2246
Biochemistry 1991, 30, 2246-2255
A,, Torchinsky, Y.,& Ueno, H. (1985) in Transaminases (Christen, P., & Metzler, D. E., Eds.) pp 38-108, John Wiley and Sons, New York. Likos, J. J., Ueno, H., Feldhaus, R. W., & Metzler, D. E. ( 1 982) Biochemistry 21, 4377-4386. Miles, E. W. (1 978) in Enzyme Actioated Irreversible Inhibitors (Seiler, N., Jung, M. J., & Koch-Weser, J., Eds.) pp 73-85, ElsevierINorth-Holland, Amsterdam. Morino, Y., & Okamoto, M. (1973) Biochem. Biophys. Res. Commun. 50, 1061-1067. Mueckler, M. M., & Pitot, H. C. (1985) J . Biol. Chem. 260, 12993-12997.
Peraino, C., Bunville, L. G., & Tahmisian, T. N. (1969) J . Biol. Chem. 244, 2241-2249. Sanada, Y., Shiotani, T., Okuno, E., & Katunuma, N. (1976) Eur. J . Biochem. 69, 507-5 15. Silverman, R. B., & Invergo, B. J. (1986) Biochemistry 25, 68 17-6820. Simmaco, M., John, R. A,, Barra, D., & Bossa, F. (1986) FEES Lett. 199, 39-42. Ueno, H., Likos, J. J., & Metzler, D. E. (1982) Biochemistry 21, 4387-4393. Williams, J. A., Bridge, G., Fowler, L. J., & John, R. A. (1982) Biochem. J . 201, 221-225.
Interaction of Human Placental Ribonuclease with Placental Ribonuclease Inhibitor? Robert Shapiro and Bert L. Vallee* Center for Biochemical and Biophysical Sciences and Medicine, Harvard Medical School, Boston, Massachusetts 021 I5 Received August 6, 1990; Revised Manuscript Received October 9, I990
ABSTRACT: The interactions of human placental ribonuclease inhibitor (PRI) with bovine pancreatic ri-
bonuclease (RNase) A and human angiogenin, a plasma protein that induces blood vessel formation, have been characterized in detail in earlier studies. However, studies on the interaction of PRI with the RNase(s) indigenous to placenta have not been performed previously, nor have any placental RNases been identified. In the present work, the major human placental RNase (PR) was purified to homogeneity by a five-step procedure and was obtained in a yield of 110 pg/kg of tissue. The placental content of angiogenin was also examined and was found to be at least 10-fold lower than that of PR. On the basis of its amino acid composition, amino-terminal sequence, and catalytic properties, PR appears to be identical with an RNase previously isolated from eosinophils (eosinophil-derived neurotoxin), liver, and urine. The apparent second-order rate constant of association for the PR-PRI complex, measured by examining the competition between PR and angiogenin for PRI, is 1.9 X lo8 M-' s-'. The rate constant for dissociation of the complex, determined by HPLC measurement of the rate of release of PR from its complex with P R I in the presence of a scavenger for free PRI, is 1.8 X lo-' s-l. Thus the Ki value for the PRsPRI complex is 9 X M, similar to that obtained with angiogenin, and 40-fold lower than that measured with RNase A. Complex formation causes a small red shift in the protein fluorescence emission spectrum, with no significant change in overall intensity. The fluorescence quantum yield of PR and the Stern-Volmer constant for fluorescence quenching by acrylamide are both high, possibly due to the presence of an unusual posttranslationally modified tryptophan residue at position 7 in the primary sequence.
