AN ENZYMATIC SPECTROPIIOTOMETRIC METIIOD FOR THE
   DETERMINATION OF PHENYLALANINE 1N BLOOD

BERT N. LA Du, M.D., PH.D., AND PATRICIA J. MICHAEL, B.A.
           BETHESDA, MD.

T HERE is a need for a simple rapid method to measure phenylalanine in
   serum or plasma, not only as an aid in the diagnosis of phenylketonuria, but
also in the evaluation of the effectiveness of a diet low in phenylalanine in the
treatment of patients with this condition. The method most commonly used to
measure phenylalanine in plasma is that of Udenfriend and Cooper,* in which
plasma is deproteinized and treated with Xtreptococcus faecalis t,o decarboxylat,c
r,-phenylalanine to phenylethylamine.  The latter is then determined colorimet-
rically after reaction with methyl orange.  Although this method has been used
successfully by many workers,2-6 it requires relatively large amount,s of plasma
and takes several hours to perform each set of analyses. Furthermore, extremel)
careful t,echnique is necessary to avoid an appreciable and variable blank from
methyl orange.

  Another method for the determination of serum phenylalanine is that of
Berry,7 by paper chromatography. This met,hod, although simple, has limitations
in quantitative analysis inherent in any method which depends upon estimat,ion
of the intensity of colored spots on paper chromatograms.

< ====\-
\        CH,CHCOOH
   y I       c \ =-,>H~c-coOII
           NH,            CI

                 Snake venom
                r,-amino acid oxidaw
   L-phenylalanine              > phenylpyruvic acid

<-
==$
\            -cc=0
     d  G
     `B'
     0' `0
< ======\ -1    I
\  y-7-c -c=o

          H

          Arsrnate borate > ISor;lte-mol complex of phenylpyruvie acid

  From the National Institute of Arthritis and Metabolic Diseases, National Institutes of
Health, U. S. Public Health Service, Bethesda, &Id.
   Received for publication June 22, 1959.

                              491


492                   LA J1U ANI) MICHAEL

J. Lab. & Clin. Merl.
    March. 1960

A simple ral)id m&hod is described which permits the analysis of plasma
L-phenylalanine in 0.1 ml. of serum without deproteinization.  The method uti-
lizes L-amino acid osidase from snake venom to oxidize I,-phenylalanine to
phenylpyruvic acid. In the presence of arsenate and borate ions the resulting
Lu-keto acid is rapidly converted to an enol-borate complex which has a high ab-
sorption in ultraviolet light. Catalase is added to protect the ol-keto acid I'rom
peroxide formed in the oxidative deamination by L-amino acid osidase.

Enol-borate compleses have been employed by Knox and Pitt" to determine
p-hydroxyphcnylpyruvic acid osidase activity in mammalian liver preparations.
More recently these complescs have been used to measure the act,ivity of several
cnzgmes which either form or break down the aromatic a-kcto acids related to
phenylalanine, tyrosine, tryptophan, and histidine."

&IETHOD

   Becqents.-
  1. Phosphate buffer :  0.231 sotliurn phosphate buffer, pH ti.5.
  2. Arsenate-phosphate buft'er :  %.OM sodium arsenate was dissolved in the O.%hl
phosphate buffer, and the final pH was readjusted to 6.5 with tlilute HCl.
  3. Borate-arsenate reagent:  l.OM borate, dissolved in 2.OM arsenate, plI 6.3, (61.8
Gm. boric acid + 624 Gm. sodium arsenate [Na,HAsO,. iH,O]) was adjusted to pH 6.5
with HCl and malle up to 1 I,.

  4. Snake venom Id-amino acid oxidase (the venom of Crotctlus atlnmanttus") : a suspeI,-
sion of dried venom in water was nrade which contained 10 mg. per milliliter.  This was cen-
trifuged, and the clear supernatant solution was removed and used. The enzyme solution
is kept at 0 to 5" C. and remains active wit.h little loss of activit,y for several days. The
dry venom is stored in the refrigerator and maintains high enzyme activity for several
months.

  The activity of the L-amino acitl oxidase preparation can be assayed hy using 0.1 ml.
of the standard phenylalanine solution in place of serum in the directions given belolv.
   5. Catalase:  crystalline beef liver catalase,t diluted 1:s \yith 0.2X phosphate buffer,
pH 6.5.  The enzyme solution is kept at 0 to 5" C. and can be used for several rveeks.
  6. Standard phenylalanine solution:  L-phenylalanine, 1 pill per milliliter (0.1 ml. con-
tains 16.5 ~g).
   7. Blood serum:  Blood is drawn ant1 allowed to clot.  (Some samples of heparinized
plasma develop turbidity during analysis which makes them unsuitable for this determination.)

