INTRODUCTION

Top Abstract Introduction Materials and Methods Results Discussion References

The polycationic polyamines putrescine, spermidine, and spermine are present in all living organisms and engage in noncovalent interactions with a wide variety of cellular targets, including nucleic acids, proteins, and phospholipids (26, 29, 39). These interactions affect various processes of cell growth. For instance, preferential stimulation or inhibition of the in vivo synthesis of specific proteins is one of the important functions of polyamines in cell growth and regulation of differentiation (see reference 14 and references therein). Due to its four positive charges at physiological pH, spermine is the most effective of the naturally occurring polyamines both in regulating the in vitro translation process at several levels and in decreasing (but not abolishing) the Mg²⁺ requirements for protein synthesis (8, 11, 25-27, 39). We have previously demonstrated (3) that in an Escherichia coli cell-free system, spermine at 6 mM Mg²⁺ displays a concentration-dependent allosteric biphasic activity on ribosomal peptidyltransferase. In agreement with this, accumulation of excess polyamines causes inhibition of cell growth or a decrease in cell viability, primarily through inhibition of protein synthesis (9). On the other hand, blockage of polyamine synthesis by mutations or by inhibitors leads to a virtual cessation of growth, unless exogenous polyamines are provided (29). Accumulating evidence suggests that these inhibitors may be useful therapeutic agents for treatment of a variety of diseases, including cancer (23, 29).

N¹,N¹²-Bis(ethyl)spermine, the most active derivative in depleting intracellular polyamine pools, not only negatively regulates the synthesis of ornithine and S-adenosylmethionine decarboxylases (30, 32), but also induces spermidine/spermine N¹-acetyltransferase and substitutes for the functions of spermine in several aspects (4, 6, 10, 12, 23, 29). This analogue, with minor exceptions (24), enters mammalian cells via the polyamine transport system (1). Other analogues, such as dibenzyl polyamine analogues, are substrates for an uptake system distinct from the natural polyamine transport system (2).

The use of specific transport inhibitors is a useful strategy to understand the differential specificity of polyamine transporters. A great deal of effort has been invested in probing the molecular details of polyamine recognition by the transporters and in the isolation and expression of genes encoding the constituents of uptake systems. Although attempts in mammalian systems have not yet been successful, the bacterial polyamine transporters have been cloned and expressed. In E. coli cells, the proteins encoded by potABCD and potFGHI operons constitute the spermidine- or spermine-preferential and the putrescine-specific uptake systems, respectively (13). Another transport system encoded by potE also catalyzes putrescine uptake; however, its ability is significantly lower than those of the potABCD and potFGHI systems. The substrate specificity of the two transport systems is determined by a polyamine-binding protein in the periplasm: PotF for putrescine and PotD for putrescine or spermidine. The amino acid residues in PotD and PotF involved in the interaction with polyamines have been revealed by mutational and x-ray analysis (20, 37, 40). Recently, we have studied in E. coli the interactions of acetyl polyamines with their transporters (17) or peptidyltransferase (18), and we have evaluated the significance of polyamine primary and secondary amino groups, as well as that of chain flexibility, as determinants of these bacterial functions. Since acetyl polyamines per se have no pharmaceutical significance, it was of interest to extend our knowledge by using a series of spermine analogues which are known to have an antiproliferative effect on eukaryotic cells. An investigation of their effects on prokaryotic cells could have an impact not only on basic research, but also on interpretation of potential symbiotic relationships between prokaryotic and eukaryotic cells.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results Discussion References

Materials. GTP (disodium salt), poly(U), ATP (disodium salt), phenylalanine, puromycin dihydrochloride, heterogeneous tRNA from E. coli W, spermine tetrahydrochloride, N¹-acetylspermine, and vitamin B₁ were obtained from Sigma Chemical Co. (St. Louis, Mo.). L-Phenyl-[2,3-³H]alanine, [¹⁴C]spermine tetrahydrochloride, and [1,4-¹⁴C]putrescine dihydrochloride were purchased from Amersham (Arlington Heights, Ill.). Cellulose acetate filters (type Sepraphore III) were obtained from Gelman Sciences (Ann Arbor, Mich.), and cellulose nitrate filters (type HA, 24-nm diameter, 0.45-µm pore size) were obtained from Millipore (Bedford, Mass.). Casamino Acids were obtained from Difco Laboratories (Detroit, Mich.).