D e s p i t e several decades of study [see Levy and Karpetsky ( 198 l)], human ribonucleases (RNases) * have only recently begun to be characterized as molecular entities. Five structurally and functionally distinct species have now been identified, including pancreatic RNase (Weickmann et al., 1981; Beintema et al., 1984), the "nonsecretory" RNase from liver, urine, and eosinophils (Cranston et al., 1980; Gleich et al., 1986; Beintema et al., 1988a; Sorrentino et al., 1988), and three additional forms found in blood (see Discussion). These RNases are closely related: their amino acid sequences are homologous and all cleave preferentially on the 3' side of pyrimidines, albeit at different rates and with somewhat different specificities. 'This work was supported by funds from Hoechst, A.G., under agreements with Harvard University. *Address correspondence to this author at the Center for Biochemical and Biophysical Sciences and Medicine, Seeley G . Mudd Building, Room 106, 250 Longwood Ave., Boston, MA 021 15.
0006-296019 110430-2246$02.50/0
Contrasting with the considerable amount of knowledge obtained concerning the primary structures and, in some instances, functional properties of these enzymes is the relative scarcity of information available concerning their distribution throughout the body. Thus, blood, urine, pancreas, and liver remain the only sources from which human RNases have been purified to homogeneity and definitively identified. The nature of the RNases in other tissues has been explored thus far I Abbreviations: RNase@), ribonuclease(s); RNase A, bovine pancreatic ribonuclease A; PRI, placental ribonuclease inhibitor; PR, placental ribonuclease; TSE, 20 mM Tris, pH 7.5, containing 0.25 M sucrose and 1 mM EDTA; CM, carboxymethyl; C18, octadecylsilane; HPLC, high-performance liquid chromatography;TFA, trifluoroacetic acid; SDS-PAGE, sodium dodecyl sulfate-polyacrylamide gel electrophoresis; Mes, 2-morpholinoethanesulfonicacid; CpA, cytidylyl-3',5'adenosine; UpA, uridylyl-3',5'-adenosine; HSA, human serum albumin; pHMB, p-(hydr0xymercuri)benzoate; EDN, eosinophil-derived neurotoxin; PTH, phenylthiohydantoin; ECP, eosinophil cationic protein.
0 199 1 American Chemical Society
Inhibition of Human Placental Ribonuclease primarily by indirect means, involving the examination of RNase contents of crude extracts by immunological (Neuwelt et al., 1977, 1978; Weickmann & Glitz, 1982; Morita et al., 1986; Sorrentino et al., 1988) or enzymatic (Neuwelt et al., 1978; Weickmann & Glitz, 1982) assays. The results of these studies suggest that both the pancreatic- and liver-type enzymes are widely distributed, with the former predominating in pancreas, brain, kidney, and heart and the latter in liver, spleen, lung, and stomach. In addition to RNases, mammalian tissues contain cytoplasmic RNase inhibitor proteins that bind to both the pancreatic- and liver-type enzymes (Roth, 1967; Blackburn & Moore, 1982). Placenta is a rich source of such an inhibitor (Blackburn et al., 1977), and the interactions of human placental RNase inhibitor (PRI) with bovine pancreatic RNase A and human angiogenin, an RNase homologue, have been examined in detail (Blackburn & Moore, 1982; Lee et al., 1989a,b). However, studies on the interaction of PRI with human tissue RNases, in particular, with the RNase(s) indigenous to placenta, have not been performed previously. In the present report, we describe the isolation and characterization of the major RNase from human placenta. This enzyme has catalytic and structural properties distinct from those of pancreatic RNase and appears to be identical with the predominant liver RNase (Sorrentino et al., 1988). The dissociation constant for its complex with PRI, calculated from the rate constants for association and dissociation, is extremely low, 9 X M. This value is about 40-fold lower than that measured with RNase A and similar to that obtained with angiogenin (Lee et al., 1989b). Changes in protein fluorescence accompanying placental RNase (PR).PRI complex formation have also been examined. EXPERIMENTAL PROCEDURES Materials. Recombinant human angiogenin was prepared as described (Kurachi et al., 1988) and was quantitated by amino acid analysis. UTP-agarose was