bovine globulin, 0.1 mL of Tween polyoxyethylene (40) sorbitan ... PLA2 gamma-globulin; TR-FIA, time-resolved fluoroimmunoassay; and IR-PLA2 ...
CLIN. CHEM. 29/10, 1777-1780
(1983)
Time-Resolved Fluoroimmunoassayof Human Pancreatic Phospholipase A2 U. Eskola,1’2 limo J. Nevaiainen,1’ and Timo N.-E. Lovgren3
Jarkko
We describe an immunofluorometric assay for human pancreatic phospholipase A2 based on time-resolved fluorescence. The labeled antibody technique in combination with
the time-resolved 1-s fluorometric detection of the europium label, which essentially eliminates all background fluorescence, resulted in a high sensitivity (20 ng/L) and a wide (5000-fold) linear range. Nonspecific binding was minimized by treating the solid-phase antibody with NaSCN before coating, to remove endogenous antigen, and by immunosorbent purification of the antibody before labeling with europium. This is a one-incubation, multi-site, solid-phase assay on polystyrene microtiter strips, even though a polyclonal antibody was used. As measured by this assay, activity of immunoreactive phospholipase A2 was found to be above normal
in sera of patients suffering from acute pancreatitis.
AddItional
Keyphrases:
pancreatic disease
solid-phase assay
reference interval Phospholipase A2 (PLA2; phosphatide 2-acyihydrolase; EC 3.1.1.4) is a digestive enzyme synthesized by pancreatic acinar cells and secreted as an enzyrnatically inactive proenzyme (prophospholipase A2) into the duodenum, where trypsin converts it into the active form. During an attack of acute pancreatitis this enzyme may be activated within the pancreas and released into the pancreatic interstitium and peripancreatic tissue and carried via vascular channels to various organs of the body. It has been postulated that phospholipase A2 might be a key enzyme in the pathogenesis of pancreatic damage and other tissue injury in acute pancreatitis (1, 2). Diagnosis of acute pancreatitis is difficult because no current laboratory test is specific for the disease. The demonstration of increased serum amylase activity has served as a sign of pancreatic damage for over haifa century (3). However, it is quite clear that pancreatic disease is not the sole cause of increased serum amylase (4).
Several immunoassays, which measure the concentrations of pancreatic enzymes in serum instead of their catalytic activity, have recently been recommended for use in the diagnosis of pancreatic disease (4-10). We have described the purification of human pancreatic secretory phospholipase A2 from cadaver pancreas in the accompanying paper (11). This enzyme is localized in the pancreatic acinar cells in the normal pancreas and in glands affected by acute and chronic pancreatitis (12). 1 Departments of Pathology and 2 Clinical Chemistry,University of Turku, Kiinamyllynkatu 10, 20520 Turku 52, Finland. 3Wallac Biochemical Laboratory, Turku, Finland. 4Address correspondence to this author. 5 abbreviations: PLA2, phospholipase A2; T-buffer, Tris HCI (50 mmol/L, pH 7.4) containing 9 g of NaC1 and 0.5 g of NaN3 per liter; TB-buffer,the same, plus 5 g of bovine serum albumin per liter; TBBTE-buffer, same as TB-buffer, plus 0.5 g of bovine globulin, 0.1 mL of Tween polyoxyethylene (40) sorbitan monostearate, and 40 tmol of EDTA per liter; anti-PLA2, antiPLA2 gamma-globulin; TR-FIA, time-resolved fluoroimmunoassay; and IR-PLA2, immunoreactive PLA2. Received June 1, 1983; accepted July 19, 1983.
The purpose of the present study was to develop an immunoassay for the measurement of PLA2 in sera of patients suffering from pancreatic disease and thus improve the laboratory diagnosis of diseases such as acute and chronic pancreatitis and pancreatic cancer.
