the use of both bile salts and colipase for the following reasons. The effect of bile salts in the absence of colipase is strongly concentration dependent (10, 11,22 ...
CLIN. CHEM. 31/2, 257-260 (1985)
Measurement of Lipase Activity by a Differential pH Technique F. Ceriotti, P. A. Bonini, M. Murone, L Barenghl, M. Luzzana,1 A. Mosca,1 M. Rlpamonti,1 and L Rossi-Bemardi’ This is a new electrochemical method for determination of lipase activity in biological fluids, including serum, plasma, and duodenal juice. Advantages of turbidimetric methodsshort reaction time, and small sample and reagent volumes-are combined with those of titrimetric methods: measurement of absolute activity (i.e., no standardization required), saturated substrate conditions, and direct measurement of reaction products. The proposed method is easy, inexpensive, and takes only 3 mm. Precision is good: CV = 3.74% within day and 7.3% between days at the clinicaldecision concentration, CV = 1.86% within day and 4.65% between days for above-normal lipase activities. The standard curve is linear up to 4500 U/L. Results (y) correlate well with those by turbidimetry (c): y = 0.9287x - 65.3 (r = 0.9719). Reference values are between 0 and 130 U/L.
oxygen electrode (18). So far, however, as Hockeborn and Rick (19) stated in 1982, “simple methods that give reproducible quantitative results have not yet been developed” for lipase determination. The differential pH technique, already applied to measurement of serum glucose (20), seems to combine the following advantages of photometric and titrimetric methods: substrate under optimal conditions, short measurement time, low reagent cost, and simple instrumentation. In the experiments we report here, we used an olive oil emulsion in the reaction mixture, at a concentration sufficient to saturate the enzyme. This provided linear reaction kinetics after a short lag phase, typically about 30 s. Because of the high sensitivity of the differential pH methodology, measurement time can be shortened to about 1 mm, thus providing a rapid and accurate method that is also suitable for emergency determinations.
AddItIonal Keyphrases: electrochemistry reference interval turbidimetry compared dure
Materials and Methods
.
duodenal juice
emergencyproce-
Apparatus
Measurement of lipase activity (EC 3.1.1.3) is considered by many to be the most clinically useful assay in the diagnosis of acute pancreatitis. In fact, in this disease, the increase in serum lipase activity is more specific and frequently more pronounced and prolonged than that of amylase (1-3). Despite these diagnostic advantages, lipase measurement has not been favored for several methodological reasons: long analysis time, nonspecificity of some of the
substrates used, and difficulties related to use of an emulsified substrate. Among the different methods developed to overcome these problems, most are modifications of the Cherry-Crandall procedure (4), in which fatty acids released from an olive oil emulsion are titrated, either with use of an indicator (5-7) or electrometrically (8). These methods are time consuming, having incubation times from 1 to 24 h, and therefore not suitable for emergency determinations. A better method is the kinetic determination of lipase by use of a pH-stat apparatus (9, 10). Turbidimetric and nephelometric methods (11-13) are used in routine determinations because of their rapidity; however, because of instrumental limitations, the substrate concentrations used are far below the Km value of the reaction, making standardization of the procedure less reliable. Several colonmetric techniques have also been proposed, based either on determination of the released fatty acids as copper soaps (14) or on cleavage of some synthetic substrate to yield a colored compound (15, 16). The former is very complicated; the others are nonspecific and require inhibition of the esterase activity. Methods involving fluorescent substrates (17) also suffer from low substrate specificity. Recently, a procedure involving lipoxygenase has been proposed, involving an
Laboratorio di Analisi, H. S. Raffaele, Istituto di Ricovero e Cura a Carattere Scientifico, Milano, Italy. ‘Dipartimento di Scienze e Tecnologie Biomediche, Universita’ di Milano. Received August 1, 1984; accepted September
25, 1984.