prepared as described by Lamed et al. (1973). Sources of all other materials employed have been listed elsewhere (Shapiro et al., 1986a,b, 1987). Purification of RNase from Human Placenta. Placentas were obtained within 2 h after delivery, washed with TSE (20 mM Tris, 0.25 M sucrose, 1 mM EDTA, pH 7.5), and cut into small pieces. These pieces were then blotted on paper towels in order to remove as much blood as possible and stored at -70 OC. After thawing, tissue (600-900 g) was homogenized in TSE (2 mL/g) in a blender. All subsequent operations were performed at 4 OC unless indicated otherwise. Cell debris was removed by centrifugation for 30 min at 16000g. The supernatant was brought to pH 3 with glacial acetic acid and frozen at -20 OC. After 5 days, the homogenate was thawed and centrifuged as above. The supernatant was then filtered through Whatman 934-AH glass microfiber paper and dialyzed vs 20 mM sodium phosphate, pH 6.0. The retentate was brought to pH 6.0 with sodium hydroxide, centrifuged and filtered as above, and loaded onto a 2.5 X 35 cm column of CM-52 cation-exchange resin that had been equilibrated with 20 mM sodium phosphate, pH 6.0. Elution was achieved with a 4-L gradient from 0 to 1 M NaCl in the same buffer at a flow rate of 150 mL/h. Fractions from the major peak of RNase activity were pooled and dialyzed into 0.1 M Tris, pH 7.5, in an Amicon ultrafiltration device equipped with a YM 5 membrane. The sample (5 mL) was then centrifuged (1 5600g, 15 min) and the supernatant loaded onto a 1.6 X 98 cm column of Sephadex (3-50 (superfine) that had been equilibrated with 0.1 M Tris, pH 7.5, at a flow rate of 5 mL/h.
Biochemistry, Vol. 30, No. 8, 1991
2247
Fractions containing RNase activity were pooled and dialyzed into 25 mM piperazine, pH 5.3, in an Amicon ultrafiltration device. This material was then applied to a 0.8 X 5 cm column of UTP-agarose at a flow rate of 5 mL/h. After the column was washed with the piperazine buffer, RNase was eluted with 0.25 M sodium phosphate, pH 5.45 (Smith et a]., 1978). Fractions containing RNase activity were pooled, dialyzed into 20 mM sodium acetate, pH 5.5, and then applied to a Mono S cation-exchange column (5 X 50 mm; Pharmacia) equilibrated in the same buffer. Elution was achieved with a 50-min gradient from 0 to 0.4 M NaCl in the acetate buffer at 0.8 mL/min at ambient temperature. Material in the peak containing RNase activity, eluting at 36 min, was then chromatographed on a Synchropak C18 HPLC column (Synchrom Inc., Linden, IN) (10-pm particle size, 4.6 X 250 mm) as described in the legend for Figure 2. The final preparation was quantitated by amino acid analysis. Isolation of Angiogeninfrom Human Placenta. Angiogenin was isolated from placental homogenate by a modification of procedures previously employed for its purification from HT-29 cell conditioned medium (Fett et al., 1985) and plasma (Shapiro et al., 1987). Homogenate was acidified, frozen and thawed, concentrated, dialyzed, and chromatographed on CM-52 cation-exchange resin (50 mL of resin/kg of tissue) essentially as described (Fett et al., 1985). Material that binds to the CM-52 resin in 0.1 M sodium phosphate, pH 6.6, and elutes with 1 M NaCl was dialyzed vs water, lyophilized, and then chromatographed on phenyl and C18 reversed-phase HPLC columns (Shapiro et al., 1987). Physicochemical Characterization. SDS-PAGE was performed by using 15% gels as described by Laemmli (1970). Gels were silver-stained with a kit from ICN. Amino acid analyses (PicoTag method; Waters Associates) and N-terminal sequence analyses were performed as described (Strydom et al., 1985). UV spectra were recorded on a Varian 219 spectrophotometer. RNase Assays. Activity toward high molecular weight wheat germ RNA was measured by a modification of the precipitation assay of Blank and Dekker (1981). Reaction mixtures containing 33 mM Hepes, pH 7.3, 33 mM NaCI, 0.6 mg of RNA, 30 pg of HSA, and sample (diluted in 0.01% HSA) in a volume of 300 pL were incubated at 37 OC for 30 min. The reaction was stopped by addition of 700 p L of ice-cold 3.4% perchloric acid, and after 10 min on ice the mixtures were centrifuged at 15600g for 10 min at 4 OC. The absorbance of the supernatants was measured