IPHENYLALANINE METABOLISMANDFOLIC ACIDANTAGONISTS*

BY THEODOREL. GOODFRIENDt AND SEYMOURKAUFMAN

(From the Endocrinology Branch, National Cancer Institute, and the Laboratory of CellularPharmacology, National Institute of Mental Health, Bethesda, Md.)

(Submitted for publication March 10, -1961; accepted April 7, 1961)

Amethopterin and related folic antagonists in-hibit the reduction of folic acid to tetrahydro folicacid (2). The clinical effects of these drugs arethought to be related in turn to the role of tetra-hydro folic acid in single-carbon transfer reactionsduring nucleic acid synthesis (3). Recently, pteri-dines such as folic acid have been implicated inanother area of metabolism, the conversion ofphenylalanine to tyrosine. In this conversion thepteridine is thought to act as a cofactor which isreduced in a separate enzymatic reaction, thentakes part in the actual hydroxylation of phenyl-alanine. The reaction sequence as it is currentlydeduced from studies in vitro is presented in Fig-ure 1. Enzymatic activities corresponding tophenylalanine hydroxylase ("rat enzyme") andthe cofactor-reducing enzyme ("sheep enzyme"),as well as the cofactor itself, have been observedin normal human liver (4).

It has been shown that folic acid antagonistssuch as amethopterin inhibit tyrosine synthesis ina partially purified enzyme system, and in vivo inthe rat (5). The site of inhibition is believed to bethe enzymatic reduction of the pteridine cofactor.The work reported here is an investigation of theeffects of amethopterin on phenylalanine metabo-lism in man, and the possible relevance of this ef-fect to the clinical properties of folic antagonists.Experiments were also performed on rodents toexplore the possible interrelationship of folic acidmetabolism, phenylalanine metabolism, and cancerchemotherapy.

METHODS

Patients receiving amethopterin (Methotrexate) forthe treatment of histologically-proven metastatic chorio-

* A preliminary report of the clinical findings describedhere has been presented at the Forty-Fourth AnnualMeeting of the Federation of American Societies forExperimental Biology, Chicago, Ill., April 13, 1960 (1).

t Present address: Department of Medicine, Washing-ton University Medical Center, St. Louis, Mo.

carcinoma or leukemia were studied before, during andafter therapy. Amethopterin was given in either of twodosage schedules: patients with carcinoma received 20 to25 mg intramuscularly daily for 5 days in repeatedcourses (6) ; patients with leukemia received 5 to 10mg orally daily, as continuous therapy. Several pa-tients were studied who received 6-mercaptopurine orallyin daily doses of 100 to 200 mg and several who receivedvincaleukoblastine, 5 to 10 mg intravenously, each dayfor 3 days (7). Six adults were studied who did nothave known malignant disease and did not receive chem-otherapy.

Phenylalanine metabolism was measured by an oralphenylalanine tolerance test. The DL or L -form of the

PHENYLALANINEHYDROXYLATION

9 + 0 + REDUCED TPNQ + 0, + COFACTOR

CH2CHNN2COON

PHENYLALANINERAT" 'SHEP Folic

ON ENZYME ENZYME Antagonistsz (Phenytof/al e

Hydroxylast)

CH2

COO" OXIDIZEDTYROSINE + H20 + COFACTOR TPNH

FIG. 1. SCHEMATIC REPRESENTATIONOF THE ENZYMESYSTEM FOR HYDROXYLATION OF PHENYLALANINE IN

MAMMALIANLIVER ANDTHE LOCUSOF INHIBITION BY FOLICANTAGONISTS.

amino acid was given after an overnight fast in a doseof 0.1 g per kg body weight mixed in orange-ice. Bloodwas drawn at 0, 2, 4, and 6 hours after the phenylalaninewas given. The serum was analyzed for phenylalanineby the method of La Du and Michael (8), and severaldeterminations confirmed by the method of Kaufmanand Levenberg (5). Tyrosine was measured by themethod of Udenfriend and Cooper (9).

