Inhibition of Ribonuclease P Activity by Retinoids*

The effect of two naturally occurring (retinol and all-trans retinoic acid) and two synthetic (isotretinoin and acitretin) analogs of vitamin A (retinoids) on tRNA biogenesis was investigated employing the RNase P ofDictyostelium discoideum as an in vitroexperimental system. RNase P is an ubiquitous and essential enzyme that endonucleolytically cleaves all tRNA precursors to produce the mature 5′ end. All retinoids tested revealed a dose-dependent inhibition of RNase P activity, indicating that these compounds may have a direct effect on tRNA biogenesis. Detailed kinetic analysis showed that all retinoids behave as classical competitive inhibitors. The K i values determined were 1475 μmfor retinol, 15 μm for all-trans retinoic acid, 20 μm for isotretinoin, and 8.0 μmfor acitretin. On the basis of these values acitretin is a 184, 2.5, and 1.9 times more potent inhibitor, as compared with retinol, isotretinoin, and all-trans retinoic acid, respectively. Taking into account that retinoids share no structural similarities to precursor tRNA, it is suggested that their kinetic behavior reflects allosteric interactions of these compounds with hydrophobic site(s) ofD. discoideum RNase P.

Retinoids, a group of natural and synthetic analogs of vitamin A, play an essential role in vision, growth, and reproduction as well as exhibiting striking effects on cell proliferation, differentiation, and pattern formation during development (1,2). The discovery that members of the steroid/thyroid hormone receptor superfamily are nuclear retinoic acid-binding proteins tremendously improves our understanding of the mechanisms that mediate the regulatory action of retinoids on gene expression (3)(4)(5)(6). The retinoid receptors are ligand-activated, DNAbinding, trans-acting, transcription factors (7,8).
Due to their ability to regulate cell differentiation and suppress or reverse the malignant phenotype, retinoids have a potential use as chemopreventive and chemotherapeutic agents in cancers of skin and other organs (9 -11). Moreover, oral synthetic retinoids are presently successfully applied in the management of severe and recalcitrant dermatoses, which were previously regarded as frustrating therapeutic problems (12)(13)(14).
RNase P is a key enzyme in tRNA biogenesis, which cleaves all tRNA precursors endonucleolytically to produce the mature 5Ј end. RNase P enzymes are composed of both RNA and protein (15). In vitro, RNA subunits of bacterial enzymes are catalytically active in the absence of protein (16) and are the only known RNA catalysts naturally devoted to act in trans (17). Catalytic activity of RNA subunits, in the absence of protein subunits, has not been demonstrated so far for archaea and eukaryotes RNase P. However, the catalytic center of these RNase P enzymes most likely is associated with the RNA subunits. Eukaryotic RNase P activity has been detected in nuclei, mitochondria, and chloroplasts (15,18). Recently, the partial purification and characterization of RNase P from the slime mold Dictyostelium discoideum has been reported (18,19).
In the present study we have examined the effect of natural and synthetic retinoids on RNase P using a cell free system from the slime mold D. discoideum as an in vitro model. D. discoideum RNase P is a nuclear enzyme that has low buoyant density in Cs 2 SO 4 gradients possibly due to the occurrence of an unidentified component such as a fatty acid or lipid (19). This hypothesis is consistent with the existence of hydrophobic site(s) on D. discoideum RNase P and perhaps justifies the role of retinoids as potential inhibitors, being hydrophobic molecules acting in the nucleus.