The apparatus used is the same as we previously described (20), except that it has been thermostated at 30 #{176}C and its differential amplifier has been connected to a potentiometric recorder to monitor the progress of the reaction. Data acquisition, slope calculation, and linearity check were performed by a 32-bit personal microcomputer (DLT 16000; Delta Instrumentation, 20090 Segrate, Italy). For the turbidimetric measurements, we used a photometer (Beckman Model 42) with a temperature-controlled cuvette at 30 #{176}C.
Reagents Acacia solution. Dissolve 10 g of acacia (Merck cat. no. 4282) in 100 mL of a solution containing 0.1 mol of NaC1 and 1 g of NaN3 per liter. Stock olive oil emulsion. To remove any free fatty acids from olive oil, add 20 g of neutral alumina (aluminum oxide, 90% active, neutral; Merck art. 1077) to 200 mL of commercial olive oil and stir for 20 mm with a magnetic stirrer. Allow the alumina to settle, decant the oil, and
repeat the treatment with fresh alumina two times; centrifuge to completely remove the alumina (14). Stored at -20 #{176}C, the purified oil remains stable for several months. Into a 50-mL glass beaker weigh 4 g of purified oil and 16 g of acacia solution. Sonicate 10 times for about 1 s each time at full power (we used a Soniprep 150 MSE sonicator, from Cellai SRL Struinentazione Scientifica, 20134 Milano, Italy), which will produce a whitish emulsion with some large oil drops on the surface. Place this emulsion in 5-mL glass vials and sonicate each one three times for 10 s each; avoid heating the emulsion. As verified with a particle analyzer (Coulter S Plus), the size distribution of the emulsion droplets remained constant during seven months of storage at 4 #{176}C. However, the total number of particles counted in the range 2-2O 10-15 L decreased by approximately 30%. Working olive oil emulsion. Add 150 pL of 0.1 mollL NaOH to 5 mL of stock olive oil emulsion. The pH should be
CLINICALCHEMISTRY, Vol. 31, No. 2, 1985 257
in the range 9.1-9.3. at 4#{176}C. Working triolein weigh 4 g of triolein
This emulsion isstableforthree days emulsion.
Into a 50-mL
(glycerin-trioleat,
Fluka
glass beaker cat. no. 92859)
and 16 g of acacia solution. Sonicate and titrate as for the olive oil emulsion. Working buffer. Prepare a solution containing 10 mmol of glycine, 100 mmol of NaC1, and 1 g of NaN3 per liter and titrate to pH 9.2 with 1 mol/L NaOH. Kept in air-tight bottles to avoid absorbing C02, which would lower the pH, the solution should be stable for longer than one month at room temperature. Turbidimetric method. For comparison, we also determined lipase with the “Monotest 10 Lipase” kit (Boehringer Mannheim, ref. 263346), following the manufacturer’s instructions. This method isbased on the turbidimetric procedure of Ziegenhorn et al. (12).
