- Email: [email protected]
5'-Nucleotide phosphodiesterase isoenzymes in human serum: Quantitative measurement and some biochemical properties
275
Clinica Chimica Acta, 164 (1987) 275-284 Elsevier
CCA 03773
5 ‘-Nucleotide phosphodiesterase isoenzymes in human serum: quantitative measurement and some biochemical properties * Jiirgen Liithje and Adaling Institut (Received
Ogilvie
ftirPhysiologische Chemie, Universitiit Erlangen-Niimberg,
30 June 1986; revision received
Key words: 5’-Nucleotide
phosphodiesterase
12 December isozyme;
1986; accepted
Erlangen (FRG)
after revision
Phosphodiesterase;
Isoenzyme;
5 January Human
1987) serum
Summary
A method based on native PAGE is used for the quantitative measurement of 5’-nucleotide phosphodiesterase isoenzymes (5’-NPD; EC 3.1.4.1) in human serum. In contrast to other techniques this method works with a commercially available substrate. In sera of healthy donors four isozymes could be separated, designated as 5’-NPD-I, 5’-NPD-II, 5’-NPD-III and 5’-NPD-IV in the reverse order of their electrophoretic mobility. When low amounts of serum were applied to the gel, the separation between all four activities was sufficient enough to allow their quantitation. Higher amounts of serum impaired the separation between the isoenzymes I and II, However, even when using high amounts of serum, the quantitation of these activities was possible when taking advantage of some of their biochemical properties which are described herein.
Introduction
5’-Nucleotide phosphodiesterase (5’-NPD) is an exonuclease hydrolyzing DNA or RNA starting from the 3’-end to give 5’-nucleotides. The enzyme is widely
* This work is dedicated to Prof. Dr. Walter Kersten on the occasion of his 60th birthday. Abbreviations: Bis, N, N’-methylene-bis-acrylamide; DEAE-, diethylaminoethyl-; 5’-NPD, 5’-nucleotide phosphodiesterase (The isozymes of 5’-NPD were designated as 5’-NPD-I, 5’-NPD-II.. etc. Similar the expressions isozyme I, isozyme II.. . etc. or the ‘first activity’, the ‘second activity’. . etc. were used); PAGE, Polyacrylamide gel electrophoresis; PNPpT, p-nitrophenyl thymidine 5’-phosphate; S-300, Sephacryl-300; TEMED, N, N, N’, N’-tetramethyl-ethylene-diamine; Tris, Tris-(hydroxymethyl)-aminomethane. Correspondence to: Dr. Jtirgen Llithje, Institut fiir Physiologische Chemie, Fahrstrasse 17, D-8520 Erlangen, FRG.
0009-8981/87/$03.50
0 1987 Elsevier Science Publishers
B.V. (Biomedical
Division)
276
distributed in nature [1,2]. Among human organs the highest specific activities are found in kidney, pancreas, liver and small intestine [3]. In body fluids the activity was highest in seminal plasma, followed by serum, bile, cerebrospinal fluid and urine [3]. Variation of 5’-NPD-activity in human sera was reported by Lo et al using a fluorescent method [4] and by Hynie et al applying a chromogenic technique [5]. Tsou and coworkers also described the isoenzyme distribution of the 5’-nucleotide phosphodiesterase in human sera [6,7]. Normal human serum reveals four isozymes which have been identified as 5’-NPD-I, 5’-NPD-II 5’-NPD-III, and 5’-NPD-IV in the reverse order of their electrophoretic mobility. 5’-NPD-III and 5’-NPD-IV usually account for more than 80% of the total serum 5’-NPD activity [6]. In patients with primary hepatocellular carcinoma (PHC), and liver metastases an additional faster moving isozyme (5’-NPD-V) was seen [7,8]. Since then a lot of work on the 5’-NPD-V has been published showing that this isoenzyme is useful in detecting liver metastasis in various forms of cancer. These studies have recently been reviewed [6]. In contrast to the fifth isoenzyme very little information is available on the other four enzymes. They are present in sera of healthy donors and, under pathological conditions, their pattern may change. Due to the lack of a quantitative method, which can be performed in clinical laboratories, very little information about these enzymes is available. In pilot studies we have observed marked alterations of the isozyme activities I-IV in various cases of disease. This paper describes a gel electrophoretic method for the quantitative measurement of the isoenzymes and gives information on those of their biochemical properties which are helpful for their determination. Materials and methods Reagents
Acrylamide (p.A.) was purchased from Serva, Heidelberg, FRG. N, N ‘-Methylene-bis-acrylamide (Bis) and N, N, N ‘, N ‘-tetramethyl-ethylenediamine (TEMED) were from Bio-RAD Labs., Richmond, CA, USA. Ammoniumperoxidisulfat and Tris-(hydroxymethyl)-aminomethane, both of analytical grade, were purchased from Roth, Karlsruhe, FRG. n-Butanol and salts were from Merck, Darmstadt, FRG. p-Nitrophenyl thymidine 5’-phosphate (PNPpT) was obtained from Sigma, St. Louis, MO, USA. Diethylaminoethylcellulose (DE 52) was from Whatman, UK, and Sephacryl 300 (S 300) from Pharmacia, Uppsala, Sweden. Serum specimens were obtained from healthy laboratory volunteers or from the University Hospital Erlangen-Niimberg. Separation of 5’-NPD isoenzymes by gel electrophoresis
Separation of isozymes was performed by a 7% polyacrylamide gel according to the conditions described by Tsou and Lo [6]. Serum samples (50 ~1) were mixed with 10 ~1 of 99% glycerol (p.A.) and 3 11 of 450 mmol/l Tris-boric acid (pH 9.4). After centrifugation (18 000 X g, 5 rnin, 20 o C) to remove any particulate material the samples were loaded onto the gel using a microsyringe. Electrophoresis was
