonate reabsorption was not substantially augmentedand bicarbonaturia developed. To confirm that a fallin GFR occurs during metabolic alkalosis, additionalclearance studies were performed. Awake rats werestudied before and after induction of metabolic al-kalosis associated with varying amounts of potassium

Portions of this work were presented at the National Meet-ings of the American Federation for Clinical Research, SanFrancisco, CA 25-27 April 1981 and Wash. DC, 7-10 May1982 (1981. Clin. Res. 29: 458a. (Abstr.) and 1982. Clin. Res.30: 456a. (Abstr.)) and at the VIII International Congress ofNephrology, Athens, Greece, 7-12 June 1982 (1981. Ab-stracts of the VIII International Congress of Nephrology. 31;34.).

Address all correspondence to Dr. Cogan.Received for publication 2 August 1982 and in revised

form 10 January 1983.

and chloride depletion. In all cases, the rise in bloodbicarbonate concentration was inversely proportionalto a reduction in GFR; filtered bicarbonate load re-mained normal. In conclusion, a reduction in GFR isproposed as being critical for maintaining chronicmetabolic alkalosis in the rat. Constancy of the filteredbicarbonate load allows normal rates of renal bicar-bonate reabsorption to maintain the alkalotic state.

INTRODUCTION

The pathogenesis of chronic metabolic alkalosis maybe divided into two phases (1). In the first, generationphase, there is net hydrogen ion loss from the body orbicarbonate gain so that plasma bicarbonate concen-tration and pH are increased. In the second, mainte-nance phase, all the filtered bicarbonate is reabsorbedand hence the alkalosis is sustained. This contrasts tothe normal renal response to an elevated bicarbonateconcentration in which excess bicarbonate is promptlyexcreted so that the plasma bicarbonate concentrationis quickly normalized (2).

Metabolic alkalosis is usually associated with potas-sium and chloride deficiency, which in turn might beresponsible for maintaining the alkalosis in two ways.To accommodate the increased filtered bicarbonateload, a proportional increase in bicarbonate reabsorp-tion might occur, predominantly in the proximal neph-ron. It is assumed that glomerular filtration rate (GFR)1would remain normal in this case. Alternatively, as theplasma bicarbonate concentration rises, maintenanceof metabolic alkalosis might be accomplished by a re-

' Abbreviations used in this paper: DOCA, deoxycorti-costerone acetate; EDD, electrolyte-deficient diet; FFDS,fat-free dry solid; GFR, whole kidney glomerular filtrationrate; SNGFR, single nephron GFR.

J. Clin. Invest. © The American Society for Clinical Investigation, Inc. - 0021-9738/83/05/1141/20 $1.00Volume 71 May 1983 1141-1160

1141

ciprocal reduction in GFR. In this case, there is noincrement in filtered bicarbonate load. Consequently,a normal (not enhanced) rate of proximal acidificationwould suffice to prevent excessive distal bicarbonatedelivery and bicarbonaturia.

Differentiation of these two pathophysiological al-ternatives rests in measuring absolute rates of bicar-bonate filtration and proximal reabsorption duringchronic metabolic alkalosis. Such measurements havenot been previously performed. Indeed, there is evendisagreement over whether fractional proximal bicar-bonate reabsorption during chronic metabolic alkalosisis increased (3), decreased (4), or normal (5).

The principal purpose of the free-flow micropunc-ture experiments in this study was to compare the sin-gle nephron glomerular filtration rate (SNGFR) andabsolute proximal bicarbonate reabsorption in ratswith chronic metabolic alkalosis accompanied by ex-tracellular volume (chloride) and potassium depletionto values in normal euvolemic animals. Another pur-pose was to examine the response of proximal acidi-fication to an increase in filtration rate. These latterexperiments attempted to define the mechanismwhereby isohydric expansion during metabolic alka-losis results in a bicarbonate diuresis and correction ofthe alkalotic state (6, 7).

The micropuncture studies revealed that SNGFRwas reduced during chronic metabolic alkalosis. Atcomparable filtered bicarbonate loads, absolute prox-imal bicarbonate reabsorption during alkalosis was nohigher than normal. Proximal bicarbonate reabsorp-tion did not rise when SNGFRwas caused to increase.

Since proximal bicarbonate reabsorption was notenhanced during metabolic alkalosis, the fall in filtra-tion rate appeared to be critical for maintaining thealkalotic state. Additional clearance studies weretherefore performed in awake rats to confirm that GFRfalls in chronic metabolic alkalosis.

The clearance studies confirmed the micropuncturefindings. As plasma bicarbonate concentration rose inconscious animals with varying amounts of potassiumand chloride depletion, there was a proportional de-crease in GFR. This inverse relationship continuedwhen the alkalosis was partially or fully repaired bypotassium and/or chloride repletion. Under all con-ditions of induction and repair of chronic metabolicalkalosis, filtered bicarbonate loads did not exceed nor-mal values. Thus, a fall in GFRpermitted normal ratesof renal acidification to maintain the alkalotic state.

METHODSMicropuncture protocols43 Munich-Wistar rats (Timco Breeding Laboratories,

Houston, TX) weighing 218±4 g were studied using free-flow micropuncture techniques.

Chronic metabolic alkalosis. To induce chronic meta-bolic alkalosis, animals were maintained for 11-14 d on astandard liquid electrolyte-deficient diet (EDD) of 40ml/d, supplemented with Na2SO4, 2.6 meq/d (3). They alsoreceived 80 mMNaHCO3drinking water and were injecteddaily with 0.5 mg deoxycorticosterone acetate (DOCA)(Organon Inc., W. Orange, NJ) intramuscularly. The dietwas withheld for 24 h before study to permit sulfate excre-tion. The first period of micropuncture followed surgicalpreparation of these animals. Surgically induced plasma vol-ume losses (8, 9) were replaced using homologous plasmaobtained from identically pretreated rats (10). It was foundby extensive preliminary experience that failure to replaceplasma losses precluded micropuncture because of the severevolume depletion sustained. The measured total CO2 con-centration of the infused plasma was 42.2±0.9 mM(n = 34).The infusion rate was 35 Ml/min for the 45-60 min of surgicalpreparation, for a total infusion of 1% body wt. The infusionrate was then reduced to 4 gl/min for the duration of thestudy. Using this protocol, plasma volume was maintainedconstant, as reflected by stability of the hematocrit: the he-matocrit immediately following arterial catherization, butbefore major surgery, was 49.8±0.5 vol%, while the hema-tocrit during the first period of micropuncture was 49.5±0.6vol% (n = 27).

Chronic metabolic alkalosis plus isohydric expansion.Following the first micropuncture period, 10 rats underwentisohydric expansion, similar to the protocol of Cohen (6).They received a 5% body wt/h infusion of a colloid-freesolution that approximated the plasma electrolyte compo-sition during alkalosis. It contained (in milliequivalents perliter): NaCl, 88; NaHCO3, 45; Na2HPO4, 4; KCI, 2; MgSO4,1; CaCl2, 1.8. The second period of micropuncture was per-formed in the last half of the second hour of the infusion,thus after 7.5-10% body wt total expansion. To define theindependent role of extracellular volume status on proximalacidification during expansion, seven of these rats subse-quently underwent aortic constriction while volume expan-sion continued. The suprarenal aorta was constricted to amean arterial pressure of 74±1 mmHgto return SNGFRtothe level obtained during the first, preexpansion period. Thethird period of micropuncture was then performed followinga 0.5-h stabilization period.

Another group of six animals were isohydrically expandedwith homologous plasma to increase SNGFRfollowing thefirst micropuncture period. At the same SNGFRfollowingplasma expansion compared with colloid-free expansion, theeffect of differences in peritubular Starling forces on prox-imal anion reabsorption could be examined. The second pe-riod commenced following 1.5 h of plasma expansion at 2%body wt/h, thus, after a total of 3-4% body wt plasma ex-pansion.

Chronic metabolic alkalosis plus acetazolamide. In sixanimals, the second period of micropuncture followed 0.5h of carbonic anhydrase inhibition with acetazolamide. Ac-etazolamide was given in a loading dose of 50 mg/kg, andthen infused at 50 mg/kg per h in 125 mMNaHCO3/25 mMKHCO3at 0.1 ml/min. When carbonic anhydrase is inhib-ited, normal anion concentration gradients collapse (i.e., tu-bular fluid/plasma anion concentrations approximate unity)(11). Under these conditions, both carbonic anhydrase-in-sensitive proximal bicarbonate reabsorption as well as non-diffusional proximal chloride reabsorption can be quanti-tated (12).

Chronic metabolic alkalosis plus severe aortic constric-tion. To ascertain whether the end-proximal luminal totalCO2 concentration in the animals with chronic metabolic

1142 M. C. Cogan and F.-Y. Liu

alkalosis had achieved a stable, steady-state value or whethera further reduction in concentration were possible, fivechronically alkalotic animals were prepared for micropunc-ture as above. They were micropunctured during a singleperiod only, after severe aortic constriction had been usedfor 20-30 min. Mean arterial pressure was reduced to 64-71mmHgto markedly diminish SNGFRand tubular flow rate.

Acute metabolic alkalosis. This group was studied in or-der to examine proximal reabsorption: (a) at a higher SNGFRthan could be obtained in chronic metabolic alkalosis evenfollowing volume expansion; and (b) in an alkalotic statewith normal potassium stores. 16 rats received an infusionof isotonic bicarbonate (125 mMNaHCO3/25 mMKHCO3)at 5%body wt/h. Micropuncture began 1.5 h following com-mencement of infusion, for a total of 7.5-10% body wt ex-pansion.

