Journal of Clinical InvestigationVol. 42, No. 12, 1963

CYSTINURIA: METABOLISMOF THE DISULFIDE OF CYSTEINEANDHOMOCYSTEINE*

By GEORGEW. FRIMPTERt

(From the Department of Medicine, The New York Hospital-Cornell University MedicalCenter, New York, N. Y.)

(Submitted for publication June 14, 1963; accepted September 6, 1963)

The isolation and identification of the asym-metrical disulfide of L-cysteine and L-homocysteine,hereafter referred to as "mixed disulfide," werepreviously described (1). At that time the aminoacid had been found only in the urine of patientswith cystinuria.

This paper attempts to explain the origin andsignificance of the mixed disulfide. In approach-ing this work, I first assessed the natural occur-rence of mixed disulfide. To increase the com-pound's sensitivity to detection in a complicatedarea of the chromatogram, it was rendered radio-active by the administration of L-methionine-S35.The radioactive mixed disulfide was isolated fromthe urine of a patient with cystinuria after he wasfed L-methionine-S35, and the specific activitiesof the two sulfur atoms were compared to eachother and to cystine sulfur. The mixed disulfide,synthesized with S35 was administered to a pa-tient with cystinuria, and its fate was studied.Finally, to define the role of the kidney in me-tabolism of the compound, concentrations in renalvenous blood were compared with those in arterialblood.

METHODS

Free amino acids of urine and plasma were measuredby the method of Spackman, Stein, and Moore (2). '5-labeled, mixed disulfide is 100 + 3% recoverable whenadded to urine or to plasma prior to picric acid precipi-tation of proteins (3). The compound is, however, un-stable under the conditions suggested by Stein and Moorefor oxidation of cysteine to cystine for chromatography(4). Accordingly, specimens were analyzed without thisstep, except when examination for the possible occur-rence of homocysteine was desirable. Radioactive mixeddisulfide, when added to plasma, dialyzed from cellophanein a manner consistent with its being free and unboundto plasma proteins (5). Fractions of the column efflu-ent were collected by a stream-splitting technique (6).

* This study was supported by grants H-4148 andFR-47 from the National Institutes of Health, U. S.Public Health Service.

t Senior Research Fellow, NewYork Heart Association.

Radioactivity in fractions of column effluent was de-termined by liquid scintillation counting. Two-ml frac-tions were poured directly into counting vials, and thefraction collection tubes were rinsed into the countingvials with 10 ml of a dioxane phosphor system.1 Thismethod yielded an over-all counting efficiency of 30%o.

For identification of mixed disulfide in urine, fractionswere collected by the stream-splitting technique. Thepooled fractions containing the material were desalted onDowex 50 X 4 (H+) (7) and evaporated to dryness.The specimens were then subj ected to oxidation withperformic acid (8), and the resultant cysteic and homo-cysteic acids were determined by 2-dimensional paperchromatography, with butanol: acetic acid: water (120:30: 50) versus butanol: pyridine: water (1: 1: 1) (9).The Ninhydrin colored spots were eluted with an ethanoliccopper sulfate solution and compared with standards in aspectrophotometer (10).

Fifty Ac of L-methionine-S' with a SA of 17 ,uc permg was administered orally to a cystinuria patient whowas fasting and had just voided urine. This patient hadno evidence of urinary calculi and was known to havenormal renal function, including normal inulin clear-ance. During the 24 hours following administration ofthe isotope, urine was collected fractionally at 1.25 hours,5.25 hours, and 6.0 hours; the urine from 6 to 24 hourswas collected in one vessel. Samples of each specimenwere analyzed for total radioactivity by liquid scintilla-tion counting. Cystine and mixed disulfide were isolatedchromatographically on the 150-cm column, as previouslydescribed, from portions of the 0 to 1.25-, 5.25 to 6-, and6 to 24-hour specimens. These fractions were desaltedand lyophilized. Performic acid oxidation split the S-Sbond (8), and the specimens were dried on the rotaryevaporator. Five-tenths ml of water was added to theflasks, and several different microliter amounts, in dupli-cate, were spotted on Whatman 3 filter paper forchromatography of the cysteic and homocysteic acids.Multiple amounts of standard solutions of these com-pounds were used for comparison. Descending chroma-tography with butanol: acetic acid: water (120: 30: 50)was run for 40 hours. This procedure resulted in thecysteic and homocysteic acids' running 185 and 220 mm,respectively, from the origin. Addition of nonoxidized,standard solutions of cystine and mixed disulfide revealed

1 Naphthalene, 50 g; 2,5-diphenyloxazole, 7 g; 1,4-bis-2- (4-methyl-5-phenyloxazolyl) benzene, 50 mg; and enoughdioxane to make 1 L.

