Antibacterial Activity Evaluation of Microcin J25 against Diarrheagenic Escherichia coli

Abstract

The inhibitory activities of known microcins were evaluated against some diarrheagenic Escherichia coli strains. Some antibacterial properties of microcin J25, the most active one, were studied. A rapid two-step purification was performed. The MIC and the minimum bactericidal concentration of J25 against E. coli O157:H7 were 1 and 100 μg ml⁻¹, respectively. A 10⁴-CFU ml⁻¹ contamination by this strain was destroyed in milk and meat extract by 6.25 μg of J25 ml⁻¹ and in half-diluted egg yolk by 50 μg of J25 ml⁻¹.

Diarrheagenic Escherichia coli (DEC) strains are currently associated with food-related illnesses (13). The notorious DEC serotype O157:H7 has been the cause of several large food-related epidemics in Europe, North America, and Japan, mostly after consumption of meat or dairy products (9). Physicochemical (3, 4, 7) and biological (2, 8) methods for the control of DEC have been described, mainly with E. coli O157:H7 as the target strain.

Microcins and colicins are classical bacteriocins produced by Enterobacteriaceae that inhibit E. coli and closely related strains (12). Unlike most colicins, microcins are secreted peptides with low molecular masses (<10 kDa). Their synthesis is nonlethal for the producing strains and not mediated by conditions inducing the SOS system (12). Because of their small size, they are resistant to some proteases and fairly thermostable. So far, six microcins have been described, namely, B17 (19), C7 (6), D93 (10), E492 (18), H47 (5), and J25 (1); colicin V (Col V) can also be considered a microcin (12).

The objective of the present study was to evaluate the inhibitory activity of microcins, particularly J25, against DEC strains. Our results showed that all the tested strains were inhibited by at least two microcins. The antagonistic activity of purified microcin J25, the most active one under our experimental conditions, was evaluated. This microcin is a cyclic peptide of 21 unmodified amino acid residues (1) which apparently block cell division (15). The operon coding for production, export, and immunity was recently described (16, 17).

Bacterial strains, preservation, and growth conditions.

Microcin producers were recombinant E. coli strains (MC 4100 for B17, C7, and Col V; VCS 257 for E492; RYC 1000 for H47; and KI 3110 for J25), with each strain harboring the genes required for synthesis, export, and immunity for a single microcin. Producers of B17, C7, E492, J25, and Col V were supplied by F. Moreno (Unidad de Genetica Molecular, Hospital Ramon y Cajal, Madrid, Spain), and the producer of H47 was supplied by M. Laviña (Instituto de Investigaciones Biologicas Clemente Estable, Ministerio de Educacion y Cultura, Montevideo, Uruguay). DEC strains were six collection strains purchased from Institut Pasteur (Paris, France), named CIP 52.168, CIP 52.170, CIP 52.172, CIP 62.23, CIP 62.24, and CIP 103.571 (respective serotypes: O111:H12, O55:H6, O26:H11, O119, O125, and O157:H7) and nine clinical isolates obtained from different patients with acute diarrhea, which were serotypes O26, O55 (three strains), O86, and O111 and three undetermined serotypes (X₁, X₂, and X₃).

Brain heart infusion (BHI), Mueller-Hinton broth, nutrient broth (NB), and bacteriological products were purchased from Biokar Diagnostics (Beauvais, France). M63 minimal medium was prepared as described by Miller (11) and supplemented with glucose (0.2%), thiamine (0.01%), and Casamino Acids (0.1%). The media for plating were solidified with 12 or 6 g of E-type agar (soft agar) liter⁻¹. Unless otherwise stated, bacterial cultures were propagated in BHI at 37°C. Working cultures were maintained at 4°C on BHI agar slants. Stock cultures were stored at −80°C.

Inhibition assays.

M63 agar (10 ml) was overlaid with 5 ml of M63 soft agar seeded with 10⁷ CFU of the target strain ml⁻¹. Sterile glass rings (4-mm inside diameter [i.d.]) were placed on the soft agar and filled with 20 μl of filter-sterilized samples to be tested. The plates were incubated for 24 h at 37°C and clear inhibition zones were measured. Inhibition assays were done in triplicate. The results of the inhibition of DEC strains by culture supernatants of microcin producer strains are given in Table 1. No strain was resistant to all the microcins, but the strains showed various sensitivities to the microcins.

TABLE 1.

Inhibition of DEC strains by microcins

Microcina	Inhibitionb of:
	E. coli collection strains						E. coli clinical isolates
	O111:H12	O55:H6	O26:H11	O119	O125	O157:H7	O26	O551	O552	O553	O86	O111	X1	X2	X3
B17	−	−	+	+	+	−	−	−	−	+	+	+	−	−	+
C7	−	−	−	−	+	+	+	+	++	++	+	++	++	+	+
E492	++	++	++	+	−	+	+	−	−	−	−	+	+	++	+
H47	−	++	−	+	++	+	−	−	−	−	+	++	+	+	++
J25	++	−	++	−	−	++	++	++	++	++	+	++	++	++	++
Col V	−	−	−	−	−	+	−	−	−	−	−	+	−	−	−

^aTwenty microliters of a culture of the producer strain in M63 medium (18 h, 37°C). 

^bInhibition was evaluated by measuring the average diameter (d_a) of clear zones. −, d_a < 5 mm, no inhibition; +, 5 mm < d_a < 10 mm; ++, d_a > 10 mm. 