Recoveries of added phenylalanine and tyrosine were

quantitative from sera of patients before, during and af-ter therapy. The decay constants for serum phenylala-nine were calculated for the 2- to 4-hour, and 4- to 6-hour

1743

THEODOREL. GOODFRIENDAND SEYMOURKAUFMAN

intervals, from the equation

2.303 100(C2 - C4) 10(C4 - C6)- 2~logso or2 C2 IC4

where C2, C4, and C6 represent the increment of serumphenylalanine above the fasting level at 2, 4, and 6hours, respectively. The greater of the two values for kwas included in Table I. These calculations assume thatserum phenylalanine fell in accordance with a first orderrate equation after an oral dose. The assumption is sup-ported by the agreement of the k values for each timeinterval after the second hour; i.e., the logarithms of thephenylalanine values at 2, 4, and 6 hours fell on astraight line when plotted against time. The otherparameters of phenylalanine tolerance were the maximalserum phenylalanine achieved during the test and the in-crement in plasma tyrosine during the test. Each pa-tient provided her own control before or after therapy.Several tests were performed in which oral L-tyrosinewas used in place of phenylalanine, and the data weretreated in the same way.

In some studies urine was collected in 6-hour sampleson the days of the tolerance tests and assayed for phenyl-pyruvic acid by the method of Frilling and Sydnes (10).

The therapeutic response to the drugs was assessedby hormonal and radiological criteria in the cases ofchoriocarcinoma, as described by Hertz and co-workers(6), and by hematological criteria in the cases of leu-kemia. Drug toxicity was assessed clinically. Theparameters for toxicity included mucous membrane ul-ceration, skin eruptions, and bone marrow suppressionas reflected in peripheral blood counts.

Several tissues were assayed for the presence of theenzyme, phenylalanine hydroxylase, by using the twoindependent methods described by Kaufman (11). Thetissue samples were homogenized in either 0.01 M aceticacid, or in 0.01 M potassium phosphate buffer, pH 7.0.Simultaneous assays were performed on rat liver as aday-to-day check of the assay procedure.

Folic deficiency in Sprague-Dawley rats was inducedby feeding weanling animals either a commercial folic-deficient diet with 2 per cent succinylsulfathiazole (Sulfa-suxidine, Nutritional Biochemicals) for 7 weeks, or ahigh-fat, folic-deficient diet for 16 weeks. The latterdiet, developed by Silverman and Pitney, prevents theweight loss usually seen in folic deficiency in rats (12).Controls were given the former diet plus 20 mg of folicacid per kg of diet. Folic deficiency in the first groupof animals was documented grossly by failure to gainweight, and by low total white blood cell counts. Inthe group fed the high-fat diet, deficiency was docu-mented by the elevated urinary excretion of formimino-glutamic acid, as measured by a microbiologic assay.The livers were assayed for folic acid plus precursorswith the digestion procedure of Silverman and Gardiner(13), and a microbiological assay employing Pediococ-cus cerezvsnae. The same livers were assayed for thecofactor of phenylalanine hydroxylase by the method ofKaufman (14). Assays were performed in the linearrange of the cofactor determination.

J

U)

0

0cr-

0 20C

0.150

0.1 00

0.050

Hrs. 2 4 6 8

SERUMPHENYLALANINE( LM / ml. )

4

Hrs. 2 4 6 8

PLASMATYROSINEl iM/ml.)

FIG. 2. A REPRESENTATIVE PAIR OF ORAL PHENYLALA-

NINE TOLERANCETESTS PLOTTED ON LINEAR COORDINATES

(PATIENT 1, TABLE I).

A pilot study was performed to test whether changesin the phenylalanine or tyrosine content of the diet af-fected the response to amethopterin of transplantablemouse leukemia. The diets contained either 1 per centL-phenylalanine and 1 per cent tyrosine (normal), 3.5per cent L-phenylalanine and 1 per cent L-tyrosine (high-phenylalanine), or 1 per cent L-phenylalanine and no ty-rosine (low tyrosine). The diets were formulated ac-cording to Franklin, Stokstad, Belt and Jukes (15), ex-cept that the casein hydrolysate was treated to removearomatic amino acids 1 and supplemented with tryptophanand methionine. Survival times were recorded, and adose-response curve to amethopterin was determined foreach diet after inoculation of the mice with strain L-1210transplantable leukemia. The general procedure of Goldin,Venditti, Humphreys and Mantel was employed (16).

