Journal of Clinical InvestigationVol. 41, No. 3, 1962

THE EFFECTS OF PHOSPHOPROTEINSON ACID BALANCEINNORMALSUBECTS*

By EDWARDJ. LENNONj JACOB LEMANN, JR. AND ARNOLDS. RELMANWITHTHE TECHNICAL ASSISTANCE OF HELEN P. CONNORS

(From the Evans Memorial Department of Clinical Research, Massachusetts Memorial Hos-pitals, and the Department of Medicine, Boston University School of Medicine,

Boston, Mass.)

(Submitted for publication September 11, 1961; accepted November 6, 1961)

Identification and measurement of the dietarysources of fixed endogenous acid have been ham-pered by the chemical complexity of naturallyoccurring foods. Recent studies from this labora-tory (1) have demonstrated the feasibility ofidentifying and quantitating the sources of fixedacid production when a chemically defined, syn-thetic diet is utilized and body composition re-mains approximately constant. With the specialdiet used in those studies, the processes resultingin net endogenous acid production in the steadystate were identified as: a) the release of acidduring the oxidation of neutral sulfur to inorganicsulfate, and b) the net production of organic acidsfrom neutral foodstuffs. In addition, c) the datasuggested that a third source of acid was relatedto the phosphorus content of the dietary protein,and it was proposed that acid was released duringthe oxidation of the cations neutralizing phosphateesters, or during the hydrolysis of phosphate di-esters in the protein, or both. However, theamount of organic phosphorus in the basic dietwas relatively small and analysis of its contribu-tions to acid production was therefore limited.

Accordingly, in the present experiments thepostulated role of phosphorus was tested morecritically by the administration of extra loads ofthe purified soy phosphoprotein used in the basicdiet. The effects were compared with those offeeding equivalent loads of a more complex natu-ral protein (contained in cooked ground beef-steak). The data to be presented support ourprevious conclusions that a predictable amount of

* Supported by U.S. Public Health Service GrantA-3140. Submitted in honor of Chester S. Keefer, M.D.,and the Golden Anniversary of the Evans Memorial De-partment of Clinical Research, Boston, Mass.

f During tenure as a Trainee of the National Instituteof Arthritis and Metabolic Diseases. Present address:Milwaukee County Hospital, Milwaukee 13, Wis.

acid production is associated with the metabolismof phosphorus in the purified protein used in thebasic diet. On the other hand, the net acidifyingeffect of the meat could be completely accountedfor by the measured increase in the oxidation ofsulfur and the production of organic acids withoutconsidering the phosphorus content. This sug-gests either that the phosphate residues in themeat were in a form that does not release acid orthat the acid released was neutralized by thesimultaneous production of an equivalent quan-tity of alkali that might be expected to have de-rived from the meat. Taken together, the resultsof these studies are consistent with the view thatthe amount of acid produced in association withthe metabolism of organic phosphorus dependsupon the chemical form of the phosphate residues.Accurate, independent assessment of endogenousacid production is valid, and measurement of thenet external acid balance possible, only whenpurified, well characterized food sources are used.

METHODS

Eight balance studies of 12 to 17 days' duration werecarried out in 5 healthy, ambulatory, male medical stu-dents. Each study began with a 2-day period of adapta-tion to the diet. There followed a control period of 3 to7 days, an experimental period lasting 3 days, and a re-covery period of 5 to 7 days. During the experimentalperiod in four studies the subjects were fed an additionalload of soy phosphoprotein (1.5 g per kg per day). Inthe other four studies the subjects were fed an equivalentamount of nitrogen in the form of ground beefsteak (6 gof meat per kg per day). The extra protein nitrogenload in both groups was approximately 0.2 g per kg perday.

The basic liquid formula diet was identical with thatpreviously described (1). The purified soy phosphopro-tein used as the sole source of nitrogen (sometimes here-inafter referred to by the manufacturer's name, "J-pro-tein") was virtually free of inorganic cations and anions

637

E. J. LENNON, J. LEMANN, JR. AND A. S. RELMAN

C4 t C 0 G o o co co

O e) el; CC) CCi eC C C - ~C 0 0 0Uj t j iW0 0 0 00m %C 0 lC

m0oo0o0oO 0 co0 00

- ~ ~ ~~I

z;~~ ~~~ -;o4o ~r~ (I Co5esoOOr4 -. . * -.. .4 . ..4 .

