DETERMINATION OF THE PHYLOGENETIC RELATEDNESS OF CRONOBACTER SPP. ISOLATED FROM POWDERED INFANT FORMULA RETAILED IN NIGERIA USING PAN–GENOMIC DNA MICROARRAY OF THE PHYLOGENETIC RELATEDNESS OF CRONOBACTER SPP. ISOLATED FROM POWDERED INFANT FORMULA RETAILED IN NIGERIA USING PAN–GENOMIC DNA MICROARRAY.”

Cronobacter spp. are emerging, opportunistic, food-borne pathogens associated with infections like meningitis, necrotizing enterocolitis and septicaemia in premature and immunocompromised neonates and infants. The phylogenetic relatedness of three Cronobacter species isolated from powdered infant formula retailed in Nigeria was carried out using a Pan-Genomic DNA Microarray constituting 19,287 independent genes representing 15 Cronobacter genomes and 18 plasmids and 2,371 virulence genes of phylogenetically related Gram-negative bacteria. The hybridization results showed that Cronobacter malonaticus (CS14) and Cronobacter sakazakii (CS17 and CS124) clustered with powdered infant formula environmental and clinical strains of C. malonaticus and C. sakazakii isolated from countries like Jordan, Czech Republic, Ireland and USA with a significant relatedness greater than 80%. The sequence types of C. malonaticus CS14 was ST303 and C. sakakakii CS17 and CS124 were ST304 and ST296, respectively. Some virulence genes (integrase of Shigella flexnerri bacteriophage X, hypothetical protein z1655, dihydrofolate reductase, and formate acetyltransferase 1) were detected in CS124 and CS17. Adequate regulatory measures should be applied to monitor imported and locally produced powdered infant formulae to prevent contamination with Cronobacter spp. and other food borne pathogens to ensure the safety of vulnerable neonates and infants.


Microarray Hybridization
The microarray hybridization of DNA from the Cronobacter spp. was carried out using the Affymetrix MyGeneChip Custom Array (Affymetrix design number: FDACRONOa520845F). Hybridizations were performed according to the Affymetrix GeneChip Expression Analysis Technical Manual for the 49-format array (Bolstad et al., 2003). Five μg of the genomic DNA was fragmented by incubating at 37°C for 1 min in a 20 μl of reaction containing 1× One-Phor-All Plus Buffer (GE Healthcare) and 0.01 U DNase I (GE Healthcare). The fragmentation was heatinactivated at 99°C for 15 min. The fragmented DNA was 3'-end labelled by adding 4 μl of 5× terminal transferase buffer (Promega), 1 μl of 1 mM biotin-11-ddATP (PerkinElmer NEL508), and 2 μl (60 u) of terminal transferase enzyme (Promega). Labelling was carried out for 4 h at 37°C followed by heat inactivation at 98°C for 1 min. Hybridizations were performed according to the Affymetrix GeneChip Expression Analysis Technical Manual for the 49-format array (Affymetrix, 2014). Briefly, 146 µl of a hybridization buffer comprised of 100 µl of 2X hybridization buffer, 3.3 µl of a 3nM B2 oligonucleotide solution, 2 µl each of a 10 mg /ml Salmon DNA and 50 mg/ml Bovine Serum Albumin (BSA) solutions, and 15.5 µl of Dimethyl sulfoxide (DMSO) (SIGMA-ALDRICH, Inc. St. Louis, MO) per reaction followed by denaturation at 98°C for 1 minute. The denatured samples were added onto the Affymetrix arrays, which were then incubated at 45°C, with rotation (60 rpm) for 16 h in a hybridization oven. Following hybridization, wash and stain procedures were carried out on an Affymetrix FS-450 fluidics station using the mini_prok2v1_450 fluidics. Reagents for washing and staining were prepared according to the GeneChip® Expression Analysis Technical Manual (Affymetrix, 2014). The following exceptions were made to the wash and stain procedure: Streptavidin solution mix (vial 1) was replaced with SAPE solution mix (LIFE TECHNOLOGIES, Grand Island, NY). Arrays were scanned using Affymetrix GeneChip® Scanner 3000 running AGCC software (Tall et al., 2015).

