Effect of acute hypercapnia on renal and proximal tubular total carbon dioxide reabsorption in the acetazolamide treated rat

Effect of acute hypercapnia on renal and

proximal tubular total carbon dioxide

reabsorption in the acetazolamide-treated rat.

J Winaver, … , K A Walker, R T Kunau Jr

J Clin Invest.

1986;

77(2)

:465-473.

https://doi.org/10.1172/JCI112325

.

The present study evaluates the effect of acute hypercapnia on renal total CO2 (tCO2)

reabsorption after inhibition of renal carbonic anhydrase. Simultaneous renal clearance

studies and free-flow micropuncture studies of the superficial proximal tubule were

performed on plasma-repleted Sprague-Dawley rats treated with acetazolamide, 50 mg/kg

body weight. Acute hypercapnia (arterial PCO2, 120 mmHg; blood pH, 7.02) was induced

by ventilation with a 10% CO2-90% O2 gas mixture. Control rats (PCO2, 49.5 mmHg, pH

7.34) were ventilated with room air. The renal fractional excretion of tCO2 was

approximately 20% lower in the hypercapnic group compared with the rats given

acetazolamide alone. Acute hypercapnia reduced the fractional delivery of tCO2 to the late

proximal tubule by a comparable amount. The absolute proximal reabsorption of tCO2 was

increased by hypercapnia to 410 +/- 47 vs. 170 +/- 74 pmol X min-1, P less than 0.05. The

single nephron glomerular filtration rate was 32.6 +/- 0.7 nl X min-1 in the hypercapnic group

and 43.8 +/- 1.7 nl X min-1 in the rats given acetazolamide only, P less than 0.01. Acute

hypercapnia enhances renal sympathetic nerve activity. To eliminate this effect, additional

experiments were performed in which the experimental kidney was denervated before

study. Denervation prevented the change in the single nephron filtration rate during acute

hypercapnia, but absolute and fractional proximal tCO2 reabsorption […]

Research Article

Effect

of Acute

Hypercapnia

on

Renal and Proximal Tubular Total Carbon

Dioxide Reabsorption

in

the

Acetazolamide-treated

Rat

Joseph Winaver, KathleenA.Walker, and Robert T.Kunau,Jr.

Department ofMedicine, TheUniversityof Texas Health Science Center, SanAntonio, Texas 78284

Abstract

The present study evaluates the effect of acute hypercapniaon renal total CO2 (tCO2) reabsorption afterinhibition ofrenal car-bonicanhydrase.

Simultaneous renal clearance studies and free-flow micro-puncture studies of the superficial proximal tubule were per-formed on plasma-repleted Sprague-Dawley rats treated with acetazolamide, 50mg/kg bodyweight. Acute hypercapnia (ar-terialPco2, 120

mmHg,

blood pH, 7.02)wasinduced by ven-tilation witha10% COr90% 02 gas mixture. Control rats (Pco2, 49.5 mmHg, pH 7.34) were ventilated with room air. The renal fractional excretion of tCO2 was -20% lower in the hypercapnic group compared with the rats given acetazolamide alone. Acute hypercapnia reduced the fractional delivery of tCO2 to the late proximal tubule by a comparable amount. The absolute proximal reabsorption of tCO2 was increased by hypercapnia to 410±47 vs.170±74 pmol *

min-',

P<0.05. The single nephron glomer-ular filtration rate was 32.6±0.7 nl *min-' in the hypercapnic group and 43.8±1.7 nl-

min-'

in the rats given acetazolamide

only,P <0.01.

Acutehypercapnia enhances renal sympathetic nerve activity. Toeliminate this effect, additional experiments were performed inwhich the experimental kidney was denervated before study. Denervation prevented the change in the single nephron filtration rate during acute

hypercapnia,

but absolute and fractional prox-imaltCO2 reabsorption remained elevated in comparison to de-nervated controls.

TheconcentrationofH2CO3 in the late proximal tubule, cal-culated from the measured luminalpH and bicarbonate concen-tration and theestimated corticalPco2, washigherin the

hy-percapnicgroup, which was afinding

compatible

with H2CO3 cyclingfromlumenintoproximaltubularcell, whichprovideda sourceof hydrogen ions for secretion.

Introduction

Arterial bloodCO2tension(Pco2)isconsideredtobeoneof the

major factors regulatingtherateoftubularhydrogen ion(H+)

secretion (1). Many clearance studies have documented that

Partsof this study werepresentedatthe 16th annual meetingof the American SocietyofNephrology,Washington,DC, 1983.

AddressreprintrequeststoDr. Kunau. Dr.Winaver's present address is theFaculty ofMedicine,RappaportFamilyInstitute for Medical Re-search, Haifa,Israel.

Receivedforpublication 24 October 1984 and in revisedform 17 October 1985.

1.Abbreviations usedinthis paper: ACTZ, acetazolamide; AHC,acute hypercapnia; DNX, denervated; FE, fractionalexcretion;GFR,

glomer-acute hypercapnia (AHC)' enhances the maximal bicarbonate reabsorptive capacity of the kidney and decreases fractional bi-carbonate excretion (2-5). It hasalso been shownthat AHCcan augment renal bicarbonate reabsorption in the presence ofphar-macologic inhibitors of renal carbonic anhydrase such as acet-azolamide (ACTZ). Thus, it has been suggested that carbonic anhydrase independent bicarbonate reabsorption is stimulated byhypercapnia (5-7).

Rector (8) has argued that, in the absence of carbonic an-hydrase, the uncatalyzed rate of CO2 hydration or hydroxylation cannotsupplysufficientH'toaccount for all of the bicarbonate reabsorbed under these circumstances. A number of possible means have been suggested as being responsible for maintenance ofH'secretion after inhibition of carbonic anhydrase. Of several possible mechanisms, recent studies have been interpreted to suggest that cycling of carbonic acid (H2CO3) from lumen into cellis a means by whichH'may be continuously made available for secretion in the proximal tubule of the ACTZ-treated rat undercontrol circumstances and after the induction of metabolic acidosis oralkalosis (9).

Inaddition to a variety of systemic effects, AHC can reduce theglomerular filtration rate (GFR) and renal blood flow (10-14). The reduction in renal blood flow noted after AHC can be abolished by renal denervation, but not by maneuvers that block renin production (14). These findings suggest that the change in renal hemodynamics after AHC is related to increased renal sympathetic nerve activity. The findings of Anderson et al. (14) also suggest that the reduction in renal blood flow after AHC is proportionately greater than the change in the GFR. As a con-sequence, the filtration fraction and the peritubular capillary protein concentration are both increased. Recently, Cogan et al. (15) have demonstrated that proximal total CO2 (tCO2)

reab-sorptionduring inhibition of carbonic anhydrase may be mod-ified by the peritubular protein concentration (15). Thus, any increase in proximal tCO2 reabsorption after AHC in the ACTZ-treated rat may, at least in part, be an indirect result of the changein peritubularprotein concentration.

Inaddition totheabove observations, Shah et al. (16) and Chan(17)haveshown that alpha adrenergic stimulation increases and denervation reduces proximal bicarbonate reabsorption. Thus, AHC may result in an increase in renal bicarbonate

reab-sorptionviaadirect effect of the AHC per se at the tubular level or as aconsequence of stimulation of renalsympatheticnerve

activity.

