CLIN. CHEM. 41/6, 872-880 (1995)
#{149}
Laboratory
Management
and
Utilization
Multienzyme Control Serum (Seraclear-HE) Containing Human Enzymes from
Established Cell Lines and Other Sources. 1: Preparation and Properties
Akira
Eto,”3
Atsushi
Shild,’
Yasushi
Chikaura,2
Tetsuya
We have developed
a new multienzyme
control serum,
Seraclear-HE,
which was designed
to function not only
as an accuracy and precision control serum but also as
an intermethod
calibrator for unifying interlaboratory
clinical enzyme data in terms of reference
method values.
Seraclear-HE
contains as analytes the following enzymes
of human origin only: aspartate
aminotransferase
(AST,
EC 2.6.1.1) and lactate dehydrogenase
(LD, EC 1.1.1.27)
from erythrocytes;
alanine aminotransferase
(ALT, EC
2.6.1.2) from a hepatoma
cell line; alkaline phosphatase
(ALP, EC 3.1 .3.1)from an amnion cell line; creatine kinase
(CK, EC 2.7.3.2) from an embryo kidney cell line; y-glutamyltransferase
(GGT, EC 2.3.2.2) from a macrophage
cell line; and amylase (AMY, EC 3.2.1.1) from urine and
saliva. The seven partly purified enzymes
were lyophilized in partially delipidated
human serum containing
sucrose
(50 g/L), pyridoxal 5’-phosphate
(30 mmoVL),
and other stabilizers. The material is stable for at least 2
years at temperatures
1 0#{176}C.
For two concentrations
of
this preparation,
reference
method values (mainly International
Federation
of Clinical Chemistry
and Japan
Society of Clinical Chemistry) obtained at both 30#{176}C
and
37#{176}C
are assigned.
Indexing Terms:
va!ues/intermethod
activity/reference material/reference
comparison/calibration/quality contml
enzyme
Unification
of interlaboratory
clinical enzyme data
cannot be achieved by method standardization
alone,
but requires
stable and well-characterized
enzyme reference materials
(ERMs) carrying a certified Reference
Method value.4 Extensive
work has been carried out,
mainly in the European
Communities
(1, 2) and in the
US, to establish
reference
materials
for some routinely
measured
enzymes of diagnostic
interest.
The Community Bureau
of Reference
of the European
Communities, in particular,
has made available
Certified Reference Materials
(CRM) for alkaline phosphatase
(ALP,
‘The Chemo-Sero-Therapeutic
Research
SUKEN”), and 2Depment
of Medical
College of Medical Science, Kumamoto 860,
3Address for correspondence:
668 Ohkubo,
mamoto 860, Japan. Fax 81-96-345- 1345.
Institute
(“KAKET-
Technology,
Japan.
Shimizumachi,
Ginkyo
Ku-
4Nonstandard
abbreviations: ERM, enzyme reference material; CRM, Certified Reference Material; JSCC, Japan Society of
Clinical Chemistry; IFCC, International
Federation of Clinical
Chemistry;
SFBC, Societe Francaise de Biologie Clinique; AST,
aspartate aminotransferase;
ALT, alanine aminotransferase;
CK,
creatine kinase; ALP, alkaline phosphatase; LD, lactate dehydrogenase; GGT, -glutamyltransferase;
AMY, amylase; PBS, phosphate-buffered
saline; FBS, fetal bovine serum; BSA, bovine
serum albumin; and DMEM, Dulbecco’s modified Eagle’s medium.
Received
April 11, 1994; accepted
March 23, 1995.
872 CLINICAL CHEMISTRY, Vol. 41, No. 6, 1995
Oka,’
and Naomi
I. Nakano2
EC 3.1.3.1)
(3), ‘y-glutamyltransferase
(GGT,
EC
2.3.2.2) (4, 5), creatine kinase (CK, EC 2.7.3.2) (6 ), and
alanine
aminotransferase
(ALT, EC 2.6.1.2) (7), and
the US National Institute
of Standards
and Technology
(Gaithersburg,
MD) supplies
reference
material
RM
8430 (8) for aspartate
aminotransferase
(AST, EC
2.6.1.1). These preparations
generally
contain a single
enzyme
analyte
of animal
origin in a well-defined
matrix;
RM 8430 and the CRM for CK (CRM 299),
however,
are from human erythrocytes
and placenta,
respectively.
These materials,
carrying
the reference
method values of International
Federation
of Clinical
Chemistry
(IFCC),
are, according
to BCR information,
mainly
intended
for the transfer
of values determined
by IFCC Reference
Methods between laboratories
and
for assigning
values to secondary
ERMs by the IFCC
Reference Methods; they are not primarily
intended for
use in routine enzyme assays as calibrators.
Considerable
efforts have also been devoted in Japan
in the last decade or so to minimize
interlaboratory
variation
of clinical enzyme data. The Japan Society of
Clinical
Chemistry
(JSCC)
has established
recommended methods for five enzymes: AST (9), ALT (10),
CK(11 ), ALP (12), and lactate dehydrogenase
(LD, EC
1.1.1.27) (13). Enzymes
of animal origin were initially
sought for use in ERMs or intermethod
calibrators,
but
without success (14). In Japan, laboratories
can select
a reagent for an enzyme analyte from a variety of kits,
which are based on various methodologies
from many
suppliers.
Further,
no restrictions
have been placed on
the reporting
unit and temperature
of measurement.
These points are reflected in the results of the comprehensive survey of 1992 carried out by the Japan Medical Association
for several
enzymes;
the survey included results of all methods,
units, and temperatures
of measurement
(Table 1). All of the four lyophilized
samples
assayed
showed similar
overall interlaboratory variation
(CV,%) for the raw data. Table 1 shows
such results for two samples
(1 & 2) only and for the
ratios for two sample pairs (2 + 4 & 2 + 1). The greater
the variety in the methodology
and reporting
unit for
an enzyme analyte,
such as ALP, amylase (AMY, EC
3.2.1.1), and cholinesterase
(EC 3.1.1.8), the greater
the difference
between
the CVs for the raw data and
those for the ratio. The remarkably
lower CVs for the
ratios indicate
that the interlaboratory
variation
is
largely systematic
error rather than random error and
can be corrected
with use of appropriate
calibrators,
i.e., those containing
enzymes
for which the intermethod changes
are comparable
with those observed
for measuring
the enzymes
in human
serum.