performed at 15 mA (constant current) for 2 h and then at 10 mA for a further 3 h. The separation was stopped when albumin (dyed with bromophenol blue) had reached the area marked on the plates (between 11 and 12 cm). The slab gel was cut into individual tracks, which were sliced transversely into pieces 3 mm in width. The gel slices were transferred into plastic caps containing 0.7 ml of Tris-HCl buffer (100 mmol/l, pH 8.6) and 4 mmol/l MgCl,. For measurement of phosphodiesterase activity the reaction was started with the chromogenic substrate p-nitrophenyl thymidine 5’-phosphate (PNPpT) at a final concentration of 4.5 mmol/l, if not indicated otherwise. Incubation was performed at 37°C under gentle rocking usually for 16 h. Absorbance was measured in a sp~trophotometer (UV-210 A, Shimadzu Seisakusho LTD., Kyoto, Japan) at 405 m. Enzyme activity was expressed as the increase in light absorbance during the incubation time. Our experiments revealed that all the 5’-NPD isoenzymes were located within the first 7 cm of the gel. Therefore, it was unnecessary to slice the whole track. Column chromatography DEAE-cellulose chromatography and gel filtration with Sephacryl 300 were performed as described with the exception that a 50 mmol/l T&s-HCl buffer (pH 8.0) was used [ll]. Chromatography was performed at 4*C. The fractions (12 and 8 ml, respectively) were assayed for phosph~i~ter~e activity using PNPpT as a substrate. The assay contained, in a total volume of 0.4 ml, 0.32 ml of the sample, 0.04 ml of Tris-HCI (1 mol/l, pH 8.6), 5 mmol/l MgCl, and substrate at 12 mmol/l. Incubation was performed at 37*C. The absorption (405 run) of eightfold diluted portions of the incubation assay were measured at various times (multipoint determination). Enzyme activity was expressed as A E,,/h. Samples obtained from colurmr chromatography were concentrated tenfold by ultrafiltration (Amicon, model 8 MC) before applying to gel electrophoresis. A Diaflo-membrane PM 10 was used, having a molecular exclusion limit of M, 10000. n-Butanol extraction 0.5 ml of the sample to be extracted was shaken with 0.25 ml of n-butanol for 1 min at room temperature (4 X 15 s, Whirl mix). After ~nt~fugation (18000 x g, 5 min, 20 o C) the cheese-like supernatant was discarded. The procedure was repeated with the aqueous phase obtained from the first extraction. After careful removal of the butanol layer the aqueous sample was used for gel electrophoresis experiments. The removal of traces of n-butanol from the sample by dialysis was found to be unnecessary.
Separation of 5’-2WD isoenzymes by nutiue PAGE and their ~nt~tat~on Serum from a healthy donor was separated on a 7% poiya~l~de gel under non-denat~a~g conditions. The gel slices were assayed for phosphodiesterase
278
E 405
a IV
i
5
15
25
4
Slice
Number
Fig. 1.5’-NPD isoenqme pattern of human serum. Serum (20 ~1) of a healthy donor (a) or of a patient with acute myeolic leukemia (b) was separated by native PAGE. The gel was sliced and 5’-NPD activity in the gel pieces was determined as described under ‘Materials and Methods’. Enzyme activity is expressed as the increase of absorbance at 405 nm during the incubation of 16 h.
activity as described under ‘Materials and Methods’. Figure la shows four isozymes which were identified as 5’-NPD-I, 5’-NPD-II, 5’-NPD-III, and 5’-NPD-IV in the reverse order of their electrophoretic mobility. The reproducibility of the isoenzyme positions in repeated gels was excellent for the 5’-NPDs I, III and IV. The first isozyme was found in slice numbers 4 or 5, 5’-NPD-III in 15 or 16 and the fourth isoenzyme in position I9 or 20. These positions referred to the peaks of the activities, respectively. 5’-NPD-II showed a greater variability, the peak of this activity was usually detected in gel slices 7, 8, 9, or 10, making it sometimes difficult to differentiate 5’-NPD-I from 5’-NPD-II. Fig. lb shows the 5’nucleotide phosphodiesterase isoenzyme pattern of a pathological serum obtained from a patient with acute myeloic leukemia. A fifth, very fast moving isozyme emerged in position 23. This 5’-NPD-V has been intensively studied by Tsou and coworkers (61. We could demonstrate that this technique was able to quantitate isozyme activities. All the four activities exhibited linear kinetics for at least 16 hours (not shown). Experiments with different amounts of serum also revealed linear relationships as far as separation was suffi~~t. Figure 2 demonstrates that the activities clearly separated when 8 or I6 nl of serum were used. Higher amounts, however, impaired
279
Slice
Number
Fig. 2. Effect of the amount of the serum applied to the gel on the 5’-NPD isozyme pattern. The amounts of serum are given in the boxes. Enzyme activity is expressed as described in the legend of Fig. 1.
the separation between the two slow moving activities (I and II), whereas the 5’-NPD isozymes III and IV remained well distinguishable. Concerning the influence of storage conditions on isozyme activities, experiments with sera from several donors revealed different stabilities of the four isoenzymes during storage at - 20 ’ C. Whereas the activities of 5’-NPDs I, III and IV remained stable for months, isozyme II lost part of its activity. Especially, after repeatedly freezing and thawing of the serum, a severe loss of this activity was observed. The other isozymes revealed no significant alterations.
