Journal of Clinical InvestigationVol. 42, No. 2, 1963

AMIINNO ACID EXTRACTIONANDAMMONIAMETABOLISMBYTHE HUMANKIDNEY DURINGTHE PROLONGED

ADMINISTRATION OF AMMONIUMCHLORIDE*

By EDWARDE. OWENAND ROSCOER. ROBINSON

(Front the Departmienit of Medicine, Duke University Medical Cciiter, anld the Veterans Ad-ministration Hospital, Durham, N. C.)

(Submitted for publication June 25, 1962; accepted November 1, 1962)

According to current concepts, renal ammoniasynthesis may be attributed to the renal extractionand catabolism of certain plasma amino acids.In 1943, Van Slyke and associates demonstratedin the dog that the amide nitrogen of glutaminewas removed from arterial plasma in quantitiessufficient to account for approximately 60 percent of the urinary ammonia excreted during met-abolic acidosis (1). It was proposed that the re-maining 40 per cent of urinary ammonia could beaccounted for by the renal uptake of plasmaa-amino nitrogen. In agreement with this hy-pothesis, many investigators have since shownthat the administration of several amino acidsother than glutamine is associated with an in-creased urinary ammonia excretion (2-4). Load-ing experiments of this type have indicated that-besides glutamine-glycine, alanine, asparagine,leucine. and histidine may serve as precursors ofurine ammonia. The demonstration of appropri-ate enzyme systems within renal tissue capable offorming ammonia from these substrates has beenoffered as additional support for the thesis thatseveral plasma amino acids may participate nor-mally in the renal production of ammonia. Nev-ertheless, the specific amino acids that serve asrenal ammonia precursors have not been identi-fied clearly in conditions associated with normalacid-base balance and normal plasma amino acidconcentrations. The present study was under-taken for two principal reasons: first, to charac-terize the' typical patterns of uptake or release

* Investigation supported in part by U. S. Public HealthService research grants A-5930 and A-6306 from the Na-tional Institutes of Health, Bethesda, Md., and by theVeterans Administration Clinical Investigator Program;presented in part at the annual meeting of the SouthernSection, Amer. Fed. Clin. Research, New Orleans, La.,January 18, 1962.

of individual amino acids by the normal humankidney, and second, to determine whether or notthe adaptive increase of renal ammonia productionduring metabolic acidosis could be attributed tochances in the renal extraction of certain plasmaamino acids.

METHODS

Twelve renal clearance experiments were carried outon eleven healthy male volunteers between 21 and 38years old. All studies were conducted with them in therecumbent position after an overnight fast. To facili-tate the collection of timed urine specimens, urine flowsof approximately 5 ml per minute were assured by theoral ingestion of 500 ml of tap water 1 hour before thetest, followed immediately by a constant intravenous in-fusion of 5 per cent dextrose in water. After appropri-ate priming doses, an- intravenous maintainance infusionof a solution of inulin and para-aminohippurate (PAH)was begun 45 minutes before the first clearance periodvia a Bowman constant-infusion pump. During theequilibration period, a no. 7 or 8 cardiac catheter wasintroduced into a renal vein through either the rightbrachial or femoral vein under fluoroscopic control. Anindwelling Cournand arterial needle was placed in theleft brachial artery. Three sequential clearance periods,each 20 to 40 minutes long, were obtained during all ex-periments. Simultaneous arterial and renal venous bloodsamples were drawn for analysis at the approximate mid-point of each urine collection period. Urine was col-lected without the aid of bladder catheterization byvoiding into a flask containing a small amount of min-eral oil.

Six of the 12 experiments were performed during thepresence of a mild metabolic acidosis induced by thedaily oral ingestion of 6 to 8 g of nonenteric coated am-monium chloride for 6 to 9 days before the procedure.The first clearance period was obtained 8 to 12 hoursafter administration of the last dose of ammonium chlo-ride. In one subject, G.J., experiments were performedboth before and after the prolonged administration ofammonium chloride.

The free amino acid content of urine and arterial andrenal venous plasma was determined by the ion exchangecolumn chromatographic technique of Spackman, Stein,

263

EDWARDE. OWENAND ROSCOER. ROBINSON

and Moore (5) with a Beckman Spinco model 120amino acid analyzer. With the exception of glutamine,an accuracy of + 3 per cent was obtained regularly withthis procedure when 0.1 to 0.3 Amole of an amino acidwas placed on the column for analysis. The total quan-tity of most free amino acids placed on the columns fellwithin this range when plasma and urine samples wereprepared for analysis by the following procedures.

Plasma. Protein precipitation was carried out im-mediately after collection by the addition of 50 ml of 1per cent picric acid to 10 ml of plasma. After centrifu-gation, excess picrate was removed from the supernatantfluid by filtration of a 50-ml sample through a 2 X 2-cmcolumn bed of Dowex 2 X 8 resin. The column bed wasthen washed with 15 ml of 0.02 N HCl, and the totaleluate was lyophilized and reconstituted in 4.2 ml ofcitrate buffer at pH 2.2. These samples were stored at-200 C until analysis, when 2-ml samples were placedon the columns.

Since glutamine cannot be separated reliably fromasparagine under the usual conditions of chromatography(6), arterial and renal venous plasma concentrations ofglutamine were also measured by an enzymatic assaytechnique using a specific glutaminase obtained fromE. coli, strain W. The enzyme was prepared accordingto Meister, Levintow, Greenfield, and Abendschein (7)and the glutamine assay was carried out in duplicate asdescribed by Segal and Wyngaarden (8). Duplicate de-terminations agreed within ± 8 per cent and recoveries ofglutamine added to plasma ranged between 94 and 106 percent consistently. Plasma glutamine values obtained byenzymatic assay were between 20 and 30 per cent higherthan the glutamine-asparagine concentrations measuredby column chromatography.