The presentfreeflow micropuncture and clearance studies were undertaken to evaluate the effect of AHC on renal and proximal tubular tCO2 reabsorption in the ACTZ-treated rat, andtoexamine thepotentialcontribution of the increased sym-patheticnervousactivityinduced by AHContCO2 reabsorption

under these conditions.

ular filtration rate;SNGFR, single nephron glomerularfiltrationrate; tCO2,totalC02; TF,tubularfluid; TF/P,tubular fluidtoplasma ratio; U/P,urinetoplasma ratio.

Total CarbonDioxideReabsorption after Hypercapniaand Acetazolamide 465 J.Clin.Invest.

©TheAmericanSocietyfor Clinical

Investigation,

Inc.

Methods

25 Sprague-Dawley rats weighing 300-360 g were studied in four ex-perimental groups. The rats were fed a standard rat chow diet and had free access to food and water before the experiments.

Anesthesia was induced by Inactin (100 mg/kg body weight, BYK Guldens, Konstanz, Federal Republic of Germany). The animals were placed onathermostatically controlled heated table and body temperature was maintained between 370 and 380C. Polyethylene catheters (PE 50) were' inserted into the right femoral vein and artery for infusion of the various solutions and for blood pressure monitoring and sampling of arterial blood, respectively.

Atracheostomy was performed and the animals were connected to aHarvard rodent respirator. The rate and depth of ventilation were adjusted so as to keep arterial blood Pco2 between 40 and 50 mmHg in the control animals ventilated with room air in the presence of ACTZ (groups A and D). AHC was induced in groups B and C by connecting theventilator to a 10% C02, 90% 02 gas mixture.

All animals were infused with homologous isooncotic plasma obtained onthesame day from another Sprague-Dawley rat. The plasma was infused from the start of the experiment at a rate of 20 Ml/min. This amountof plasma resulted in a hematocrit at the conclusion of the study that wassimilar to that obtained before the abdominal and neck surgery. Inaddition, the rats received a noncolloid solution, either a modified Ringer solution (NaCl, 115 meq/liter; NaHCO3, 25 meq/liter; KCI, 4

meqjliter)

innormocapnic animals (groups A and D) or normal saline (0.9% NaCl) in the hypercapnic animals (groupsBand C). These solutions weregiven at a rate of 25Ml/minsimultaneously with the plasma infusion. Normal saline rather than Ringer's solution was administered to the hypercapnic animals to minimize the increase in plasma bicarbonate levelinduced by the hypercapnia, thus reducing the variability in filtered load among the experimental groups (15). An indwelling polyethelene catheter wasinserted into the bladder for collection of the bladder urine in groupsAand B. The clearance data reported for these groups represent thefindings from both kidneys. In groups C and D, where denervation of the left kidneywasperformed, urine was collected separately from eachkidney.

Micropuncture protocol. A ventral abdominal approach was used to expose the left kidney for micropuncture. Care was taken to minimize blood loss during the surgical preparation by clamping the bleeding points. Theleft kidney was separated from the surrounding tissue after ligation of the adrenal vessels. The kidney was placed in a lucite cup surrounded by oil-soaked cotton and bathed with mineral oil bubbled continuously with 5%CO2 gas and maintained at 370 to 380C. An arterial blood pressure of 100 mmHg or less at this point led to the exclusion of the animal from further study.

After completion of surgical preparation, a loading dose of 240MCi 3H-methoxyinulin (New England Nuclear, Boston, MA) was given as a bolusfollowed byasustaining infusion of the same amount per hour in thenoncolloid solutions listed above. 30 min after the initiation of the inulin infusion, the modified Ringer solution was discontinued and ACTZ, 50 mg/kg body weight, was given as a bolus injection followed byacontinuous infusion of the same amount per hour. ACTZ was dis-solved in a 300mMNaHCO3 solution and was given at a rate of 25Ml/

min inthis solution in groupsAandDtoprevent reduction in plasma bicarbonate level due to ACTZ-induced urinary losses of bicarbonate. Inthehypercapnic animals, groups B and C, ACTZ was dissolved in 50 Ml of 300 mMNaHCO3 solutionandthen mixedwith 3 mlofsaline solution and infused at a rate of 25 Ml/min. Replacement ofurinary losses ofbicarbonate was not necessary in groups B and C.

30 minafter ACTZ was begun, an arterial blood sample of 400JA was taken for determination of pH and tCO2, and a urine collection was begun into tared vials under mineral oil. A second blood sample was obtainedatthe endof theexperiment. 4-6 late proximal convoluted tubuleswereidentified byintratubular injection of 0.3% FD&C solution (Keystone Aniline & Chemical Co., Chicago, IL) with sharpened micro-pipettes(5-6MAmOD).

Thefollowing experimental groups were studied:

(a) GroupA(n= 7), ACTZ, and normal Pco2. This experimental groupwasventilated withroomair. Plasma Pco2wasmaintained between 40 and 50 mmHg by appropriate adjustment of the respirator. 4-6 timed tubularfluid collectionswereobtained from lateproximal convoluted tubules by sharpened glass micropipettes (10-II_jlmOD). Sampleswere immediately transferred to a chamber containing water-equilibrated mineral oil diffused with a 10% Co2, 90% 02 gas mixture and kept there until analysis.Asimultaneous urine collectionwasobtained during the micropuncture period for clearance data.

(b) GroupB(n=7) and ACTZ plusAHC. This group was ventilated with, 10%CO2, 90%O 02 gas mixture during the experiment. Late proximal tubular samples and urinewereobtainedasin group A. The sample-containing pipettes were transferredtoan oil chamberasin group A, with the exception that 15% C02 was in the chamber.

(c) Group C(n=6), ACTZ plus AHC and denervation. This exper-imental group was treated identically to groupBwith the following ex-ceptions: after separation of the left kidney from the surrounding tissue, apolyethylene catheterwasinsertedinto the leftureter.Renal denervation was then performed according to the method described by Bello-Reuss andGottschalk (18).Therenal arterywasexposed and the adventitia wasstripped. The renal nerves were dissected and cut free. The renal artery was then wrapped witha narrowband of filter paper, previously soaked in a solution of 90% alcohol, 10% phenol, for 20-30 min. Care wastaken to prevent contact between the phenol solution andthesurface ofthekidney during the procedure. Tubular sampleswereobtainedas in theprevious group. Urine was collected separately from the denervated (DNX) left kidney and the untouched right kidney.

(d)GroupD(n =5), ACTZ, normalPco2, and denervation. This group served asacontrol forgroup Ctoevaluate the effect ofDNXper se onbicarbonate transport in the presence of ACTZ. The animalswere

ventilated as in groupAwith room air. Denervationwasperformedas described in group C. Tubular sampleswereobtainedaspreviously de-scribed.

