- Email: [email protected]
Human seminal deoxyribonuclease I (DNase I): purification, enzymological and immunological characterization and origin
ELSEVIER
Clinica Chimica Acta 218 0993) 5-16
Human seminal deoxyribonuclease I (DNase I): purification, enzymological and immunological characterization and origin Toshihiro Yasudaa, Kazumi Sawazaki a, Daita Nadano a, Haruo Takeshita a, Masao Nakanaga b, Koichiro Kishi *a aDepartment of Legal Medicine. bDepartmentof Internal Medicine and Medical Genetics, Fukui Medical School. Matsuoka-cho, Fukui 910.1!, Japan
(Received 8 October 1992; revi~iionreceived 8 March 1993; accepted 12 March 1993)
Abstract Deoxyribonuclease 1 (DNase 1) was purified fre~ the semen of a 3S-year-old male and then characterized. The catalytic properties of the purified enzyme closely resembled those of DNase I purified from the urine of ~his individual and the following other similarities were observed: molecular masses, iodoacetic acid inactivation kinetics, desialylated isoenzyme patterns. However, the behavior of the purified enzymes determined on several different lectinaffinity chromatography columns differed, which suggests that organ.specific glycosylation of DNase I occurs. Multiple forms of the purified seminal DNase ! were demonstrated, each of which had a different pl value seperated by isoelectric focusing, which is compatible with the reported existence of genetic polymorphism of seminal DNase ! (Sawazaki et al., Forensic Sci lnt 1992;57:39-44). Furthermore, enzymological and immunological comparisons of purified seminal and urinary and partially purified prostatic DNases 1 indicated that the prostate may be one of seminal enzyme source tissues. Key words: Deoxyribonucleas¢; Human semen; Purification; Multii)licity; Prostate
1. Introduction Deoxyribonuclease I (DNase I, EC 3.1.21.1) is located preferentially in the pancreas, liver and kidney [1-3]. The value of serum DNase I activity has been reported to be useful in predicating the therapeutic efficacy in patients with cancer [4]. It has * Corresponding author. 0009-8981/93/~ ~.00 © 1993 ElsevierSciencePublishers B.V. All rights reserved. SSD! 0009.898i(93)05556-T
6
7". Yasuda et al./Clin. Ch#n. Acta 218 (1993) 5-16
been suggested that the enzymatic activities of DNase I in vivo are important for DNA metabolism in view of the hydrolytic activity and presence of DNase I in both duodenal and pancreatic juices. However, the distribution of DNase I activity in other tissues raises doubts that its major physiological role i~ a digestive function [5]. In particular, it is worth emphasizing that a human urine-derived interleukin-I i~lhibitor is functionally diverse from, but highly homologous with, DNase I [6]. Therefore, the extra- and intracellular functions of DNase I, as well as the digestive role, are of interest and should be elucidated. The isoenzyme patterns of human serum and urinary DNase I from different individuals were separated clearly into ten groups, which shows the existence of genetic polymorphism [2,7]. In our preliminary survey, the seminal DNase I type of a man was found to correlate with his serum and urinary enzyme types [8]. Deoxyribonuclease activities have been detected in human seminal plasma [9], but, to our knowledge, there is little biochemical information about seminal DNase I. In this report, we describe the purification and characterization of the seminal and prostatic DNase I enzymes, their comparison with the urinary enzyme and the origin of seminal DNase I. 2. Materials ~ad methods
Z !. Chemicals and biological materials All the chemicals used were of reagent grade or the purest grade available commercially. DEAE-Sepharose CL.6B, phenyl-Sepharose CL-4B and Sephadex G-75 were obtained from Pharmacia LKB (Uppsala, Sweden); hydroxyapatite and polyethylene glycol 6,000 (PEG) were from Nacalai Tesque (Kyoto, Japan); CIostridium perfringens sialidase, iodoacetic acid, phenylmethylsulfonylfluoride (PMSF) and soybean trypsin inhibitor (SBTI) were from Sigma (St. Louis, Me); agaroses conjugated with concanavalin A (Con A), lentil lectin (LCA), caster bean iectin (RCAI20) and wheatgerm lectin (WGA) were from Seikagaku-Kogyo (Tokyo, Japan). The human semen and urine used for purification of the enzyme were collected from a healthy 38-year-old male (DNas© 1 phenotype I). The spermatozoa were removed by centrifugation at 3,000 × g for 5 rain and the resulting seminal plasma was stored at -80°C until required for use. Donors supplied samples obtained by masturbation or by withdrawal during sexual intercourse. The ejaculate was collected in clean, wide-mouthed plastic containers with screw caps. These samples were presented at the laboratory within 12 h of collection and immediately frozen at -80°C until analyzed. Human urinary DNase I was purified as described previously [10,11]. Rabbit antibodies against human urinary DNase ! (anti-DNase I) and spleen DNase I! (anti-DNase II) were prepared as described previously [11,12]. A piece of prostate was obtained at autopsy from a 20-year-old drowned male within 20 h postmortem and stored at -80°C until required for use. 2.2. DNasc I activity assay and other analytical methods Deoxyribonuclease I activity was determined using the single radial enzyme diffusion (SRED) method as described previously [3,12], One unit of the DNase I activity was defined as described previously [I 1].
T. Yasuda et al. / Clin. Chim, Acta 218 (1993) 5-16
7
The DNase 1-active fractions were subjected to polyacrylamidegel electrophoresis in the presence of 0.1% (w/v) sodium dodecyl sulfate (SDS-PAGE) using a 12% polyacrylamide gel, as described by Laemmli [13], and the enzyme was detected by immunoblotting with anti-DNase I [141. The DNase I was also subjected to 7.5% PAGE followed by activity staining, as described previously [3]. The final preparation was also assayed for other enzymes as follows: DNase II, two types of human ribonuclease (RNase) and phosphodiesterase I were measured as described by Yasuda et al. [12], Mizuta et al. [!5] and Ito et al. 116], respectively. 2.3. Purification of DNase I from human seminal plasma and human prostate All the purification steps were carried out at 4°C. The seminal plasma (12 ml) was mixed with 20 ml 25 mmol/l Tris-HCl buffer, pH 7.5 (buffer l), which contained 10 manol/I CaCI2 and i mmol/l PMSF, and was then dialyzed against this buffer. The dialyzed materials were applied to a DEAE-Sepharose CL-6B column (1.6 x 30 cm) pre-equilibrated with this buffer, and the adsorbed materials were eluted with a NaCI linear gradient (0-1.0 tool/i) in the same buffer at 30 ml/h (fraction size, 28 ml). The pooled active fractions were dialyzed against 10 mmol/I Tris-HCi buffer, pH 7.5, which contained 1 tool/! ammonium sulfate, followed by dialysis against 10 mmol/i potassium phosphate buffer, pH 6.7 (buffer ll), which contained 1 mol/l ammonium sulfate. The dialyzed materials were applied to a phenyl-Sepharose CL,4B column (1.6 x 15 cm) pre-equilibrated with this buffer and the adsorbed materials were eluted first with an ammonium sulfate linear reverse gradient (I.0-0 tool/i) in buffer I1 and then with buffer II alone at 20 ml/h (fraction size, 18 ml). The activefractions eluted with buffer II alone were collected and dialyzed against 5 mmol/l potassium phosphate buffer, pH 6.7, and applied to a hydroxyapatite column (1.0 x 10 cm) pre-equilibrated with this buffer, after which the column was washed with 100 ml buffer II at 15 ml/h (fraction size, 7.5 ml). Most of the enzyme activities passed through the column and the eluate was concentrated with PEG; the concentrated materials were dialyzed against buffer I, which contained 0.25 tool/! NaCI, and then subjected to gel filtration on a Sephadex G-75 column (!.6 x 90 cm? preequilibrated with this buffer, which also contained 1 mmoi/I CaCI2. Five-milliliter aliquots of each were collected at 7 ml/h. This gel filtration procedure wa~ repeated and the DNase I-active fractions obtained were collected and dialyzed against 20 mmol/l sodium phosphate, buffer, pH 7.2 (buffer Ill), which contained 0.15 mol/I NaCI. The dialyzed materials were applied to a Con A-agarose column (I x 5 cm) pre-equilibrated with this buffer and the adsorbed enzyme was eluted with 0.2 mol/I methyl et.D-mannopyranoside in this buffer at 7.5 ml/h (fraction size, 5 ml). The DNase I-active fractions were pooled and dialyzed against buffer l, which contained 1 mmol/l CaCl2, and then concentrated by ultrafiltration for use in the subsequent experiments described below. A human prostate (about 2.5 g) was cut into small pieces and homogenized in 10 ml buffer I, which contained 10 mmol/i CaCI2, 2.5 mmol/l PMSF and 30 mgJl SBTI, centrifuged at 15,000 x g for 20 rain and the supernatants were dialyzed against this buffer without SBTI. From the dialyzed materials, the prostate DNase I was purified by chromatography on successive DEAE-Sepharose CL-6B, phenyl-Sepharose CL4B and hydroxyapatite columns under conditions similar to those described above. The final passed-through fractions from the hydroxyapatite column were dialyzed