RESULTS

Phenylalanine tolerance tests. Table I presentsthe results of the clinical study and includes thetoxic and therapeutic effects of amethopterin, andthe parameters of phenylalanine tolerance. Insome patients repeated courses of therapy andrepeated phenylalanine tolerance tests were per-formed. Every patient demonstrated impairedphenylalanine tolerance after amethopterin. Thiswas manifest by a higher maximal serum phenyl-alanine level, a slower decay from this maximum(smaller k), and a smaller increment in blood ty-rosine. A typical study is presented graphicallyin Figure 2, and phenylalanine values for severalother studies are plotted on semilogarithmic co-ordinates in Figure 3.

The shortest time interval between the first in-jection of amethopterin and the performance of atolerance test was 12 hours, (Patients 5, 6), and

1 Supplied as compound MJ 309B by Mead, Johnsonand Co., Evansville, Ind.

1744

PHEXYLALANINEMETABOLISMAND FOLIC ACID ANTAGONISTS

>4

z a

C)00

- o_ -o -° - N N N oo0+ + + + + + ± + +

~ NOC4 OCr- N - Co00o to oc U 0

_

0

0o ON 0 0N 0 0 ^ COc 14U 0% O>r- C o o or

0 0 0 0 0 0z zz zz z

ON NOo oN 0+ + +

0C e0 00 004CC f--U) 00

'CO 00-l NO4 0)4 0rC u--

i-~-sN (D.0oN or - NN_-- NO r-c

NU) 'VI) o0 ZN t- uU)tt U) tNo oo 0000 00 0 00

66 000 00 00 00 00 60

uLr-.- ocIU)Nr4 m L 00000 00066 66

U- cN oor- o-oOcC uo-O "oCN 4m Ir, r C4o\1)o 0t to1 0u u -X st m C) o o- Co X0 o om M C7, 4 o X- r -oe 0t t,i 0 c O C0 °40 C>0 e 00 0X 0 0o - 00Ch ONCON 0 t- t-Ut) 0 ))04c041 41) 00 - f- o0 00 0: t- -

66c_66-66 -006 600 6600 66060 6600 6' 660 6 6 6 6O _

0040- u)U)0 O.)U)ue4i) 4N MLu) N r-mm 000OO0 1) -e - - t- N t sN ens 00C' a, Uf- t-_ 4 O000o Ur N co tU-. 41)0 4U1N ,t NC-_ NoC u-1_ 0 t- 0NC 4 ) N e4)n i)m0C-0 !0oC - t N N't m) ON -M ten mr- - " 4 -* 4 ) N N - t1) t) N N N666 666666660606 66606066 6660 6 6 6 6 6 6 6 6__~~~~~ ___ '_ :: .:42 -Z _: .4 __ : : -1 .4 4 '4 _ _:

:4. ~_ 1-)-4a~z lz9 Z ~ Q ml ~- Cza ~ a a 9 : -, -) . -)

t- t- It -t o000 40O(Ds 0 CCm m No 40Cr--r--t- Nt ' t t aON o \ t t s1it)41) U\ U- ) t$4 U.4 LL) )U), 1l) CCd CO'C 40d "ie4 6 1044 4)0 " X 0 0 C4r 0CCC U)

-. U) U. U) U)U)1in UL in U) 1U 0 Lo 1U) uU) U) 1r) U) in in;^- C1 O6 60r-i C6 6 6 6 Ncs C0 00 0 _ 00 4\0F:~~ m o - -