1956

DISULFIDE METABOLISMIN CYSTINURIA

that these were separated from the cysteic and homo-cysteic acids. No unreacted cystine or mixed disulfidewas encountered on the paper chromatograms. Aftercolor development, spots were eluted, and micromolarvalues were determined for cysteic and homocysteic acidsby comparison with standards. The duplicate spots wereeluted with 80% ethanol; radioactivity was determinedby liquid scintillation counting. Recovery studies showedthat the Ninhydrin reaction did not appreciably interferewith counting, and an over-all efficiency of 25%vwasobtained.

For the infusion of the mixed disulfide into the tailvein of a rat, 20 mg was dissolved in 35 ml of 0.6%o sa-line with pH adjusted to 7.4 with sodium bicarbonate.A similar solution without amino acid was infused, andurine specimens were collected. The solution containingthe amino acid was then substituted.

To study metabolism of the compound in a patient withcystinuria, radioactive mixed disulfide was prepared withS35 on the cysteine sulfur. This was done by sodium re-duction of L-cystine-S3 2 in liquid NH3 (11) followed byreaction with L-homocysteine prepared from L-homocys-teine thiolactone hydrochloride 3 (12). The reaction mix-ture was brought to pH 7.5 with NaHCO3 and oxy-genated at room temperature for 3 hours. The mixeddisulfide, S' labeled, was chromatographically separatedon an Amberlite IR-120 column from the reaction's otherproducts, cystine-S35 and homocystine without label (1).To minimize losses, it was elected to separate the ma-terial from the citrate buffer with an ion-exchange resin,rather than to attempt crystallization. Accordingly, frac-tions containing mixed disulfide were passed into a smallcolumn of Dowex 2 OH- washed with water; the com-pound was removed with N acetic acid (13) and lyophi-lized. Dissolved in water, samples of the material wereassayed for radioactivity and amino acid content. Ap-proximately 22.9 Ac was diluted in 100 ml of saline, auto-claved, and infused intravenously over a 4-hour period.Chromatography of this solution revealed that it alsocontained cystine, with 40%o of the radioactivity, andhomocystine, which was not radioactive. Urine was col-lected from the beginning of the infusion. Urine sulfatewas precipitated with benzidine (14) and counted onplanchettes with a gas-flow counter. The over-all re-covery and efficiency were 13.7%.

Renal extraction of amino acids was measured bycomparing renal venous plasma concentrations to ar-terial plasma concentrations, simultaneously obtained.Left renal vein catheterization was performed via the fe-moral vein.4 Peripheral arterial plasma concentration

2 Obtained from Schwarz Bioresearch, Inc., Orange-burg, New York.

3 Obtained from Nutritional Biochemicals Corp., Cleve-land, Ohio.

4 Certain restrictions wvere placed on the selection ofpatients for this study. Since the procedure is not with-out hazard, only patients with two well-functioning kid-neys were used. Patients with calculi in the renal pelviswere avoided because they are potential candidates for

was taken to be representative of renal arterial plasmaconcentration. The left renal vein blood was diluted byan unknown amount of left spermatic vein blood. Bloodwas prepared by the method of Brigham, Stein, andMoore for cysteine (as carboxymethylcysteine) andcystine (15). The carboxymethylcysteine peaks weresuperimposed upon small methionine sulfoxide peaks.The latter were evaluated separately in plasma processedaccording to the technique of Stein and Moore, without,however, the oxidation step (4), and deducted from thearea of the carboxymethylcysteine peaks.

RESULTS

The occurrence of mnixed disilfide in the nor-mnal human. Normal urine does not contain mixeddisulfide in amounts detectable by the Spackman,Stein, and AMoore technique (2). This is largelybecause there is insufficient material present; anadditional factor might be, however, that this areaof the normal chromatogram is somewhat compli-cated by the number of compounds present intrace amounts (Figure 1).

To evaluate the possibility that the amino acidoccurs normally but in trace concentrations, meas-tures were taken to increase the sensitivity for de-tection. Fifty ,tc of L-methionine-S35 was ad-ministered orally in a single dose to two subjectswithout liver or kidney disease. Blood, obtainedat 6 and 24 hours, was immediately processed forion-exchange chromatography. Urine was col-lected for 24 hours and preserved by refrigeration

surgery. Patients had not recently suffered any compli-cation, such as the passage of stones, since the lattermight compromise the desirability of giving the patienta methionine load to increase blood concentrations of themixed disulfide. Patients were preferably large people,so that a large catheter might be introduced to facilitatewithdrawal of blood. Finally, the patients were volun-teers for the procedure. In a preliminary study, a pa-tient was not given a methionine load, and significantpeaks of mixed disulfide were not detected. In anotherpatient, the renal vein blood was obtained at the timeof pyelolithotomy. Insignificant a-v differences were ob-tained; however, it became apparent that the kidney wasnot functioning, and nephrectomy was later required.The patient from whom data are reported was patientF. C. in Table II. This patient, who filled all the criterialisted above, excreted the highest amount of mixed di-sulfide and might have the highest blood levels. Onelimitation of the technique is that leaving the relativelylarge and stiff catheter in place for an extended periodto perform concurrent renal clearance studies seemedunwise. To minimize any effect of water extraction, uri-nary flow was kept low by restricting fluids, and theurinary bladder was not catheterized for clearance studies.