Materials and Methods Materials Sephadex G-50, Sepharose 6B, and activated CH-Sepharose 4B were purchased from Pharmacia, Uppsala, Sweden. Ethylenediaininetetraacetic acid (EDTA) was obtained from Merck AG, Darmstadt, F.R.G. Bovine serum albumin and bovine globulin were purchased from Sigma Chemical Co., St. Louis, MO 63178. Europium oxide, Eu203, was obtained from Fluga AG, Buchs, Switzerland. The chloride salt of Eu was prepared by dissolving Eu203 in hydrochloric acid and evaporating the excess acid. Aminophenyl-EDTA-Eu was synthesized by a modification of the method described by Sundberg eta!. (13). A corresponding isothiocyanate derivative of the Eu-chelate was prepared by the reaction with thiophosgene. 2-Naphthoyltrifiuoroacetone was synthesized according to the method described by Reid and Calvin (14). Other chemicals used were of analytical grade. Human pancreatic PLA2 was purified from pancreatic tissue, and antiserum to human pancreatic PLA2 was raised in rabbits (11). We used three buffers in this assay.5 The washing solution contained 9 g of NaCl and 0.5 g of NaN3 per liter.
Methods Coating
of microtit.er
strip
wells
with
anti-human
PLA2.
The gamma-globulin fraction was prepared from the rabbit anti-human PLA2 antiserum (anti-PLA2) by precipitation with an 180 g/L solution of Na2SO4 at 25#{176}C. The anti-PLA2 gamma-globulin (3.5 mg) was incubated in 1 mL of T-buffer containing 3.5 mol of NaSCN per liter for 10 mm at room temperature. The gamma-globulin was recovered on a 0.9 x 28cm column of Sephadex G-100 equilibrated with T-buffer. The microtiter strip wells (polystyrene, 12 wells per strip; Eflab Oy, Helsinki, Finland) were coated with NaSCNtreated anti-PLA2 gamma-globulin (5 zg/mL) by adsorption in 0.2 mL of sodium carbonate buffer (0.1 mol/L, pH 9.5) for 20 h at room temperature. After washing (washing device; Nunc-Immuno Wash 12, Nunc, Denmark) three times with isotonic saline, the coated wells were saturated by adding 0.3 mL of TB-buffer. After incubation for 2 h at room temperature, the strips were ready for use. The coated strips could be stored at 4 #{176}C for at least two months in TB-buffer. They were washed twice with saline solution before use. Isolation
of specific
antibodies
to PLA2
by the immunosor-
bent technique. The immunosorbent for the isolation of specific antibodies to PLA2 was prepared by conjugation of human pancreatic PLA2 to activated CH-Sepharose 4B according to instructions supplied by the manufacturer. Resorption was carried out with 3.5 mol/L NaSCN (pH 7.4). A 20-mg fraction of the total gamma-globulin yielded 370 g of specific antibody. Labeling of anti-human PLA2. The isothiocyanate derivaCLINICAL
CHEMISTRY,
Vol. 29, No. 10, 1983
1777
tive of the Eu-chelate at 50-fold molar excess was allowed to react with immunosorbent-purified anti-human PLA2 for 20 h at 4#{176}C. A column 1 cm in diameter filled to 5.5 cm with Sephadex G-50 and for a further 52 cm with Sepharose 6B was used to separate the conjugated protein fraction from excess reagent. Collected 1-mL fractions were monitored by absorbance at 280 rim and fluorescence (see below). The incorporation ratio of Eu to IgG was 3.4. The protein was considered as 100% recovered. Preparation (IR-PLA2-free
of human serum).
serum
free of immunoreactive
PLA2
The immunosorbent gel was prepared by coupling 30 mg of anti-PLA2 (total immunoglobulin fraction) to 1 g of activated CH-Sepharose 4B. Ten milliliters of normal human serum was treated with 1 g of gel. Time-resolved fluoroimmunoassay (TR-FIA). For the assay, 25 L of serum samples or 25 L of human PLA2 standard in LR-PLA2-free serum was pipetted into the microtiter wells. Then 25 ng of Eu-labeled anti-human PLA2 in 175 p1 of TBBTE-buffer (diluted in buffer shortly before use to bind any dissociated free Eu) was added to each weil. After gentle shaking, the strips were incubated for 1.5 h at room temperature and then washed three times with saline. Measurement of fluorescence. Eu was dissociated from the second antibody on the solid phase and measured in solution
as a 2-naphthoyltrifluoroacetone chelate, with use of singlephoton-counting time-resolved fluorometry as described by Soini and Kojola (15). Total counting time was 1 s. The background signal in a strip without Eu was 890 counts/s (CV 4.4%). This was subtracted from all readings. The variation in the wells of one microtiter strip was analyzed by using a Eu standard of 0.1 nmol/L, which gave a signal of 90500 counts/s (CV 1.3%).