0 0U) 0nU)) 0U)U) 0U)0 O04) 4O1) 0U) OQ QON NN N r-- ' 1 C4 C1N--C")9 b1

- me0n

-.1

xC1C

000 C0 U)O 0 0 e(0I U) (N CN U) C

Nt N - U)I-

U) U)6 6

x

0 CC

0I

5n

x>4C)X

~~~~~~~~1 O 6

CC 6 CC66

0 0vQ Y X0 u 0 00(d 0Q 0u 0v 0 - 0 u 0 0U 0 6-£'5>-0 U *- to u- *- * C-, u ZI- C)0 C).000Iu

- 40 40) ; 41) C 0- 00 0o0 - N 40) + U) C

>40N,+0!Cd

0Q t .

0 -

1745

En

I-

0X7 X

0

-W 0

rn

CC

r.H.1-C

660 "t,,Jtc

xd v.O~5.cd

r_

6q .0CC 0

a) w :,

0+-

~,,zV-

C

'° O 4C)4)0

5c)

10

bC04=

" V.-C)

.0

00-

b0

Cecr

CC3

00

CJ-

-44

U)4

0CJi

.4= C)

Q CC.h

0-.W.C)0f>U)CFC

11I >4080.0*0C)

- 0.0r

11 C.C jC)0beii s 2

CCogU) b4

0T0

0

.C_

r-

._

._I

ul V) (A cn U) V) U) U) :0v v V 4) u v v w u

.>. >4 >. >4 >4 >.

THEODOREL. GOODFRIENDAND SEYMOURKAUFMAN

0

t-00 CN00 0%214 m00 00 0 006 o _ o

0 u) .d m

O uo Lo6 6 6 6

o

4. m

O, es

0% c20

.4 .2 .

.2.4 .2 .- M -, .

N6 26if 1cooC 11)

1-4 10- - 0

)2) di) 2)w q0 0 0 00 0t0 0 0 0 0 0z z z z z z

0 00 0Z z

)00

000 -. eN mc.. 1.4 N N C4N C4

Oz

cn4P-

Id0o

EN.I2

:dwz

-J.4-J

zIxQ. 0

V)0.

PHENYLKETONURIC3.0 -

2.0

1.0 ROZYGOTES

0.5 -OJ 0~2A~After0 Amethopterin

0.2NORMALS

0.1 7

\05 Before/Therapy

1.022 3 4 5 6 7

HOURS

FIG. 3. ORAL PHENYLALANINE TOLERANCE TESTS,PLOTTED ON SEMILOGARITHMIC COORDINATES. The curveslabeled 1, 2, 3, and 1A, 2A, 3A are from the correspond-ing patients in Table I. The suffix "A" indicates testsperformed after amethopterin. The curves labeled "heter-ozygotes" and "normals" are from the data of Hsia andPaine (17). "Heterozygotes" are subjects whose chil-dren had phenylpyruvic oligophrenia. The curve labeled"phenylketonuric" is a phenylalanine test performed dur-ing the current study on a subject with known phenyl-pyruvic oligophrenia. All values are corrected for thefasting phenylalanine level except those for the phenyl-ketonuric subject.

a marked change in phenylalanine tolerance wasevident at that time. The changes in phenylala-nine tolerance were observed 2 weeks after the lastinjection of amethopterin.

The control tests (before therapy) in patientswith cancer did not differ from those in patientswithout cancer. The fasting levels of phenylala-nine and tyrosine did not differ in these groups,nor did they change after therapy. Also includedin the table and in Figure 3 are the mean valuesreported by Hsia and Paine for similar tests innormal subjects and presumed heterozygotes forthe gene producing phenylpyruvic oligophrenia(17). One test in the current study was done' ona patient known to have phenylpyruvic oligo-phrenia.

Oral tyrosine tolerance tests were performedin three patients before and after therapy.Amethopterin did not alter the absorption or me-tabolism of tyrosine as measured by the maximal

1746

a)

2)0d

O22+00

00

222)0

2o m

v Cd 0

-o~

K

a2)20

0

4Cd a

_C w

024

&.

_u a

0

> ow

lz(-I

m¢4;Q

Ea0A

a)

2)

0

I2

a20222

.0

.400r.be

.0

24.)20

e

el

1.C.)I--,v

IS1.:L

PHENYLALANINEMETABOLISMAND FOLIC ACID ANTAGONISTS

plasmna level andl the rate of decay of plasma ty-rosine.

Routine liver function tests (serum bilirubin,alkaline phosphatase, and glutamic-oxalacetictransaminase levels) failed to show consistentchanges after amethopterin therapy.

Patients who received 6-mercaptopurine or

vincaleukoblastine demonstrated toxic side effectsto the drugs which resembled those that followedamethopterin, but none of these patients showedimpaired phenylalanine tolerance (Patients 2, 10,13-16).

Within Table I the patients are arrangedroughly in order of decreasing tumor response totherapy. No correlation was found between thisresponse and the degree of impairment of phenyl-alanine tolerance. Furthermore, no correlationwas observed between toxicity from amethopterinand phenylalanine tolerance.

The urine of Patients 1, 3-6, 8, 9, 11 and 12was tested for phenylpyruvic acid before and afteramethopterin. In those patients who received theDL-phenylalanine mixture, there was detectablephenylpyruvic acid in the urine, but the excretiondid not rise after therapy. In the patients who