at 260 nm in a 1-cm cuvette. This assay was used throughout the isolation procedures. Changes in absorbance at 260 nm were linearly dependent on RNase concentration up to a value of -2.0. Activity toward polyhomoribonucleotides was determined by the method of Zimmerman and Sandeen (1965) as described (Shapiro et al., 1986b). Activity toward dinucleotides was measured either by continuous spectrophotometric assay (Witzel & Barnard, 1962) or by HPLC quantitation of products and starting material (Shapiro et al., 1986a,b). The spectrophotometric method was employed when CpA or UpA was the substrate; the HPLC method was used with all other dinucleotides. Unless specific otherwise, assays were performed with 0.1 mM substrate in 33 mM Mes, pH 6.0, containing 33 mM NaCl at 25 OC. Determination of Nucleotide at 3' Termini of PR Digest of R N A . A modification of previous methods (Farkas & Marks, 1968; Shapiro et al., 1986a) was employed. Wheat germ RNA was incubated with 0.1 nM PR for 40 min at 37
2248
Biochemistry, Vol. 30, No. 8, 1991
Shapiro and Vallee
OC as described above. Perchloric acid soluble products were treated with alkaline phosphatase and base-hydrolyzed as described (Shapiro et al., 1986a). This procedure should convert to free nucleoside any nucleotide donating a 3’phosphate to a cleaved bond, while the remainder are converted to a mixture of 2’- and 3’-nucleotides. The distribution of nucleosides and nucleotides was determined by C18 HPLC (Shapiro et al., 1986a). Fluorescence Measurements. Fluorescence measurements were performed at 25 OC with a Perkin-Elmer Model MPF-3 fluorescence spectrophotometer equipped with a Model 150 xenon lamp and a Hitachi Model QPD33 recorder. Emission spectra and quenching measurements were obtained as described by Lee et al. ( I 989a). Fluorescence quenching by acrylamide was analyzed by Stern-Volmer plots of Fo/Fvs acrylamide concentration, where Fo and F a r e the fluorescence intensities in the absence and presence of quencher, respectively. The Stern-Volmer quenching constant, K,,,was obtained from the initial slope of these plots. Samples (0.08 pM PR or 2 pM indole) were in 10 mM Tris, pH 7.5. Excitation was at 295 nm in order to minimize any contribution of tyrosines in PR. Emission was monitored at 340 nm for PR and at 348 nm for indole. Interaction of P R with PRI. The apparent second-order rate constant for association of PRI with PR was determined by examining the competition between PR and human angiogenin for PRI. PR (4.2 nM) was mixed with 0.6-1.4 equiv of angiogenin at 25 OC in 0.1 M Mes, pH 6.0, containing 0.1 M NaCl and 1 mM EDTA. PRI was then added to a final concentration of 4.2 nM. After 15 s, the activity of free PR was measured by adding the dinucleotide CpA (final concentration 100 pM) and continuously monitoring the decrease in absorbance at 286 nm. Angiogenin does not cleave CpA at a detectable rate. No significant dissociation of angiogenin (Lee et al., 1989b) or PR (see below) from their complexes with PRI occurs during the course of these assays. The apparent second-order rate constant of association was then calculated by using the equation
Table I: Purification of Human Placental RNase
= k a , A In ([RIT/[RIF)/ln ([AIT/[AIF) where ka,Ais the apparent second-order rate constant for association of PRI with angiogenin (1.8 X lo8 M-’ s-’ a nd [RIT, [R],, [AIT, and [AIF are the total PR, free PR, total angiogenin, and free angiogenin concentrations, respectively (Lee et al., 1989a). The PRI stock solution was quantitated by a titration plot that measures inhibition of PR in this assay system, assuming a 1:1 stoichiometry. This plot was linear, indicating that the PRI concentration was well above Ki. The rate constant for dissociation of the PR-PRI complex was measured by a modification of a previously described procedure (Lee et al., 1989b). PR (50 nM) and PRI (75 nM) were incubated for 20 min at 25 OC in 0.1 M Mes, pH 6, containing 0.1 M NaCl and 10 pg/mL HSA. Angiogenin was then added (14 pM final concentration) as scavenger for any free PRI dissociating from the complex, and the mixture was incubated at 25 OC. At various times, aliquots were removed and assayed for PR activity by using the CpA assay described above.