Clearance protocols95 female Sprague-Dawley rats weighing 200-220 g were

studied in the awake state. Each animal was studied twiceby clearance techniques: when normal and following theinduction of metabolic alkalosis (groups I-IV); or when al-kalotic and following the partial or complete repair of al-kalosis (groups V-VIII).

Animals in the first four groups were used to study theinduction of metabolic alkalosis. They were allowed freeaccess to Purina rat chow (Ralston Purina Co., St. Louis, MO)and water before the first clearance study. Following thefirst clearance period of these normal rats, the 40 ml/d stan-dard EDDwas provided, divided into twice daily feedings.The diet was supplemented with various salts to generate

metabolic alkalosis (groups II-IV) or to maintain a normalcontrol state (group I), as shown in the top of Table I. Halfof the animals in each group (except IVc) received 80 mMNaHCO3drinking water while the other half received dis-tilled water. During this period, rats were housed in meta-bolic cages under balance conditions. Urine was collectedunder oil with thymol as a preservative.

Group I: control. 20 rats were maintained for 9 d (groupIA) or for 21 d (group IB) on a normal intake of NaCl (1.5meq/d) and KCI (1.5 meq/d).

Group II: severe K and Cl deficiency. 20 rats were givenNa2SO4 (2.6 meq/d) and were injected intramuscularly dailywith 0.5 mg DOCAin sesame oil for 8 d. This diet was usedto induce a high degree of K and Cl depletion and severemetabolic alkalosis, as in the rats used for micropuncture.

Group III: moderate K and Cl deficiency. K replaced Nain the diet (K2SO4, 2.6 meq/d) and in the drinking water(80 mMKHCO3) of 10 rats. Less Cl deficiency also ensued(13). This diet was used to ameliorate the K depletion ingroup II.

Group IV: K deficiency. In 25 rats, dietary NaCl withDOCA(group IVA) or without DOCA(groups IVB and IVc)was used to induce K depletion without Cl depletion. Thesubgroups with pure diet-induced K deficiency, unenhancedby DOCA, necessitated a longer treatment period. Theselatter diets were designed to be equivalent to group I on thebasis of similar Na content (group IVB) or Cl content(group IVc).

The second clearance period was performed on each an-imal following completion of the above protocols.

The next four study groups were used to study the repairof metabolic alkalosis. In each group of five rats, K and Cl

TABLE IDietary Protocols for Clearance Studies

No. Quantity No. receivingof days HCO- in

Group Treatment n on diet Na' K+ Cl- So4 drinking water

meq/dInduction

I Control A EDD, NaCl, KCI 10 9 1.5 1.5 3.0 - 5B EDD, NaCi, KCI 10 21 1.5 1.5 3.0 - 5

II Severe K and Cl EDD, Na2SO4, DOCA 20 8 2.6 - - 2.6 10deficiency

III Moderate K and Cl EDD, K2SO4, DOCA 10 10 - 2.6 - 2.6 5deficiency

IV K deficiency A EDD, NaCi, DOCA 10 10 1.3 - 1.3 - 5B EDD, NaCl 10 21 1.3 - 1.3 - 5C EDD, high NaCI 5 21 2.6 - 2.6 - 5

RepairInitial preparation EDD, Na2SO4, DOCA 20 7 2.6 - - 2.6 20

V No K and Cl EDD, Na2SO4, DOCA 5 8 2.6 - 2.6 5replacement

VI K replacement EDD, K2SO4, DOCA 5 7 - 2.6 - 2.6 5

VII Cl replacement EDD, NaCI, DOCA 5 8 2.6 - 2.6 - 5

VIII K and Cl replacement EDD, KCI, DOCA 5 7 - 2.6 2.6 5

Metabolic Alkalosis in the Rat 1143

depletion with alkalosis was first induced with Na2SO4 andDOCA(the same protocol used in group II). The initial base-line clearance period was then performed. A second diet wasthen instituted for 7-8 d in order to continue the alkaloticstate (group V), to partially repair the alkalosis with K orCl (groups VI and VII), or to fully repair the alkalosis withKCI (group VIII), as shown in the bottom of Table 1. Allgroups received 80 mMNaHCO3drinking water.

Group V: no K or Cl replacement. The previous Na2SO4and DOCAregime was continued.

Group VI: K replacement. K replaced Na in the diet(K2SO4) and drinking water (KHCO3).

Group VII: Cl replacement. Cl replaced the S04 in thediet.

Group VIII: K and Cl replacement. KCI replaced Na2SO4.The second clearance period was performed following com-

pletion of the above protocols.

ProceduresMicropuncture techniques. The free-flow micropuncture

methodology from this laboratory has been previously pub-lished (10-12).

Clearance techniques. For the first clearance period,each rat was anesthetized with 50 mg/kg body wt Brevitali.p. (methohexital sodium, Eli Lilly & Co., Indianapolis, IN).The femoral vein and artery were catheterized (PE-50). Therat was allowed to awaken (30-90 min). It was then placedin a plexiglass cage small enough to prevent catheter dis-lodgement, as previously described (8, 9). Care was takenduring the rest of the experiment to avoid disturbing theanimal. When the animal was fully conscious, a [methoxy-3H]inulin (Nuclear Chicago Corp., Chicago, IL) infusion wasbegun. Following a prime of 6 ,ACi, a maintenance infusionof 16 ACi/h (in bicarbonate Ringer's solution at 0.8 ml/h)was started. 1 h elapsed for equilibration. Five consecutiveurine collections were then performed with intermittent ar-terial blood sampling for inulin, protein, Na, K, and C1 con-centrations. At the end of the clearance studies, blood wasobtained for blood gas determination. Finally, 0.75 gCi 1251human serum albumin was injected for plasma volume mea-surement (8, 9).

Each rat was then reanesthetized to remove the cathetersand to suture the femoral incision. The animal was placedin a metabolic cage for its assigned 1-3-wk dietary regime.

For the second clearance period, the same procedurescited above were repeated, except that the contralateral fem-oral artery and vein were used for cannulation.

Following the second clearance study, a small piece ofmuscle from the thigh of each animal was excised for sub-sequent determination of muscle [K]. To define muscle [K]for the first clearance period of each group, two separatesets of rats (n = 10 each) were utilized. They were main-tained on identical initial diets to those used in the induction(groups I-IV) or repair (groups V-VIII) groups before ex-cision of muscle.

AnalysisMeasurements were made of [3H]inulin concentration by

liquid scintillation counting (Tri Carb 460C, Packard In-strument Co., Downers Grove, IL), of plasma and urine so-dium and potassium concentration by flame photometry(model 343, Instrumentation Laboratory, Inc., Lexington,MA), of plasma and urine chloride concentration ampero-metrically (Buchler-Cotlove chloridometer, Buchler Instru-

ments, Inc., Fort Lee, NJ), of plasma osmolality by vaporpressure osmometry (Wescor, Inc., Logan, UT), and ofplasma protein concentration by refractometry (AmericanOptical Co., Keene, NH). Arterial pH and Pco2 were mea-sured by a blood gas analyzer (model 165 Corning Medicaland Scientific, Corning Glass Works, Medfield, MA). TotalCO2concentration of an aliquot of 20-40 nl of tubule fluid,plasma, or urine in the micropuncture studies was measuredby microcalorimetry (Picapnotherm) (14). Within the phys-iologic pH range, the total CO2 in a sample represents bi-carbonate plus dissolved CO2 gas. To document bicarbona-turia, urine total CO2 was measured using a Natelson mi-crogasometer (model 600, Scientific Instruments, Inc.,Springfield, MA) from two rats in each of the groups studiedby clearance techniques that were receiving bicarbonate inthe drinking water. Tubule fluid chloride concentrationwas measured by the electrometric titration method ofRamsay (15).

For the muscle potassium measurements, the muscle wasminced and then dried at 80°C for at least 48 h to constantweight. Fat-free dry solid (FFDS) weight was defined afterfat extraction with ether and subsequent redrying and re-weighing. Finally, 0.46 N nitric acid digestion was carriedout for 48 h with constant rotation of the samples. The su-pernatant [K] was then measured.

To ascertain whether morphological changes in the prox-imal tubule had occurred due to kaliopenia, kidneys fromalkalotic animals (n = 2) were compared with normal con-trols (n = 2). Kidneys were fixed in Bouin's solution, stainedwith hematoxylin and eosin, and examined by light micros-copy.

CalculationsSNGFRand GFRwere estimated from the single nephron

and whole kidney inulin clearances. Absolute and fractionalrates of water, total C02, and chloride reabsorption and thelogarithmic mean luminal total CO2concentration were cal-culated as previously described (10-12).2 The mean lengthof the accessible proximal convoluted tubule in the Munich-Wistar rat is 4.5-5.0 mm.

Results are expressed as mean±SEM. Statistical signifi-cance was assessed using the paired t test for results obtainedin the same animal or the unpaired t test for comparisonsbetween groups. Multiple linear regression analysis was per-formed using standard techniques (16).

RESULTSProximal bicarbonate reabsorption during meta-

bolic alkalosis. As shown in Table II (line 1), animals

2 Occasional rats had insufficient surface glomeruli fordirect glomerular ultrafiltrate analysis. In such cases, theglomerular ultrafiltrate total CO2 concentration was esti-mated by multiplying the measured plasma water total CO2concentration (corrected for protein content) by an effective"Donnan factor" of 1.06. In other experiments when Bow-man spaces were present, measurements during alkalosisverified that the ratio of the glomerular ultrafiltrate totalCO2 concentration to the simultaneously obtained plasmawater total CO2concentration was 1.06±0.01 (n = 52). Thiscoefficient is slightly higher than previously reported in nor-mal animals, 1.05±0.01 (10), perhaps due to the increasedanionic protein charge density during alkalosis. The com-parable factor for chloride during alkalosis was 1.05±0.01(n = 48).