Prooedwre.-

  Three 1.2 ml. quartz Beckman cuvettes with a 1 cm. light path are used for the
determinations.  The additions to each of the cuvettes are as followfi:

                                  01           El            IL
   PO, buffer, 0.2M, pH 6.5::           0.39         0.39      0.:;9
   Arsenate-phosphate, pH 6.5:         0.5                       -
    1M borate in 2M arsenate:      OZ         0.5          0.5
   Catalase, 1:5 dilution:                        0.01         0.01
   Venom, 10 mg. per milliliter:         0.10         0.10         0.10
     (Serum added later.)

  The contents are mixed and the duplicate experimental cuvettes, E, and E,, are read
against the control cuvette, C? at 308 mp, in the Beckman spectrophotometer.$  If the addi-

  *From the Ross hllen Rel>til<l Institute. Silver Springs, Fla.
   ?From the Worthington Biochemical Corporation, Freehold, PIT. J.
  fIf more tllan 0.1 ml. of wr11111 is to he dllccl, tire volunle of phosphate buffer n,ust he
reduced proportionately.

_Model DU with ultraviolet attachment.


Volu~ 55 ~~EARURERIE:Nl' OF PHENYLALANINE IS PLASMA AND BLOOI)
Number 3                                          493

tious have been madc correctly, this reading shoulcl be approximately zero. Then 0.1 ml. of
serum is added to cuvettes C,, E,, and E?, mixed, and readings are taken at 1 or 2 minute
intervals for 10 minutes at 308 m,u, to he certain that the enzyme activity is adequate, and
that the reaction is essentially complete, as indicated by finding no further change in the
optical density reading. Final readings are made at 308, 330, and 350 mn at this time.

Calculatiom-

  Readings are taken at 330 ant1 330 mp, as well as at 305 rn~, to correct for tho ab-
sorption of the keto acids derived from tyrosine and tryptophan, since the latter 2 amino
acids are also oxidized by the venom L-amino acid oxidase.  The correction applied is based
upon the relative absorption of the enol-borate complexes of these keto acids at 308, 330,
and 350 mp:

                          RELATIVE OPTICAL DENSITY READINGS AT
                    DIFFERENT WAVELENGTHS (Illa)
  KETO ACIDS OF               308         330,         350
X = tyrosine                  X        0.60X       0.06X
Y = phenylalanine               Y         O.lOY
Z = tryptophan               0.852        1.412

Let the reading at 308 mp = A; 330 mp = B ; 330 mp = C

Then: A = X + Y + 0.852
    B = 0.6X + O.lY + 1.412
    c = 0.06X + z

      Solving these equations gives:
            X = 2.38B - 0.238A - 3.16C                 (1)
             Y = 1.23A - 2.26B + 2.15C
             Z = 1.192 - O.lJB + 0.014,4 = C - 0.06X    Ii,'
The absorption at 308 rnp caused by phenylpyruvie acid is giveu by solving for Y iu
equation (2) :

Y

__ x 10 = ng phenylalanine per milliliter of serum (if 0.1 ml. of serum were
0.031
analyzed).  One microgram phenylalanine under these conditions of 1.1 ml., etc., at
308 ma reads 0.031 O.D. units.                             x

  By the same means, solving for X in equation (1) ant1 dividing: -om x 10 = pg
tyrosine per milliliter of plasma (1 pg tyrosine under these conditions reads 0.045 at 308

mp); and, similarly, using equation (3):      x 10 = fig tryptophan per milliliter of
serum (1 pg of tryptophan reads 0.033 at 350 mp).

