- Email: [email protected]
Human seminal ribonuclease. Immunological quantitation of cross-reactive enzymes in serum, urine and seminal plasma
Biochimica et Biophysica Acta,
998 (1989) 97-101
97
Elsevier BBAPRO 33443
Human seminal ribonuclease. Immunological quantitation of cross-reactive enzymes in serum, urine and seminal plasma S a l v a t o r e S o r r e n t i n o 1, R o c c o D e P r i s c o i a n d M a s s i m o L i b o n a t i 2 i Dipartimento di Chimica Organica e Biologica, Facoltd di Scienze, Universitd di Napoli, Napoli, end 2 Istituto di Chimica Biologica, Facoltit di Medicina e Chirurgia, Universitd di Verona, Verona (Italy)
(Received 12 June 1989)
Key words: Ribonuclease;RNAase; RIA; (Human seminalplasma)
The distribution of secretory-type ribonudease in human serum, urine and seminal plasma has been studied by immunolo|,~cal measurements. Inhibition of enzyme activity by antibodies against pure human seminal RNAase shows that a cross-reactive enzyme is predominant (90%) in seminal plasma and is a significant component (70-80%) in urine and serum. A competitive binding radioimmunoassay has been developed by using specific antibodies and 12sl-labelled RNAase as radioligand. The procedure, very sensitive, reproducible and specific, has been used to determine seminal RNAase levels in seminal plasma samples from 48 healthy individuals (age range, 20-58 years). The mean concentration of the enzyme was found to be 6.6 ttg/ml (S.D. :l: 1.9).
Introduction Measurable levels of RNAase activity in human serum and urine were suggested to be indicative of a state oil disease, including cancer [1-4]. However, while some authors [1,5,6] specifically linked hi'gh serum levels of pancreatic RNAase with pancreatic cancer, others, using similar methods, did not find such a correlation [7,8]. On the other hand, serum RNAase activity was found to be significantly elevated in patients with prostatic cancer [9]. These contradictions could be explained by taking into account that human serum and other body fluids contain a variety of different soluble ribonacleases [10-12]. Both secretory (pancreatic-type, 'with high pH optimum and preference for poly(C) as substrate) and non-secretory (liver-type, with a low pH optimum and a preference for poly(U) as substrate) RNAases, as classified by Sierakowska and Shugar [13], have been found in serum, urine, and seminal plasma. Five ribonucleases have been partially purified by Akagi et al. [14] from normal serum. From human urine, Cranston et al. [15] and lwama et al. [16] purified and characterized two different ribonucleases. Sugiyama et
Abbreviations: Mops, 4-morpholinepropanesulphonicacid; GARGG, goat anti-rabbit ~,-globulin. Correspondence: S. Sorrentino, Dipartimento di Chimica Organica e Biologica, Via Mezzocannone 16; 1-80134Napofi, Italy.
al. [17] also showed the presence of four RNAases in human urine. Lee et al. [18] isolated and characterized four ribonucleases from human seminal plasma. At the same time, a secretory-type RNAase (a basic pratein produced in the prostate [19], very active on double-stranded RNA) has been purified and characterized by us from human seminal plasma [19,20]. Immunological [21,22] and structural [23] studies indicate that human secretory ribonucleases may be identical gene products, except for organ-specific post-translational modifications. Moreover, it is quite difficult to discriminate secretory from non-secretory ribonucleases by simple enzymatic measurements, whereas a radioimmunological procedure could reproduc~bly permit the quantitation of a single structural form of RNAase as a protein entity. By using radioimmunological assays, Weickman et al. [24] demonstrated, for instance, that high circulating levels of secretory (pancreatic-type) RNAase do not serve as specific markers for cancers of the pancreas or other organs, but are rather related to kidney malfunction. Recently, Sorrentino et al. [25] also showed by RIA procedures that no relationship can be found between serum levels of non-secretory (liver-type) RNAase and any particular disease. Here we describe (a) the cross-reactivity of antibodies against human seminal RNAase with secretory-type enzymes present in human body fluids, and (b) the RIA quantitation of the seminal enzyme in normal seminal plasma. A preliminary report on these studies has been presented [26].
0167-4838/89/$03.50 © 1989 ElsevierScience Publishers B.V.(Biomedical Division)
98 Materials and Methods
Materials Human seminal RNAase was purified from normal seminal plasma as described [19]. In all experiments the pooled fractions, corresponding to the glycosylated and non-glycosylated forms of the enzyme, were used. The purity of the enzyme preparations was checked electrophoretically and by amino acid analysis [19]. Bovine pancreatic RNAase A (type XlI-A), bovine serum albumin, and egg albumin were puchased from Sigma. Double-stranded poly(A) • poly(U) was a Miles product. High molecular weight ribosomal RNA was purified from wheat germ as described [27]. Bovine milk lactoperoxidase and goat anti-rabbit V-globulin (GARGG) were from Calbiochem. Radioiodine (carrier-free a25I) was purchased from Amersham.
