Clinica Chimica Acta, 78 (1977) 235-242 @ Elsevier/North-Holland Biomedical Press
RIBONUCLEASE ACTIVITY FLUID, AND URINE
*, V. WEINBERGER
Division of Nephrology, (Canada) (Received
IN HUMAN SERUM, CEREBROSPINAL
and B. TATTRIE
Ottawa Civic Hospital, 1053 Carling Avenue, Ottawa, Ontario
Poly(C)-avid ribonucleases of molecular weight 33 000 are present in the serum, cerebrospinal fluid and urine of humans. Purified human urinary ribonuclease was used to produce a monospecific antibody in rabbits. The antibody was capable of: (i) inhibiting the enzyme activities in the serum, CSF, and urine; (ii) reacting with antigens in the serum and CSF. The antigens in the serum, CSF and urine were found to be immunologically identical. Immunoelectrophoresis data suggested that the urinary and CSF RNAase are chemically identical. Succesful renal transplantation reduced elevated serum RNAase to normal levels. The data suggest that the most likely source of both urinary and CSF ribonuclease activity is the blood stream.
In 1975 we reported on the isolation, purification, and properties of a ribonuclease from normal human urine [l]. The enzyme was found to have a molecular weight of 33 000 and a marked avidity for the synthetic substrate poly(C). In that same year Schmukler et al.  described the isolation and purification from normal plasma of a poly(C)-avid ribonuclease whose molecular weight was 32 000 . Recently we have documented, in human cerebrospinal fluid, the presence of ribonuclease activities with molecular weights of 33 000, 21 000, and 15 000. The occurrence of ribonucleases with molecular weight of 33 000 in the plasma, urine, and cerebrospinal fluid led us to investigate the relationship between the activities in these biological fluids.
* To whom correspondence
should be addressed.
Ribonuclease activity was measured according to the method of Zimmerman and Sandeen [ 31 as modified by Fink et al. [ 41 and Rabin and Weinberger [ 11. Antibody against purified human urinary ribonuclease (HUR) was produced in rabbits by injecting 5 mg of the purified enzyme three times at intervals of 3 weeks. Three weeks after the last injection the animals were exsanguinated and the antiserum was fractionated with ammonium sulfate in order to obtain the gamma globulin [ 51. Normal rabbit serum was treated in the same manner for control experiments. The antisera, thus prepared, were used in the study of purified HUR, uremic serum, cerebrospinal fluid. Chromatography of uremic serum and pooled normal human cerebrospinal fluid was carried out on columns of Sephadex gel G-100 (2.5 X 90 cm) with appropriate markers. The molecular weights of the ribonuclease activities were calculated according to the method of Andrews . Immunological
The double-diffusion gel studies were carried out as previously described [ 71. The inhibition studies were carried out exactly according to Fuchs et al.  in the following manner. The antibody and the ribonuclease antigen were incubated in a volume of 0.6 ml (0.05 M phosphate, pH 6.5, and 0.9% NaCl) at 37°C for 30 min and then overnight at 4°C. The mixture was centrifuged at 17 000 X g for 20 min and the supernatant was stored in the cold and used for assay and further study. The ammonium sulfate treatment of the supernatant and the acidification of the supernatants were carried out in manner identical to Fuchs et al. [ 51. Control antiserum was used in all experiments. Fractionation of uremic serum and cerebrospinal fluid for gel diffusion Cerebrospinal fluid. 500 ml of CSF was obtained aseptically, post-opera-
tively from patients who had undergone pituitary surgery. The fluids were obtained during uneventful recovery periods. The pooled CSF was dialyzed against potassium phosphate buffer, pH 6.5, concentrated against polyethylene glycol and dialyzed exhaustively against 0.1 M acetic acid sodium acetate buffer, pH 4.7. The resultant solution was then precipitated with ammonium sulfate (50%) to remove albumin and then dialyzed against 0.05 M potassium phosphate buffer, pH 6.5, and concentrated to 3.4 ml using polyethylene glycol. The final solution had an RNAase activity of 2.7 X lo5 units/ml. This solution was used from the double-diffusion gel studies. Uremic serum. Blood was obtained from patients undergoing chronic hemodialysis. The serum was initially fractionated with ammonium sulfate (40%) in the cold to precipitate the globulins. The supernatant was dialyzed against 0.1 M acetic acid sodium acetate, pH 4.7, and fractionated with ammonium sulfate 50% to precipitate the albumin. The supernatant was then retained and dialyzed against 0.05 M potassium phosphate buffer, pH 6.5, and concentrated with polyethylene glycol. The resultant solution was used for the doublediffusion gel studies.
