Aminoglycoside binding to human and bacterial A-Site rRNA decoding region constructs.

The 16S bacterial ribosomal A-site decoding rRNA region is thought to be the pharmacological target for the aminoglycoside antibiotics. The clinical utility of aminoglycosides could possibly depend on the preferential binding of these drugs to the prokaryotic A-site versus the corresponding A-site from eukaryotes. However, quantitative aminoglycoside binding experiments reported here on prokaryotic and eukaryotic A-site RNA constructs show that there is little in the way of differential binding affinities of aminoglycosides for the two targets. The largest difference in affinity is 4-fold in the case of neomycin, with the prokaryotic A-site construct exhibiting the higher binding affinity. Mutational studies revealed that decoding region constructs retaining elements of non-Watson-Crick (WC) base pairing, specifically bound aminoglycosides with affinities in the muM range. These studies are consistent with the idea that aminoglycoside antibiotics can specifically bind to RNA molecules as long as the latter have non-A form structural elements allowing access of aminoglycosides to the narrow major groove.

Reduced levels of retinyl esters and vitamin A in human renal cancers.

Clinical and preclinical studies suggest that retinoids can inhibit the growth of a small percentage of human renal cancers (RCs), although the majority of RCs both in vitro and in vivo are retinoid resistant. Our recent studies indicate that the metabolism of retinol to retinyl esters is greatly reduced in human carcinoma cell lines of the oral cavity, skin, and breast as compared with their normal epithelial counterparts, suggesting that human carcinoma cells are retinoid deficient relative to normal epithelial cells. We considered whether retinoid resistance in RCs was related to an abnormality in retinoid metabolism. The metabolism of [3H]retinol and of [3H]retinoic acid (RA) was examined in RC cell lines and normal human kidney (NK) epithelial cells cultured in media, in RA, or in RA plus IFN-alpha. The expression of LRAT (lecithin:retinol acyltransferase) was assessed by Northern and Western analysis. Retinol and retinyl ester levels were determined in tissue samples of normal human kidney and renal cell carcinoma. NK cells esterified all of the 50 nM [3H]retinol in which they were cultured. In contrast, six of the seven RC cell lines metabolized only trace amounts of [3H]retinol to [3H]retinyl esters. Consistent with this relative lack of [3H]retinol esterification by the tumor cells, the tumor cells exhibited LRAT transcripts of aberrantly low sizes relative to those in normal epithelial cells. Moreover, the NK cells expressed abundant levels of LRAT protein by Western analysis, whereas the RC cells did not express LRAT protein. When samples of human kidney tumor tissue were compared with samples of normal kidney tissue from patients who had undergone surgery for primary RC, the normal kidney tissues contained much higher levels of retinol and retinyl esters (approximately 0.5-2 microg/gram wet weight) than the tumor tissues in all seven patients examined. Culture of the RC lines in IFN-alpha plus all-trans-RA, a combination therapy used clinically, resulted in higher intracellular levels of [3H]retinol and [3H]retinyl esters. The metabolism of [3H]RA was also examined in these RC lines versus NK cells. Although the NK epithelial cells metabolized [3H]RA, the majority of the RC lines metabolized [3H]RA at a much slower rate. Most of the RC lines metabolized only 10-30% of the 50 nM [3H]RA over 6 h of culture. These data indicate that RCs both in vitro and in vivo are retinol and retinyl ester deficient relative to the normal human kidney, and they suggest that the aberrant differentiation of the neoplastic renal cells results in part from a defect in retinoid metabolism.

Two histidine residues are essential for catalysis by lecithin retinol acyl transferase.

Lecithin retinol acyl transferase (LRAT) is a novel membrane bound enzyme that catalyzes the formation of retinyl esters from vitamin A and lecithin. The enzyme is both essential for vision and for the general mobilization of vitamin A. The sequence of LRAT defines it as a novel enzyme unrelated to any other protein of known function. LRAT possesses a catalytically essential active site cysteine residue. The enzyme also contains six histidine residues. It is shown here that two of these residues (H57 and H163) are essential for catalysis. A mechanistic hypothesis is presented to account for these observations.

Esterification of all-trans-retinol in normal human epithelial cell strains and carcinoma lines from oral cavity, skin and breast: reduced expression of lecithin:retinol acyltransferase in carcinoma lines.