It also appeared that each microcin had a different spectrum of activity. Under our experimental conditions, Col V was the most ineffective product. On the other hand, J25, which had inhibitory activity against 12 of the 15 DEC strains (including all the clinical isolates), was apparently the most active one. Thus, we purified it and carried out further studies of some of its growth-inhibiting properties.

Purification of microcin J25 and analysis of its antagonistic properties against DEC.

A culture (18 h, 37°C) of the J25 producer in M63 medium was centrifuged (13,000 × g, 30 min, 10°C) and heated (10 min, 120°C). Crude extract was prepared from heated supernatant by ammonium sulfate precipitation (95% saturation). The pellet was resuspended in water (1/80 of the initial culture volume) and dialyzed (with a membrane cutoff of 1,000) against water (24 h, 10°C). The crude microcin was then purified by reversed-phase high-pressure liquid chromatography (RP-HPLC). Filtered samples (2 ml) were subjected to semipreparative RP-HPLC (Delta Pak C₁₈ column, 300 by 19 mm [i.d.]). The mobile phase (12 ml min⁻¹) was 10 mM ammonium acetate buffer (pH 6.0) (eluent A) and acetonitrile (eluent B). The gradient was 0 to 20% eluent B in 4 min, 20 to 56% eluent B in 36 min, and 56 to 100% eluent B in 9 min. Optical density was measured at 215 nm. The peak fraction of J25 (retention time, 24 min) was collected and freeze-dried. Purified J25 (J25_p) was subjected to an additional RP-HPLC analysis (Delta Pak C₁₈ column, 300 by 3.9 mm [i.d.]) to check its purity. By peak area integration, J25_p had a purity of 96.7%. All the J25_p fractions obtained by semipreparative RP-HPLC were pooled, freeze-dried, and stored at 4°C. Before utilization, J25_p was dissolved in a water-acetonitrile mixture (60:40, vol/vol) in order to obtain an initial solution with a concentration of 4 mg ml⁻¹.

The MIC was determined by the standard broth macrodilution method as described by Sahm and Washington (14). We checked that the highest concentration of acetonitrile in Mueller-Hinton broth was not affecting the growth of tested bacteria. The minimum bactericidal concentration (MBC) was determined from tubes showing complete inhibition. An NB agar plate was seeded on the surface with 0.1 ml from each clear tube and incubated (24 h, 37°C). The MBC was defined as the lowest concentration in the tubes giving no growth on an NB plate afterwards.

The results of the inhibition assays for DEC strains with J25_p are reported in Table 2. This microcin inhibited all the DEC strains tested with varying effectiveness. Most of the strains (11 of 15) had high sensitivities to J25_p (inhibition zones were >20 mm with 2 μg of drug), clinical isolate O86 showed medium sensitivity to J25_p, and 3 strains displayed low sensitivities to J25_p (inhibition zones were obtained only with 20 μg of drug). Microcin J25_p inhibited the growth of E. coli O157:H7 in a dose-dependent manner with 1 to 1,000 μg ml⁻¹ (results not shown). With the same target strain, the MIC and the MBC of J25_p were 1 and 100 μg ml⁻¹, respectively.

TABLE 2.

Inhibition of DEC strains by purified J25

DEC strain	Inhibitory activitya of:
	2 μg	20 μg
Collection strains
O111:H12	21.0	25.4
O55:H6	—b	5
O26:H11	22.1	25.9
O119	—	5
O125	—	8
O157:H7	22.3	27.0
Clinical isolates
O26	22.0	26.4
O551	21.5	24.2
O552	21.2	24.8
O553	20.6	24.3
O86	14.1	16.1
O111	22.0	26.2
X1	22.8	26.4
X2	21.1	25.0
X2	22.2	26.5

^aInhibitory activity of J25_p (diluted in 20 μl of sterile saline water) is expressed as the average width of clear zones, in millimeters. 

^b—, no inhibition zone. 

Ability of microcin J25 to inhibit E. coli O157:H7 in biological products.

The inhibition activity of J25_p against E. coli O157:H7 was tested in three products: sterile skim milk, diluted egg yolk (1:1 mixture of sterile egg yolk and sterile water), and meat extract (50 g of sterile water was added to 100 g of mincemeat and mixed for 5 min in a stomacher at maximum speed; the supernatant was harvested by centrifugation and then filter sterilized before use). Product samples of 500 μl mixed with 20 μl of an appropriate dilution of the initial J25_p solution in sterile saline water (8.5 g liter⁻¹) were inoculated with 5 μl of an E. coli O157:H7 culture in M63 medium (12 h, 37°C) diluted with sterile M63 medium in order to obtain the desired quantity of cells. After 24 h of incubation (37°C), the presence of viable cells in 100-μl samples was checked using a spiral plating method on BHI agar with incubation (37°C, 24 h). Assays were done three times.

When the three products contained 6.25 μg of J25_p ml⁻¹ and an E. coli O157:H7 concentration of 10³ CFU ml⁻¹, no viable bacteria were detected after 24 h (0 CFU in three 100-μl product samples). With an O157:H7 concentration of 10⁴ CFU ml⁻¹, we obtained the same results (i.e., no bacteria were detected) in milk and meat extract, but complete inhibition required 50 μg of J25_p ml⁻¹ in egg yolk, due to the richness of the medium.

The data presented here show the possibility of using microcins to control DEC strains, including serotype O157:H7.