The usual isoenzyme pattern of human serum, when incubation with substrate was performed at a concentration of 4.5 mmol/l, is shown in Fig. 3b. Reduction of the PNPpT concentration to one tenth (0.45 mmol/l) caused the apparent loss of the isozyme II activity (Fig. 3a). Increasing the substrate concentration up to 15 mmol/l resulted in a huge activity peak (Fig. 3~). The dependence of the elution profile on substrate concentration is obviously due to K, differences. A kinetic analysis of 5’-NPD-II after purification from the other isozymes by means of DEAE-cellulose chromatography (Fig. 4) revealed a K, of 30 mmol/l. Since the variation of the substrate concentration in the range from 0.5 to 15 mmol/l had no influence on the activities I, III and IV, we suggest that their Michaelis constants are
280
&OS 0.5
5
15
St ice
25
Number
Fig. 3. Influence of substrate concentration on the isoenzyme pattern. In three experiments, serum (20 ~1) from the same donor was separated by get electrophoresis. The gel slices were incubated with (a) 0.45 mmol/l, (b) 4.5 mmol/l, (c) 15 mmol/l of pnitrophenyl these 5’-phosphate. Incubation was performed for 16 h and activities are expressed as in Fig. 1.
smaller than 0.5 mmol/l. It is obvious that the use of the substrate at a low concentration is an useful way for the quantitation of 5’-NPD-I, in spite of its moderate separation from the second isozyme.
I
1
10
m
Fraction
30
LO
50
60
I
Number
Fig. 4. DEAE-cellulose chromatography of human serum. Serum (30 ml) was diluted fivefold with elution buffer and applied to a DF.AE-cellulose column. Elution was performed with a linear salt gradient. Fractions were assayed for 5’-NPD activity (0 -0). Enzyme activity is the increase of light absorbance (405 nm) during the incubation time of 1 h. Assay conditions are given under ‘Materials and Methods’.
281
Separation of serum 5’-NPD isozymes by anion exchange and gel filtration chromatography
Fractionation of human serum on a DEAE-cellulose column resulted in the separation of two 5’-NPD activities (Fig. 4). One activity did not bind to the colmmr (‘non-binding activity’), whereas the other had a strong affinity to the anion exchanger and was only eluted with NaCl at SO-120 mrnol/l (‘binding activity’). Both enzyme preparations were characterized by native gel electrophoresis. Figure 5a demonstrates that the non-binding phosphodiesterase could be attributed to the 5’-NPD-II, since the maximum activity was found in gel position 9, being typical for the second isozyme. The fact that two samples from different regions of the non-binding activity (fractions 7,8 and fraction 17, see Fig. 4) revealed an identical migration behavior on the gel (not shown), suggested that this activity was electrophoretically homogeneous. The characterization of the binding activity, however, displayed three activities on the gel. They were localized in typical positions of the 5’-NPD isoenzymes I, III, and IV, respectively (Fig. 5a). By using gel filtration we could separate a tiny peak (which eluted first) from a huge phosphodiesterase activity. Gel electrophoresis revealed that the small activity
E LO5
2 4
2
b 0.5
0.2
5
15
25
SI ice Number Fig. 5. Ch~acte~ation of 5’-NPD activities from the DEAE-ceIlulose and gel f&ration steps by native PAGE. Concentrated samples from the cohunn chromato~aphy steps were applied to gel electrophoresis. 5’-NPD activities in the gel slices were measured as described in the text. Enzyme activity is expressed as described in the legend of Fig. 1. a. DEAE-cellulose step (Fig. 4): aiiquots from fractions 7 and 8 of the ‘non-binding activity’ (0 0) and from fractions 29 and 30 of the ‘bindiug activity’ (ol) were applied onto different slots of the native gel. b. Sephacryl-300 step: aliquots from the small peak 0) and from the major activity (O0) were applied onto different slo& of the native gel. (O-
282
could be attributed to the 5’-NPD-I, whereas the other enzyme activity consisted of two isoenzymes (III and IV, Fig. 5b).
still
Effect of n-butanol treatment on the 5’-NPD isozyme pattern
The 5’-NPD isozyme pattern of butanol-treated serum was compared with that of untreated serum (Fig. 6a). Butanol extraction of serum caused the loss of the 5’-NPD-III activity, whereas the fourth isozyme was still detectable at its typical position. The slow moving enzymes I and II were not distinguishable (Fig. 6a). To find out more about the fate of these isozymes, we repeated the experiment with a slight modification. The enzyme activity was measured at a lower substrate concentration. Under this condition the activity of 5’-NPD-II is negligible (see Fig. 3). After butanol extraction the isozyme activities I and III were lost (Fig. 6b). The findings of both experiments suggested that the 5’-NPD-II activity was not affected by the extraction procedure. These observations were confirmed by experiments with partially purified isoenzymes. Serum was fractionated by DEAE-cellulose chromatography (Fig. 4). The first, non-binding activity (5’-NPD-II) was not affected by the butanol treatment. However, after extraction of the binding activity, the isozymes I and III disappeared. 5’-NPD-IV was not influenced by this procedure.
ELOS0.7 -
s .> ..I-
flL -
y
0.1 -
z e
5
15
25
15
25
aJ z
0.9
5 2 $
0.6
E
5
ice Number Fig, 6. Effect treatment on 5’-NPD isoenzyme pattern of serum. The extraction is under ‘Materials and a. of butanol-treated Gel pieces were incubated with substrate at a 0) and of untreated serum (0 -0). serum (Oconcentration of 4.5 mmol/l. b. A similar experiment as in a, however, with substrate at 0.45 mmol/l.
