Regulation of Renal Blood Flow by Plasma Chloride

CHRISTOPHERS. WILCOX, Department of Medicine and Clinical Pharmacology,Harvard Medical School, The Brigham and Women's Hospital,Boston, Massachusetts 02115

A B S T R A C T Micropuncture studies have shown thatglomerular filtration rate (GFR) falls in response to arise in Na' or Cl- concentrations in the loop of Henle,whereas studies in isolated kidneys have shown thatGFR falls in response to osmotic diuresis. To definethe separate effects of an acute increase in plasma so-dium (PNa), chloride (Pci) or osmolality (Posmoi), changesin renal blood flow (RBF) and GFR were measuredduring intrarenal infusions of hypertonic NaCl,NaHCO3, Na acetate, dextrose, NH4Cl or NH4acetateto denervated kidneys. The infusions raised Posmol atthe experimental kidney by 30-45 mosmol. RBF in-creased abruptly by 10-30% with all hypertonic in-fusions indicating that an acute increase in plasma to-nicity causes renal vasodilatation. Renal vasodilatationpersisted or increased further during infusion of dex-trose, NaHCO3and Na acetate, but GFR was un-changed. In contrast, during infusion of the two Cl-containing solutions, vasodilatation was reversed after1-5 min and RBFand GFRdecreased (P < 0.01) belowpreinfusion levels. Prior salt depletion doubled thevasoconstriction seen with hypertonic NaCl infusions.Overall, changes in RBF were unrelated to changes inPNa or fractional Na or fluid reabsorption but corre-lated with changes in P(:I (r = -0.91) and fractionalCl- reabsorption (r = 0.94). The intrafemoral arterialinfusion of the two Cl-containing solutions did not in-crease femoral vascular resistance. In conclusion, hy-perchloremia produces a progressive renal vasocon-striction and fall in GFR that is independent of therenal nerves, is potentiated by prior salt depletion andis related to tubular Cl- reabsorption. Chloride-in-duced vasoconstriction appears specific for the renalvessels.

INTRODUCTIONThe maintenance of fluid and electrolyte balance bythe kidney is achieved by integrated adjustments inglomerular filtration and tubular reabsorption. When

Received for publication 21 July 1982 and in revised form24 November 1982.

the adequacy of the extracellular volume (ECV)' isthreatened by salt depletion, impaired proximal tubulereabsorption or a high perfusion pressure, renal va-soconstriction restricts the volume of filtrate deliveredto the tubules (1-5). A mechanism that can relate glo-merular filtration to distal fluid delivery has been iden-tified in micropuncture experiments in which singlenephron glomerular filtration rate (SNGFR) has beenfound to fall during a selective increase in the Na+ orCl- concentrations or osmolality of early distal tubulefluid (6-8).

In previous experiments, we (9) and others (10, 11)showed that intrarenal infusion of hypertonic NaCIsolution causes transient vasodilatation followed bysustained vasoconstriction. Schnermann et al. (8) dem-onstrated that retrograde injection of Cl-containingsolutions into the distal tubule regularly elicited a de-crease in SNGFR, whereas similar injections of Na-containing solutions did not. In contrast, in experi-ments on the isolated kidney, Nizet (12-14) concludedthat there was no specific effect of increased Na+ orCl- concentration on renal vascular resistance but thatthe GFR varied in direct proportion to tubular fluidreabsorption. To settle these differences between theresponses of the isolated kidney and the nephron per-fused retrogradely, changes in renal vascular resistanceof the intact kidney occurring in response to an acuteincrease in plasma sodium (PNa), chloride (Pc1), or os-molality (Posmoi) were investigated.

METHODSPreparationExperiments were performed on 48 male greyhounds

weighing 28±2 kg (mean±SEM) and fed a diet of meat anddog biscuits; food and water were withheld for 12 h beforeexperiments.

' Abbreviations used in this paper: ECV, extracellularvolume; FBF, femoral blood flow; FF, filtration fraction;GFR, glomerular filtration rate; Hct, hematocrit; MAP,mean arterial pressure; PS, plasma solid; R/F, fractionalreabsorption; RPF, renal plasma flow; RBF, renal blood flow;SNGFR, single nephron glomerular filtration rate.

726 J. Clin. Invest. © The American Society for Clinical Investigation, Inc. - 0021-9738/83/03/0726/10 $1.00Volume 71 March 1983 726-735

Dogs were anesthetized with intravenous pentobarbitoland were ventilated mechanically (arterial blood pco2, 32-42 mmHg). Mean arterial blood pressure (MAP) was re-corded via a cannula in a femoral artery (as the dampedelectrical output from a pressure transducer [Statham In-struments, Inc., Oxnard, CA]).

The "experimental" kidney was denervated by autotrans-plantation to the neck of the same dog by anastomosis to thecarotid artery and jugular vein as described previously (15,16). The artery to the experimental kidney was encircled byan electromagnetic flow probe connected to a pulsed-logicblood flow meter (Biotronex Laboratory, Inc., Kensington,MD). About 1 in. distal to the flow probe, the artery wascannulated with an 18-gauge needle to infuse solution I (seebelow). A tributary of the jugular vein was cannulated toobtain samples of renal venous blood. The ureter was can-nulated and urine flow rate was displayed graphically by adrop counter (Devices Ltd., London, England) and its vol-ume measured gravimetrically. The experimental kidneywas placed in a box filled with light liquid paraffin mineraloil and maintained at 370C by a water jacket. The ureterof the contralateral kidney was cannulated, but this kidneywas otherwise undisturbed.