Urine. The pH of 10-ml samples of urine was ad-justed to between 11.5 and 12 by the drop-by-drop addi-tion of 4 N NaOH. After 6 hours of vacuum desicca-tion, the urine samples were acidified to pH 2.0 to 2.2with 6 N HCl and their volume readjusted to 10 ml bythe addition of citrate buffer at pH 2.2. The sampleswere then stored at -20° C until analysis. The appro-priate sample volume to be placed on the column forfree amino acid analysis was calculated by the methodof Stein (9).

Blood and urine ammonia was measured by a modifi-cation of the microdiffusion method of Brown, Duda,Korkes, and Handler (10). Inulin was determined by theresorcinol method of Roe as modified by Schreiner (11).PAH was measured by the technique of Selkurt (12).The pH of whole blood and urine was determined with aCambridge model R pH-meter equipped with an en-closed glass electrode. Measurements were made atroom temperautre and were corrected to 370 C withRosenthal's factor (13). The total CO2 content of ar-terial blood was determined by the method of Van Slykeand Neill (14). Assuming a pK of 6.11, the plasma CO2tension was calculated from the line chart of Van Slykeand Sendroy (15). Oxygen content and saturation ofarterial blood were determined by the spectrophoto-

metric method of Hickam and Frayser (16). The he-matocrit of arterial blood was measured in duplicate bythe method of Wintrobe (17).

Calculations. Glomerular filtration rate was measuredby the clearance of inulin. True or total renal plasmaflow was estimated from the clearance and extraction ofPAH. Total renal blood flow was calculated from thisvalue and the peripheral arterial hematocrit (18).

Estimates of the net uptake or release of each freeamino acid across the renal circulation were obtainedby multiplying the arteriorenal venous amino acid dif-ference by total renal plasma flow. Because urinaryamino acid excretion rates were not measured in allstudies, these values were not used in the calculation ofnet renal uptake or release. Because of their small mag-nitude, however, their use in such calculations would nothave affected materially the directional patterns of up-take or release that were obtained from the use of A-Vdifferences alone. The total amino acid nitrogen valuesused in similar balance calculations reflect the summa-tion of the nitrogen content of each amino acid meas-ured by column chromatography.

Renal venous ammonia release was calculated from theA-V difference of ammonia and the total renal bloodflow. The total bidirectional release of ammonia wasobtained by summing the values for urine ammonia ex-cretion and renal venous ammonia release.

RESULTS

Experiments in healthy human subjects withoutprior administration of ammonium chloride. Forall clearance periods in these 6 experiments, glo-merular filtration rate and total renal blood flowaveraged 157 and 1,491 ml per minute, respec-tively. Extraction ratios for PAH remainedsteady throughout each procedure and averaged88 + 5 per cent. The plasma bicarbonate concen-tration averaged 25.5 ± 1.3 mmoles per L, plasmapCO2 was 46 + 3.9 mmHg, and arterial pH,7.35 +.02 units. A modest decrease of arterialpH and elevation of arterial pCO2 was observed inone subject, but could be attributed to the presenceof mild pulmonary hypoventilation induced bysleep. Urine pH remained constant in each sub-ject and averaged 5.98 +.33 for the entire group.

Table I lists the individual values of each of the6 patients for renal blood flow, urine pH, ammoniarelease into urine and renal venous blood, and thearterial and renal venous concentrations of am-monia and those amino acids whose net extractionor release appeared of definite magnitude. In ad-dition, the average arterial plasma concentrationsof all 19 free amino acids are summarized for theentire group in Table II, with the appropriate

264

RENAL EXTRACTIONOF AMINO ACIDS DURING AMMONIUMCHLORIDE ADMINISTRATION 265

o m m

.00< o +

Ili n Ue

eM If U)t-6. 6 c

- eM

o- o4 6

0_

+

If) ee)

+

' 00%o o6

In)

t-

+l

0% - -uif 0-

r-~ 6 r.-

_i _

0 0 ef

C1o _

0

rq 0% 'ICeli 4 eMCN eM Ce

U) -,

o6 00 4I eq - 00

CN -0 0%If) r~ I)

6 - )

o t0 0r-Z 06 6

%0 Uf) 0- - 0

+

0- V

_- NU)to6If)

(0 r- 0%.6 . .4

-4

oo00q

00 Vef) Ili if;ei

r- O M.z 11)

1- _- (N1

Vc 00 r-'6 -

eM e en

+

Zr- 0., eM00 (,, 6

+

00 0, 00%d6 00

00

o41

0%00in

Vo'I'I

tt 3E-30C 20 CD0 E0o

)0 U cl cd0 U0 00 C

cd - Y; 3 E co 0 3 ce 1. 0 1- U -93

> > >

0co0

0 0C0c 02 02t M E =

H

-e

(.d 4-j

CC

U)

4 L._

CO-j

_Cld

0) <

U) U)o .

0)

C'sO

4._

cO 00

C' 't=

,aa

-o -~c d

4C C0

M 0

0 Q

0

.0 .o C

2= E-

iC0C CO*- 4_& J

4C) 0)00

0)

o E0-r 3 o @o

0 UCcd(U)

00) C. CO*

C.)

C)

* -*--°4°4.I

Cd

U)

coU)

0

T4

U)|

7)

Wr.Im.ju

._

Un

r.