1144 M. G. Cogan and F.-Y. Liu

2- -oz +1 +I QE -4 o Z

9_

ooI)

+1-H +1O +l

rw el)

_ _

6 0 o

t- Go-

+1 +1800 0)0c

oa-Z +l

E c4

~~~~+IV) +IC)n +1

+1 o Z o Zo o eO t

eq 1-

q CI

C +I o +1

co agoOo Z0 00 00

+e0 +1 Z ZCq -

- eq -

+1 +l (A +l CA (n16 1 t-

141-

W eq co CZO+1 +I d) +1cn V)

2 CO- -O COz

C-: +1

C-4£ 0-

Ci -

-- +1 8 + o16 CO

CD- 0q o06 o

+1+1j + +1°kO V VV 0 V

It W C4 V C C

0--

11

:C 0

C) c

esw*

r11

£-

c

Cd

c

0

+.

4.

(A

+_

C4 Ci Il

+1 Z +1 Z +1'IT0 16

eq q eq

+1 c +l cn +1,,M Z CS ;Z _-to _

to

+I cn +I V)+0) +100CO

C

Ce+10)Ce

Ti 8 o1 z 3C?6 0

9 6

0 ZC C

+1 +1

IV

_o11

0

._

xC

CU

+

0

114)

CU

N

CU._

0

co-4

Co

._L

C2 a

E_V)

._ .~co

CUCUV )-d ._

co

o

> >

~C

X CdO E =

4 6. M.

4) °4.4.

ZQ 00

0; Cd WI

CU 1 ~Q +c_ U4)

d-

Metabolic Alkalosis in the Rat 1145

-a0

co;

wz

2-.

6

8R

co

.b

0n

4)

VC3Q

0.

0o

4-

4)

+1 U.) +18 8 + q +IC/) +10Z v O0 0Z t

I

I

I

I

maintained on the electrolyte-deficient diet supple-mented with Na2SO4 and given DOCAwere alkalotic(arterial pH 7.57±0.01 and plasma total CO2concen-tration 41±1 mM) and potassium depleted (plasmapotassium 2.1±0.1 meq/liter).

As shown in Table III (lines 1 and 8), the alkaloticanimals had a markedly reduced SNGFRcomparedwith euvolemic (similarly plasma repleted) normalanimals (10) (27.4±1.5 vs. 51.6±1.6 nl/min). GFRwasalso greatly decreased (0.45±0.05 ml/min) comparedwith normal (1.42±0.03 ml/min) (Table IV). Becauseof the reciprocal differences in Bowman's space totalCO2 concentration (46.5±0.9 in alkalotic rats vs.27.7±0.9 mMin normal controls), single nephron fil-tered total CO2 loads in the two groups were similar.Despite significant extracellular volume and potassiumdepletion, absolute and fractional proximal total CO2reabcorption in the alkalotic animals (981±49 pmol/

min and 0.77±0.01) were no higher than normal values(1,081±57 pmol/min and 0.77±0.02). As illustrated inFig. 1, absolute proximal total CO2reabsorption in thealkalotic animals (open circles) overlapped the normalrange (shaded area) as a function of filtered total CO2load. Distal total CO2 delivery rates in the alkaloticand normal euvolemic groups were thus similar(297±27 vs. 329±36 pmol/min).

That proximal bicarbonate reabsorption was notstimulated during chronic metabolic alkalosis was sur-prising considering that the mean luminal bicarbonateconcentration (Fig. 2) was substantially higher (32mM, open circle) than during normal euvolemia (17mM, closed triangle). In previous studies, proximalbicarbonate reabsorption was stimulated when themean luminal bicarbonate concentration increased,either as a result of increased flow rate (shown by thetriangles and squares in Fig. 2 [10]), or increased lu-

TABLE IIIFiltration and Proximal Reabsorption of H20, Total CO2, and Chloride

Bowman's space Filtered load End-proximal

SNGFR [tCO] [CI] tCO2 Cl V [tCOJ [Cl]

nl/min mM meq/liter pmol/min peq/min nl/min mM meqlliter

Chronic metabolic 27.4±1.5° 46.5±0.9 99.0±2.0 1,276±69 2,785±184 13.2±0.9 21.5±1.1 123.3±2.1alkalosis (n = 22)

+ Isohydric expansion 35.8±2.4 43.1±1.3 103.4±1.7 1,547±109 3,702+259 22.5±1.8 25.7±1.4 120.3±2.5(n = 10) P < 0.01t 0.001 0.01 0.05 0.005 0.001 NS NS

+ Aortic constriction 22.6±1.2 41.0±1.3 106.4±2.2 922±36 2,418±169 13.1±1.3 16.0±1.5 127.2±2.8(n = 7) P < 0.001 0.05 0.001 0.001 0.001 0.001 0.001 0.05

P NS§ 0.005 0.001 0.005 NS 0.025 0.001 0.05

+ Plasma expansion 37.7±2.1 41.2±1.9 106.0±3.4 1,543±86 3,988±218 20.7±1.4 21.4±2.2 128.8±4.0(n = 6) P < 0.0051 0.05 NS 0.01 0.01 0.001 NS NS

+ Acetazolamide 27.2±2.1 46.1±2.1 98.1±2.9 1,241±80 2,695±260 20.1±1.8 49.1±2.4 105.7±3.7(n = 6) P < NSt NS NS NS NS 0.05 0.001 0.005

+ Severe aortic 11.1±1.5 40.6±1.9 100.1±4.4 444±53 1,126±169 4.0±0.6 7.6±1.0 143.5±4.2constriction (n = 5)

Acute metabolic alkalosis 46.4±1.9 47.6±1.0 89.4±0.9 2,250±115 4,075±251 29.7±1.6 39.1±1.0 103.2±1.2(n= 16)

Normal: euvolemia 51.6±1.6 27.7±0.9 - 1,412±82 - 31.8±1.1 10.3±0.9(n = 12)11

Normal: hydropenia 24.1±1.4 25.0±0.9 116.1±1.4 605±40 2,736±155 11.1±0.5 4.4±0.5 140.1±2.0(n = 11)T

tCO2, total CO2; V, end-proximal flow rate.Mean±SEM.

I P values for second period compared with first period (chronic metabolic alkalosis).§ P values for aortic constriction period compared with isohydric expansion period.11 Normal euvolemic values for comparison from reference (10).T Normal hydropenic values for comparison from reference (12).

1146 M. G. Cogan and F.-Y. Liu

minal bicarbonate concentration when microperfusiontechniques were used (17, 18). Lack of stimulation ofacidification by an increase in luminal buffer concen-tration during alkalosis might have been due to analteration in the luminal gradient attainable for totalCO2concentration or pH. According to this hypothesis,the measured end-proximal total CO2 concentrationof 21.5±1.1 mMmay have represented a rate-limiting,minimal luminal bicarbonate concentration, substan-tially above the value that normal hydropenic animalscan achieve of 4.4±0.5 mM(Table III, line 9) (12).Severe aortic constriction was therefore used duringchronic metabolic alkalosis and SNGFRfell to 11.1±1.5nl/min. The gradient limitation hypothesis was re-jected by finding that the end-proximal total CO2con-centration could be reduced to 7.6±1.0 mMunderthese conditions (Table III, line 6).

In addition, there was a remote possibility of phys-

ical damage to the proximal nephron induced by po-tassium deficiency. This was excluded by light mi-croscopy. Proximal tubules of alkalotic rats were mor-phologically unchanged from normal controls, inaccord with others (13).

The bicarbonate reabsorptive process during met-abolic alkalosis was further characterized by assessingthe response to carbonic anhydrase inhibition (TableIII, line 5). As had been found previously in normalanimals (11), bicarbonate reabsorption during acet-azolamide administration was inhibited by 80% in al-kalosis. Absolute proximal total CO2 reabsorption wasalso similar in magnitude (268±31 pmol/min) to thatobserved previously (291±16 pmol/min) when a com-parable filtered total CO2load was achieved by plasmaloading normal animals during carbonic anhydrase in-hibition. In each case, due to either increased filteredbicarbonate concentration or flow rate during carbonic

TABLE III (Continued)

Distal delivery Absolute proximal reabsorption Fractional proximal reabsorption

tCO, Cl tCO, Cl H20 tCO, Cl f1,0

pmol/min peq/min pmol/min peq/min ni/min

297±27 1,650±132 981±49 1,088±75 14.3±0.8 0.77±0.01 0.40±0.01 0.52±0.01

582±57 2,700±235 961±64 979±85 13.4±0.9 0.63±0.02 0.27±0.02 0.38±0.020.001 0.001 NS NS NS 0.001 0.001 0.001

228±33 1,665±171 694±26 752±52 9.5±0.4 0.78±0.03 0.32±0.03 0.43±0.030.001 0.001 0.01 0.05 0.025 0.01 0.025 NSNS 0.005 0.001 0.025 0.001 NS 0.005 0.001

434±56 2,686±179 1,098±52 1,347±68 17.1±0.9 0.71±0.02 0.34±0.02 0.44±0.020.05 0.001 0.025 0.05 0.05 0.05 0.05 0.005

973±64 2,092±235 268±31 542±73 7.2±0.6 0.22±0.02 0.20±0.02 0.27±0.020.001 0.01 0.001 0.025 0.005 0.001 0.025 0.005

28±4 546±107 416±50 550±112 7.1±1.0 0.94±0.01 0.47±0.05 0.64±0.02

1,170±76 2,949±232 1,090±64 1,150±82 16.9±0.8 0.49±0.02 0.28±0.02 0.37±0.02

329±36 - 1,081±57 - 19.0±1.1 0.77±0.02 0.37±0.01

48±6 1,639±151 557±38 1,214±93 13.0±1.0 0.92±0.01 0.44±0.01 0.54±0.01

Metabolic Alkalosis in the Rat 1147

TABLE IVGFRand Urinary Electrolyte Excretion Rates in Micropuncture Studies

GFR V U^V U,V UaV UcxV

mi/mn $d/min nq/min nmo/min

Chronic metabolic alkalosis 0.45±0.05° 1.4±0.2 66±11 29±5 17±3 2±1(n = 22)

+ Isohydric expansion (n = 10) 0.63±0.04 12.6±2.3 2,941±495 15±3 1,769±415 593±111P < 0.0011 0.005 0.001 NS 0.025 0.005

+ Aortic constriction (n = 7) 0.39±0.02 1.8±0.3 285±96 7±1 67±18 16±9P < 0.025t 0.05 0.025 NS 0.05 0.025P < NS§ 0.05 0.05 NS NS NS

+ Plasma expansion (n = 6) 0.65±0.08 11.7±6.7 2,510±1,392 19±5 1,604±972 655±437P < 0.005t 0.05 0.05 NS 0.05 0.05

+ Acetazolamide (n = 6) 0.54±0.05 54.5±2.3 11,257±524 607±135 222±79 8,419±1,083P NSt 0.001 0.001 0.05 0.05 0.001

Acute metabolic alkalosis 1.07±0.03 44.6±5.8 8,104±1,095 3,599±205 354±83 7,423±679(n = 16)

V, urinary volume excretion rate; UN.V, urinary sodium excretion rate; UKV, urinary potassium excretion rate; UcOV, urinary total CO2excretion rate; UcV, urinary chloride excretion rate.

Mean±SEM.I P values for second period compared with first period (chronic metabolic alkalosis).§ P values for aortic constriction period compared with isohydric expansion period.

*oO

00

0

500 1000FILTERED TOTAL C02

1500 2000LOAD (pmol/min)

FIGURE 1 Absolute proximal total CO2 reabsorption as a function of filtered total CO2 loadduring chronic metabolic alkalosis. Shaded area encompasses the previously described proximaltotal CO2reabsorption by normal euvolemic and isohydrically expanded animals (10). Proximalbicarbonate reabsorption during metabolic alkalosis (circles) was not stimulated over normalvalues.

1148 M. G. Cogan and F.-Y. Liu