Drugs. N¹-Ethylspermine, N¹,N¹²-bis(ethyl)spermine, and N⁴,N⁹-bis(ethyl)spermine were synthesized by reduction of the corresponding acetyl analogues of spermine with lithium aluminum hydride for 12 h in refluxing tetrahydrofuran (THF). N¹-Acetylspermine, N¹,N¹²-diacetylspermine, N⁴,N⁹-diacetylspermine, and N⁴,N⁹-dibenzylspermine were synthesized by the application of a general methodology (22), with N-tritylamino acids used as building blocks.

Biochemical preparations. Salt-washed (0.5 M NaCl) and polyamine-depleted ribosomes were obtained from E. coli B cells, as described previously (15). Partially purified translation factors (FWR fraction) and crude acetyl-(Ac)[³H]Phe-tRNA charged with 16.3 pmol of [³H]Phe (86 kcpm total) per A₂₆₀ unit were prepared as reported previously (15). Complex C, i.e., the Ac[³H]Phe-tRNA-poly(U)-ribosome complex, was prepared and purified through adsorption on cellulose nitrate filters, as described elsewhere (18). The radioactivity trapped on the filters was counted in a liquid scintillation spectrometer. Controls without poly(U) were included in each experiment, and the values obtained were subtracted.

Cell culture. E. coli B cells were grown aerobically in M9 medium (48 mM Na₂HPO₄, 22 mM KH₂PO₄, 9 mM NaCl, 19 mM NH₄Cl), supplemented with 0.03 mM FeCl₃, 0.1 mM CaCl₂, 1 mM MgSO₄, 0.01 mM vitamin B₁, 0.6% glucose, and 0.1% Casamino Acids, at 37°C in shaking Erlenmeyer flasks. Polyamine analogues (100 µM, total concentration) were added at the time of culture initiation, and growth was followed by measuring the A₅₄₀. Polyamine pool depletion was tested by treating cells for up to 24 h with analogues. Control cells were washed with ice-cold phosphate-buffered saline prior to the performance of uptake assays with radiolabeled polyamines.

Polyamine pool analysis. Cells were sampled at the end of each treatment period, washed, and pelleted for extraction with 0.6 N perchloric acid. Each supernatant, after dansylation (28), was analyzed for dansyl polyamines or polyamine analogues by reverse-phase high-performance liquid chromatography (HPLC) by using a Beckman C₁₈-Ultrasphere ODS (5-µm spherical packing, 25 cm by 4.6 mm) column heated at 50°C. For the separation of derivatized polyamines, an initial 3-min isocratic elution with 45% aqueous CH₃CN was followed by an 18-min linear gradient to methanol and then by methanol for 9 min. Results were expressed as micromoles of polyamine or polyamine analogue per gram of total protein. The concentration of proteins was determined by the method of Lowry et al. (21).

Kinetic analysis of polyamine uptake. Polyamine uptake by intact cells was induced and kinetically analyzed as described previously (17), with various concentrations of [¹⁴C]putrescine or [¹⁴C]spermine in the presence or in the absence of specified concentrations of polyamine analogues.

Peptide bond formation assay and first-order analysis. The peptidyltransferase activity of ribosomes was assessed by the puromycin reaction carried out at 25°C in the presence of 6 mM Mg²⁺ by using complex C adsorbed on a cellulose nitrate filter. Under these conditions, the reaction between complex C and excess puromycin (S) proceeds as an irreversible pseudo-first-order reaction (18):
 

(1)

(2)

(3)

View larger version (20K): [in this window] [in a new window]   FIG. 1.   Kinetic model for AcPhe-puromycin synthesis carried out in the presence of spermine analogues, with complex C formed in the presence of the FWR fraction. C, complex C; S, puromycin; I and I(A), spermine analogue bound to the inhibition site and activation site of complex C, respectively; C', CI', CI'(A), and CI(A)I', ribosomal complexes after their reaction with puromycin.

	RESULTS

Top Abstract Introduction Materials and Methods Results Discussion References

Effect of spermine analogues on ribosomal functions. (i) Effect on the stability of complex C. The stability of complex C, i.e., the AcPhe-tRNA-poly(U)-ribosome complex, was examined at 6 mM Mg²⁺ by incubation of complex C at 25°C with or without spermine analogues and monitoring the percentage of surviving complex C. In accordance with previous results (16, 18), we observed that 58% of complex C formed in the absence of translation factors (FWR fraction) was inactivated after 20 min of incubation in the absence of spermine or spermine analogues. However, the stability of complex C was greatly enhanced by the presence of spermine analogues, in a dose-dependent manner. Complex C formed in the presence of the FWR fraction exhibited high stability, and therefore the presence of spermine analogues caused only a slight improvement.