0

E

+ F

.5r_

¢

00

z

c o3mC°1 0L m:

0

.; 'd4tn

EDWARDE. OWENAND ROSCOER. ROBINSON

00 00 0 0% 00 %0os oE 6 '6o

o in 00O rz

C; CS6- t_ ,

o o U

_14

6- 00

- t. 0%_,+N U) 00

\~0 -

00 U)1 00

c_ 4

U) U) 4

+

vo 0% U

- 00t- 0 %0

+

-4 00

'444t'c" '0U)~

e#)'IO

00(DU)

*4 4. 9

4. 0 .1

-0

o o 0

6-

o6 4~+ s +

ON (Neq

d 0

00 U

u) u) d'+

O 0% (N4 U)

0 ~ . t.

en~ 00 0

C4 00 00

U) 440 0 r)

t. . 00

6 o

_4 , U)

:.+

+

+

4 .60

I)(N;

00U)

0

0%

C13 - (3)4- '.- 4) 4- 4)

- 4 E- a-E - z- E -- C4 E E +

. > 0 C13¢n 0 0 co.~ ~-4%- - C

266

44

0

E,i

4._

V._¢

44

0

4)r.

w

4)

v

44

E-acd t-&

44.-

44

0.EC:

it O0(, N

Ne- cl

0 0 o 3

V)

RENALEXTRACTION OF AMINO ACIDS DURINGAMMONIUMCHLORIDE ADMINISTRATION 267

average values for either their net renal uptake orrelease. In these studies, urine ammonia excre-tion averaged 37.8 + 25.6 /Amoles per minute, re-nal venous ammonia release was 37.3 ± 24.6Ixmoles per minute, total renal ammonia release,77.4 + 28 emoles per minute, the arterial bloodammonia concentration, 5.2 + 1.6 1.moles per 100ml, and the venous blood ammonia level averaged7.8 ± 2.2 ,moles per 100 ml.

Of those amino acids-glutamine, proline, andglycine-that exhibited the most definite figuresof net uptake from plasma by the kidney (TablesI and II), only glutamine was extracted consist-ently by all 6 subjects. Net glutamine uptakeaveraged 47 + 14 lumoles per minute for the en-tire group. Net renal extraction of proline oc-curred in 5 of 6 studies, and glycine exhibited netuptake in 4 of 6 experiments (Table I). Althoughvaline also demonstrated net extraction duringmost experiments, the average net uptake of thisamino acid was much less than that of either glu-tamine, proline, or glycine (Table II).

Serine, glutamic acid, alanine, cystine, and ar-ginine comprised the group of amino acids thatclearly exhibited average values for net releaseinto renal venous plasma (Tables I and II). Of

TABLE II

Net renal uptake or release of free amino acids in healthyhuman subjects*

Arterial plasma Net uptake (+) orconcentration net release (-)

Amino acids Average Range Average Range

l.moles/100 ml imoles/minuteTaurine 6.1 4.8- 8.7 +2.5 +15.2 to -5.7Threonine 10.6 8.0-13.3 - 0.1 + 9.0 to - 7.2Serine 8.8 7.2-10.1 -30.1 - 8.2 to -84.5Glutaminet 64.9 52.3-71.0 +47.4 +68.0 to +25.0Proline 17.4 11.6-24.8 + 9.8 +30.3 to - 8.1Glutamic 12.4 5.2-17.4 - 5.8 + 3.9 to -15.9Glycine 18.4 15.7-24.5 + 4.2 +15.8 to -10.0Alanine 26.1 20.0-35.6 -16.5 + 3.4 to -31.3Valine 18.2 10.9-30.6 + 2.2 +22.3 to - 7.0Half-cystine 8.4 5.8-11.9 -10.8 0.0 to -17.9Methionine 1.4 1.1- 2.5 + 0.4 + 5.0 to - 3.5Isoleucine 4.3 1.7- 8.6 + 2.0 + 4.0 to - 0.8Leucine 7.9 4.6-12.9 - 2.4 + 4.5 to - 6.0Tyrosine 3.3 2.0- 4.8 - 4.0 + 0.4 to -11.1Phenylalanine 3.6 2.8- 4.8 - 0.2 + 3.4 to - 6.5Ornithine 4.9 3.2- 6.9 - 0.8 + 0.2 to - 1.8Lysine 10.4 10.3-10.6 - 2.0 + 1.3 to - 5.4Histidine 6.8 6.1- 7.2 0.0 + 4.0 to - 3.6Arginine 6.4 4.4- 8.6 - 6.0 + 1.8 to -16.1

* Net uptake or release was calculated by multiplying each arterio-renal venous amino acid difference by total renal plasma flow.

t Values listed for glutamine were obtained from enzymatic analysisrather than from column chromatography.

these, only serine was added to renal venous bloodby all subjects (Table I), and usually in consid-erable quantity, averaging 30 ± 27 /Amoles perminute. Net venous release of the other aminoacids in this group was noted in at least 4 of 6experiments, and in some instances (alanine andcystine), their addition to venous plasma was ofrespectable proportions (Tables I and II). Glu-tamic acid also exhibited net release in 4 of 6 stud-ies, averaging 5.8 /.moles per minute for all sub-jects. Although leucine and tyrosine were usuallyadded to renal venous plasma, the average valuesfor their net release were smaller than those ofeither serine, glutamic acid, alanine, cystine, or ar-ginine (Table II).

The arteriorenal venous differences of the re-maining amino acids-taurine, threonine, methio-nine, isoleucine, phenylalanine, ornithine, lysine,and histidine-were so small or variable that noconclusions could be drawn as to the presence ofeither net uptake or net release (Table II).