8
T. Yasuda et aL/Ciin. Chim. Acta 218 (1993) 5-16
against buffer I, which contained 1 mmol/! CaCI2, concentrated by uitrafiltration and then characterized. 2.4. Inhibitory effects of anti-human urinary DNase I and inactivation by iodoacetic acid The purified enzyme (about 5 x 10-3 units) in 50/~l 25 mmol/l potassium phosphate buffer, pH 6.7, and 50-;d aliquots of diluted seminal plasma with the same activity were mixed with different amounts of the IgG antibody fraction, and kept at 4°C overnight. The samples were centrifuged at 10,000 × g for 10 rain, and the DNase i activity remaining in each supernatant was detected by the zymogram method after PAGE, as described above. Inactivation of the enzymes by iodoacetic acid was performed essentially according to the method of Price et al. [17]. The enzyme activities remaining were determined by the SRED method. 2.5. Seminal DNase I detected by isoelectricfocusing and the zymogram method Isoelectric focusing in a thin layer of polyacrylamide gel (IEF-PAGE, pH 3.5-5) was performed as described previously [18,19], after which the DNase I was detected using the dried agarose film-overlay (DAFO) method [18,20]. The enzymes were treated with an equal volume of sialidase (10 units/ml in 50 mmol/I sodium acetate buffer, pH 5.0) as described previously [18]. 2.6. Affinity of seminal DNase l for lectin columns The purified enzyme (about 2 x 10-2 units) was dialyzed against buffer !11, which contained 0.15 mol/I NaCI. The dialyzed materials were applied to separate lectin-agarose columns (I ml each) with Con A, LCA, RCAI20 and WGA, which were all equilibrated with this buffer, washed with 5 ml this buffer (unbound fraction) and the adsorbed enzyme was eluted with 5 ml 0.2 mol/i methyl or.Dmannopyranoside, 0.1 mol/I D-lactose, 0.2 mol/l D-galactose and 0.2 mol/I N-acetyl. D-glucosamine, respectively in this buffer (bound fraction). The enzyme activities in these fractions were determined using the SRED method. 3. Iteselts
3.1. Purification of DNase ! from human seminal plasma The yield and purification were estimated to be about 33% and in excess of 8,485. fold, respectively (Table I). The final product yielded a single band when subjected to SDS-PAGE under reducing conditions when detected by immunostaining with anti-DNase I, but one major and several minor bands were detected by silverstaining. No DNase 11, secretory- or non-secretory-type RNase or phosphodiesterase 1 activities were detected in the final preparation. The final preparation migrated as a single band both in substrate-cast gel under non-reducing conditions and when subjected to SDS-PAGE followed by immunostaining detection with anti-DNase I under reducing conditions. Moreover, it showed electrophorctic mobilities similar to those of the urinary enzyme purified from the same individual when analyzed with the PAGE and SDS-PAGE techniques (~:ig. i).
T. Yasuda el ai./ Clin. Chim. Acta 218 (1993) S-16 Table 1 Summary of the procedures for purification of DNase I from human seminal plasma Purification step
Seminal plasmac I. DEAE-SepharoseCL-6B 2. PhenyI-SepharoseCL-4B 3. Hydroxyapatite 4. First Sephadex G-75 5. Second Sephadex G-75 6. Concanavalin A-agarose
Proteina (rag)
103.880 8.710 1.231 0.090 0.034 0.010 <0.004 d
Activityb
Purification (fold)
Total activity (units)
Specific Yield activity (%) (units/mg)
1.750 1.730 1.506 1.018 0.923 0.819 0.577
0.017 0.199 1.233 11.311 27.147 81.900 > 144.250
100 98.9 86. I 58.2 52.7 46.9 32.9
1.0 I 1.7 72.5 665.4 1,596.9 4,817.6 >8,485.3
aProtein concentration was determined using the Bio-Rad protein assay kit with bovine serum albumin as a standard. bEnzyme activity was measured by the SRED method as described in the text. CStarting from 12 ml of semen collected from a 38-year.old male. dprotein concentration was below the sensitivity of the method. The seminal DNase I had a molecular mass of 38.0 kDa determined by SDSPAGE under reducing conditions and of 34.0 kDa by gel filtration on a Sephadex G-75 column in its native state. These masses were found to be virtually identical to those determined for urinary DNase l.
3,2. Catalytic properties in the presence of both 10 mmol/I MgCI2 and I mmol/I CaCI2, the pH activity profile of the seminal DNase i exhibited a bell-shape with an optimal pH of about 6,8, with no activity below pH 4.5 or above pH 9.5. The effect of ionic strength on the enzyme activity was also examined using SRED at pH 6.5 in the presence of both 10 mmol/I MgCI2 and I mmol/I CaCI2 by adding NaCI, The activity decreased markedly as the ionic strength increased and almost disappeared when it reached 0.5. The requirements for various divalent cations (5 mmol/I as the chloride form) were also examined at pH 6,5, The enzyme was active in the presence of Co 2+, Mn 2+ or Mg 2+, but not in the absence of divalent cations; 10 mmol/I EDTA, I mmol/I EGTA and 100 mg/I G-actin abolished the activity, even in the presence of 10 mmol/I MgCI 2. The enzyme hydrolyzed native DNA three to four times faster than thermally denatured DNA. These catalytic properties of the seminal DNase I were very similar to those of the urinary enzyme purified from the same individual. The similar pH dependencies of the activities of the purified seminal and urinary enzymes are shown in Fig. 2. 3.3. Inhibitory effects of anti-human urinary DNase I and inactivation by iodoacetic acid The DNase I activity (about 5 × 10 -3 units) of the purified enzyme and seminal plasma were abolished by less than 50 ~g lgG derived from 5 ~l anti-urinary DNase
T. Yasuda et al./ Clin. Chim. ,4eta 218 (1993) 5-16
IO
C
B
A kDa 80.0 ""
49.5
IBm
32.5 27.5 ~"
1
1
2
1
2
Fig. !. Electrophoretic patterns of purified seminal (lane I) and urinary DNases I (lane 2). The SDSPAGE patterns detected by silver-staining (A) and immunoblotting (B). Silver-staining was accomplished using Silver Stain DC (Daiichi Pure Chemicals, Tokyo). After electrophoresis on 12% polyacrylamide gel, the protein was transferred onto a Durapore membrane (Millipore, Bedford, MA) by electroblotting and visualized by immunostaining with anti-DNase I, as described previously [14,18]. The molecular weights indicated in the figure were estimated using a prestained SDS-PAGE standard kit (Bio-Rad). (C) The PAGE patterns detected by the activity staining. After electrophoresis on 7.5% polyacrylamide gel, DNase ! activity was detected as previously described [3]. The arrow indicates the bromophenol blue migration position.
i I
!
I
I
!
I
I
IO0
so
o
60
pH
80
-,o:o
Fig, 2. Typical demonstration of the similar catalytic properties of the seminal (@), prostatic (A) and urinary (E~ DNases I and the effect of,~H on their activities. The pH dependency in the presence of both 10 mmol/I MECI2 and I mmol/I CaCI: was investigated using the SRED method. The reaction buffers (0.1 mol/I each) used were sodium acetate (pH 4.5-5,5), sodium cacodylate (pH 6,0-7.0), Tris-HCI (pH 7.5-9,0) and sodium glycine (pH 9.5). The activity is expressed as a percentage of that at optimal pH (6.5).