283
Discussion
In this paper we demonstrate the qualitative and quantitative measurement of the 5’nucleotide phosphodiesterase (5’-NPD) isoenzymes. After gel electrophoresis we have detected the isozyme activities by using the commercially available, chromogenic substrate p-nitrophenyl thymidine 5’-phosphate (PNPpT). To overcome the problem of product diffusion of the soluble p-nitrophenol in the gel we have applied a gel slice technique, which ahowed the incubation of gel pieces for times long enough to detect even weak enzyme activities. In contrast to this method, other workers used substrates which have to be chemically synthesized. The indigogenic method described by Tsou et al [7] is based on the production of an insoluble, stable indigo dye, so that whole gel rods can be incubated for long times without having the problem of product diffusion. The product of another very sensitive, fluorescent technique, 4-methylumbelliferone, is soluble. Therefore, only short incubation times are possible, as long as whole rods or lanes are incubated [9]. In combination with a slice technique this method would be very powerful, however with the disadvantage that the substrate is not commercially available (to our knowledge) and that special equipment (fluorometer) is needed. Our method works with a standard spectrophotometer. We have detected four 5’-NPD isoenzymes in sera from healthy donors. This is in good accordance with the findings of Tsou and coworkers [6]. Under certain pathological conditions, especially in malignant states of the liver, Tsou described the emergence of a fifth, fast-moving isoenzyme [L&10], which has also been confirmed by our results. Furthermore there are marked changes in the pattern of the other isoenzymes. The linearity of the kinetics as well as the experiments with variable serum concentrations demonstrate that the technique is able to determine the activities of the isozymes III and IV, respectively. The activities of 5’-NPD-I and -11 can be measured as long as low concentrations of serum are applied. Higher concentrations lead to unsufficient separation of these activities (Fig. 2). Even when higher serum concentrations are required, it is possible to differentiate between the isozymes I and II by taking advantage of the bi~he~cal properties of these isoenzymes. Firstly, the activity of 5’-NPD-I can be quantitat~ when low concentrations of the substrate are used. Under these conditions, the activity of isozyme II is negligible (Fig. 3). Secondly, 5’-NPD-II can be measured separately after fractionation of the serum on a DEAE-cellulose column (Fig. 4). And thirdly, 5’-NPD-II can specifically be measured in a serum which has been treated with n-butanol, since after this the activities I and III are lost (Fig. 6). Hawley and coworkers tried to isolate as many as possible of these isozymes detected by Tsou [12] and found only two activities hydrolyzing phosphodiester bonds in human serum. Unfortunately, they used butanol extraction as the first step of their purification procedure. Their results can be explained by our finding that butanol treatment causes the loss of two of the four serum 5’-NPD activities. Nevertheless, n-butanol treatment may be a useful tool at a later step. For example the partially purified preparation obtained from gel filtration, which contained the
284
isozymes III and IV (Fig. 5b), could be purified from 5’-NPD-III by butanol extraction. The effect of butanol extraction on the 5’-NPD isozyme activities can also be used as a tool for the selective measurement of specific isozymes in unfractionated serum. For example, the 5’-NPD activity of n-butanol-treated serum should only represent the activities of the isozymes II and IV. Furthermore, when low substrate concentrations are used, the serum activity should only be due to 5’-NPD-IV. Systematic work on differential measurements of single isozymes in unfractionated sera is in progress. The quantitation of specific isozymes in unfractionated serum without the need for gel electrophoresis would facilitate the evaluation of the clinical significance of the 5’-nucleotide phosphodiesterase isoenzymes in human sera. Acknowledgements This work was supported by the Deutsche Forschungsgemeinschaft. We gratefully acknowledge the continuous support by Prof. Dr. W. Kersten and the excellent technical assistance of Herbert BrGnner. We thank Prof. Dr. W. Domschke, Dr. Siegfried and Dr. Waldherr from the University Hospital of Erlangen-Niirnberg for supplying us with patholo~cal sera. References 1 Razzel WE. Plant tissue phosphodiesterase activities. Biochem Biophys Res Commun 1966;22:243-241. 2 Khorana HG. Phosphodiesterases. In: Boyer PD, Lardy H, Myrback K, eds. The enzymes, 2nd ed. vol V. New York; Academic Press, 1961;79-94. 3 Haugen HF, Skrede S. Nucleotide pyrophosphatase and phosphodiesterase. I. Organ distribution and activities in body fluids. Clii Chem 1977;23:1531-1537. 4 Lo KW, Ferrar W, Fmeman W, Tsou KC. Fluorometric assay of serum 5’-nucleotide phosphodiesterase in normal humans and cancer patients. Anal B&hem 1972;47:609-613. 5 Hynie J, Meuffels M, Poznanski WJ. Determination of phosph~ies~rase I activity in human blood serum. Clin Chem 1975;21:1383-1387. 6 Tsou KC, Lo KW. 5’-Nucleotide phosphodiesterase and liver cancer. Methods in cancer research, Vol. XIX. New York: Academic Press, 1982;273-300. 7 Tsou KC, Ledis S, Lo KW, McCoy MG. S’-Nucleotide phosphodiesterase isoenzyme pattern in the sera of human hepatoma patients. J Histochem Cytochem 1973;21:402. 8 Tsou KC, McCoy MC, Enterline HT, Herberman R, Wahner H. 5’-Nucleotide phosphodiesterase isoenzymes and hepatic cancer in NCI-Mayo Clinic panel sera. J Nat1 Cancer Inst 1973;51:2005-2006. 9 Lo KW, Aoyagi S, Tsou KC. A fluorogenic method for the detection of human serum 5’-nucleotide phosphodiesterase isozymes after polyacrylamide gel electrophoresis. Anal Biochem 1981;117:24-27. 10 Tsou KC, Ledis S, McCoy MC. 5’“Nucleotide phosphodiesterase isoenzyme pattern in the serum of human hepatoma. Cancer Res 1973;33:2215-2217. 11 Ltithje J, Ogilvie A. Catabolism of ApsA and Ap,A in human plasma. Purification and characterization of a glycoprotein complex with 5’-nucleotide phosphodiesterase activity. Eur J Biochem 1985;149:119-127. 12 Hawley DM, Crisp M, Hades ME. Tissue specificity of human phosph~iesterase. 1. Blood serum. Clin Chim Acta 1983:130:31-37.