Toevaluate the effectiveness of the denervation procedure,aseparate group of eight ratswasstudied.Four ratsunderwent the denervation procedure as described for groups C and D, and fourratsweresham operated. 72 hlater thekidneyswereremoved andplacedimmediately in dry ice. Tissue norepinephrine contentwassubsequently determined (19). Three of the four denervated kidneys had undetectable levels of norepinephrine.Thefourthcontained 30 ng/g kidney. The norepineph-rine levels in the sham-operated kidneyswere250, 365, 411, and 376 ng/g. We conclude that the denervation procedure usedwassuccessful. Ina separate series of studies, in rats prepared as in groups C and D, theintratubular pHwasmeasured withamicro(7-9MmOD) pH elec-trode. Theelectrode was manufactured as described by Puacacco and Carter(20) withtheexceptionthattheU02content wasreduced from 4to2%. No electrode was used that did not haveaslope of at least 58 mV/pH unitwhendetermined in vitro in a series of standard phosphate buffers (pH 4, 6.84, and 7.384) at 370C. After its insertion into and removalfrom the tubule, the electrode was again tested in vitro. If the electrodepotential deviated by >2 mV from the value obtained before tubularpuncture,thedatawerediscarded. The pH electrode was con-nected to a model 616 digital electrometer (input resistance 2 X 1014Q) (Keithley Instruments, Inc., Cleveland, OH) through anAg/AgCl half cell(WPI,NewHaven, CT).Theelectrometerwasconnected to a7132A thermalrecorder(Hewlett-Packard Co., San Diego, CA). This arrange-mentpermitted us to read the potential of the electrode by the digital readoutof the electrometer, yet graphically observe the characteristics andtimecourseofthe electroderesponse onthe chartrecorder.Acalomel reference electrode, connected to the low input of the electrometer, and the cut endoftherat's tailwere placed into a beaker containing Ringer's solution.

mea-surements were madeinlateproximal siteslocated asdescribed above. Two measurements were made in each of six rats.

Analysis. The volumeofeach tubularfluidsample wasdetermined by measuring the lengthofthefluidcolumnafter transfer to a calibrated, constant bore quartz capillary. Great care was taken to maintain C02-equilibratedoil at both endsofthe quartzcapillary. The concentration of 3H-methoxy inulinin known volumesoftubularfluidsamples, plasma, andurinewas measured byscintillation counting usingaLS7000liquid scintillationcounter(Beckman Instruments, Inc., Palo Alto, CA), as pre-viously described from this laboratory (21). The tCO2 concentration in tubularfluidsamples was measured bymicrocalorimetry (22).The vol-umeofsamples usedfor this determinationwas14.0-19.0 nl.Detailsof thegeneral methodology used to handle the tubularfluidsamples were reported previously (21).

PlasmaandurinarytCO2 were measured by a 960 carbondioxide analyzer(Corning Medical andScientific, Medfield, MA). Urineand arterial blood pHweremeasuredwithanExpandomatic Type IVpH meter(Beckman Instruments, Inc.). Plasma Pco2 was calculated from tCO2and pHusingpK.of6.1 and aof0.0301. Plasma Na+ and K+ concentrationswere measured byflamephotometry. Plasma solids were measured by refractometry. The latter values were identical in all ex-perimentalgroups.Therefore, no correction for plasma water was made inthecalculation ofTF/Pratio of inulin andtCO2.

Calculations. Whole kidney GFR was calculated in the standard fashion. Singlenephron glomerular filtration rate (SNGFR) =

_TF/P1.

Xtubularfluid flow rate, where

_TF/PIn

is the tubular fluid to plasma inulin ratio.The SNGFR is expressed innl*min-'.Fraction of filtered tCO2 deliveredtoendof proximalconvolution or to final urine=TF/ P or U/PtCOJIn X 100%, where TF/P or U/P tCO2/In is the tubular fluidorurinetoplasmatCO2/inulin ratio. The absolute reabsorption of tCO2 in theproximal convoluted tubulewascalculatedasthedifference betweentheamountfilteredandtheamountdeliveredtothe lateproximal puncturesiteandis expressed in pmol*min-'.

Statistics. One-way analysis of variancewas used to evaluatestatistical differencesamongtheexperimentalgroups. The Duncan test was used toevaluatethelevelof significancebetween each two groups that were compared(23). Thistest wasappliedonly when the null hypothesis of nodifference between the four experimental groups was rejected. The pairedt test wasused to compare clearance data between the control andexperimental kidneys inthe DNXanimals(groups C and D). The data areexpressedasmean±SEM.

Results

Asummaryof the systemic and acid base data in the fourgroups is presented in TableI. A mild degreeof respiratory acidosis (Pco249.5 and50.8mmHg)wasobservedingroupsAand D, respectively. Increasing thetidal volume and/orrate of venti-lationdidnotresultinafurtherdecrementinthearterial blood

Pco2. Similar results were obtained previously in the dog(5)

andare presumably dueto the inhibitory effect of ACTZ on

erythrocyte carbonic anhydrase. Ventilation with a 10%C02,

90% 02mixture during carbonic anhydrase inhibition(groups BandC)resulted ina severedegreeof hypercapnic acidosis of

similar degreein both groups (Table I). PlasmatCO2 was

sig-nificantlyelevatedin the lattergroupscompared with their

con-trol counterparts (groups A and D). Calculated plasma bicar-bonatewasalso significantly higherinthe hypercapnicgroups

compared with the normocapnic animals. This increment in plasmabicarbonateoccurreddespitethefactthatasaline solution wasinfused inthehypercapnic animalsandreflectsthe

nonbi-carbonate buffer capacity of the plasma and extracellular fluid (24).Thefourgroupswerenotstatistically different withrespect totheirbody weights,finalblood hematocrit values, and arterial bloodpressures.Thelatter findingisof interestsince in thedog,

hypercapnia may result ina significant fall in systemic blood pressure(25).Finally, plasmasodium concentrationswerenot statistically differentin groupsAandDfrom thoseinB and C

(133.3±1.19 meq/liter and 132.4±0.94 meq/liter, respectively, P=NS). However, AHC(groups BandC)resulted ina

signif-icantincrease in plasma potassium level (4.65±0.09 meq/liter compared with 3.13±0.05 meq/liter in groups A and D (P <0.001).

The clearance data fromgroupsAandBarelistedonTable II.AHC resulted inapromptandsignificant reduction in urinary flowrate.Total GFRwasnotstatistically different between the twogroups.Urinary toplasma tCO2 ratiowassignificantly

de-creased in the hypercapnicanimals(ACTZ, 9.05+0.16;ACTZ plushighPCO2,6.42±0.12,P<0.001). Thefractional excretion

TableI.Systemic and Acid-base Data

Group Pco2 pH PlasmatCO2 Plasma HCO5 MBP Body weight Hct

mmHg mM/liter mM/liter mmHg g %

A. ACTZ* (n=7) 49.5 7.34 27.6 26.1 118 329 48.0

1.1 0.01 0.5 0.5 5 7 0.8

B. ACTZ+AHC (n=7) 120.1 7.02 33.8 30.2 125 334 48.9

0.8 0.004 0.3 0.3 2 7 0.5

C. ACTZ+AHC+DNX(n=6) 119.6 7.00 32.0 28.4 122 327 47.0

1.8 0.006 0.4 0.4 3 4 0.7

D. ACTZ+DNX(n=_{5) 50.8 7.33 27.2} 25.7 110 335 47.8

3.4 0.02 0.6 0.6 3 6 0.9

B vs.A P<0.01 <0.01 <0.01 <0.001 NS NS NS

Cvs.B PNS NS <0.05 <0.01 NS NS NS

Cvs.A P<0.01 <0.01 <0.01 <0.01 NS NS NS

D vs.C P<0.01 <0.01 <0.01 <0.01 NS NS NS

D vs.A PNS NS NS NS NS NS NS

MBP,meanarterial bloodpressure; n,number ofrats.Resultsonthis and subsequent tablesareexpressedasthe mean±SEM.