Microarray data analysis
Probe set intensities for each gene represented on the microarray were summarized using the Robust MultiArray Averaging (RMA) function in the Affymetrix package of R-Bioconductor as described by Bolstad et al. (2003). The RMA summarization of probe level data was done by carrying out three individual treatments on all of the experimental data (CEL file). The probe specific correction of the perfect match (PM) probes was done using a model based on the observed intensities being the sum of signal and noise. Second, quantile normalization was performed on the corrected PM probe intensities. Finally, a median polishing algorithm was used to summarize the background-corrected, normalized probe intensities to generate a final probe set value.

Calculating gene differences and generating dendrograms
Robust MultiArray Averaging-summarized probe set intensities were compared across all strains for each gene. If the same gene in different strains had an RMA intensity difference greater than eightfold (log2 = 3), then that gene was considered to be "different." With this criterion, a strain versus strain gene-difference matrix was generated; where the difference matrix represents the number of genes/alleles that differs between any two isolates. Gene-difference matrices were converted to dendrograms using the hclust function in the base package as well as the phylo function in the ape package of R-Bioconductor. Hierarchical clustering was performed using the RMA-summarized probe set intensities using the MADE4 package of R-Bioconductor. Phylogenetic trees were made using the nearest neighbour-joining method via the MEGA 5

Results
The strains analysed in this study were from different sources and different countries as shown on Table 1.   The microarray analysis was able to correctly identify the Cronobacter strains to each species epiphet. For example, C. malonaticus (CS 14) was identified as C. malonaticus and clustered with environmental and clinical strains of C. malonaticus which were isolated from different countries like Jordan, Czech Republic, and USA (Figure 1), with a significant relatedness of more than 80% as shown by Pearson's correlation coefficient analysis (Table 2). Accordingly, microarray analysis of the C. sakazakii strains CS 17 (ST304) and CS 124 (ST296) correctly identified these strains as C. sakazakii and phylogenetically placed them within the C. sakazakii species cluster alongside a C. sakazakii clinical strain CDC 1121-73 ( Figure 1). These results also suggest that strains possessing ST304 and ST296 share phylogeny with ST64 strains and may represent a new clonal complex. The larger C. sakazakii cluster contained isolates from different countries like Ireland and USA with a significant relatedness of more than 80% as shown in Table 2. The relatedness of CS 14, CS 17, CS 124 and other closely related Gram-negative bacteria ranged from 33 -47% as shown in Table 2.  strains CS 17 and CS 124 had 369 and 386 genes respectively that were different from clinical strain CDC1121-73A_2 that was isolated from bronchial wash (Table 3).      Table 6 shows the phage related genes acquired by CS 17 from C. sakazakii 2151.