Differential pH Procedure Start up. Switch on the instrument each morning and fill it with about 200 mL of fresh buffer. After two to five blank cycles, a drift < 0.0003 pH/mm will ordinarily be observed, and measurements can begin. Reagent blank measurement. Add 75 p.L of emulsion to the buffer; this generates a drift of about 10 iO pH units/mm. We decreased this drift to about 5. i0 pH units/mm by adding 50 L of heat-inactivated (1 h at 56#{176}C) pooled serum to the reaction mixture. Thus, to measure lipase activity in a reagent blank, add a sample of inactivated serum at the beginning of each analytical series and use the measured drift as the blank drift. Correct all measurements of patients’ sera for blank drift. Typical measurement. Figure 1 shows the sequence of events in a measurement. For activities up to 500 U/L, use a lag phase of 1 mm and a measurement time of 1 mm. For greater activities (up to 4500 U/L), use shorter lag phase and measurement times to keep the measured pH change within 0.1 pH unit. Cakulation. Lipase activity is expressed as micromoles of free fatty acids released per minute. At pH 9.2 all the fatty acids produced are fully dissociated (21, 23), stoichiometri-
I
9.180i
2
1 TIME
3
(mm)
Fig. 1. Change in pH during a typical measurement sequence 1,5O-iL sample, after injection into the mixingchamber,which contains0.75 mL of working buffer, is aspirated into both electrodes. 2. to the remaning 400 L of sample-buffer mixture, 75 L of working olive oil emulsionis added.3. aspiration of the sample + emulsion into channel 1. 4, linear phase of the reaction.5, washing of electrodes and mixing chamber with fresh buffer
258
CLINICAL
CHEMISTRY,
Vol. 31, No. 2, 1985
cally releasing 1 mol of proton per mol of fatty acid. If the pH change during the reaction is less than 0.1 pH unit the buffer power can be assumed to be constant, and the number of protons released can be derived from the following equation (20): (1)
[HI=13pH
where /3 is the buffer power, as determined from the pH change produced by adding 50 /LLof 0.1 molIL HC1 to 10 mL of buffer. The lipase activity of a serum sample is computed as follows: Activity
=
d [H]/min
(2)
.
where d is the dilution ratio. Combining equations 1 and 2 yields: Activity = d f3 pWmin
(3)
For example, when a 50-L sample is injected in the mixer, diluted 15-fold (final volume 0.75 mL), and gives a pHJmin of 0.0180 with a buffer power of 4060 moL1L, the lipase activity is calculated as follows: Activity
=
15 4060 0.018 .
-
=
1096 UJL
The effect of adding the emulsion can be ignored because the increased dilution of the sample is compensated by the corresponding decrease of buffer power. The emulsion itself has negligible buffering power.
Results and Discussion Effect of bile salts and colipase. In our method we have avoided the use of both bile salts and colipase for the following reasons. The effect of bile salts in the absence of
colipase is strongly concentration dependent (10, 11,22,2426). Thus they might have a little activating effect at low concentrations but at slightly higher concentrations intensely inhibit lipase activity. Moreover, there are large differences among the various kinds of bile salts (22-24), and an acidic shift of pH optimum in the presence of bile salts, with and without colipase, has also been reported (24, 29). In the absence of colipase, the only advantage in using bile salts should be a linearization of the kinetics involved (2,10,22,24,30,31) by preventing interfacial denaturation. However, albumin has the same protective effect (30), and under our assay conditions we have always obtained linear kinetics. The action of colipase in the absence of bile salts is uncertain: several investigators (24, 31, 32) have reported an activating effect, ranging from 30 to 80%; others (25) observed no effect. The ability of colipase to counteract inhibition by bile salts is well known (20,25,26, 31-33); it seems to act by restoring lipase interfacial adsorption onto oil micelles (28,29,33). When coupled to bile salts, colipase appears to shorten the lag phase and improve the linearity of the reaction kinetics (28, 32). Therefore, colipase should be used only with bile salts. Our assay conditions, in which linear kinetics and a short lag phase are obtained without including these components, thus produce a procedure that is less expensive and in which any variability that would be caused by bile salts is eliminated. Effect of cations. With preparation of pure lipase and reaction mixtures containing bile salts, many effects ofCa2 on lipase activity have been observed: linearization of the kinetics (22), elimination of lag phase and acidic shift of the pH optimum due to bile salts (27), and strong activation of the lipase hydrolytic activity (21,34). This activating effect seems to be due to a lipase-Ca2 binding that facilitates the adsorption of lipase onto lipid micelles. However, in systems
without bile salts and with serum lipase samples, no effect of Ca2 was observed (5, 6,10). We checked the Ca2 effect by adding to the working buffer CaCl2 in increasing final concentrations (0.2 to 100 mmol/L), and noted only a very small activation effect (about 10% increase in lipase activity for a 500-fold difference in Ca2 concentration). Including Mg2 at the same concentrations inhibited lipase activity by as much as 20% when Mg2 exceeded 10 mmol/L. The appropriate concentration of NaC1 is very important for complete ionization of the free fatty acids released (22), but NaC1 seems also to act directly on lipase activity when bile salts are present (24,35); lipase activation was maximal when NaC1 was 100 mmol/L, but with a little change between 50 and 200 mniol/L. KC1 added in various concentration to the working buffer increased lipase activity only slightly, but when used instead of NaCl in the working buffer it decreased the lipase activity by about 35%. Buffer and pH optimum. We used a glycine buffer because Tris buffer with the same pH and buffer power inhibited lipase activity by about 50%. We found a broad pH optimum curve, with a maximum at pH 9.2, which does not significantly differ from the maxima reported by Tietz and Fiereck (8), pH 8.8; Rick (9), pH 8.6; and Hockeborn and Rick (19), pH 8.5-9.0. Effect of different substrates and substrate concentration.