(ii) Effect on poly(U)-directed Ac[³H]Phe-tRNA binding to ribosomes. In the absence of spermine or spermine analogues, the optimum concentration of Mg²⁺ for AcPhe-tRNA binding to poly(U)-programmed ribosomes has been found in our laboratory to equal 10 mM (see reference 3 and references therein). Addition of spermine or spermine analogues at each optimal concentration resulted in a lowering of the Mg²⁺ optimum to about 6 mM (data not shown). When translation factors were omitted from the standard reaction mixture (Fig. 2A), addition of spermine or spermine analogues at 6 mM Mg²⁺ caused substantial stimulation of AcPhe-tRNA binding. The stimulatory effect was moderated when the reaction mixture included the FWR fraction (Fig. 2B).

View larger version (14K): [in this window] [in a new window]   FIG. 2.   Effect of spermine and spermine analogues on poly(U)-directed Ac[3H]Phe-tRNA binding to ribosomes. The binding mixture (25 µl) was prepared in the absence (A) or presence (B) of the FWR fraction (10 µg of protein) and contained 100 mM Tris-HCl (pH 7.2), 100 mM NH4+, 6 mM Mg2+ (acetate), 25.6 A260 units of washed ribosomes per ml, 320 µg of poly(U) per ml, and, finally, spermine (), N4,N9-dibenzylspermine (), N1-ethylspermine (), N4,N9-bis(ethyl)spermine (), or N1,N12-bis(ethyl)spermine () at the indicated concentrations. The time course of the reaction was monitored up to 30 min at 25°C. The values of bound Ac[3H]Phe-tRNA were estimated from the maximum level of binding curves.

(iii) Effect on peptide bond formation. The effects of spermine and spermine analogues on peptide bond formation were compared in an E. coli cell-free system by using the puromycin reaction as a model reaction (38).

View larger version (15K): [in this window] [in a new window]   FIG. 3.   Variation of kmax/k3 as a function of spermine-analogue concentration. Complex C, formed in the presence of the FWR fraction at spermine or spermine analogue concentrations such as those used during the puromycin reaction, was adsorbed on a cellulose nitrate filter, washed, and then reacted with the appropriate concentration of puromycin in reaction buffer containing 6 mM Mg2+ and spermine (), N4,N9-dibenzylspermine (), N4,N9-bis(ethyl)spermine (), N1-ethylspermine (), or N1,N12-bis(ethyl)spermine () as indicated. The kmax value was calculated by fitting the experimental data to equation 3 by nonlinear regression. Similarly, the Ks and k3 values for the control experiment (no amine added) were estimated by using equation 2 and were found to be 660 ± 58 µM and 2.10 ± 0.22 min1, respectively.

Polyamine uptake. According to recent observations in our laboratory (17), spermine transport in E. coli cells obeys simple Michaelis-Menten kinetics (Fig. 4, lower line) with V_max and K_t values of 83.3 ± 2.4 nmol · min¹ per g of protein and 25.16 ± 1.58 µM, respectively. Competitive inhibition of spermine uptake was observed when E. coli cells were treated with ethyl or benzyl analogues of spermine (Fig. 4). From the K_i values presented in Table 3, it is obvious that the carrier affinity is highest for N⁴,N⁹-dibenzylspermine, then lower for N¹-ethylspermine and N⁴,N⁹-bis(ethyl)-spermine, and lowest for N¹,N¹²-bis(ethyl)spermine.

View larger version (18K): [in this window] [in a new window]   FIG. 4.   Double-reciprocal plots of [14C]spermine uptake by E. coli B cells, in the presence or absence of various alkyl and benzyl analogues of spermine. E. coli B cells grown in supplemented M9 medium were washed and suspended in buffer A to yield a protein concentration equal to 0.1 mg/ml. The [14C]spermine uptake was determined () in the absence of analogues or in the presence of 100 µM N1,N12-bis(ethyl)spermine (), N4,N9-bis(ethyl)spermine (), N1-ethylspermine (), and N4,N9-dibenzylspermine (). Each point represents the mean value of four individual measurements.

View larger version (20K): [in this window] [in a new window]   FIG. 5.   Double-reciprocal plots of [14C]putrescine uptake by E. coli B cells, in the presence or absence of various ethyl and benzyl analogues of spermine. The incubation mixture of cells was prepared as described in the legend to Fig. 4. The [14C]putrescine uptake was determined in the absence of analogues () or in the presence of 100 µM N1,N12-bis(ethyl)spermine (), N4,N9-bis(ethyl)spermine (), N1-ethylspermine (), and N4,N9-dibenzylspermine (). Inset, detail of the kinetic plots at high concentrations of [14C]putrescine.