liberates any bound RNase, increased activity toward wheat germ RNA nearly 10-fold. A similar enhancement of activity was observed following acidification and freeze-thawing of the homogenate; addition of pHMB to an aliquot of the treated homogenate did not increase activity further. RNase was isolated from acidified/freeze-thawed homogenate by a five-step procedure summarized in Table I: CM52 cation-exchange chromatography, gel filtration on Sephadex G-50, affinity chromatography on UTP-agarose, Mono S cation-exchange HPLC, and C18 HPLC (see Experimental Procedures). About 70% of the RNase activity eluted from CM-52 as a single major peak (Figure 1). Subsequent gel filtration, affinity chromatography, and Mono S HPLC revealed a single peak of activity in each case. C18 HPLC (Figure 2), however, yielded two peaks, at 26.4 and 28.0 min, which contained RNase activity. Both the specific activities and the amino acid compositions of material in the two peaks were indistinguishable. Only material from the second, larger peak was employed in subsequent studies. The final preparation-1 10 pg/kg of starting material-contained 28% of the RNase activity measured in the initial homogenate after acidification and freeze-thawing. Assuming that the activity in the treated homogenate was primarily due to the species subsequently purified, the overall extent of purification was about 60000-fold. The specific activity of this protein (PR) toward high molecular weight wheat germ RNA is 70%of that measured with RNase A under the standard assay conditions.
ka.R
RESULTS Purification of RNase from Human Placenta. Most of the RNase activity in placental homogenate is present in “latent” form, presumably complexed with PRI, as observed previously (Blackburn et al., 1977). Thus, incubation of an aliquot of homogenate with 1 mM pHMB, a treatment that inactivates both free and complexed PRI (Blackburn et al., 1977) and
proteina (total mg) 22 000 16 800
RNase activityb (total units) 29 200 253 000
step vol (mL) extract 2000 acidification/ 2050 freeze-thawing 62.4 158 000 CM-52 195 13.1 135 000 Sephadex G-50 18.5 UTP-agarose 8.1 0.77 93 200 0.28 75 000 3.2 Mono S 0.10 70 300 C18 3.2 Protein quantitations were obtained by amino acid analysis for the UTP-agarose, Mono S, and C18 samples and by the procedure of Bradford (1976) for the remainder. bRNase activity was determined with wheat germ RNA as substrate as described in the text. RNase A (1 25 pg) was included as a standard for each set of assays performed. One unit of activity is defined as the activity of 1 ng of RNase A. 1.51
FRACTION NUMBER FIGURE 1:
Chromatograph of placental homogenate on CM-52. Homogenate from 900 g of placenta, prepared for chromatography as described in the text, was applied to a 2.5 X 40 cm column of CM-52 that had been equilibrated with 20 m M sodium phosphate, p H 6.0. After it was loaded, the column was washed with 200 mL of equilibration buffer. Elution was achieved with a 4-L linear gradient of 0-1.0 M NaCl in the same buffer. Fifteen-milliliter fractions were collected and assayed for activity toward wheat germ R N A at pH 7.3 (see Experimental Procedures). The activities shown represent values obtained with 5 p L of a 20-fold diluted aliquot from each fraction. Less than 10%of the RNase activity loaded onto the column eluted in the drop-through fractions (not shown). Fractions 57-69 were pooled for further purification.