MATERIALS AND METHODS
Assay for RNase P Activity-Enzyme assays were carried out at 37°C in 20 l of buffer D (50 mM Tris/HCl, pH 7.6, 10 mM NH 4 Cl, 5 mM MgCl 2 , and 5 mM dithiothreitol) containing 2-5 fmol of tRNA substrate (an in vitro labeled transcript of the Schizosaccharomyces pombe tRNA Ser gene supSI) and 1.3 g of protein from the RNase P fraction. Stock solutions of retinoids (see Fig. 1), which were kindly supplied by Roche Hellas S.A. (Athens, Greece), were prepared in 100% Me 2 SO. Based on high pressure liquid chromatography (one single peak) and NMR spectroscopy (400 MHz NMR Bruker instrument), all retinoids used in the present study appeared to be highly pure. When retinoids were used, enzyme assays were carried out at 37°C in 20 l of buffer D in the presence of 10% Me 2 SO. The reactions were stopped by the addition of 5 l of stop dye (80% formamide, 50 mM EDTA, 0.1% bromphenol blue, 0.1% xylene cyanol). Reaction products were resolved on a denaturing 10% polyacrylamide/8 M urea gel and visualized by autoradiography without drying. Activity was quantified by Cerenkov counting of excised gel slices.
Enzyme Purification-Growth of D. discoideum cells (strain AX2 wild type) and cell breakage were essentially carried out as described previously (19). Broken cells were centrifuged for 10 min at 8000 ϫ g. The supernatant was removed and spun at 100,000 ϫ g for 1 h, yielding a 15-ml S-100 fraction. The S-100 fraction was loaded onto a DE 52 column (40 ml; 2.1 cm ϫ 24 cm) that had been equilibrated with AK 50 buffer. The column was washed with the same buffer until the A 280 dropped almost to zero. RNase P was then eluted with a 120-ml linear gradient of 50 -300 mM KCl in buffer A at a flow rate of 2 ml/min. Activity was eluted at 130 -150 mM KCl. The active fractions were pooled and dialyzed overnight against 4 liters of AK 50 buffer. The RNase P fraction was loaded onto a second DE 52 column (20 ml; 2.1 cm ϫ 24 cm) and eluted as described above. The active fractions were pooled and dialyzed against buffer D. At this step the purification was estimated to be 20-fold over S-100 fraction. A final gel filtration column was added to improve the overall purification of the RNase P. 0.6 ml of purified RNase P was loaded onto a Sephacryl S-300 column (24 ml; 1 cm ϫ 32 cm; fractionation range for proteins 1 ϫ 10 4 to 1.5 ϫ 10 6 daltons) and eluted with 40 ml of buffer D at a flow rate of 0.4 ml/min. Activity elutes with the void volume of the column. The active fractions were pooled, concentrated in dialysis bags with PEG 20.000, dialyzed against buffer D, and stored at Ϫ20°C in the presence of 50% glycerol. The overall purification was estimated to be 820-fold over the S-100 fraction.
Assay for RNase A Activity-Enzyme assays were carried out at room temperature in 20 l of buffer D containing 5 fmol of tRNA (an in vitro labeled transcript of the S. pombe tRNA Ser gene supSI) and 0.1 units of RNase A from bovine pancreas (Sigma), in the presence of 10% Me 2 SO. * This work was supported in part by the Greek Government (General Secretariat of Research and Technology, Ministry of Development). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
The reactions were stopped by addition of 5 l of stop dye, and reaction products were resolved on a denaturing 20% polyacrylamide/8 M urea gel and visualized by autoradiography without drying.

RESULTS
The present data show that all retinoids investigated here ( Fig. 1) are capable of inhibiting the RNase P activity of D. discoideum. The substrate for RNase P assays was an in vitro labeled transcript of the S. pombe tRNA Ser gene supSI (20).
The concentration of all-trans retinoic acid (IC 50 ) at which the product formation is reduced by 50% is equal to 80 M (Fig.  2, A and B). Although not as effective as all-trans retinoic acid, retinol also inhibited RNase P activity; the IC 50 was equal to 500 M (Fig. 2, C and D). The dose response curves for the synthetic retinoids isotretinoin and acitretin were similar to those observed with the natural retinoids (Fig. 3). The IC 50 values were 60 M (Fig. 3, A and B) and 40 M (Fig. 3, C and D), respectively. It is obvious from these results that acitretin is the strongest inhibitor among the retinoids tested. It is important to note that RNase P purified through the second DE 52 column was inhibited by retinoids to the same extent as RNase P purified through the S-300 column.
The type of inhibition of D. discoideum RNase P activity by the natural and synthetic retinoids was further elucidated by detailed kinetic analysis. Because the yields of RNase P after gel filtration chromatography are low, for the kinetic analysis we used RNase P fractions obtained from DEAE-cellulose chromatography. Due to the high hydrophobicity of retinoids, all assays were carried out in the presence of 10% Me 2 SO. At this concentration, Me 2 SO marginally affects (K i ϭ 4 M) the catalytic parameters of D. discoideum RNase P. The initial velocity in the presence or absence of retinoids was determined from the initial slopes of time plots (not shown). Fig. 4 shows double reciprocal plots with increasing concentrations of acitretin. The lowest line in Fig. 4 represents the data obtained in the absence of inhibitor and Me 2 SO (control). The calculated apparent K m (K m,app ) and the apparent V max (V max,app ) values from this plot are 240 nM and 3 pmol/min, respectively. These values are in agreement with previous values reported from our laboratory (19). In the presence of 10% Me 2 SO the K m,app and the V max,app values are 250 nM and 3 pmol/min, respectively (Fig. 4). The slopes of the lines in Fig. 4 were replotted against the concentration of acitretin, and the results are shown in the top panel of Fig.  4. The linearity of this replot is indicative of simple competitive inhibition and leads to the graphical determination of K i ϭ 8.0 M from the negative intercept of the line with the I-axis. Further evidence for simple competitive kinetics comes from the Dixon plot, which is shown in Fig. 5. When the slopes of the lines of Fig.  4 are replotted against 1/[pre-tRNA] (Fig. 5, top panel) they give a straight line passing through the origin. This is further evidence of simple competitive inhibition (21). Hill plots were obtained by calculating the log[vЈ/(v 0 -vЈ)] values at each concentration of pre-tRNA and plotting them as a function of logarithmic retinoid concentration. The molecular interaction coefficient (n) for retinoids was given by the slope of these plots. Fig. 6 shows the Hill plot for acitretin obtained at 100 nM pre-tRNA. The n value calculated from this plot is about equal to 1. These results support the notion that the inhibition of D. discoideum RNase P activity by retinoids involves only one retinoid molecule.
The same kinetic examination was carried out for isotretinoin, all-trans retinoic acid, and retinol. All these compounds showed simple competitive inhibition with molecular interaction coefficient equal to 1. The corresponding K i values are given in Table  I.
To address the question of whether retinoids may be capable of affecting also other nucleases, we investigated the effect of retinol, all-trans retinoic acid, isotretinoin, and acitretin on RNase A from bovine pancreas and found that these compounds exert no effect on RNase A activity (data not shown).

