2. Gennari, C., Galli, M.

, and Montagnani,

M., Urinary cyclic adenosine monophosphate

in young adults and elderly subjects. J. Clin.
Pat hot. 29,69 (1976).
3. Madsen, N. S.,Badawi, I., and Skovsted,
L., A simple competitive protein-binding

assay for adenosine-3,5-monophosphate in

plasma and urine. Acta Endocri not. 81, 208
(1975).

4. Latner, A. L., and Pridhoe, K., A simplified competitive protein-binding assay for

adenosine 3,5-monophosphate in plasma.

Clin. Chim. Acta 48, 353 (1973).
5. Neelson, F. A., Drezner, M. K., Birch, B.
M., and Lebovitz, H. A., Urinary cyclic
adenosine monophosphate as an aid in the
diagnosis of hyperparathyroidism. Lancet i,
631 (1973).
6. Schmidt-Gayk, H., and Roher, H. D.,

Urinary excretion of cyclic adenosinemonophosphate in the detection and diagnosis of

primary hyperparathyroidism. Surg. Gynecot. Obstet. 137, 439 (1973).
7. Murad, F., and Pak, Ch. Y. C., Urinary
excretion of adenosine 3,5-monophosphate
and guanosine 3,5-monophosphate. N. Engi.
J. Med. 286, 1382 (1972).
8. Taylor, A. L.,Davis, B. B., Pawlson, G. L.,
et al., Factors influencing the urinary excretion of 3,5-adenosine monophosphate in
humans. J. Clin. Endocrinol.
30, 316
(1970).
Harry

Husdan

Rudolph Vogi
Dimitrios
Abraham
Departments
of Medicine
Clinical Biochemistry
Toronto

Western

Hospital

Oreopoulos

Rapoport
and
and

University of Toronto,
Toronto, Ontario, Canada M5T 2S8

Amino Acid Sequence in Bovine

Serum Albumin
To the Editor:
In this journal, Peters reviewed recent
progress in the understanding of the
structure and biosynthesis of albumin
(1). Therein, the complete amino acid
sequence of bovine serum albumin as
reported by Brown (2) is reproduced.
The sequences of nine peptides obtained
during an investigation
of tryptic peptides from bovine serum albumin enhancing the effect of bradykinin
(3, 4)
can generally
be fitted
into this sequence, but a few discrepancies
exist.

In the complete amino acid sequence

of bovine serum albumin (2) the fragment Asp - Leu - Gly - Glu - Glu - HisPhe-Lys- (residues 13-20) is identical in
amino acid composition with our peptide C-V, viz., Asp-(G1u2,Gly,Leu,Phe,
His)-Lys. Because this peptide was
formed by the action of trypsin, the Cterminal amino acid is lysine. The residue of a basic amino acid (lysine) at po-

sition 12 also corresponds with the specific action of trypsin. We determined

aspartic acidasthe N-terminal amino

agreement with the proposed sequence

of the fragment 180-184, as is also with
theresidue
ofa basicamino acidatpo-

acid of peptide C-V. The isoelectric

sition 179 (lysine).

point which can be calculated (5) from

fragment 13-20 (p1
4.2) fits with the
mobilities of peptide C-V by high-voltage electrophoresis at pH 4.00 and 7.00.
Therefore,

as published,

the aspartyl

and glutamyl residues must be present

as the free acids. The analytical results
for this fragment are in full agreement
with the results of Shearer et al. (6).
The fragment -Leu-Val-Asn-GluLeu-Thr-Glu-Phe-Ala-Lys(residues
42-5 1) is identical in amino acid composition with our peptide B-Vu, viz.,
Leu-(Asx,Thr,Glx2,Ala,Val,Leu,Phe)
Lys. Because this peptide was formed by
the action of trypsin, the C-terminal
amino

acid is lysine.