Through a cannula inserted into a brachial vein, a solutionof two-thirds 0.154 M NaCl solution and one-third 1.10 Mmannitol was infused at a rate of 0.04 ml min-' kg bodywt-' (solution 2). The markers used to estimate GFR andrenal plasma flow rate (RPF) were added to this infusion.During surgery, the animals received 500 ml of isooncotic(6 g 100 ml-'), dextran (average molecular weight 110,000)in 0.154 MNaCl solution (Dextraven 110, Fisons, Ltd., Lon-don, England). Infusion I (intrarenal) was delivered at 0.1ml min-' body wt kg-'. It was initially 0.154 M NaCl so-lution. 1 h after completing surgery, urine was collectedfrom each kidney for a 15-min collection period and bloodsamples (10-15 ml) were drawn from the femoral artery andthe renal vein at the midpoint. Dextraven 110 was infusedto replace blood losses.

Experimental protocolEffects of hypertonic intrarenal infusions on renal func-

tion. The infusion into the artery of the experimental kid-ney was changed from 0.154 M NaCI to either NaCl,NaHCO3, Na acetate, NH4Cl NH4acetate or dextrose at1.232 Mfor 30 min, but the rate of infusion was not changed.Urine was collected for the last 15 min during which bloodsamples were drawn. Thereafter, the infusion was returnedto 0.154 MNaCI solution for a further 30 min and urine wascollected for the last 15 min with a blood sample at themidpoint.

Effects of prior salt depletion on renal hemodynamicsduring hypertonic NaCl infusion. In five experiments, thedogs received furosemide (80 mg d-' with the food) for 2d to cause ECV depletion. The second dose was given 18 hbefore study and no further food or fluid was given. Thesedogs received the intrarenal infusions of hypertonic NaCIaccording to the protocol described above.

Effects of intraarterial infusions of hypertonic solutionson femoral blood flow. In four dogs, the hypertonic solu-tions used in protocol A were infused into a femoral arteryat 0.1 ml min-' kg body wt-'. These infusions were given sfor '-15 min with 0.154 MNaCl infused for 15 min betweeneach one. Femoral blood flow (FBF) was assessed by a non-cannulating blood flow meter placed around the femoral fartery at least 1 in. proximal to the site of the infusion.

Laboratory methods

Details of the analytical methods have been describedpreviously (16). Measurements of Na+ and K+ concentrationswere made with an EEL flame photometer, Cl with a Corn-ing chloride meter (Corning Medical, Corning Glass Works,Corning, NY), osmol with an Advanced osmometer (Ad-vanced Instruments, Inc., Needham Heights, MA), pH, pco2and P02 of arterial and renal venous blood with a Radi-ometer blood gas analyzer (Radiometer Co., Copenhagen,Denmark) and plasma HCO3 calculated from a standardnormogram (Siggaard-Andersen). Hematocrit (Hct) was es-timated as the mean of two measurements with a micro-hematocrit centrifuge. Plasma solid concentration (PS; in-dicating plasma protein concentration) was estimated bydessicating a weighed quantity of plasma to constant weightand reweighing with correction for the measured contentof ions. The radioactivity of arterial and renal venous plasmaand urine were counted with a Packard autogamma counter(Packard Instruments, Inc., Downers Grove, IL).

Calculation of results. The RPF was calculated as theclearance of Hippuran corrected for renal extraction. Therenal blood flow (RBF) was calculated from RPF and Hct.The electromagnetic flow meter was used to record the timecourse of any changes in RBF. The GFRwas calculated asthe clearance of EDTAand the filtration fraction (FF) cal-culated as GFR/RPF. Paired comparisons were used andvalues of 2P < 0.05 were taken as statistically significant.

RESULTS

Blood pressure and composition before and duringhypertonic infusions. Table I details the initial valuesof MAPand blood composition at the experimentalkidney and the changes produced by the hypertonicinfusions. Since there were no differences in initialvalues between groups, the mean starting values werepooled. The MAPwas not altered during any infusion.The Posmoj at the experimental kidney was increased-30-45 mosmol. Infusions containing Na+ or Cl- in-

creased PNa or Pci -20-25 mM. Infusion of NaCl,dextrose and especially NH4Cl produced a systemicmetabolic acidosis; infusions of NaHCO3, and Na ac-etate produced a systemic metabolic alkalosis. How-ever, there were no differences between the pH in therenal vein and the systemic circulation except duringintrarenal infusion of NaHCO3. The systemic plasmahydrogen ion concentration (PH) was higher during theinfusion of NH4-containing solutions than during thecorresponding Na-containing solutions.

Steady-state changes in RBF, GFRand renal clear-ance during intrarenal hypertonic infusions. Therenal function of the experimental kidney during theinitial infusion of 0.154 MNaCl and the relationshipof its function to that of the control kidney werestrictly comparable to that reported previously (16).There were no differences between the RBF, RPF, orfractional excretion of sodium (FENa) or chloride(FEc1) between the experimental and control kidneys,

Regulation of RBF by Chloride 727

TABLE IBlood Pressure and Composition

(:hange (tliring infusion of 1.232 M solution

Starting valne Na(C NI41, Nalt(:C) Na Acetate Dextrose NH4 Acetate(n = 48) (n = 17) (n = 6) (n = 5) (n = 6) (n = 9) (n = 5)