Total ammonia nitrogen release by the kidneysduring these experiments averaged 1.05 ± 0.36mg per minute (Table V). Of this amount, 49.5per cent (0.52 ± 0.34 mgper minute) was releasedinto renal venous blood and the remaining 50.5 percent, (0.53 + 0.18 mg per minute) was excretedin urine. Despite the addition of more free aminoacids to renal venous plasma than were extractedfrom arterial plasma, the net balance of totalamino acid nitrogen across the renal circulationwas positive (net uptake) in 5 of 6 subjects, av-eraging 0.53 ± 0.59 mg per minute for the entiregroup. From these data it is apparent that thetotal renal release of ammonia nitrogen cannot beaccounted for solely by an analysis of the netrenal extraction of total amino acid nitrogen fromplasma. Similarly, the average net uptake of gly-cine and the amide nitrogen of glutamine (0.72 +0.22 mg per minute) could account for only 70per cent of the total renal ammonia nitrogen re-lease. The average net extraction of glycine andglutamine total nitrogen (1.39 ± 0.41 mgper min-ute), however, was more than sufficient (132 percent) to provide the total amount of ammonianitrogen released by the kidney. In fact, the re-nal extraction of glutamine total nitrogen (1.33 +0.41 mg per minute) was the major determinantof the modest positivity (net uptake) of totalamino acid nitrogen extraction observed in these

EDWARDE. OWENAND ROSCOER. ROBINSON

Mt 1-14 0 0% 0 009 c9 0 _ C9 -_

_- _ . _ -

00 t- (

o. el 06

cli C r

. 0_ d

00 0'ICe 6

CSN +

+

4. . 4o -f U)

oo o +

\0 - -U)

U) '0 +

U)_

) -:

~6 o6 6

en C., -.116 4 6+

to .4U- -U

Ul)6~

4-

~6+

+

00 - -C- cli 4

-

-

1-

Oo0

00

-4

to6

U) - _

+

Ild 00 C1I4(9~~~~~~~~~~~~~~~

4 eU 'IC

+

'0 d' '0

00

00

0%00

'0

U)