1957

GEORGEW. FRIMPTER

A NORMALURINE BMH 8 9 FANCONI SYNDROME

- -. -

350 400 4-0 300F 3303SO 400 40so 3.0 4" X

C Y.YR.. 42 * WIL"SS DSEASE

Il ait"Om f*f

350400m430a500'~~~~~~~~~~~~~~"InSWI 400 430 50D

Ad 354460.tfZ t k

380~400450 300 3460 430 500

FIG. 1. A) NORMALURINE. Several Ninhydrin-positive substances are present in trace amounts. B) SPONTANE-OUS APPEARANCEOF MIXED DISULFIDE PEAK IN A CHILD WITH THE FANCONI SYNDROME(TABLE I, PATIENT 1). Notethat there is a relatively high reading at 440 m/A (lower tracing) compared to isoleucine and leucine. C) WILSON'SDISEASE. The unlabeled peak between isoleucine and leucine was shown by isolation and further chemical tests to bethe mixed disulfide. D) TYPICAL APPEARANCEOF MIXED DISULFIDE AND CYSTINE PEAKS IN CYSTINURIA. E) CANINECYSTINURIA. The peak at 480 ml with relatively high reading at 440 my (lower of 3 tracings) is probably the mixeddisulfide.

and the addition of thymol crystals. After chro-matographic fractionation, 2.0 ml-fractions wereexamined for radioactivity. Peaks of radioac-tivity were found for sulfate, taurine, cystine, cys-tathionine, and methionine.5 Chromatography ofthe urine from one of these subjects showed asignificant peak corresponding to the mixed di-sulfide. After corrections for efficiency and de-cay, 1.3 x 104 counts above background werefound in the mixed disulfide area. This, in the24 hours following administration, indicated thatthe subject excreted 0.012%o of the administeredmethionine as the mixed disulfide. Urine fromthe other subject revealed a few counts abovebackground in the area under question, but these

5 That these compounds were radioactive was, of course,expected. The relative amounts of radioactivity con-tained in the peaks were not quantitatively determined.In finding the relatively high degree of radioactivity con-tained in the mixed disulfide, as opposed to the low radio-activity anticipated, technical limitations were encoun-tered (for example, when large amounts of urine werechromatographed, amino acids and unknown substances inthe counting vial precipitated). In addition, the rela-tive amounts of these compounds excreted would de-pend on various factors, including the distribution ofdietary sulfur, which was not studied. That these com-pounds were radioactive is important, however, as evi-dence that the methionine S53 was metabolized.

were not statistically significant. None of theplasma specimens contained any radioactivityeither in the area of the mixed disulfide or of ho-mocystine, whether or not air-oxidation of the-SH groups was performed (4).

To increase the concentration of the mixed di-sulfide for its determination in plasma, 12 g ofL-methionine was given in divided oral doses on2 successive days to a subject without liver orkidney disease. On the morning of the secondday of nonradioactive methionine loading, he wasgiven 50 uc of L-methionine-S35. Urine was col-lected throughout the day. Blood specimens weredrawn at 2, 4, 6, and 24 hours to minimize thepossibility of missing the peak of mixed disulfideactivity in the blood plasma. The picric acid fil-trates from each specimen were pooled, and 10/moles of nonradioactive mixed disulfide was addedas a carrier. Chromatography of 58.3 ml of thepooled plasma filtrate and counting as describedrevealed a significant radioactive peak (6.8 X 103uncorrected counts above background) of themixed disulfide. S35 peaks were also found forsulfate, taurine, cystine, cystathionine, and methio-nine (see footnote 5).