Immunology Antibodies against purified human seminal ribonuclease were induced in female New Zeland rabbits by intradermal injection. 0.5 ml of RNAase solution (200 pg/ml) was mixed with an equal volume of complete Freund's adjuvant and injected at multiple sites along the back of each rabbit. After 21 days, three subsequent booster injections at the same dosage were given at 1-week intervals. One week after the last injection, blood was collected by cardiac puncture. Immunoglobulins were partially purified by precipitation from serum 45% saturated with ammonium sulphate. The precipitated material was dissolved in 10 mM Tris-HCl (pH 7.5)/0.15 M NaCI, dialysed against the same buffer and stored at - 20 ° C.
Radioiodination Purified seminal RNAase was radiolabelled with 125I by a modification of the lactoperoxidase method described by Gospodarowicz [28]. The reagents were added rapidly in the following order and amounts: RNAase, 12/~g in 50/~1; 0.5 M phosphate buffer (pH 7.5) in 50/~1; 125INa, 400 pCi in 20/~1; lactoperoxidase, 4/~g in 2/~1. The reaction was initiated by adding 5/~1 of a 1 : 15 000 dilution of 30% hydrogen peroxide. 5/zl of the same hydrogen peroxide dilution were added four times at 2 rain intervals. The labelled enzyme was purified by gel filtration on Sephadex G-25.
Radioimmunological assay The RIA procedure for human seminal RNAase was performed according to Weickmann and Glitz [12]. The reaction mixtures (400 /~1) contained: buffer (0.05 M Tris-HC1, 0.15 M NaCI, 1 mg/ml of egg albumin), 20 pg of non-immune rabbit serum, iodinated seminal RNAase (about 4.104 cpm), varying dilutions of standard RNAase solution or seminal plasma as competing antigen, and an appropriate dilution of antibodies
against seminal RNAase. To each sample, after 16 h incubation in capped Biovials at 4°C, 0.5 units of GARGG (in 200 ~1 of buffer) were added to fully precipitate the rabbit globulins. The total radioactivity of each sample was then measured. After 4 h at 4 ° C, the free and bound forms were separated by centrifugation at 9000 rpm for 15 min in a Sorvall GSA rotor. The supernatants were removed and the precipitates washed with 0.6 ml of cold buffer. Radioactivity in the precipitates was then counted. All determinations were carded out in duplicate.
Human body fluid samples Serum and urine samples obtained from 100 male healthy volunteers were pooled and centrifuged. The supernatants were dialysed at 4 o C against distilled water with cut-off Mr 8000 membi~.mes, concentrated by lyophilization and stored at - 2 0 ° C . Human seminal plasma samples, obtained from 48 healthy volunteers, were frozen shortly after collection and stored at - 20 o C until used.
Enzyme assays Wheat germ high molecular weight RNA and synthetic double-stranded poly(A)-poly(U) were used as substrates. Ribonuclease activity toward wheat germ RNA was measured by the formation of perchloric acid-soluble nucleotides according to Elson and Glitz [29], except that the reaction buffer contained 0.05 M Mops (pH 7.8), 0.1 M NaCI, and bovine serum albumin (1 mg/ml).. Inhibition of RNAase activity by antibodies against seminal ribonuclease was measured as described [12,21,25]. Constant quantities (containing about one enzyme unit) of human serum, urine or seminal plasma were mixed with varying amounts of antibody. Protein concentration in the assay mixtures was kept constant by adding non-immune rabbit y-globulins. After 10 min at 37 o C, enzyme assay was initiated with the addition of RNA. With poly(A), poly(U) as substrate, ribonuclease activity was assayed spectrophotometrically by measuring the increase in absorbance at 260 nm due to enzyme action, as already described [19].
Protein determination Protein concentration was measured according to Lowry et al. [30] using bovine RNAase A as standard. Results and Discussion
Inhibition of RNAase activity by antibodies Antibodies against pure human seminal RNAase were used to study the distribution of cross-reactive (secretory-type) enzymes in serum, urine and seminal plasma. Fig. 1 shows that most of the ribonucleases (90%) present in seminal plasma are structurally very similar or identical to the purified seminal enzyme, and also
99 !
i
|
|
!
|
|
100
.~ loo
->- 8 0
~ 40
&
n
~-~
80
"'~
60
,
n- 4 0
~E •
¢Y
0 I
I
1
10
I
I
10 2
10 3
I
104
I
105
I
~
o
10 6
1
nl antibody added
10 102 103 Ribonudeese ( n g / t u b e )
104
105
Fig. 1. Inhibition of ribonuclease activity by antibodies against seminal RNAase. Immunoglobulins were directly added to the standard RNAase assay mixture; non-specific 7-globulins were also included in the reaction mixtures te keep protein concentration constant. Assays were performed in duplicate, and data are expressed in semi-log plot. 0, purified seminal RNAase; o , seminal plasma; zx, urine; n, serum; A, bovine pancreatic RNAase A.