received L-phenylalanine, no phenylpyrtuvic acidwas detected before or after therapy.

Assay of phenylalanine hydroxylase in various

tissues. Phenylalanine hydroxylase assays were

performed on the following numbers of separatetissue specimens: 1 human choriocarcinoma (2strains) grown as heterologous transplants inthe cheek pouch of the hamster (18), humanchoriocarcinoma excised at surgery; 1 humanterm placenta; 1 human placenta aborted after 18weeks' gestation; 3 rat placenta approximately 3days from term; 4 rat mammary gland duringpregnancy; 2 rat mammarygland during lactation;20 immature rat uterus after 6 days' stimulation byexogenous estrogen; 3 rat small intestine; 1 rabbitmarrow; 1 rabbit skin. These tissues were

chosen for assay because of their known sensitivityto the effects of folic antagonists. There was no

phenylalanine hydroxylase activity detected in any

of the specimens tested. Thus, liver is the onlytissue in which significant phenylalanine hydroxy-lase activity has been detected.

Effect of folic acid-deficient diet on cofactorcontent of liver. Table II presents the findingsof an experiment to determine the effect of folic

TABLE II

Effects of folic-deficient diets on the content of cofactor for phenylalanine hydroxylase in rat liver

Peripheral Folic acid CofactorWBC FIGLU* content of content

Rat On diet Weight count excretion livert of livert

days g jimoles/day jIg/g wet liver U/g wet liver

Standard diet + folic acid1 51 201 4,700 10.0 0.672 51 188 7.5 0.733 51 198 9.5 0.894 51 211 5,200 8.6 0.815 51 188 10.0 0.716 51 175 4,300 8.0 0.84

Standard diet, deficient7 51 183 2,400 0.48 0.368 51 124 3,100 0.47 0.509 51 165 1,500 0.50 0.57

10 51 123 2,200 0.30 0.5511 51 116 2,600 0.49 0.49

High-fat diet, deficient12 120 180 133 0.56 0.7713 120 170 113 0.46 0.8014 120 180 113 0.49 0.69

* Formiminoglutamic acid; normal urinary excretion for the rat is less than 5 jsmoles/day.t Measured as citrovorum factor after autolysis to convert "prefolic" and folic acids to citrovorum factor (13).1 One unit is the amount of cofactor activity leading to the formation of 1 jumole tyrosine in a standard assay con-

taining phenylalanine, phenylalanine hydroxylase, and "sheep enzyme" (14).

1 747

THEODOREL. GOODERIENDAND SEYMOURKAUFMAN

acid deficiency on the level of cofactor for p)henyl-alanine hydroxylase in rat liver. There was a36 per cent reduction in the mean level of co-factor activity in only those rats which demon-strated inanition. The rats fed a high-fat diet toprevent inanition failed to show any reduction ofcofactor content.

The pilot study using mouse leukemia failed toyield evidence for an effect of various dietaryphenylalanine: tyrosine ratios on prolongation ofsurvival time by amethopterin.

DISCUSSION

The complete phenylalanine hydroxylase systemrequires two enzymes (named for their conveni-ent animal sources), TPNH, oxygen, phenyl-alanine, and a cofactor (Figure 1). The exactstructure of the cofactor is not known, but it hasmany of the chemical characteristics of a pteridine.Several synthetic tetrahydro pteridines includingtetrahydro folic acid can substitute for the naturalcofactor in vitro.

The second enzyme ("sheep liver enzyme"),catalyzes the reduction of the cofactor by TPNH(4). In the presence of reduced cofactor andoxygen, phenylalanine is converted to tyrosine ina reaction catalyzed by phenylalanine hydroxylase(i"rat liver enzyme"). Folic antagonists arethought to block the enzymatic reduction of co-factor in a manner analogous to their inhibition ofthe activation of folic acid coenzymes by folic acidreductase (2). The experiments described heresuggest that the scheme of phenylalanine hydroxy-lation derived from in vitro investigation appliesas well to the human.

The inhibition of phenylalanine metabolism byamethopterin therapy is incomplete. There is,possibly, nonenzymatic reduction of the cofactor,unaffected by amethopterin, which permits phenyl-alanine metabolism to proceed at a reduced rateafter inhibition of sheep liver enzyme. The ob-served inhibition of phenylalanine metabolism isof the same order of magnitude as the impairedphenylalanine metabolism found by Hsia andPaine in heterozygotes for phenylpyruvic oligo-phrenia (17) but far less than that seen in thehomozygote. It should be emphasized thatwhereas amethopterin blocks sheep liver enzyme,patients with phenylpyruvic oligophrenia lack the