When exogenous [(3)H]retinol (vitamin A) was added to culture medium, normal human epithelial cells from the oral cavity, skin, lung and breast took up and esterified essentially all of the [(3)H]retinol within a few hours. As shown by [(3)H]retinol pulse-chase experiments, normal epithelial cells then slowly hydrolyzed the [(3)H]retinyl esters to [(3)H]retinol, some of which was then oxidized to [(3)H]retinoic acid (RA) over a period of several days. In contrast, cultured normal human fibroblasts and human umbilical vein endothelial cells (HUVEC) did not esterify significant amounts of [(3)H]retinol; this lack of [(3)H]retinol esterification was correlated with a lack of expression of lecithin:retinol acyltransferase (LRAT) transcripts in normal fibroblast and HUVEC strains. These results indicate that normal, differentiated cell types differ in their ability to esterify retinol. Human carcinoma cells (neoplastically transformed epithelial cells) of the oral cavity, skin and breast did not esterify much [(3)H]retinol and showed greatly reduced LRAT expression. Transcripts of the neutral, bile salt-independent retinyl ester hydrolase and the bile salt-dependent retinyl ester hydrolase were undetectable in all of the normal cell types, including the epithelial cells. These experiments suggest that retinoid-deficiency in the tumor cells could develop because of the lack of retinyl esters, a storage form of retinol.

RNA aptamers that specifically bind to a 16S ribosomal RNA decoding region construct.

RNA-RNA recognition is a critical process in controlling many key biological events, such as translation and ribozyme functions. The recognition process governing RNA-RNA interactions can involve complementary Watson-Crick (WC) base pair binding, or can involve binding through tertiary structural interaction. Hence, it is of interest to determine which of the RNA-RNA binding events might emerge through an in vitro selection process. The A-site of the 16S rRNA decoding region was chosen as the target, both because it possesses several different RNA structural motifs, and because it is the rRNA site where codon/anticodon recognition occurs requiring recognition of both mRNA and tRNA. It is shown here that a single family of RNA molecules can be readily selected from two different sizes of RNA library. The tightest binding aptamer to the A-site 16S rRNA construct, 109.2-3, has its consensus sequences confined to a stem-loop region, which contains three nucleotides complementary to three of the four nucleotides in the stem-loop region of the A-site 16S rRNA. Point mutations on each of the three nucleotides on the stem-loop of the aptamer abolish its binding capacity. These studies suggest that the RNA aptamer 109.2-3 interacts with the simple 27 nt A-site decoding region of 16S rRNA through their respective stem-loops. The most probable mode of interaction is through complementary WC base pairing, commonly referred to as a loop-loop 'kissing' motif. High affinity binding to the other structural motifs in the decoding region were not observed.

Isoprenylation/methylation and transducin function.

Freshly prepared proteolyzed (deprenylated) T beta gamma and material isolated from retina are inert with respect to activating T alpha in the presence of R* in detergent and in disk membranes. In addition, proteolyzed T beta gamma is also incapable of supporting the pertussis toxin-catalyzed ADP ribosylation of T alpha-GDP. These experiments show that isoprenylation/methylation is essential for the fruitful interactions between T alpha and T beta gamma at the membrane. When tested for its ability to support GTP-for-GDP exchange catalyzed by R*, demethylated T beta gamma proved to be approximately 50% as active as methylated T beta gamma in photoreceptor disk membranes (Fig. 3) and in reconstituted liposomes containing rhodopsin. In detergent, no difference was observed between methylated and demethylated T beta gamma, suggesting no role at all for the methyl group in functional interactions between T alpha, T beta gamma, and R*. The twofold activity difference observed in membranes can be accounted for by the twofold lessened affinity of the demethylated T beta gamma, compared with its methylated counterpart, for membranes in the presence of R* and T alpha. It is interesting to note that a substantially larger difference (> 10-fold) in the relative binding of methylated versus demethylated T beta gamma to membranes is observed in the absence of R* and T alpha. However, R* has a substantial affinity for T alpha beta gamma, and the influence of R* and T alpha greatly reduces any differences resulting from the presence or absence of a methyl group on T beta gamma. The results from studies of demethylated T beta gamma demonstrate that specific lipid-receptor interactions are unlikely to play a critical role in the rhodopsin-transducin system, and further show that the effect of methylation is probably due to the increased hydrophobicity of methylated T beta gamma versus its unmethylated counterpart. These studies are, of course, relevant to heterotrimeric G proteins, and specifically to the interactions of receptor (R*) with T alpha and T beta gamma. If a hydrophobic lipid-lipid mechanism is operative, the state of methylation would be expected to have a more profound effect on the membrane-associative properties of farnesylated proteins, but not on those of geranylgeranylated proteins. The increased hydrophobicity of the C20 geranylgeranyl group relative to the C15 farnesyl group will compensate for the loss of the methyl substituent. The results obtained in the transducin-rhodopsin system can be contrasted with the effect of gamma-subunit methylation on effector enzyme activation. In the case of the geranylgeranylated beta 1 gamma 2, methylation proved to have only a small effect on PIPLC beta activation (Fig. 4B). An approximately 25% diminution in efficacy, but not potency, was observed for the demethylated geranylgeranylated beta 1 gamma 2 versus its methylated counterpart. This again shows that specific lipid-protein interactions are unimportant. The effect of methylation on membrane binding would be expected to be small, given that beta 1 gamma 2 is geranylgeranylated. It is of interest to compare these results with those found with methylated and unmethylated T beta gamma as activators of PIPLC beta. In this instance there was a large effect noted, with methylated T beta gamma being at least 10-fold more potent than its unmethylated counterpart with respect to activating either enzyme (Fig. 4A). This result is readily understandable in light of the role of methylation in selectively enhancing hydrophobicity of farnesylated proteins as opposed to geranyl-geranylated proteins. Similar results were obtained for the activation of PI3K, further strengthening the conclusion that it is lipid-lipid interactions that direct beta gamma subunit membrane association. (ABSTRACT TRUNCATED)