-t 00' - oo 0 00C CON'0 0000 C 00'-

(xo 0 0 e rC)0.C o o C IOn

c> o ON 01 m 4C0,orN >w

0% 0 0 0%400'.t. . .. . . .

zh _ tr + Xo 0 0oOin 00 0 C 00 °

?.0) h + a s va O

0) o ro-CC)C~ECeC%e\0- V O)%+ b -.0) _ E e S N C t

O0 0000tl)1C4'0%

o to _ CS o o o oo o o o o co int- '0 C) -t W) Cl) Ctl In)

oC CC) so U

:C '

% 0

0% ~1

C

Cl)

C' C'

-

-! -0. .o

00

C'e- C-

C) C4 Cr

00 co' 00

CwC)r

: 0'" 0'- '- WI

C4 m in 0- C0 W)

O Q

638

.9,aN

w

CO

z

u

co

:d

Iz

1-0

,ssO. C)0e

zz+7-0

0

00

o

E

o.

Co

CO

CO

E

0%

co£

-

7, C

zcd_

'O

CO

_~

cdV

co q.,+co

.r

o

.w

ooo

O

._

0d

0

C

.0 . o

,.0 a:

t I+ a

* 4.0=

0A.

1.1

ItY i, -.v.,

-law 4-1o 3~

*1.

j b

.1

I

-1 - - I -I - - - - -

'.11

j,

'j,

PHOSPHOPROTEINANDACID BALANCE

oo0 - %

0 00 C14 0%m 00 m e-4

U)% Ul) LO

U) 00 0 rq

oo t- No

o \0 \ U)

0

*0

00C,

0% 4 V-i C-

U) C4 C14 to

ei '' i-

O LO to

0 1 0

U)4 Ce4-

0e m

eti

eWi CNIu

U)me4

zss-

C^ V VX 00t

- W U) W4

-t 00 a 0e

c t-0i o oo

\)U)

- o1 't m

\000 C 00 xu rU0

oef VEoX U)V>4

on)C 00 00O

_0U U)U U) U)

C\ ( C

C 4 ) U)4 C

oo t; t It if ot if;

U)U)U+ )

Ul)0

V-

d4

-

00

I- Ul)

0 oN4 el;

4 _.

00

e

00

r-

I-4 NO 0)

to to toU) U)

\0 \0o

4

-d

_q \

cN

t- t- t- t-

en o 4 Ul) -

0 00 V% V0 U)

U) U) U) U)

00 O0 0-_ N m1- _- _.

639

?3

U

Cd

u

z

b-0

bOwutbo .

A. 0

0(

zE4

.6

z

Ca

(14

0

U

ci

Ez :.

-o

V~ -

~-00

o 0%

0

0'

o~ X

0

E -

u

; _%

CU

Cd

>~ 0

0 cu*

8 U

co

cc

C 0

z E

U)

+j

0.co

E cu

V- (12

-,oO a;

C 0_ i*r D-4-

E. J. LENNON, J. LEMANN, JR. AND A. S. RELMAN

other than phosphate.' Because alkali was required todisperse this protein, and because calcium and magnesiumwere not otherwise provided in the basic diet, an equi-molar mixture of Ca(OH), and Mg(OH)2 was added.This provided milliequivalents of alkali equal to 1.8times the millimoles of phosphorus in the diet protein.As discussed in a previous study (1), this quantity of al-kali appeared approximately to neutralize that fractionof acid production from the basic diet not measured asurinary sulfate plus organic acids (and therefore pre-sumed to be associated with the metabolism of the dietaryphosphorus). Although the basic diet, with this quan-tity of alkali, was maintained constant throughout eachbalance study, no additional alkali was added to the extrasoy protein or meat fed during the loading periods.