Discussion
Infants are the most vulnerable group of the human population and so attempt to protect them from health hazards should be done with utmost priority. PIF is not a sterile product but it should be free from all potential pathogens because neonates and infants possess under-developed immune systems and lack a competing intestinal flora (Townsend et al., 2008). Because infant formula products are primarily imported into the Nigerian market, the relatedness of the isolated Nigerian strains was examined in relation to other strains from other countries. Out of the 154 samples of PIF analysed in this study, Cronobacter species was isolated from 2 per cent (3). Cronobacter sakazakii was isolated from 1.30% while C. malonaticus was isolated from 0.65% of samples.
Both of these Cronobacter spp. have been isolated from PIF samples from different countries around the world (Farmer, 2015). Gicova  sakazakii which were isolated from the environment of a group of European PIF manufacturing facilities. The results of that study showed that the microarray could separate 25 C. sakazakii ST4 strains into two distinct subclades which suggested that there may be two evolutionary lineages associated with ST4 strains. The microarray analysis also showed that these two lineages differed in a total of 95 unique genes, of which many were phage-related genes (seven related genes) and 17 of these unique genes were associated with the pESA3-encoded type six secretion system (T6SS) gene cluster as described by Franco et al. (2011a).
The three (3) isolates in this study were compared directly with nearest neighbors and other Cronobacter species. The results showed that CS 17 and CS 124 had significant relatedness (> 80%) to C. sakazakii strains isolated from blood, CSF and breast abscess from different countries such as USA and Ireland. The CS 14 also had greater than 80% relatedness to the clinical isolates of C. malonaticus from countries like Czech Republic, USA and Jordan. The number of gene differences is based on strain-to-strain comparisons and gene difference is defined as an eightfold difference in the RMA-summarized probe set intensities for each gene (Tall et al., 2015). Speculatively, the difference in the Cronobacter strains isolated in this study may be due to bacterial adaptation to Nigerian tropical environment or the acquisition of genes from indigenous bacteria. This study also illustrates the global nature and spread of Cronobacter spp., in infant formula products which may be produced in one part of the world and consumed in another part.
Microarray analysis of C. malonaticus strain CS 14 and C. sakazakii strain CS 17 showed that these strains had acquired some phage related genes which were found in Cronobacter dublinensis while C. sakazakii strain CS 17 had acquired some phage related genes which were found in C. sakazakii 2155. Whole-genome analyses have revealed that many bacterial genomes contain foreign genes, especially phage genes (Ochman et al., 2000). The phage genes in bacterial genomes include genes for virulence or fitness factors such as extracellular toxins, super antigens, lipopolysaccharide-modifying enzymes, and proteins conferring serum resistance, etc. Pathogenicity islands which contain one or more virulence genes, are present in the genomes of pathogenic bacteria but are absent from the non-pathogenic variant of the same species and often exist in the size range of 10-200 kb (Schmidt and Hensel, 2004). They contain clusters of functionally related genes necessary for virulence in bacteria. Salmonella spp. contains a wide variety of mobile genetic elements from pathogenicity islands to conjugative transposons, (Kelly et al., 2009). One of these pathogenicity island gene, formate acetyl transferase 1, was found in CS 17 (C. sakazakii). A pathogenicity island gene from E. coli 0157:H7, hypothetical protein z1655, was also found in CS 124 (C. sakazakii). The presence of these pathogenicity islands genes in CS 17 and CS 124 could enhance their virulence. The presence of reputable virulence genes in the C. sakazakii isolates indicates the potential risk of consumption of these Cronobacter contaminated powdered infant formula (PIF) by neonates and infants; hence the need for intensive and continuous monitoring of potential pathogens in powdered infant milk formula to ensure the safety of vulnerable infants.
Antibiotic resistance gene (dihydrofolate reductase) from E. coli was also detected in C. sakazakii strains CS 17 and CS 124. Resistance to clinically relevant, front-line antimicrobials such as fluoroquinolones, extended-spectrum β-lactams (including extended-spectrum cephalosporins) has been reported among E. coli strains and they are believed to be an important reservoir of transferable antimicrobial resistance genes (Singh et al., 2005). The transfer of this antibiotic resistance to indigenous non-resistant bacteria could contribute to an increase in the rate of resistance of bacteria to drugs especially in the Nigerian environment where there is no regulation on the use of antibiotics.

Conclusion
Infant formula producers must enforce the use of guidelines aimed at decreasing the risks of product contamination with foodborne pathogens. The control of primary populations of Cronobacter spp. during the PIF production process and prevention of post processing contamination can be ensured by using suitable microbiological guidelines for quality control and assurance. Sanitary practices for the preparation of infant formula in both the home and hospitals should be carefully controlled through the regular creation of the awareness that PIF are not sterile but that they may contain potential pathogens. The use of hygienic measures during preparation and reconstitution of PIF are essential. The risk of foodborne illness in neonates and infants fed infant formula can be reduced if guidelines for the preparation, storage and handling of PIF are strictly adhered to (Silano et