MAP, mmHg 164±22 -3±15 +8±15 -3±7 -8±12 -1±4 -4±7P,.,,,,,1, mosmol kg-' 310.2±11.0 +41.6±13.7° +45.5±16.9° +39.0±12.7° +33.7±7.9° +30.3±12.2° +37.6±7.5°PNa,, mmol 1' 147.4±5.8 +22.4±7.4° -1.0±0.9 +21.0±4.2° +18.5±2.3° -6.2±1.7° -0.8±1.3PK, mmol 1' 3.64±0.54 -0.3±0.30 +1.0±0.3° -0.3±0.21 -0.9±0.6t -0.2±0.2t +0.6±0.1°P(.I, mmol 1' 110.4±15.5 +22.2±5.4° +27.8±7.1° -4.4±+1.1 -4.7±2.7§ -6.0±2.7° -6.6±3.31PIl, nmrol 1' 47.6±1.8 +4.8±4.7° +17.6±7.0° -16.0±4.3° -7.8±6.31 +4.3±1.7° +0.5±4.2P(:t2, mmHg 36.7±8.9 +1.2±4.0 +0.3±8.5 +4.5±2.1 +0.8±6.5 -0.5±4.4 -1.5±6.6Pt2t mmHg 83.1±15.2 +1.5±5.1 +4.3±9.8 +0.5±5.0 +5.0±6.6 -1.3±4.4 +1.0±2.6Ptto.,3, mmol l' 18.9±0.8 -2.4±2.3° -5.2±2.7 +8.1±0.6° +3.2±2.8 t -2.2+1.2° -1.9±1.11PS, g 100 g-' 5.74±0.72 -0.53±0.39° +0.06±0.08 -0.28±0.201 -0.29±0.271 -0.08±0.14 +0.21±0.22HOt, % 48.9±7.7 -4.1±3.1° +1.8±1.8 -1.6±0.91 -0.2±3.7 -3.7±1.5° +4.2±2.5i

Column 1 shows the mean (±SD) initial value for the entire group (n = 48) and subsequent columns show the mean (±SD) absolutechanges produced in the blood perfusing the experimental kidney 15-30 min after starting intrarenal arterial infusions.'P < 0.001. IP < 0.05; §P < 0.01;

but the GFRand FF were slightly lower (average, 7%;P < 0.01) at the experimental kidney.

The changes in renal hemodynamics 15-30 min af-ter starting infusions of the four hypertonic solutionsthat did not contain Cl- are shown in Table II. Therewere significant increases in RBFand RPF for all saveNH4 acetate. There were no significant changes inGFRand FF decreased with each infusion. All thesechanges in RBF and RPF were reversible, except thatthere was a persistent fall in FF after the infusion ofthe two acetate-containing solutions. Thus, hypertonicinfusions that did not contain Cl- or NH4' led to re-versible renal vasodilatation. The changes in renal he-modynamics that occurred after 15-30 min of infusionof hypertonic solutions containing Cl- were quite dif-ferent (Table II). RBF, RPF, and GFR were all re-duced significantly and these changes were greaterduring the infusion of NH4C1 than NaCl (P < 0.05).The FF was maintained during infusion of NaCl, butit fell during NH4Cl. These changes (except for thefall in GFR with NH4Cl) had reversed within 15-30min after stopping the infusions. Thus, hypertonic Cl-infusions produced renal vasoconstriction and a fall inGFR that was more pronounced when NH4+ was thecation.

In Fig. 1 are shown the percentage changes in RBFof the experimental kidney alone and relative to thecontrol kidney during the different infusions. Duringthe infusion of Na acetate, dextrose, and NaHCO3,RBF increased, whereas during the infusion of the twoCl-containing solutions, it decreased. These changesin RBF were also seen when the RBF of the experi-mental kidney was expressed as a fraction of its control.

This indicates that the changes in the composition ofblood perfusing the experimental kidney caused thechanges in vascular resistance. However, the magni-tude of the changes in RBF were generally reducedwhen expressed as a fraction of the control-kidney,indicating that the infusions caused similar, althoughsmaller, changes in RBF of the control kidneys, pre-sumably in response to the small changes in compo-sition of the systemic blood.

There was no significant correlation between thegroup mean changes in RBF and changes in PN.(r = 0.18; n = 6; NS). However, there was a very closecorrelation between group mean changes in RBF andchanges in PcI (r = -0.91; n = 6; P < 0.01).

Relationships between changes in RBF and frac-tional reabsorption of fluid, Na' or Cl-. To examinethe hypothesis that changes in GFR or RBF duringosmotic diuresis are mediated by inhibition of tubularfluid reabsorption (12-14), the group mean changesin RBFand GFRproduced by the hypertonic infusionswere compared to changes in fractional reabsorptionof fluid (R/FH2O). No significant correlations were ap-parent between group mean changes in R/FH.O andeither RBF (Fig. 2A) or GFR (r = 0.37; NS).

During osmotic diuresis or during administration ofNH4Cl or hypertonic NaCI administration, there isconsiderable inhibition of proximal tubular reabsorp-tion of fluid, Na+ and Cl- (17-22). Consequently, thechanges in the fractional reabsorption of sodium (R/FN.) or chloride (R/Fcl) during these infusions indicatethe delivery of filtered Na+ and Cl- to the loop ofHenle. To examine the hypothesis that changes in GFRor RBF were related to changes in the delivery of

728 C. S. Wilcox

TABLE IIChanges in Renal Hemodynamics during Hypertonic Intrarenal Infusions

Before % Changecompared

Solution with RBF RPF GFR FF

NaHCO3 During +9.7±8.5 +11.0±8.8 -9.7±10.5 -21.4±12.7(n = 5) P < 0.05 P < 0.05 NS P < 0.05

After +4.5±11.2 +7.3±11.6 +2.2±14.5 -4.4±9.3NS NS NS NS

Na Acetate During +27.6±17.7 +27.6±15.3 -13.7±25.3 -29.0±12.6(n = 6) P < 0.02 P < 0.01 NS P < 0.005

After +5.6±18.7 +6.8±17.6 -10.0±11.9 -16.4±12.8NS NS NS P < 0.025

Dextrose During +14.7±13.1 +20.5±13.9 -8.4±21.0 -21.3±12.7(n = 9) P < 0.01 P < 0.005 NS P < 0.05

After -8.8±14.6 -9.0±14.3 +2.2±14.5 -4.4±9.3NS NS NS NS

NH4 Acetate During +4.8±26.4 -5.2±20.1 -25.6±28.0 -23.8±7.8(n = 5) NS NS NS P < 0.01