Clinica Chimica Acta, 164 (1987) 275-284 Elsevier
CCA 03773
5 ‘-Nucleotide phosphodiesterase isoenzymes in human serum: quantitative measurement and some biochemical properties * Jiirgen Liithje and Adaling Institut (Received
Ogilvie
ftirPhysiologische Chemie, Universitiit Erlangen-Niimberg,
30 June 1986; revision received
Key words: 5’-Nucleotide
phosphodiesterase
12 December isozyme;
1986; accepted
Erlangen (FRG)
after revision
Phosphodiesterase;
Isoenzyme;
5 January Human
1987) serum
Summary
A method based on native PAGE is used for the quantitative measurement of 5’-nucleotide phosphodiesterase isoenzymes (5’-NPD; EC 3.1.4.1) in human serum. In contrast to other techniques this method works with a commercially available substrate. In sera of healthy donors four isozymes could be separated, designated as 5’-NPD-I, 5’-NPD-II, 5’-NPD-III and 5’-NPD-IV in the reverse order of their electrophoretic mobility. When low amounts of serum were applied to the gel, the separation between all four activities was sufficient enough to allow their quantitation. Higher amounts of serum impaired the separation between the isoenzymes I and II, However, even when using high amounts of serum, the quantitation of these activities was possible when taking advantage of some of their biochemical properties which are described herein.
Introduction
5’-Nucleotide phosphodiesterase (5’-NPD) is an exonuclease hydrolyzing DNA or RNA starting from the 3’-end to give 5’-nucleotides. The enzyme is widely
* This work is dedicated to Prof. Dr. Walter Kersten on the occasion of his 60th birthday. Abbreviations: Bis, N, N’-methylene-bis-acrylamide; DEAE-, diethylaminoethyl-; 5’-NPD, 5’-nucleotide phosphodiesterase (The isozymes of 5’-NPD were designated as 5’-NPD-I, 5’-NPD-II.. etc. Similar the expressions isozyme I, isozyme II.. . etc. or the ‘first activity’, the ‘second activity’. . etc. were used); PAGE, Polyacrylamide gel electrophoresis; PNPpT, p-nitrophenyl thymidine 5’-phosphate; S-300, Sephacryl-300; TEMED, N, N, N’, N’-tetramethyl-ethylene-diamine; Tris, Tris-(hydroxymethyl)-aminomethane. Correspondence to: Dr. Jtirgen Llithje, Institut fiir Physiologische Chemie, Fahrstrasse 17, D-8520 Erlangen, FRG.
0009-8981/87/$03.50
0 1987 Elsevier Science Publishers
B.V. (Biomedical
Division)
276
distributed in nature [1,2]. Among human organs the highest specific activities are found in kidney, pancreas, liver and small intestine [3]. In body fluids the activity was highest in seminal plasma, followed by serum, bile, cerebrospinal fluid and urine [3]. Variation of 5’-NPD-activity in human sera was reported by Lo et al using a fluorescent method [4] and by Hynie et al applying a chromogenic technique [5]. Tsou and coworkers also described the isoenzyme distribution of the 5’-nucleotide phosphodiesterase in human sera [6,7]. Normal human serum reveals four isozymes which have been identified as 5’-NPD-I, 5’-NPD-II 5’-NPD-III, and 5’-NPD-IV in the reverse order of their electrophoretic mobility. 5’-NPD-III and 5’-NPD-IV usually account for more than 80% of the total serum 5’-NPD activity [6]. In patients with primary hepatocellular carcinoma (PHC), and liver metastases an additional faster moving isozyme (5’-NPD-V) was seen [7,8]. Since then a lot of work on the 5’-NPD-V has been published showing that this isoenzyme is useful in detecting liver metastasis in various forms of cancer. These studies have recently been reviewed [6]. In contrast to the fifth isoenzyme very little information is available on the other four enzymes. They are present in sera of healthy donors and, under pathological conditions, their pattern may change. Due to the lack of a quantitative method, which can be performed in clinical laboratories, very little information about these enzymes is available. In pilot studies we have observed marked alterations of the isozyme activities I-IV in various cases of disease. This paper describes a gel electrophoretic method for the quantitative measurement of the isoenzymes and gives information on those of their biochemical properties which are helpful for their determination. Materials and methods Reagents
Acrylamide (p.A.) was purchased from Serva, Heidelberg, FRG. N, N ‘-Methylene-bis-acrylamide (Bis) and N, N, N ‘, N ‘-tetramethyl-ethylenediamine (TEMED) were from Bio-RAD Labs., Richmond, CA, USA. Ammoniumperoxidisulfat and Tris-(hydroxymethyl)-aminomethane, both of analytical grade, were purchased from Roth, Karlsruhe, FRG. n-Butanol and salts were from Merck, Darmstadt, FRG. p-Nitrophenyl thymidine 5’-phosphate (PNPpT) was obtained from Sigma, St. Louis, MO, USA. Diethylaminoethylcellulose (DE 52) was from Whatman, UK, and Sephacryl 300 (S 300) from Pharmacia, Uppsala, Sweden. Serum specimens were obtained from healthy laboratory volunteers or from the University Hospital Erlangen-Niimberg. Separation of 5’-NPD isoenzymes by gel electrophoresis
Separation of isozymes was performed by a 7% polyacrylamide gel according to the conditions described by Tsou and Lo [6]. Serum samples (50 ~1) were mixed with 10 ~1 of 99% glycerol (p.A.) and 3 11 of 450 mmol/l Tris-boric acid (pH 9.4). After centrifugation (18 000 X g, 5 rnin, 20 o C) to remove any particulate material the samples were loaded onto the gel using a microsyringe. Electrophoresis was