1500-N

0

I00< c

c0 -,

Oz\

j0- 50

Jo 1000

< c

OZ

CL Uua a 500-D

It

j OnJ uO

V) <D Lfi

<t O0

0

1500

1000 F-

500H

0 10 20 30Luminal Log Mean [Total C02] (mM)

40

FIGURE 2 Absolute proximal total CO2 reabsorption as a function of luminal logarithmic mean

total CO2 concentration. Values are shown for volume contracted (A) and euvolemic normalanimals (A), for contracted (l) and euvolemic acidotic animals (U) (10), and for animals withchronic metabolic alkalosis (0).

anhydrase inhibition, total CO2 reabsorption was >70%higher than the normal hydropenic value during car-

bonic anhydrase inhibition of 158 pmol/min (11).These higher rates surpassed reasonable estimates ofthe proton secretory rate due to the uncatalyzed rateof hydration and hydroxylation of CO2 (11).

The next series of experiments examined the re-

sponse of proximal acidification to an increase in flowrate and extracellular volume expansion. Cohen (6) hasshown that isohydric expansion of alkalotic dogs in-duced bicarbonaturia and repaired the alkalosis. It isunknown whether this reduction in fractional renalbicarbonate reabsorption was a consequence of a

change in proximal acidification. 10 alkalotic animalswere expanded with a colloid-free isohydric solutionin a protocol similar to Cohen's. Other alkalotic ani-mals were isohydrically expanded with homologousplasma. In both cases, SNGFRrose by '30%, but therewas little change (-2 to 12%) in absolute total CO2reabsorption in the two groups, as shown in Table III(lines 2 and 4). Hence, fractional total CO2 reabsorp-tion was diminished and distal total CO2 delivery in-creased (Fig. 3). Similar to the studies of Cohen (6),bicarbonaturia developed. Whole kidney bicarbonateexcretion was substantial (Table IV), representing2±1% of the filtered load in both groups. The glo-merular ultrafiltrate total CO2 concentration fell by

- 10%.

Enhanced distal total CO2delivery during isohydricexpansion might have been the consequence of an in-creased filtered total CO2 load in a setting of reab-sorptive saturation. Alternatively, volume expansionmay have independently inhibited total CO2 reab-sorption that otherwise would have paralleled thehigher load. If the latter explanation were true, a sub-sequent reduction in filtered total CO2 load duringpersisting expansion should be associated with contin-ued suppression of proximal total CO2 reabsorption.Hence, at similar flow rates, fractional proximal totalCO2 reabsorption postexpansion would be less thanpreexpansion. Therefore, aortic constriction was usedto return SNGFRto the preexpansion level (22.6±1.2vs. 24.0±1.4 nl/min in the seven paired studies), whileisohydric expansion continued. Because of the inter-vening reduction in glomerular ultrafiltrate total CO2concentration, the filtered total CO2load during aorticconstriction was somewhat less than the preexpansionlevel even though SNGFRwas similar (Table III, line3). The absolute rate of total CO2 reabsorption was

therefore less in the third period than in the first. Theimportant observation, however, is that glomerulotu-bular balance was restored: fractional proximal totalCO2 reabsorption and distal total CO2 delivery re-

turned to preexpansion levels (Fig. 3). In addition,bicarbonaturia disappeared (Table IV), despite per-

Metabolic Alkalosis in the Rat 1149

c0

0.-h-00e00

0~.ir

C~

F- .7-(o'C

SE

0

0~

-50

N0u-J410I.-

Co EI -

z

40-

201

401-

30h-cIE

0zU,

20

10

* Lp< 0.0I

-4

0 -0 m

0.8 r' A00 a

z

en

600 1-

_ >

400 -

3o5 0

-0

m

20

m

Vi

Chronic lsohydric lsohydric Chronic Isohydric lsohydricMetabolic Expansion Expansion Metabolic Expansion ExpansionAlkolosis + Alkalosis +

Aortic AorticConstriction Constriction

FIGURE 3 Paired changes in Bowman's space total CO2 concentration, SNGFR, fractionalproximal total CO2reabsorption and distal total CO2delivery during chronic metabolic alkalosis,following isohydric expansion, and with aortic constriction during persistent expansion.

sisting volume expansion (total 15% body wt expan-sion). These findings suggest that isohydric expansionhad diminished fractional proximal total CO2 reab-sorption and had enhanced distal total CO2 deliveryas a consequence of increasing flow rate in the settingof static reabsorption. An inhibitory effect on net bi-carbonate transport by the increased extracellular vol-ume status per se was not observed.

To examine whether a further increase in filteredtotal CO2 load and in mean luminal total CO2 con-centration would stimulate total CO2 reabsorption,normal animals were infused with isotonic bicarbon-ate. A higher SNGFR(46.4±1.9 nl/min) and hencefiltered total CO2 load could be achieved in this acutemodel of metabolic alkalosis than had been possiblein chronic metabolic alkalosis even after isohydric ex-pansion (Table III, line 7). However, little augmen-tation in absolute total CO2 reabsorption (1,090±69pmol/min) was found, as shown in Fig. 4A (closedsquare), even at the markedly higher filtered load andmean luminal bicarbonate concentration of 43.2 mM.Thus, the fraction of total CO2 reabsorbed in the su-perficial proximal convoluted tubule fell to 0.49±0.02.

The very high distal bicarbonate delivery was asso-ciated with substantial bicarbonaturia (Table IV).However, it should be noted that total renal bicarbon-ate reabsorption was 0.86±0.01 of the filtered load(Fig. 4B, closed square), indicating a large bicarbonatereabsorptive capacity by nephron segments distal tothe superficial proximal convoluted tubule and/or bydeeper nephrons. Nevertheless, a qualitative relation-ship in all groups between distal bicarbonate deliveryand urinary bicarbonate excretion clearly existed, ashas been previously found (10, 11).

Proximal chloride and water reabsorption duringmetabolic alkalosis. In previous studies, absoluteproximal chloride reabsorption was divided into twocomponents: a passive component, dependent princi-pally on the magnitude of the lumen-to-blood chlorideconcentration gradient; and an active transport (orpossibly convection) component. The active compo-nent can be quantitated when the chloride gradientis minimized by the use of carbonic anhydrase inhi-bition (12). As shown in Fig. 5A, absolute proximalchloride reabsorption during metabolic alkalosis (closedsymbols) was similar to normal hydropenic controls

1150 M. G. Cogan and F.-Y. Liu

I I I

I I lI I II I I

PROXIMALTUBULEA

/ .

. . i. .-

500 1000 1500 2000

Filtered Total C02 Load (pmol/min)

-202500

c

E

0

E

0

G.

0

(o.D0

v

cren

-60

Z0

cr.-

.D

4~

40

30

WHOLEKIDNEY B

40 50

Filtered Total CO2 Load (Lnrpl/min)

FIGURE 4 Absolute proximal (top panel) and whole kidney (bottom panel) total CO2 reab-sorption as a function of filtered total CO2load in metabolic alkalosis. Represented are animalsduring acute metabolic alkalosis (U); during chronic metabolic alkalosis (0); and during chronicmetabolic alkalosis with subsequent plasma expansion (A), with subsequent isohydric colloid-free expansion (V), with isohydric expansion and aortic constriction (®), or with severe aorticconstriction (X). Shaded area is the normal proximal and whole kidney total CO2 reabsorptionrates (10).

(open symbols) (12) before (circles) and after (dia-monds) acetazolamide. It is important for this analysisthat comparisons be made at similar flow rates becauseof the flow dependence of both total and acetazol-amide-insensitive absolute proximal chloride reabsorp-tion (11). Mean values for SNGFRin the four groups

ranged from 24.1 to 27.4 nl/min.The passive diffusional chloride reabsorptive com-

ponent appeared normal. Both the principal electro-chemical driving force, the chloride gradient (-25meq/liter), as well as the proximal chloride perme-ability (estimated by the slopes of the lines in Fig. 5A)were similar in alkalotic and control rats.

The active component of proximal chloride reab-sorption can be estimated from the vertical interceptsof Fig. 5A, when the diffusive driving force is mini-

Metabolic Alkalosis in the Rat 1151

c

0E

1500-

CL

c:

0

0(,

1000-

-00

0

I-- 500-Ex

.0a-a)

on oor

1500

C-C-

Ij I 000

X Z0 o

w aL 500O

1jU)0 mm Ld

0

O Ecr

< (L_* L

OZCK O

cr-

0 CDDO

a menw

1500r

lF A

Normal

Ikalosis

IF

10 20 30A [Cl-] (meq/1iter)

(End-Proximal - Bowman's Space)

B

1250 1-

1000L

2000 3000 4000FILTERED CI LOAD (peq/min)

FIGURE 5 (A) Absolute proximal chloride reabsorption as a function of the chloride concen-tration gradient (end-proximal minus Bowman's space). Values both before (circles) and after(diamonds) acetazolamide administration are shown for normal hydropenic animals (filledsymbols) (12) and animals with chronic metabolic alkalosis (open symbols). Mean SNGFRinthe four groups were similar. (B) Absolute proximal chloride reabsorption as a function ofSNGFRand filtered Cl load during chronic metabolic alkalosis (0) and following isohydricplasma expansion (0) or isohydric colloid-free Ringer expansion (U).