duodenal juice samples with inactivated serum to avoid denaturing the sample and impairing the lipase activity. Precision and linearity. Within-day variation was determined by analyzing, 10 consecutive times, two serum sam-
ples with different activities. For one serum the results ranged from 259 to 293 U/L (mean 274 U/L, SD 10.26 U/L, CV 3.74%). For the other, measured lipase activity ranged from 933 to 990 U/L (mean 965 U/L, SD 17.98 U/L, CV 1.86%). Between-day variation was determined by analyzing aliquota of the same sample on 12 consecutive days. For a serum with slightly above-normal lipase activity, the mean was 296 U/L (range 258 to 342 U/L, SD 20.87 UIL, and CV 7.3%). For a serum with greater activity, the mean was 989 UIL (range 931 to 1039 U/L, SD 46.04 UIL, and CV 4.65%). We checked the linearity of the dilution curve by assaying dilutions of serum from a patient with pancreatitis with a normal serum (Figure 3). The curve was linear up to at least the 4500 UIL, the activity of the undiluted serum from the pancreatitis patient. Specificity. Measurements performed on eight serum samples with activities from 300 to 2500 UIL, before and after heating for 1 h at 56#{176}C to inactivate the pancreatic thermolabile lipase (9), showed complete parallelism of results by both methods and support the specificity of our method.
Increasing quantities of the working olive oil emulsion were added to the mixing chamber of the apparatus and the initial rate of the reaction was plotted as a function of the added volume (Figure 2). We used olive oil because it is less expensive than triolein and is analytically as useful (10,19). In fact, we observed a good correlation (r = 0.997) between serum lipase activity as determined with olive oil (y) or triolein (x) as substrate (y = 0.9858x - 24; n = 14). We preferred to use an emulsion volume of 75 /.LL and a sample volume of 50 pL to measure maximum lipase activity in sera or in duodenal juice diluted with inactivated sera. Under these conditions the enzymic reaction shows a prolonged linear phase, even with sera having high lipase activities. Collection,
treatment,
and
stability
of sample.
Sample
Comparison
with the turbidimetric
procedure.
We assayed
35 sera by our method (y) and by the turbidimetric Monotest 10 Lipase test (x). The following regression equation was obtained: y = 0.9287x - 65.3 (r = 0.9719) and = 95.4. Reference values. We assayed 92 serum samples obtained from fasting inpatients and outpatients (50 men, 42 women, ages 20-65 years) who had normal values for amylase, alkaline phosphatase, y-glutamyltransferase, cholinesterase, aspartate aminotranferase, alamne aminotranferase, and lactate dehydrogenase. The observed lipase values ranged from 0 to 130 U/L and, in agreement with the findings of others (6, 10, 11, 36), showed a non-gaussian distribution of serum lipase activity (Figure 4). We therefore suggest that 130 U/L at 30#{176}C may be a clinically useful
upper limit for the normal reference
dilution is a critical step in lipase determination (33). In our assay conditions, diluting sera with isotonicsaline caused significant loss of activity, with high dilutions almost completely inactivating samples of pancreatic or duodenal juices. Adding bovine serum albumin to the buffer partly prevents this loss. We also observed a marked decrease in enzyme stability when the protein content of the medium was low. We recommend diluting high-activity serum and
interval.