performed at 15 mA (constant current) for 2 h and then at 10 mA for a further 3 h. The separation was stopped when albumin (dyed with bromophenol blue) had reached the area marked on the plates (between 11 and 12 cm). The slab gel was cut into individual tracks, which were sliced transversely into pieces 3 mm in width. The gel slices were transferred into plastic caps containing 0.7 ml of Tris-HCl buffer (100 mmol/l, pH 8.6) and 4 mmol/l MgCl,. For measurement of phosphodiesterase activity the reaction was started with the chromogenic substrate p-nitrophenyl thymidine 5’-phosphate (PNPpT) at a final concentration of 4.5 mmol/l, if not indicated otherwise. Incubation was performed at 37°C under gentle rocking usually for 16 h. Absorbance was measured in a sp~trophotometer (UV-210 A, Shimadzu Seisakusho LTD., Kyoto, Japan) at 405 m. Enzyme activity was expressed as the increase in light absorbance during the incubation time. Our experiments revealed that all the 5’-NPD isoenzymes were located within the first 7 cm of the gel. Therefore, it was unnecessary to slice the whole track. Column chromatography DEAE-cellulose chromatography and gel filtration with Sephacryl 300 were performed as described with the exception that a 50 mmol/l T&s-HCl buffer (pH 8.0) was used [ll]. Chromatography was performed at 4*C. The fractions (12 and 8 ml, respectively) were assayed for phosph~i~ter~e activity using PNPpT as a substrate. The assay contained, in a total volume of 0.4 ml, 0.32 ml of the sample, 0.04 ml of Tris-HCI (1 mol/l, pH 8.6), 5 mmol/l MgCl, and substrate at 12 mmol/l. Incubation was performed at 37*C. The absorption (405 run) of eightfold diluted portions of the incubation assay were measured at various times (multipoint determination). Enzyme activity was expressed as A E,,/h. Samples obtained from colurmr chromatography were concentrated tenfold by ultrafiltration (Amicon, model 8 MC) before applying to gel electrophoresis. A Diaflo-membrane PM 10 was used, having a molecular exclusion limit of M, 10000. n-Butanol extraction 0.5 ml of the sample to be extracted was shaken with 0.25 ml of n-butanol for 1 min at room temperature (4 X 15 s, Whirl mix). After ~nt~fugation (18000 x g, 5 min, 20 o C) the cheese-like supernatant was discarded. The procedure was repeated with the aqueous phase obtained from the first extraction. After careful removal of the butanol layer the aqueous sample was used for gel electrophoresis experiments. The removal of traces of n-butanol from the sample by dialysis was found to be unnecessary.
Separation of 5’-2WD isoenzymes by nutiue PAGE and their ~nt~tat~on Serum from a healthy donor was separated on a 7% poiya~l~de gel under non-denat~a~g conditions. The gel slices were assayed for phosphodiesterase
278
E 405
a IV
i
5
15
25
4
Slice
Number
Fig. 1.5’-NPD isoenqme pattern of human serum. Serum (20 ~1) of a healthy donor (a) or of a patient with acute myeolic leukemia (b) was separated by native PAGE. The gel was sliced and 5’-NPD activity in the gel pieces was determined as described under ‘Materials and Methods’. Enzyme activity is expressed as the increase of absorbance at 405 nm during the incubation of 16 h.
activity as described under ‘Materials and Methods’. Figure la shows four isozymes which were identified as 5’-NPD-I, 5’-NPD-II, 5’-NPD-III, and 5’-NPD-IV in the reverse order of their electrophoretic mobility. The reproducibility of the isoenzyme positions in repeated gels was excellent for the 5’-NPDs I, III and IV. The first isozyme was found in slice numbers 4 or 5, 5’-NPD-III in 15 or 16 and the fourth isoenzyme in position I9 or 20. These positions referred to the peaks of the activities, respectively. 5’-NPD-II showed a greater variability, the peak of this activity was usually detected in gel slices 7, 8, 9, or 10, making it sometimes difficult to differentiate 5’-NPD-I from 5’-NPD-II. Fig. lb shows the 5’nucleotide phosphodiesterase isoenzyme pattern of a pathological serum obtained from a patient with acute myeloic leukemia. A fifth, very fast moving isozyme emerged in position 23. This 5’-NPD-V has been intensively studied by Tsou and coworkers (61. We could demonstrate that this technique was able to quantitate isozyme activities. All the four activities exhibited linear kinetics for at least 16 hours (not shown). Experiments with different amounts of serum also revealed linear relationships as far as separation was suffi~~t. Figure 2 demonstrates that the activities clearly separated when 8 or I6 nl of serum were used. Higher amounts, however, impaired
279
Slice
Number
Fig. 2. Effect of the amount of the serum applied to the gel on the 5’-NPD isozyme pattern. The amounts of serum are given in the boxes. Enzyme activity is expressed as described in the legend of Fig. 1.
the separation between the two slow moving activities (I and II), whereas the 5’-NPD isozymes III and IV remained well distinguishable. Concerning the influence of storage conditions on isozyme activities, experiments with sera from several donors revealed different stabilities of the four isoenzymes during storage at - 20 ’ C. Whereas the activities of 5’-NPDs I, III and IV remained stable for months, isozyme II lost part of its activity. Especially, after repeatedly freezing and thawing of the serum, a severe loss of this activity was observed. The other isozymes revealed no significant alterations.
