DNA extract preparation and phylogenetic group identification

The studied E. coli were 34 antibiotic-resistant strains isolated from cases of subclinical bovine mastitis in western Michoacán state, Mexico²³. Total DNA extracts were prepared from these isolates²⁴. An established protocol was used to identify the phylogenetic groups of each strain¹⁰. The primers used for fragment amplification were:

ChuA.1 (5'-GACGAACCAACGGTCAGGAT-3');

ChuA.2 (5'-TGCCGCCAGTACCAAAGACA-3');

YjaA.1 (5'-TGAAGTGTCAGGAGACGCTG-3');

YjaA.2 (5'-ATGGAGAATGCGTTCCTCAAC-3');

TspE4C2.1 (5'-GAGTAATGTCGGGGCATTCA-3');

TspE4C2.2 (5'-CGCGCCAACAAAGTATTACG-3').

The PCR reactions were run using a 25 µl volume containing: 1X PCR Master mix (Promega, Madison, WI, USA), 1 µM of each primer and 2 µl bacterial DNA extract (50 ng). Fragment amplification was done in a C1000 thermocycler (Bio-Rad, Mexico City, Mexico) under the following conditions: initial denaturation at 94 °C for 5 min; 30 cycles as follows, denaturation at 94 °C for 30 s, alignment at 55 °C for 30 s and extension at 72 °C for 30 s; and final extension at 72 °C for 5 min. The amplified products were separated on 1.5% agarose gels and viewed with a Universal Hood II Gel Doc (Bio-Rad).

Biofilm production

Biofilm production was grown according to an established protocol²⁵ with some modifications. Bacteria were seeded in Petri dishes with trypticase soy agar (TSA, BD, Mexico City, Mexico). These cultures were used to inoculate tubes with 2 ml trypticase soy broth (TSB) containing 0.25% glucose, which were incubated overnight at 37 °C under constant agitation. In each tube the cultures were diluted to the 0.5 MacFarland standard with TSB containing 0.25% glucose. From these bacterial dilutions 200 μl were deposited in triplicate in 96-well ELISA plates and incubated for 24 h at 37 °C without stirring. Non-inoculated culture medium was used as a negative control, and the biofilm-forming bacterium Pseudomonas aeruginosa ATCC 27853 was used as a positive control. After incubation the culture medium was discarded, the wells washed three times with sterile saline, and the plate allowed to dry at 60 °C for 1 h to fix the cells. One percent (1%) violet crystal (200 µl) was added to the wells and the plate left at room temperature for 20 min. It was washed with running water until no color evolution was observed and allowed to dry at room temperature. The violet crystal was solubilized by adding 200 µl 96% ethanol to each well and stirring, and transferring the supernatant to a microcentrifuge tube. This process was repeated, tube content completed to 1 ml and optical density (OD) of the suspension read at 570 nm with a SmartSpec Plus spectrophotometer (Bio-Rad). This test was repeated four times in triplicate and average OD calculated from the results. The optical density cutoff (ODc) was calculated using the average OD of the negative control and increasing the value by three times its standard deviation, producing a value of 0.22. The biofilm-forming bacteria were classified into four groups: strong= OD> 4xODc; moderate= 2xODc<OD≤4xODc; weak= ODc<OD≤2xODc; and negative= OD≤ODc²⁵.

Detection of genes associated to biofilm formation

The presence of two genes associated with biofilm formation, fimA (main subunit of type I fimbria) and csgA (main subunit of curli fimbria), was analyzed by PCR. A 119 bp fimA fragment was amplified with fimA-F (5'-CTCTGGCAATCGTTGTTCTGTCG-3') and fimA-R (5'-GCAAGCGGCGTTAACAACTTCC-3'), and a 178 bp csgA fragment was amplified with csgA-F (5'-GATCTGACCCAACGTGGCTTCG-3') and csgA-R (5'-GATGAGCGGTCGCGTTGTTACC-3')²⁶. The reactions were run in a 25 µl volume containing 1X PCR Master mix (Promega), 1 µM of each primer, and 2 µl total DNA extract. Amplification conditions were: initial denaturation at 95 °C for 5 min; 30 cycles of 94 °C for 30 s, 55 °C for 30 s and 72 °C for 30 s; and a final extension cycle at 72 °C for 10 min. The amplified products were separated on 2% agarose gels and viewed as described above.

Determination of swarming type motility

Bacterial motility was measured following an established protocol¹⁵. Briefly, swarming motility was determined by allowing the cultures to grow overnight in Lysogeny Broth (LB), and adding 5 µl swarming agar (LB + 0.5% agar) containing 0.5% glucose. After incubating the cultures for 12 h at 37 °C, bacteria displacement diameter was measured. Three measurements were taken and averaged.

Rep-PCR of bovine mastitis-causing E. coli

Genetic diversity was analyzed by purifying genomic DNA according to standard protocols²⁷. The purified DNA was amplified in 25 µl reaction mixture containing 12.5 µl 2X Master mix (Promega), 3 mM MgCl₂, 5% DMSO, 0.16 µg/µl bovine serum albumin, 100 ng DNA and 2 µM of the primer (GTG)₅ (5'-GTGGTGGTGGTGGTG-3'). Amplification was done in a C-1000 thermocycler (Bio-Rad) under the following sequence: initial denaturation 95 °C for 2 min; 30 cycles as follows, denaturation (94 °C/3 s and 92 °C/30 s), alignment (40 °C/1 min) and extension (65 °C/8 min); and final extension at 65 °C for 8 min²⁸. The amplified products were separated on 1.5% agarose gels and viewed with a Universal Hood II Gel Doc (Bio-Rad).