Occurrence of mixed disulfide in diseases otherthan cystinuria. To determine the specificity of

1958

DISULFIDE METABOLISMIN CYSTINURIA

the mixed disulfide for cystinuria, urine from sev-eral patients with other amino-acidurias was ex-amined by chromatography. Occasionally, a smallpeak corresponding to mixed disulfide was seenin urine specimens obtained from patients withthe Fanconi syndrome (Figure 1, B) or those withWilson's disease (Figure 1, C). These patientshad a marked generalized amino-aciduria. In achild with the Fanconi syndrome (Patient 1) andan adult with Wilson's disease (Patient 2), theamounts of the compound were sufficiently largeto calculate excretion and permit collection offractions containing the mixed disulfide and fur-ther identify them (Table I). In two other pa-tients with the Fanconi syndrome, oral methionineloading was followed by disulfide excretion suffi-cient to permit chromatographic isolation and sub-sequent confirmation of the identity of thecompound. An adult (Patient 3) received 15 g ofL-methionine in divided doses daily for 2 days,and urine was collected during the second 24hours. A child (Patient 4) was given a doseequivalent on a weight basis. The urine of Pa-tient 3 had a small peak in the area correspondingto the mixed disulfide before methionine adminis-tration, but that of Patient 4 had none. In bothof these patients there was a considerable peakcorresponding to the mixed disulfide after methio-nine loading (Table I). Fractions containing thesubstance were pooled, desalted, and subjected toperformic acid oxidation and paper chromatog-raphy. Equimolar amounts of cysteic acid andhomocysteic acid were demonstrated in the frac-tions from urine of all 4 patients.

Ion-exchange chromatography of deproteinizedplasma from patients with uremia has been de-

TABLE I

Asymmetrical disulfide in the urine of patients withdiseases other than cystinuria

Patient S04 Asymmetrical disulfide

g/day1. Child, Fanconi syndrome 1.30 mg per 100 ml

2. Adult, Wilson's disease 1.07 mg per 100 ml

3. Adult, Fanconi syndromeBefore methionine loading 3.11 25.9 mg per 24 hoursAfter methionine loading 10.6 72.1 mg per 24 hours

4. Child, Fanconi syndromeBefore methionine loading 0.79 0After methionine loading 2.29 12.9 mg per 24 hours

scribed by Woods, Rubin, and Luckey (16), whoobserved that "several unidentified Ninhydrin re-active substances" were present in the plasma ofpatients with uremia. In this laboratory, chro-matograms of the deproteinized plasma from pa-tients with uremia have consistently shown a peakbetween isoleucine and leucine. To establish moredefinitely that this peak was the mixed disulfide,50 uc of L-methionine-S35 was administered orallyto a patient without amino-aciduria who had aterminal chronic renal disease (blood urea ni-trogen 240 mg per 100 ml). Blood specimensdrawn at 2, 4, 6, and 24 hours were deproteinized,a carrier of nonradioactive mixed disulfide wasadded, and the pooled specimen was chromato-graphed. The peak between isoleucine and leucinewas significantly radioactive. Amounts insufficientfor further chemical analysis were obtained.However, the presence of radioactivity in thecharacteristic peak is presumptive evidence thatthis was the mixed disulfide.

Occurrence of mixed disulfide in human andcanine cystinuria. The mixed disulfide was in-variably found in the urine of 15 patients withcystinuria (Figure 1, D). Fifteen to 224 mgwereexcreted daily (Table II). There was no con-sistent relationship to the amount of cystine, orni-thine, lysine, or arginine excreted. Deproteinizedplasma from these patients with cystinuria did notcontain the mixed disulfide in amounts detectableby the Spackman, Stein, and Moore technique.Urine from a Dachshund with cystine calculicontained large amounts of cystine, ornithine, andlysine, but no arginine. There was a peak withthe characteristics of mixed disulfide (Figure 1,E).

Origin and metabolic fate of mixed disulfide.To determine the origin of the mixed disulfide, 50,uc of L-methionine-S35 was administered orallyto a patient with cystinuria, and urine was col-lected as the patient voided. The specific activi-ties of the cysteic and homocysteic acids processedfrom mixed disulfide and from cystine in the urineare shown in Table III. Radioactive mixed di-sulfide was not detected in deproteinized plasmaobtained when considerable radioactive mixed di-sulfide was found in urine. No radioactivity wasfound in the homocystine area of either plasma orurine chromatograms. Plasma specimens, how-ever, revealed radioactive sulfate, taurine, cystine,

1959

GEORGEW. FRIMPTER

TABLE II

Amino acid excretion in 15 patients with cystinuria

AsymmetricalPatient Glycine Cystine disulfide Ornithine Lysine Arginine

mg/24 hoursJ.L. 66.7 849 69.3 330 2,334 1,048S.G.* 58.9 1,067 74.9 550 2,456 1,162F.G.* 82.8 765 100 534 2,217 931F.C.* 107 755 224 585 2,308 1,531W.H. 191 618 36.7 31.6 1,292 475I.H. 150.4 557 39.6 214 1,280 447N.S. 67.8 234 15.2 138 624 140H.F. 26.0 708 25.1 163 283 80A.S. 122 666 58.4 326 435 552J.P. 47.1 424 32.7 264 891 555J.G. 78.9 646 37.2M.F.t 754 602 89.5

mg/100 mlJ.KI( 2.81 31.9 7.62 36.6 151 66.1G.S.1 1.97 27.1 1.79 9.76 80.6 35.5E. C.t15.7 24.3 6.1 26.7 113 169

* Unrestricted diet (see text).t This patient also demonstrated a defect of glycine excretion and possible tubular excretion of cystine (17). All

other amino acids were excreted in normal amounts by these patients.I Twenty-four-hour urine collections were not obtained on these patients.

cystathionine, and methionine. With correctionsfor efficiency, decay, and color of the specimens,15% of the administered counts was excretedwithin 24 hours.