This
property is referred to as commutability
of ERMs (16).
Table 1. Comparability of interlaboratory
enzyme data In Japan.8
Overall interlaboratory
CV, %
Raw datab
Enzyme
Sample 1
clinical
Ratios
Sample 2
Sample
Sample 21
2:4
AST
12.2
8.8
1.3
ALT
LD
12.1
14.2
11.1
13.3
1.8
1.1
ALP
CK
GGT
58.1
25.3
65.3
20.4
2.2
2.6
16.1
18.1
47.4
17.6
47.6
1.6
1.7
8.0
4.4
1.5
6.4
AMY
CHE
201
204
-
4.9
-
From the report on the results of the 26th (1992) comprehensive proficiency (or contro survey by the Japan Medical Association in which >2000
laboratories participated (15). Corrected data are presented:Outliers exceeding ± 3 SD of the overall mean of the raw data or that of the ratio of one
sample to another calculated for each participating laboratory are excluded.
I) Of the four survey samples, the supplier of samples 1 & 3 is different from
that of samples 2 & 4.
CHE, cholinesterase.
contains
albumin
the
range
of concentrations
(Abnormal).
Both
of
are marketed
by Nippon Shoji, Osaka, Japan,
under
the trade name Seraclear-HE.
Here we describe
the preparation,
including
the
selection
of the cell lines, and properties
of the Seraclear-HE
multienzyme
control serum. Although
it was
marketed
in Japan in 1990 as a control serum for use
with wet reagent
systems,
it was from the beginning
intended
to serve as a possible candidate
for a future
secondary
or working
ERM, whereby
values obtained
by various routine wet reagent
systems
could be converted
to the reference
method value. In a subsequent
report
(17) we will describe
the results
of such an
evaluation
of these materials,
i.e., as an intermethod
calibrator
such routine
data.
for transfer
methods
a reference
for unification
method
to several
of interlaboratory
Materials and Methods
Materials
Current
consensus
in Japan
holds that the ideal
enzyme calibrators
should contain purified human enzymes, although
not all enzymes of human origin are
necessarily
commutable
with the enzymes
in serum.
For practical
purposes,
many Japanese
medical technologists prefer multienzyme
control sera, so that they
can obtain results
for many enzymes
simultaneously
with a multichannel
analyzer.
Accordingly,
we have
looked into enzymes of human origin, avoiding ethical
and legal problems
for developing
ultimately
a secondary ERM that can be used as an intermethod
calibrator
for routine
standardization
of interlaboratory
data in
terms of the reference method value. The utility of such
material
is treated
in detail
in the accompanying
report in this series (17). For similar purposes
in the
past, multienzyme
reference
materials
have been supplemented
with partially
purified
enzymes
from human organs (18, 19).
We have chosen human
erythrocytes
as the source
material
for AST and LD, urine and saliva for AMY.
For ALT, GGT, ALP, and CK we used human cell lines
that produce
high enzyme activity,
in expectation
of
obtaining
desirable
performance
as an intermethod
calibrator,
such as those described
by Rej et al. (16, 20).
Using these enzymes of human origin, we have developed two types of human
enzyme preparations.
One
type
abnormal
these
seven
enzymes
(BSA) solution,
marketed
Industries,
Osaka, Japan
in
bovine
by Wako
under the
serum
Pure
Chemical
trade
name Enzyme
Reference.
The detailed
preparation
of
human enzymes from cell culture and their properties
are described
elsewhere
(21),
as are details
of the
production
and properties
of the prototype.
The other
preparation
consists of two formulations
in a human
serum base. One is supplemented,
to a small extent,
with the same enzymes as those of Enzyme Reference,
to make a normal range control (Normal);
the other is
supplemented
with the same set of enzymes but in an
Dextran
sulfate,
saccharose,
EDTA disodium
salt,
ammonium
sulfate,
n-butanol,
acetone,
sodium
chloride, manganese
chloride,
zinc chloride,
pyridoxal
5’phosphate,
folic acid, and calcium
chloride
were obtained from Wako. Insulin, transferrin,
ethanolamine,
sodium
seler4te
and trypsin
were purchased
from
Sigma Chemical Co. (St. Louis, MO). DEAE-Toyopearl
was from Tosoh (Tokyo, Japan); 5’-AMP Sepharose
was
from Pharmacia
LKB Biotechnology
(Tokyo, Japan).
Dulbecco’s modified Eagle’s medium (DMEM), Eagle’s
minimum
essential
medium,
and RPMI 1640 were
purchased
from Nissui
Pharmaceutical
Co. (Tokyo,
Japan). Fetal bovine serum (FBS) was purchased
from
CC Labs. (Cleveland,
OH).
Cell strains.
The availability
of the human
cell collection at the Chemo-Sero-Therapeutic
Research
Institute (“KAKETSUKEN,”
Kumamoto,
Japan)
permitted
us to study systematically
enzymes of human cell lines.
The collection contains >50 human cell lines, including
diploid cell lines, 12 human hepatoma
cell strains,
and
8 fused cell strains. The cell lines of the collection have
been either produced
in KAKETSUKEN
or donated
through
cooperative
agencies,
such as the Japanese
National
Institute
of Health and some Japanese
universities.
All strains
used in this study were our
original
ones except for FL strain. All were preserved
in liquid nitrogen
before use.
Apparatus.
We used the following apparatus:
a KAKETSUIKEN
original
50-L fermentor
(Fig.
1), a
Wheaton
type-3 cell roller (Millville,
NJ), a Hitachi
(Tokyo, Japan) Model 736 clinical automatic
analyzer
and Model 750 flame spectrophotometer,
a Model UV260 double-beam
spectrophotometer
with a TCC-260
Peltier thermoelectric
controller
(Shimadzu,
Kyoto, Japan), a Hitachi
18 PR centrifuge
with a RPR9-2 rotor,
a Branson
(Danbury,
CT) Sonifier
450 with a horn set
at 30% power 15 mm, an E-type
viscometer
(Tokyokeiki,
Tokyo, Japan),
a DA-100 pycnometer
(Kyoto
Electronics,
Kyoto, Japan), and Ultipor AB3NX7PSH4
CLINICAL CHEMISTRY, Vol. 41, No. 6. 1995 873
Fig. 1. KAKETSUKENoriginal 50-L fermentor.
(pore size 0.45 m) and Posidyne AB3NFZ7PH4
size 0.22 m) filters from P.T.M. Corp. (Cortland,
(pore
NY).