Ii
7". Yasuda et aL / Clin. Chim. ,4cta 218 (1993) ~-!6
~2
'
'
A
"5
0__.1 I
0
30
I
I
60 0 30 Reaction time (rain)
I
60
Fig. 3. The kinetics of the seminal (A) and urinary (B) DNase I activities by iodoacetic acid. The purified enzyme was incubated with 10 mmol/I (O) and 20 mmol/I (O) iodoacetie acid in the presence of 10 mmol/I CuCI2 at 37°C. The conditions and determination of the remaining activities are described in the text. The logarithm of the percentage of remaining activity was plotted against the reaction time and the leastsquare line was drawn.
I, but not by that from anti-DNase I1. These results indicate that the seminal DNase I has strong immunological reactivity with antibody to the urinary enzyme. The seminal DNase I was inactivated rapidly by 10 mmol/i iodoacetic acid in the presence of 10 mmol/I CuCI2 at 37°C (Fig. 3). When the logarithm of the percentage of remaining activity was plotted against the reaction time, a straight line was obtained, which indicates that the iodoacetic acid inactivation reaction obeyed firstorder kinetics and the first-order rate constant for the reaction was calculated to be 0.026 min "1. A similar plot was also obtained with the purified urinary DNase I, which indicates that the iodoacetic acid inactivation kinetics of the two enzymes were essentially identical. The inactivation reaction of bovine pancreatic DNase i reported by Price et al. [17] was similar to our results.
£4. Seminal DNase ! isoenzymes on IEF-PAGE gel detected by the zymogram method The purified seminal DNase I was separated into several bands with different pi values on the gel region corresponding to pH 3.5-4.0 (Fig. 4). Desialylation of the enzyme diminished the anodal bands and shifted them towards the cathode. The major bands of the purified seminal and urinary enzymes before and after sialidase digestion were similar. These findings indicate that variations in the sialic acid content of the seminal DNase I contribute in part to the charge multiplicity of the enzyme. 3.£ Lectin affinities of the seminal and urinary DNases I The affinities of the purified seminal and urinary enzymes for four different iectins were determined (Table 2). The strongest affinity of enzymes was observed with Con
12
7". Yasuda et al./ Clin. Chim. Acta 218 (1993) 5-16
B
A pl 3 . 7 5 ='~
e~
O O
4.15=,.-
Q 1
2
t, I
1
2
Fig. 4. The ]EF-PAGE patterns of purified seminal and urinary DNases ! before (A) and after (B) treatment with sialidase, The separated DNase 1 isoenzymes were detected by activity staining using the DAFO method described in previous papers [18,20]. The pl values indicated in the figure were estimated using low pl calibration kit (Pharmacia LKB), Lane I, seminal DNase l; lane 2, urinary DNase !, The anode is at the top,
A, to which 88% of the seminal and 71% of the urinary enzyme bound, whereas they exhibited low affinities (5-43% bound) for the other lectins. However, the affinities of the seminal DNase ! for several lectins were found to differ from those of the urinary enzyme, which indicates that some of the carbohydrate moieties of these enzymes differ.
Table 2 Binding of ~emen. and urine.derived DNase ! to lectin.atZ]nity columns Lectinoaprose
Con A-agarose LCA-agarose RCAi20-agarose WGA-agarose
Bound fraction' (%) Seminal DNase I
Urinary DNase I
87,8 30,7 (32.7) b 43,2 6.5 (8,3) b
70,5 5,2 (13.5) b 26,8 24,9 (23.5)b
The seminal enzyme fractions eluted From the second gel filtration on Sephadex G-75 were applied to four different lectin-agarose columns, The urinary enzyme was purified from the urine collected from the same individual and subjected to the same column chromatographic series as the seminal enzyme. Thirty-five microliters of each enzyme solution (about 4 x 10-2 units) was treated with 35/AI sialidase (10 units/ml) at 37°C overnight and then applied to each column, The chromatographic conditions and the activity assays for the unbound and bound fractions are described in Materials and Methods. 'Binding of the enzyme activity to lectin-agarose is expressed relative to the total activity. bValues in parentheses are those for the desialylated enzymes,
T. Yasudaet al./Clin. Ckim. Acta 218 (1993) 5-16
13
3.6. DNase I activity levels in semen samples and their correlation with phenotypes The DNase I types determined from semen samples have been correlated with the types determined from corresponding blood and urine samples [8]. In order to clarify whether the DNase I activity levels in semen correlated with the phenotypes, the activities in semen samples from 45 unrelated healthy Japanese men of three major type,,, (1, 1-2 and 2), which comprise about 98% of the Japanese population [2], were quantified using S R E D (see the details in the legend to Table 3). The activities of semen samples collected periodically from the same men at separate times were also quantified and the levels were found to be constant with a mean intra-individual variation with a standard deviation of about 18%. Furthermore, as the activity levels in semen samples preincubated at 37°C for a few hours remained virtually unchanged, proteases that may be present in seminal plasma may not affect DNase I inactivation under the conditions used in this study. The DNase I activity levels of the three groups (types 1, 1-2 and 2) were determined and there were no significant differences among them by testing with one-way analysis of variance (Table 3). 3. 7. Partial purification and characterization o f human prostatic DNase I Human seminal plasma contains the combined secretions of the accessory glands and of these the prostate has been demonstrated to contribute substantially to the seminal plasma composition [22]. The DNase I activity of the crude extract from human prostate was lower than that of the semen and the content of the enzyme was estimated to be 2.92 × 10 -4 units/g wet weight (2.51 × 10 -4 units/mg protein). The activity of the partially purified prostatic DNase I was equivalent to 25% of the total activity o f the crude extract and about 100-fold purification was achieved. The partially purified enzyme demonstrated optimal activity at about pH 6.5 in the presence of 10 mmol/I MgCI2 and l mmol/i CeCI2 and no activity below pH 4.5 or above pH 9.5 (Fig. 2). At pH 6.5, the enzyme was active in the presence of either
Table 3 The distribution of DNase I activity levelsa in semen samples from ,*5 individualswith three different phenotypes Phenotype
Total no. in group
Range (× 10-2 units/mg)
Mean value .*- S.D, (× 10-2 units/mg)
I I-2
15 15
2 Total
15 45
1.21-7.78 1.65-4.67 1.13-8.10
3.34 ± 2.09 3.19 4. 0.96 3.09 4. 1.90
1.13-8.10
3.21 4. 1,71b
After centrifugation, the resulting supernatants from semen samples were diluted 40.fold with 50 mmol/I Tris.maleate buffer (pH 6,5) containing I0 mmol/I MgCI 2 and ! mmol/I CaCI2. Enzyme activity of each
diluted sample was measured by the SRED method. All assays were performedat least in triplicate. The DNase I of each sample was phenotyped as described previously [8,18]. aDNase ! activities of each sample are expressed as units/rag protein. bMean value (units/rag protein) of all the samples corresponded to 354.0 -)- 171.0 units/I semen.