Table II. Urinary Excretion Data

Group V GFR U/P tCO2 Urine pH FEtCO2

id-min-' ml-min-'* 100 g %

body weighf'

A. ACTZ 74.9 0.53 9.1 7.84 39.04 (n=7) 4.1 0.02 0.2 0.01 1.48

B. ACTZ 51.6 0.48 6.4 7.57 20.23

+AHC 3.7 0.01 0.1 0.02 0.76 (n=7)

A vs.B P<0.005 NS <0.001 <0.001 <0.001

V, urine flow rate; GFR represents results from both kidneys.

(FE) of tCO2 was 39.04±1.48% in the normocapnic animals treated with ACTZ. This value is in general agreement with otherpublished data regarding the effects of inhibition of renal carbonic anhydrase activityinthe rat (26, 27), and demonstrates that -60% of the filteredload oftCO2 was reabsorbed. AHC resultedinasubstantial reductionin fractionalexcretion (FE)

tCO2to 20.23±0.76%,P <0.001 (vs. group A), indicating that

tCO2 reabsorption in the presence of ACTZ was sensitive to changes inarterialblood CO2 tension.

The micropuncture datafromthesuperficialproximal con-voluted tubule are presented in Table III. Duringadministration ofACTZ, group A, the lateproximal

TF/Pj.

was1.43±0.03 and the SNGFR 43.8±1.7 nl/min. The late proximal TF/P tCO2

ratio was 1.19±0.08.Inhibitionof renalcarbonicanhydrase re-sulted in fractional delivery oftCO2 of 85±6% to the end of

proximal convoluted tubule,demonstrating that only 15% of

filteredload oftCO2wasreabsorbed in theproximal convoluted

tubule under thesecircumstances.These dataaresimilartothose

obtainedbyothers(15,28) in plasmarepletedrats.

Induction of AHC during carbonic anhydrase inhibition (group B) was associated withtwomajor findings. First, under these conditions, the SNGFR was significantly decreased to 32.6±0.7 nl-min-',P<0.01, and the TF/P1n was elevated to

1.70±0.08, P < 0.01, both compared with group A. Second, fractional tCO2 reabsorption in the proximal tubule was mark-edly enhanced by AHC; i.e., fractional delivery of tCO2 to the end oftheproximal tubule was 62±4%,P < 0.01, compared with group A. The decrement in whole kidney tCO2 excretion after AHC is therefore reflected by a corresponding change in fractional tCO2 reabsorption in the superficial proximal tubule. The fact that the difference in the TF/P tCO2 ratio between groups A and B did not reach statistical significance may be related to the simultaneous increment in proximal fluid reab-sorption in group B after AHC.

The single nephron filtered load of tCO2 in groupsAandB werecomparable (Table III). The absolutereabsorptionof tCO2 in group B, 410±47 pmol-min-',wassignificantlygreater than

thatobserved in group A, 170±74 pmol * min-', P<0.05. To evaluate thecontribution ofrenalsympatheticnerveactivityto the renal responsetohypercapnia, additional studies,groups C and D, wereperformedin which the leftkidneywasDNXbefore thestudy.InductionofAHCafter denervation (group C) resulted

inarterialblood pH and Pco2valuessimilarto that noted in

groupB,Table I. There was a mild decrease in plasma tCO2as well as in the calculated plasma bicarbonate concentration in group Ccomparedwithgroup B.

Clearance data comparing the left DNX kidney with the right, untouched kidney in group C are presented in Table IV.

Single kidney GFR was mildly elevated in the DNX

kidney

(0.28±0.01 ml/min compared with the innervated kidney,

0.24±0.01 ml/min,P<0.01). Thisresultis of interest in view

of thefindingthat there was no significantdifference in total kidney GFR between groupsAand B.The reasonfor this

dis-crepancyisunclear, butitmay bethatachangeinwholekidney GFR, resulting from sympathetic overactivityafterAHC,is

rel-TableIII.Micropuncture Data

TF/P

Group TF/P,. SNGFR FLtC02 TFtC02 tCO2 ARtCO2 FDtCO2 nlmin-' pmol- min-' mM/liter pmol- min-' %

A. ACTZ(n=7) 1.43 43.8 1,189 33.3 1.19 170 85

0.03 1.7 43 2.8 0.08 74 6

B. ACTZ+AHC (n=₇₎ 1.70 32.6 1,085 35.3 1.05 410 62

0.08 0.7 28 2.1 0.06 47 4

C. ACTZ+AHC+DNX(n = 6) 1.47 47.7 1,527 26.22 0.83 653 58

0.06 2.0 71 1.6 0.05 102 5

D. ACTZ+DNX(n=_{5) 1.38} 44.7 1,210 29.7 1.08 262 78

0.05 2.8 109 1.7 0.05 45 3

Bvs.A P < 0.01 <0.01 NS NS NS <0.05 <0.01

Cvs.B P < 0.01 <0.01 <0.01 <0.05 <0.05 <0.05 NS Cvs.A PNS NS <0.01 <0.05 <0.05 <0.01 <0.01

D vs.C PNS NS <0.01 NS <0.05 <0.01 <0.05

Dvs.A PNS NS NS NS NS NS NS

Table IV Comparison of Clearance Data

from the Innervated and DNX Kidneys in Group C

Innervated DNX

P GFR (ml *min-'*100 g body

weight-') 0.24±0.01 0.28±0.01 <0.01 FEtCO2 (%) 18.0±1.3 22.7±0.7 <0.02

ativelysmall and can be depicted only when the DNX kidney is compared with the untouched kidney in the same hypercapnic animal. There was a slight but significant difference in FEtCO2 between the two kidneys (innervated 18.0±1.3%; DNX, 22.7±0.7%, P < 0.02). Both values are similar to the value ob-served in group B. The fractional excretionof tCO2 in the DNX kidney of group D, ACTZ alone, was 35.8±2.9%, which was significantly greater than the 22.7±0.7% noted in the DNX kid-neyin group C. Thus, renal sympathetic activity was not re-sponsible for the enhanced renal tCO2 reabsorption induced by the AHC.

Table III alsolists the values obtained from micropuncture resultsfrom groups C and D, in which the left (experimental) kidney was DNX before study. DNX of the experimental kidney didnotaffectthe response to ACTZ noted in the proximal tubule, group D vs. group A. Theobservation that DNX did not affect proximal tubular function in the absence of AHC (group D vs. group A)is of interest. These findings suggest that renal DNX may notaffect the SNGFR and proximal tubular fluid reab-sorption if baseline sympathetic nerve activity is not increased. Renal DNX prevented a reduction in SNGFR and an in-crease in proximal

TF/Pln

ratio from occurring in response to AHC, group B vs. group C. Despite the DNX, an increase in

proximal tCO2 reabsorption was still observed in response to AHC, group C vs. group D. The TF/P tCO2 ratio in group C, 0.83±0.05, was less than noted in group D, 1.08±0.05, P < 0.05. Fractional delivery oftCO2 to the late proximal tubule after AHC, group C, was 20% less than in group D, which was

ap-proximatelythe samedifference as was noted between groups Aand B, the groups in which renal innervation was intact. The absoluteproximal reabsorption of tCO2 was significantly greater

afterAHC, group C, 653±102 pmol*

min-',

than in the dener-vated control, groupD,

262+45,

P<0.01. Thesingle nephron

filteredloadoftCO2 wassignificantly greater in group C than in group D, TableIII.However,after correctingtheplasma tCO2

concentrationforthearterial Pco2 in thetwogroups, thefiltered

load of bicarbonate was not different in the two groups,

1,352±52, group C, vs. 1,150±88

pmol*min-',

group D,

P=NS.