After +26.0±29.9 +40.8±49.4 -9.1±27.9 -33.6±12.7NS NS NS P < 0.005

NaCl During -14.4±9.2 -7.2±10.3 -9.4±11.4 -1.9±7.6(n = 17) P < 0.001 P < 0.01 P < 0.005 NS

After -5.9±13.0 +1.6±16.9 +0.5±15.9 +1.2±17.1NS NS NS NS

NH4Cl During -44.4±23.7 -48.0±21.1 -62.2±14.6 -26.4±6.2(n = 6) P < 0.01 P < 0.005 P < 0.001 P < 0.001

After -1.0±25.3 -5.0±25.9 -20.7±18.7 -6.0±16.4NS NS P < 0.05 NS

Mean (±SD) percentage changes in RPF, GFR, FF, and RBF 15-30 min after starting and 15-30 min aftercompletion of the infusion of 1.232 Msolutions compared to values before these infusions. The mean±SD initialvalues for the entire group were as follows: RBF, 9.43±3.12 ml min-' kg body wt-'; RPF, 4.73±1.51 ml min-'kg body wt-'; GFR, 1.39±0.40 ml min-' kg body wt-'; FF, 29.6±5.9%.

filtered Na+ (6) or C1- (8) to the loop of Henle, therelationships between RBF and either R/FNa or R/Fc,are presented in Fig. 2B and C. There were no sig-nificant correlations between group mean changes inR/FNa and RBF (Fig. 2B) or GFR (r = 0.45; NS). Incontrast, there were very close relationships betweengroup mean changes in R/Fc, and RBF (Fig. 2C) orGFR (r = 0.81; P < 0.05).

Since all infusions increased Posmoi, yet some led torenal vasodilatation and others to renal vasoconstric-tion, it is clear that the steady-state changes in renalvascular resistance were independent of PoSmol.

Time course of changes in RBFduring hypertonicinfusions. Fig. 3 shows the records of RBF, MAP, andurine flow from two experiments during the infusionof hypertonic solutions of Na acetate (A) and NH4C1(B). Within the first 2-3 min of starting both infusions,there was an increase in RBF without a rise in MAPindicating renal vasodilatation. However, during thesubsequent 30 min, opposite changes in renal vascularresistance occurred. Thus, with Na acetate, RBF in-creased progressively from 220 to 380 ml min-'. Incontrast, with NH4Cl, RBF fell sharply from 480 to190 ml min-'. Both changes were reversed by -15

Regulation of RBF by Chloride 729

-60 0

6) ;Na ACETATE (I

(9)

(5)

NN4ACETATE (S)

Na CI (17) 1

NN4a (6)

I .40

1<0.02

:IF

(h)EXPERMIEENTAL KIDNEYCONTROL

-40 0 .20

(6)

P .01 (a)

P <0.05 (5)

ns (S)

P <0.001 (is)

P <o.1 (S) U

j4

FIGURE 1 Mean (±SEM) values with number of observationsin parentheses for (a) percentage change in RBF at the ex-perimental kidney during the period 15-30 min after start-ing 1.232 M intrarenal infusions of the solutions shown and(b) percentage changes in renal blood flow at the experi-mental kidney compared to the control kidney during theseinfusions. The P value refers to results of paired t tests. Themean (±SD) starting values for the entire group were: RBF,9.43 ± 3.12 ml min' kg body wt-'; experimental/controlRBF, 97.7±17.1%.

min after stopping the hypertonic infusions. The in-fusion of Na acetate reduced urine flow rate, whereasthe NH4Cl increased it initially, although by the endof the infusion when GFR had fallen sharply, urineflow rate also fell.

The time course of changes in RBF were assessed

from the electromagnetic flow meter records. Per-centage changes in RBF at 1-2 min and at 30 minafter starting the infusions are shown in Table III.Within 1-2 min of starting the infusions of the dif-ferent hypertonic solutions, RBF increased in 45 of 46experiments. Compared with the 10.3% mean increasein RBF after 2 min of dextrose infusion, the rise of35.2% with Na acetate was significantly greater(P < 0.02), but the early increase in RBFwith the othersolutions was not different. Thereafter, RBF increasedfurther during the infusion of dextrose and NaHCO3;with Na acetate, the RBF increased further in six ofseven experiments. During NH4acetate, there was alarge initial rise in RBF that returned towards (but notbelow) preinfusion values.

The infusion of the two Cl-containing solutions pro-duced an early increase in RBF that was strictly com-parable to that seen with the other hypertonic solu-tions. In contrast, however, in each of the 21 experi-ments, RBFdeclined over the period 1-30 min duringthe infusion of the Cl-containing solutions. By 30 min,RBF was reduced significantly below preinfusion lev-els with both NaCl and NH4Cl; the fall in RBF wasgreater with NH4Cl than with NaCl (P < 0.01).

Effects of prior salt depletion on hemodynamic re-sponse to hypertonic NaCl infusions. The effects ofprior salt and fluid depletion on the renal hemody-namic response to hypertonic NaCl infusions was as-sessed by comparing the changes produced by intra-renal infusion of hypertonic NaCl in 17 "euvolemic"dogs with those of five "Na-depleted" dogs. As shownin Table IV, the Na-depleted dogs had a higher FFand a tendency for a lower RPF before the infusion.

MEAN±SEMo NaCI 0 N"CIa NaACETATE A NNHC03o DEXTROSE * % ACETATE

r =O.SS® Rs

I }

-Yo -A 6 -itoAR/F %H20

r=0.0

i401)

+Y .±4?-40

-20 -10 0 .10R*o N 0/ 0

r= 0.94p..:.Ol

I/TA4+

-30 -20 -10 0 *10AR/FcI 0/0

FIGURE 2 Group mean (±SEM) values for percentage changes in renal blood flow are shownas a function of changes in: (A) fractional water reabsorption, (B) fractional sodium reabsorption,and (C) fractional chloride reabsorption.