For determination of the effects of the mixeddisulfide per se, it was infused into the tail veinof a rat. The infusion resulted in an increase intaurine excretion relative to creatinine from 0.94umole per mg of creatinine to 2.94 Mumoles per mgof creatinine. Except for the increased taurineexcretion, no amino-aciduria was induced by themixed disulfide.

To explore the possibility that the mixed di-sulfide is excreted in cystinuria consequent to theabsence of some mechanism for its normal dis-posal, S35-labeled, mixed disulfide was adminis-

TABLE III

Specific activities of S35 from urinary cystineand asymmetrical disulfide

From cystine From asymmetrical disulfide

Homocysteic acidCysteic Cysteic Homocysteic

Specimen acid acid acid Cysteic acid

hours cpm peiismole

0-1.25 330 982 8835 9.05.25-6 330 417 2630 6.36-24 190 274 333 1.2

tered to a patient. Radioactivity was found largelyin urine sulfate precipitated with benzidine. Thedistribution of radioactivity in the chromatogramfrom the urine in the first 24 hours after the be-ginning of collection is shown in Table IV. In the24 hours following administration, the patient ex-creted 19.57% of the radioactivity. Mixed di-sulfide in the urine contributed only 1.12% of thetotal urinary radioactivity, although mixed di-sulfide constituted 60% of administered radioac-tivity.

Renal extraction of amino acids. Becauseplasma concentrations of the mixed disulfide couldnot be easily and accurately determined, studying

TABLE IV

Distribution of administered and urinary S35

AdministeredCystine (S35-S35) ............ 2.025 X 107 countsMixed disulfide (S35-S) ..... 3.053 X 107 counts

Excreted radioactivity

24-hour urine

Mixed disulfideCystineSulfate

CountsX 106

0.1182.956.75

%ofadminis-

tered0.2335.85

13.40

%oftotal

excreted1.12

30.1068.78

1960

DISULFIDE METABOLISM IN CYSTINURIA

TABLE V

Renal extraction of amino acids

A-V*Amino acid A

Proline .041Glycine .092Isoleucine .029Phenylalanine .016Methionine .076Cystine .059Cysteine .39Mixed disulfide .42

* Arterial concentration-renal venous concentration

arterial concentration.

the renal plasma clearance of the compound was

impossible. To estimate the clearance of themixed disulfide, renal venous plasma concentra-tions of amino acids were compared to those of a

simultaneously obtained arterial specimen. Thepatient was given orally 5 g of L-methionine 2hours before the obtaining of blood specimens toassure the presence of detectable amounts of mixeddisulfide. The results of this experiment are

shown in Table V. [In calculation of A-V/A, theareas under the curves for the various amino-acidpeaks were used directly (identical amounts were

chromatographed) rather than first calculatingconcentrations, a procedure which tends to "roundoff" numbers with a loss of sensitivity.] Cystineappeared to be extracted similarly to other aminoacids, whereas cysteine was extracted to a super-

normal degree. In addition, the mixed disulfideappeared to be cleared to a high degree. The re-

sults do not take into account water extraction bythe kidney, which should not, however, invalidatethe comparison between the various amino acids.

DISCUSSION

The experiments reported here permit com-

ment on the significance of the new amino acid,the disulfide of L-cysteine and L-homocysteine.The demonstration of radioactive mixed disul-fide in the urine of normal subjects after ad-ministering L-methionine-S35 suggests that thecompound is normally excreted in trace amounts.When methionine turnover was increased by oralloading, the disulfide was also demonstrated in theplasma of a normal subject.

The spontaneous occurrence of a peak with the

characteristics of the mixed disulfide in urines ofpatients with a generalized amino-aciduria alsosuggests that this is a "normal" amino acid whichescapes into the urine in detectable quantities un-der special conditions. The identification of thispeak by chromatographic demonstration of equi-molar amounts of the oxidation products, cysteicacid and homocysteic acid, is evidence that thispeak was the mixed disulfide. No other sulfur-containing amino acid has been reported to oc-ctupy the same position as the mixed disulfide inthe Spackman, Stein, and Moore system (2).Furthermore, the increase in magnitude of thepeak following an oral load of methionine con-firms the nature of the peak. The suggestion ofpresence of the mixed disulfide in the plasma ofpatients with advanced renal insufficiency furtherindicates that the amino acid has a wide distribu-tion in trace amounts.