Themeasurement oftCO2in tubularfluid by microcalori-metryreflectsnotonly measurement of bicarbonate in the sam-ple, but thatof dissolved CO2aswell. Withoutprecise knowledge

of the renal corticalPco2,it isnotpossibleto correctthe

mea-suredtubularfluid tCO2 for the latter value.Inviewof the dif-ference between the systemic and renal cortical Pco2 and an

amplificationof this differenceduring respiratory acidosis(28), correctionof the tubularfluid tCO2 concentration fortherenal

corticalPco2 will result in agreaterchangethancorrection of plasma tCO2 for arterialblood Pco2.Consequently,

proximal

TF/P bicarbonate ratioswould be lower than the TF/P

tCO2

ratiosshownwith proportionately greater reductions in this ratio in AHC rats becauseof the difference in renal cortical Pco2 as mentioned above. Given theseconsiderations, it is clear that the rates offractional and absolute tCO2reabsorption demonstrated abovereflectcorresponding changes in the reabsorptiverateof bicarbonate.

Itis unlikely that the increase in single nephron filtered load of tCO2 in group C, incontrast togroupD, contributes signifi-cantly tothe differencein absoluteproximal absorption noted between the two groups. Variations in the filtered load of tCO2 do not,of themselves, alter fractional proximal tCO2 reabsorption in the ACTZ-treated rat (15). Accordingly, <20% of the differ-encein absolutetCO2 reabsorption, in group C in comparison with group D, can beattributed to the higher filtered load present in theformer group. Further, thedifferences in proximal tCO2 reabsorption cannot be attributedtodifferences in tubular fluid flowrate(15) asthey were identical in thesetwogroups.

Intratubular pH measurements wereperformed in six late proximal convolutions of three normocapnic, ACTZ-treated, DNX animals, and in six tubules of three hypercapnic, DNX

animals treated with ACTZ, Table V. Mean plasma pH and tCO2 in thenormocapnicgroup were 7.35±0.01 and 25.7±0.33 mM,respectively; they were 7.05±0.01 and 31.7±0.88 mM, re-spectively, in the hypercapnic group. These valuesare not

sta-tisticallydifferent from those obtained in experimentalgroups C and D, the groups that were handled identicallytothose in whichthe pH measurements were made. Itis assumed,therefore, that the in situ pH measurements obtainedin these animals accurately reflect the insitu pH in groups C and D in which determination oftCO2 transport was made. The in situ late proximal intratubular pH was 6.90±0.01 in the animals treated with ACTZ only.This value issignificantly different (P<0.005) from the valueof 6.58±0.02 obtained in theproximalsegment

oftheanimals treated with ACTZ and subjectedtoAHC.

Discussion

Thefindings of the present studyconfirm earlier observations

(5-7)which demonstrate that renal bicarbonatereabsorptionin the presenceofaninhibitor ofcarbonic anhydrasemay be sig-nificantlyincreased byelevation ofthearterialPco2.Anincrease

of-20% infractionaltCO2

reabsorption

wasnoted after AHC in the present studies, indicatingthat carbonic anhydrase in-dependent bicarbonate reabsorption was

quite

sensitive to changes in arterial Pco2. Our data also demonstrate that the

effect of AHC,inthisexperimental setting,is reflectedbya cor-TableV.LateProximal LuminalpH Measurements

ACTZ+AHC ACTZ

Rat 1 6.62* 6.84

6.55 6.84

Rat2 6.51 6.95

6.57 6.94

Rat3 6.61 6.92

6.65 6.91

Mean 6.58±0.02 P<0.005 6.90±0.01

*Each measurement was obtained from a separate tubule.

responding change in tCO2 reabsorption in the proximal con-voluted tubule. This observation,however, does not exclude an effect of AHC in other nephron segments or nephron groups under these circumstances. Such a major impact of AHC on proximal bicarbonatereabsorption is of interest in view of the finding that in the normal rat, when renalcarbonic anhydrase activity is intact, AHC induces onlyamildeffectonproximal bicarbonate reabsorption,orhas noeffectat all(29-32). Finally, ourdata show thatAHC can have amarked influenceon the SNGFR andfluid reabsorption ofthesuperficial proximaltubule. Thesealterations appear to be mediatedby renal sympathetic

nerve stimulation, secondary to theAHC. However, theincrease in tCO2 reabsorption in the superficial proximal tubule after

AHC was unrelated toindirect ordirect effects oftheincrease

in sympathetic activity, as renal denervation prevented AHC from affecting either the SNGFR andproximal fluid reabsorp-tion;yetproximaltCO2reabsorptionremainedsubstantially el-evated.

Theincrease in tCO2 reabsorption in the innervated group B, in comparison with group A, may beattributabletoseveral

factors, and not only theresult of the direct effectofan acute increase in Pco2 on proximal H'transport. Cogan et al. (15) have suggested thatproximal tubular tCO2reabsorptionin the ACTZ-treated rat may vary directlywith theperitubular capillary

protein concentration. If AHC in the innervated group B in-creased thefiltration fraction insuperficial nephronsas a con-sequenceofenhanced renalsympatheticnerveactivity(14), such anincreasecouldcontribute tothe increase in thetCO2

reab-sorption in thisgroup. Second, renal

sympathetic activity,

of

itself, can directly stimulate proximal

tCO2

reabsorption

and couldcontributetotheincreasenoted in groupB

(16,

17).

In contrast tothe findings in group B, thecontributionof these indirect and direct effects ofrenal sympathetic

activity

would not beresponsible for theincrease in tCO2reabsorption noted after AHC in theDNX state. Denervation preventeda reduction in SNGFRafter AHC, andwepresume, based upon otherstudies ( 14), that no change in renalplasma flowoccurred in these rats. Similarly, any direct effect ofsympathetic nerve

activityontCO2reabsorptionwould be removedbythe dener-vation. The results in theDNXstudies, therefore, indicate that theeffects ofAHConproximaltCO2reabsorptionare notonly

mediated via stimulation of renalsympatheticactivity.

After the administration ofan inhibitor ofrenal carbonic anhydrase, fractional tCO2deliverytothe end of thesuperficial

proximal tubule remains remarkably constant when examined in the normal or plasma expanded rat orafter theinductionof metabolic acidosis or metabolic alkalosis (15, 33). Thus, as a consequence of theconsiderablevariabilityof the filtered load oftCO2 orsomederivative thereof, noted in thesediverse ex-perimental settings,theconstancy of fractionaltCO2reabsorption is associated with large differences in absolute reabsorption. In contrasttotheobservations cited above, the present experiments indicate that AHC increases tCO2 reabsorption in the proximal tubule of the ACTZ-treated rat independent of an increase in thefiltered load.