730 C. S. Wilcox

(a) EXPERIMENTALKIDNEY

DEXTROSE

NaNCO3

+40

+20

ARBFO/A

0

-20

-4j

11

1.2321 S11111 ACETATE

a

mlimin

_i -

TIME min ..........................- -

256MEANIF

mmHg

TRANSPLANT

RIFml min

SEAN IFmmHg

~Iu Feb -ii- 1 1

1 11.2320 Swills ACETTtE

EINIE lISPS

TINE m

|1-.2312 0": II.CL I

I

I

ini

25 1- o- -

II

Uin

1l.2322 ff- RL I

FIGURE 3 Records of two experiments showing the mean RBF recorded with a flow meter,mean BP and urine drop rate. An intrarenal isotonic NaCl infusion was switched to a 1.232-M infusion of (a) Na acetate and (b) NH4Cl during the periods shown.

During NaCl infusion, the falls in RBF, RPF, and GFRin the salt-depleted dogs were more than twice as greatas in the euvolemic dogs, but there were no changesin FF or blood pressure (BP) in either group. Thus,salt depletion augmented the renal vasoconstrictionthat developed with hypertonic NaCl infusions.

FBF during intraarterial infusions of hypertonicsolutions. In four experiments, the same six hyper-tonic solutions were infused into the femoral artery.All six hypertonic solutions led to an abrupt increasein FBF that was maintained for the 15-min infusionperiods. Since mean BP did not change, femoral vas-

cular resistance declined during the infusion of eachof the solutions.

DISCUSSION

Infusion of hypertonic solutions into the renal arterycaused a biphasic response in renal vascular resistance.The hyperosmolality invariably led to an abrupt renalvasodilatation. Acute vasodilatation, in response tohyperosmolality, is seen at other vessels besides thekidney (9-11, 23, 24), including the aorta (25), intactand isolated muscle arteries (26-30), the pulmonary(31), the mesenteric (10) and the coronary (30), al-though this response is not always fully maintained(30). Gazitua et al. (32) contrasted the changes in renalvascular resistance produced by hypertonic dextroseand NaCl. As in the present study, they foundthat both produced an initial fall in resistance but,whereas this vasodilatation was maintained during the

3-min infusion of dextrose, vasoconstriction developedwith NaCl. In the present study, only NaHCO3failedto elicit a significant early rise in RBF. This may

have been due to alkalosis since this has been shownto potentiate calcium-activated vascular contraction(33, 34).

The addition of hypertonic solutions to blood causes

an abrupt shrinkage of erythrocytes that restricts theirdeformality and increases the blood viscosity (35, 36).In the present experiment, however, an increase inplasma tonicity led to an abrupt reduction in renalvascular resistance and therefore, changes in bloodviscosity could only have limited the increase in RBF.

Late changes in renal vascular resistance. Theuniform early renal vasodilatation during intrarenalinfusions of various hypertonic solutions was suc-

ceeded by strikingly different patterns of renal vas-

cular resistance that depended on the composition ofthe infusate. The infusion of dextrose, NaHCO3, andNa acetate led to persistent or progressive renal va-

sodilatation; NH4acetate did not change renal vascularresistance consistently. Thus, hypertonic solutions can

cause a sustained renal vasodilatation regardless ofwhether they increase PNa. In marked contrast, in eachof the 21 experiments during which Pc1 was increased,renal vasoconstriction occurred after the initial vaso-

dilatation.Weshowed previously that the degree of renal va-

sodilatation that occurred during more prolonged in-fusions of hypertonic NaCl was dependent upon theaccompanying metabolic acidosis (9). In the present

Regulation of RBF by Chloride 731

EINIE RsOps

TABLE IIIChanges in RBF during Intrarenal Arterial Infusions of Hypertonic Solutions

Percentage changes in RBF during infusion of hypertonic solution

n After 1 min of infusion After 30 min of infusion

Dextrose 7 +10.3±5.1 +14.4±6.0During vs. before P < 0.005 P < 0.00130 min vs. 1 min P < 0.05

Na Acetate 8 +35.2±23.5During vs. before P < 0.005 +50.7±23.430 min vs. 1 min NS P < 0.001

NaHCO3 5 +10.8±11.7 +20.8±16.1During vs. before NS P < 0.0530 min vs. I min P < 0.05

NH4 Acetate 5 +25.6±20.5 +2.3±29.3During vs. before P < 0.05 NS30 min vs. 1 min NS

NaCl 14 +11.4±4.9 -12.5±12.6During vs. before P < 0.001 P < 0.00530 min vs. 1 min P < 0.001

NH4Cl 7 +18.6±11.3 -41.3±22.5During vs. before P < 0.005 P < 0.00530 min vs. I min P < 0.001

Data represent changes in RBF (mean±SD) at the experimental kidney as measuredwith an electromagnetic flow meter. Values refer to percentage changes at 1 and 30min after starting the infusions of 1.232 M solutions compared to values recorded im-mediately before the hypertonic infusions. The P values refer to results of paired t tests;n = number of experiments.

TABLE IVEffects of Prior Salt Depletion on Renal Vascular Response

to Intrarenal Hypertonic NaCl Infusion

Prior salt %Change during hypertonicdepletion Value before NaCi infusion

RBF, No 9.43±3.12 NS -14+9ml min- kg-' Yes 9.31±1.21 -29±13

RPF, No 4.73±1.51 NS -7±10 <0.05ml min' kg-' Yes 3.88±1.18 -20±11

GFR, No 1.37±0.38 -S 9±11<.0

ml min' kg-' Yes 1.43±0.52 -25±8

FF, No 29.6±5.9 <0005 -2±8 NS% ~~~Yes 36.4±3.2 -5±9

MAP, No 164±23 NS -2+10 NSmmHg Yes 165±15 -1±6

Mean±SD values for renal clearance and mean arterial pressure in dogs without priorsalt depletion (n = 17) compared to those with salt depletion (n = 5). Shown are thevalues recorded before the hypertonic infusion and the percentage changes occurring15-30 min during the infusion.