4000
0.050
a,.
3000
I-
0.040
a,. I-
C.,
0.030
2000 ‘U
U) 0.020 -J
1000
0.010
25
50 EMULSION
75
100
(ph
0 20 40 60 80 100% sample Fig. 2. Rate of hydrolysis 30 s after addition of various amounts of 100 80 60 40 20 0 %diluent emulsifiedoil (final concentration 35 g/L) (0) pancreaticjuice diluted with heat-inactivated serum, activity890 uL;(#{149}) Fig. 3. Linearity of lipase activity determination with differential pH serum sample, activity3000 U/L method
CLINICALCHEMISTRY, Vol.31, No. 2, 1985 259
n
24
20 a,.
16
2 ‘U
o
12
‘U U,.
8 4
0
20
40 60 80 100 120 LIPASE ACTIVITY (u/L) Fig. 4. Frequency distribution of lipase activity in 92 subjects as measured with the proposed pH method
References 1. Song H, Tietz NW, Tan C. Usefulness of serum lipase, esterase, and amylase in the diagnosis of pancreatitis-a comparison. Clin Chem 16, 264-268 (1970). 2. Proeless HF, Wright BW. Lipoxigenic micromethod for specific determination of lipase activity in serum and duodenal fluid. Clin Chem 23, 522-531 (1977). 3. Kreutzer HH, PenningsAW, Punt JMHM, Verduin PA. Further studies on determination of lipase activity. Gun Chim Acta 60, 273279 (1975). 4. Cherry IS, Crandall LA. The specificity of pancreatic lipase: Its appearance in the blood after pancreatic injury. Am J Physiol 100, 266-273 (1932). 5. Henry RJ, Berkman S, Sobel C. On the determination of “pancreatitis lipase” in serum. Gun Chem 3, 77-89 (1957). 6. MacDonald RP, Lefave RO. Serum lipase determination with an olive oil substrate using a three-hour incubation period. Clin Chem 8, 509-519 (1962). 7. Roe JM, Byler RE. Serum lipase determination using a one-hour period of hydrolysis. Anal Biochem 6, 451-460 (1963). 8. Tietz NW, Fiereck EA. A specific method for serum lipase determination. Clin Chim Acts 13, 352-358 (1966). 9. Rick W. Kinetischer Test zur Bestimmung der Serum-Lipaseaktivitat. Z Kim Chem Kim Biochem 7, 530-539 (1969). 10. Tietz NW, Rapique EV. Proposed standard method for measuring lipase activity in serum by a continuous sampling technique. Clin Chem 19, 1268-1275 (1973). 11. Shihabi ZK, Bishop C. Simplified turbidimetnc assay for lipase activity. Clin Chem 17, 1150-1153 (1971). 12. Ziegenhorn J, Neumann V, Knitsch KW, Zwez W. Determination of serum lipase. Clin Chem 25, 1067 (1979). 13. Arzoglou PL, Ferad G, Khalfa K, Metais P. Conditions for the assay ofpancreatic lipasefrom human plasma using a nephelometnc technique.Gun Chim Acts 119,329-335 (1982). 14. DirstinePH, Sobel C, Henry RJ. A new rapidmethod forthe determinationofserum lipase. Clin Chem 14, 1097-1106 (1968).