The mixed disulfide was invariably found inthe urine of patients with cystinuria. In 15 pa-tients, the amounts excreted bore no correlation ona molar basis to other amino acids excreted. Theamount of mixed disulfide excreted appeared tobe related to methionine content of the diet. Thiswas not specifically studied; however, those pa-tients who were not treated with a low methionine(400 mg) diet excreted higher amounts of disul-fide (Table II). The compound's tendency to beunstable was noted, although this was not specifi-cally studied. After administration of L-methio-nine-S35 to a patient with cystinuria, there werea few counts above background in the homocystinearea of one chromatogram of urine. The largeamount of cystine present compared to the smallamount of mixed disulfide might suggest that themixed disulfide arises by disulfide interchange be-tween cystine and homocystine, exhausting thelatter. However, that homocystine has not beenfound in this laboratory in freshly collected urinesmilitates against this origin of the mixed disulfide.If the compound is merely another diamino aminoacid cleared from the plasma of the cystinuriapatient at a rate similar to the glomerular filtra-tion rate, it should have sufficient concentrationin the blood plasma, at least in the normal indi-vidual, to yield an easily detectable Ninhydrinpeak. If, however, the mixed disulfide is clearedfrom plasma at a rate similar to the renal plasmaflow, then difficulty might be anticipated in ap-

1961

GEORGEW. FRIMPTER

preciating the peak in plasma." This would implyrenal tubular secretion or renal production of theamino acid. Renal tubular secretion of aminoacids in cystinuria has been suggested by otherstudies in this laboratory (17). Renal tubularsecretion of amino acids under loading conditionshas been shown in dogs by Webber, Brown, andPitts (18) and demonstrated in human fl-amino-isobutyric acid excretion by -Armstrong and co-workers ( 19).

The origin of mixed disulfide from methioninein cystinuria is confirmed by the study in whichL-methionine-S35 was administered to a patient.The urine obtained 1.25 hours after feeding themethionine-S35 had the highest specific activity ofthe mixed disulfide of all the specimens examined.The homocysteine sulfur had initially 9 times thespecific activity of the cysteine sulfur, but withtime these specific activities tended to becomeequal. Reasons for this might include, first, thecontribution to the cysteine side of mixed di-sulfide by cysteine from other sources, and second,dilution of the labeled cysteine in a larger bodypool than the pool for dilution of labeled homo-cysteine. Similarly, the specific activity of thecysteic acid from mixed disulfide was alwayshigher than the specific activity of the cysteic acidfrom cystine. This might also be explained onthe basis that cystine labeled with S35 was dilutedin a larger body pool than was the mixed disulfide.

Mixed disulfide was metabolized by the rat.The contamination of the S35 mixed disulfide withS35 cystine limits the interpretation of the experi-ment on iv administration to the patient withcystinuria. The rather large amounts of S35 ex-creted as sulfate, however, and the relativelysmall amount (1.12%o) excreted as mixed disul-fide are compatible with metabolism of mixed di-

6 For example, in one patient the daily urinary excre-tion of mixed disulfide was 225 mg or 886 /moles, andthe inulin clearance was 141 ml per minute. Substitutingin the formula, C = UV/P, C = 203,040 ml per 24 hours;UV= 886 gmoles per 24 hours; P = 886/203,040 =.00444mole per ml or 0.132 jSmole in a 30-ml plasma specimenprepared with picric acid. This amount is easily detectable,although not accurately measured by the technique em-ployed. The renal plasma flow in this patient was notmeasured but may have been approximately 5 times theinulin clearance. Then substituting in the formula, C =

UV/P, P = 0.00087 /Amole per ml or 0.0022 Amole in a30-ml plasma specimen. This small amount would notbe detected by the method used.

sulfide to sulfate. The fraction of S35 excretedin 24 hours by the cystinuria patient who receivedS35 mixed disulfide and S35 cystine was similar tothe fraction of S35 excreted in 24 hours by thepatient with cystinuria who received L-methionineS35. The possibility of a difference in the rateof metabolism of mixed disulfide between thenormal person and the patient with cystinuriacannot be answered from the available data.

The intermediary metabolism of the sulfur-containing amino acids, as currently understood,does not provide for production of the mixed di-sulfide. Free cysteine has been demonstrated indeproteinized blood plasma (15). The presentstudy failed to demonstrate free homocysteine indeproteinized plasma or urine. Neither homo-cystine nor the mixed disulfide was produced byaeration at pH 7 of deproteinized plasma fronteither normal subjects or from patients with cys-tinuria after loading with L-methionine-S35. Ho-mocysteine may, of course, be present in theplasma in amounts that are not detectable by thesemethods. The studies suggest that mixed disul-fide is produced in the tissues and cleared rapidlyfrom the blood by the kidneys in cystinuria. Thesite or sites of production cannot be determinedfrom present data.