732 C. S. Wilcox

series, PH was higher and RBF was lower during theinfusions of the two NH4-containing solutions thanduring the infusion of the corresponding Na-contain-ing solutions. This suggests that the vasoconstrictionoccurring with NH4' infusions could be related tometabolic acidosis (21, 37). The same pattern ofchanges in renal vascular resistance was apparentwhen the RBFat the experimental kidney was factoredby that at the control kidney. This suggests that therenal vasoconstriction during the infusion of Cl-con-taining solutions was related to the local increasein Pci that was potentiated by an accompanyingrise in PH.

Prior salt depletion doubled the renal vasoconstric-tion that occurred during intrarenal hypertonic NaClinfusion. Gerber et al. (10) showed that acute volumeexpansion with isotonic saline abolished renal vaso-constriction during hypertonic NaCl infusions. An un-controlled salt intake might explain why renal vaso-constriction was not seen with hypertonic NaCl infu-sion in one previous study (24). Several other studieshave demonstrated vasoconstriction during infusion ofhypertonic NaCl solutions in intact and isolated kid-neys (9-11, 32).

The intrarenal infusion of all the hypertonic solu-tions that did not contain Cl- (except for NH4acetate)produced an increase in RPF but no significantchanges in GFR; the FF fell in every case indicatingthat the renal vasodilatation probably affected afferentand efferent arterioles (38). In contrast, the GFRwasreduced during the intrarenal infusion of the two Cl-containing solutions. With NaCl, the fall in GFRwasproportional to the fall in RPF and the FF remainedunchanged, suggesting that the vasoconstriction af-fected afferent and efferent arterioles. With NH4Cl,however, the FF fell considerably suggesting an ad-ditional effect on the glomerular ultrafiltration coef-ficient (39).

There are several similarities between the Cl-in-duced fall in RBF and GFRobserved in these exper-iments and the "tubulo-glomerular feedback" mech-anism identified at the single nephron level and thatis initiated by tubular Cl- reabsorption (8, 40-42).Some experiments have indicated that tubular fluidosmolality may also regulate the nephron feedbackresponse although this remains controversial (4, 43)."Feedback inhibition" of SNGFRoccurs during in-fusion of hypertonic NaCl but not hypertonic NaHCO3(44), is augmented by prior dietary salt restriction (45),and is independent of the nervous system (46). In oneseries, feedback inhibition of SNGFRwas accompa-nied by a rise in FF (47) but in a recent study, Ichikawahas shown that although feedback inhibition of SNGFRis accompanied by comparable increases in afferentand efferent arteriolar resistance, the FF falls because

SNGFRis reduced not only by the fall in plasma flowrate, but also by a large fall in the glomerular capillaryultrafiltration coefficient (39). In the present series,sustained renal vasoconstriction was specifically re-lated to hyperchloremia, was related to Cl- reabsorp-tion, was initiated by an intrarenal mechanism thatwas independent of the nervous system, was poten-tiated by prior salt depletion, was specific for the kid-ney, and was accompanied by a fall in GFR. Weob-served no change in FF during Cl-induced renal va-soconstriction with NaCl infusion, but a fall in FFduring NH4Cl. The NH4Cl produced a hyperchloremicacidosis, which has been found to impair proximal tu-bule Na+-K+ transport (21, 22), leads to enhanced de-livery of filtrate to the distal nephron (48) and in-creases UNaV (21, 23, 37). The greater vasoconstrictionapparent with NH4Cl than with NaCl in the presentseries and the potentiation of NaCl-induced renal va-soconstriction by metabolic acidosis observed previ-ously (49) may be related to inhibition of proximalfluid reabsorption during metabolic acidosis with aconsequent increase in distal Cl- delivery and poten-tiation of the feedback response. Thus, the presentstudies extend the micropuncture findings to the wholekidney level, demonstrate the specificity of hyper-chloremic vasoconstriction for the kidney and indicatea physiological role for the feedback mechanism inregulating renal vascular resistance during hyper-chloremia and metabolic acidosis.

The predominant effects of hypertonic infusions onsalt and water reabsorption should occur in the prox-imal tubule. In the isolated tubule, an increase in theNaCl concentration of the perfusate inhibits fluid reab-sorption (20). In the intact animal, NH4Cl or hyper-tonic NaCl inhibits proximal reabsorption of fluid, Na+and Cl- and increases the early distal NaCl concen-tration (48, 50) whereas hypertonic NaHCO3decreasesearly distal Cl- concentration (51). Thus, infusions ofNH4C1or NaCl will increase substantially the fractionof filtrate delivered to the loop of Henle (17, 21, 22).Accordingly, changes in R/FNa or R/Fc, should indi-cate changes in distal delivery of filtered Na+ and Cl-.The changes in RBF and GFR were independent ofchanges in R/FNa but were very closely correlated withchanges in R/Fc, (Fig. 2B and 2C), suggesting thatrenal vascular resistance was related to the deliveryof Cl-, but not Na+, to the loop of Henle. Nizet foundthat the GFRfell during osmotic diuresis or inhibitionof proximal-tubular function but that this was relatedto a fall in R/FHo and was independent of Cl- reab-sorption (12-14). His studies were conducted duringa considerable osmotic diuresis that may have in-creased the intratubular and intrarenal pressures andcould have led to passive falls in GFRand RBF. In ourstudies, the osmotic diuresis was less severe and

Regulation of RBF by Chloride 733

changes in GFR or RBF were independent ofR/FHso (Fig. 2A). Whereas Nizet (12) found thatNa2SO4, but not NaCl, reduced RBF, Sadowski (11)obtained opposite results using rates of osmolar infu-sion similar to those in the present experiments. Thus,the mechanisms identified by Nizet to regulate GFRand RBF probably operate at high rates of osmoticdiuresis. The present study demonstrates that the reg-ulation of RBF and GFRby Cl- can override the ef-fects of hyperosmolality on the renal circulation.