260
CLINICAL CHEMISTRY, Vol. 31, No. 2. 1985
15. Whitaker JF. A rapid and specific method for the determination of pancreatic lipase in serum and urine. Clin Chim Acts 44, 133-138 (1973). 16. Kurook 5, Kitamura T. Properties of serum lipase in patients with various pancreatic diseases. Analysis by a new serum lipase assay method (the BALB-DTNB method) in combination with gelfiltration and isoelectrofocusing technique. J Biochem 84, 14591466 (1978). 17. Guibault GG, Hienserman J. Fluorimetnic substrate for sulfatase and lipase. Anal Chem 41, 2006-2009 (1969). 18. Gabriel RJ, Knoblock EC, Koch TR. Measurement of serum lipase with the oxygen electrode. Clin Chem 27, 163-165 (1981). 19. Hockeborn M, Rick W. Zur Bestimmung der katalytischen Activitat der Lipase mit dem kontinuierlichen titrimetrischen Test. J Clin Chem Clin Biochem 20, 773-785 (1982). 20. Luzzana M, Dossi G, Mosca A, eta!. Measurement of glucose in plasma by a differential pH technique. Gun Chem 29,80-85(1983). 21. Benzoana G. Sur le role des ions calcium durant l’hydrolyse des triglycerides insolubles par la lipase pancreatique en presence de sels biliares. Biochim Biophys Acts 151,137-146 (1968). 22. Benzoana G, Desnuelle P. Action of some effectors on the hydrolysis of long-chain triglycerides by pancreatic lipase. Biochim Biophys Acts 164, 47-58 (1968). 23. Patton JS, Carey MC. Inhibition of human pancreatic lipasecolipase activity by mixed bile salts-phospholipid micelles, Am J Physiol 241, G328-G336 (1981). 24. Borgstom B, Erlanson C. Pancreatic lipase and co-lipase. Interaction and effects of bile salts and other detergents. Eur J Biochem 37, 60-68 (1973). 25. Morgan RGH, Hoffman NE. The interaction of lipase, lipase cofactor and bile salts in tryglyceride hydrolysis. Biochim Biophys Acts 28, 143-148 (1971). 26. Borgstom B, Earlson C. Pancreatic juice co-lipase:Physiological importance. Biochim Biophys Acts 242, 509-513 (1971). 27. Brown JB, Belmonte AA, Melius P. Effect of divalent cations and sodium taurocholate on pancreatic lipase activity with gum arabic emulsified tributyrylglycerol substrates. Biochim Biophys Acts 486, 313-321 (1977). 28. Lairon D, Nalbone G, Lafont H, et al. Possible roles of lipids and colipase in lipase adsorption. JAm Chem Soc 17,5263-5269(1978). 29. Vandermeers A, Vandermeers-Piret MC, Rathe J, Christophe J. Effect of colipase on adsorption and activity of rat pancreatic lipase on emulsified tributynin in presence of bile salts. FEBS Left 49, 334-337
(1975).
30. Brockeroff H. On the bile salts and proteins as cofactor of lipase. JBmoi Chem 246, 5828-5831 (1971). 31. Momsen WE, Brockman HL. Effect of colipase and taurodeoxycholate on the catalytic and physical properties of pancreatic lipa8e B at an oil-water interface. J Biol Chem 521, 378-383 (1976). 32. Julien R, Canioni P, Rathelot J, et al. Studies on bovine pancreatic lipase. Biochim Biophys Acts 280, 215-224 (1972). 33. Chapus C, Sari H, Semeriva M, Desnuelle P. Role of colipase in the interfacial adsorption of pancreatic lipase at hydrophylic interfacies. FEBS Left 58, 155-158 (1975). 34. Lainon D, Nalbone G, Lafont H, et al. Effects of bile lipids on the adsorption and the activity of pancreatic lipase on triacylglycerol emulsions. Biochim Biophys Acts 618, 119-128 (1980). 35. Benzoana G, Desnuelle P. Etude cinetique de l’action de la lipase pancreatique sur de triglycerides en emulsion. Essai d’enzimologie en milieu heterogene. Biochim Biophys Acts 105, 121-136 (1965). 36. Verduin PA, Punt JHM, Kreutzer HH. Studies on the determination of lipase activity. Gun Ghim Acts 46, 11-19 (1973).