Assay Methods
Routine enzyme assays were performed
with the
clinical analyzer
and the following commercial
reagent
kits: AST, ALT, ALP, and U) with Santest
reagent
from Sanko Junyaku
(Tokyo, Japan);
CK with an
N-Assay
CPK reagent
from Nittobo
Medical (Tokyo,
Japan); GOT with BMY y-GT reagent
from Boehringer
Mannheim
(Tokyo, Japan);
and AMY with a kit from
Nippon Shoji in which p-nitrophenyl-a-D-maltoheptaoside is the substrate.
The final materials
were assayed
for various contaminating
enzymes,
including:
(a) glutamate
dehydrogenase
(EC 1.4.1.3),
assayed
with
Monotest
GLDH, and glucose-6-phosphate
dehydrogenase (EC 1.1.1.49), with an ultraviolet-based
rate kit
from Boehringer
Mannheim;
(b) acid phosphatase
(EC
3.1.3.2), assayed with N-Assay ACP (Nittobo); and (c)
triacyiglycerol
lipase (lipase, EC 3.1.1.3), assayed with
Nescauto
Lipase VE (Nippon Shoji).
Sodium
content was measured
by flame spectrophotometry.
Protein was assayed by the method of Lowry
et al. (22) with BSA as the standard.
Enzyme
Production
Screening
cell strains
liquid
rithmically
growing
cells were then homogenized
in
PBS, pH 7.4, with the hypersonic
homogenizer;
the
homogenates
were then analyzed
for ALP, AST, ALT,
LD, CK, and GOT activities
with the clinical analyzer.
Enzyme
activities
of the screened
cells, analyzed
at
37#{176}C,
were expressed
in U/b6 cells.
Mass production
of human cell strains.
Large-scale
culture methods
were developed
by adapting
the stationary
culture
of human
cells to fermentor
culture
conditions.
The human amnion cell strain (FL), human
kidney
cell strain (EK), and human
macrophage
cell
strain (M4) were cultivated
in RPMI 1640 medium
supplemented
with FBS, 70 milL.
The hepatoma
cell
strain
(KN) was propagated
in a roller bottle with
gentle stirring
in a cell roller. The medium
used for
roller bottle culture
was DMEM supplemented
with
FBS, 70 milL. All cultures
were performed
at 37#{176}C.
Enzyme
purification.
In brief, AST and LD from
human
erythrocytes
were separated
with a 5’-AMP
Sepharose
4B column.
AST from the breakthrough
fraction was partly purified according to the method of
Rej et al. (23) with some modifications.
LD eluted from
the column was further purified on DEAE-Toyopearl
as
described elsewhere
(21 ). ALT
from cultured hepatoma
cells was partially
purified according to the protocol of
Kojima (24). ALP was purified
from spinner-cultured
FL cells according
to the procedures
of Duncan et al.
(25). We purified
CK from a spinner-cultured
human
embryo cell strain as described
by Eppenberger
et al.
(26). GOT was purified
from spinner-cultured
human
macrophage
cells according to the procedures
of Schiele
et al. (27) with some modification.
As described
by
Takeuchi
et al. (28), we purified pancreatic
and salivary AMY from urine and saliva, respectively.
The
details
of the purification
methods
for these enzymes
and their properties
have been described
elsewhere
(21). The overall yields of these enzymes
were >60%
from the starting
material.
All work was carried out at
4#{176}C
unless otherwise
specified. As before, the activities
of these partially
purified
enzymes
in PBS, pH 7.4,
were measured
with the clinical automated
analyzer.
Assays
of isoenzymes.
Electrophoresis
was carried
out as described
previously
(21) for all enzymes.
ALP
was further studied
with respect to heat resistance
and
phenylalanine
inhibition.
The partially
purified
ALP
was heated at 65#{176}C
for 10 mm in 10 mmol/L Tris-HC1
buffer
(pH 7.5) containing
1 mmol/L MgC12 and 0.1
mmol/L ZnCl2. The same ALP samples were tested for
inhibition
with 5 mmol/L L-phenylalanine.
cell strains
for enzyme
production.
The
were taken from cell containers
filled with
nitrogen,
then
maintained
and
propagated
in
each growth medium. The cells were detached from the
wall of a Roux bottle with a trypsin-EDTA
solution and
suspended
in phosphate-buffered
saline (PBS; KC1 0.2
g/L, KH2PO4 0.2 g/L, Na2HPO4
12 H20 2.89 g/L, NaC1
8.0 g/L, pH 7.4) after low-speed
centrifugation.
The
number of viable cells that excluded trypan blue in a
Burker-Turk
hemocytometer
was counted.
The loga874 CLINICAL CHEMISTRY, Vol. 41, No. 6,
1995
Preparation
of the Control Materials
Preparation
of the matrix.
Frozen
human
plasma
obtained
from healthy
donors was purchased
from a
licensed company in the US. The plasma tested negative for hepatitis
B, HIV (AIDS), and adult T-cell
leukemia virus at our laboratory.
Plasma samples were
thawed,
pooled, and then treated
with CaC12 (final
concentration,
50 mmol/L).
After removal
of fibrin,
partial delipidation
with dextran sulfate (final concentration 1 g/L) decreased
the 3-lipoprotein
concentration
Table 2. BasV and stabilizers of Seraclear-HE.
AddItive
Concentration
Sucrose
50 g/L
Pyridoxal
5’-phosphate
CaCI2
30 mmoVL
0.5 mmoVL
0.5 mmoVL
0.1 mmoVL
MgCI2
ZnCI2
Folic acid
0.5 mmoVL
a Partially delipidated serum (dialyzed against 10 mmoVL HEPES buffer, pH
-Removal of Fibrin with CaC12
-Dextran Sulfate Treatment
against HEPES Buffer
-Measurement of Endogenous Enzyme Activity
7.4).
of Stabilizers and Preservatives
to 10 mg/L. This “partially
delipidated
was then dialyzed
against
10 mmol/L
HEPES
buffer,
pH 7.4. Finally,
we added sucrose,
CaC12, MgC12, ZnC12, and pyridoxal
5’-phosphate
to
give the final concentrations
indicated
in Table 2.
Preparation
of final materials.
“Multienzyme
control
serum” was prepared
in two concentrations
in modified
human
serum
base, one supplemented
with “normal”
amounts
of the human
enzymes,
the other supplemented with amounts
calculated
to make an abnormal
LFa1
Product
range control.