ACKNOWLEDGMENTS

The author gratefully acknowledges a Pharmaceutical Man-ufacturer's Faculty Award and a grant from the NationalInstitutes of Health AM27952. Wethank Miss Paula Dolanfor typing the manuscript.

REFERENCES

1. Thurau, K. 1964. Renal hemodynamics. Am. J. Med. 36:698-719.

2. McDougal, W. S., and F. S. Wright. 1972. Defect inproximal and distal sodium transport in post-obstructivediuresis. Kidney Int. 2: 304-317.

3. Ploth, D. W., J. Schnermann, H. Dalheim, M. Hermle,and E. Schmidmeier. 1977. Autoregulation and tubulo-glomerular feedback in normotensive and hypertensiverats. Kidney Int. 12: 253-267.

4. Wright, F. S., and J. P. Briggs. 1977. Feedback regu-lation of glomerular filtration rate. Am. J. Physiol. 233:Fl-F7.

5. Tucker, B. J., R. W. Steiner, L. C. Gushwa, and R. C.Blantz. 1978. Studies on the tubulo-glomerular feedbacksystem in the rat. The mechanism of reduction in fil-tration rate with benzolamide. J. Clin. Invest. 62: 993-1004.

6. Thurau, K., and J. Schnermann. 1965. Die Natriumkon-zentration an der Macula densa Zellen als regulierenderFaktar fur das Glomerulum filtrat. Klin. Wochenschr.43: 410-413.

7. Guyton, A. C., J. B. Langston, and L. G. Navar. 1964.Theory for renal autoregulation by feedback of the jux-taglomerular apparatus. Circ. Res. 15(Suppl.): 187-196.

8. Schnermann, J., D. W. Ploth, and M. Hermle. 1976.Activation of tubuloglomerular feedback by chloridetransport. Pfluegers Arch. Eur. J. Physiol. 362: 229-240.

9. Nashat, F. S., J. W. Tappin, and C. S. Wilcox. 1976. Therenal blood flow and the glomerular filtration rate ofanesthetized dogs during acute changes in plasma so-dium concentration. J. Physiol. 256: 731-745.

10. Gerber, J. G., R. A. Branch, A. S. Nies, J. W. Hollifield,and J. F. Gerkens. 1979. Influence of hypertonic salineon canine renal blood flow and renin release. Am. J.Physiol. 237: F441-F446.

11. Sadowski, J. 1972. Effects of renal artery infusion ofvarious hypertonic solutions on the renal blood flow andrenal handling of PAHin the dog. Pfluegers Arch. Eur.J. Physiol. 334: 85-102.

12. Nizet, A. 1979. Influence of replacement of chloride bysulphate upon urine excretion and glomerular filtrationrate in blood-perfused dog kidneys. Pfluegers Arch. Eur.J. Physiol. 382: 217-223.

13. Nizet, A. 1981. Connection between the changes in tu-bular reabsorption and in glomerular filtration rate in-duced by replacement of plasma sodium and chlorideby isotonic mannitol. Arch. Int. Physiol. Biochim. 89:35-40.

14. Nizet, A. 1981. Influence of uranyl nitrate upon tubularreabsorption and glomerular filtration in blood-perfusedisolated dog kidneys. Pfluegers Arch. Eur. J. Physiol.391: 296-300.

15. Wilcox, C. S. 1976. A denervated dog's kidney prepa-ration for studying the actions of drugs or hormones onrenal function and renal lymph formation. J. Physiol.265: 6-7P.

16. Wilcox, C. S. 1978. The effect of increasing the plasmamagnesium concentration on renin release from thedog's kidney: interactions with calcium and sodium. J.Physiol. 284: 203-217.

17. Giebisch, G., R. M. Klose, and E. E. Windhager. 1964.Micropuncture study of hypertonic sodium chlorideloading in the rat. Am. J. Physiol. 206: 687-693.

18. Koch, K. M., T. Dume, H. H. Krause, and B. Ochwadt.1967. Intratubularer Druck, glomerularer Capillardruckund Glomerulumfiltrat wahrend mannitoldiurese. Pfiue-gers Arch. Eur. J. Physiol. 295: 72-79.

19. Dubb, J., M. Goldberg, and Z. S. Agus. 1977. Tubulareffects of acute metabolic acidosis in the rat. J. Lab.Clin. Med. 90: 318-323.

20. Corman, B., S. Carriere, C. Le Grimellec, and J. Car-dinal. 1980. Proximal tubular response to variations inextracellular sodium concentration. Am. J. Physiol. 238:F256-F260.

21. Wilcox, C. S. 1981. The renal response to salt loading.In Advances in Medical Science; Kidney and BodyFluids. L. Takas, editor. Pergamon Press, InternationalUnion Physiol. Sci., 11, pp. 621-629.

22. Cemerikic, D., C. S. Wilcox, and G. Giebisch. 1982. In-tracellular potential and K+ activity in rat kidney prox-imal tubular cells in acidosis and K+ depletion. J.Membr. Biol. In press.

23. Harvey, R. B. 1960. Vascular resistance changes pro-duced by hyperosmotic solutions. Am. J. Physiol. 199:31-34.

24. Navar, L. G., A. C. Guyton, and J. B. Langston. 1966.Effect of alterations in plasma osmolality on renal bloodflow autoregulation. Am. J. Physiol. 211: 1387-1392.

25. Bohr, D. F., D. C. Brodie, and D. H. Cheu. 1953. Effectof electrolytes on arterial muscle contraction. Circula-tion. 17: 746-749.