Lecithin retinol acyltransferase contains cysteine residues essential for catalysis.

Lecithin retinol acyltransferase (LRAT) is an essential enzyme in vitamin A metabolism and mobilization. The membrane-bound enzyme catalyzes the transfer of an acyl group from the sn-1 position of lecithin to vitamin A to generate retinyl esters. The sequence of LRAT is novel and hence does not suggest a mechanistic class to which the enzyme belongs. However, the activity of the enzyme is exceedingly sensitive to affinity labeling and group-specific reagents directed toward thiol groups. LRAT from human retinal pigment epithelium has cysteine residues at positions 161, 168, 182, and 208. Site-specific mutagenic studies show that C182 and C208 can be converted to alanines with little affect on activity. The activities of the C161A and C168A mutants are virtually nil. Moreover, while C168S is substantially active, C161S possesses only a few percent of the activity of wild-type (WT) LRAT. Also, pH-rate profiles show that C168S has virtually the same profile as WT LRAT, while C161S shows an aberrant profile quite unlike that of WT LRAT. Therefore, LRAT is a thiol acyltransferase and C161 may be the essential nucleophilic residue critical for catalysis.

Specific binding of Hoechst 33258 to site 1 thymidylate synthase mRNA.

The translational initiator codon in thymidylate synthetase (TS) mRNA is located in a stem-loop structure with a CC bubble. TS is an important target for anticancer drugs. Aminoglycoside antibiotics have been shown to specifically bind to TS mRNA site 1 constructs and, furthermore, specific binding requires the non-duplex CC bubble region. It is shown here that DNA intercalating agents and DNA minor groove-binding drugs also bind to a TS mRNA site 1 construct. This binding is competitive with aminoglycosides, suggesting that the binding sites overlap. Hoechst 33258 binds with a dissociation constant of 60 nM, a value significantly lower than the approximately 1 microM values found for aminoglycosides. Footprinting and direct binding studies show that the CC bubble is important for binding of the Hoechst compound. However, the exact structure of the bubble is unimportant. Interestingly, mutations in regions adjacent to the bulge also affect binding. These studies point to the important role of non-duplex RNA structures in binding of the DNA minor groove binder Hoechst 33258.

The specific binding of small molecule isoprenoids to rhoGDP dissociation inhibitor (rhoGDI).

The activities of small G-proteins are in part regulated by their interactions with GDI proteins. This binding is thought to be dependent on the C-terminal isoprenoid modification (geranylgeranyl or farnesyl) of these proteins. G-proteins are generally isoprenylated/methylated at their C-terminal cysteine residues. A quantitative fluorescence assay is reported here to evaluate the specificity of binding of rhoGDI. A rhodamine-labeled geranylgeranylated/methylated cysteine derivative is used to measure its binding to rhoGDI. Saturable binding in the low micromolar range is found with various geranylgeranylated/farnesylated analogues. Interestingly, the carboxymethylated derivatives bound significantly better than their free acid counterparts, suggesting that the state of methylation of the analogues is important for binding. The binding is also selective with respect to isoprenoid. Analogues containing hydrophobic modifications other than geranylgeranyl or farnesyl do not bind with significant affinities. These data demonstrate a substantial degree of specificity in the binding of isoprenoids to a protein important in signal transduction.

Specificity in the binding of aminoglycosides to HIV-RRE RNA.