In two of the studies with meat the basic diet was so-dium-free; in the remaining six balances the basic dietcontained 1.5 mmoles of NaCl per kg per day; 1.0mmoles per kg per day of KCl was added to all diets.

Details of the balance technique used in this laboratoryand references to most of the analytical methods have al-ready been published (1, 2). Titratable acid and organicacids in the urine were measured at 370 C with a Radi-ometer automatic titrator. Blood pH was determinedwith the Astrup capillary glass electrode at 37° C. Ina few balances, urinary uric acid was determined by aphosphotungstic acid method.

RESULTS

Complete analytical data from one subjectgiven a soy protein load are shown in Table I;similar data for a subject loaded with beefsteakare presented in Table II. Figures 1 and 2present some pertinent analytical and derived datafrom two other subjects loaded with soy phos-phoprotein and meat, respectively.

A. Control observations. During control peri-ods body weight and all of the measured con-stituents of plasma and urine remained relativelyconstant. Stools were small and infrequent andhad a low content of electrolytes. With but fewexceptions there were no significant positive ornegative balances of any of these constituents.

In confirmation of previous observations in thesteady state (1), it was noted that the net ex-cretion of acid in the urine (titratable acid +ammonium - bicarbonate) was closely matchedby the simultaneous endogenous production ofacid as measured by the sum of the urine in-organic sulfate plus organic acids.2

1 J-protein-100, kindly made available by the J. R. ShortMilling Company, Chicago, Ill.

2When no inorganic alkali was added to the basic dietin previous experiments, acid excretion exceeded sulfate

In the lower half of Figures 1 and 2 the netexternal acid balance is plotted in the mannerconventionally used for inorganic constituents.Acid production (inorganic sulfate + organicacid) is plotted downward from the zero axis.Net acid excretion is plotted upward from theheavy line indicating acid production. Thus,clear areas below the zero axis indicate positiveacid balance, while extensions of the shaded areaabove the zero line define negative acid balance.Although there were minor daily fluctuations, itis apparent from these figures and from the datain Tables I and II that the subjects were in ap-proximately zero net acid balance during the con-trol periods. All eight balances considered, themean per cent difference between the cumulativeacid production and acid excretion [(production- excretion)/production] was + 1.8 + 6.4 (SD)per cent.

B. Effects of soy phosphoraprotein loading.As shown in Table I and Figure 1, administra-tion of extra soy protein was followed by an in-crease in the excretion of ammonium and titrat-able acid. Net acid excretion reached a maximumvalue almost double that of the control on thethird and final day of loading. This graduallydiminished during the recovery period, returningto baseline values by the end of the study. ThepH of 24-hour urine specimens remained essen-tially unchanged. Serum CO2 content alsoshowed no significant change. Urinary inorganicsulfate and organic acids increased, but thechanges were smaller and less sustained than thosein acid excretion. In the subject of Table I, asin the others fed soy protein, phosphorus excre-tion increased only slightly, if at all. Stool cal-cium, phosphorus, and potassium excretion in-creased slightly during the recovery period.There were no consistent or significant changesin the urinary excretion or the net balance ofsodium, potassium, or chloride.

In all four subjects loaded with J-protein thepattern of response was quite similar in all re-spects to that described above. Despite approxi-

plus organic acids by an amount in milliequivalents ap-proximately equal to 1.8 times the millimoles of proteinphosphorus. In the present studies, in which an equiva-lent amount of alkali was added as calcium and mag-nesium hydroxide, it would be expected that this dis-crepancy would be eliminated.

A40

PHOSPHOPROTEINANDACID BALANCE

*-@

SERUMCONTENmM/L

URINEpH

ACIDBALANCEmEq/Doy

33T 31

29

5.55.0

- 20

0

+20E,+40

+60

+80

+100

041R.-I I_/

*1[I 2 31 1 169FI 9J0III2I131

333129

5.5

5.0

20-

20+

40+

60+

80+

100+

DAYS

FIG. 1. NET EXTERNAL ACID BALANCE, SERUM CO,CONTENT, AND URINE PH IN A SUBJECT GIVEN A SUPPLE-

MENT OF J-PROTEIN FOR 3 DAYS. In the constructionof the acid balance, total acid production is plotted down-ward from the zero line, and net acid excretion in theurine is plotted as a hatched column back up from theheavy line indicating production. Production is calcu-lated (except during the loading period) as urinary sul-fate + organic acids; excretion is urine ammonium +titratable acid - bicarbonate. During the loading periodacid derived from the fed phosphorus is included in thecalculation of total production. This increment is de-fined by the distance between the heavy line indicatingtotal production and the dotted line parallel to it. Seetext for further details.