The study of renal extraction of amino acidshelps to explain the difficulty in determiningplasma concentrations despite large urinary ex-cretion of the mixed disulfide. The compound isapparently extracted from the plasma at a rategreater than any other amino acid.

A by-product of this study was the observationthat cystine (-S-S-) is handled by the kidneyssimilarly to other amino acids but cysteine (-SH)is extracted to a high degree. Consideration ofthis finding is interesting in light of the experi-ments of Brand, Cahill, and Harris, who demon-strated that feeding of cystine does not result inincreased cystine excretion in cystinuria but thatthe fed cystine was found as urinary sulfate (20).Also, Brigham, Stein, and Moore (15) andFrimpter and co-workers (17) have shown thatcysteine is virtually absent from urine. Thesedata give further support to the previous sugges-tion (17) that plasma cysteine is oxidized to cys-tine before its appearance as urinary cystine. Thesmall extraction of cystine, like that of mostamino acids, demonstrates a high degree of reab-

1962

DISULFIDE METABOLISM IN CYSTINURIA

sorption by the renal tubules and agrees with theobservation of Rosenberg, Downing, and Segalthat cystine does nlot compete with lysine, arginine,and ornithine for uptake by rat kidney slices, al-though the three basic amino acids do compete(21). Crawhill, Scowen, and Watts have reporteda remarkable and prompt reduction in urinary cys-tine in cystinuria by the administration of D-peni-cillamine (22, 23). The cystine was accounted forby the excretion of the disulfide of L-cysteineand D-penicillamine. The present report suggeststhat the disulfide of cysteine and penicillamine(22, 23) arises through reaction of the sulfhydrylof cysteine with the sulfhydryl of penicillamine(p3, 8,-dimethylcysteine) in the kidney or urinarytract. This reaction may, of course, be non-enzymatic and occurs readily it vitro, particularlyat pH over 7 and in the presence of air. Unfor-tunately in the present study, plasma filtrate ade-quate for chromatography of the basic amino acidsdid not remain.

The results of the present study are most com-patible with asymmetrical disulfide's being a "nor-mal" metabolite that exists in trace amounts inhuman blood plasma and urine. It is probablymetabolized by the normal subject, largely to sul-fate. In patients with cystinuria, large amountsare excreted into the urine. The exact mecha-nism of excretion has not been defined, but theobservations suggest an abnormally high renal ex-traction of plasma mixed disulfide. This studydoes not take into account the possibility thatdifferent rates of formation of mixed disulfidemay exist in the normal subject and in patientswith cystinuria. It neither rules out nor confirmsformation of cystine and the mixed disulfide inthe kidneys or in the urinary tract.

SUMMARY

1. The asymmetrical disulfide of L-cysteine andL-homocysteine is uniformly excreted in the urineof patients with cystinuria. This amino acid wasfound in the urine of a dog with cystinuria andcystine stone disease. The asymmetrical disulfidewas occasionally seen in the urine of patients withgeneralized amino-aciduria because of the Fanconisyndrome or Wilson's disease. The amount wasincreased by feeding L-methionine to two patientswith the Fanconii syndrome. The disulfide waspresunmPtively demonstrate(l in the plasma of pa-

tients with advanced renal insufficiency. Theseexperiments demonstrated that the mixed di-sulfide has a wide distribution in trace amounts.

2. The asymmetrical disulfide, labeled with S35,was excreted by a patient with cystinuria after ad-ministration of L-methionine-S35. The homo-cysteine sulfur was initially of higher specificactivity than the cysteine sulfur.

3. Intravenously administered mixed disulfidewas metabolized by a rat and apparently by apatient with cystinuria.

4. It is proposed that the mixed disulfide arisesthrough a mechanism whereby cysteine combineswith homocysteine. The data suggest that theamino acid may be a normal metabolite. In cysti-nuria and in the generalized amino-acidurias,mixed disulfide seems to be cleared rapidly fromthe plasma by the kidney and appears in the urine.

5. Determination of renal venous and arterialconcentrations of amino acids in cystinuria re-vealed that cystine is extracted to a degree similarto amino acids not considered to be involved incystinuria. The plasma cysteine, however, is ex-tracted to a high degree. This study suggests thaturinary cystine arises largely from plasma cysteinein cystinuria. The asymmetrical disulfide alsoappears to be extracted to a high degree, althoughrenal production of the amino acid is not ruledout.