Netbicarbonate transport in the proximal convoluted tubule

is felt tobe a function of the secretion of H+ coupled with a smallpassive backleak of bicarbonate (34).Inthis segment, H+ secretion isprimarilyaresult ofaNa+/H+antiporterlocated in theluminal membrane (35). Inthe absence ofan inhibitor of

carbonic anhydrase, this segment demonstrates a remarkable

capacity

toreabsorbbicarbonate,particularlyifthe tubularflow

rateand the luminalbicarbonate concentrationareoptimal (36, 37). ACTZ treatment is felt to reduce proximal bicarbonate reabsorption byreducingthecellularsupplyof H'available for

secretion, and, atleast in part, byallowing luminalH2CO3 to accumulate, thereby inducingapH gradientbetween cell and lumen unfavorable forproton secretion. The continued

reab-sorptionof bicarbonate in theproximaltubule after ACTZ has been ascribedtoseveral mechanisms (8, 38). Some of this bi-carbonatecanbe attributedtothe

H'

supplied by the uncatalyzed

hydrationandhydroxylationofCO2reactions,butthe numbers sosuppliedareinsufficientto accountfor thequantityof bicar-bonatereabsorbed(27). Coganetal.(27)haverecentlyestimated the contribution oftheuncatalyzed reactionsto proximal bi-carbonatereabsorption, takinginto consideration the effects of alterations in intracellular bicarbonate concentration. Their es-timates wouldpredictthat theuncatalyzedrateof

H'

secretion isinsufficientto accountfor therateof carbonicanhydrase

in-dependent

reabsorption

observed unless the intracellularpHwere inexcessof8.5.Althoughit has been shown that the intracellular pH becomesmorealkalineduringcarbonicanhydraseinhibition

(39, 40),anintracellularpHof thismagnitude, togetherwithan

intratubular aciddisequilibriumpH, would result inapH gra-dientacrosstheapicalmembraneof -2 pHunits.Theauthors

consideredaH' gradient of this magnitudetobeunlikely.

Todetermine ifanincrease in cell Pco2as aconsequence ofthe AHC contributed

significantly

totheincreaseinproximal

bicarbonate

reabsorption

noted in groupC, in comparisonto groupD,wehaveextended the calculations ofCoganetal. (27)

using slightly different kinetic constants (41). By computer

modeling(seeAppendix),we wereabletoestimatetherate of cellularH'

production

bytheuncatalyzed reactionsas afunction ofboth intracellularpHandPco2rangingfrom 40to180mmHg,

at a constantintracellularbicarbonateconcentrationof25mM. The absoluterateoftCO2 reabsorptionobservedafterAHC in group C is substantially higherthan thatpredicted bythe un-catalyzedreactions,evenwhenthecorticalPco2is -60mmHg higherthan thearterialPco2.Thus, unlessone assumes an ex-tremelyalkaline intracellularpH(pH of 9.0at acorticalPco2 of120or 180mmHg), theanticipatedrate oftheuncatalyzed

reactions cannot account for the magnitude ofthe proximal bicarbonate reabsorptionobservedafterAHC.

Anumberof years ago, Rector(8)proposed another

mech-anismthat would account foracontinuous supply of H+to a protonsecretoryprocess in the presenceof carbonic anhydrase inhibition(8). He suggested that the H2CO3, which accumulated in the lumenafterACTZ,mightcyclefromthelumen intothe

cell,whereit woulddissociate.Thus, H+ would beavailable for

secretionnotdependentuponcontinuousproduction fromthe

uncatalyzedhydrationandhydroxylationofCO2. Cogan (9)has recently argued that the directrelationshipbetweenluminal bi-carbonate concentration and bicarbonate reabsorption in the presence of ACTZ is theresult ofaproximalH+secretory mech-anism that is limited by the lumentocell pHgradient. Inhis

view,H+secretion will vary directly with the luminal bicarbonate concentration and,at a constantluminal pH, the luminal H2CO3 concentration will parallel theluminalbicarbonate

concentra-tion.Theproportionalchange inH2CO3 concentrationwith lu-minalbicarbonate concentration will allowmoreH2CO3to enter the cell at high luminal bicarbonate concentrations, thus

ac-countingformorebicarbonatetobereabsorbed thanwhen the luminal bicarbonate concentration is lowered.

studies cannot be attributed precisely to the same mechanism. Based upon approximations of the renal cortical Pco2 (below), luminal bicarbonate concentrations in the AHC groups would be approximately equal to (group B) or below (groupC) their appropriate controls, groups A and D, respectively. Further, in contrast tothehypothesis put forward by Cogan (9) to explain parallel changes inluminal H2CO3 and bicarbonate concentra-tions at a constant luminal pH, our data indicate that, at least under our experimental circumstances, the luminal pH is not fixed. Yet, our data are compatible with luminal H2CO3 as a source for H' for secretion and with an increase in luminal H2CO3 after AHC. The luminal pH in the rats treated as in group D was6.90. AssumingacorticalPco2of 60mmHg(33), andusing a luminal bicarbonate concentration of 27.9 mM and a

pK.

of3.57, the calculatedluminal H2CO3 concentrationat

thispH was 13 uM. No experimental data are available to

in-dicatethecorticalPco2in the hypercapnic animals in group C, butitlikely can be expected to exceed 150 mmHg (33). If we assume acortical Pco2 of 180 mmHg, a luminal H2CO3

con-centration of20

gM

can be calculated from the luminal pH (6.58) and corrected luminal bicarbonate concentration.2 The

permeability oftheapical membrane of the proximal convoluted tubule toH2CO3 and theH2CO3 concentration gradient across this membrane are unknown. Nevertheless, the higher luminal H2CO3 in the AHC rats is consistent with a greater potential source forH' production derived from H2CO3,whichmoves from lumen to cell. Our studies, unfortunately, do not permit a precisedetermination of the amount of H2CO3 that has tra-versed theluminal membranes into the cell and contributed to H+secretion. Theoretically, this could be all of the H2CO3 pro-duced by H+secretionthatdidnotnoncatalyticallybreak down in the lumen. Luminal pH measurements and calculated H2CO3

concentrationsare notreportedforratsstudiedasgroupsAand B. The direct andindirect consequences of renal sympathetic stimulation afterAHC(group B) prevents examination of the

singular effects oftheincreasein Pco2 per seontCO2 reabsorp-tion, andlikelythese parameters as well.