The

residueof a

basic amino acid at position 41 (lysine)

also corresponds with the specific action
of trypsin. We determined leucine as the
N-terminal amino acid of peptide B-WI.
The calculated isoelectric point of
fragment 42-51 is p1
4.2. However,
our peptide B-VII was almost neutral at
pH 7.00.Therefore, only one of the three
w-carboxylgroups
can be free. This
might be either the aspartyl residue or
one of the two glutamyl residues. If the

aspartyl residue 44 is an amide as published, then also one of the two giutamyl
residues (45 or 48) must be amidated. In
that case, the recalculated
isoelectric
point (p1
6.85) is in agreement with
our results (p1 - 7.0).

The fragment -Ala-Glu-Phe-ValGlu-Val-Thr-Lys(residues 224-231) is

in amino acid composition identical with
our
peptide
B-Ill,
viz.,
Ala(Thr,Giu,Gln,Va12,Phe)-Lys.
The residues of a basic amino acid at positions
223 and 231 (lysine) correspond with the
specific action of trypsin. We determined alanine as the N-terminal amino
acid of peptide B-Ill. The calculated
isoelectric point of fragment 224-231 is

p1

4.2. However, we found peptide

B-Ill to be almost neutral at pH 7.00.

Therefore, one of the two glutamyl
residues present at positions 225 and 228
must be amidated.
The published sequence of fragment
232-238, viz., -Leu-Val-Thr-Asp-LeuThr-Lys-, is in full agreement with our
analytical results (amino acid composition, isoelectric point, N- and C-terminal amino acid) of peptide B-IV, viz.,
Leu-(Asp,Thr2,Val,Leu)-Lys.
The same applies to the published

sequence of fragment
-Ala-Asp-Leu-Ala-Lyspeptide

256-260, viz.,
in relation to

C-Il,viz., Ala-(Asp,Ala,Leu)-

Lys. The residue of a basic amino acid

(arginine)
at position 225 is in agreement with our results.
Because peptide C-IV enhances the
effect of bradykinin, its amino acid sequence was determined by phenylisothiocyanide

degradation

as well as by

The fragment -Thr-Cys-Val-Ala- massspectrometry. All our data (Thr-

Asp-Glu-Ser-His-Ala-Gly-Cys-Glu-Lys(residues 52-64) is identical in amino

acid composition with our peptide B-V,
viz., Thr-(Asx,
Ser, G1x2, Gly, Ala2,
Cys2,Val,His)-Lys.
Our results, namely

the amino acid residues at position 51

Lys, 52 Thr, and 64 Lys, and a disulfide
bridge in this fragment, agree with the
published structure. The calculated
isoelectric point of fragment 52-64 is p1
4.2. This value is lower than is to be
expected from the mobilities of our
peptide B-V on high-voltage electrophoresis at pH 4.00 and 7.00.Therefore,
one of the carboxylgroups must be amidated. If residue 56 is Asn instead of
Asp, the calculated isoelectric point (p1
5.4) of the modified fragment 52-64
corresponds with that of peptide B-V.
One fragment of the published se-

quence,viz., -Glu-lle-Ala-Arg- (residues

140-143)

is in good agreement

with our

peptide D-IV: Glu-(Ala,Ile)-Arg.

However, because D-IV is a tryptic peptide
the amino acid residue in position 139
must be lysine or arginine instead of
tyrosine as published.
Because peptide D-V enhances the
effect of bradykinin,
its amino acid sequence was determined
by phenylisothiocyanide
degradation as well as by
mass spectrometry.
All our data, viz.,
Ile-Glu-Thr-Met-Arg,
are in good

Pro-Val-Ser-Glu-Lys) are in good

agreement with the published sequence
of fragment 464-469 except for the interchange of amino acid residues serine
and glutamic acid at positions 467 and
468 in the published sequence. In my
opinion our sequence is correct. Firstly,
the sequence of peptide C-IV was determined by two different methods.
Secondly, we have isolated a tryptic
peptide from rabbit albumin enhancing
the effect of bradykinin with the same
sequence as C-IV and, finally, the sequence of the corresponding fragment
(467-472) of human serum albumin is
-Thr-Pro-Val-Ser-Asp-Lys(7), in
which the serine is also followed by a
dicarboxylic amino acid.
In summary, the following corrections
of the sequence determined by Brown
are suggested:
Residue 45 or 48 Gin instead of Glu,
Residue 56 Asn instead of Asp,
Residue 139 either
stead of Tyr,