Fig. 2. Competitive binding radioimmunoassay for human seminal ribonudease. Inhibition of binding of 1251-1abdled human seminal RNAase by antibodies against human seminal RNAase. Each point is the average of three determinations obtained with three different RNAase preparations labelled at three different times over a 9 month period. Data are presented in semi-log plot. Purified uniabelled human seminal RNAase (@), and bovine RNAase A (&) are shown as competitors.
that a significant proportion (70-80%) of the RNAases contained in serum and urine are immunologically very similar to the secretory-type RNAase from seminal plasma. These results are in agreement with previous studies on immunological properties of secretory RNAases from human organs and body fluids [12,22,31]. Moreover, on the basis of sequence analyses, seminal RNAase was shown to be very similar to the secretory. (pancreatic-type) ribonudeases from urine, serum and pancreas [23]. The most striking differences among these enzyme proteins concern their glycosylation state and the presence, in urine and seminal RNAase, of a threonine residue at the C-terminus (position 128), which is absent in the pancreatic enzyme [32]. Fig. 1 also shows that structurally distinct (uninhibited) enzymes exist in various amounts in semen, urine and serum. The immunological procedure used by us for the assay of a particular RNAase in body fluids is more specific than a simple enzymatic determination. On the other hand, inhibition of RNAase activity by antibodies may also result from antibody binding (without precipitation) at posititions close to the catalytic site of the enzyme protein. This may be deduced by the inhibition pattern of bovine pancreatic RNAase A, as displayed in Fig. 1. Even such a distantly related protein, provided a large excess of antibody is used, was able to give a significant cross-reaction. These data are in agreement with the results obtained by Weickmann et al. [21] with antibodies raised against human pancreatic ribonudease.
inhibition pattern of the enzymatic activity assay (Fig. 1) is absent in the RIA procedure, and therefore appears to be devoid of significance. In contrast to the antibody inhibition of the enzymatic activity, the competitive binding radioimmunological assay depends on antibody recognition of several different antigenic determinants of the RNAase molecule, and is thus specific for a unique structural form of the enzyme. Moreover, the RIA procedure was reproducible over a 9 month period with three different RNAase preparations labelled at three different times.
Secretory RNAase levels in normal seminal plasma The amounts of the secretory-type seminal RNAase present in seminal plasma of 48 healthy individuals (age range, 20-58 years) was determined with the competitive binding RIA procedure described above. Fig. 3 shows the relationship between RNAase concentration and age. Notwithstanding the scatter in values, a slight upward trend of RNAase concentration with age can be noted. The arithmetic mean of 6.6/~g/n~ (S.D. + 1.9)
12 ~o
00
@~
- - -""
2
Radioimmunoassay for human seminal ribonuclease The RIA quantitation of pure human seminal RNAase is shown in F;g. 2. Inhibition of antibody binding of 125I-labeled RNAase by purified RNAase is quite sensitive. Accurate quantitations from 5 to 100 ng are possible. The cross-reaction with bovine RNAase A observed by using high amounts of antibody in the
I
l
I
20 Age ( y e a r s )
l
40
!
i
60
Fig. 3. RIA quantitation of secretory seminal RNAase in human seminal plasma from normal subjects. Each point is the average of four independent determinations performed with different dilutions of seminal plasma. ~ , best linear fit of the data determined according to the method of least squares; . . . . . . , S.D. from the mean.
100
r
]
800 J-
0 0
/
o
"-" 6ooi-
0 °
I-
I
...6~o o
I
0 -.-"" 0
OoOO-- 1
400
'~
.....- -" " - 1
o
-
I
I
°- I
20
I
40
I
i
I
_~
60
Age (yeors)
Fig. 4. Enzymatic measurements of secretory seminal RNAase in human seminal plasma from normal subjects. RNAase activity, expressed as enzymatic units, was measured with poly(A)-poly(U) as substrate. Reaction mixtures (1 ml of 0.15 M NaCI/0.015 M sodium citrate (pH 7)) contained 0.1 mM substrate and 1-2 /~1 of seminal plasma. Assays were performed in duplicate. - - , best linear fit of the data determined according to the method of least squares; . . . . . . , S.D. from the mean.
verified, this observation could have important implications in the early diagnosis of the desease. Consequently, by using antibodies against seminal RNAase, we studied the content of cross-reactive enzymes in serum and urine samples from male volunteers. We demonstrated (Fig. 1) that a large proportion of the ribonucleases present in male serum and urine is structurally very similar, if not identical, to the secretory seminal enzyme and thus indistinguishable from a possible secretory RNAase of prostatic origin. On this basis, we think that a RIA quantitation in human seminal plasma only could be the most logical means through which to investigate the relationship, if any, between high RNAase levels and prostatic cancer. Therefore, we developed and used the described radioimmunological assay to quantitate the seminal secretory-type RNAase in human seminal plasma from normal subjects. This procedure could be a useful method to study the potential use of seminal RNAase as a tumour marker.
Acknowledgements obtained from the 48 determinations, was in good agreement with the average value of 6.3/tg/ml obtained in four different measurements performed on a pool of the 48 seminal plasma samples. The same 48 samples of human seminal plasma used in the RIA were also used for enzymatic activity assays with poly(A), poly(U) as substrate (Fig. 4). The marked activity towards double-stranded RNA shown by seminal RNAase [19,20], which is a property also common to the human pancreatic enzyme [23], was used as a tool to discriminate between the secretory seminal RNAase and other ribonucleases [18] present in crude seminal plasma. The data shown in Fig. 4 are almost in line with the RIA measurements. Here also enzyme units show a tendency to increase with age, and the mean value of 440 units/ml (S.D. +_ 120), corresponding to about 7.3 ttg/ml of pure enzyme, is not far from the average value of 6.6 /tg/ml obtained with the RIA determinations. However, the spectrophotometric assay with doublestranded poly(A)- poly(U), although more specific than enzymic assays performed with single-stranded RNA, poly(C) or poly(U) as substrates, may not be always exactly reproducible, depending on many variables inchiding possible differences in molecular size of different batches of the substrate.