26. Haddy, J. F. 1960. Local effects of sodium, calcium andmagnesium upon small and large blood vessels of thedog forelimb. Circ. Res. 8: 57-70.

27. Raizner, A. E., J. C. Costin, R. P. Croke, J. B. Bishop,T. V. Inglesby, and N. S. Skinner. 1973. Reflex, systemicand local hemodynamic alterations with experimentalhyperosmolality. Am. J. Physiol. 224: 1327-1333.

28. Arvill, A., B. Johansson, and 0. Jonsson. 1969. Effectsof hyperosmolarity on the volume of vascular smoothmuscle cells and the relation between cell volume andmuscle activity. Acta Physiol. Scand. 75: 484-495.

29. Mellander, S., and J. Lundvall. 1971. Role of tissue hy-perosmolality in exercise hyperemia. Circ. Res. 28:139-I45.

30. Gazitua, S., J. B. Scott, B. Swindall, and F. J. Haddy.1971. Resistance responses to local changes in plasmaosmolality in three vascular beds. Am. J. Physiol. 220:384-391.

31. Hauge, A., and G. Bo. 1971. Blood hyperosmolality and

734 C. S. Wilcox

pulmonary vascular resistance in the cat. Circ. Res. 28:371-376.

32. Gazitua, S., J. B. Scott, C. C. Chou, and F. J. Haddy.1969. Effect of osmolarity on canine renal vascular re-sistance. Am. J. Physiol. 217: 1216-1223.

33. Young, D. B., and H. H. Rostorfer. 1973. Blood flow andfiltration rate responses to alterations in renal arterialosmolarity. Am. J. Physiol. 225: 1003-1008.

33. Betz, E., and M. Csornai. 1978. Action and interactionof perivascular H+, K+ and Ca++ on pial arteries.Pfluegers Arch. Eur. J. Physiol. 374: 67-72.

34. Grun, G., U. Weder, R. Volkmann, and A. Fleckenstein.1972. The mutual antagonism between H and Ca ionsin the control of vascular tone and autoregulation: Neu-tralization of Ca-deficiency by experimental alkalosis.Pfluegers Arch. Eur. J. Physiol. R9: 338-339.

35. Meiselman, H. J., E. W. Merrill, E. R. Gillilan, G. A.Pelletier, and E. W. Salzman. 1967. Influence of plasmaosmolarity on the rheology of human blood. J. Appl.Physiol. 22: 772-781.

36. Schmid-Schonbein, H., R. E. Wells, and J. Goldstone.1973. Effect of ultrafiltration and plasma osmolalityupon the flow properties of blood: A possible mechanismfor control of blood flow in the renal medullary vasarecta. Pfluegers Arch. Eur. J. Physiol. 338: 93-114.

37. Gordon, D., F. S. Nashat, and C. S. Wilcox. 1981. Ananalysis of the regulation of sodium excretion duringinduced changes in plasma sodium concentration inanesthetized dogs. J. Physiol. 314: 531-545.

38. Baylis, C., W. M. Deen, B. D. Myers, and B. M. Brenner.1976. Effects of some vasodilator drugs on transcapillaryfluid exchange in renal cortex. Am. J. Physiol. 230:1148-1158.

39. Ichikawa, I. 1982. Hemodynamic influence of altereddistal salt delivery on glomerular microcirculation. Kid-ney Int. 22: S109-S113.

40. Schnermann, J., F. S. Wright, J. M. Davis, W. Stackel-berg, and G. Grill. 1970. Regulation of superficial neph-ron filtration rate by tubulo-glomerular feedback. Pflue-gers Arch. Eur. J. Physiol. 318: 147-175.

41. Wright, F. S., and A. E. G. Persson. 1974. Effect ofchanges in distal transepithelial potential difference onfeedback control of filtration. Kidney Int. 6: 1114a.

42. Wright, F. S., and J. Schnermann. 1974. Interferencewith feedback control of glomerular filtration rate byfurosemide, triflorin and cyanide. J. Clin. Invest. 53:1695-1708.

43. Navar, L. G., P. D. Bell, C. E. Thomas, and D. W. Ploth.1978. Influence of perfusate osmolality on stop-flow pres-sure feedback responses in the dog. Am. J. Physiol. 235:F352-F358.

44. Schnermann, J., J. Briggs, and F. S. Wright. 1981. Feed-back-mediated reduction of glomerular filtration rateduring infusion of hypertonic saline. Kidney Int. 20:462-468.

45. Dev, B., C. Drescher, and J. Schnermann. 1974. Reset-ting of tubuloglomerular feedback sensitivity by dietarysalt intake. Pfluegers Arch. Eur. J. Physiol. 346: 263-277.

46. Schnermann, J., N. Stowe, S. Yarimizu, M. Magnusson,and C. Tingwald. 1977. Feedback control of glomerularfiltration rate in isolated, blood-perfused dog kidneys.Am. J. Physiol. 233: F217-F224.

47. Briggs, J. P., and F. S. Wright. 1979. Feedback controlof glomerular filtration rate: Site of the effector mech-anism. Am. J. Physiol. 236: F40-F47.

48. Levine, D. Z., and L. A. Nash. 1973. Effect of chronicacidosis on proximal tubular H20 and HCO3reabsorp-tion. Am. J. Physiol. 225: 380-384.

49. Nashat, F. S., J. W. Tappin, C. S. Wilcox. 1976. Plasmasodium concentration and sodium excretion in the anes-thetized dog. J. Physiol. 254: 183-202.

50. Malnic, G., A. Mello-Aires, and F. Lacaz-Viera. 1970.Chloride excretion in nephrons of rat kidney during al-terations of acid-base equilibrium. Am. J. Physiol. 218:20-29.

51. Rector, D., and J. P. Clapp. 1962. Evidence for activechloride reabsorption in the distal tubule of the rat. J.Clin. Invest. 41: 101-108.

Regulation of RBF by Chloride 735