hydroxylating enzyme itself (rat liver enzyme).The "sheep" enzyme and cofactor are present innormal amounts in the liver of patients with thisdisease (4).

There is little resemblance in the toxic state fol-lowing folic antagonists to the clinical picture ofphenylpyruvic oligophrenia. No evidence wasobtained for characteristic mental or dermatologicchanges. In addition, the patients with drug-in-duced impairment of phenylalanine metabolismdid not excrete phenylpyruvic acid in the urine.This is consistent with the finding of Armstrongand Low, who failed to detect phenylpyruvic acidin the urine unless the serum phenylalanine levelexceeded 0.91 Mmoles per ml (19).

The results of our study, with regard to thespeed of onset and duration of changes in phenyl-alanine tolerance after amethopterin, are in ac-cordance with those found by Condit for folic re-ductase activity (20). He measured the excre-tion of citrovorum factor after a test dose of folicacid and found an effect within 3 days of an in-jection of amethopterin. The effect was evidentfor 2 to 10 weeks after the last dose of amethop-terin, depending on the dosage schedule. Char-ache, Condit and Humphreys (21) found per-sistence of amethopterin, detected by direct assayof human liver, 3 months after the last dose ofamethopterin.

Three experiments were performed to assessthe importance of the "phenylalanine effect" ofamethopterin in chemotherapy, and all three werenegative. First, there was no correlation betweenthe degree of impaired phenylalanine toleranceand the clinical response to treatment. Second,there was no effect from feeding additional phenyl-alanine or omitting dietary tyrosine on amethop-terin therapy in mouse leukemia. These dietaryalterations were designed to accentuate the prob-able metabolic consequences of inhibition of phenyl-alanine hydroxylation. Third, there was no evi-dence for the presence of phenylalanine hydroxy-lase in tumor tissue or in normal tissues (otherthan liver) which are known to be affected by folicantagonists.

The folic-deficient diet failed to reduce the he-patic level of cofactor for phenylalanine hydroxy-lase except when inanition was also induced. Thefolic acid level was consistently reduced. This

1748

PHENYLALANINEMETABOLISMAND FOLIC ACID ANTAGONISTS

may indicate that the cofactor is not derived fromfolic acid, or that the cofactor is preferentially re-tained in the liver under conditions of folic de-ficiency, or both.

Hydroxylation is the first step in the sequenceof reactions that convert phenylalanine to CO2via tyrosine and homogentisic acid. The reactionmay be considered, therefore, part of a degrada-tive pathway for excessive phenylalanine, as wellas a part of several synthetic pathways which in-clude tyrosine. Folic antagonists also inhibit thedegradation of the amino acid histidine by block-ing the metabolism of formiminoglutamic acid(FIGLU) (22). The excretion of this inter-mediate has been used as a test of folic deficiencyor folic antagonism. There is no evidence thatFIGLU excretion correlates with the therapeuticeffects of amethopterin. Thus, three biochemicaltests are available for the assessment of folic an-tagonism in mammals: the failure to excrete nor-mal amounts of citrovorum factor after a test doseof folic acid, the excretion of abnormally largeamounts of formiminoglutamic acid after the feed-ing of histidine, and the impairment of phenylala-nine metabolism during an oral tolerance test.

SUMMARY

1. Amethopterin therapy interferes with phenyl-alanine metabolism in man, as measured by theoral phenyalanine tolerance test.

2. No evidence could be obtained for a rela-tionship between impaired phenylalanine metabo-lism and the therapeutic or toxic effects of ame-thopterin in man or in the mouse.