Conclusions Many different assay procedures and substrates have been used to define a relationship between RNAase levels and the presence and extent of various forms of cancer [1-9]. One such report from Chu and co-workers [9] linked high serum levels of prostatic RNAase, determined by enzymatic assay, with prostatic cancer. If
This work was supported in part by the Progetto Finalizzato 'Oncologia' (contratto, No. 87.01332.44) and in part by the Progetto Finalizzato 'Biotecnologie' of the C.N.R., Rome, Italy.
References 1 Reddi, K.K. and Holland, J.F. (1976) Proc. Natl. Acad. Sci. USA 73, 2308-2310. 2 Neuwelt, ILA., Schmukler, M, Niziak, M.S., Jewett, P.B. and Levy, C.C. (1977) Biochem. J. 163, 419-426. 3 Sheid, B., Lu, T., Pedrinan, L. and Nelson, J.H. (1977) Cancer (Phila) 39, 2204-2208. 4 Reddi, K.K. (1978) Clin. Biochem. 11, 133-134. 5 Rennet, I.G., Mock, A., Reitherman, R. and Douglas, A.P. (1978) Gastroenterology 74, 1142. 6 Warshaw, A.L., Lee, K.-H., Wood, W.C. and Cohen, A.M. (1980) Am. J. Surg. 139, 27-32. 7 Peterson, L.M. (1979) Froc. Natl. Acad. Sci. USA 76, 2630-2634. 8 Isaacs, P. (1981) Digestion 22, 101-107. 9 Chu, T.M, Wang, M.C., Kuciel, R., Valenzuela, L. and Murphy, G.P. (1977) Cancer Treat. Rep. 61, 193-200. 10 Blank, A. and Dekker, C.A. (1981) Biochemistry 20, 2261-2267. 11 Thomas, J.M. and Hodes, M.E. (1981) Clin. Chim. Acta 11, 199-209. 12 Weickmann, J.L. and (;ritz, D.G. (1982) J. Biol. Chem. 257, 8705-8710. 13 Sierakowska, H. and Shugar, D. (1977) Progr. Nucleic. Acids Res. Mol. Biol. 20, 59-130. 14 Akagi, K., Murai, K,, Hirao, N. and Yamanaka, M. (1976) Biochim. Biophys. Acta 442, 368-378. 15 Cranston, J.W, Perirli, F., Crisp, ER. and Hixon, C.V. (1980) Biochim. Biophys. Acta 616, 239-25~. 16 Iwama, M., Kunihiro, M., Ohgi, K. and Irie, M. (1981) 3. Biochem. 89, 1005-1016. 17 Sugiyama, R.H., Blank, A. and Dekker, C.A. (1981) Biochemistry 20, 2268-2274. 18 Lee, C.-L., Li, S.-L., Li, C.-Y. and Chu, T.M. (1983) Biochem. J. 215, 605-612.
101 19 De Prisco, R., Sorrentino, S., Leone, E., Libonati, M. (1984) Biochim. Biophys. Acta 788, 356-363. 20 Sorrentino, S., Lavitrano, M., De Prisco, R. and Libonati, M. (1985) Biochim. Biophys. Acta 827, 135-139. 21 Weickmann, J.L., Elson, M. and Glitz, D.G. (1981) Biochemistry 20, 1272-1278. 22 Morita, T., Niwata, Y., Ohgi, K., Ogawa, M. and Irie, M. (1986) J. Biochem. (Tokyo) 99, 17-25. 23 Beintema, J.J., Blank, A., Schieven, G.L., Dekker, C.A., Sorrentino, S. and Libonati, M. (1988) Biochem. J. 255, 501-505. 24 Weickmann, J.Lo, O|son, E.M. and Glitz, D.G. (1984) Cancer Res. 44, 1682-1687. 25 Sorrentino, S., Tucker, G.K. and Glitz, D.G. (1988) J. Biol. Chem. 263, 16125-16131.
26 Sorrentino, S., De Prisco, R. and Libonati, M. (1986) 3rd International Conference on Human Tumor Markers, Naples, Italy, April 23-26, Abstr. P.4.8, p. 244. 27 Glitz, D.G. and DeLker, C.A. (1963) Biochemistry 2, 1185-1192. 28 Gospodarowicz, D. (1973) J. Biol. Chem. 248, 5042-5049. 29 Elson, M. and Glitz, D.G. (1975) Biochemistry 14, 1471-1476. 30 Lowry, O.H., Rosebrough, N.J., Farr, A.L. and Randall, R.J. (1951) J. Biol. Chem. 193, 265-275. 31 Kurihara, M., Ogawa, M., Ohta, T., Kurokawa, E., Kitahara, T., Kosaki, G., Watanabe, T. and Wada, H. (1982) Cancer Res. 42, 4836-4841. 32 Beintema, J.J., Wietzes, P., Weickmann, J.L. and Glitz, D.G. (1984) Anal. Biochem. 136, 48-64.