mately twofold increase in acid excretion, in no

instance was there a detectable reduction in serum

CO2 content. In each study the increment in theexcretion of organic acids plus urinary sulfate was

significantly less than that in acid excretion. Inthe upper half of Table III the total incre-ment in acid excretion during and after the phos-phoprotein load is compared with the total in-crease in the excretion of inorganic sulfate plusorganic acids. These quantities were calculatedby substracting the mean control values in eachcase from the daily observed values, until bothproduction and excretion had returned to theiroriginal baselines. It is evident from Table IIIthat in each of the four subjects fed soy proteinthe total increase in acid excretion considerablyexceeded the total increase in sulfate plus organicacids. The increase in acid excretion ranged

from 96 to 231 mEq, while the increase in theexcretion of sulfate plus organic acids was only37 to 144 mEq. The mean per cent differencewas 121 + 44 (SD) per cent.

C. Effects of loading with beefsteak. As shownin Table II and Figure 2, feeding of the beef-steak was also followed by an increase in theexcretion of ammonium and titratable acid. Inthe study summarized in the table the total ex-cretion of acid reached a peak of 97 mEq perday on the third and final day of loading, repre-senting an increase of more than 50 per cent abovecontrol. Acid excretion gradually returned tocontrol levels by the second or third day of re-covery. As with phosphoprotein loading, therewas no significant change in serum CO2 contentor urine pH. The urinary excretion of inorganicsulfate and organic acids rose promptly with thefeeding of meat and dropped back to control levelson the first day of recovery. In contrast to theresults with the soy protein, the total incrementin the excretion of inorganic sulfate and or-ganic acids after meat feeding was virtually equalto the total increment in acid excretion. Urinaryphosphorus excretion approximately doubledduring the period of loading, and there was alsoa slight increase in potassium excretion, reflect-ing the increased intake. Sodium and chlorideexcretion were not discernibly affected. Becauseof small and infrequent stools no attempt wasmade in this study to divide the scanty fecal elec-trolyte excretion into separate periods.

TABLE III

Comparison of total increase in acid excretion with total in-crease in excretion of inorganic sulfate and organic acids

for subjects fed soy protein and meat loads

A BTotal Total Differenceincrease increas Difernc

in acid in S04+ (B-Axi0Subject excretion org. acids B

mEq mEq %Soy protein

R.R. 96 37 -160P.R. 150 69 -117F.G. 204 83 -146A.S. 231 144 - 60

Mean -121 444

BeefsteakF.G. 152 137 -11P.R. 107 113 + 5R.R. 127 135 + 6G.S. 106 129 +18

Mean +4.5 412

641

I

E. J. LENNON, J. LEMANN, JR. AND A. S. RELMAN

DAYS

FIG. 2. NET EXTERNAL ACID BALANCE, SERUM CO2CONTENT, AND URINE PH IN A SUBJECT GIVEN A SUPPLE-

MENTOF MEAT. The acid balance is constructed in thesame manner as in Figure 1, except that urine inorganicsulfate and organic acids are taken to be the sole sources

of acid production throughout the study. See text forfurther details.

In the other three subjects the pattern of re-

sponse was almost identical. In each instanceacid excretion increased by 50 to 100 per centwithout detectable change in serum CO2 contentor urine pH. There was a prompt increase inthe excretion of sulfate and organic acids, limitedto a period of loading, which closely matched thetotal increment in acid excretion. Restriction ofsodium in the diet of two subjects did not affectthis pattern, and the feeding of meat producedno significant change in the low control rates ofsodium excretion. In two balances in which itwas possible to divide stools into periods, therewas a slight increase in fecal excretion of calcium,potassium, and phosphate during the loading or

recovery periods.In the lower half of Table III the total incre-

ments in acid excretion during and after meat

feeding are compared with the total increase inthe excretion of inorganic sulfate plus organicacids. These were calculated in the manner

already described. It is apparent from the table,in contrast to the soy protein experiments shown

in the upper half, that there was no significant orconsistent difference between these quantities forthe beefsteak experiments. The mean per centdifference between acid excretion and the sum ofsulfate and organic acids was only 4.5 + 12 (SDper cent.'