Both preparations
were sterilized
by
Fig. 2. Production process of Seraclear-HE.
ifitration
through the micropore filters. The total number of vials of one lot was 10 000-11 000 each (Normal
and Abnormal).
For a quality check, we weighed
>30
of Nippon Shoji before value assignment.
No reference
vials in each lot after filling (3.0 mL). The bottles were
method was available
for AMY.
immediately
cooled to below -35#{176}C
and lyophilized
Stability
of final materials.
Seraclear-HE,
Normal
(final pressure
0.1 mmHg, final temperature
25#{176}C).
At
and Abnormal
(lot 1), have been stored at 10#{176}C
for 5
the end of the process, sterile nitrogen
gas was introyears, during which time samples have been taken for
duced into the lyophilization
chamber to ambient presanalysis
at predetermined
intervals.
We measured
six
sure, the vials were closed with stoppers,
and then
enzymes by the IFCC and SFBC Reference
Methods as
were sealed with aluminum
caps. The residual
moismentioned
above at 30#{176}C
for 4.5 years. The stability
of
ture was about 1% by weight.
Homogeneity
studies
the enzyme activity in reconstituted
material
at 10#{176}C
were carried
out throughout
the filling process
by
and 37#{176}C
within the same working day was examined
weighing
the reconstituted
material
and measuring
by measuring
with the analyzer,
in duplicate,
at 1-h
the seven enzyme activities
in duplicate
for 15 to 20
intervals
for 6 h. The exposed samples were stored at
vials taken at random
from the manufactured
batch.
-20#{176}C
before assay.
Measuring
the enzyme activities
in our laboratory
by
Results
using the commercial
reagent kits at 37#{176}C
showed that
the losses in activities
during
lyophilization
were
Choice of Source of Added Enzymes
within 5% for all seven enzymes.
The production
proPreliminary
experiments
showed that all human cell
cess for Seraclear-HE
is summarized
in Fig. 2.
lines tested exhibited
very high AST and LD activity,
Assignment
of Reference
Method values. Values for
but more than half of the AST obtained from these cell
five enzymes-ALT,
AST, ALP, GOT, and CK-were
lines was of the mitochondrial
type (m-AST). In condetermined
by the IFCC Reference
Methods
at 30#{176}C, trast, nearly all of the AST activity from erythrocytes
and also at 37#{176}C
with the same reagent mixtures
as for
was that of the cytosolic fraction (s-AST). Therefore,
we
the Reference Method (30#{176}C),
on the double-beam
specpurified AST from human
erythrocytes
and obtained
trophotometer,
the reaction
temperature
being conLD as a byproduct
of this AST purification.
Nearly all
trolled by the Peltier thermoelectric
controller.
Simicell lines showed faint ALT activity, but only hepatoma
larly,
the reference
method
values
for LD were
cell lines, e.g., KN, HuH-6, and Hep G2, produced ALT
determined
by the recommended
method of the Societe
at relatively
high amounts
(Table 3). Thus, we chose
Fran#{231}aisede Biologie Cliique
(SFBC). The details of
the highest ALT-producing
hepatoma
cell line (KN) as
the procedures
for these measurements
were as dethe ALT production
source.
scribed previously
(29). For five enzymes,
i.e., ALT,
Although
many cell lines produce a high concentraAST, ALP, CK, and LD, the JSCC Reference
Method
tion of ALP, the kinetic properties
are unlike those of
values also were determined
by Nippon
Shoji. Both
the isoenzymes
in human
serum (30), being instead
Reference
Method values for each analyte
were conheat-stable
and L-phenylalanine-sensitive
tumor types.
At the first screening,
however,
ALP from the FL cell
stantly cross-checked
between our laboratory
and that
of the serum
clear serum”
CLINICAL CHEMISTRY, Vol. 41, No. 6, 1995 875
Table 3. Enzyme productivity of human cell strains.
Enzyme activity, kU/L
LD
CK
OngO
Cell strain
ALP
AST
ALT
HeLa
580
100
2
1480
110
FL
560
83
1
1020
30
10
28
GOT
3
KN
15
71
24
440
59
HuH-6
<1
48
12
220
3
2
M4
<1
34
<1
430
12
53
1
31
<1
260
4
<1
21
21
44
<1
530
8
12
IMR-32
5
44
Hep G2
8
80
HEL
R-66
EK
a
59
1.0
<1
13
Mb
440
280
<1
510
140
<1
930
<1
3.
t
$
I
$1
19
HeLa, FL, R-66, and IMR-32 were obtained from Japanese National
Institute of Health. Hep G2 was purchased from RIKEN Cell Bank (RCB).
HuH-6 was supplied from J. Sato (former Prof. of Okayama University of
Medicine).
had been shown to be a heat-labile
liver-type
isoenzyme
(Table 4). Accordingly,
we chose FL cells as
an ALP production
cell line and attempted
to grow the
cells in suspension
culture for large-scale
production.
Several human
cell lines showed a relatively
high
concentration
of CK activities;
moreover,
the major
isoenzymes
in these cells were about the same as those
of human CK-MM and CK-BB on agarose gel electrophoresis
(data not shown). We chose a human embryo
cell line (EK) as the production
cell source because of
its highest CK expression
(Table 3).
Several
human
hepatoma
cell lines may produce
GOT in relatively
high amounts
(31). We selected
a
human
macrophage
cell line (M4), which exhibited
the
highest
GOT expression
among our cell bank collection
of human cells.
Despite reports on AMY-producing
tumors (32), all
of the human
cell lines used in the AMY expression
study showed activities
too weak to be used for enzyme
production.
On the other hand, salivary AMY activity
was highest in saliva. We found human urine to be a
moderately
rich source of pancreatic
AMY. Therefore,
we decided to produce
AMY from human
urine and
saliva.
Overall, although
the electrophoresis
patterns
of the
selected enzymes were somewhat
different
from those
of human sera, especially
for CK and ALP (Fig. 3), the
Michaelis
constants
for the enzymes
almost coincided
with those of human serum (21).
CK-BB
patterns for CK (left) and ALP (right): (1)
protein markers; (2) patient’s serum 1; (3) Seraclear-HE,
(4) Seraclear-HE, Abnormal; (5) Enzyme Reference; and (6)
patient’s serum 2.