998 (1989) 97-101
97
Elsevier BBAPRO 33443
Human seminal ribonuclease. Immunological quantitation of cross-reactive enzymes in serum, urine and seminal plasma S a l v a t o r e S o r r e n t i n o 1, R o c c o D e P r i s c o i a n d M a s s i m o L i b o n a t i 2 i Dipartimento di Chimica Organica e Biologica, Facoltd di Scienze, Universitd di Napoli, Napoli, end 2 Istituto di Chimica Biologica, Facoltit di Medicina e Chirurgia, Universitd di Verona, Verona (Italy)
(Received 12 June 1989)
Key words: Ribonuclease;RNAase; RIA; (Human seminalplasma)
The distribution of secretory-type ribonudease in human serum, urine and seminal plasma has been studied by immunolo|,~cal measurements. Inhibition of enzyme activity by antibodies against pure human seminal RNAase shows that a cross-reactive enzyme is predominant (90%) in seminal plasma and is a significant component (70-80%) in urine and serum. A competitive binding radioimmunoassay has been developed by using specific antibodies and 12sl-labelled RNAase as radioligand. The procedure, very sensitive, reproducible and specific, has been used to determine seminal RNAase levels in seminal plasma samples from 48 healthy individuals (age range, 20-58 years). The mean concentration of the enzyme was found to be 6.6 ttg/ml (S.D. :l: 1.9).
Introduction Measurable levels of RNAase activity in human serum and urine were suggested to be indicative of a state oil disease, including cancer [1-4]. However, while some authors [1,5,6] specifically linked hi'gh serum levels of pancreatic RNAase with pancreatic cancer, others, using similar methods, did not find such a correlation [7,8]. On the other hand, serum RNAase activity was found to be significantly elevated in patients with prostatic cancer [9]. These contradictions could be explained by taking into account that human serum and other body fluids contain a variety of different soluble ribonacleases [10-12]. Both secretory (pancreatic-type, 'with high pH optimum and preference for poly(C) as substrate) and non-secretory (liver-type, with a low pH optimum and a preference for poly(U) as substrate) RNAases, as classified by Sierakowska and Shugar [13], have been found in serum, urine, and seminal plasma. Five ribonucleases have been partially purified by Akagi et al. [14] from normal serum. From human urine, Cranston et al. [15] and lwama et al. [16] purified and characterized two different ribonucleases. Sugiyama et
Abbreviations: Mops, 4-morpholinepropanesulphonicacid; GARGG, goat anti-rabbit ~,-globulin. Correspondence: S. Sorrentino, Dipartimento di Chimica Organica e Biologica, Via Mezzocannone 16; 1-80134Napofi, Italy.
al. [17] also showed the presence of four RNAases in human urine. Lee et al. [18] isolated and characterized four ribonucleases from human seminal plasma. At the same time, a secretory-type RNAase (a basic pratein produced in the prostate [19], very active on double-stranded RNA) has been purified and characterized by us from human seminal plasma [19,20]. Immunological [21,22] and structural [23] studies indicate that human secretory ribonucleases may be identical gene products, except for organ-specific post-translational modifications. Moreover, it is quite difficult to discriminate secretory from non-secretory ribonucleases by simple enzymatic measurements, whereas a radioimmunological procedure could reproduc~bly permit the quantitation of a single structural form of RNAase as a protein entity. By using radioimmunological assays, Weickman et al. [24] demonstrated, for instance, that high circulating levels of secretory (pancreatic-type) RNAase do not serve as specific markers for cancers of the pancreas or other organs, but are rather related to kidney malfunction. Recently, Sorrentino et al. [25] also showed by RIA procedures that no relationship can be found between serum levels of non-secretory (liver-type) RNAase and any particular disease. Here we describe (a) the cross-reactivity of antibodies against human seminal RNAase with secretory-type enzymes present in human body fluids, and (b) the RIA quantitation of the seminal enzyme in normal seminal plasma. A preliminary report on these studies has been presented [26].
0167-4838/89/$03.50 © 1989 ElsevierScience Publishers B.V.(Biomedical Division)
98 Materials and Methods
Materials Human seminal RNAase was purified from normal seminal plasma as described [19]. In all experiments the pooled fractions, corresponding to the glycosylated and non-glycosylated forms of the enzyme, were used. The purity of the enzyme preparations was checked electrophoretically and by amino acid analysis [19]. Bovine pancreatic RNAase A (type XlI-A), bovine serum albumin, and egg albumin were puchased from Sigma. Double-stranded poly(A) • poly(U) was a Miles product. High molecular weight ribosomal RNA was purified from wheat germ as described [27]. Bovine milk lactoperoxidase and goat anti-rabbit V-globulin (GARGG) were from Calbiochem. Radioiodine (carrier-free a25I) was purchased from Amersham.