14
7". Yasuda et al. / Clin. Chim. Acta 218 (1993) 5-16
MgCI2, MnCI2 or COC12 (5 mmol/l each) and activities of the purified enzyme and crude extract were abolished by anti-DNase l, but not anti-DNase II. These characteristics were very similar to those of the urinary and seminal enzymes and these findings strongly suggest that the prostate is one of the source tissues of the seminal DNase I. 4. Discussion
Although Quinn first reported the existence of DNases ! and !I in human semen in 1968 and DNase was suggested to cause a rapid decrease in the motility of human spermatozoa |9], these enzymes have not been characterized in detail. Therefore, we have attempted to purify and characterize the DNase I from human seminal plasma. Multiplicity of bovine pancreatic DNase I has been studied extensively [23] and that demonstrated by human urinary DNase ! has been suggested to be attributed to variations in the primary structure and/or differences in the sialic acid content [I !]. In general, in order to elucidate the molecular multiplicity, heterogeneity and genetic control of an enzyme, purificatiori from the enzyme source collected from a single individual is essential. We have established that, in the case of urinary DNases I and II, enzyme preparations from a single individual were needed to investigate their biochemical and genetic multiplicities [11,12]. Therefore, our seminal DNase ! was obtained and purified from the semen collected from an individual. The purified enzyme and original semen from this individual yielded very similar patterns consisting of several bands with different pl values when analyzed using ! EF-PAGE and zymogram detection before and after sialidase digestion (Fig. 4). it is clear, therefore, that this multiplicity results from intra- and not inter-individual variation, and the assumption that the multiplicity of sem'nal DNase ! may be a result of partial proteolyti¢ degradation of the intact enzyme, as suggested by Polakoski et al. [22] is not credible. The urinary and seminal DNases I from this individual exhibited the same chromatographic behavior, mobilities with a si,lgle band when subjected to PAGE with activity stainin~ and SDS-PAGE with immunostaining, catalytic properties, inhibitory effects of G-actin and anti-DNase I, molecular masses and iodoacteic acid inactivation kinetics. These findings indicate that seminal and urinary DNases I may be the same gene products; however, the slight differences between them, which the lectin-affinity analyses revealed (Table III), may be caused by organ-specific post-translational modifications, which has been observed with human secretory-type RNases [16,24-26]. As both enzymes originated from the same individual, the difference could not be attributed to inter-individual variation. Yamashita et al. [27] indicated that the behavior of RNases subjected to various types of lectin-affinity chromatography provided valuable information about organspecific glycosylation. Our lectin-affinity analysis results of the seminal and urinary enzymes suggest that seminal DNase I may undergo glycosylation to a different extent and/or in a different manner from the urinary enzyme.
$. Acknowledgments We thank Miss E, Tenjo and Mrs. F. Nakamura for technical and secretarial assistance. This work was supported in part by a grant from the Sagawa Traffic and
7". Yasuda et al./Clin. Chlm. Acta 218 (1993) 5-16
15
Social F o u n d a t i o n , and G r a n t s - i n - A i d for Scientific R e s e a r c h f r o m the Ministry o f E d u c a t i o n , Science a n d C u l t u r e o f J a p a n (04152052, 04836007 a n d 04770356).
6. References I 2
3 4
$ 6 7 8
9 10
Love JD, Hewitt RR. The relationship between human serum and human pancreatic DNase !. J Biol Chem 1979;254:!2588- ! 2594. Kishi K, Yasuda T, Ikehara Y, Sawazaki K, Sate W, iida R. Human serum deoxyribonuclease I (DNase !) polymorphism: pattern similarities among isozymes from serum, urine, kidney, liver, and pancreas. Am J Hum Genet 1990;47:i21-126. Nadano D, Yasuda T, Kishi K. Purification and characterization of genetically polymorphic deoxyribonuclease I from human kidney. J Biochem 1991;I10:321-323. Economidou-Karaoglou A, Opsomer M, Petit G, Lans M, Taper HS, Roberfroid M. Characteristic variations of serum alkaline DNase activity in relation to response to therapy and tumor prognosis in human lung cancer. Eur J Cancer Clin Oncol 1988;24:1337-1343. Laskowski M. Deoxyribonuclease !. In: Boyer PD, ed. The enzymes. Vol. 4, 3rd ed. New York: Academic Press, 1971;289-311. Rosenstreich DL, Tu JH, Kinkade PR et al. A human urine-derived interleukin I inhibitor: homology with deoxyribonuclease I. J Exp Mad 1988;168:1767-1779. Kishi K, Yasuda T, Awazu S, Mizuta K. Genetic polymorphism of human urine deoxyribonuclease I. Hum Genet 1989;81:295-297. Sawazaki K, Yausda T, Nadano D et al. A new individualization marker of semen: deoxyribonuclease ! (DNase l) polymorphism. Forensic Sci Int 1992;57:3q-44. Quinn PJ. r~oxyribonuclease activity in semen. J Reprod Fertil 1968;17:35-39. [to K, Minamiura N, Yamamoto T. Human urine DNase l: immunological identity with human pancreatic DNase I, and enzymic and proteocbemical properties of the enzyme. J Biochem
1984;95:1399-1406. II 12 13 14
15 16 17 18
19
20
Yasuda T, Awazu S, Sate W, [ida R, Tanaka Y, Kishi K. Human genetically polymorphic deoxyribonuclease: purification, characterization and multiplicity of urine deoxyribonuc]ease I. J Biochem 1990;108:393-398. Yasuda T, Nadano D, Awazu S, Kishi K. Human urine deoxyribonucleaseII (DNase II) isoenzymes: a novel immunoaffinity purification, biochemical multiplicity, genetic heterogeneity and broad distribution among tissues and body fluids. Biochim Biophys Acta 1992;1119:i85-193. Laemmli UK. Cleavage of structural proteins during the assembly of the head of bacteriophage 1"4. Nature 1970;227:680-685, Yasuda T, Nadano D, Tanaka Y, Kishi K. Specific identification of human ribonucleases by antibodies produced against two synthetic peptides corresponding to the N- and C.terminal amino.acid sequences of human urinary secretory-type ribonuclease, Biochim Biophys Acta 1992; 1121:331-334, Mizuta K, Awazu S, Yasuda T, Kishi K. Purification and characterization of three ribonucleases from human kidney: comparison with urine ribonucleases. Arch Biochem Biophys 1990;281:144-151. Ire K, Yamamoto T, Minamiura N. Phosphodiesterase I in human urine: purification and characterization of the enzyme. J Biochem 1987',102:359-367. Price PA, Moore S, Stein WH. Alkylation of a histidine residue at the active site of bovine pancreatic deoxyribonuclease. J Biol Chem 1969;244:924-928. Yasuda T, Mizuta K, lkehara Y, Kishi K. Genetic analysis of human deoxyribonuclease I by immunoblotting and the zymogram method following isoelectric focusing. Anal Biochem 1989;183:84-88. Yasuda T, Nadano D, Tanaka Y, Sate W, Nakanaga M, Kishi K. Practical utility of an antibody specific to a synthetic peptide corresponding to the N-terminal amino acid sequence of human urine DNase ! for genetic analysis of human urine DNase ! isozymes. Biochem Int 1990;22:699-705. Yasuda 1", Nadano D, Tenjo E, Takeshita H, Kishi K. The zymogram method for detection of ribonucleases after isoelectric focusing: analysis of multiple forms of human, bovine, and microbial enzymes. Anal Biochem 1992;206:172-177.
16
T. Yasuda et al. / Clin. Chim. Acta ~!8 (1993) 5-16
21 Yasuda T, Nadano D, Sawazaki K, Kishi K. Genetic polymorphism of human deoxylibonuclease ll (DNase ll): low activity levels in urine and leukocytes are due to an autosomal recessive allele. Ann Hum Genet 1992;56:1-10. 22 Polakoski KL, $yner FN, Zaneveld LID. Biochemistry of human seminal plasma. In: Hafez ESE, ed. Human semen and fertility regulation in men. St. Louis, MO: C.V. Mosby, 1976;133-152. 2] Liao T-H. Multiple forms of deoxyribonuclease !. Moi Cell Biochem 1981;34:15-22. 24 Beintema JJ, Blank A, Schieven GL, Dekker CA, Sorrentino S, Libonati M. Differences in glycosylation pattern of human secretory ribonucleases. Biochem J 1988,255:501-505. 25 Yasuda T, Mizuta K, Sato W, Kishi K. Purification and characterization of a ribonuclease from human spleen: immunological and enzymological comparison with nonsecretory ribonuclease from human urine. Eur J Biochem 1990;191:523-529. 26 Yasuda T, Mizuta K, Kishi K. Purification and characterization of two ribonucleases from human erythrocytes: immunological and enzymological comparison with ribonucleases from human urine. Arch Biochem Biophys 1990;279:130-137. 27 Yamashita K, Hitoi A, Irie M, Kobata A. Fractionation by lectin-affinity chromatography indicates that the giycosylation of most ribonucleases in human viscera and body fluids is organ specific. Arch Biochem Biophys 1986;250:263-266.