Fig. 3. Electrophoretic
serum
Normal;
line
Production of Control Materials
Culture condition
of the production
cell lines. The
human
macrophage
cell strain
(M4) was routinely
cultured
in spinner vessels and was easy to adapt to
fermentor
culture.
In contrast,
FL and EK cells were
anchorage-dependent
cell strains
in nature,
which
meant that their adaptations
to fermenter
culture
were
not as easy as for M4. These fermentor
cultures were
grown to concentrations
of >0.9 x i0 to 2.5 x i0
viable
cells per milliliter
in RPMI medium
supplemented
with FBS (50-70
mL/L)
in the presence
of
enough oxygen in the 50-L fermenter
at 37#{176}C.
The hepatoma
cell strain (KN) was kept in DMEM
supplemented
with FBS,
70 mLIL. Because
the KM
strain could not be adapted
to suspension
culture, we
cultivated
the cells in roller bottles in an incubation
room at 37#{176}C.
The maximum
cell density of the KN
cells was 5.5 x 106 viable cells per bottle.
Enzyme purification
and heterogeneity
studies.
The
specific activity
of each purified
enzyme is given in
Table 5. Interfering
enzymes in Seraclear-HE
were not
significant:
glutamate
dehydrogenase,
0 U/L; acid
phosphatase,
3.7 UIL; glucose-6-phosphate
dehydrogenase 0.7 U/L; and lipase, 22 U/L.
The 3-mL aliquots
of material
dispensed
into vials
had a mean mass of 3.129 g (SD = 0.001 g, n = 32).
There
was no evidence
for any general
trend in the
variation
in the mass throughout
the filling procedure
Table 4. So me characte ristics of ALP fro m FL cell strain.
Sera from
HoLe cell
Pregnant
woman
33
84
46
34
33
95
52
44
99
96
FL cell
% inhibition
by L-phenylalanine, 10 mmoVL
% heat inactivation
at 56#{176}C
for15 mm
Km,
b
mrnoVLa
With p-nltrophenyl phosphate substrate.
ND, not determined.
876 CLINICAL CHEMISTRY, Vol. 41, No. 6, 1995
0.36
160
NDL
Normal
child
0.29
Patient with
hopatlc disorder
0.40
Table 5. Specific activities of purified enzymes.
Enzyme
Purity, kU/g of
total protein
Source
AST
Erythrocytes
LD
50
120
Erythrocytes
Table 9. Postreconstitutlon stability of six enzymes of
Seraclear-HE (Abnormal) stored at 37#{176}C.
% of original activity remaIning
after reconstitution
Storage, h
1
ALP
AST
ALT
ID
CK
OCT
100
99
98
100
98
98
100
2
100
99
99
99
98
99
Embryo kidney cell line
Macrophage cell line
200
3
101
98
98
99
92
98
99
99
98
99
89
98
Salivary AMY
Saliva
5
100
98
97
99
78
97
Pancreatic AMY
Urine
1000
200
6
101
98
97
99
71
98
ALT
Hepatoma cell line
ALP
FL
CK
GGT
Table 6. Homogeneity
80
100
(Table 6). Table 7 shows the variability
of enzyme
activities
measured
in 20 reconstituted
specimens.
The
between-vial CVs of these enzymes were <2.5%, i.e.,
of the vial-filling process.
Mass, g
Abnormal
(n
=
Normal
(n =
32)
Maximum
Minimum
3.131
3.128
3.130
3.126
Range
0.003
0.004
Mean
3.129
3.129
SD
0.001
0.032
0.001
CV, %
0.032
Filling machine: Bausch-StrObel(Type FFV-4010)
substantially
components
Stability
equal to the within-vial
and between-run
of variance.
study. The long-term
stability
data are
in Table 8, each value being the mean of
presented
duplicate
analyses.
After
storage
at 10#{176}C
for 5 years,
the six enzymes maintained
nearly full activities.
The
stability
data indicate that the material
was stable in
lyophilized
form for at least 2 years when <10#{176}C.
After
reconstitution
with water at room temperature
and
at 37#{176}C,
the materials
were stable throughout
a routine
working
day except for the CK (Table 9).
After reconstitution,
CK activity
in liquid form destorage
________________________________________________
Table 7. Catalytic concentration of each enzyme
measured in reconstituted material.
Normal
i,normai
Enzyme, U/L
at 37#{176}C
ALP
41.0
41.0
AST
ALT
LO
235
GGT
42.0
84.0
CK
AMY
Sodium, mmol/L
Lot no. 10, n
Mean
107
=
102
123
creased at 37#{176}C
within
CV, %
Mean
2.4
656
0.6
2.1
1.2
128
115
0.9
1.1
744
1.0
1.7
1.0
108
320
1.6
0.6
1.5
0.4
469
135
0.6
0.3
CV,
1.1
3 h.
The viscosity of the control material
was slightly lower than that of human
plasma. The density and water content were about the
same as for human plasma (Table 10).
Assigned
values. Analytical
values
shown
in the
package insert were determined
according
to the recPhysicochemical
properties.
ommended
methods
of IFCC and SFBC at 30#{176}C
and
also at 37#{176}C
(Table 11). The variability
of the measurements by the IFCC
described
elsewhere
and SFBC
(29).
% actlvlty
Analyte
0
1
3
103
100
99
6
24
48
100
99
100
99
100
98
100
99
100
100
99
101
100
100
104
100
102
101
101
100
100
103
102
100
100
101
102
101
100
Normal
100
100
ALP
Abnormal
Normal
100
101
102
100
105
Abnormal
100
Normal
100
a
(lot 1) at 10#{176}C.
100
Abnormal
GGT
was to develop a secondary
as an intermethod
calibrator
after storage for months
Normal
ALT
C K
are
Discussion
Table 8. Stability of lyophilized Seraclear-HE
LD
methods
20.
The goal of our study
ERM that could be used
AST
reference
97
99
Abnormal
100
100
101
Normal
100
100
100
97
98
98
Abnormal
100
100
100
100
100
100
97
97
100
98
Normal
Abnormal
97
100
101
101
97
100
Determined by the IFCC Reference Methods at 30#{176}C
except for LD, which was determined by the SFBC Reference Method at 3OC.
CLINICAL CHEMISTRY, Vol. 41, No. 6, 1995
100
98
99
877
properties of Seraclear-HE.