Immunology Antibodies against purified human seminal ribonuclease were induced in female New Zeland rabbits by intradermal injection. 0.5 ml of RNAase solution (200 pg/ml) was mixed with an equal volume of complete Freund's adjuvant and injected at multiple sites along the back of each rabbit. After 21 days, three subsequent booster injections at the same dosage were given at 1-week intervals. One week after the last injection, blood was collected by cardiac puncture. Immunoglobulins were partially purified by precipitation from serum 45% saturated with ammonium sulphate. The precipitated material was dissolved in 10 mM Tris-HCl (pH 7.5)/0.15 M NaCI, dialysed against the same buffer and stored at - 20 ° C.
Radioiodination Purified seminal RNAase was radiolabelled with 125I by a modification of the lactoperoxidase method described by Gospodarowicz [28]. The reagents were added rapidly in the following order and amounts: RNAase, 12/~g in 50/~1; 0.5 M phosphate buffer (pH 7.5) in 50/~1; 125INa, 400 pCi in 20/~1; lactoperoxidase, 4/~g in 2/~1. The reaction was initiated by adding 5/~1 of a 1 : 15 000 dilution of 30% hydrogen peroxide. 5/zl of the same hydrogen peroxide dilution were added four times at 2 rain intervals. The labelled enzyme was purified by gel filtration on Sephadex G-25.
Radioimmunological assay The RIA procedure for human seminal RNAase was performed according to Weickmann and Glitz [12]. The reaction mixtures (400 /~1) contained: buffer (0.05 M Tris-HC1, 0.15 M NaCI, 1 mg/ml of egg albumin), 20 pg of non-immune rabbit serum, iodinated seminal RNAase (about 4.104 cpm), varying dilutions of standard RNAase solution or seminal plasma as competing antigen, and an appropriate dilution of antibodies
against seminal RNAase. To each sample, after 16 h incubation in capped Biovials at 4°C, 0.5 units of GARGG (in 200 ~1 of buffer) were added to fully precipitate the rabbit globulins. The total radioactivity of each sample was then measured. After 4 h at 4 ° C, the free and bound forms were separated by centrifugation at 9000 rpm for 15 min in a Sorvall GSA rotor. The supernatants were removed and the precipitates washed with 0.6 ml of cold buffer. Radioactivity in the precipitates was then counted. All determinations were carded out in duplicate.
Human body fluid samples Serum and urine samples obtained from 100 male healthy volunteers were pooled and centrifuged. The supernatants were dialysed at 4 o C against distilled water with cut-off Mr 8000 membi~.mes, concentrated by lyophilization and stored at - 2 0 ° C . Human seminal plasma samples, obtained from 48 healthy volunteers, were frozen shortly after collection and stored at - 20 o C until used.
Enzyme assays Wheat germ high molecular weight RNA and synthetic double-stranded poly(A)-poly(U) were used as substrates. Ribonuclease activity toward wheat germ RNA was measured by the formation of perchloric acid-soluble nucleotides according to Elson and Glitz [29], except that the reaction buffer contained 0.05 M Mops (pH 7.8), 0.1 M NaCI, and bovine serum albumin (1 mg/ml).. Inhibition of RNAase activity by antibodies against seminal ribonuclease was measured as described [12,21,25]. Constant quantities (containing about one enzyme unit) of human serum, urine or seminal plasma were mixed with varying amounts of antibody. Protein concentration in the assay mixtures was kept constant by adding non-immune rabbit y-globulins. After 10 min at 37 o C, enzyme assay was initiated with the addition of RNA. With poly(A), poly(U) as substrate, ribonuclease activity was assayed spectrophotometrically by measuring the increase in absorbance at 260 nm due to enzyme action, as already described [19].
Protein determination Protein concentration was measured according to Lowry et al. [30] using bovine RNAase A as standard. Results and Discussion
Inhibition of RNAase activity by antibodies Antibodies against pure human seminal RNAase were used to study the distribution of cross-reactive (secretory-type) enzymes in serum, urine and seminal plasma. Fig. 1 shows that most of the ribonucleases (90%) present in seminal plasma are structurally very similar or identical to the purified seminal enzyme, and also
99 !
i
|
|
!
|
|
100
.~ loo
->- 8 0
~ 40
&
n
~-~
80
"'~
60
,
n- 4 0
~E •
¢Y
0 I
I
1
10
I
I
10 2
10 3
I
104
I
105
I
~
o
10 6
1
nl antibody added
10 102 103 Ribonudeese ( n g / t u b e )
104
105
Fig. 1. Inhibition of ribonuclease activity by antibodies against seminal RNAase. Immunoglobulins were directly added to the standard RNAase assay mixture; non-specific 7-globulins were also included in the reaction mixtures te keep protein concentration constant. Assays were performed in duplicate, and data are expressed in semi-log plot. 0, purified seminal RNAase; o , seminal plasma; zx, urine; n, serum; A, bovine pancreatic RNAase A.