DISCUSSION

The most striking finding in the four studieswith the purified soy phosphoprotein was the largediscrepancy between the total increase in acid ex-cretion and the total increase in urinary inorganicsulfate plus organic acids (Table III). Previousexperience (1), as well as the observations in thecontrol periods in all of the present balances, indi-cates that with this balance technique discrep-ancies of this magnitude between acid excretionand the total measured production of fixed en-dogenous acid are not to be expected. The mostreasonable conclusion to be drawn is that the ex-tra phosphoprotein provided a third source of en-dogenous acid production in addition to the oxi-dation of sulfur to sulfate and the production oforganic acids. If this third source of acid is as-sumed to be related to the phosphorus content ofthe phosphoprotein, the discrepancies noted abovecan be satisfactorily explained. The total load ofextra organic phosphorus ranged from 43 to 53mmoles, whereas the absolute discrepancies be-tween acid excretion and the measured sources ofproduction (sulfate + organic acids) ranged from59 to 121 mEq. If the discrepancy in each caseis divided by the extra phosphorus fed, the ap-parent acid derived per mmole of phosphorusranges from 1.4 to 2.3 mEq (mean, 1.9 + 0.3).This calculation therefore is consistent with ourprevious conclusions that: a) endogenous acidproduction is associated with the metabolism ofphosphorus in the purified soy phosphoproteinused in these studies, and b) the acid so releasedappears to approximate the theoretical amount tobe expected from the hydrolysis of the organicphosphate residues in the protein and from themetabolism of the combustible cations associatedwith these residues. A more detailed discussionof this subject is given elsewhere (1, 3), but in thefollowing paragraphs a brief attempt will be madeto summarize the relevant considerations dealingwith the role of organic phosphate in endogenousacid formation.

642

PHOSPHOPROTEINAND ACID BALANCE

Although it has been generally assumed that themetabolism of all organic phosphorus in the dietgives rise to acid, it has been pointed out byChristensen (4) and by Hunt (5) that the phos-phorus present in natural foods as inorganic saltsof orthophosphate monoesters should not be ex-pected to yield acid. This is demonstrated by thefollowing equation in which the inorganic cationneutralizing the orthophosphate residue in a pro-tein is assumed to be all K+:

C00- NH3+\/ I0

(C) -0-P-0- + K+ + H20

NH2 0- + 02+ K+

0

H2N-C-NH2 + C02 + HP04- + 2 K+ [1]The problem is more complex than this, how-

ever, because Perlmann (6) has shown that na-tural phosphoproteins probably also contain diesterand pyrophosphate structures as well as ortho-phosphate monoesters. Furthermore, diesterstructures are, of course, known to occur in phos-pholipids and nucleic acids and other compoundsabundant in natural foods. In their natural state,neutralized at least in part by inorganic cations,these various forms of phosphorus would havediffering effects on net acid production, and hencecalculation of the latter quantity would requirethat the nature of the phosphate linkages be knownin detail. Although such information is not avail-able for the soy protein used in these studies orfor most other phosphoproteins, this difficulty iseliminated by the use of the protein in its iso-electric form, free of all inorganic cations. This isexplained by the following three reactions whichschematically show the hydrolysis and combustionof an isoelectric phosphoprotein containing eachof the three possible types of phosphate residues:

Orthophosphate monoester

C00- NH3+0

(C) -0-P-OH + H20

\ +02°NH3+ 0-

0

H2N-C-TNH2 + CO2 + H20 + HP04- + 2 H+ [2]