Table 10. Physlcochemical
DensIty, kgfL
Human
Viscosity,
CPb
Water content,
1.0251
1.56
0.8912
Seraclear-HE
(Normal)
1.0355
1.43
0.8772
Seraclear-HE
1.0340
1.30
0.8865
kg/I.0
plasma
(Abnormal)
a
b
Determined by a pycnometer.
Determined by Ubbelohde’s viscometer.
#{176}
Determined by Karl
Fischer procedure.
to help transfer
Reference
Method values to routine
enzyme analyses
in which catalytic activity concentration is determined.
Such an ERM must (a) contain
enzyme analytes
that are commutable
with the counterpart
enzymes in human sera between the Reference
Method and each of routine methods in which they are
to be used, (b) be sufficiently
stable and homogeneous,
(c) carry reliable Reference
Method values, and (d) be
available
in large batches
at reasonable
cost. Seraclear-HE
was developed
with particular
attention
to
each of these properties
and has been made commercially available to all clinical laboratories
in Japan as a
multienzyme
control serum. For adequate
commutability of enzymes,
we chose human
erythrocytes
as the
source of enzymes
for AST (23) and LD (33). Most of
other enzymes required cannot be adequately
prepared
from sources
other than human
organs.
Thus, we
explored the possibility
of utilizing established
human
cell strains
as reliable
production
sources of human
enzymes (34). Human cell strains are readily available,
legally dispensable,
and free from pathogens
and can
be maintained
for quality-control
purposes.
Moreover,
cell strains
can be produced
in large quantities
with
constant
properties,
which eliminates
lot-to-lot variations commonly associated
with the variability
of each
organ. Thus, the human
cell strains
are considered
more suitable
sources
of human
enzymes
than are
human
organs
themselves.
Although
human
cell
strains would be the most prospective
source of human
enzymes,
the actual
selection
of a cell strain
that
produces
the desired
enzyme is laborious
work. We
examined
>50 cell lines of human
origin for their
enzyme productivities.
The cell lines differed from each
other not only in morphology
and growth kinetics,
but
also in enzyme productivities,
especially
for ALP, ALT,
GOT, and CK activity (Table 3). After a long, discouraging screening
test, we finally identified
human cell
strains
suitable
for the production
of ALT, ALP, CK,
and GOT. The Michaelis
constants
of the enzymes
produced
were almost the same as those from human
serum (21), but the isoenzyme patterns of CK and ALP
differed from those of the enzymes in normal human
serum
(Fig. 3). The ALP from human
amnion
cells,
used in Seraclear-HE,
has properties
that resemble
those of human
placenta
and intestine.
We are now
screening
hepatocyte-origin
strains,
osteosarcoma
strains, and others for cell lines that produce ALP with
properties
that more closely resemble
those of ALP
from human
liver or bone. Some cell lines showed
relatively
high CK activities,
but the major CK isoenzyme produced
by these cell lines was CK-BB, well
known to be a heat-labile
isoenzyme.
Table 9 shows the
significant
decrease
of CK activity with storage as a
reflection
of the high concentration
of CK-BB isoenzyme of the material.
The EK cell strain we selected for
CK production
also produces
mainly
CK-BB isoenzyme. Recently, we found a cell strain that produced a
high activity of CK-MM, but the total CK productivity
of that cell strain was lower than that of the EK strain.
These results suggest that human cultured cells do not
always
ensure
enzymes
with properties
similar to
those of the counterpart
enzymes
in normal
human
serum.
Not only the CK-MM but also the ALT of our human
cell culture
systems
are relatively
low producing;
we
are still trying to find highly productive
systems
for
these enzymes.
The specific enzyme productivity
of a
cell line directly
influences
the cost of the product
through
determining
the size of production
equipment
and the amount
of FBS consumed
per unit of enzyme
activity
produced.
Culturing
human
cells generally
requires FBS-which,
however, raises cultivation
costs
drastically
and makes more difficult the purification
of
the cultured
enzyme. Therefore,
we have made great
efforts to reduce the amount
of FBS used in culture
media at our production
laboratory.
Identifying
possible areas of cost reduction
is important
in the manufacture
of these cell-cultured
enzymes.
We are now
trying to produce suitable
isoenzymes
by using moreproductive
strains
obtained
by cell cloning and the
Table 11. Comparison of Refe rence Method va lues of Se raclea r-HE (lot 1).
Normal
Abnormal
IFCC
Enzyme, U/I.
ALP
AST
ALT
LD
a
30#{176}C
24.9
29.3
21.8
135
37#{176}C
37.2
44.8
31.0
220#{176}
CK
46.6
78.3
GGT
33.4
45.4
Determined by SFB C Reference Method.
878
IFCC
JSCC
CLINICAL CHEMISTRY, Vol. 41, No. 6, 1995
30#{176}C
93.3
24.6
19.7
53.6
46.0
-
37#{176}C
121
38.8
29.1
94.0
77.3
-
JSCC
37#{176}C
30#{176}C
37#{176}C
165
101
81.2
618
207
114
961
577
93.5
78.1
237
790
144
111
415
239
381
237
376
107
147
30#{176}C
154
-
-
gene recombinant
techniques
recombinant
GOT (36) and
of Tanase (35). Human
CK (37) have also been
reportedly
expressed
at high enough
amounts
to be a
source
for a standard
ERM.
The matrix
of Seraclear-HE
is human
delipidated
serum (Table 12, Fig. 2), and the added enzymes have
not been highly purified.
Thus, the final product
contained not only the seven enzyme analytes
but also
some enzyme
dehydrogenase
contaminants-of
activity
might
which
interfere.
only glutamate
At this stage
of
purification,
the specific activities
in this multienzyme
control serum seemed to be sufficient
for intermethod
calibration
for some enzymes,
as will be shown in
subsequent
reports.
Therefore,
we agree that ultrapurity need not be mandatory
for enzymes used in reference material
preparations
(25). The ideal matrix
for
an ERM would be one whose composition
is fully
defined and would exactly match the composition
of the
clinical
specimen.
A reasonable
matrix
would be a
serum albumin solution or artificial
serum-like
material. At first, therefore,
we used the BSA-based
Enzyme
Reference
material
a primary
ERM (21). For development of a secondary
ERM, we tried to prepare
enzyme
analyte-free
human
serum
for the base matrix.
However, the base serum of Seraclear-HE
is not free from
endogenous
enzymes,
because the techniques
used to
Table 12. Major constituents
Constituent
ALP, U/L#{176}
AST, U/LU
ALT, U/LU
LD, U/Lb
LD-1, %
LD-2, %
LD-3, %
LD-4, %
of Seraclear-HE
(lot 1).