Lys or Arg

in-

Residue 225 or 228 Gln instead of

Glu,
Residue 467 Ser instead of Glu and
residue 468 Glu instead of Ser.
References
I. Peters, T., Jr., Serum albumin:

Recent

progress in the understanding of its structure

CLINICALCHEMISTRY, Vol. 23, No. 7, 1977

1361

and biosynthesis. Clin. Chem. 23, 5 (1977).

Review.

2. Brown, J. R., Structure of bovine serum

albumin. Fed. Proc. Fed. Am. Soc. Exp. Biol.
34, 591 (1975). Abstract 2105.
3. Weijers, R. N. M., Thesis, University of
Utrecht, The Netherlands, 1971.
4. Weijers, R., Hagel, P., Das, B. ., and van
der Meer, C., Tryptic
peptides from rabbit
and bovine albumin enhancing the effect of
bradykinin. Biochim. Biophys. Acta 279,331
(1972).
5. Segel, I. H., Biochemical
Calculations,
John Wiley & Sons, Inc., New York, N. Y.,
1968, pp 139-149, 347-349.
6. Shearer, W. I., Bradshaw, R. A., Gurd, F.
R. N., and Peters, T., Jr., The amino acid sequence and copper (11)-binding properties of
peptide (1-24) of bovine serum albumin. J.
Biol. Chem. 242, 5451 (1967).
7. Meloun, B., Mor#{225}vek,
L,, and Kostka, V.,
Complete amino acid sequence of human
serum albumin. FEBS Lett. 58, 134 (1975).

R. N. M. Weijers
Laboratory
Onze Lieve
Amsterdam,

of Clinical Chemistry
Voouwe Gasthuis
The Netherlands

tenninal cyanogen bromide peptides of bevine plasma albumin. Arch. Biochem. Riophys. 153,627 (1972).
3. Meloun, B., Moravek, L., and Kostka,V.,
Complete amino acid sequence of human
serum albumin. FEBS Lett. 58, 134 (1975).
4. Spencer, E. M., Amino acid sequence of
the alanyl peptide from cyanogen bromide
cleavage

plasma

of bovine

albumin.

Arch.

Biochem. Biophys. 165,80 (1974).

5. Brown, J. R., personal communication,
March 1977.
6. Brown, J. R., personal communication,
September

The Mary Imogene

(affiliated

with

12

Peters, Jr.

Bassett

Hospital

Enzyme Immunoassayof

To the Editor:

careful work in both laboratories.

With regard to the minor corrections
Dr. Weijers suggested for the sequence
reported in this journal (1), only the last
one is clearly indicated.
1. Glu at residues 45 and 48 and Asp
at 56 were confirmed
by King and
Spencer (2). The sequence of human

albumin is also in agreement (3).

2. Tyr at residue 139 was confirmed
by Spencer (4) and by homology with

termining phenytoin and phenobarbital

in serum by adapting the EMIT system
(Syva, Palo Alto, Calif. 94304) to the
CentrifiChem (Union Carbide, Rye, N.
Y. 10580).
Both drugs can be assayed after one
dilution of standards and samples. All
reagents are prepared according to kit
instructionsexcept as stated below.

Controls: Syva-AED
Control
Procedure: Pipet 200 zl of calibrators
and sample serum into 12 X 75 mm
polypropylene tubes. To each of these
add 400 Ml of EMIT AED Buffer(pH
7.9). The diluted specimens and standards are then transferred
to specimen

specificity,
cleaves after Tyr13g.
3. Direct confirmation
of the absence
of amide residues at Glu2 and Glu2ss is

cups for the pipettor.

albumin

lacking, but Dr. Brown notes that his

peptide 224-23 1, corresponding to Dr.
Weijers peptide B-Ill, was acidic rather
than neutral (5).
4. Subsequent to his initial publication of the bovine albumin
sequence
Brown has agreed that residues 467-468

should be Ser-Glu rather than Glu-Ser

(6),

so that

this correction

should

made in the sequence reported

be

(1).