Results Ribonuclease inhibition The antibody against purified human urinary ribonuclease (Abnca) was capable of inhibiting the ribonuclease activity as shown in Fig. 1. The results of the experiments clearly indicate that both soluble and insoluble antigen-antibody complexes are formed and that a point of equivalence could be defined. Under the condition of the experiments, it appeared that 1~1 of purified antiRNAase globulin could inhibit 270 units of enzyme activity. If ribonuclease is added to an assay mixture containing both poly(C)substrate and antibody it was found that 2 1.11of purified anti-RNAase globulin would be necessary in order to inhibit 1 unit of enzyme activity. Hence the substrate has a greater avidity for the enzyme compared with the antibody and appears to be over 500 fold. This effect of substrate on the antigen-antibody reaction has been described for the other ribonucleases [El]. Studies of uremic serum Our studies have shown that the serum ribonuclease level is approximately 1041 + 247 U/ml in normal individuals without evidence of disease and with a glomerular filtration rate of 90 ml/min/1.73 m*. As the glomerular filtration falls the serum RNAase increases. Patients undergoing hemodialysis have levels between 10 000-35 000 U/ml . Fig. 2 shows the results of Sephadex gel filtration of human uremic serum. Two major peaks of ribonuclease activity were found; one peak has a molecular weight of 33 000 (a) and the other of 18 000 (b). A shoulder of activity could also be seen at 29 000. The AbHUR was found capable of inhibiting the activity in all these peaks. Thus the AbHUR was capable of blocking the activity of all
pg RIBONUCLEASE Fig. 1. Inhibition of purified human ribonuclease by rabbit antiserum. Details Of method in text. a, RNAase with control antiserum: 0, RNAase with antibody; 0, RNAase with antibody. suPematant, A, RNAase with antibody, (NH4)2S04 acidified; A, RNAase with antibody. (NH4)2S04 supernatant: supernatant with acidification.
200, E z 3 IOO-
VOLUME Fig. wt.
poly(C)-avid ribonuclease activity. Uremic serum which was fractionated with ammonium sulfate to remove albumin and globulin was found to form precipitation lines when reacted with Ah HUR in double-diffusion gel experiments. Fig. 3 shows that the sera of different patients react with the AbHUR and that the precipitation lines are identical. Fig. 4 shows that the precipitation lines formed using purified human urinary ribonuclease and uremic serum are also identical. Thus one may conclude that the antigens in urine and uremic sera are similar.
paration Fig. A,
Double-diffusion and of
urinary D contain
Studies of AbHUR with cerebrospinal fluid Studies in our laboratory have shown that there is present in the cerebrospinal fluid of humans a measurable level of ribonuclease activity. The normal level is 269 + 95 U/ml. Fig. 5 shows the results of Sephadex gel chromatography of concentrated pooled normal CSF. Major peaks of activity were found at mol. wt. 33 000 and 21000 and a minor peak at 15 000. The Abnua was capable of inhibiting all peaks of activity. The fractionated cerebrospinal fluid was capable of forming precipitin lines with Abnna as shown in Fig. 6. The precipitin lines formed using purified HUR and fractionated cerebrospinal fluid showed reactions of identity. Immunoelectrophoresis showed that the migration of the CSF antigen and the purified HUR were identical. The data suggest that the CSF antigen and the HUR are not only immunologically identical but are chemically identical. Unfortunately because of low levels of CSF antigen the precipitin lines were too faint for adequate photographic reproduction. Renal transplantation and serum RNAase Fig. 7 demonstrates that the pre-tr,ansplant serum RNAase level was near 25 000 U/ml and within 18 h the level fell to 2000 U/ml. Approximately 57 X lo6 U of enzyme activity was recovered from the urine. Later in the course of treatment when rejection supervened, the serum ribonuclease levels began to rise.
( ml 1
Fig. 5. Ribonuclease activity determined by Sephadex gel G-100 chromatography of human cerebrospinal fluid.
Fig. 6. Double-diffusion gel studies. Centrr well contains rabbit anti-human urinary ribonuclease. Wells A, C, and I? contain purified human urinary rihonuclease and wells B. and D contain human CSF. Preparation of the CSE‘ for the gel diffusion studies are detailed in the text.