Clinica Chimica Acta 218 0993) 5-16
Human seminal deoxyribonuclease I (DNase I): purification, enzymological and immunological characterization and origin Toshihiro Yasudaa, Kazumi Sawazaki a, Daita Nadano a, Haruo Takeshita a, Masao Nakanaga b, Koichiro Kishi *a aDepartment of Legal Medicine. bDepartmentof Internal Medicine and Medical Genetics, Fukui Medical School. Matsuoka-cho, Fukui 910.1!, Japan
(Received 8 October 1992; revi~iionreceived 8 March 1993; accepted 12 March 1993)
Abstract Deoxyribonuclease 1 (DNase 1) was purified fre~ the semen of a 3S-year-old male and then characterized. The catalytic properties of the purified enzyme closely resembled those of DNase I purified from the urine of ~his individual and the following other similarities were observed: molecular masses, iodoacetic acid inactivation kinetics, desialylated isoenzyme patterns. However, the behavior of the purified enzymes determined on several different lectinaffinity chromatography columns differed, which suggests that organ.specific glycosylation of DNase I occurs. Multiple forms of the purified seminal DNase ! were demonstrated, each of which had a different pl value seperated by isoelectric focusing, which is compatible with the reported existence of genetic polymorphism of seminal DNase ! (Sawazaki et al., Forensic Sci lnt 1992;57:39-44). Furthermore, enzymological and immunological comparisons of purified seminal and urinary and partially purified prostatic DNases 1 indicated that the prostate may be one of seminal enzyme source tissues. Key words: Deoxyribonucleas¢; Human semen; Purification; Multii)licity; Prostate
1. Introduction Deoxyribonuclease I (DNase I, EC 3.1.21.1) is located preferentially in the pancreas, liver and kidney [1-3]. The value of serum DNase I activity has been reported to be useful in predicating the therapeutic efficacy in patients with cancer [4]. It has * Corresponding author. 0009-8981/93/~ ~.00 © 1993 ElsevierSciencePublishers B.V. All rights reserved. SSD! 0009.898i(93)05556-T
6
7". Yasuda et al./Clin. Ch#n. Acta 218 (1993) 5-16
been suggested that the enzymatic activities of DNase I in vivo are important for DNA metabolism in view of the hydrolytic activity and presence of DNase I in both duodenal and pancreatic juices. However, the distribution of DNase I activity in other tissues raises doubts that its major physiological role i~ a digestive function [5]. In particular, it is worth emphasizing that a human urine-derived interleukin-I i~lhibitor is functionally diverse from, but highly homologous with, DNase I [6]. Therefore, the extra- and intracellular functions of DNase I, as well as the digestive role, are of interest and should be elucidated. The isoenzyme patterns of human serum and urinary DNase I from different individuals were separated clearly into ten groups, which shows the existence of genetic polymorphism [2,7]. In our preliminary survey, the seminal DNase I type of a man was found to correlate with his serum and urinary enzyme types [8]. Deoxyribonuclease activities have been detected in human seminal plasma [9], but, to our knowledge, there is little biochemical information about seminal DNase I. In this report, we describe the purification and characterization of the seminal and prostatic DNase I enzymes, their comparison with the urinary enzyme and the origin of seminal DNase I. 2. Materials ~ad methods
Z !. Chemicals and biological materials All the chemicals used were of reagent grade or the purest grade available commercially. DEAE-Sepharose CL.6B, phenyl-Sepharose CL-4B and Sephadex G-75 were obtained from Pharmacia LKB (Uppsala, Sweden); hydroxyapatite and polyethylene glycol 6,000 (PEG) were from Nacalai Tesque (Kyoto, Japan); CIostridium perfringens sialidase, iodoacetic acid, phenylmethylsulfonylfluoride (PMSF) and soybean trypsin inhibitor (SBTI) were from Sigma (St. Louis, Me); agaroses conjugated with concanavalin A (Con A), lentil lectin (LCA), caster bean iectin (RCAI20) and wheatgerm lectin (WGA) were from Seikagaku-Kogyo (Tokyo, Japan). The human semen and urine used for purification of the enzyme were collected from a healthy 38-year-old male (DNas© 1 phenotype I). The spermatozoa were removed by centrifugation at 3,000 × g for 5 rain and the resulting seminal plasma was stored at -80°C until required for use. Donors supplied samples obtained by masturbation or by withdrawal during sexual intercourse. The ejaculate was collected in clean, wide-mouthed plastic containers with screw caps. These samples were presented at the laboratory within 12 h of collection and immediately frozen at -80°C until analyzed. Human urinary DNase I was purified as described previously [10,11]. Rabbit antibodies against human urinary DNase ! (anti-DNase I) and spleen DNase I! (anti-DNase II) were prepared as described previously [11,12]. A piece of prostate was obtained at autopsy from a 20-year-old drowned male within 20 h postmortem and stored at -80°C until required for use. 2.2. DNasc I activity assay and other analytical methods Deoxyribonuclease I activity was determined using the single radial enzyme diffusion (SRED) method as described previously [3,12], One unit of the DNase I activity was defined as described previously [I 1].
T. Yasuda et al. / Clin. Chim, Acta 218 (1993) 5-16
7
The DNase 1-active fractions were subjected to polyacrylamidegel electrophoresis in the presence of 0.1% (w/v) sodium dodecyl sulfate (SDS-PAGE) using a 12% polyacrylamide gel, as described by Laemmli [13], and the enzyme was detected by immunoblotting with anti-DNase I [141. The DNase I was also subjected to 7.5% PAGE followed by activity staining, as described previously [3]. The final preparation was also assayed for other enzymes as follows: DNase II, two types of human ribonuclease (RNase) and phosphodiesterase I were measured as described by Yasuda et al. [12], Mizuta et al. [!5] and Ito et al. 116], respectively. 2.3. Purification of DNase I from human seminal plasma and human prostate All the purification steps were carried out at 4°C. The seminal plasma (12 ml) was mixed with 20 ml 25 mmol/l Tris-HCl buffer, pH 7.5 (buffer l), which contained 10 manol/I CaCI2 and i mmol/l PMSF, and was then dialyzed against this buffer. The dialyzed materials were applied to a DEAE-Sepharose CL-6B column (1.6 x 30 cm) pre-equilibrated with this buffer, and the adsorbed materials were eluted with a NaCI linear gradient (0-1.0 tool/i) in the same buffer at 30 ml/h (fraction size, 28 ml). The pooled active fractions were dialyzed against 10 mmol/I Tris-HCi buffer, pH 7.5, which contained 1 tool/! ammonium sulfate, followed by dialysis against 10 mmol/i potassium phosphate buffer, pH 6.7 (buffer ll), which contained 1 mol/l ammonium sulfate. The dialyzed materials were applied to a phenyl-Sepharose CL,4B column (1.6 x 15 cm) pre-equilibrated with this buffer and the adsorbed materials were eluted first with an ammonium sulfate linear reverse gradient (I.0-0 tool/i) in buffer I1 and then with buffer II alone at 20 ml/h (fraction size, 18 ml). The activefractions eluted with buffer II alone were collected and dialyzed against 5 mmol/l potassium phosphate buffer, pH 6.7, and applied to a hydroxyapatite column (1.0 x 10 cm) pre-equilibrated with this buffer, after which the column was washed with 100 ml buffer II at 15 ml/h (fraction size, 7.5 ml). Most of the enzyme activities passed through the column and the eluate was concentrated with PEG; the concentrated materials were dialyzed against buffer I, which contained 0.25 tool/! NaCI, and then subjected to gel filtration on a Sephadex G-75 column (!.6 x 90 cm? preequilibrated with this buffer, which also contained 1 mmoi/I CaCI2. Five-milliliter aliquots of each were collected at 7 ml/h. This gel filtration procedure wa~ repeated and the DNase I-active fractions obtained were collected and dialyzed against 20 mmol/l sodium phosphate, buffer, pH 7.2 (buffer Ill), which contained 0.15 mol/I NaCI. The dialyzed materials were applied to a Con A-agarose column (I x 5 cm) pre-equilibrated with this buffer and the adsorbed enzyme was eluted with 0.2 mol/I methyl et.D-mannopyranoside in this buffer at 7.5 ml/h (fraction size, 5 ml). The DNase I-active fractions were pooled and dialyzed against buffer l, which contained 1 mmol/l CaCl2, and then concentrated by ultrafiltration for use in the subsequent experiments described below. A human prostate (about 2.5 g) was cut into small pieces and homogenized in 10 ml buffer I, which contained 10 mmol/i CaCI2, 2.5 mmol/l PMSF and 30 mgJl SBTI, centrifuged at 15,000 x g for 20 rain and the supernatants were dialyzed against this buffer without SBTI. From the dialyzed materials, the prostate DNase I was purified by chromatography on successive DEAE-Sepharose CL-6B, phenyl-Sepharose CL4B and hydroxyapatite columns under conditions similar to those described above. The final passed-through fractions from the hydroxyapatite column were dialyzed
