Two acetyl analogues of spermidine and five analogues of spermine were used to determine the structural specificity of the polyamine transport system in Escherichia coli by measuring their ability to compete with [14C]putrescine or [14C]spermine for uptake, as well as to inhibit cell growth, and, finally, to affect the intracellular polyamine pools. Spermine uptake follows simple Michaelis-Menten kinetics (Kt = 24.58 +/- 2.24 microM). In contrast, the putrescine uptake system involves two saturable Michaelis-Menten carriers exhibiting different affinity towards putrescine (Kt = 3.63 +/- 0.43 microM, Kt' = 0.61 +/- 0.10 microM). From the Ki values, it is inferred that N1-5-amino-2-nitrobenzoylspermine is the most effective competitive inhibitor followed by N1-acetylspermine, and then N1,N12-diacetylspermine. N1-acetylspermidine and N8-acetylspermidine also inhibit competitively the uptake of spermine, the latter being the most effective inhibitor. In addition, the above-mentioned analogues inhibit identically one of the carriers of putrescine uptake, suggesting the existence of a common transporter for both putrescine and spermine. The order of analogue potency, regarding the other carrier of putrescine is as follows: N1,N12-diacetylspermine approximately N1-5-amino-2-nitro-benzoylspermine > N1-acetylspermine. Both N1-acetylspermidine (Ki = 753 +/- 25 microM, Ki' = 128 +/- 5 microM) and N8-acetylspermidine (Ki = 22.4 +/- 0.4 microM, Ki' = 279 +/- 3 microM) also cause competitive inhibition of putrescine uptake, however with inverse affinity towards the putrescine carriers. Neither N4,N9-diacetylspermine, nor N1,N4-bis(beta-alanyl)diaminobutane affect the uptake of any polyamine. Interestingly, none of the acetyl analogues of spermine has a measurable effect on cell growth and cellular polyamine pools, although some of them are accumulated in cells. Based on these findings, the relative significance of the primary and secondary amines and of the chain flexibility as determinants of cellular uptake are discussed.