mized. Previous studies in normal and acidotic rats chloride reabsorption in metabolic alkalosis was vir-found this component of reabsorption to be an isochlo-ric process (11). In agreement, the -20% reductionfrom normal in the vertical intercept in Fig. 5A wasproportional to the fall in the blood chloride concen-tration during metabolic alkalosis.

Thus, absolute proximal chloride reabsorption dur-ing metabolic alkalosis was similar to or slightly lessthan normal hydropenic controls. Fractional proximal

tually identical to normal hydropenic controls, in ac-cord with others (4, 5). Low distal chloride deliveryrates and, hence, avid renal chloride reabsorption ap-pears to result from both the low GFR and hypo-chloremia in chronic metabolic alkalosis.

Sodium chloride reabsorption had previously beenfound to increase when filtration rate was raised byplasma expansion, but not by colloid-free expansion

1152 M. C. Cogan and F.-Y. Liu

(10, 11). As shown in Fig. 5B, absolute proximal chlo-ride reabsorption increased with plasma expansion(closed circle), but not with colloid-free expansion(closed square) during metabolic alkalosis. Looked atin a different way, at the higher flow rate (36 nl/min),the greater plasma and peritubular protein concentra-tion during plasma expansion than during colloid-freeexpansion was associated with a significant accentua-tion in absolute proximal chloride reabsorption(1,347±68 vs. 979±85 neq/min, P < 0.05). Absoluteproximal water reabsorption was also higher duringplasma expansion (17.1±0.9 vs. 13.4±0.9 nl/min, P< 0.05). However, absolute proximal total CO2 reab-sorption was not significantly different between thetwo groups. Thus, the peritubular Starling forces havean important influence over proximal-sodium chloridereabsorption, although not over proximal sodium bi-carbonate reabsorption (10, 11, 19, 20).

Despite the difference in proximal sodium and chlo-ride reabsorption during the two types of expansion,urinary sodium and chloride excretion rates were com-parable (Table IV, lines 2 and 4). Therefore, nephronsegments other than the superficial proximal convo-luted tubule played an important role in modulatingrenal sodium chloride excretion during volume expan-sion in metabolic alkalosis. Further support for thisconcept comes from an analysis of chloride and bi-carbonate reabsorption during expansion. Whole kid-ney chloride reabsorption was stimulated comparedwith bicarbonate reabsorption by colloid-free expan-sion, resulting in partial repair of the alkalosis (6, 7).Nevertheless, in the superficial proximal convolutedtubule, there was no enhancement in the ratio of chlo-ride/bicarbonate reabsorbed during isohydric expan-sion compared with the base-line alkalotic state (1.02vs. 1.11).

Clearance studies

Induction of alkalosis. The previous micropunc-ture studies found absolute proximal bicarbonate reab-sorption to be normal. Reabsorption was not stimulatedwhen flow rate was made to increase toward normal.The maintenance of metabolic alkalosis in this settingwas dependent upon a reduction in SNGFR. To con-firm that GFR falls during chronic metabolic alkalosisin unanesthetized rats, additional clearance studieswere performed.

As shown in Table V, the control group (group I)maintained on the artificial diet with NaCi and KCIsustained no significant alteration in acid-base status,GFR, plasma volume, muscle [K] or plasma [K]. In thisand the following groups, a change in muscle [K] was

used as an index of K depletion and a change in plasmavolume as a reflection of Cl depletion. There was nodifference in the results of rats receiving supplementalbicarbonate in the drinking water from those not re-ceiving extra bicarbonate. Rats in this and other groupstypically ingested and excreted 1.0 meq/d of bicar-bonate with a mean urinary bicarbonate concentrationof 50±5 meq/liter. Also, there were no significant dif-ferences in the rats maintained for 9 d (groups IA) com-pared with those maintained for 21 d (groups IB) SOthat results were combined for these groups inTable V.

The Na2SO4 and DOCA-treated group II developeda significant metabolic alkalosis (plasma bicarbonateconcentration rose from 24.6±0.4 to 39.2±1.1 meq/liter), as shown in Table V. There was also marked Kdepletion (muscle [K+] fell from 42.1±0.5 to 22.7±0.7meq/100 g FFDS) and C1 depletion (plasma volumedeclined from 8.6±0.5 to 6.2±0.1 ml). Weight loss of7% compared with controls occurred. Supplementalbicarbonate sufficient to induce bicarbonaturia causedno greater degree of alkalosis in group II. Thus, theplasma bicarbonate level stabilized at the same levelwhether the excess blood bicarbonate was generatedendogenously or was provided exogenously. In bothcases, the rise in blood bicarbonate concentration wasaccompanied by a proportional fall in GFR, from2.9±0.1 to 1.8±0.1 ml/min (P < 0.001). This inverserelationship between plasma bicarbonate concentra-tion and GFR is illustrated in Fig. 6 (the mean valuefor all normal animals in groups I-IV during the firststudy period is represented by the large black circlewith error bars, while the individual group II animalswith severe metabolic alkalosis in the second studyperiod are represented by squares).

Since the changes in plasma bicarbonate concentrateand GFR were reciprocal, the filtered load of bicar-bonate before and after induction of alkalosis was iden-tical (Table V). Thus, a normal rate of renal hydrogenion secretion was sufficient to maintain the high plasmabicarbonate concentration of group II.

To ameliorate the K deficiency, K salts replaced Nasalts in the diet and drinking water in group III. Incomparison to group II, group III sustained less of areduction in muscle [K] and plasma volume (13). Amore moderate degree of alkalosis resulted (plasmabicarbonate concentration 29.5±0.3 meq/liter). As hadbeen observed with group II, supplemental bicarbon-ate had no additional effect on acid-base or electrolyteparameters, so the results are combined in Table V.Mirroring the smaller increment in plasma bicarbonateconcentration, the decline in GFRwas also less (hex-agons in Fig. 6). Again, filtered bicarbonate loads be-fore and after alkalosis were comparable.

Metabolic Alkalosis in the Rat 1153

0-1 -

eq 0e'10+11-

+1

t--

UN gCl +l

CO -C6 6

+1

C-

,ItC+1

0>00C6 +10 +l

(N -( O

+1

+1-i l

0 066

(N +°l

Co -( O

+1

0-+1

F

00( +c9 +l

-

m cli( O

+1

+10-4 -

0 0

C +°

-

+1

CDio6

+1

C +°

-N

c- +1

-Q

+1

_C+1

Co +1

(N +1

(N -

L-0o

+l

0

+1.

+1c_)

o6+1

~-+1

cl C)C +1

Co -

04+1

(N +1

(N +l

Co -

+-

+l

(N-

+1

+1

+1a11) co

+1

O +1

C-N CO

+1

c'i+1

-o

Co +1

a'S_(N +°l

0

+1

1O

+1.

COCoi+1

+1

CC+1

° +1

C-

(NCO

+1

co

+1

(N +1

(N +1

~-0in C)

+1

t-0

+1

-+1

+1

O. ItDQO o

- +1

CC+1

C-

_NO1

+1

CC -

+1

(N +1

-

cs+l

Co-

t-

CO_

+1.

< co U

-0_

aCU

-0Cd

)>CI V V

w- - -_). () ) )"0Q "0Q ~

10

00 -

-0o+1

C -(NO

+1

- C

( +1(N +1

CC-+1

cq

ci+

+1

'- +1

( -t +1

+1

(NO

C CO

~c000

C6 6

C +1

Co -

t~ 0+l

t-0

+1

(N -Co -

(NO

+1

C-o

C. 6

0 +l

0

oq -c.16

+1

( O+1

as+

C9 +°l

+1

Co-

t.-0t

+1

-+1

-

+1

+1

(N-

+1

C,,)

C6 6

C +1

C-0C +1

1- 0

+1

010- +1

4-1

+1

0-E +o

Co -

+1

+1

00

CoD +1

o +1

If

I-0c+

+l

C- 1(N

+1

CC;

+1

+l

[~+1

I--

cli

+1

0-0

(N -

C1 +l

(NO

C)+1

(N -

(N +l

-0

C -

r.o+l

ci

+1

r.

::

C)_ CU

c 0 -)0) E ..

u - @0 0,) C)

r. @ Y m~ )

6.

Q0.o 0X 0)

.; Z X _

CL > 5: > >0)

1154 M. C. Cogan and F.-Y. Liu

+1

Co -

( O+1

0-6 +1

": +°lI

sr -

+1

IC

-

a6ci

.10d +1

+1

mN

+1

CC+1

C -

+1

°° +°l

's+l

's+l

YcE

Ecli

E

0. _

o Cl

o -_

;X

Cli.0O

Cl

C)

I-61

V)

IC

14)

Q

c

CIA)

u

.1

INC

;s

:

CO)

CO

CZ

-0

r.'CUCdO

C)

0 "0

Q~ o

+l06o6CU)V V

la0o0~~~~~~T0o

2 3GFR (ml /min)

FIGURE 6 Relationship of plasma bicarbonate concentration to GFRduring clearance studiesin normal controls and in chronic metabolic alkalosis. The mean±SEMfor the data of all 75normal animals in the first clearance period is represented by the large black circle with error

bars (+). Shown are the individual data in the second clearance period for group II (0); group

III (0); group IVA (V); group IVB (tx); and group IVc (°).

The three subsets of animals comprising group IVwere used to induce isolated K deficiency without Cldeficiency. Each subgroup sustained a reduction inmuscle [K+] while plasma volume did not change.Weight loss compared with controls was -4%. A mod-erate alkalosis occurred in the three subgroups. As il-lustrated in Fig. 6 (triangles and diamonds), the coun-

terbalancing fall in GFR (2.3±0.1 ml/min in each)kept filtered bicarbonate load constant. Animals givenexcess bicarbonate were no more alkalotic than thosereceiving just water so that the data have again beencombined. Neither the usage of DOCA(group IVA)nor the higher NaCl diet with a slightly higher plasmavolume in group IVc than in IVB disrupted the inverserelationship between changes in plasma bicarbonateconcentration and GFR.