8
T. Yasuda et aL/Ciin. Chim. Acta 218 (1993) 5-16
against buffer I, which contained 1 mmol/! CaCI2, concentrated by uitrafiltration and then characterized. 2.4. Inhibitory effects of anti-human urinary DNase I and inactivation by iodoacetic acid The purified enzyme (about 5 x 10-3 units) in 50/~l 25 mmol/l potassium phosphate buffer, pH 6.7, and 50-;d aliquots of diluted seminal plasma with the same activity were mixed with different amounts of the IgG antibody fraction, and kept at 4°C overnight. The samples were centrifuged at 10,000 × g for 10 rain, and the DNase i activity remaining in each supernatant was detected by the zymogram method after PAGE, as described above. Inactivation of the enzymes by iodoacetic acid was performed essentially according to the method of Price et al. [17]. The enzyme activities remaining were determined by the SRED method. 2.5. Seminal DNase I detected by isoelectricfocusing and the zymogram method Isoelectric focusing in a thin layer of polyacrylamide gel (IEF-PAGE, pH 3.5-5) was performed as described previously [18,19], after which the DNase I was detected using the dried agarose film-overlay (DAFO) method [18,20]. The enzymes were treated with an equal volume of sialidase (10 units/ml in 50 mmol/I sodium acetate buffer, pH 5.0) as described previously [18]. 2.6. Affinity of seminal DNase l for lectin columns The purified enzyme (about 2 x 10-2 units) was dialyzed against buffer !11, which contained 0.15 mol/I NaCI. The dialyzed materials were applied to separate lectin-agarose columns (I ml each) with Con A, LCA, RCAI20 and WGA, which were all equilibrated with this buffer, washed with 5 ml this buffer (unbound fraction) and the adsorbed enzyme was eluted with 5 ml 0.2 mol/i methyl or.Dmannopyranoside, 0.1 mol/I D-lactose, 0.2 mol/l D-galactose and 0.2 mol/I N-acetyl. D-glucosamine, respectively in this buffer (bound fraction). The enzyme activities in these fractions were determined using the SRED method. 3. Iteselts
3.1. Purification of DNase ! from human seminal plasma The yield and purification were estimated to be about 33% and in excess of 8,485. fold, respectively (Table I). The final product yielded a single band when subjected to SDS-PAGE under reducing conditions when detected by immunostaining with anti-DNase I, but one major and several minor bands were detected by silverstaining. No DNase 11, secretory- or non-secretory-type RNase or phosphodiesterase 1 activities were detected in the final preparation. The final preparation migrated as a single band both in substrate-cast gel under non-reducing conditions and when subjected to SDS-PAGE followed by immunostaining detection with anti-DNase I under reducing conditions. Moreover, it showed electrophorctic mobilities similar to those of the urinary enzyme purified from the same individual when analyzed with the PAGE and SDS-PAGE techniques (~:ig. i).
T. Yasuda el ai./ Clin. Chim. Acta 218 (1993) S-16 Table 1 Summary of the procedures for purification of DNase I from human seminal plasma Purification step
Seminal plasmac I. DEAE-SepharoseCL-6B 2. PhenyI-SepharoseCL-4B 3. Hydroxyapatite 4. First Sephadex G-75 5. Second Sephadex G-75 6. Concanavalin A-agarose
Proteina (rag)
103.880 8.710 1.231 0.090 0.034 0.010 <0.004 d
Activityb
Purification (fold)
Total activity (units)
Specific Yield activity (%) (units/mg)
1.750 1.730 1.506 1.018 0.923 0.819 0.577
0.017 0.199 1.233 11.311 27.147 81.900 > 144.250
100 98.9 86. I 58.2 52.7 46.9 32.9
1.0 I 1.7 72.5 665.4 1,596.9 4,817.6 >8,485.3
aProtein concentration was determined using the Bio-Rad protein assay kit with bovine serum albumin as a standard. bEnzyme activity was measured by the SRED method as described in the text. CStarting from 12 ml of semen collected from a 38-year.old male. dprotein concentration was below the sensitivity of the method. The seminal DNase I had a molecular mass of 38.0 kDa determined by SDSPAGE under reducing conditions and of 34.0 kDa by gel filtration on a Sephadex G-75 column in its native state. These masses were found to be virtually identical to those determined for urinary DNase l.
3,2. Catalytic properties in the presence of both 10 mmol/I MgCI2 and I mmol/I CaCI2, the pH activity profile of the seminal DNase i exhibited a bell-shape with an optimal pH of about 6,8, with no activity below pH 4.5 or above pH 9.5. The effect of ionic strength on the enzyme activity was also examined using SRED at pH 6.5 in the presence of both 10 mmol/I MgCI2 and I mmol/I CaCI2 by adding NaCI, The activity decreased markedly as the ionic strength increased and almost disappeared when it reached 0.5. The requirements for various divalent cations (5 mmol/I as the chloride form) were also examined at pH 6,5, The enzyme was active in the presence of Co 2+, Mn 2+ or Mg 2+, but not in the absence of divalent cations; 10 mmol/I EDTA, I mmol/I EGTA and 100 mg/I G-actin abolished the activity, even in the presence of 10 mmol/I MgCI 2. The enzyme hydrolyzed native DNA three to four times faster than thermally denatured DNA. These catalytic properties of the seminal DNase I were very similar to those of the urinary enzyme purified from the same individual. The similar pH dependencies of the activities of the purified seminal and urinary enzymes are shown in Fig. 2. 3.3. Inhibitory effects of anti-human urinary DNase I and inactivation by iodoacetic acid The DNase I activity (about 5 × 10 -3 units) of the purified enzyme and seminal plasma were abolished by less than 50 ~g lgG derived from 5 ~l anti-urinary DNase
T. Yasuda et al./ Clin. Chim. ,4eta 218 (1993) 5-16
IO
C
B
A kDa 80.0 ""
49.5
IBm
32.5 27.5 ~"
1
1
2
1
2
Fig. !. Electrophoretic patterns of purified seminal (lane I) and urinary DNases I (lane 2). The SDSPAGE patterns detected by silver-staining (A) and immunoblotting (B). Silver-staining was accomplished using Silver Stain DC (Daiichi Pure Chemicals, Tokyo). After electrophoresis on 12% polyacrylamide gel, the protein was transferred onto a Durapore membrane (Millipore, Bedford, MA) by electroblotting and visualized by immunostaining with anti-DNase I, as described previously [14,18]. The molecular weights indicated in the figure were estimated using a prestained SDS-PAGE standard kit (Bio-Rad). (C) The PAGE patterns detected by the activity staining. After electrophoresis on 7.5% polyacrylamide gel, DNase ! activity was detected as previously described [3]. The arrow indicates the bromophenol blue migration position.
i I
!
I
I
!
I
I
IO0
so
o
60
pH
80
-,o:o
Fig, 2. Typical demonstration of the similar catalytic properties of the seminal (@), prostatic (A) and urinary (E~ DNases I and the effect of,~H on their activities. The pH dependency in the presence of both 10 mmol/I MECI2 and I mmol/I CaCI: was investigated using the SRED method. The reaction buffers (0.1 mol/I each) used were sodium acetate (pH 4.5-5,5), sodium cacodylate (pH 6,0-7.0), Tris-HCI (pH 7.5-9,0) and sodium glycine (pH 9.5). The activity is expressed as a percentage of that at optimal pH (6.5).