Normal
Abnormal
165
101
81.2
618
32.8
34.8
21.8
3.2
24.9
29.3
21.8
135
24.8
41.4
16.9
3.6
LD-5,%
11.9
6.9
GGT, U/LU
33.4
107
CK, U/L0
CK-BB, %
46.6
2
7
239
45
2
At albumin
CK-MB,
position,
%
%
5
CK-MM, %
AMY, U/L#{176}
Pancreatic, %
Salivary,%
86
45.3
57.7
2
51
215
53.2
42.3
46.8
Total protein,
g/L
48
40
Albumin,
Total cholesterol, g/L
Triglycerides,
mg/L
28
0.25
80
22
0.21
80
frLipoprotein,
mg/L
20
Phospholipids,
g/L
g/L
0.59
0
0.53
Freefatty acids, mg/L
4.1
3.6
Total lipids,
3.16
3.16
pH
T at660
U
b
g/L
7.58
nm, %
95.95
7.57
96.30
IFCC method at 30#{176}C.
SFBC method at 30#{176}C.
Activity at 37#{176}C
wIth p-nitrophenyl-a-o-maltoheptaoslde
substrate.
the endogenous
analytes
significantly
altered
protein
composition
(data not shown).
A
well-defined
matrix
such as human
serum
albumin
solution
is easier to reproduce,
but human
serumbased materials
may provide a better matrix for enzyme analytes.
The physicochemical
properties
of the
two types of preparations-Enzyme
Reference
and Seraclear-HE-are
similar to those of human serum as a
whole, but the viscosity of Enzyme Reference is slightly
lower than the average viscosity for human sera.
Homogeneity,
as determined
by the filling mass as
well as by the Na concentration
and catalytic activity
concentrations,
appears
to be similar
to that of the
CRMs supplied by the BCR. Our stability data, extending now for 4 years by the manual reference
methods of
either IFCC or SFBC, indicate
that the preparations
are stable for at least 2 years at 10#{176}C.
At the time that
these preparations
were introduced
to the Japanese
clinical laboratories
(1990), most of control sera carried
numerous
methodand instrument-dependent
assigned values for each analyte. As H#{248}rder
and Rej (38)
have strongly
suggested,
assignment
of target values
only in terms of reference
methods
would not only
reduce the amount of excess work for the manufacturers, but also potentially
influence the move towards the
remove
the
serum
standardization
of clinical
enzyme
data.
Because
the
JSCC Reference
Methods
for five clinically
important
enzymes
were recently
approved
as recommended
methods, we were prompted
to assign the IFCC as well
as JSCC Reference Method values at 30#{176}C
and 37#{176}C
to
Seraclear-HE.
Although
every effort has been made to
follow the protocols of the Reference
Methods
as carefully as possible
with constant
cross-checks
between
our laboratory
and that of Nippon Shoji, these values
are not certified-there
being no official body to do so in
Japan. Thus we introduced
Seraclear-HE
as a control
serum that would be useful not only in precision control
but also for accuracy
control for laboratories
whose
routine methods are based upon these reference
methods. Seraclear-HE
is now widely used in Japan: (a) as
an interlaboratory
survey
material,
(b) to verify the
transferability
of the reference
methods,
(c) to transfer
Reference
Methods
into individual
laboratories
to
make possible
intermethod
comparison
between
their
routine method values and the Reference
Method values, and (d) for accuracy control in laboratories
using
reagents
based on either of the Reference
Methods or
as a calibrator
to correct bias arising
from day-to-day
variability
of enzyme assays such as instrumentor
reagent-related
factors that affect both patients’
sera
and calibrators
equally.
To be a candidate
secondary
ERM,
Seraclear-HE
must be demonstrated
for each analyte to be commutable between
each of the analytical
systems (reagentinstrument
pairs)
in routine
use and the Reference
Method.
Such
evaluations
are under study, and the
accompanying
report (17) deals with the results of such
an evaluation
for AST and ALT. We are hopeful
that,
except for ALP, this preparation
can be of value in
transfer
of the Reference
Method values to several of
CLINICAL CHEMISTRY, Vol. 41, No. 6, 1995 879
the various
routine
reagent chemistry.
methods
that
are
based
on wet
We express our thanks to Chozo Hayashi and Yoji Marui
(Saiseikai Nakatsu Hospital) for valuable advice and discussion.
We also thank
Hiroki Tanaka
(Nippon Shoji) for his collaboration.
References
1. Siest G, Oster T, Visvikis A, Thioudellet C, Angele C, Schiele
F, Weilman M. Application of gene transfer technologies to the
production of enzyme reference materials: example of-glutamyltransferase [Review]. Clin Chem 1993;39:1573-89.
2. Moss DW, &hiele F, Colinet E. Reference materials for clinical
enzymology: the work of the community bureau of reference of the
European Community. Clin Chem 1986;32:556-8.
3. Schiele F, Muller J, Colinet E, Siest G. Certification
of an
enzyme reference material for alkaline phosphatase (CRM 371).
Cliii Biochem 1991;24:159-68.
4. Schiele F, Muller J, Colinet E, Siest G. Production
and
certification of an enzyme reference material for y-glutainyltransferase (CRM 319). Part 1: preparation and characterization. Clin
Chem 1987;33:1971-7.
5. Schiele F, Muller
J, Colinet E, Siest G. Production
and
certification of an enzyme reference material for ‘y-glutamyltransferase (CRM 319).Part 2: certification
campaign.
Clin Chem
1987;33:1978-82.
6. Mathieu M, Steghens J, H#{248}rder
M, Moss DW, Colinet E.
Profflis C. A reference preparation of creatune kinase BB isoenzyme. Clin Chem 1993;39:1894-8.
7. Schiele
F,
Brettschneider
Muller
J,
Colinet
E,
Siest
G, Arzoglou
P,
H, et al. Interlaboratory
study of the IFCC
performed with use of a
method for alanune aminotransferase
partly purified reference material. Clin Chem 1992;38:2365-71.
8. Bowers GN Jr, McComb RB, Syed D, Edwards CC, Paule RC,
Greenberg N, et al. Primary human aspartate aminotransferase
reference material. The catalytic activity concentration of AST in
RM 8430 as measured by the IFCC method for AST in 10 enzyme
standardization laboratories. Clin Chem 1988;34:450-9.