Inhibition of Human Placental Ribonuclease
20
40
ELUTION TIME (MIN) FIGURE 2: Chromatograph of Mono S purified placental RNase on a Synchropak C18 HPLC column. Full-scale absorbance is 0.045. Solvent A was 0.1% TFA in water. Solvent B was 0.08% TFA in a 3:2:2 2-propanol/acetonitrile/watermixture. Elution was achieved with a 90-min linear gradient from 25% to 75% solvent B at 0.8 mL/min. Fractions were collected a t I-min intervals and assayed for RNase activity as described in the text. The proteins eluting at 26.4 and 28.0 min accounted for >90% of the RNase activity loaded onto the column.
Purification of Angiogenin from Placenta. The purification procedures just described were designed for the isolation of acid-stable RNases that are active toward high molecular weight RNA. Angiogenin, a blood vessel inducing protein that has ribonucleolytic activity (Shapiro et al., 1986b) and binds tightly to PRI (Shapiro & Vallee, 1987), is relatively inactive toward this substrate and therefore would not have been detected on this basis. In order to examine the content of this protein in placenta, an alternative protocol was employed (see Experimental Procedures), based on the known chromatographic properties of the protein. This procedure yielded about 6 pg/kg of tissue of a protein having the same phenyl and C18 HPLC elution times as angiogenin. Its identity as angiogenin was then established by SDS-PAGE and amino acid analysis (not shown). Some of this protein may have derived from blood present in the placental homogenate rather than from the tissue itself normal adult plasma contains -400 pg/L angiogenin (K. A. Olson, personal communication), while levels in fetal blood have not been determined. Regardless of the ultimate origin of the angiogenin isolated from placenta, it appears to be at least 10-fold less abundant than PR, assuming that the yield of angiogenin during the purification is not significantly lower than that for PR. [The yield of angiogenin from plasma utilizing the same procedure is 20-40% (unpublished results).] Physicochemical Characterization of PR. SDS-PAGE of the final C18 HPLC purified PR preparation revealed a single diffuse band of apparent molecular weight 18000 that stains poorly with silver (not shown). The amino acid composition of this protein (Table 11) is indistinguishable from those of human liver RNase (Sorrentino et al., 1988), RNase Us from urine (Beintema et al., 1988a), and eosinophil-derived neurotoxin (EDN) (Rosenberg et al., 1989; Hamann et al., 1989). Automated Edman degradation of the amino-terminal 22 residues indicates a sequence identical with that previously reported for the liver, urine, and eosinophil RNases (Gleich et al., 1986; Beintema et al., 1988a; Sorrentino et al., 1988):
Biochemistry, Vol. 30, No. 8, 1991 2249 Table 11: Amino Acid Compositions of Human Placental and Liver RNase and RNase U, amino acid placental RNase liver RNase” RNase U.6 Asx 21.5 21.2 21 Glx 14.0 14.9 14 Ser 7.3 8.0 6 GIY 2.4 2.9 2 His 4.6 5.0 5 ‘4% 8.1 8.3 8 Thr 12.6 13.0 12 Ala 6.4 6.7 6 Pro 12.0 12.6 12 TYr 4.0 4.2 4 Val 7.8 11.5 9 Met 4.0 3.4 4 Ile 5.6 6.8 7 Leu 4.8 5.5 5 Phe 5.1 5.4 5 LYS 4.I 4.6 4 TrP 1 .o ndc 1 CYS nd nd 8 “Calculated from mole percent values of Sorrentino et al. (1988) assuming 134 residues. bFrom sequence (Beintema et al., 1988). CNot determined. Table 111: Activities of Human PR and Bovine RNase A toward Dinucleotides k,,/K,“ (M-I s-I) dinucleotide PR RNase A 294 000 4 590 000 144000 3 230 000 541 476 000 133 141 000 20 139 000 11 24 300 2.5 12 200 41 300