The usual isoenzyme pattern of human serum, when incubation with substrate was performed at a concentration of 4.5 mmol/l, is shown in Fig. 3b. Reduction of the PNPpT concentration to one tenth (0.45 mmol/l) caused the apparent loss of the isozyme II activity (Fig. 3a). Increasing the substrate concentration up to 15 mmol/l resulted in a huge activity peak (Fig. 3~). The dependence of the elution profile on substrate concentration is obviously due to K, differences. A kinetic analysis of 5’-NPD-II after purification from the other isozymes by means of DEAE-cellulose chromatography (Fig. 4) revealed a K, of 30 mmol/l. Since the variation of the substrate concentration in the range from 0.5 to 15 mmol/l had no influence on the activities I, III and IV, we suggest that their Michaelis constants are
280
&OS 0.5
5
15
St ice
25
Number
Fig. 3. Influence of substrate concentration on the isoenzyme pattern. In three experiments, serum (20 ~1) from the same donor was separated by get electrophoresis. The gel slices were incubated with (a) 0.45 mmol/l, (b) 4.5 mmol/l, (c) 15 mmol/l of pnitrophenyl these 5’-phosphate. Incubation was performed for 16 h and activities are expressed as in Fig. 1.
smaller than 0.5 mmol/l. It is obvious that the use of the substrate at a low concentration is an useful way for the quantitation of 5’-NPD-I, in spite of its moderate separation from the second isozyme.
I
1
10
m
Fraction
30
LO
50
60
I
Number
Fig. 4. DEAE-cellulose chromatography of human serum. Serum (30 ml) was diluted fivefold with elution buffer and applied to a DF.AE-cellulose column. Elution was performed with a linear salt gradient. Fractions were assayed for 5’-NPD activity (0 -0). Enzyme activity is the increase of light absorbance (405 nm) during the incubation time of 1 h. Assay conditions are given under ‘Materials and Methods’.
281
Separation of serum 5’-NPD isozymes by anion exchange and gel filtration chromatography
Fractionation of human serum on a DEAE-cellulose column resulted in the separation of two 5’-NPD activities (Fig. 4). One activity did not bind to the colmmr (‘non-binding activity’), whereas the other had a strong affinity to the anion exchanger and was only eluted with NaCl at SO-120 mrnol/l (‘binding activity’). Both enzyme preparations were characterized by native gel electrophoresis. Figure 5a demonstrates that the non-binding phosphodiesterase could be attributed to the 5’-NPD-II, since the maximum activity was found in gel position 9, being typical for the second isozyme. The fact that two samples from different regions of the non-binding activity (fractions 7,8 and fraction 17, see Fig. 4) revealed an identical migration behavior on the gel (not shown), suggested that this activity was electrophoretically homogeneous. The characterization of the binding activity, however, displayed three activities on the gel. They were localized in typical positions of the 5’-NPD isoenzymes I, III, and IV, respectively (Fig. 5a). By using gel filtration we could separate a tiny peak (which eluted first) from a huge phosphodiesterase activity. Gel electrophoresis revealed that the small activity
E LO5
2 4
2
b 0.5
0.2
5
15
25
SI ice Number Fig. 5. Ch~acte~ation of 5’-NPD activities from the DEAE-ceIlulose and gel f&ration steps by native PAGE. Concentrated samples from the cohunn chromato~aphy steps were applied to gel electrophoresis. 5’-NPD activities in the gel slices were measured as described in the text. Enzyme activity is expressed as described in the legend of Fig. 1. a. DEAE-cellulose step (Fig. 4): aiiquots from fractions 7 and 8 of the ‘non-binding activity’ (0 0) and from fractions 29 and 30 of the ‘bindiug activity’ (ol) were applied onto different slots of the native gel. b. Sephacryl-300 step: aliquots from the small peak 0) and from the major activity (O0) were applied onto different slo& of the native gel. (O-
282
could be attributed to the 5’-NPD-I, whereas the other enzyme activity consisted of two isoenzymes (III and IV, Fig. 5b).
still
Effect of n-butanol treatment on the 5’-NPD isozyme pattern
The 5’-NPD isozyme pattern of butanol-treated serum was compared with that of untreated serum (Fig. 6a). Butanol extraction of serum caused the loss of the 5’-NPD-III activity, whereas the fourth isozyme was still detectable at its typical position. The slow moving enzymes I and II were not distinguishable (Fig. 6a). To find out more about the fate of these isozymes, we repeated the experiment with a slight modification. The enzyme activity was measured at a lower substrate concentration. Under this condition the activity of 5’-NPD-II is negligible (see Fig. 3). After butanol extraction the isozyme activities I and III were lost (Fig. 6b). The findings of both experiments suggested that the 5’-NPD-II activity was not affected by the extraction procedure. These observations were confirmed by experiments with partially purified isoenzymes. Serum was fractionated by DEAE-cellulose chromatography (Fig. 4). The first, non-binding activity (5’-NPD-II) was not affected by the butanol treatment. However, after extraction of the binding activity, the isozymes I and III disappeared. 5’-NPD-IV was not influenced by this procedure.
ELOS0.7 -
s .> ..I-
flL -
y
0.1 -
z e
5
15
25
15
25
aJ z
0.9
5 2 $
0.6
E
5
ice Number Fig, 6. Effect treatment on 5’-NPD isoenzyme pattern of serum. The extraction is under ‘Materials and a. of butanol-treated Gel pieces were incubated with substrate at a 0) and of untreated serum (0 -0). serum (Oconcentration of 4.5 mmol/l. b. A similar experiment as in a, however, with substrate at 0.45 mmol/l.