Quantitative studies of the binding of neomycin B to RRE constructs are carried out to determine the relationship between non-Watson Crick base-paired elements in the RNA and aminoglycoside binding. The RRE region contains two unpaired domains containing a single base bulge and a bubble structure, respectively. Deletion of the single base bulge has no effect on neomycin binding as the site of aminoglycoside binding is localized to the bubble region. Converting the bubble region into an A-form duplex gradually abolishes neomycin B binding in 3-5-fold steps in affinity over a 75-fold range. Thus, the binding of aminoglycoside is favored at domains in RNA that are nonduplex in nature, but aminoglycoside binding is only graded-specific in that affinities are enhanced gradually as the structure further deviates from a duplex form. It is likely that high-affinity aminoglycoside binding does not occur in duplex RNA because the major groove is too narrow to allow for aminoglycoside access and that structural perturbations that allow widening of the groove facilitate access. However, these interactions are only graded-specific with respect to both aminoglycoside structure and RNA domain structure.

Aminoglycoside antibiotics are able to specifically bind the 5'-untranslated region of thymidylate synthase messenger RNA.

The translational initiation codon for thymidylate synthase (TS) mRNA is located in a unique stem-loop structure which contains an internal cytosine-cytosine (CC) bubble. This stem-loop structure is thought to be important in the regulation of TS translation, which is itself an important target for anticancer drugs, such as 5-fluorouracil. Internal bubble or bulge structures are candidate receptors for the aminoglycoside antibiotics. It is shown here that aminoglycosides bind in a specific and saturable fashion with dissociation constants of approximately 1 microM to a TS mRNA site 1 construct and that the binding site for the aminoglycosides is located in the CC bubble region. In fact, the CC bubble, when grafted into other stem-loop structures, confers aminoglycoside binding on them. These studies reveal an additional binding domain for aminoglycosides and also suggest how novel anti-cancer drugs might be designed that affect TS mRNA translation rather than enzyme function.

Molecular and biochemical characterization of lecithin retinol acyltransferase.

The enzyme responsible for conversion of all-trans-retinol into retinyl esters, the lecithin retinol acyltransferase (LRAT) has been characterized at the molecular level. The cDNA coding for this protein was cloned and its amino acid sequence deduced. LRAT is composed of a polypeptide of 230 amino acid residues with a calculated mass of 25.3 kDa. Tissue distribution analysis by Northern blot showed expression of a 5.0-kilobase transcript in the human retinal pigment epithelium as well as in other tissues that are known for their high LRAT activity and vitamin A processing. Affinity labeling experiments using specific compounds with high affinity for LRAT and monospecific polyclonal antibodies raised in rabbits against two peptide sequences for LRAT confirmed the molecular mass of LRAT as a 25-kDa protein. High performance liquid chromatography analysis of the reaction product formed by HEK-293 cells transfected with LRAT cDNA confirmed the ability of the transfected cells to convert [3H]all-trans-retinol into authentic [3H]all-trans-retinyl palmitate as chemically determined.

Lack of effect of RPE65 removal on the enzymatic processing of all-trans-retinol into 11-cis-retinol in vitro.

RPE65 is a major membrane associated protein found in the vertebrate retinal pigment epithelium (RPE). Various studies have shown this protein to be essential for visual function, possibly at the level of the processing of retinoids. The pigment epithelium is the anatomical site in which the visual chromophore 11-cis retinal is generated. The two critical RPE enzymes in the isomerization pathway are lecithin retinol acyl transferase (LRAT) and isomerohydrolase, which processes all-trans-retinyl esters into 11-cis-retinol. Both enzymes are membrane bound. It is shown here that RPE65 can be largely extracted (90-95%) from RPE membranes by 1 M KCl by itself, or with added detergent CHAPS. The almost quantitative extraction of RPE65 from RPE membranes has little or no effect on in vitro LRAT and isomerohydrolase activities in quantitative enzymatic assays using RPE membranes, suggesting that RPE65 may not have an important role to play in the enzymatic processing of all-trans-retinol into 11-cis-retinol in vitro.

A high-throughput fluorescence screen to monitor the specific binding of antagonists to RNA targets.

Since RNA molecules can form intricate three-dimensional structures, it should be possible to design specific, high-affinity antagonists directed against these structures. To begin to explore the validity of this possibility, high-throughput screening methods are required to assay for RNA antagonists. A fluorescence quenching technique is described here in a 96-well plate format which is capable of screening chemical diversity libraries. A pyrene-containing aminoglycoside analog is used to accurately monitor antagonist binding to a prokaryotic 16S rRNA A-site decoding region construct. This rRNA region comprises the natural target for aminoglycoside antibiotics. The fluorescence technique reported here should be generally adaptable to monitor the binding of structurally novel antagonists to any selected RNA target.