ACKNOWLEDGMENTS

Several physicians, especially Drs. Victor Marshall,Russell Lavengood, J. Edwin Drew, John W. Draper,John Woodward, Joel Clark, and Alexander Bearn,kindly permitted me to study their patients. Dr. JayMeltzer suggested the administration of L-methionineto his patient with the Fanconi syndrome and co-operated in the performance of the study, as did Dr.W. W. McCrory with his patient. I am indebted toDr. Myron Schaeffer and Dr. Felix 0. Kolb for theurine of the dog with cystinuria. Dr. M. Halpern kindlyperformed the renal vein catheterization.

Drs. Vincent du Vigneaud, Melvin Horwith, WilliamH. Stein, Ralph E. Peterson, Abraham Mazur, andDavid D. Thompson contributed immeasurably by theirstimulating interest and advice.

The skillful technical assistance of Miss Audrey Bass,Miss Naomi Schechter, and Mrs. Marie Mongelli isgratefully acknowledged.

REFERENCES

1. Frimpter, G. W. The disufide of L-cysteine andL-homocysteine in urine of patients with cystinuria.J. biol. Chem. 1961, 236, PC 51.

1963

GEORGEW. FRIMPTER

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

3. Hamilton, P. B., and D. D. Van Slyke. The gaso-metric determination of free amino acids in bloodfiltrates by the Ninhydrin-carbon dioxide method.J. biol. Chem. 1943, 150, 231.

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

5. McManamy, R. H., C. C. Lund, and J. L. Oncley.Unbound amino acid concentrations in human bloodplasmas. J. clin. Invest. 1957, 36, 1672.

6. WVoods, K. R., and R. L. Engle, Jr. Automaticanalysis of amino acids. Ann. N. Y. Acad. Sci.1960, 87, 764.

7. Thompson, J. F., C. J. Morris, and R. K. Gering.Purification of plant amino acids for paper chroma-tography. Analyt. Chem. 1958, 31, 1028.

8. Schram, E., S. Moore, and E. J. Bigwood. Chro-matographic determination of cystine as cysteicacid. Biochem. J. 1954, 57, 33.

9. Smith, I. Chromatographic Techniques. London,Heinemann, 1958.

10. Giri, K. V., A. N. Radhakrishnan, and C. S. Vaidy-anathan. Some factors influencing the quantitativedetermination of amino acids separated by circu-lar paper chromatography. Analyt. Chem. 1952,24, 1677.

11. Du Vigneaud, V., L. F. Andrieth, and H. J. Loring.The reduction of cystine in liquid ammonia bymetallic sodium. J. Amer. chem. Soc. 1930, 52,4500.

12. DuVigneaud, V., W. I. Patterson, and M. Hunt.Opening of the ring of the thiolactone of homo-cysteine. J. biol. Chem. 1938, 126, 217.

13. Dreze, A., S. Moore, and E. J. Bigwood. On thedesalting of amino acids by ion exchange. Analyt.chem. Acta. 1954, 11, 554.

14. Power, M. H., and E. G. Wakefield. A volumetricbenzidine method for the determination of inor-ganic and ethereal sulfate in serum. J. biol. Chem.1938, 123, 665.

15. Brigham, M. P., W. H. Stein, and S. Moore. Theconcentrations of cysteine and cystine in humanblood plasma. J. clin. Invest. 1960, 39, 1633.

16. Woods, K. R., A. L. Rubin, and E. H. Luckey.Effects of dialysis with the artificial kidney onplasma amino acids in uremic patients. Trans.Amer. Soc. artificial intern. Organs 1961, 7, 83.

17. Frimpter, G. WV., M. Horwith, E. Furth, R. E. Fel-lows, and D. D. Thompson. Inulin and endogenousamino acid renal clearances in cystinuria: evidencefor tubular secretion. J. clin. Invest. 1962, 41,281.

18. WVebber, W. A., J. L. Brown, and R. F. Pitts. In-teractions of amino acids in renal tubular trans-port. Amer. J. Physiol. 1961, 200, 380.

19. Armstrong, M. D., K. Yates, Y. Kakimoto, K.Taniguchi, and T. Kappe. Excretion of j-amino-isobutyric acid by man. J. biol. Chem. 1963, 238,1447.

20. Brand, E., G. F. Cahill, and M. M. Harris. Cystinuria.II. The metabolism of cystine, cysteine, methio-nine, and glutathione. J. biol. Chem. 1935, 109,69.

21. Rosenberg, L. E., S. J. Downing, and S. Segal.Competitive inhibition of dibasic amino acid trans-port in rat kidney. J. biol. Chem. 1962, 237,2265.

22. Crawhill, J. C., E. F. Scowen, and R. W. E. Watts.Effect of penicillamine on cystinuria. Brit. med. J.1963, 1, 588.

23. Watts, R. W. E., J. C. Crawhill, and E. F. Scowen.Diminished cystine excretion in cystinuria associ-ated with administration of D (-) -penicillamine(abstract). Clin. Res. 1963, 11, 230.

1964