U)(9

-

~~~~~~~~~ ~ ~ ~~~+(C)Z ) En O = C) cd a 0v- -I

M

°Cd rn w

O&- w M M

_Q > Q 4- 4. V > V 4-- V 0 > 0.e o _ oe o o do o

U0 0 50 0 0d~ 0 =)

-t -g -=-g-.~=

¢4 4) ¢o

c d

cis

4>

V)H

4)

_

J44U L.

' U)

(U (e 4

2 -U

cd *> d

(L) D-

4)J

cdD.

4U) (UU) 4-00 )

4)0'=r H

cro

0

= U4=X.Z0 4r34i.$. () -)

° )0 U

Sw

U) - 0 0q

4-00 t- t.

-0

E 0c()OU00 4)-~

r. E 404.-J

268

4)r.

4.-clt= co

0

Nc (s r-

cl eli O.

t4 00 6

+

.14 U 00r=- U); -4+-

44.5055'4o0U)

C)0

0.

¢i

4)

.0

0

0

0s

- 4)

00

-

4)

-4

4)

0SU).E

0SS

cd

z

000

. !00E

:E

3 M

- -I:

0

00

U)0

r -U

1- i1

b

RENALEXTRACTIONOF AMINO ACIDS DURINGAMMONIUMCHLORIDE ADMINISTRATION 269

- \0 _o~

EDWARDE. OWENAND ROSCOER. ROBINSON

subjects. If glutamine total nitrogen was ex-cluded from this calculation, an average net addi-tion of total amino acid nitrogen to renal venousblood occurred that averaged 0.80 ±+.61 mg perminute.

It should be emphasized that the amino acidnitrogen balances listed above were obtained onlyacross the renal circulation. They do not repre-sent total balance calculations, since urinary aminoacid excretion rates were not included. Althoughurine amino acid excretion was measured in twosubjects, the values were normal and their in-clusion in the calculations would not have ap-preciably altered the individual amino acid bal-ances (net uptake or release) across the renalcirculation.

Experiments in healthy human subjects duringmild ammonium chloride acidosis. Values for glo-merular filtration rate and total renal blood flowwere within accepted normal limits during all sixstudies. Filtration rate averaged 125 ml per min-ute and total renal blood flow 1,366 ml per minute.PAH extraction ratios were almost identical tothose of the nonammonium chloride subjects and

remained constant throughout each individualstudy. Plasma bicarbonate concentrations av-eraged 22 + 2 mmoles per L, arterial pCO2, 395 mmHg, and arterial blood pH, 7.33 + 0.05.Urine pH was stable throughout each experiment,averaging 5.36 + 0.20 for the entire group.

The individual values of renal blood flow, urinepH, renal venous ammonia release, urine ammoniaexcretion, and the arterial and renal venous con-centrations of ammonia and certain amino acidsare listed in Table III for each subject. As be-fore, the group averages for the arterial plasmaconcentration, urine excretion rate, and net up-take or release of all 19 amino acids are sum-marized in Table IV. As anticipated, the pro-longed administration of ammonium chloride wasassociated with a higher average value for totalrenal ammonia release (139.1 + 42 /emoles perminute) than was observed in those subjects whodid not ingest the drug. In these subjects, thearterial blood ammonia averaged 4.4 + 1.3 pmolesper 100 ml, the renal venous ammonia concen-tration, 7.5 2.4 Mmoles per 100 ml, and theurine ammonia excretion, 96.8 + 31.9 Mumoles per

TABLE IV

Net renal uptake or release and urinary excretion of free amino acids in healthy humansubjects after prolonged oral ingestion of ammonium chloride*

Arterial plasma conc. Net uptake ( +) or net release (-) Urinary excretiont

Amino acids Average Range Average Range Average Range

jumoles/lOO ml pmoles/minule Mmoles/minuteTaurine 6.4 3.7-10.4 + 4.2 + 13.3 to - 2.9 1.63 0.98-2.49Threonine 7.8 6.0-11.4 + 0.1 + 5.4 to - 5.2 0.13 0.10-0.15Serine 5.8 3.9- 7.9 - 24.3 - 16.4 to -48.6 0.22 0.16-0.31Glutaminet 40.3 31.7-46.3 +102.5 +203.8 to +70.1 0.37 0.26-0.73Proline 14.7 10.0-20.8 + 11.5 + 33.6 to + 2.0 TraceGlutamic 9.2 6.3-12.1 + 11.5 + 15.3 to + 8.6 TraceGlycine 15.7 12.8-20.8 + 5.8 + 12.7 to + 0.7 0.63 0.27-1.19Alanine 23.2 13.5-30.8 - 15.2 - 1.9 to -24.4 0.24 0.15-0.37Valine 19.0 15.1-26.8 + 11.7 + 40.1 to - 2.5 0.09 0.03-0.14Half-cystine 10.3 6.8-14.4 - 9.0 - 0.2 to -22.1 0.21 0.15-0.27Methionine 1.3 1.0- 1.6 - 0.1 + 1.8 to - 0.9 TraceIsoleucine 3.6 2.8- 5.1 + 0.1 + 1.0 to - 1.4 TraceLeucine 7.6 4.4-12.2 + 0.2 + 4.0 to - 2.9 0.04 Tr-0.08Tyrosine 3.7 2.6- 4.6 + 0.2 + 4.6 to - 3.1 0.11 0.06-0.13Phenylalanine 3.7 3.1- 5.1 0 + 1.5 to - 1.0 0.08 0.03-0.11Ornithine 5.2 4.1- 6.1 + 2.7 + 7.3 to - 1.6 TraceLysine 12.9 11.8-14.4 + 4.6 + 24.7 to - 7.3 0.29 0.07-0.64Histidine 5.9 5.1- 7.5 + 2.5 + 8.6 to - 4.2 0.55 0.34-1.23Arginine 4.9 3.2- 6.4 - 5.2 - 2.4 to -12.3 Trace

* Net renal uptake and release were calculated from A-V differences multipled by the total renal plasma flow.t Measurable amounts of 1-methyl and 3-methyl histidine, a-aminobutyric, phosphoethanolamine, and ethanolamine

were present in urine and accounted for an average 20 per cent of total amino acid nitrogen excretion. They have notbeen listed because of the lack of corresponding plasma values.

t Plasma values were measured by enzymatic assay. Urine values were determined by chromatography and repre-sent the sum total of glutamine and asparagine concentrations, most of it asparagine.

270

RENAL EXTRACTION OF AMINO ACIDS DURINGAMMONIUMCHLORIDEADMINISTRATION 271

AMINOACID

TAURINE

THREONINE

SERINE

GLUTAMINE

PROLI NE

GLUTAMIC

GLYCINE

ALANINE

VALINE

HALF- CYSTINE

METHIONINEISOLEUCINE

LEUCINE

TYROSINEPHENYLALANINE

ORNITHINE

LYSINEHISTIDINEARGININE

AVERAGE NET UPTAKE

rIE

z WITHOUT AMMONIUMCHLORIDE

M AFTER AMMONIUMCHLORIDE

U

U

I

I

+ 100 + 75 + 50 + 25

jumoles / min.

AVERAGENET RELEASE

J

)

0 - 25 - 50

jamoles / min.

-75

FIG. 1. CHANGESOF NET RENAL UPTAKEOR RELEASE OF AMINO ACIDSADMINISTRATION.

minute. Since the renal venous release of am-monia was essentially the same both before andafter ammonium chloride administration (37.3 +24.6 and 42.3 + 14.9 /moles per minute, re-spectively) the elevated total ammonia releaseduring mild acidosis appeared to be largely re-lated to an increased urinary excretion of am-monia. This rise was highly significant statisti-cally (p < 0.01), and could not be explainedsatisfactorily by the relatively small reduction ofurine pH that occurred after ammonium chlorideadministration (Table III).

Values from both experimental groups for theaverage net uptake or release of each amino acidhave been illustrated in Figure 1 for purposes ofgraphic comparison. Ammonium chloride ad-ministration was unassociated with a significantchange in the plasma levels of the various aminoacids (Tables III and IV), with the exception ofglutamine. In general, plasma amino acid con-

INDUCED BY AMMONIUMCHLORIDE