The binding site of a specific aminoglycoside binding RNA molecule.

A small (40 nucleotides) stem-loop derivative (J6f1) of a specific aminoglycoside-binding RNA aptamer, containing a 3 nt and a 1 nt bulge, has previously been shown to stoichiometrically bind tobramycin with a dissociation constant of approximately 5 nM [Hamasaki, K., Killian, J., Cho, J. and Rando, R. R. (1997) Biochemistry 36, 1367-1371]. This construct can strongly discriminate among similar aminoglycosides with respect to binding. A combination of chemical interference studies, chemical modification studies, and mutational studies are performed to define the aminoglycoside binding site of J6f1. Recognition of the aminoglycoside by J6f1 involves contacts with nucleotide bases, rather than with the phosphate backbone. The binding site 1 comprised of part of the stem-loop region. The two bulges are also essential for high affinity and stoichiometric binding of tobramycin. These bulges are probably important for prying open the double helical region, thereby allowing the aminoglycoside access to the nucleotide bases.

Analogs of farnesylcysteine induce apoptosis in HL-60 cells.

S-Farnesyl-thioacetic acid (FTA), a competitive inhibitor of isoprenylated protein methyltransferase, potently suppressed the growth of HL-60 cells and induced apoptosis, as evidenced by the development of increased annexin-V binding, decreased binding of DNA dyes and internucleosomal DNA degradation. FTA did not impair the membrane association of ras proteins, conversely, it brought about a decrease in the proportion of ras present in the cytosolic fraction. Farnesylated molecules which are weak inhibitors of the methyltransferase also induced DNA laddering and reduced the proportion of cytosolic ras. These findings suggest that neither inhibition of isoprenylated protein methylation nor impairment of ras membrane association are essential for apoptosis induced by farnesylcysteine analogs.

Regulation of isomerohydrolase activity in the visual cycle.

While the overall biosynthetic pathway leading from all-trans-retinoids to 11-cis-retinoids in the visual cycle is understood, little is known about which step(s) may be rate-limiting and how control is exerted. One possible target for control is the isomerohydrolase, which processes all-trans-retinyl esters into 11-cis-retinol. The basal rate of 11-cis-retinol synthesis from all-trans-retinyl esters is extremely slow using bovine retinal pigment epithelial membranes [3.5 pmol of 11-cis-retinol min-1 (mg of protein)-1], and only small amounts of 11-cis-retinyl ester are formed. However, the addition of retinol binding proteins stimulates 11-cis-retinol formation by a factor of approximately 13. Specific protein-protein interactions are probably unimportant because bovine serum albumin and the physiologically relevant cellular retinaldehyde binding protein (CRALBP) both stimulate 11-cis-retinol formation to the same extent, although CRALBP does so at much lower concentrations. The relatively rapid rate of isomerization in the presence of binding proteins [44.3 pmol of 11-cis-retinol min-1 (mg of protein)-1] suggests that the rate-limiting enzyme in the visual cycle need not be the isomerohydrolase. Also, 11-cis-retinol is shown to inhibit isomerohydrolase, providing a simple mechanism for regulation of the visual cycle and the stimulating effect of binding proteins.

Minimal RNA constructs that specifically bind aminoglycoside antibiotics with high affinities.

RNA molecules are the functional targets for aminoglycosides. In order to approach an understanding of the rules which underlie aminoglycoside-RNA recognition, high-affinity RNA aptamers have been prepared which discriminate among various aminoglycosides [Wang et al. (1996) Biochemistry 35, 12338-12346]. One of these aptamers, J6, which is 109 nts in length, binds the aminoglycoside tobramycin stoichiometrically with a dissociation constant of 0.77 +/- 0.03 nM. Aminoglycosides, similar in structure to tobramycin, bind with affinities diminished by 10(3)-10(4) compared to tobramycin. Experiments are reported here which are designed to reveal the nature of the tobramycin binding domain of J6. A small (40 nts) stem-loop derivative of J6, containing a 3 nt and a 1 nt bulge, stoichiometrically binds tobramycin with a dissociation constant of approximately 5 nM. This construct can strongly discriminate between similar aminoglycosides with respect to binding. Elimination of either the three or the single nucleotide bulge eliminates specific aminoglycoside binding. The structure of the loop region is also critical. These studies demonstrate that simplified RNA molecules can be generated which bind aminoglycosides specifically and with high affinities.