3. Pronounced folic deficiency in the rat issometimes accompanied by a slight reduction inthe level of the cofactor for phenylalanine hydroxy-lase.

4. These findings have been discussed with ref-erence to the proposed scheme of phenylalaninemetabolism and the enzymatic effects of amethop-terin in man.

ACKNOWLEDGMENT

The authors are indebted to Dr. Roy Hertz and Dr.Delbert M. Bergenstal for their many helpful suggestions,to Dr. Milton Silverman for supplying some of the folic-deficient rats, to Dr. Abraham Goldin and Mr. StewartHumphreys for assistance in the mouse-leukemia experi-nlent, to Dr, Patil Condit and Dr. William Harrington

for reading the manuscript, and to Mr. Richard Funkand Mr. Frederick Dhyse for their valuable technicalassistance.

REFERENCES

1. Goodfriend, T. L. Impaired conversion of phenylala-nine to tyrosine after folic-antagonist therapy inwomen. Fed. Proc. 1960, 19, 6.

2. Futterman, S., and Silverman, M. The "inactivation"of folic acid by liver. J. biol. Chem. 1957, 224,31.

3. Skipper, H. E., Mitchell, J. H., Jr., and Bennett, L.L., Jr. Inhibition of nucleic acid synthesis by folicacid antagonists. Cancer Res. 1950, 10, 510.

4. Kaufman, S. Phenylalanine hydroxylation cofactorin phenylketonuria. Science 1958, 128, 1506.

5. Kaufman, S., and Levenberg, B. Further studies onthe phenylalanine-hydroxylation cofactor. J. biol.Chem. 1959, 234, 2683.

6. Hertz, R., Bergenstal, D. M., Lipsett, M. B., Price,E. B., and Hilbish, T. F. Chemotherapy of chorio-carcinoma and related trophoblastic tumors inwomen. J. Amer. med. Ass. 1958, 168, 845.

7. Hertz, R. Suppression of human choriocarcinomamaintained in the hamster cheek-pouch by extractsand alkaloids of Vinca rosea. Proc. Soc. exp.Biol. (N. Y.) 1960, 105, 281.

8. La Du, B. N., and Michael, P. J. An enzymaticspectrophotometric method for the determinationof phenylalanine in blood. J. Lab. clin. Med. 1960,55, 491.

9. Udenfriend, S., and Cooper, J. R. The chemical es-timation of tyrosine and tyramine. J. biol. Chem.1952, 196, 227.

10. Folling, A., and Sydnes, S. A method for the esti-mation of phenylpyruvic acid in urine with someexamples of its use in dietary treatment of phenyl-pyruvic oligophrenia. Scand. J. cdin. Lab. Invest.1958, 10, 355.

11. Kaufman, S. The enzymatic conversion of phenyl-alanine to tyrosine. J. biol. Chem. 1957, 226, 511.

12. Silverman, M., and Pitney, A. J. Dietary methionineand the excretion of formiminoglutamic acid by therat. J. biol. Chem. 1958, 233, 1179.

13. Silverman, M., and Gardiner, R. C. Labile citro-vorum factor in urine. J. Bact. 1956, 71, 433.

14. Kaufman, S. A new cofactor required for the en-zymatic conversion of phenylalanine to tyrosine.J. biol. Chem. 1958, 230, 931.

15. Franklin, A. L., Stokstad, E. L. R., Belt, M., andJukes, T. H. Biochemical experiments with asynthetic preparation having an action antagonisticto that of pteroylglutamic acid. J. biol. Chem.1947, 169, 427.

16. Goldin, A., Venditti, J., Humphreys, S. R., andMantel, N. Quantitative evaluation of chemo-therapeutic agents against advanced leukemia inmice. J. nat. Cancer Inst. 1958, 21, 495.

1 749

THEODOREL. GOODFRIENDAND SEYMOURKAUFMAN

17. Hsia, D. Y., and Paine, R. S. Phenylketonuria; de-tection of the heterozygous carrier. J. ment. Def.Res. 1957, 1, 53.

18. Hertz, R. Choriocarcinoma of women maintained inserial passage in hamster and rat. Proc. Soc. exp.

Biol. (N. Y.) 1959, 102, 77.19. Armstrong, M. D., and Low, N. L. Studies on phenyl-

ketonuria. VIII. Relation between age, serum

phenylalanine level, and phenylpyruvic acid ex-

cretion. Proc. Soc. exp. Biol. (N. Y.) 1957, 94,142.

20. Condit, P. T. Studies on the folic acid vitamins.III. The duration of the effects of folic acid antag-onists in man. Cancer 1960, 13, 229.

21. Charache, S., Condit, P. T., and Humphreys, S.Studies on the folic acid vitamins. IV. The per-

sistence of amethopterin in mammalian tissues.Cancer 1960, 13, 236.

22. Hiatt, H. H., Goldstein, M., and Tabor, H. Urinaryexcretion of formiminoglutamic acid by humansubjects after antifolic acid therapy. J. clin. In-vest. 1958, 37, 829.

1750