9. Japan Society of Clinical Chemistry. Recommendation for
measuring enzyme activity in human serum. Aspartate aminotransferase (1989-08-30). Jpn J Clin Chem 1989;18:226-49.
10. Japan Society of Clinical Chemistry. Recommendation
for
measuring enzyme activity in human serum. Alanine aminotransferase
(1989-08-30).
Jpn J Clin Chem 1989;18:250-62.
11. Japan Society of Clinical Chemistry. Recommendation
for
measuring
enzyme activity in human serum. Creatine kinase
(1989-08-30).
Jpn J ClinChem 1990;19:184-208.
12. Japan Society of Clinical Chemistry.
Recommendation
for
measuring
enzyme activity in human serum. Alkaline phosphatase (1989-08-30).
Jpn J Cliii Chem 1990;19:209-27.
13. Japan Society of Clinical Chemistry. Recommendation
for
measuring enzyme activity in human serum. Lactate dehydrogenase (1989-08-30).
Jpn J Clin Chem 1990;19:228-46.
14. Kinoshita
N, Marui Y, Kohda K, Hayashi
C, Miyai K.
Evaluation
of a new multienzyme reference. The 6th ISQC(Osaka) Conference Manual 1987;66.
15. Japan Medical Association. Report on the results of the 26th
control, 1992:98-148.
16. Fasce CF Jr, Rej R, Copeland HW, Vanderlinde
RE. A
discussion of enzyme reference materials: application and specifications [Review]. Clin Chem 1973;19:5-9.
17. Nakano NI, Eto A, Chikaura Y, Oishi T. Multienzyme control
serum (Seraclear-HE)
containing
human enzymes from established cell lines and other sources. 2: Evaluation as candidate
working enzyme reference material for alanine and aspartate
aminotransferases.
Chin Chem 1995;41:881-91.
18. Koedam JC, Steentjes GM, Buitenhuls 5, Schmidt E, Klauke
R. Production and certification of secondary enzyme reference
880 CLINICALCHEMISTRY. Vol.41,No.6,1995
materials (ERMs). Part 1: preparation
of the sera and some of
their properties.
Cliii Chem 1986;32:1901-5.
19. Koedam JG, Van Dreumel HJ, Ham WVD, Steentjes GM,
Terlingen JEA. Production and certification of secondary enzyme
reference materials (ERMs). Part 2: the need for meaningful
protocols that assure photometric accuracy and a well-defined
process for value assignment. Cliii Chem 1986;32:1906-13.
20. Rej R, Jenny RW, Bretaudiere
JP. Quality control in clinical
chemistry: characterization
of reference materials. Talanta 1984;
3 1:851-62.
21. Marui Y, Hayashi C, Matsuda Y, Matsuura S, Eto A, Ohba Y,
et al. Multienzyme
reference material from established
human
cell lines and human sources. Clin Chim Acts; in press.
22. Lowry OH, Rosebrough
NJ, Farr AL, Randall RJ. Protein
measurement with the Folrn phenol reagent. J Biol Chem 1951;
193:265-75.
23. Rej R, Vanderlinde RE, Fasce CF Jr. An L-aspartate:2oxoglutarate
aminotransferase
reference material from human
erythrocytes:
preparation
and characterization.
Clin Chem 1972;
18:373-83.
24. Kojima H. Large scale preparation
of crystalline
alanine
aminotransferase
from pig heart and some of its properties.
Kumamoto Med J 1979;32:23-34.
25. Duncan PH, McKneally SS, MacNeil ML, Fast DM, Bayse
DD. Development of a reference material for alkaline phosphatase. Chin Chem 1984;30:93-7.
26. Eppenberger HM, Dawson DM, Kaplan NO. The comparative
enzymology of creatine kinase. 1. Isolation and characterization
from chicken and rabbit tissues. J Biol Chem 1967;242:204-9.
27. Schiele F, Siest G, Moss DW, Colunet E. The certification
of
the catalytic concentration of gamma-glutamyltransferase
in a
reconstituted
lyophilized material CRM No 319. BCR Information, 1986.
28. Takeuchi T, Fujiki H, Kanieya T. Characterization
of amylanes produced by tumors. Chin Chem 1981;27:556-9.
29. Nakano N, Eto A, Chikaura Y, Matsubara T, Koga K, Kawata
K. Measurement of enzyme activity of control materials containing human enzymes by IFCC reference methods. Jpn J Chin
Pathol 1991;39:981-93.
30. Nitowsky HM, Herz F. Hormonal regulation of alkaline
phosphatase
in dispersed cell cultures.
Biochem Biophys Res
Comm 1963;11:261-6.
31. Visvikis A, Goergen J-L, Oster T, Bagrel D, Wellman M, Marc
A, et al. Gamma-glutamyltransferase
from human hepatoma cell
lines: purification
and cell culture of HepG2 on microcarriers.
Cliii Chim Acts 1990;191:221-32.
32. Ueda M, Kobayashi M, Taketa K, Sato J. Ectopic production
of a salivary type amylase by adenocarcinoma cells: demonstration by a culture technique. Cliii Chim Acts 1977;80:105-11.
33. BuhI SN, Copeland WH, Fasce CF Jr, Kowalski P, Richards
AH, Vanderlinde
RE. Proficiency testing: purification
of lactate
1 and results of its use as a reference material in
the New York State program. Chin Chem 1977;23:1000-11.
dehydrogenase
34. Eto A, Shiki A, Harada S, Oka T, Sako M. Control sera
containing
human enzymes and their preparation. Japanese
patent 1989; Sho 64-11000.
35. Tanase S. DNA sequence: alanine aminotransferase.
DNA
data bank Japan 1992; access no. D10355.
36. Visvikis A, Thioudellet C, Oster T, Fournel-Gigleux 5, Wellman M, Siest G. High-level expression of enzymatically
active
mature human y-glutamyltransferase
in transgenic V79 Chinese
hamster cell. Proc Natl Acad Sci U S A 1991;88:7361-5.
37. Friedman
DL, Kesterson
R, Puleo P, Wu AHB, Perryman
MB. Recombinant
creatine kinase proteins and proposed standards for creatine kinase isoenzyme and subform assays. Clin
Chem 1993;39:1598-601.
38. H#{248}rder
M, Rej R. Requirements
and functions of reference
materials for enzymes. Prog Chin Enzymol, Vol. 2, 1983;29-34.