DISCUSSION
The study of the action of retinoids is of essential importance because these compounds are involved in gene expression, cell growth, and differentiation (1,2), represent the drugs of choice for a wide spectrum of severe and recalcitrant skin disorders (12)(13)(14), and are capable of reversing or suppressing carcinogenesis in many tissues (9 -11).
Retinoids act through binding to nuclear receptors that belong to the steroid/thyroid hormone superfamily. It has been proposed that target genes are regulated by these receptors that act as transcription factors activated by retinoids (3-6). However, there is also evidence indicating that retinoid action can be mediated through mechanisms not involving the retinoid nuclear receptors (22,23). In the present study it is demonstrated that the important ribozyme RNase P isolated from D. discoideum is competitively inhibited by natural and synthetic retinoids.
All retinoids tested revealed a dose-dependent inhibition of RNase P activity, indicating that these compounds may have a direct effect on tRNA biogenesis. Detailed kinetic analysis showed that the type of inhibition of retinol, all-trans retinoic acid, isotretinoin, and acitretin is simply competitive. According to this finding we could assign the potency of inhibitors solely on the basis of the K i value. Thus, acitretin is 184, 2.5, and 1.9 times more potent than retinol, isotretinoin, and alltrans retinoic acid, respectively.
An interesting observation is that substitution of the CH 3 O group for hydrogen in the P position of the aromatic ring of the parent compound (all-trans retinoic acid) (Fig. 1) enhances the inhibitory potency of the retinoid. Because the CH 3 O group increases the hydrophobicity of acitretin, it is possible that this compound may fit better into hydrophobic site(s) of the RNase P. It has been suggested that the low buoyant density of RNase P in Cs 2 SO 4 gradients (19) can be attributed to the occurrence of an unidentified component such as a fatty acid or lipid. This hypothesis is consistent with the existence of hydrophobic site(s) on RNase P. On the other hand, retinol, which is a very weak inhibitor (K i ϭ 1475 M), has a hydroxyl group in place of the carboxyl one (Fig. 1), indicating that the absence of the latter affects the binding of the analog with the enzyme. It is also interesting that among the stereoisomers (Fig. 1) all-trans retinoic acid (K i ϭ 15 M) is stronger inhibitor than isotretinoin (13-cis retinoic acid) (K i ϭ 20 M), suggesting that the stereochemical properties play an important role in the inhibitory behavior of these molecules.
A significant point of interest emerged from the Hill plot analysis (Fig. 6). Only one binding site (n Х 1) for retinoids seems to be necessary for the inhibition of RNase P reaction. Because of the kinetic nature of this observation, we cannot exclude the existence of additional retinoid-binding sites on RNase P holoenzyme, which are not involved in the kinetic model of inhibition.
Because the retinoids examined in the present study reveal no structural similarity to the substrate (pre-tRNA) of RNase P, the competitive character of the inhibition suggests that these compounds may bind to allosteric inhibition sites of the enzyme. Finally, it is possible that the inhibition of D. discoideum RNase P could have been caused by a retinoid-specific receptor complex; the receptor could have been co-eluted with RNase P activity during the purification procedure. However, this hypothesis can be ruled out, because RNase P activity during the purification through the Sephacryl S-300 column elutes with the void volume of the column, with the enzyme behaving as a protein with very high molecular mass (see "Materials and Methods"), whereas retinoic acid receptors, if present in the extract, will elute in later fractions well apart from the RNase P activity, because of their low molecular mass (Ͻ50 kDa) (3). Furthermore, it could be speculated that the inhibition of RNase P activity results from retinoids titrating out a co-activator component from RNase P, for example a bivalent cation like Mg 2ϩ ; nevertheless, this possibility is definitely ruled out because in this case the inhibition pattern is of partial noncompetitive type. 1 Moreover, other types of RNase P co-activators have not been reported so far. 1 A. Tekos, C. Stathopoulos, and D. Drainas, unpublished results.