Ifweassume thatH2CO3 cycling may provide acontinuous

sourceofH+for secretionin the ACTZ-treated state, theability ofAHCto increase H+ secretion relative to the control state requires explanation. The increasein proximaltCO2reabsorption afterAHC presumably reflects stimulation ofthe Na+/H+

an-tiporterin theluminalmembrane,or analteredluminal mem-branepermeabilitytoH+.Theactivity of this

antiporter

is dic-tated,in part,by the pHgradientacrosstheluminalmembrane. In addition, Aronson et al. (42) have demonstrated in brush border membranevesiclesthatareductioninintravesicular

pH

stimulates the activity oftheNa+/H+

antiporter

in a manner that suggests allosteric interaction of H+ with anactivator or

modifier site. Consequently,aparallel increasein theintracellular

and intratubularconcentrationsofH+may

potentially

enhance Na+/H+exchange withoutameasurablechangeinpH

gradient

acrosstheluminalmembrane.On the otherhand,if AHCwere todisproportionatelyincrease intracellularH+concentration in theproximaltubule oftheACTZ-treatedratrelativetothe in-tratubular H+ concentration, a more favorable pH

gradient

would beestablished across the luminal membrane andan in-crease inantiporter activity and tCO2

reabsorption might

be

2. TheH2CO3 concentrationswerecalculated from lateproximalvalues. Whether the sameconcentrations would exist along the entirelength of theproximal tubulewillrequire additional study.

anticipated.Ifa morefavorablepH gradientweregenerated be-tween cell and lumen by AHC, this may not be reflected by

bloodtolumenpH gradients.ThepH gradientsbetween blood and lumenwerevirtuallyidenticalingroups C andD.

Insummary, the present studies indicate that AHCcan

dra-matically

increase renal tCO2

reabsorption

in the absence of renal carbonic anhydraseactivity. An increase in tCO2

reab-sorption

under these circumstances canbe shown to occurin theproximalconvoluted tubule. It issuggestedthatH2CO3,

cy-cling

from luminalto

cell,

provides

acontinuoussourceof H+ forsecretion.

Appendix

Thesourceofintracellular protongeneration duringcarbonicanhydrase inhibition is theuncatalyzedreaction which consists of thefollowing:

ki

(1)CO2 hydration: CO2+ H20 -H2CO3 H'+HCO3;and(2) CO2

k-i _k

hydroxylation:H20 OH-+H'

CO2

+OH- =

HCO3.

k-2

Thelatterreaction,which makesasignificantcontributiontothe disappearanceofCO2in aqueous solutionsatpH> 7.8(41)isan ad-ditionalsourceof intracellular H'supply;atpH> 10,it dominates the uncatalyzed reaction. The relative contribution of the hydrationand hydroxylationofCO2will be,therefore,pHdependent. The maximal rateofH+ ionsupplyby the uncatalyzed reaction intheproximalnephron is:

(dH+)/(dt)=

([CO2]k

-

_[H2CO31k-1

+

_[OH-][CO2]k2

-[HCO3JkJ2

xVc,

where

V,

=_{the volumeof proximal tubule cells and}

_[CO21,

_[H2CO31,

[OH-], and[HCO-] referto theintracellular concentrations of these

species.

Coganetal.(27)have shownthat therateof the uncatalyzedH+ion supply (dH+)/(dt) depends mainlyonintracellular pHandbicarbonate concentration. Their calculation is basedontheassumptionthat the intracellular carbonic acid concentrationisinequilibriumwith intra-cellular [H+]i and bicarbonate

[HCO3Ji.

Thus:

[H2CO3Ji

=

_([H+]i

x[HCO-]i)/(2.7X I0O- M). Furthermore, the intracellular Pco2 is con-stantandis similartothecorticalPco2occurring

during

hypercapnia.

Todetermine the influence of variations inPco2ontheH+

generation

bytheuncatalyzedreaction,weextended their calculationstoinclude corticalPco2as anadditional variable.

Using

acomputer modelwe were ableto_predicttheeffect of alterations in corticalPco2

ranging

from 40

to180mmHgonH+ionsupply

independent

of

changes

inintracellular pHandbicarbonateconcentration. Themodel usesthesametubular

geometiy

dataand kineticsasusedby

Cogan

etal.

(27),

with the

exception

that

k,

and k1 werereplacedasfollows:k1 = 4.8

min-';

kL1

= 1.92

x103min'1.Theseconstants wereconsideredbyMaren(42)tobemore accurate at37°C.Fig. I showsathree-dimensional model

relating

the

rateofH+ionproductiontoboth intracellularpHandPco2atan

[HCO3]j

of 25 mM. It is clear that theH+ion

production by

the

uncatalyzed

reactionincreasesas afunctionof intracellular

pH

andPco2.A

com-parisonofH+secretoryratesatintracellular

[HCO3J,

of 15 and 30mM rather than 25mMindicated that intracellular

_[HCO3]J

startstobean

importantfactor,albeitquantitatively minor,onlywhentheintracellular pHislow,i.e.,<7.00.

Wenowexamine thepredictionsofthemodel in relationtothedata obtained inour_study

during

AHC,group C.In groupC,arterial Pco2 was 120mmHg andplasmatCO2was 32.0 mM.Calculated

plasma

bicarbonate concentration is therefore 28.4 mM. The

single

nephron

filtered load of bicarbonatewas1,355

pmol

-min-'.Tocalculateabsolute

proximalbicarbonate

reabsorption

from

proximal

tCO2 concentration of 26.2 mM,it is necessarytoobtain theexacttubularPco2thatwas

400

200

/30000

7.20

Figure1.Computermodelrelating H' secretion (DH)tointracellular Pco2andpH.

mmHg;(2) tubularPco2is identicaltoarterial Pco2,i.e., 120 mmHg. Thecalculatedratesof absoluteproximalbicarbonatereabsorptionunder theseconditionsare679pmol *min',at aPco2 of 180 mmHg, and 621 pmol *min-',at aPco2 of 120 mmHg.

Assuming that proximal bicarbonatereabsorptionis mediated only by H'secretion,ourmodel shows that in ordertoobserveratesof H+ secretion of themagnitudeobserved in groupC,anintracellular pHover 9.00at anintracellularPco2of either 120or 180 mmHg isrequired. Since suchextremelyalkaline intracellular pH values ofovernineare

unlikely,it is clear thatduringhypercapniaaswellasnormocapnia (27),

anadditionalsourceofintracellularH'mustexistto accountfor the observedrateofproximal bicarbonate reabsorption.

Acknowledgments

The authors aregratefulto Dr. F.Wiener for constructing the computer model, and toJill Scotch and R. Snyder Weiss fortheir excellent secretarial assistance.

This study was supported by National Institutes of Health grants HL 18875,HL31485,andAM 17387.

References

1. Warnock, D.G., and F. C. Rector. 1981. Renal acidification mechanism.InTheKidney.B. M. BrennerandF.C. Rector, Jr., editors. W. B.SaundersCo.,Philadelphia. 440-494.

2.Relman,A.S.,B.Esteten,and W. B. Schwartz.1953. Theregulation ofrenalbicarbonatereabsorptionby plasma carbon dioxide tension. J. Clin. Invest. 32:972-978.

3. Brazeau, P., andA.Gilman. 1953. Effect ofplasma CO2 tension

onrenal tubularreabsorptionof bicarbonate. Am. J.Physiol. 175:33-38.

4.Dorman, P.J.,W. J.Sullivan, and R. E.Pitts. 1954. The renal response to acuterespiratory acidosis. J. Clin. Invest. 54:82-90.

5.Rector, F.C., Jr.,D. W.Seldin,A.D.Roberts, Jr.,and J. S. Smith. 1960. The role ofplasmaCO2 tension and carbonicanhydrase activity in the renalreabsorptionof bicarbonate. J. Clin.Invest.39:1706-1721. 6.Seldin,D.W., R. M.Portwood, F. C.Rector,and R. Cade. 1959. Characteristics of renal bicarbonatereabsorptioninman.J.Clin. Invest. 38:1663-1671.