centrations were similar to those of the nonam-monium chloride group (Tables II and IV) andagreed closely with the values reported by Steinand Moore for peripheral venous plasma (6).During mild acidosis, however, the arterial plasmaglutamine concentration averaged 40 ± 5 lxmolesper 100 ml, a value significantly lower (p < 0.01)than the 65 + 7.5 Mmoles per 100 ml obtained inthe nonammonium chloride group. A significantdifference (p < 0.05) was also noted in the netrenal extraction of glutamine from arterial plasma(Figure 1). During mild acidosis, glutamine ex-traction averaged 102 + 53 umoles per minute(Table IV), in contrast to 47 ± 14 jumoles perminute (Table II) in those subjects who did notreceive ammonium chloride. As before, a simi-lar percentage of the total renal release of am-monia could be accounted for by the renal ex-traction of glutamine amide nitrogen (74 per cent).As in the subjects who did not receive ammonium

i i

EDWARDE. OWENAND ROSCOER. ROBINSON

TABLE V

Changes of net renal uptake or release of ammonia and amino acid nitrogen associatedwith the prolonged administration of ammonium chloride

Without Afterammonium ammonium

Determination chloride chloride

mg/min mg/min

Average renal release of ammonia Renal venous release 0.52 0.59nitrogen Urine excretion 0.53 1.36

Total renal release 1.05 1.95Average net renal uptake of total

amino acid nitrogen* 0.53 2.86Average net renal uptake of Glutamine total nitrogen 1.32 2.88

glutamine nitrogen Glutamine amide nitrogen 0.66 1.44Average net renal release of amino acid nitrogen

(glutamine total nitrogen excluded) 0.79 0.02

* The total nitrogen content of glutamine was used in this caluclation.

chloride, however, the net uptake of glutamineamide nitrogen could account for all of the am-monia excreted in urine.

Although after ammonium chloride the net re-nal uptake of proline, glycine, and valine aver-aged 12 4, 6 5, and 12 17 /Lmoles per min-ute, respectively, these somewhat higher extrac-tion values were not significantly different fromthose of the nonacidotic subjects (Figure 1). Incontrast, however, ammonium chloride adminis-tration was associated with a highly significant(p < 0.01) directional change in the net balanceof glutamic acid across the renal circulation (Fig-ure 1). An average net release (6 + 9 /umolesper minute) of glutamic acid was observed in thenonammonium chloride group (Table II). Dur-ing acidosis, however, glutamic acid was extracted(net uptake) consistently (12 3 Mmoles perminute) from arterial plasma by the kidney.Average figures for the net uptake or release ofthe remaining amino acids were essentially thesame in both experimental groups (Figure 1).

The average net balances of total amino acidand ammonia nitrogen across the renal circula-tion from both experimental groups have beencompared in Table V. After ammonium chloride,the average net uptake of total amino acid ni-trogen (2.86 + 1.38 mg per minute) was signifi-cantly higher (p < 0.01) than that observed inthe absence of drug administration (0.53 + 0.59mg per minute), and was more than sufficient(147 per cent) to account for the total renal re-lease of ammonia nitrogen (1.95 + 0.60 mg perminute). As shown in Figure 1, higher extrac-

tion values of the total nitrogen content of gluta-mine and glutamic acid were largely responsiblefor this increased uptake of total amino acidnitrogen.

DISCUSSION

Previous measurements in nonacidotic dogshave suggested that arteriorenal venous differ-ences of total free amino acid nitrogen are sosmall and variable that no conclusion may beformed as to the presence of either net renal up-take or release (19). The present observationsdemonstrate that such differences are also smallacross the human kidney (Table V). However,the consistent finding of values indicative of netrenal extraction suggests that these small dif-ferences may well be significant. Regardless ofthe validity of small arteriorenal venous differ-ences of total free amino acid nitrogen, the pres-ent results indicate clearly that several free aminoacids exhibit definite patterns of either net renaluptake or release that are not apparent from de-terminations of total free amino acid nitrogenalone. Characteristically, the healthy human kid-ney extracts glutamine, proline, glycine, and va-line from arterial plasma, whereas serine, glu-tamic acid, cystine, leucine, and arginine are usu-ally added to renal venous blood. Glutamine isthe only free amino acid extracted consistently bythe kidney from arterial plasma, and serine is theonly amino acid regularly released into renal ve-nous blood.

The precise metabolic activities of the kidneythat are responsible for either the net uptake or

272

RENIAL EXTRACTIONOF AMINO ACIDS DURING AMMONIUMCHLORIDE ADMINISTRATION 273

release of certain amino acids are not identifiedby the present experiments. Nevertheless, quanti-tative differences between the net extraction or re-lease of individual amino acids during conditionsof normal acid-base balance and mild metabolicacidosis do permit certain statements to be maderegarding the relationship of renal amino acidmetabolism to ammonia production. It has beenwidely accepted for some time that the amide ni-trogen of glutamine is the major precursor ofurine ammonia. In addition, previous loadingexperiments have indicated that glycine, alanine,leucine, histidine, and asparagine may also serveas presursor materials for the renal production ofammonia (2-4). Since glutamine and glycinewere the only members of this group that ex-hibited net extraction by the kidney, the presentresults can definitely implicate only two of thesesix amino acids as possible ammonia precursorsat normal plasma amino acid levels. Net aspara-gine utilization, however, has not been excludedwith certainty since the analytic technique used inthis study prohibited its accurate measurement(6).

Under normal conditions of acid-base balance,the net renal uptake of glutamine amide nitrogenwas sufficient to account for 63 per cent of thetotal amount of ammonia nitrogen released by thekidneys. The net extraction of glycine nitrogenwas adequate to provide an additional 6 per centof the total bidirectional release of ammonia. Thenet removal rates of other amino acids implicatedas ammonia precursors were insufficient to ac-count for the remaining total release of ammonia.

The remaining 30 per cent of the ammoniaadded to urine, or renal venous blood, or both,may be accounted for by consideration of the pos-sible fates of the glutamic acid arising from theenzymatic deamidation of glutamine by glutami-nase. This process yields one mole of both glu-tamic acid and ammonia from each mole of avail-able glutamine. If the a-amino nitrogen of theglutamic acid produced by this reaction is not usedin either ammonia synthesis or other metabolicreactions, a stoichiometric relationship betweenglutamine uptake and glutamic acid release shouldbe found. That such is not the case is illustratedby the finding that only small amounts of freeglutamic acid are actually released into eitherurine or renal venous blood. Although the present