Pyrophosphate ester

NH3+/ 0 0-

II(CH) -0-P-O-P-OH +2 H20

0H3 0 + 21 02NH3+0

H2N-C-NH2 + H20 + 2 HP04- + 4 H+ [3]

Orthophosphate diester

C00- NH3+0

A.-I

(CH) -O-P-0- (CH2) + H200- NH3+ +202

0

H2N-C-NH2+ 2 H20+ 2 C02 + HP04- + 2 H+ [4]

It can be seen that, regardless of the form inwhich phosphorus exists in the protein, the theo-retical ratio of acid released to the phosphoruspresent is the same. The actual value of this ra-tio will depend upon the extent to which the sec-ondary hydrogen is dissociated from phosphoricacid, and this in turn depends upon the pH of themedium in which these reactions take place. If thereactions are arbitrarily assumed to occur at pH7.4, 1.8 mEqof acid would be released per mmoleof phosphorus (pK'2 of phosphoric acid = 6.8),but assumption of any other reasonable physio-logic pH would not markedly alter this value.The actual mean ratio (1.9 + 0.3) calculatedfrom the four balances with J-protein is a reason-ably satisfactory confirmation of this prediction.

In Figure 1, which shows the net external acidbalance in one of the studies with J-protein load-ing, the significance of the acid contribution relatedto the soy phosphoprotein load is demonstrated.During the 3-day loading period, the distance be-tween the heavy line indicating acid production andthe dotted line parallel to it defines that part of theproduction derived from the extra phosphorus, as-sumed that this amounts to 1.8 mEqper mmole ofextra phosphorus fed. It is obvious from thisfigure that, if phosphorus had been neglected in theconstruction of the balance and if acid productionhad actually been at the level indicated by thedotted line, the net balance of acid would have re-

mained approximately zero on the first two load-

643

E. J. LENNON, J. LEMANN, JR. AND A. S. RELMAN

ing days and would then have become significantlynegative on the third day. As a result there wouldhave been an unexplained cumulative negative bal-ance. With the phosphorus taken into considera-tion, however, the observed changes in net acidexcretion fit into a physiologically more rationalpattern, and the total acid excretion at the end ofthe study approximately equals the acid production.

Turning now to a consideration of the studieswith meat, it is apparent from the data in TableIII that there is no necessity to assume that acidis derived from the ingestion of extra organicphosphorus. Despite the fact that the amountsof beef fed contained more total phosphorus (55to 100 mmoles) than did the load of soy protein(43 to 53 mmoles), there was no discrepancy be-tween the total increment in measured acid ex-cretion and the total increase in acid productioncalculated simply from the urinary inorganic sul-fate and organic acids. As shown in Figure 2 itis possible to construct a satisfactory acid balancediagram without taking the extra phosphorus loadinto account. The acid retained during the load-ing period was virtually all excreted during therecovery period, and cumulative net balance atthe end of the study was not significantly differentfrom zero.

This difference between the acid productionfrom the purified soy phosphoprotein and thatfrom beefsteak indicates either that no acid is de-rived from the hydrolysis and metabolism of thephosphate residues in the meat and their associ-ated cations, or that any acid arising from thissource is neutralized by the generation of anequivalent amount of alkali during the metabolismof combustible anionic groups in the meat. If, forexample, the phosphorus in the meat were mainlyin the form of completely dissociated orthophos-phate residues, neutralized by inorganic cationssuch as K+, no net acid would be released duringmetabolism, as illustrated in Equation 1. It wouldseem more likely, however, that acidogenic phos-phodiester or pyrophosphate structures do exist inthe meat protein or in the phospholipids of themeat, but their effect is balanced by the alkali de-rived from combustion of carboxyl groups neu-tralized by inorganic cations.