nals by more than 15-fold. ‘Freating the antiserum used for coating with NaSCN lowered the zero-sample signal from i05 to 400 counts/s when 100 rig of labeled IgG was used. Normal human serum contains immunoreactive PLA2. Treatment of normal human serum with immunosorbent lowered the signal observed with serum to the level recorded with the buffer. At higher doses, standards made in IRPLA2-free serum gave somewhat lower signals than standards in buffer; therefore, we always used serum-based standards. The detection limit was calculated to be about 20 ng/L, this limit being defined at the PLA2 concentration corre-
sponding to the mean fluorescence signal of 12 aliquots of zero sample (one strip) plus 3 SD (376 ± 147 counts/s). The corresponding signal for the 1 .eg/L standard was 7519 counts/s.
The range of values that can be measured with the assay is 0.02-100 g/L. Figure 2 shows the precision profile calculated from 12 replicates (one strip) for each PLA2 concentration. The CV varied from 5 to 12.0%, depending on dose. Specificity of the antiserum was assessed by fluoroimmunoassay after fractionation of a crude preparation of human pancreas PLA2 obtained by purification with ammonium sulfate (11) with liquid chromatography on a cationexchange column. Only one peak was obtained (Figure 3). Analytical recovery of added human PLA2-standards (10,
12
10 C8
Results Figure 1 shows dose-response curve for human PLA2 in IR-PLA2-free serum. The ratio between the maximum signal and the zero-sample signal is 635. Purification of labeled antiserum by immunosorbance improved the specific sig-
>4 LI
01
10
50
25
Phosphohpase 2
0 C
100 ng ImL
Fig. 2. Precision profile of TR-FIA of human pancreatic PLA2 SD, g/L (#{149}), and cv, % (0), were calculatedfrom 12 replicates for each phospholipaseA2 concentration
1o
0 2
a) In In C J 0 0
a)
10
io
Phophoipo.,
A2
8 -
0 C
a; 0 (0
a) 0 J U.
2
102 0
002
01
Phospholipase
1
10
100
A2 ng / mL
Fig. 1. Standard curve for human pancreatic phospholipase A2 The dashedhonzontalline denotesthe backgroundlevelplus3 SD
1778 CLINICALCHEMISTRY, Vol. 29, No. 10, 1983
10
20 Fraction
30
40
50
number
Fig. 3. Specificity of pttospholipase A2 TR-FIA A crudepreparationof human pancreaswas fractionated by ‘high-performance” ion-exchangechromatography and fractions wereanalyzed byTR-FIA.Onlyone pealcwas obtained (-). The chromatogramwas recorded by the instrumentat 280 nm. Solid line: the NaCI gradient
50, and 100 .tgfL) from the serum of four normal healthy persons was 95.4% (n = 14, CV 10.9%). Analyses made after twofold dilution of serum samples having high IR-PLA2 values (range, 14.8-84.4 rigfL) with IR-PLA2-free serum (x) correlated well with the results obtained for undiluted samples (y): y = 1.8x + 4.4; n 19; r 0.95. Serial-dilution experiments with IR-PLA2-free serum were also made for three serum samples having different IRPLA2 concentrations. Excellent correlation (r = 1.00) was found in each case (Figure 4). The normal reference interval for IR-PLA2 concentration was determined from data on 29 serum samples taken from 14 healthy women and 15 men. No sex-related difference was found. The normal range, defined as three standard deviations from the mean (6.5 j.tg/L), was 0.68-12.4 .ug/L =
=
5).
(Figure
We made some attempts to correlate the concentration of IR-PLA2 with clinicalfindings by analyzing 13 serum samples from patients suffering from acute pancreatitis. The samples were selected on the basis of the clinical diagnosis and an increased serum amylase concentration (>309 U/L) in the first sample taken after hospital admission. With one exception, the ffi-PLA2 values were clearly above normal (Table 1). The mean IR-PLA2 value for the pancreatitis group (41.2 g/L, SD 26.6 ig/L) was highly significantly (p