Fig. 2. Competitive binding radioimmunoassay for human seminal ribonudease. Inhibition of binding of 1251-1abdled human seminal RNAase by antibodies against human seminal RNAase. Each point is the average of three determinations obtained with three different RNAase preparations labelled at three different times over a 9 month period. Data are presented in semi-log plot. Purified uniabelled human seminal RNAase (@), and bovine RNAase A (&) are shown as competitors.
that a significant proportion (70-80%) of the RNAases contained in serum and urine are immunologically very similar to the secretory-type RNAase from seminal plasma. These results are in agreement with previous studies on immunological properties of secretory RNAases from human organs and body fluids [12,22,31]. Moreover, on the basis of sequence analyses, seminal RNAase was shown to be very similar to the secretory. (pancreatic-type) ribonudeases from urine, serum and pancreas [23]. The most striking differences among these enzyme proteins concern their glycosylation state and the presence, in urine and seminal RNAase, of a threonine residue at the C-terminus (position 128), which is absent in the pancreatic enzyme [32]. Fig. 1 also shows that structurally distinct (uninhibited) enzymes exist in various amounts in semen, urine and serum. The immunological procedure used by us for the assay of a particular RNAase in body fluids is more specific than a simple enzymatic determination. On the other hand, inhibition of RNAase activity by antibodies may also result from antibody binding (without precipitation) at posititions close to the catalytic site of the enzyme protein. This may be deduced by the inhibition pattern of bovine pancreatic RNAase A, as displayed in Fig. 1. Even such a distantly related protein, provided a large excess of antibody is used, was able to give a significant cross-reaction. These data are in agreement with the results obtained by Weickmann et al. [21] with antibodies raised against human pancreatic ribonudease.
inhibition pattern of the enzymatic activity assay (Fig. 1) is absent in the RIA procedure, and therefore appears to be devoid of significance. In contrast to the antibody inhibition of the enzymatic activity, the competitive binding radioimmunological assay depends on antibody recognition of several different antigenic determinants of the RNAase molecule, and is thus specific for a unique structural form of the enzyme. Moreover, the RIA procedure was reproducible over a 9 month period with three different RNAase preparations labelled at three different times.
Secretory RNAase levels in normal seminal plasma The amounts of the secretory-type seminal RNAase present in seminal plasma of 48 healthy individuals (age range, 20-58 years) was determined with the competitive binding RIA procedure described above. Fig. 3 shows the relationship between RNAase concentration and age. Notwithstanding the scatter in values, a slight upward trend of RNAase concentration with age can be noted. The arithmetic mean of 6.6/~g/n~ (S.D. + 1.9)
12 ~o
00
@~
- - -""
2
Radioimmunoassay for human seminal ribonuclease The RIA quantitation of pure human seminal RNAase is shown in F;g. 2. Inhibition of antibody binding of 125I-labeled RNAase by purified RNAase is quite sensitive. Accurate quantitations from 5 to 100 ng are possible. The cross-reaction with bovine RNAase A observed by using high amounts of antibody in the
I
l
I
20 Age ( y e a r s )
l
40
!
i
60
Fig. 3. RIA quantitation of secretory seminal RNAase in human seminal plasma from normal subjects. Each point is the average of four independent determinations performed with different dilutions of seminal plasma. ~ , best linear fit of the data determined according to the method of least squares; . . . . . . , S.D. from the mean.
100
r
]
800 J-
0 0
/
o
"-" 6ooi-
0 °
I-
I
...6~o o
I
0 -.-"" 0
OoOO-- 1
400
'~
.....- -" " - 1
o
-
I
I
°- I
20
I
40
I
i
I
_~
60
Age (yeors)
Fig. 4. Enzymatic measurements of secretory seminal RNAase in human seminal plasma from normal subjects. RNAase activity, expressed as enzymatic units, was measured with poly(A)-poly(U) as substrate. Reaction mixtures (1 ml of 0.15 M NaCI/0.015 M sodium citrate (pH 7)) contained 0.1 mM substrate and 1-2 /~1 of seminal plasma. Assays were performed in duplicate. - - , best linear fit of the data determined according to the method of least squares; . . . . . . , S.D. from the mean.
verified, this observation could have important implications in the early diagnosis of the desease. Consequently, by using antibodies against seminal RNAase, we studied the content of cross-reactive enzymes in serum and urine samples from male volunteers. We demonstrated (Fig. 1) that a large proportion of the ribonucleases present in male serum and urine is structurally very similar, if not identical, to the secretory seminal enzyme and thus indistinguishable from a possible secretory RNAase of prostatic origin. On this basis, we think that a RIA quantitation in human seminal plasma only could be the most logical means through which to investigate the relationship, if any, between high RNAase levels and prostatic cancer. Therefore, we developed and used the described radioimmunological assay to quantitate the seminal secretory-type RNAase in human seminal plasma from normal subjects. This procedure could be a useful method to study the potential use of seminal RNAase as a tumour marker.
Acknowledgements obtained from the 48 determinations, was in good agreement with the average value of 6.3/tg/ml obtained in four different measurements performed on a pool of the 48 seminal plasma samples. The same 48 samples of human seminal plasma used in the RIA were also used for enzymatic activity assays with poly(A), poly(U) as substrate (Fig. 4). The marked activity towards double-stranded RNA shown by seminal RNAase [19,20], which is a property also common to the human pancreatic enzyme [23], was used as a tool to discriminate between the secretory seminal RNAase and other ribonucleases [18] present in crude seminal plasma. The data shown in Fig. 4 are almost in line with the RIA measurements. Here also enzyme units show a tendency to increase with age, and the mean value of 440 units/ml (S.D. +_ 120), corresponding to about 7.3 ttg/ml of pure enzyme, is not far from the average value of 6.6 /tg/ml obtained with the RIA determinations. However, the spectrophotometric assay with doublestranded poly(A)- poly(U), although more specific than enzymic assays performed with single-stranded RNA, poly(C) or poly(U) as substrates, may not be always exactly reproducible, depending on many variables inchiding possible differences in molecular size of different batches of the substrate.