It should be noted that the ingested, rather thanthe excreted, phosphorus was used in the calcu-lation of acid production. As pointed out previ-

ously ( 1 ), the present balance technique assumesthat body composition is relatively constant dur-ing the period of study, and in healthy subjects in-gesting an adequate diet such an assumption seemsreasonable. Theoretically, therefore, total excre-tion of phosphorus should equal intake, and eitherquantity should be suitable for the calculation ofacid production. In the present studies, however,total urinary and fecal phosphorus differed fromtotal intake, so that the apparent final cumulativebalances varied from + 79 to - 93 mmoles, with-out any consistent difference between the studieswith meat and with J-protein. The most likelyexplanation for the variability in apparent balanceseems to be the inaccuracies inherent in the de-termination of fecal phosphorus when stools arescanty and the collection periods relatively short.Under these circumstances, phosphorus intake isclearly the proper basis for the calculation of acidproduction attributable to the metabolism of theJ-protein.

The increased acid production from the J-proteinaverages 56 mEq per 100 g of protein (3.9 mEqper g nitrogen), compared with an increase of 10mEqper 100 g of beef (2.9 mEqper g nitrogen).Approximately two-thirds of the increase in acidproduction after meat feeding was accounted forby an increased urinary excretion of inorganicsulfate, whereas this accounted for only about one-fifth of the increased acid from the soy proteinload. All of the remaining increase in acid pro-duction due to the meat and an additional 17 to 33per cent of the increased acid production due tosoy protein was accounted for by an increase inthe excretion of urinary organic acids. In onestudy of each type, serial measurements of uri-nary uric acid excretion revealed small incrementsduring the loading periods of approximately 2 to3 mEq per day. Increased uric acid productionwould therefore account for no more than one-fifthof the total increase in organic acid. The identityof the remaining major fraction of the incrementin organic acids remains undetermined, but it isclear that this increment could not be predicted onthe basis of the known composition of the meat orsoy protein. Since only the sulfur oxidized to sul-fate generates endogenous acid, it is apparent thatthe acid accounted for by sulfur metabolism is alsonot strictly predictable from diet composition.

These considerations make it apparent that the

644

PHOSPHOPROTEINANDACID BALANCE

net effect of protein metabolism on acid balancemay be expected to vary considerably from one

protein to another. The total amount of net acidreleased will depend upon: 1) the quantity andnature of the phosphate groups, 2) the relativeproportion of combustible cationic and anionicgroups (mainly - NH3+ and - COO, respec-

tively), 3) the quantity of organic sulfur oxidizedto sulfate, and 4) the net effect of the protein inaugmenting organic acid production.

SUMMARYAND CONCLUSIONS

Large supplements of purified soy phospho-protein and of beefsteak were fed to normal sub-jects and the effects on acid balance compared.The soy protein generated 3.9 mEq of acid per g

of nitrogen and the beefsteak 2.9 mEq per g

of nitrogen.The results show that a significant fraction of

the net endogenous acid produced from the metab-olism of the soy protein was predictably relatedto its phosphorus content, but no such source ofacid could be demonstrated from meat. It is sug-

gested that the production of acid associated withthe presence of phosphate groups in a protein willbe determined by both the structural relationships

of the phosphate and the nature of the associatedcations.

The remainder of the acid production from thesoy phosphoprotein and all of the acid arisingfrom the meat was accounted for by increasedoxidation of sulfur and increased endogenous pro-duction of organic acids.

REFERENCES

1. Relman, A. S., Lennon, E. J., and Lemann, J., Jr.Endogenous production of fixed acid and the meas-urement of the net balance of acid in normal sub-jects. J. clin. Invest. 1961, 40, 1621.

2. Lemann, J., Jr., and Relman, A. S. The relation ofsulfur metabolism to acid-base balance and electro-lyte excretion: The effects of DL-methionine innormal man. J. clin. Invest. 1959, 38, 2215.

3. Relman, A. S. Endogenous acid production and themeasurement of the acid balance in health and dis-ease. Honyman Gillespie Lecture, Univ. of Edin-burgh, 1961. To be published.

4. Christensen, H. N. Diagnostic Biochemistry; Quanti-tative Distributions of Body Constituents and TheirPhysiological Interpretation. New York, OxfordUniv. Press, 1959.

5. Hunt, J. N. The influence of dietary sulphur on theurinary output of acid in man. Clin. Sci. 1956, 15,119.

6. Perlmann, G. E. The nature of phosphorus linkagesin phosphoproteins. Advanc. Protein Chem. 1955,10, 1.

645