studies do not provide a definitive explanation ofits fate, several possibilities must be considered.First, glutamic acid may contribute directly to re-nal ammonia production via its direct deaminationto a-ketoglutarate and ammonia by glutamic de-hydrogenase. Second, it may participate in vari-ous transamination reactions leading to the pro-duction of other amino acids which, in turn, areeither released unchanged into renal venous blood,or deaminated directly to form ammonia and theappropriate keto-acid. A metabolic pathway ofthis type provides one possible explanation for thenet venous release of such amino acids as alanineand serine. Third, it may be released into eitherblood or urine in a conjugated form undetectedby the chromatographic technique used in thisstudy. Fourth, it may be used directly in thesynthesis of various peptides or proteins by kidneytissue. Although the results of these experimentsare compatible with a proposal that both thea-amino and amide nitrogens of glutamine mayserve as potential sources of renal ammonia, it isfelt that complete utilization of glutamic acid forthis purpose is unlikely for two principal reasons.First, the observed bidirectional release of am-monia is significantly less than would be expectedif the a-amino and amide nitrogens of glutaminewere both used completely for its synthesis. Sec-ond, previous balance studies during the adminis-tration of glutamic acid have suggested that thisamino acid does not contribute significantly torenal ammonia production (2, 3). Nevertheless,the possibility does exist that at least a portion ofthe glutamic acid arising from glutamine deamida-tion is used in the formation of ammonia. It isobvious that further experiments are needed toclarify the exact role of glutamic acid in the renalsynthesis of ammonia.

Although the possible transamination reactionsof the glutamic acid derived from glutamine pro-vide a potential source of the amino acids added torenal venous blood, another possibility should bementioned. If current estimates of the normalprotein content of glomerular filtrate are valid,the intratubular digestion of reabsorbed proteincould also contribute to the observed patterns ofnet amino acid release. If so, however, only asmall amount of reabsorbed protein must be de-graded completely in view of the small quantityof nitrogen actually added to renal venous blood

EDWARDE. OWENAND ROSCOER. ROBINSON

as free amino acids. In addition, the dissimilaritybetween the amino acid composition of plasmaproteins and the character of the free amino acidsreleased into renal venous blood suggests that thedegradation of reabsorbed protein by proteolyticenzyme systems- is not solely responsible for theirrelease.*The twofold adaptive rise of total renal am-

monia production after ammonium chloride ad-ministration is in agreement with many previousobservations (20, 21). Since renal venous am-monia release did not rise appreciably, this re-sponse was almost solely related to an increasedurine excretion of ammonia. This observationraises the important question of whether or notthe factors controlling renal venous ammonia re-lease also participate in the adaptive response toammonium chloride administration. Additionalstudies are urgently required to evaluate this pos-sibility more fully.

After prolonged ammonium chloride adminis-tration, glutamine, valine, and glutamic acid wereextracted by the kidney in increased amounts.Of these, only the increased net removal valuesof glutamine and glutamic acid were statisticallysignificant, and only glutamine has been shownby previous loading experiments to be associatedwith an increased excretion of urine ammonia.Other amino acids such as glycine that have beensuggested as renal ammonia precursors were notextracted in increased amounts. As under normalconditions of acid-base balance, most (74 per cent)of the total bidirectional release of ammonia couldbe explained by the increased extraction of glu-tamine amide nitrogen. Again, the data failed toaccount for approximately 25 per cent of the totalammonia released, unless the assumption is madethat some, but not all, of the glutamic acid arisingfrom glutamine deamidation is used in renal am-monia synthesis. The lower plasma glutamineconcentration observed after ammonium chloridein association with an elevated renal extraction ofglutamine has not been described previously. Itsoccurrence suggests, but does not prove, that theincreased extraction of glutamine by the kidneyis not counterbalanced by a compensatory rise inits availability to plasma.

Recently, the characteristic patterns of net re-nal uptake or release of free amino acids have beendescribed in acidotic dogs by Shalhoub and asso-

ciates (22). In most respects, their results ap-pear to agree qualitatively with those observedduring ammonium chloride administration to hu-man subjects. They differ mainly in that mildmetabolic acidosis in the human subject is as-sociated with net renal extraction of glutamic acidrather than net release into renal venous plasma.The appearance of net glutamic acid uptake bythe human kidney after ammonium chloride ad-ministration is difficult to relate to changes of re-nal ammonia metabolism unless, as discussed pre-viously, one assumes either its direct deaminationto ammonia and a-ketoglutarate, or its transami-nation to other amino acids that may then serveas direct precursors of renal ammonia nitrogen.That such may be the case is perhaps implied bythe fact that glutamine and glutamic acid are theonly amino acids whose renal extraction was sig-nificantly responsive to an induced change of acid-base balance.

Although these experiments provide informa-tion that must be explained by any complete andfinal description of the mechanisms of renal am-monia metabolism and excretion, it is apparentthat the exact sources of urinary ammonia nitrogencannot be ascertained by the simultaneous meas-urement of urinary amino acid excretion and ar-teriorenal venous amino acid differences. In thisregard, it should be emphasized that net balancedata reflect only the algebraic summation of therates of total organ influx and efflux of a givensubstance. After ammonium chloride adminis-tration, it is conceivable that an amino acid mightbe extracted in amounts sufficient for ammoniasynthesis without an apparent change of net ex-traction if its increased total influx is, for whateverreason, also accompanied by an appropriate changeof total organ efflux. Nevertheless, if the factorsresponsible for the total efflux of amino acids intorenal venous blood remain constant after am-monium chloride administration, a contribution ofany amino acid to increased ammonia productionshould be reflected by either diminished net re-lease or increased net uptake. In the present ex-periments, the -only significant changes of thistype were those exhibited by glutamine and glu-tamic acid.