283
Discussion
In this paper we demonstrate the qualitative and quantitative measurement of the 5’nucleotide phosphodiesterase (5’-NPD) isoenzymes. After gel electrophoresis we have detected the isozyme activities by using the commercially available, chromogenic substrate p-nitrophenyl thymidine 5’-phosphate (PNPpT). To overcome the problem of product diffusion of the soluble p-nitrophenol in the gel we have applied a gel slice technique, which ahowed the incubation of gel pieces for times long enough to detect even weak enzyme activities. In contrast to this method, other workers used substrates which have to be chemically synthesized. The indigogenic method described by Tsou et al [7] is based on the production of an insoluble, stable indigo dye, so that whole gel rods can be incubated for long times without having the problem of product diffusion. The product of another very sensitive, fluorescent technique, 4-methylumbelliferone, is soluble. Therefore, only short incubation times are possible, as long as whole rods or lanes are incubated [9]. In combination with a slice technique this method would be very powerful, however with the disadvantage that the substrate is not commercially available (to our knowledge) and that special equipment (fluorometer) is needed. Our method works with a standard spectrophotometer. We have detected four 5’-NPD isoenzymes in sera from healthy donors. This is in good accordance with the findings of Tsou and coworkers [6]. Under certain pathological conditions, especially in malignant states of the liver, Tsou described the emergence of a fifth, fast-moving isoenzyme [L&10], which has also been confirmed by our results. Furthermore there are marked changes in the pattern of the other isoenzymes. The linearity of the kinetics as well as the experiments with variable serum concentrations demonstrate that the technique is able to determine the activities of the isozymes III and IV, respectively. The activities of 5’-NPD-I and -11 can be measured as long as low concentrations of serum are applied. Higher concentrations lead to unsufficient separation of these activities (Fig. 2). Even when higher serum concentrations are required, it is possible to differentiate between the isozymes I and II by taking advantage of the bi~he~cal properties of these isoenzymes. Firstly, the activity of 5’-NPD-I can be quantitat~ when low concentrations of the substrate are used. Under these conditions, the activity of isozyme II is negligible (Fig. 3). Secondly, 5’-NPD-II can be measured separately after fractionation of the serum on a DEAE-cellulose column (Fig. 4). And thirdly, 5’-NPD-II can specifically be measured in a serum which has been treated with n-butanol, since after this the activities I and III are lost (Fig. 6). Hawley and coworkers tried to isolate as many as possible of these isozymes detected by Tsou [12] and found only two activities hydrolyzing phosphodiester bonds in human serum. Unfortunately, they used butanol extraction as the first step of their purification procedure. Their results can be explained by our finding that butanol treatment causes the loss of two of the four serum 5’-NPD activities. Nevertheless, n-butanol treatment may be a useful tool at a later step. For example the partially purified preparation obtained from gel filtration, which contained the
284
isozymes III and IV (Fig. 5b), could be purified from 5’-NPD-III by butanol extraction. The effect of butanol extraction on the 5’-NPD isozyme activities can also be used as a tool for the selective measurement of specific isozymes in unfractionated serum. For example, the 5’-NPD activity of n-butanol-treated serum should only represent the activities of the isozymes II and IV. Furthermore, when low substrate concentrations are used, the serum activity should only be due to 5’-NPD-IV. Systematic work on differential measurements of single isozymes in unfractionated sera is in progress. The quantitation of specific isozymes in unfractionated serum without the need for gel electrophoresis would facilitate the evaluation of the clinical significance of the 5’-nucleotide phosphodiesterase isoenzymes in human sera. Acknowledgements This work was supported by the Deutsche Forschungsgemeinschaft. We gratefully acknowledge the continuous support by Prof. Dr. W. Kersten and the excellent technical assistance of Herbert BrGnner. We thank Prof. Dr. W. Domschke, Dr. Siegfried and Dr. Waldherr from the University Hospital of Erlangen-Niirnberg for supplying us with patholo~cal sera. References 1 Razzel WE. Plant tissue phosphodiesterase activities. Biochem Biophys Res Commun 1966;22:243-241. 2 Khorana HG. Phosphodiesterases. In: Boyer PD, Lardy H, Myrback K, eds. The enzymes, 2nd ed. vol V. New York; Academic Press, 1961;79-94. 3 Haugen HF, Skrede S. Nucleotide pyrophosphatase and phosphodiesterase. I. Organ distribution and activities in body fluids. Clii Chem 1977;23:1531-1537. 4 Lo KW, Ferrar W, Fmeman W, Tsou KC. Fluorometric assay of serum 5’-nucleotide phosphodiesterase in normal humans and cancer patients. Anal B&hem 1972;47:609-613. 5 Hynie J, Meuffels M, Poznanski WJ. Determination of phosph~ies~rase I activity in human blood serum. Clin Chem 1975;21:1383-1387. 6 Tsou KC, Lo KW. 5’-Nucleotide phosphodiesterase and liver cancer. Methods in cancer research, Vol. XIX. New York: Academic Press, 1982;273-300. 7 Tsou KC, Ledis S, Lo KW, McCoy MG. S’-Nucleotide phosphodiesterase isoenzyme pattern in the sera of human hepatoma patients. J Histochem Cytochem 1973;21:402. 8 Tsou KC, McCoy MC, Enterline HT, Herberman R, Wahner H. 5’-Nucleotide phosphodiesterase isoenzymes and hepatic cancer in NCI-Mayo Clinic panel sera. J Nat1 Cancer Inst 1973;51:2005-2006. 9 Lo KW, Aoyagi S, Tsou KC. A fluorogenic method for the detection of human serum 5’-nucleotide phosphodiesterase isozymes after polyacrylamide gel electrophoresis. Anal Biochem 1981;117:24-27. 10 Tsou KC, Ledis S, McCoy MC. 5’“Nucleotide phosphodiesterase isoenzyme pattern in the serum of human hepatoma. Cancer Res 1973;33:2215-2217. 11 Ltithje J, Ogilvie A. Catabolism of ApsA and Ap,A in human plasma. Purification and characterization of a glycoprotein complex with 5’-nucleotide phosphodiesterase activity. Eur J Biochem 1985;149:119-127. 12 Hawley DM, Crisp M, Hades ME. Tissue specificity of human phosph~iesterase. 1. Blood serum. Clin Chim Acta 1983:130:31-37.