Effect of spermine analogues on E. coli cell growth and polyamine pools. The effects of spermine analogues on the growth of cultured E. coli cells were screened at 100 µM. Surprisingly, none of the analogues examined showed any measurable effect, either stimulatory or inhibitory, on the growth of this strain. At first sight, these results seem to be in contradiction to the effects of analogues on the polyamine uptake system and peptidyltransferase activity. It is known from several reports (23, 29, 30, 35) that some of the analogues examined exert their action after internalization into the cells. In addition, inhibited uptake of [¹⁴C]putrescine or [¹⁴C]spermidine is usually taken as a criterion for the ability of the inhibitor to use the polyamine transporters. However, it could not be discounted that one or the other of the spermine analogues mentioned may block the transporters without being efficiently internalized. The ability of E. coli cells to accumulate spermine analogues was thus determined. The results revealed that none of the analogues is able to penetrate the plasma membrane. Only trace amounts of N⁴,N⁹-dibenzylspermine could be recovered in E. coli cells after a 24-h incubation with 100 µM, despite the high potency of this analogue as an uptake antagonist. Also, none of the analogues altered the intracellular concentrations of polyamines or the putrescine/spermidine molar ratio. The parent compound, spermine, exhibited the same behavior, although it was accumulated in cells at a concentration of 22 µmol/g of protein.

	DISCUSSION

Top Abstract Introduction Materials and Methods Results Discussion References

The use of alkyl and arylic derivatives of spermine for the chemotherapy of cancer and several parasitic diseases, including trypanosomiasis and leishmaniasis, has been greatly increased in the past few years (23). N,N'-Bis(ethyl) analogues of spermine have been of particular interest, and some members of this family are now in clinical trials. Given that the fate of certain pathogenic bacteria in animal infection is important in both fundamental and applied microbiology, we decided to explore the ability of a series of spermine analogues to affect the growth of cultured E. coli cells, to inhibit putrescine and spermine transport into cells, and to modulate several ribosomal functions.

We have previously demonstrated that the effects of spermine on protein synthesis are exerted at both the stages of initiation and elongation and that the charge distribution, as well as the chain flexibility of the ligand, plays a crucial role in its mode of action (3, 16, 18). From the present results, it is evident that the spermine analogues examined not only promote the formation and stabilization of the initiator ribosomal ternary complex, but also have a sparing effect on the Mg²⁺ requirements; although weaker, these effects are reminiscent of the behavior of spermine (3, 16). We have previously observed (18) that spermine selectively acylated at each amino group shows a much higher Mg²⁺ optimum, with a plateau at values over 9 mM. It is noteworthy that the analogues' potential is inversely related to their K_i values, estimated by kinetic analysis (Table 1). This suggests that the ability of analogues to stimulate the binding may be closely related to their affinity for one or more constituents of complex C. The consequence of amino group ethylation to the analogue's inhibitory activity is quite evident by the K_i and  values shown in Table 1; in going from bis-ethylated compounds to spermine, the ability of each analogue to interact with complex C increases (decrease in K_i values), whereas the reactivity of the CSI complex is concomitantly reduced (decrease in  values). It may be possible that the ethyl substituents at the nitrogens of spermine compromise some electrostatic interactions by increasing the distance between the charged sites of this compound and group of anions fixed to complex C. Thus, one might expect N⁴,N⁹-dibenzylspermine to bind more poorly than N⁴,N⁹-bis(ethyl)spermine. However, the K_i value of N⁴,N⁹-dibenzylspermine was about twofold smaller than that obtained with the second analogue. A reason for such behavior may be the existence of a hydrophobic pocket at the binding site on complex C which helps ligand anchoring. Alternatively, a restricted freedom of rotation around the neighboring carbon and nitrogen atoms, due to the introduction of benzyl groups, may stabilize the positive charges on secondary nitrogens and allow the development of strong hydrogen bond links. The second explanation is consistent with the finding of Frydman et al. (5) that the secondary amines of spermine bind more strongly to tRNA than the more electropositive primary amino groups. Despite the comparable K_i values between N⁴,N⁹-dibenzylspermine and spermine, the analogue behaved as a weaker inhibitor, since it is its high  value that increases the reactivity of the corresponding CSI complex.