SUMMARY

1. The net renal uptake or release of 19 freeamino acids and the total bidirectional release of

274

RENAL EXTRACTION OF AMINO ACIDS DURINGAMMONIUMCHLORIDE ADMINISTRATION 275

ammionia (urinary excretion plus renal venousrelease) were measured in 6 healthy subjects un-der conditions of normal acid-base balance. Typi-cally, serine, glutamic acid, alanine, cystine, andarginine were added to renal venous blood,whereas glutamine, proline, valine, and glycineusually exhibited net uptake. Only glutamine wasextracted consistently by all subjects, and serinealone was added regularly to renal venous blood.The quantitative net release of glutamic acid wasmuch less than the theoretical amount expectedfrom the complete deamidation of extracted glu-tamine.

2. The net extraction of glutamine amide ni-trogen accounted for only 63 per cent of the totalbidirectional release of ammonia nitrogen. Otheramino acids that have been implicated as renalammonia precursors were not extracted in amountssufficient to account for the remaining total am-monia release. If, however, the glutamic acidarising from glutamine deamidation is partiallyutilized for ammonia synthesis, then sufficient ni-trogen is available from extracted amino acids toaccount for the observed total ammonia release.

3. Six similar experiments were performedafter prolonged administration of ammonium chlo-ride in an attempt to relate changes of renal aminoacid extraction to the adaptive rise of renal am-monia production. With the exception of gluta-mine and glutamic acid, values for the net uptakeor release of individual free amino acids re-sembled those of the subjects who did not re-ceive ammonium chloride. Net glutamine uptakewas significantly higher, and its amide nitrogenwas extracted in amounts sufficient to account for74 per cent of the increased total renal release ofammonia. The increased extraction of glutamicacid from plasma was small and insufficient toaccount for the remaining total release of am-monia. Again, however, total renal ammonia re-lease could be satisfactorily explained by assumingthat both nitrogens of extracted glutamine arepotentially available to ammonia synthesis.

4. These data are compatible with the thesisthat glutamine serves as the major precursor ofrenal ammonia under conditions of both normalacid-base balance and mild ammonium chlorideacidosis, and that the glutamic acid arising fromglutamine deamidation may also serve as a po-tential source of renal ammonia. They also imply

that other amino acids do not contribute appreci-ably to renal ammonia production at normal plasmaamino acid levels.

ACKNOWLEDGMENT

The authors express their indebtedness to Dr. RonaldC. Greene, Department of Biochemistry, Duke Uni-versity Medical Center, for aid and advice in the per-formance of the amino acid analyses.

REFERENCES

1. Van Slyke, D. D., Phillips, R. A., Hamilton, P. B.,Archibald, R. M., Futcher, P. H., and Hiller, A.Glutamine as source material of urinary ammonia.J. biol. Chem. 1943, 150, 481.

2. Lotspeich, W. D., and Pitts, R. F. The role ofamino acids in the renal tubular secretion of am-monia. J. biol. Chem. 1947, 168, 611.

3. Kamin, H., and Handler, P. The metabolism of par-enterally administered amino acids: III. Am-monia formation. J. biol. Chem. 1951, 193, 873.

4. Madison, L. L., and Seldin, D. W. Ammonia excre-tion and renal enzymatic adaptation in human sub-jects, as disclosed by administration of precursoramino acids. J. clin. Invest. 1958, 37, 1615.

5. Spackman, D. H., Stein, W. H., and Moore, S.Automatic recording apparatus for use in thechromatography of amino acids. Anal. Chem.1958, 30, 1190.

6. Stein, W. H., and Moore, S. The free amino acidsof human blood plasma. J. biol. Chem. 1954, 211,915.

7. Meister, A., Levintow, L., Greenfield, R. E., andAbendschein, P. A. Hydrolysis and transfer re-actions catalyzed by w-amidase preparations. J.biol. Chem. 1955, 215, 441.

8. Segal, S., and Wyngaarden, J. B. Plasma glutamineand oxypurine content in patients with gout.Proc. Soc. exp. Biol. (N. Y.) 1955, 88, 342.

9. Stein, W. H. A chromatographic investigation ofthe amino acid constituents of normal urine. J.biol. Chem. 1953, 201, 45.

10. Brown, R. H., Duda, G. D., Korkes, S., and Handler,P. A colorimetric micromethod for determinationof ammonia; the ammonia content of rat tissuesand human plasma. Arch. Biochem. 1957, 66, 301.

11. Schreiner, G. E. Determination of inulin by meansof resorcinol. Proc. Soc. exp. Biol. (N. Y.)1950, 74, 117.

12. Selkurt, E. E. Measurement of renal blood flow.Meth. med. Res., 1949, 1, 191.

13. Rosenthal, T. B. The effect of temperature on thepH of blood and plasma in vitro. J. biol. Chem.1948, 173, 25.

14. Van Slyke, D. D., and Neil, J. M. The determinationof gases in blood and other solutions by vacuum

EDWARDE. OWENAND ROSCOER. ROBINSON

extraction and manometric measurements. J. biol.Chem. 1924, 61, 523.

15. Van Slyke, D. D., and Sendroy, J., Jr. Studies ofgas and electrolyte equilibria in blood. XV. Linecharts for graphic calculations by the Henderson-Hasselbalch equation, and for calculating plasmacarbon dioxide content from whole blood content.J. biol. Chem. 1928, 79, 781.

16. Hickam, J. B., and Frayser, R. Spectrophotometricdetermination of blood oxygen. J. biol. Chem.1949, 180, 457.

17. Wintrobe, M. M. Clinical Hematology, 5th ed.Philadelphia, Lea & Febiger, 1961.

18. Bradley, S. E., Bradley, G. P., Tyson, C. J., Curry,J. J., and Blake, W. D. Renal function in renaldiseases. Amer. J. Med. 1950, 8, 766.

19. Lathem, W., and Benjamin, A. D. Observations onrenal protein metabolism. Renal venous aminonitrogen content before and following hemoglobininfusions in dogs. J. exp. Med. 1956, 104, 211.

20. Sartorius, 0. W., Roemmelt, J. C., and Pitts, R. F.The renal regulation of acid-base balance in man.IV. The nature of the renal compensations in am-monium chloride acidosis. J. clin. Invest. 1949,28, 423.

21. Orloff, J., and Berliner, R. W. The mechanismof the excretion of ammonia in the dog. J. clin.Invest. 1956, 35, 223.

22. Shalhoub, R. J., Canessa-Fisher, M., Webber, W. A.,Glabman, S., and Pitts, R. F. Sources of urinaryammonia. Fed. Proc. 1961, 20, 415.

276