Repair of alkalosis. The next four groups of ex-

periments were designed to assess whether the reversalof the K and/or Cl deficiencies modulated the plasmabicarbonate concentration and GFR in a qualitativelysimilar fashion (but in opposite directions) to thatwhich occurred during the induction experiments.These rats were first made alkalotic with Na2SO4 andDOCAin a manner similar to group II. In the firstclearance period, somewhat milder degrees of K andCl depletion, alkalosis and GFRreduction were foundthan had occurred in Group II. The latter had receivedtwo rather than one surgical preparations and clear-ance studies. However, continuing Na2SO4 and DOCAfor another week in group V exacerbated the decrease

in muscle [K] and plasma volume. A greater degree ofalkalosis (plasma bicarbonate concentration 40.6±0.6meq/liter) and GFR reduction resulted (1.7±0.1 ml/min), indistinguishable from group II. As before, theinverse relationship between plasma bicarbonate con-

centration and GFRwas retained, as shown in Fig. 7(the mean value for the first alkalotic period of all ratsin groups V-VIII is represented by the large stippledcircle with error bars, while the individual results forthe second period of group V are represented bysquares; the lightly shaded area represents the range

of values obtained during the induction experimentsof groups I-IV, from Fig. 6). Filtered bicarbonate loadin group V remained normal.

Dietary K replacement without Cl replacement ingroup VI increased muscle [K] without appreciablyaffecting the plasma volume. A significant but incom-plete repair of the alkalosis occurred (bicarbonate con-

centration fell from 33.0±0.7 to 29.9±0.3 meq/liter).There was significant improvement, but without nor-

malization of the GFR (from 2.2±0.1 to 2.4±0.1 ml/min). The relationship between plasma bicarbonateconcentration and GFR in Group VI is shown by thehexagons in Fig. 7. Filtered bicarbonate load stayedat a normal level.

Selective NaCl replacement in group VII allowedan increase in plasma volume without appreciably af-fecting K status. As with K replacement, Cl repletioncaused a significant partial recovery of the alkalosis(plasma bicarbonate concentration fell from 35.0±0.6

Metabolic Alkalosis in the Rat 1155

50 H

40 F

-

0-4

-I-00IC)-E(0FL

30 H

20 H

lo 4

50r

40H

30

20

0 2

GFR (ml/min)3 4

FIGURE 7 Relationship of plasma bicarbonate concentration to GFRduring clearance studiesin chronic metabolic alkalosis and following recovery from alkalosis. The mean±SEMfor thedata of all 20 alkalotic animals in the first clearance period is represented by the large stippledcircle with error bars (+). Shown are the individual data in the second clearance period forgroup V (E); group VI (0); group VII (A); and group VIII (0). The shaded area encompassesthe range of data points from Fig. 6.

to 30.8±0.1 meq/liter). GFR also improved (from2.0±0.1 to 2.3±0.1) with maintenance of a normal fil-tered bicarbonate load. These results are illustrated bythe triangles in Fig. 7.

Finally, repletion of both K and Cl in group VIIIled to complete normalization of K and Cl stores withcomplete recovery of acid-base status (bicarbonateconcentration fell to 26.2±0.4 meq/liter) and hemo-dynamics (GFR rose to 2.7±0.1 ml/min), as shown bythe circles in Fig. 7. All values were then indistin-guishable from the controls of group I.

DISCUSSION

The maintenance of metabolic alkalosis requires thattubular bicarbonate reabsorption be increased or fil-tration rate decreased. The micropuncture studies re-

ported here demonstrated that absolute proximal bi-carbonate reabsorption was not increased in rats withchronic metabolic alkalosis. Proximal bicarbonatereabsorption was not stimulated by a high luminalbicarbonate concentration, extracellular volume andpotassium depletion, or when flow increased. The al-ternative pathophysiological mechanism for maintain-ing alkalosis was therefore suggested, that filtrationrate was diminished. SNGFRand GFRwere, in fact,found to be depressed in metabolic alkalosis so thatfiltered bicarbonate loads were normal. The 95 clear-ance studies in awake animals during the inductionand repair of metabolic alkalosis confirmed that an

elevation of the blood bicarbonate concentration wasaccompanied by a reciprocal fall in GFR.

The following discussion will first consider factorsthat control proximal bicarbonate reabsorption in met-abolic alkalosis. The observation of a reduction in GFRwill then be considered, especially in reference to theaccompanying potassium and chloride deficits that arefrequently present in metabolic alkalosis.

Determinants of proximal bicarbonate reabsorp-tion. On the basis of previous observations, the highermean luminal total CO2 concentration in metabolicalkalosis compared with normal euvolemic controls(32 vs. 17 mM) was expected to stimulate proximalbicarbonate reabsorption (10, 17, 18). That an in-creased luminal bicarbonate concentration failed toincrease acidification (Fig. 2) might be ascribed to in-trinsic saturation of the hydrogen ion secretory system;i.e., a maximum transport rate (Vmax) may have alreadybeen attained. Another possible explanation is that themagnitude of the maximal bicarbonate concentrationor pH gradient in the proximal tubule had been alteredduring metabolic alkalosis. A new, higher steady-stateluminal bicarbonate concentration may have occurred(e.g., 21.5 mMin alkalosis instead of 4-5 mMin normalcontrols), as might happen if bicarbonate back-leakhad markedly increased. Such was found not to be thecase since an end-proximal total CO2concentration as

low as 7.6 mMwas achievable when flow rate was

severely reduced.Yet another explanation is that the increase in peri-

tubular bicarbonate concentration and pH during met-

1156 M. G. Cogan and F.-Y. Liu

-

1-1

cr0E

I Fn00a:I_

ECE)(A

abolic alkalosis independently suppressed the proximalacidification process. An increase in the peritubularbicarbonate concentration might inhibit acidificationby slowing bicarbonate exit from the cell and alka-linizing the cell. If cellular pH increased, the rate ofproton secretion would be relatively depressed at anylevel of luminal bicarbonate concentration or pH.Proximal acidification might even appear saturated asa function of luminal bicarbonate concentration at asubnormal level (20). Thus, antagonism by peritubularalkalinity of the stimulatory effect of luminal alkalin-ity would be observed.

This thesis, that peritubular alkalinity suppressedproximal hydrogen ion secretion, is consonant withobservations of split droplet studies in the proximaltubule (4, 21), of in vivo microperfusion studies (20)and in vitro microperfusion studies (22) using proximaltubules, as well as of studies using the turtle bladder(23). Thus, suppression of acidification by peritubularalkalinity may explain why reabsorptive saturationwas observed in the proximal tubule as filtered bicar-bonate concentration was simultaneously increased,and similarly why a reabsorptive plateau (a "Tm") hasbeen observed in the whole kidney during bicarbonatetitration studies (2).

Whether proximal acidification would respond to anincrease in filtered bicarbonate load and in mean lu-minal bicarbonate concentration was examined in thethree expansion protocols. Bicarbonate reabsorptionfailed to keep pace with the increase in filtered bi-carbonate load. This lack of proximal glomerulotu-bular balance might have been due to an effectivesaturation of the hydrogen secretion process as notedabove, either intrinsic or induced by alkalemia. Al-ternatively, it has been suggested that extracellularvolume expansion, by promoting bicarbonate back-leak, can independently suppress net proximal bicar-bonate reabsorption (1). In the latter case, the en-hanced bicarbonate back-leak and relative inhibitionof acidification should persist if flow rate were reducedduring expansion. However, when aortic constrictionwas used and flow rate decreased, the end-proximalbicarbonate concentration fell and absolute proximalbicarbonate reabsorption assumed a value appropriateto the filtered load (Fig. 4A). Fractional proximal reab-sorption and distal bicarbonate delivery normalized.Thus, volume expansion did not prevent the return ofproximal glomerulotubular balance (Fig. 3). These re-sults are in accord with previous in vivo studies in therat and in vitro studies in the rabbit: if tubular flowwere maintained constant, extracellular volume statusand peritubular protein concentration had little im-pact on proximal bicarbonate absorption (10, 19, 20).Aortic constriction also caused disappearance of bi-carbonaturia, despite persistent extracellular volume

expansion. The failure to abolish bicarbonaturia whenCohen (7) combined aortic constriction with isohydricexpansion during metabolic alkalosis in the dog maybe due to the difference in a species studied but isotherwise unexplained.