Ii
7". Yasuda et aL / Clin. Chim. ,4cta 218 (1993) ~-!6
~2
'
'
A
"5
0__.1 I
0
30
I
I
60 0 30 Reaction time (rain)
I
60
Fig. 3. The kinetics of the seminal (A) and urinary (B) DNase I activities by iodoacetic acid. The purified enzyme was incubated with 10 mmol/I (O) and 20 mmol/I (O) iodoacetie acid in the presence of 10 mmol/I CuCI2 at 37°C. The conditions and determination of the remaining activities are described in the text. The logarithm of the percentage of remaining activity was plotted against the reaction time and the leastsquare line was drawn.
I, but not by that from anti-DNase I1. These results indicate that the seminal DNase I has strong immunological reactivity with antibody to the urinary enzyme. The seminal DNase I was inactivated rapidly by 10 mmol/i iodoacetic acid in the presence of 10 mmol/I CuCI2 at 37°C (Fig. 3). When the logarithm of the percentage of remaining activity was plotted against the reaction time, a straight line was obtained, which indicates that the iodoacetic acid inactivation reaction obeyed firstorder kinetics and the first-order rate constant for the reaction was calculated to be 0.026 min "1. A similar plot was also obtained with the purified urinary DNase I, which indicates that the iodoacetic acid inactivation kinetics of the two enzymes were essentially identical. The inactivation reaction of bovine pancreatic DNase i reported by Price et al. [17] was similar to our results.
£4. Seminal DNase ! isoenzymes on IEF-PAGE gel detected by the zymogram method The purified seminal DNase I was separated into several bands with different pi values on the gel region corresponding to pH 3.5-4.0 (Fig. 4). Desialylation of the enzyme diminished the anodal bands and shifted them towards the cathode. The major bands of the purified seminal and urinary enzymes before and after sialidase digestion were similar. These findings indicate that variations in the sialic acid content of the seminal DNase I contribute in part to the charge multiplicity of the enzyme. 3.£ Lectin affinities of the seminal and urinary DNases I The affinities of the purified seminal and urinary enzymes for four different iectins were determined (Table 2). The strongest affinity of enzymes was observed with Con
12
7". Yasuda et al./ Clin. Chim. Acta 218 (1993) 5-16
B
A pl 3 . 7 5 ='~
e~
O O
4.15=,.-
Q 1
2
t, I
1
2
Fig. 4. The ]EF-PAGE patterns of purified seminal and urinary DNases ! before (A) and after (B) treatment with sialidase, The separated DNase 1 isoenzymes were detected by activity staining using the DAFO method described in previous papers [18,20]. The pl values indicated in the figure were estimated using low pl calibration kit (Pharmacia LKB), Lane I, seminal DNase l; lane 2, urinary DNase !, The anode is at the top,
A, to which 88% of the seminal and 71% of the urinary enzyme bound, whereas they exhibited low affinities (5-43% bound) for the other lectins. However, the affinities of the seminal DNase ! for several lectins were found to differ from those of the urinary enzyme, which indicates that some of the carbohydrate moieties of these enzymes differ.
Table 2 Binding of ~emen. and urine.derived DNase ! to lectin.atZ]nity columns Lectinoaprose
Con A-agarose LCA-agarose RCAi20-agarose WGA-agarose
Bound fraction' (%) Seminal DNase I
Urinary DNase I
87,8 30,7 (32.7) b 43,2 6.5 (8,3) b
70,5 5,2 (13.5) b 26,8 24,9 (23.5)b
The seminal enzyme fractions eluted From the second gel filtration on Sephadex G-75 were applied to four different lectin-agarose columns, The urinary enzyme was purified from the urine collected from the same individual and subjected to the same column chromatographic series as the seminal enzyme. Thirty-five microliters of each enzyme solution (about 4 x 10-2 units) was treated with 35/AI sialidase (10 units/ml) at 37°C overnight and then applied to each column, The chromatographic conditions and the activity assays for the unbound and bound fractions are described in Materials and Methods. 'Binding of the enzyme activity to lectin-agarose is expressed relative to the total activity. bValues in parentheses are those for the desialylated enzymes,
T. Yasudaet al./Clin. Ckim. Acta 218 (1993) 5-16
13
3.6. DNase I activity levels in semen samples and their correlation with phenotypes The DNase I types determined from semen samples have been correlated with the types determined from corresponding blood and urine samples [8]. In order to clarify whether the DNase I activity levels in semen correlated with the phenotypes, the activities in semen samples from 45 unrelated healthy Japanese men of three major type,,, (1, 1-2 and 2), which comprise about 98% of the Japanese population [2], were quantified using S R E D (see the details in the legend to Table 3). The activities of semen samples collected periodically from the same men at separate times were also quantified and the levels were found to be constant with a mean intra-individual variation with a standard deviation of about 18%. Furthermore, as the activity levels in semen samples preincubated at 37°C for a few hours remained virtually unchanged, proteases that may be present in seminal plasma may not affect DNase I inactivation under the conditions used in this study. The DNase I activity levels of the three groups (types 1, 1-2 and 2) were determined and there were no significant differences among them by testing with one-way analysis of variance (Table 3). 3. 7. Partial purification and characterization o f human prostatic DNase I Human seminal plasma contains the combined secretions of the accessory glands and of these the prostate has been demonstrated to contribute substantially to the seminal plasma composition [22]. The DNase I activity of the crude extract from human prostate was lower than that of the semen and the content of the enzyme was estimated to be 2.92 × 10 -4 units/g wet weight (2.51 × 10 -4 units/mg protein). The activity of the partially purified prostatic DNase I was equivalent to 25% of the total activity o f the crude extract and about 100-fold purification was achieved. The partially purified enzyme demonstrated optimal activity at about pH 6.5 in the presence of 10 mmol/I MgCI2 and l mmol/i CeCI2 and no activity below pH 4.5 or above pH 9.5 (Fig. 2). At pH 6.5, the enzyme was active in the presence of either
Table 3 The distribution of DNase I activity levelsa in semen samples from ,*5 individualswith three different phenotypes Phenotype
Total no. in group
Range (× 10-2 units/mg)
Mean value .*- S.D, (× 10-2 units/mg)
I I-2
15 15
2 Total
15 45
1.21-7.78 1.65-4.67 1.13-8.10
3.34 ± 2.09 3.19 4. 0.96 3.09 4. 1.90
1.13-8.10
3.21 4. 1,71b
After centrifugation, the resulting supernatants from semen samples were diluted 40.fold with 50 mmol/I Tris.maleate buffer (pH 6,5) containing I0 mmol/I MgCI 2 and ! mmol/I CaCI2. Enzyme activity of each
diluted sample was measured by the SRED method. All assays were performedat least in triplicate. The DNase I of each sample was phenotyped as described previously [8,18]. aDNase ! activities of each sample are expressed as units/rag protein. bMean value (units/rag protein) of all the samples corresponded to 354.0 -)- 171.0 units/I semen.