References
1. Peters, T., Jr.,Serum albumin: Recent
progress in the understanding
of its structure
and biosynthesis. Clin. Chem. 23, 5 (1977).
2. King, T. P., and Spencer, E. M., Amino
acid sequences of the amino and the carboxyl
1362

CLINICAL CHEMISTRY,

lOs, T

0.25

8 for phenytoin
5 for phenobarbital

The fmal printout is used for plotting.

The zero standard absorbance value
(zS.Ao)is subtracted from values of all
other standards and samples before
plotting. The net result (A
&t0) is
then plotted on log-log graph paper
used
in an
Epilepsy Foundation of America, AED
Quality Control Program (Dr. C. E.
Pippenger,
Columbia
Presbyterian
Medical Center, New York, N. Y.
10032), in which samples containing
weighed-in quantities of drugs were sent
to us for assay. Our results correlated
well with established values for the 36
samples received. The coefficients of
correlation were: phenytoin, 0.9937;
phenobarbital,

0.9980.
Morris

London

Dolores Sanabria
David Yau
Brookdale

Hosp. Med. Center

Brooklyn,

N. 1. 11212

Reagent: Syva-EMrr-Anti-epileptic Drug (AED) Dilantin assay kit

Syva-EMIT-AED
Phenobarbital
assay kit
Standards:
Syva-AED
Calibrators

(3). Dr. Brown (6)

found that trypsin, contrary to its usual

human

absorbance

Besides the two quality controls

in this study, we also participated

We wish to describe a method for de-

To the Editor:
It is reassuring that the peptides isolated
by Weijers in 1972 confirm in almost
every respect the corresponding portions
of the sequence of bovine albumin
publishe4 independently by Dr. Brown
in 1975. Such agreement bespeaks

340 nm, corresponding

setting

prqvided with the kit.

N. Y. 13326

Phenytoin and Phenobarbftalwith

the CentrIfugal Analyzer

Dr. Peters responds:

blank, Terminal, Operate Ab-

Auto

sorbance.

Columbia

University)

Cooperstown,

Analyzer Setting:

T0
16

1975.

Theodore

A buffer blank cup after the highest

standard
Specimensinduplicates

The EMIT Reagent kit consists of

Reagent A (Antibody/Substrate)

and

Reagent B (Enzyme). Pipet 20 il of

Reagent A manually

into the sample

wells(topwell)of the transferdisc.Add

600 Ml of Reagent B to 9.9 ml of buffer

and then transfer to the reagent dish on

the pipettor. This volume will be sufficient for at least 20 samples.
Pipettor setting:
Reagent volume, 350
Sample volume, 10
Total volume, 50
Last sample plug

Erroneously Low Creatine Klnase

Activity Measurementswith Use of

the Caiblochem CPK-MB Kit with

the BeckmanTR EnzymeAnalyzer

To the Editor:

Recent reports indicate that the auto-

mated Beckman Enzyme Activity Analyzer, system TR, may yield falsely
negative results for high-activity samples because of substrate depletion (1,
2).

We would like to report another distressing observation. The system TR in

the automated mode reports lower activities than when run on the continuous
mode at the lower end of the normal
range for creatine kinase (EC 2.7.3.2).
This is a particularly frequent occurrence when the Calbiochem CPK-MB
kit is used without dithiothreitol
activation to measure CPK-MM (3). Table
1 lists some typical findings with use of
this kit. The average continuous-mode

result is five times greater than the average

automated mode result, a signifiThe cupson thetraycanbearranged
cant difference.
as follows:
What

Standards

Vol. 23, No. 7, 1977

in duplicates

causes

this

difference?

probable cause is that the instrument

The

in