Conclusions Many different assay procedures and substrates have been used to define a relationship between RNAase levels and the presence and extent of various forms of cancer [1-9]. One such report from Chu and co-workers [9] linked high serum levels of prostatic RNAase, determined by enzymatic assay, with prostatic cancer. If
This work was supported in part by the Progetto Finalizzato 'Oncologia' (contratto, No. 87.01332.44) and in part by the Progetto Finalizzato 'Biotecnologie' of the C.N.R., Rome, Italy.
References 1 Reddi, K.K. and Holland, J.F. (1976) Proc. Natl. Acad. Sci. USA 73, 2308-2310. 2 Neuwelt, ILA., Schmukler, M, Niziak, M.S., Jewett, P.B. and Levy, C.C. (1977) Biochem. J. 163, 419-426. 3 Sheid, B., Lu, T., Pedrinan, L. and Nelson, J.H. (1977) Cancer (Phila) 39, 2204-2208. 4 Reddi, K.K. (1978) Clin. Biochem. 11, 133-134. 5 Rennet, I.G., Mock, A., Reitherman, R. and Douglas, A.P. (1978) Gastroenterology 74, 1142. 6 Warshaw, A.L., Lee, K.-H., Wood, W.C. and Cohen, A.M. (1980) Am. J. Surg. 139, 27-32. 7 Peterson, L.M. (1979) Froc. Natl. Acad. Sci. USA 76, 2630-2634. 8 Isaacs, P. (1981) Digestion 22, 101-107. 9 Chu, T.M, Wang, M.C., Kuciel, R., Valenzuela, L. and Murphy, G.P. (1977) Cancer Treat. Rep. 61, 193-200. 10 Blank, A. and Dekker, C.A. (1981) Biochemistry 20, 2261-2267. 11 Thomas, J.M. and Hodes, M.E. (1981) Clin. Chim. Acta 11, 199-209. 12 Weickmann, J.L. and (;ritz, D.G. (1982) J. Biol. Chem. 257, 8705-8710. 13 Sierakowska, H. and Shugar, D. (1977) Progr. Nucleic. Acids Res. Mol. Biol. 20, 59-130. 14 Akagi, K., Murai, K,, Hirao, N. and Yamanaka, M. (1976) Biochim. Biophys. Acta 442, 368-378. 15 Cranston, J.W, Perirli, F., Crisp, ER. and Hixon, C.V. (1980) Biochim. Biophys. Acta 616, 239-25~. 16 Iwama, M., Kunihiro, M., Ohgi, K. and Irie, M. (1981) 3. Biochem. 89, 1005-1016. 17 Sugiyama, R.H., Blank, A. and Dekker, C.A. (1981) Biochemistry 20, 2268-2274. 18 Lee, C.-L., Li, S.-L., Li, C.-Y. and Chu, T.M. (1983) Biochem. J. 215, 605-612.
101 19 De Prisco, R., Sorrentino, S., Leone, E., Libonati, M. (1984) Biochim. Biophys. Acta 788, 356-363. 20 Sorrentino, S., Lavitrano, M., De Prisco, R. and Libonati, M. (1985) Biochim. Biophys. Acta 827, 135-139. 21 Weickmann, J.L., Elson, M. and Glitz, D.G. (1981) Biochemistry 20, 1272-1278. 22 Morita, T., Niwata, Y., Ohgi, K., Ogawa, M. and Irie, M. (1986) J. Biochem. (Tokyo) 99, 17-25. 23 Beintema, J.J., Blank, A., Schieven, G.L., Dekker, C.A., Sorrentino, S. and Libonati, M. (1988) Biochem. J. 255, 501-505. 24 Weickmann, J.Lo, O|son, E.M. and Glitz, D.G. (1984) Cancer Res. 44, 1682-1687. 25 Sorrentino, S., Tucker, G.K. and Glitz, D.G. (1988) J. Biol. Chem. 263, 16125-16131.
26 Sorrentino, S., De Prisco, R. and Libonati, M. (1986) 3rd International Conference on Human Tumor Markers, Naples, Italy, April 23-26, Abstr. P.4.8, p. 244. 27 Glitz, D.G. and DeLker, C.A. (1963) Biochemistry 2, 1185-1192. 28 Gospodarowicz, D. (1973) J. Biol. Chem. 248, 5042-5049. 29 Elson, M. and Glitz, D.G. (1975) Biochemistry 14, 1471-1476. 30 Lowry, O.H., Rosebrough, N.J., Farr, A.L. and Randall, R.J. (1951) J. Biol. Chem. 193, 265-275. 31 Kurihara, M., Ogawa, M., Ohta, T., Kurokawa, E., Kitahara, T., Kosaki, G., Watanabe, T. and Wada, H. (1982) Cancer Res. 42, 4836-4841. 32 Beintema, J.J., Wietzes, P., Weickmann, J.L. and Glitz, D.G. (1984) Anal. Biochem. 136, 48-64.