14
7". Yasuda et al. / Clin. Chim. Acta 218 (1993) 5-16
MgCI2, MnCI2 or COC12 (5 mmol/l each) and activities of the purified enzyme and crude extract were abolished by anti-DNase l, but not anti-DNase II. These characteristics were very similar to those of the urinary and seminal enzymes and these findings strongly suggest that the prostate is one of the source tissues of the seminal DNase I. 4. Discussion
Although Quinn first reported the existence of DNases ! and !I in human semen in 1968 and DNase was suggested to cause a rapid decrease in the motility of human spermatozoa |9], these enzymes have not been characterized in detail. Therefore, we have attempted to purify and characterize the DNase I from human seminal plasma. Multiplicity of bovine pancreatic DNase I has been studied extensively [23] and that demonstrated by human urinary DNase ! has been suggested to be attributed to variations in the primary structure and/or differences in the sialic acid content [I !]. In general, in order to elucidate the molecular multiplicity, heterogeneity and genetic control of an enzyme, purificatiori from the enzyme source collected from a single individual is essential. We have established that, in the case of urinary DNases I and II, enzyme preparations from a single individual were needed to investigate their biochemical and genetic multiplicities [11,12]. Therefore, our seminal DNase ! was obtained and purified from the semen collected from an individual. The purified enzyme and original semen from this individual yielded very similar patterns consisting of several bands with different pl values when analyzed using ! EF-PAGE and zymogram detection before and after sialidase digestion (Fig. 4). it is clear, therefore, that this multiplicity results from intra- and not inter-individual variation, and the assumption that the multiplicity of sem'nal DNase ! may be a result of partial proteolyti¢ degradation of the intact enzyme, as suggested by Polakoski et al. [22] is not credible. The urinary and seminal DNases I from this individual exhibited the same chromatographic behavior, mobilities with a si,lgle band when subjected to PAGE with activity stainin~ and SDS-PAGE with immunostaining, catalytic properties, inhibitory effects of G-actin and anti-DNase I, molecular masses and iodoacteic acid inactivation kinetics. These findings indicate that seminal and urinary DNases I may be the same gene products; however, the slight differences between them, which the lectin-affinity analyses revealed (Table III), may be caused by organ-specific post-translational modifications, which has been observed with human secretory-type RNases [16,24-26]. As both enzymes originated from the same individual, the difference could not be attributed to inter-individual variation. Yamashita et al. [27] indicated that the behavior of RNases subjected to various types of lectin-affinity chromatography provided valuable information about organspecific glycosylation. Our lectin-affinity analysis results of the seminal and urinary enzymes suggest that seminal DNase I may undergo glycosylation to a different extent and/or in a different manner from the urinary enzyme.
$. Acknowledgments We thank Miss E, Tenjo and Mrs. F. Nakamura for technical and secretarial assistance. This work was supported in part by a grant from the Sagawa Traffic and
7". Yasuda et al./Clin. Chlm. Acta 218 (1993) 5-16
15
Social F o u n d a t i o n , and G r a n t s - i n - A i d for Scientific R e s e a r c h f r o m the Ministry o f E d u c a t i o n , Science a n d C u l t u r e o f J a p a n (04152052, 04836007 a n d 04770356).
6. References I 2
3 4
$ 6 7 8
9 10
Love JD, Hewitt RR. The relationship between human serum and human pancreatic DNase !. J Biol Chem 1979;254:!2588- ! 2594. Kishi K, Yasuda T, Ikehara Y, Sawazaki K, Sate W, iida R. Human serum deoxyribonuclease I (DNase !) polymorphism: pattern similarities among isozymes from serum, urine, kidney, liver, and pancreas. Am J Hum Genet 1990;47:i21-126. Nadano D, Yasuda T, Kishi K. Purification and characterization of genetically polymorphic deoxyribonuclease I from human kidney. J Biochem 1991;I10:321-323. Economidou-Karaoglou A, Opsomer M, Petit G, Lans M, Taper HS, Roberfroid M. Characteristic variations of serum alkaline DNase activity in relation to response to therapy and tumor prognosis in human lung cancer. Eur J Cancer Clin Oncol 1988;24:1337-1343. Laskowski M. Deoxyribonuclease !. In: Boyer PD, ed. The enzymes. Vol. 4, 3rd ed. New York: Academic Press, 1971;289-311. Rosenstreich DL, Tu JH, Kinkade PR et al. A human urine-derived interleukin I inhibitor: homology with deoxyribonuclease I. J Exp Mad 1988;168:1767-1779. Kishi K, Yasuda T, Awazu S, Mizuta K. Genetic polymorphism of human urine deoxyribonuclease I. Hum Genet 1989;81:295-297. Sawazaki K, Yausda T, Nadano D et al. A new individualization marker of semen: deoxyribonuclease ! (DNase l) polymorphism. Forensic Sci Int 1992;57:3q-44. Quinn PJ. r~oxyribonuclease activity in semen. J Reprod Fertil 1968;17:35-39. [to K, Minamiura N, Yamamoto T. Human urine DNase l: immunological identity with human pancreatic DNase I, and enzymic and proteocbemical properties of the enzyme. J Biochem
1984;95:1399-1406. II 12 13 14
15 16 17 18
19
20
Yasuda T, Awazu S, Sate W, [ida R, Tanaka Y, Kishi K. Human genetically polymorphic deoxyribonuclease: purification, characterization and multiplicity of urine deoxyribonuc]ease I. J Biochem 1990;108:393-398. Yasuda T, Nadano D, Awazu S, Kishi K. Human urine deoxyribonucleaseII (DNase II) isoenzymes: a novel immunoaffinity purification, biochemical multiplicity, genetic heterogeneity and broad distribution among tissues and body fluids. Biochim Biophys Acta 1992;1119:i85-193. Laemmli UK. Cleavage of structural proteins during the assembly of the head of bacteriophage 1"4. Nature 1970;227:680-685, Yasuda T, Nadano D, Tanaka Y, Kishi K. Specific identification of human ribonucleases by antibodies produced against two synthetic peptides corresponding to the N- and C.terminal amino.acid sequences of human urinary secretory-type ribonuclease, Biochim Biophys Acta 1992; 1121:331-334, Mizuta K, Awazu S, Yasuda T, Kishi K. Purification and characterization of three ribonucleases from human kidney: comparison with urine ribonucleases. Arch Biochem Biophys 1990;281:144-151. Ire K, Yamamoto T, Minamiura N. Phosphodiesterase I in human urine: purification and characterization of the enzyme. J Biochem 1987',102:359-367. Price PA, Moore S, Stein WH. Alkylation of a histidine residue at the active site of bovine pancreatic deoxyribonuclease. J Biol Chem 1969;244:924-928. Yasuda T, Mizuta K, lkehara Y, Kishi K. Genetic analysis of human deoxyribonuclease I by immunoblotting and the zymogram method following isoelectric focusing. Anal Biochem 1989;183:84-88. Yasuda T, Nadano D, Tanaka Y, Sate W, Nakanaga M, Kishi K. Practical utility of an antibody specific to a synthetic peptide corresponding to the N-terminal amino acid sequence of human urine DNase ! for genetic analysis of human urine DNase ! isozymes. Biochem Int 1990;22:699-705. Yasuda 1", Nadano D, Tenjo E, Takeshita H, Kishi K. The zymogram method for detection of ribonucleases after isoelectric focusing: analysis of multiple forms of human, bovine, and microbial enzymes. Anal Biochem 1992;206:172-177.
16
T. Yasuda et al. / Clin. Chim. Acta ~!8 (1993) 5-16
21 Yasuda T, Nadano D, Sawazaki K, Kishi K. Genetic polymorphism of human deoxylibonuclease ll (DNase ll): low activity levels in urine and leukocytes are due to an autosomal recessive allele. Ann Hum Genet 1992;56:1-10. 22 Polakoski KL, $yner FN, Zaneveld LID. Biochemistry of human seminal plasma. In: Hafez ESE, ed. Human semen and fertility regulation in men. St. Louis, MO: C.V. Mosby, 1976;133-152. 2] Liao T-H. Multiple forms of deoxyribonuclease !. Moi Cell Biochem 1981;34:15-22. 24 Beintema JJ, Blank A, Schieven GL, Dekker CA, Sorrentino S, Libonati M. Differences in glycosylation pattern of human secretory ribonucleases. Biochem J 1988,255:501-505. 25 Yasuda T, Mizuta K, Sato W, Kishi K. Purification and characterization of a ribonuclease from human spleen: immunological and enzymological comparison with nonsecretory ribonuclease from human urine. Eur J Biochem 1990;191:523-529. 26 Yasuda T, Mizuta K, Kishi K. Purification and characterization of two ribonucleases from human erythrocytes: immunological and enzymological comparison with ribonucleases from human urine. Arch Biochem Biophys 1990;279:130-137. 27 Yamashita K, Hitoi A, Irie M, Kobata A. Fractionation by lectin-affinity chromatography indicates that the giycosylation of most ribonucleases in human viscera and body fluids is organ specific. Arch Biochem Biophys 1986;250:263-266.