In experiments carried out in the presence of the FWR fraction, addition of spermine analogue had a dose-dependent biphasic effect on peptide bond formation (Fig. 3). These results suggest that there are two separate binding sites for analogues on complex C, an activation site and an inhibition site. Once again, the polyamine analogues mimic the functions of spermine (3). The potency of each ligand depends on several features. Thus, N¹,N¹² substitution affords a high affinity to the ligand for the inhibition site (low K'_i value), but dramatically lowers the stimulatory effect of ligand on peptidyltransferase activity (low  value). This is in accordance with observations in rabbit reticulocyte translation systems (10, 36). In contrast, derivatization of internal amines by ethyl or benzyl groups has only a little influence on the ability of ligand to stimulate translation. Thus, the degree of peptidyltransferase activation by N⁴,N⁹-dibenzylspermine is similar to that obtained by spermine (Table 2), although the optimal concentration of analogue ([I]_opt) is twofold greater than that of the parent compound. Taking into account that polyamines with terminal benzyl groups show diminished activity in protein-synthesizing eukaryotic systems (36), our results can be interpreted on the basis of restricted rotation effects described above.

Since all analogues tested behave as competitive inhibitors of polyamine uptake, full characterization of their potency can be made exclusively on the basis of the K_i value, independently of whether putrescine or spermine is used as a substrate. As shown in Table 3, the analogues exhibit similar K_i values for both the carrier of spermine and carrier 1 of putrescine, suggesting the existence, in fact, of a common transporter. Moreover, the K_t value of carrier 1 is in the vicinity of that reported for the putrescine transport system encoded by the potABCD operon (19). These findings, taken together, suggest that all spermine analogues examined compete for the PotD protein. Obviously, ethylation of the terminal amino groups decreases the affinity of ligand for the uptake system, whereas ethyl or benzyl substitution of N⁴,N⁹-imino groups can be greatly tolerated. All of the data presented above are consistent with the idea that in addition to the key role of cationic centers, specific hydrophobic interactions must also largely contribute to the affinity of analogues for polyamine transporter. Furthermore, none of the analogues tested is recognized by carrier 2 of putrescine, i.e., by the uptake system encoded by potFGHI. This agrees with a recent interpretation of PotF specificity given by Vassylyev et al. (40), according to which PotF recruits a bulky side chain that protrudes deeply into the binding cavity. In conclusion, the polyamine uptake results in E. coli further support the mimetic character of ethyl and benzyl derivatives of spermine in reference to the parent compound. Finally, it is important to point out that E. coli cells, in contrast to erythrocytes and other mammalian cells (2), do not appear to have a specific uptake system for bis(benzyl)polyamine analogues; both spermine and N⁴,N⁹-dibenzylspermine compete for the PotD protein (Table 3).

Surprisingly, none of the analogues added to the medium affected E. coli cell growth. Even prolonged culture in the presence of 100 µM spermine analogue did not show any deviation from the behavior of untreated cells. Also, we failed to detect any effect of the spermine analogues on intracellular polyamine pools. Moreover, no spermine analogue was detected by HPLC analysis in exposed cells, nor did any intermediate product result from stepwise desubstitution and subsequent oxidative degradation (data not shown). Relatively, it has been demonstrated that N¹,N¹²-dialkyl analogues of spermine cannot be used as substrates in spermidine/spermine-N¹-acetyltransferase (SSAT)/polyamine oxidase (PAO)-catalyzed oxidative cleavage (30, 31). In addition, SSAT and PAO are not as active in E. coli cells as they are in eukaryotic cells (7). Traces of N⁴,N⁹-dibenzylspermine found in E. coli cells after a 24-h incubation with this analogue can be attributed to secondary diffusion effects. In combination, these observations suggest that the spermine analogues examined, although endowed with high affinity for the E. coli polyamine transport system encoded by potABCD, are not efficiently internalized into the cells. Consequently, they should be characterized as pure competitive antagonists of polyamine uptake. Thereby, the failure of SSAT induction in E. coli cells by exogenous N¹,N¹²-bis(ethyl)spermine (7) seems to be reasonable: it is rather the lack of permeation of analogue through the E. coli membrane which explains the above effect, not the inability per se to induce the enzyme.

Distinct structure-activity relationships relevant to the spermine molecule have become apparent during this study. Our findings also reinforce the notion that high heterogeneity of polyamine transport system specificity exists among living cells. This heterogeneity may be a useful feature that can be exploited in investigating potential symbiotic relationships between prokaryotic and eukaryotic cells and in orienting chemical-synthetic work to the creation of drugs which can achieve selective uptake into the desired target cells.