7. Carter, N. W., D. W. Seldin, andH.C. Teng. 1959. Effect of Diamox onplasmaandurine acid base composition duringchronic respiratory acidosis.Am.J.Physiol. 196:919-923.

8. Rector,F.C.,Jr. 1973. Acidification of theurine. In Handbook ofPhysiology.RenalPhysiology.J. Orloff andR.W.Berliner,editors. AmericanPhysiological Society,Washington,DC. 431-454.

9.Cogan, M. G. 1984. Bicarbonatereabsorption in theproximal tubuleduringcarbonicanhydrase inhibition. Ann. NY Acad. Sci. 429: 538-540.

10.Stone, J. E.,J.Wells,W. B.Draper, andR.W.Whitehead. 1958. Changes in renal blood flow indogsduring inhalation of 30%/ carbon dioxide.Am. J.Physiol. 194:115-119.

11.Farber,A.O.,J. J.Szwed,A.R.Dowell,andR. A.Strawbridge.

1976. Theacuteeffects ofrespiratoryand metabolic acidosisonrenal function in thedog.Clin. Sci. Mol. Med. 50:165-169.

12.Kittle,C. F.,H.Aoki,andE.B.Brown,Jr.1965. The role ofpH

andCO2 inthedistribution of blood flow.Surgery. 57:139-154. 13. Norman,J.N., J.Maclntyre,J. R.Shearer,I. M.Craigen,and G.Smith. 1970. Effect of carbon dioxideonrenal blood flow.Am.J. Physiol. 219:672-676.

14.Anderson,R.J., C.E.Rose, Jr.,A.S.Berns,A.L.Erickson,and P. E. Arnold. 1980. Mechanism of effect ofhypercapnic acidosis on

reninsecretion in thedog.Am. J.Physiol.238:Fl19-F125.

15.Cogan,M.G., and F. C.Rector, Jr. 1982. Determinants of prox-imalbicarbonate,chloride andwaterreabsorption duringcarbonic an-hydrase inhibition.Am.J.Physiol.242:F274-F284.

16.Shah,P. P.,M.LaPointe,C.Sabo,M.Laski,andY. L.Chan. 1983. Neuralregulationoffluid and bicarbonatereabsorptionin therat kidney.Physiologist.26:A-121.

17.Chan,Y. L. 1980.Adrenergic control of bicarbonateabsorption in proximal convoluted tubule of theratkidney.PfluegersArch. Eur.J.

Physiol.388:159-164.

18. Bello-Reuss, E., R. E. Colinders, E. Pastoriza-Munoz, R. A.

Mueller,and C.W.Gottschalk. 1975.Effectsofacuteunilateral dener-vation in therat.J.Clin.Invest.56:208-217.

19.Proll,M. A., C.W.Kamp, and W.W. Morgan. 1982. Useof liquidchromatography withelectrochemistryto measureeffects ofvarying intensities of whitelightonDOPAaccumulation inratretinas.LifeSci. 30:11-19.

20. Pucacco, L.R.,and N. W.Carter. 1976.Aglass-membrane pH microelectrode. Anal. Biochem. 73:501-512.

21.Richardson,R. M.A., andR. T.Kunau, Jr. 1982. Bicarbonate reabsorption in the papillary collectingduct: effect ofacetazolamide. Am. J.Physiol.243:F74-F80.

22. Vurek,G.G.,D. G.Warnock, andR.Corsey.1975.Measurement ofpicomoleamountsof carbon dioxidebycalorimetry.Anal. Chem.47: 765-767.

23. Duncan,D. B. 1955. MultiplerangeandmultipleFtests. Bio-metrics. 11:1-42.

24.Giebisch, G., L.Berger, andR. F.Pitts. 1955. The extrarenal responseto acuteacid-base disturbances ofrespiratory origin.J. Clin.

Invest. 34:231-245.

25.Kurtzman,N.A.1970.Relationshipof extracellular volumeand

CO2tensiontorenalbicarbonatereabsorption.Am. J.Physiol. 219:1299-1304.

anhydraseinhibitiononsuperficial and deep nephron bicarbonate reab-sorption in therat.J. Clin.Invest. 71:55-65.

27. Cogan,M.G.,D. A. Maddox,D. G. Warnock, E.T. Lin,and F.C. Rector, Jr. 1979.Effect of acetazolamideonbicarbonate reabsorp-tionintheproximal tubule ofthe rat. Am. J.Physiol. 237:F447-F454. 28. DuBose, T. D., Jr., L. R. Pucacco, D. W.Seldin,N. W.Carter, andJ. P.Kokko. 1978. Direct determinationof pCO2 in theratrenal cortex. J.Clin.Invest.62:338-348.

29. Levine, D. Z. 1971. Effect ofacute hypercapnia on proximal tubular water andbicarbonate reabsorption.Am.J.Physiol. 221:1164-1170.

30.Malnic, G.,M. DeMelloAires,andG.Giebisch. 1972. Micro-puncturestudy of renaltubularhydrogen iontransportin therat.Am. J. Physiol. 222:147-158.

31.DeMelloAires, M.,and G. Malnic. 1975. Peritubular pH and pCO2in renal tubularacidification.Am.J. Physiol. 228:1766-1774.

32. Cogan,M.G.1984.Effects ofacutealterations inpCO2on prox-imalHCO5, Cl, and H20 reabsorption.Am.J. Physiol.246:F21-F26.

33.Cogan, M.G.,andF. Y.Liu. 1983. Metabolic alkalosis in the rat.J.Clin.Invest.71:1141-1160.

34.Chan,Y.L., G.Malnic,andG.Giebisch. 1983. Passive driving forces ofproximal tubular fluid and bicarbonatetransport:gradient de-pendenceof H' secretion.Am.J.Physiol. 245:F622-F633.

35.Aronson,P.S. 1983.Mechanisms of activeH'secretion in the proximal tubule.Am.J.Physiol. 245:F647-F659.

36. Alpern,R.J., M.G. Cogan, andF. C. Rector,Jr. 1983. Flow dependence ofproximal tubular bicarbonate reabsorption.Am.J.Physiol. 245:F478-F484.

37.Alpern,R.J.,M.G.Cogan, andF.C. Rector, Jr. 1982. Effect of luminal bicarbonate concentration on proximal acidification in therat. Am.J.Physiol. 243:F53-F59.

38. Maren,T.H. 1974.Chemistryof the renalreabsorptionof bi-carbonate.Can. J.Physiol. Pharmacol. 52:1041-1050.

39.Struyvenberg, A.,R. B.Morrison, andA.S. Relman. 1982. Acid-base behavior ofseparatedcanine tubule cells. Am. J.Physiol.

214:1155-1162.

40.Kleinman,J.G.,W. W.Brown, R.A.Ware, and J. H. Schwartz. 1980.Cell pHandacid transport in renal cortical tissue. Am. J.Physiol. 239:F440-F444.

41.Maren,T. H. 1978.Carbon dioxide equilibriainthekidney: the problems of elevated carbon dioxide tension, delayed dehydrationand disequilibrium pH. Kidney Int. 14:395-405.

42.Aronson, P.S.,J.Nee, and M. A. Suhm. 1982. Modifier roleof internal H' in activating the Na-H exchanger inrenal microvillus mem-branevesicles.Nature(Lond.% 299:161-163.