RESULTS 89D DISCUSSIOS

  Specificity.-The L-amino acid oxidase of snake ccnom catalyzes the oxida-
tion of a number of amino acids.  However, only 4 of these are present in ap-
preciable quantities in serum which yields wketo acids with highly absorbing
enol-borate complexes; these are phenylalanine, tyrosine, trgptophan, and his-
tidine.  Fortunately, histidinc is not oxidized under the esperimcntal condit,ions
described above, and the relative contribution of each of the other 3 amino acids
can easily be determined from the absorption of the combined complexes at 3
different mawlengths. Therefore, this method permits the simultaneous deter-
mination of phcnylalaninc, tyrosinc, and tryptophan in swum. If t,he value of
phenylalanine alone is desired the correction for absorption caused by the keto
acids of tyrosinc and tryptophan would appear to be a disadvantage of the
method.  This is partly compensated for by the commercial availability of the


494

J. Lab. & Clin. P&l.
    Mxch, 1960

cnzyuc, its st,ability, and its high actirity.  E'urtherrnorc, the ease with which
both plren~lalanine and tyrosine can be measured is of special interest in some
studies on phcnylkctonuria; for esanrplc, in the detection of the heterozygous
carrier of this trait by the ratio between the levels of phenylalanine and tyro-
sine in t,he blood after the administration of phenylalanine as a tolerance tcst.Z

  The wide range of substrates which can be used with this enzyme has rnadc
it, possible to extend this method to rncasure a number of other aromatic amino
acid analogucs and antimetabolites, such as P-fluorophcn~lala~linc, rn-tyrosine,
monoiodotyrosinc, rnononitrotyrosine, and /3-2-thienylalaninc. These compounds
arc also rapidly oxidized by the L-amino acid osidase of snake vcnonr, and the
resulting kcto acid products can be measured as enol-borate conlplescs.lU

  1.000 -
         0 PHENYLALANINE ALONE
i   .800 -  . RECOVERY FROM PLASMA
A
r'
2   .600 -

s
E   .400 -
2
8

                           10       20       30
                       MICROGRAMS OF PHENYLALANINE
Fig. l.--C;raph of the optical density at 308 mp versus concentration of phenylalanine

n'lrrw determined in stxml;rr,l solutions and when adiled to normal plasma.  The linearity
between optical densi ly  and ghenylalaninc concentration is demonstrated.

  Reprotl~ucibilif~ mcl Llccur,clcy.-Tlle method described is well suited to

measuring the elevated blood level of pherq-lalaninc found in untreated phenyl-
kctonuric individuals and to following the effectirencss of a diet low in phenylal-
anine in lowering the sermn level in these individuals.  The specificity of the
absorption spectrum ol' the phenylpyruvic acid enol-borate complex aud the
case and rapidity ol' the analysis pcrrnit a large nunll~r of detcrrninations to
be carried out within a short period of time.  Fig. 1 demonstrates the linear
relationship between phenylalaninc concentration and optical density change
between 2 and 30 pg analyzed alone and the recover; of phenylalanine added to
normal plasma in this range.  These results are in agreenwnt with the thco-
retie values expected. However, the precision of t,he analytic method is lower
in analysts of phenylalaninc in normal serum than when the values are elevated
as in serum of an individual with pllen~lketorlnria.  In nomial serum the con-
tributions of tyrosine and tr?ptophan to the readin # at 308 nip arc appreciable.
l<ven though duplicate pairs of analyses in a series of normal serum samples
analyzed on successive days rarely disagreed by- more than 7 per cent, larger
aliquots of serum ~vo~ld be desirable if the method were to be used to measure,
with high accarac~~, the endogenous I)llcnPlalaninc, tyrosinc, and tryptophan
lcvcls.  Attempts to carry out the analyses directly 011 0.2 or 0.3 ml. of serulll
have been only partI)- successful, since a turbidity often OCCUIX For this reason,


Volume 55 MEASUREMENT OF PHENYLALANINE IS PLASMA AND BLOOD
Number 3

495

several methods were tested to deprotcinize the sample before analysis.
Preliminary experiments have been made using perchloric acid precipitation
followed by neutralization and the removal of KClO,, as described by Segal,
Blair, and Wyngaarden,`l before enzymatic assay of blood pyruvate. This
procedure for deproteinization appears to be satisfactory in eliminat,ing the
turbidity problem, and the solution which results can be used directly in the
enzymatic assay.

  Phenylnlanine Levels in Normal and I'henylketonuric I?ldividuuls.-The
serum phenylalaninc levels of 30 adult blood bank donors in a nonfasting con-
dition were found to have an average value of 1.55 mg. per cent (range, 0.84
to 2.6`4 111" b. per cent).  It is probable that the level in normal subjects in a fast-
ing condition is somewhat lower than these values.  In a smaller number of
analyses in laboratory workers in a fastin g condition, the group average is
slightly lower than the above value.