TRANSPLANT 1 69
Fig. 7. Serum
levels Pm- and Post-renal
Discussion that ribonuclease activities with molecular The results clearly inidicate fluid. Our previous weight 33 000 exist in uremic serum and cerebrospinal of molecular weight studies indicate that the urine contains a ribonuclease 33 000 [ 11. The immunological data presented shows that the antibody specific for human urinary RNAase (Ab nUR) is capable of inhibiting all the poly(C)-avid RNAase activity in uremic serum and CSF. Furthermore the gel diffusion studies show that the antigens in the urine, CSF, and serum are similar if not identical. Because the serum/plasma level of RNAase activity has been shown to increase with a fall in GFR and to decrease with a rise in GFR and because in successful renal transplantation there is a simultaneous rise in urinary RNAase and fall in plasma RNAase we have concluded that the urinary RNAase is derived from the plasma. The fall in plasma RNAase during successful renal transplantation may be in
part due to the renal degradation of ribonuclease protein and conversely the rise in plasma RNAase during renal failure may be partly due to failure of renal degradation. Maack [lo] has described the renal degradation of low molecular weight proteins such as lysozyme and insulin, and likely RNAases are similarly degraded with consequent effects on plasma enzyme levels [lo]. In some cases of end-stage renal failure we have found that urinary levels increase in ab:;-lute amount indicating that the source of the urinary enzyme must be the blood stream, without the resulting effects of the renal tubular degradative machinery. We have shown that the normal plasma level of RNAase is 1041 ?r 247 U/ml and that the normal CSF level is 269 + 95 U/ml. Under no circumstances have we found a level higher in the CSF than in the plasma. Because of the existing gradient and because of the increase in levels of CSF RNAase in certain disease states we have concluded that the CSF RNAase activity is also derived from the plasma. Furthermore the immunologic data clearly indicate that the CSF RNAase is inhibited by AbHUR and that it is “identical” to human urinary ribonuclease. Indeed it would be most unusual to expect that the central nervous system produces such large amounts of ribonuclease which then diffuses into the blood stream and then finally passes into the urine. Therefore we conclude that the ribonuclease which is found both in the urine and CSF is derived from the blood stream. The tissue source as yet is unknown. The ribonucleases therefore must be added to the list of low molecular weight proteins which are know to exist in plasma, CSF, and urine and which have been detailed by Walravens et al. [ll]. The proteins are as follows: (1) a*bT, (2) (Y+T, (3) P-trace protein, (4) /3-microglobulin, (5) post-‘?’ fraction, (6) lysozyme. In each instance the protein in the urine and CSF is derived from the plasma. Attempts have been made to measure lysozyme in biological fluids for diagnostic purposes but the results have unfortunately been non-specific [ 12-141. Recently investigators have been measuring beta microglobulin by radioimmunoassay in blood and urine in an attempt to distinguish between glomerular and tubular disorders of the kidney . We have also studied RNAase excretion and the results do provide useful information regarding the presence or absence of tubular disease. Furthermore we have also found that CSF RNAase levels are significantly elevated in certain neurological disease states such as chronic cerebrovascular disease, tumors and cord compression. It should be emphasized that the measurements of RNAase activity in these biological fluids were carried out using poly(C). To date the following are the enzymes which have been described and which have a marked affinity for poly(C): (1) the urinary ribonuclease described by Rabin and Weinberger (mol. wt. 33 000); (2) the plasma ribonuclease described by Schmukler et al. (mol. wt. 32 000); (3) the enzymes isolated from pancreas, duodenum, serum, and urine described by Bardon et al. (mol. wt. 14 000) ; (4) the epidermal ribonuclease partially purified by Melbye and Freedman . One more enzyme which also degrades poly(C) but which has a greater avidity for poly(U) is the ribonuclease purified by Kuciel and Ostrowski (mol. wt. 28 500) [ 181. It is possible that all of these enzymes circulate in the blood and are all excreted in the urine, and that the measurement of RNAase activity using poly(C) is rather non-specific. However, it appears according to our data that
the antibody produced with purified human urinary ribonuclease as the antigen was capable of inhibiting all the RNAase activity is CSF and serum and urine. It is conceivable that the monomeric form of the enzyme is 14 000 but that exists in biological fluids both in the monomeric and dimeric forms; furthermore it is possible that the enzyme has a molecular weight of 33 000 but that the lower molecular species represent active fragments produced during the isolation procedure. The resolution of these questions await further study. In any event the measurement allows the quantification of a particular group of ribonucleases which are immunologically related and which have a markedly avidity for poly(C). At present it is differentiated from the ribonucleases which are derived from the liver and the spleen, both of which have insignificant activity against poly(C) [ 19,201. Furthermore because the measurements are relatively simple to carry out and the results are clinically useful, we feel that RNAase particularly with renal disease are activity measurements in patients, recommended. Acknowledgments The authors would like to thank Miss Norma Roeheleau and Miss Ilda Carvalho for their secretarial help. This work was supported by the Medical Research Council (MRC Grant 5592) and the Ottawa Civic Hospital Research Fund. References 1
J. Med. C.B.
(Ackroyd, B. and
D. (1976) Biochim.
I. (1975) (1976)
R. and J.J.