The physiologic importance of the effect of flow rateduring metabolic alkalosis may be summarized as fol-lows. When SNGFR was incrementally increasedabove the value recorded during chronic metabolicalkalosis, the proximal acidification process behavedas if saturated (20). Progressively larger distal bicar-bonate delivery rates resulted as filtered bicarbonateloads exceeded normal levels. Although other nephronsegments were capable of reabsorbing a proportion ofthis excess bicarbonate delivery, bicarbonaturia nev-ertheless developed (Fig. 4). Since proximal bicarbon-ate reabsorption during metabolic alkalosis was notflow dependent, these studies underscore the criticalimportance of a reduction in SNGFRfor mediatingthe persistence of chronic metabolic alkalosis. The re-duction of SNGFRacted to prevent the filtered loadand, thus, distal bicarbonate delivery from exceedingnormal values. A bicarbonate diuresis with correctionof the alkalosis was thereby prevented.

Finally, a comment should be made regarding therole of potassium deficiency in proximal bicarbonatereabsorption. Clearance studies had suggested thatpotassium depletion stimulates renal bicarbonate reab-sorption (24-26). In the present studies, there was nosignificant difference in absolute proximal bicarbonatereabsorption when the isohydrically expanded, chron-ically alkalotic, hypokalemic group was comparedwith the expanded (10% body wt), acutely alkalotic,normokalemic group. However, Kunau et al. (3) foundincreased fractional bicarbonate reabsorption duringisotonic bicarbonate expansion of potassium-depletedrats compared with similarly expanded normokalemiccontrols. The apparent discrepancy in results can beeasily explained by the different filtration rates mea-sured in the two groups in the present studies. Whenthe potassium-deficient animals with isohydric volumeexpansion are compared with the acutely alkalotic an-imals, (Table III, lines 2 and 7), the difference in frac-tional bicarbonate reabsorption in the two cases (0.63vs. 0.49) was attributable to markedly different filteredloads (SNGFR). Absolute rates of bicarbonate reab-sorption were not significantly different. Thus, thepresent studies failed to observe a stimulation in ab-solute proximal bicarbonate reabsorption when potas-sium-depleted rats were compared with potassium re-plete rats with comparable extracellular volume ex-pansion and alkalemia. The difference in chronicityof the alkalemia in the two groups may have maskedan effect of potassium deficiency. However, other free-flow micropuncture studies by Levine et al. (27) have

Metabolic Alkalosis in the Rat 1157

not demonstrated any change in absolute proximalbicarbonate reabsorption when nonalkalotic, potas-sium-deficient animals were compared with nonalka-lemic controls or were acutely infused with potassium.Ullrich et al. (21) similarly found no abnormality inproximal acidification when potassium-depleted ratswere assessed with split-droplets. However, Chan etal., (28) using in vivo luminal and pericapillary mi-croperfusion have reported a 23% stimulation of prox-imal bicarbonate reabsorption in potassium-deficientalkalotic rats. Thus, the available evidence suggeststhat chronic potassium deficiency may stimulate ab-solute proximal bicarbonate reabsorption, but the ef-fect appears to be relatively small.

Importance of a reduction in GFRfor maintainingthe alkalotic state. The micropuncture studies founda fall in SNGFRand GFRand no stimulation of ab-solute proximal bicarbonate reabsorption during met-abolic alkalosis. Possible limitations, however, werethat only a single model of metabolic alkalosis wasutilized and animals were anesthesized. Therefore,clearance studies in awake rats with varying degreesof potassium and chloride depletion and metabolicalkalosis were performed.

These clearance experiments were not designed toexplore how the excess blood bicarbonate was gener-ated. Indeed, half of the animals received supplemen-tal bicarbonate to ensure that net acid excretion by thedistal nephron would not be rate limiting for increas-ing the blood bicarbonate concentration. Thus, theexcess bicarbonate permitted the maximum degree ofalkalosis to be achieved and the maximal rate of renalbicarbonate reabsorption (principally by the proximalnephron) to be expressed.

The inverse relationship between blood bicarbonateconcentration and GFR(Fig. 6) held in the consciousanimals when the metabolic alkalosis was associatedwith K depletion alone (group IV) or with both K andCl depletion of moderate (group III) or severe degree(group II).

That isolated dietary K depletion, as in groups IVBand IVc, can sustain metabolic alkalosis has long beenknown for the rat (5, 29, 30) and for man (31, 32),though not for the dog. Mineralocorticoid treatment(e.g., DOCA), as in group IIIA, is also a well establishedcause of persisting metabolic alkalosis associated withK deficiency (13, 29, 33), simulating primary hyper-aldosteronism. Yet, the mechanism by which K deple-tion maintains metabolic alkalosis has not been pre-viously established. The present studies suggest thatK depletion may maintain metabolic alkalosis in therat by diminishing GFR. Reduction in renal blood flowand GFRby selective K depletion is a well describedphenomenon in rats (34, 35) and dogs (25, 26, 36). Kdepletion apparently decreases GFRby selective renal

vasoconstriction. For instance, K depletion causes hy-perreninemia, which may induce intrarenal vasocon-striction by activating angiotension 11 (37). K depletionalso stimulates the renal synthesis of the potent vaso-constrictor thromboxane B2 (38). Recently, Linas andDickmann (39) have presented strong evidence thatboth of these systems are operative in the K-deprivedrat to increase renal vascular resistance and depressrenal blood flow (39). While some degree of K wastageis ubiquitous in clinical forms of metabolic alkalosis(1), the possible contribution of K depletion to decreaseGFR and maintain alkalosis in humans is undefinedat present. This issue is especially important in thepathophysiology of certain "chloride-resistant" met-abolic alkaloses such as primary hyperaldosteronism.The results obtained in group IIIA suggest that meta-bolic alkalosis in states of primary mineralocorticoidexcess may be mediated by a reduction in GFRratherthan enhanced tubular bicarbonate reabsorption. Fur-ther studies of GFR in K-depleted alkalotic humansis warranted.

When Cl depletion is added to K depletion, as ingroups II and III, metabolic alkalosis can be well main-tained in all species studied (3, 4, 6, 7, 13, 33, 40, 41).Cl deficiency reduced extracellular volume and plasmavolume, the latter in turn causing a fall in renal bloodflow and GFR (8, 9).3

Controversy exists regarding whether human formsof chronic metabolic alkalosis associated with both Kand Cl depletion are also maintained hemodynami-cally, by a depression in GFR. There is conflictinginformation in chronic metabolic alkalosis on whetherGFR is normal (24, 41) or reduced (31, 43). Again,further systematic studies are needed on this importantclinical pathophysiologic question.

In the clearance experiments examining the repairof alkalosis, the opposite changes in plasma bicarbon-ate concentration and GFR occurred to mirror theeffects in the induction series (Fig. 7). Selective reple-tion with K or Cl only partially repaired the alkalosisand GFR, while KCl repaired both, as has been pre-viously described (13, 30, 33, 42). The rat is more"chloride-resistant" than dog or man with respect torepair of alkalosis with NaCl (13, 29, 30).

The normality of filtered bicarbonate loads during

3 Cl depletion has also been suggested to stimulate distalhydrogen ion and K secretion by limiting anion delivery tothe distal exchange sites (33, 40, 41). However, the Cl-de-ficient groups (groups II, III, V, VI) receiving supplementalbicarbonate did not achieve a blood bicarbonate concen-tration disproportionately higher than that expected for anormal filtered bicarbonate load and normal rate of renalacidification. Thus, no enhanced net renal acidification wasrevealed by bicarbonate loading, even with DOCAadminis-tration.

1158 M. G. Cogan and F.-Y. Liu

all stages of alkalosis and repair, even in animals givensupplemented bicarbonate, supported the thesis thatthe rate of absolute renal acidification was unchangedby K and/or Cl depletion. Rather, that state of K andCl stores modulated GFR so the metabolic alkalosiswas present only when GFRwas diminished.