  The level of phenylalanine in a group of 10 phenylketonuric individuals on
a regular diet was found to have an average value of 36.8 mg. per cent (range,
26.2 to 59.1 mg. per cent). These values in norrnal and in phenylkctonuric
people are in good agreement wit,h those in the literature73 l"-"* obtained by
other methods.  The wide difference in the levels of norrnal and pl~enylkctonnric
individuals makes the analysis of serum phenylalaninc a valuable diagnostic
procedure.

  A simple enzymat,ic spectrophotometric method has been described for the
quantitative determination of phenylalanine in serum.  The method is based
upon the m(~asuremcnt of the absorption of the enol-borate complex of phcnyl-
pyruvic acid generated enzymatically from phenylalanine by L-amino acid osi-
dase of snake venom. For elevated serum levels of phenylalanine the method
is rapid, specific>, and precise.  It should be useful as a confirmatory test in the
diagnosis of suspected phenylkctonuria and in the evaluation of the effect,ivcness
of a diet low in phrnylalanint~.  The values obtained by this method agree well
with those in the literature obtained by other techniques.  Tyrosine and t,r,vy-
tophan arc also determined by this method, and a suitable modification of the
method is described which should allow an accurate estimation of these amino
acids and phcnylalaninc in normal serum.

REFERENCES

1. Udenfriend, S., and Cooper, J. R.:  Assay of n-Phenylalanine as PIlen)-letb~lami~te
   After Enzymatic Deearboxylat.ion; Application to Isotopic Stutlies, J. Biol. Chern.
    203: 953, 1953.

J. Meister, A., Udenfriend, S., and Bessman, S. P.: Diminished Phenylketonuria in Pheny-l-
   pyruvic Oligophrenia After Administration of r,-Glutamine, L-Glutamate or 1,.
   Asparagine, J. Clin. Invest. 35: 619, 1956.
3. Hsia, D. Y.-Y., Driscoll, K. W., Troll, W., and Knox, W. E.: Detection by Phenylalanine
   Tolerance Tests of Heterozygous Carriers of Phenylketonuria, Nature, London 178:
    1239, lx%.

1. Hsia, D. Y.-Y., and Paine, R. S.: Pheiiylketonuria: Detection of the Heterozygous
   Carrier, J. Ment. Deficiency Res. 1: 53, 1957.
5. Hsia, D. Y.-Y.:  Phenylketonuria: The Phenylalauine-Tyrosine Ratio in the Detection
   of the Hcterozygous Carrier, J. Ment. Deficiency Res. 2: 8, 1955.


             Y
LA DU AND MICRAEL

ti. Knox, W. E., anrl Xessinger, E. C.: The Detection in the Heterozygote of the MeTa-
    bolic Effect. of the Recessive Gene for Phenylketonuria, Am. J. Human Genet. 10:
    53, 1958.
7. Berry, H. Ii.: Paper Chromatographic Method for Estimation of Phenylalanine, Proc.
    Sot. &per. Biol. & Med. 95: 71, 1957.
8. Knox, W. E., and Pitt, B. M.: Enzymatic Catalysis of the Keto-Enol Tautomcrization
    of Phcnylpvruvie Acids, J. Biol. Chem. 225: 675, 1957.
9. Lin, E. C. C., Pitt,.B. X., Given, M., and Knox, W. I<.: The Assay of Aromatic Amino
   Acid Transammations and Keto Acid Oxidation by the Enol Borate-Tautorrierasc
    Method, J. Biol. Chem. 233: 668, 1958.
10. La Du, B. N., and Michael, P. J.: -4 New Sssay for Analogues and iintimetabolitea
    of Tyrosine and Phenylalanine. In preparation.
11. Segal, S., Blair, A. E., and Wvngaarden, J. B.:  An Enzymatic Spectrophotornetric
   Method for the Determinaiion of Pyruvic Acid in Blood, J. LAB. & CLIN. MEL).
    48: 137, 1956.
1". Hsia, D. P.-Y., and Driscoll, K. W.: Detection of the Hetorozygous Carriers of Phcnyl-
    ketonuria, Lancet 2: 1337, 1956.

13. Jervis, G. A., Block, R. J., Bolling, D., and Kanze, E.:  Chemical and llletabolic Studies
   on Phenylalanme. II. The Phenylalanine Content of the Blood and Spinal E'luid
    in Phenylpyruvic Oligophrenia, J. Biol. Chem. 134: 105, 1940.
14. Bickel, H., Gerrard, J., and Hickmans, E. M.: The Influence of Phenylalanine Intake
   on the Chemistry and Behavior of a Phcnylketonuric Child, Acta paediat. 43:
     64, 1954.