Band patterns were analyzed with the GelJ ver. 2 software²⁹. Band size was standardized using a 1 kb DNA ladder (Promega) with fragments between 250 and 10,000 bp as a reference standard. The similarity coefficients were generated by the Dice method with a 2 % tolerance between rows, and the dendrogram built using the UPGMA method.

Virulence in Galleria mellonella

In vivo virulence was measured for representative bacteria exhibiting the highest biofilm formation rates, as well as non-biofilm-forming isolates. The G. mellonella larvae weighed 150 to 200 mg, were free of apparent damage, exhibited no melaninization (necrotic spots) (Petmmal, Cuautitlán Izcalli, Mexico), and were stored at 30 °C in darkness until use. The infection assays were done using bacterial cultures grown in LB broth until the mean exponential growth phase (OD 600 nm= 0.5). These were centrifuged at 12,000 rpm/min, washed twice with 1 ml 10 mM MgSO₄, and resuspended in 1 ml 10 mM MgSO₄. Serial dilutions were made of this cell suspension until attaining a bacterial concentration of approximately 1x10⁵ CFU/ml, which was confirmed by CFU count on LB agar.

Ten G. mellonella larvae were used for each bacterium to be studied. Using an insulin syringe with a 29G gauge needle, 10 μl bacterial suspension (1 x 10⁵ CFU/ml) were injected into the final pro-leg of each larva. The negative control was ten larvae injected with 10 mM of bacteria-free MgSO₄, and an additional ten without inoculation. The inoculated and non-inoculated larvae were placed in sterile Petri dishes and incubated at 30 °C in darkness for 96 h. Survival percentages were measured at 24, 48, 72 and 96 h; larvae unresponsive to touch were recorded as dead²¹. The virulence assay for each bacterium was run at least twice independently.

Phylogenetic groups and genetic relationship of bovine mastitis- associated E. coli

Amplification of the chuA (279 bp), yjaA (211 bp) and TspE4.C2 (152 bp) fragments (Figure 1) showed the 34 isolated BMAEC strains to belong to three phylogenetic groups. Most (67.7 %; 23 isolates) belonged to group A, followed by 17.6 % (6 isolates) in group B1 and the remaining 14.7 % (5 isolates) in group D (Figure 1B).

Figure 1 Agarose gels showing PCR amplification results for A) chuA (279 bp), B) yjaA (211 bp) and C) TspE.C2 (152 bp) 

The genetic relationship analysis by rep-PCR using the (GTG)₅ oligonucleotide yielded 32 different patterns between 7 and 21 bands, with sizes ranging from 250 to 5,000 bp (Figure 2). The dendrogram based on the Dice similarity coefficient and generated by the UPGMA method indicated that the lowest similarity among the isolates was 58 % while the highest was 100 % (in 2 isolates) (Figure 3). Considering a 70 % genetic similarity level (dotted line), seven groups of bacteria were identified; the largest was group VII with 22 isolates (64.7 %), followed by group III with 4 isolates (11.8 %), group VI with 3 isolates (8.8 %), group I with 2 isolates (5.8 %), and groups II, IV and V with a single isolate each (2.9 %).

Figure 2 Agarose gel of products amplified by (GTG)₅-PCR. The base pair size of some bands is indicated by the 1 kb marker (M) 

Figure 3 Grouping band patterns for 34 bovine mastitis-associated E. coli strains, based on the Dice similarity coefficient and generated by UPGMA 

Motility, biofilm formation and associated genes

Ten of the 34 E. coli isolates were not mobile (-) and the remaining 24 were mobile. Based on their displacement diameters, twelve of these were minimally mobile (+), six were moderately mobile (++), and six were highly mobile (+++) (Figure 4, Table 1).

Figure 4 Representative examples of “swarming” type motility. A) Not mobile, B) Minimally mobile, C) Moderately mobile, D) Highly mobile 

Twenty-six (76.5 %) of the isolates formed biofilm to some degree (OD>0.22) while the remaining eight (23.5 %) did not form biofilm (OD≤0.22). Of the biofilm-forming isolates seven were classified as having a strong (OD>0.88) and moderate formation capacity (OD 0.45-0.88), and twelve as having a weak one (OD 0.23-0.44) (Table 1). Twenty-two (64.7 %) of the 34 isolates had either csgA or fimA, whereas of the 26 biofilm-forming isolates 15 had one or both of these genes. Both csgA and fimA were present in the seven non-biofilm-forming isolates.

E. coli Virulence in G. mellonella

Differences in the pathogenesis of the BMAEC isolates were analyzed by infection trials using G. mellonella larvae. Seven representative isolates were tested: five strong biofilm-forming isolates and two non-forming isolates. The biofilm-forming isolates killed 100 % of the larvae 24 h after infection (Figure 5). The two non-biofilm-forming isolates caused from 0 to 20 % mortality at 24 h, and from 10 to 50 % at 96 h. The control resulted in no mortality during the 96-h experiment.

Figure 5 Virulence in G. mellonella larvae of a control (C) and seven representative E. coli isolates: five biofilm-forming isolates (MC2, 6, 13, 19 and 35), and two non-biofilm forming isolates (MC41 and MC80)