If all the clearance studies are taken as a whole, itappeared that K depletion and Cl depletion had ap-parently independent and additive roles in affectingGFR. In the following analysis, an attempt is made toillustrate the empiric relationship of changes in GFRthat were evoked by alterations in K and/or Cl stores.For this purpose, alteration in K stores is reflected bychanges in muscle [K] (30), and alteration in Cl storesis reflected according to its impact on renal hemody-namics by changes in plasma volume (8, 9). The effectof changes in K and Cl stores on changes in GFRcanbe expressed quantitatively using multiple linearregression analysis. The equation was: AGFR = 0.68A muscle [K] + 0.28 A plasma volume + 2.2, r2 = 0.80,P < 0.001. The change in GFRwas significantly andindependently related to both changes in muscle [K]and plasma volume. The best correlation, however,was with both variables combined, as indicated in theequation above. The conclusion that K and Cl deple-tion were additive factors in maintaining metabolicalkalosis in the rat agrees with that of Luke and Lev-itan (13). The present studies suggest that K and Cldepletion exerted their impact on the severity of met-abolic alkalosis by modulation of GFR.

In conclusion, in all conditions of chronic metabolicalkalosis in the rat, the elevation in the plasma bicar-bonate concentration was found to be inversely pro-portional to the fall in GFR. Filtered bicarbonate loadsremained constant, so that a normal rate of renal acid-ification was sufficient to maintain the alkalosis.

ACKNOWLEDGMENTSKenneth R. Wongand Monika R. Mueller provided excellenttechnical assistance. Dr. Claude G. Biava performed themicroscopic examinations. Wewish to express our deep ap-preciation to Dr. Floyd C. Rector, Jr., of the University ofCalifornia, San Francisco and Professor Chiu Chuan-Lu ofthe Shanghai First Medical College and Hua San Hospital,Shanghai for their continuing advice, generous support, andencouragement.

This work was supported in part by a Clinical InvestigatorAward to Dr. Cogan (1 K08 AM01015) and by a grant (AM27045) from the National Institute of Arthritis, Diabetes,Digestive and Kidney Diseases.

REFERENCES1. Seldin, D. W., and F. C. Rector, Jr. 1972. The generation

and maintenance of metabolic alkalosis. Kidney Int. 1:306-321.

2. Pitts, R. F., and W. D. Lotspeich. 1946. Bicarbonate and

the renal regulation of acid base balance. Am. J. Physiol.147: 138-154.

3. Kunau, R. T., Jr., A. Frick, F. C. Rector, Jr., and D. W.Seldin. 1968. Micropuncture study of the proximal tu-bular factors responsible for the maintenance of alkalosisduring potassium deficiency in the rat. Clin. Sci. (Lond.).34: 223-231.

4. DeMello Aires, M., and G. Malnic. 1972. Micropuncturestudy of acidification during hypochloremic alkalosis inthe rat. Pfluegers Arch. Eur. Physiol. 331: 13-24.

5. Bank, N., and H. S. Aynedjian. 1965. A micropuncturestudy of renal bicarbonate and chloride reabsorption inhypokalemic alkalosis. Clin. Sci. (Lond.). 29: 159-170.

6. Cohen, J. J. 1968. Correction of metabolic alkalosis bythe kidney after isometric expansion of extracellularfluid. J. Clin. Invest. 47: 1181-1192.

7. Cohen, J. J. 1970. Selective Cl retention in repair ofmetabolic alkalosis without increasing filtered load. Am.J. Physiol. 218: 165-170.

8. Maddox, D. A., D. C. Price, and F. C. Rector, Jr. 1977.Effects of surgery on plasma volume and salt and waterexcretion in the rat. Am. J. Physiol. 233: F600-F606.

9. Ichikawa, I., D. A. Maddox, M. G. Cogan, and B. M.Brenner. 1978. Dynamics of glomerular ultrafiltrationin euvolemic Munich-Wistar rats. Renal Physiol. 1: 121-131.

10. Cogan, M. G., D. A. Maddox, M. S. Lucci, and F. C.Rector, Jr. 1979. Control of proximal bicarbonate reab-sorption in normal and acidotic rats. J. Clin. Invest. 64:1168-1180.

11. Cogan, M. G., and F. C. Rector, Jr. 1982. Determinantsof proximal bicarbonate, chloride, and water reabsorp-tion during carbonic anhydrase inhibition. Am. J. Phy-siol. 242: F274-F284.

12. Cogan, M. G., and F. C. Rector, Jr. 1982. Proximal reab-sorption during metabolic acidosis in the rat. Am. J.Physiol. 242: F499-F507.

13. Luke, R. G., and H. Levitan. 1967. Impaired renal con-servation of chloride and the acid-base changes asso-ciated with potassium depletion in the rat. Clin. Sci.(Lond.). 32: 511-526.

14. Vurek, G. G., D. G. Warnock, and R. Corsey. 1975.Measurement of picomole amounts of carbon dioxide bycalorimetry. Anal. Chem. 47: 765-767.

15. Ramsay, J. A., R. H. J. Brown, and P. C. Croghan. 1955.Electrometric titration of chloride in small volumes. J.Exp. Biol. 32: 822-829.

16. Yamani, T. 1967. Statistics, an Introductory Analysis.Harper & Row, Publishers, Inc., New York.

17. Ullrich, K. J., H. W. Ratke, and G. Rumrich. 1971. Therole of bicarbonate and other buffers on isotonic fluidabsorption in the proximal convolution of the rat kidney.Pfluegers Arch. Eur. Physiol. 330: 149-161.

18. Alpern, R. J., M. G. Cogan, and F. C. Rector, Jr. 1982.Effect of luminal bicarbonate concentration on bicar-bonate absorption in the rat proximal convoluted tubule.Am. J. Physiol. 243: F53-F59.

19. Berry, C. A., and M. G. Cogan. 1981. Influence of peri-tubular protein on solute absorption in the rabbit prox-imal tubule. J. Clin. Invest. 68: 506-516.

20. Alpern, R. J., M. G. Cogan, and F. C. Rector, Jr. 1983.Effects of systemic pH and volume expansion on bicar-bonate absorption in the rat proximal convoluted tubule.J. Clin. Invest. 71: 736-746.

21. Ullrich, K. J., G. Rumrich, and K. Baumann. 1975. Renalproximal tubular buffer-(glycodiazine) transport. Inho-mogeneity of local transport rate, dependence on so-

Metabolic Alkalosis in the Rat 1159

dium, effect of inhibitors, and chronic adaptation. Pflue-gers Arch. Eur. Physiol. 357: 149-163.

22. Sasaki, S., C. A. Berry, and F. C. Rector, Jr. 1982. Effectof luminal and bath HCO- concentrations and pCO2 onHCO- reabsorption in rabbit proximal convoluted tu-bules perfused in vitro. J. Clin. Invest. 70: 639-649.

23. Cohen, L. J., and P. R. Steinmetz. 1980. Control of activeproton transport in turtle urinary bladder by cell pH.J. Gen. Physiol. 76: 381-393.

24. Roberts, K. E., H. T. Randall, H. L. Sanders, and M.Hood. 1955. Effects of potassium on renal tubular reab-sorption of bicarbonate. J. Clin. Invest. 34: 666-672.

25. Giebisch, G., M. B. MacLeod, and R. F. Pitts. 1955.Effect of adrenal steroids on renal tubular reabsorptionof bicarbonate. Am. J. Physiol. 183: 377-386.

26. Kurtzman, N. A., M. G. White, and P. W. Rogers. 1973.The effect of potassium and extracellular volume onrenal bicarbonate reabsorption. Metab. Clin. Exp. 22:481-492.

27. Levine, D. Z., T. Walker, and L. A. Nash. 1973. Effectsof KCl infusions on proximal tubular function in normaland potassium-depleted rats. Kidney Int. 4: 318-325.

28. Chan, Y. L., B. Biagi, and G. Giebisch. 1982. Controlmechanisms of bicarbonate transport across the rat prox-imal convoluted tubule. Am. J. Physiol. 242: F532-F543.

29. Seldin, D. W., L. G. Welt, and J. H. Cort. 1956. The roleof sodium salts and adrenal steroids in the productionof hypokalemic alkalosis. Yale J. Biol. Med. 29: 229-247.

30. Struyvenberg, A., J. DeGraeff, and L. D. F. Lameijer.1965. The role of chloride in hypokalemic alkalosis inthe rat. J. Clin. Invest. 44: 326-338.

31. Schwartz, W. B., and A. S. Relman. 1953. Metabolic andrenal studies in chronic potassium depletion resultingfrom overuse of laxatives. J. Clin. Invest. 32: 258-271.

32. Jones, J. W., A. Sebastian, H. N. Hulter, M. Schambelan,J. M. Sutton, and E. G. Biglieri. 1982. Systemic and renal

acid-base effects of chronic dietary potassium depletionin humans. Kidney Int. 21: 402-410.

33. Kassirer, J. P., A. M. London, D. M. Goldman, andW. B. Schwartz. 1970. On the pathogenesis of metabolicalkalosis in hyperaldosteronism. Am. J. Med. 49: 306-315.

34. Holliday, M. A., and T. J. Egan. 1962. Changes in GFRand CHSObefore and after repair of K deficiency in rats.Am. J. Physiol. 202: 773-776.

35. Brunner, F. P., F. C. Rector, Jr., and D. W. Seldin. 1966.The mechanism of the urinary concentrating defect inpotassium-deficient rats. Pfluegers Arch. Eur. Physiol.290: 202-210.

36. Abbrecht, P. H. 1969. Effect of potassium deficiency onrenal function in the dog. J. Clin. Invest. 48: 432-442.

37. Linas, S. L. 1981. Mechanism of hyperreninemia in thepotassium-deficient rat. J. Clin. Invest. 68: 347-355.

38. Beck, N., and J. 0. Shaw. 1981. Thromboxane B2 andprostaglandin E2 in the K+-depleted rat kidney. Am. J.Physiol. 240: F151-F157.

39. Linas, S. L., and D. Dickmann. 1982. Mechanism of thedecreased renal blood flow in the potassium-depletedconscious rat. Kidney Int. 21: 757-764.

40. Needle, M. A., G. K. Kaloyanides, and W. B. Schwartz.1964. The effects of selective depletion on hydrochloricacid on acid-base and electrolyte equilibrium. J. Clin.Invest. 43: 1836-1846.

41. Kassirer, J. P., and W. P. Schwartz. 1966. The responseof normal man to selective depletion of hydrochloricacid. Factors in the genesis of persistent gastric alkalosis.Am. J. Med. 40: 10-18.

42. Kassirer, J. P., and W. B. Schwartz. 1966. Correction ofmetabolic alkalosis in man without repair of potassiumdeficiency. Am. J. Med. 40: 19-26.

43. Ariel, I. M., and F. Miller. 1950. Effects of hypochlo-remia upon renal function in surgical patients. Surgery(St. Louis). 28: 552-567.

1160 M. C. Cogan and F.-Y. Liu