RESULTS AND DISCUSSION

The fatty acid composition of the seven presumptive Streptococcus sp., were determined, the CFA composition of the bacterial isolates studied consisted primarily of straight-chained saturated and monounsaturated fatty acids (Table 1). Six of the 21 different fatty acid types were excluded from the final data analysis since these were detected in 25% or less of the isolates (n1 or 2/7).

Table 1:	Cellular Fatty Acid (CFA) profile generated from fatty acids extracts of seven individual fish isolates belonging to Streptococcus sp. and an ATCC 12386 (reference strain used in these studies). The table represents the mean percent composition, standard deviation, and frequency percent of cellular fatty acids
*For the fatty acid composition the number to the left of the colon is the number of carbon atoms and the number to the right is the number of double bonds. *The portion of the double bound can be located by counting methyl w end of the carbon *chain A cis isomer is indicated by the suffix c. *SD: Stander deviation. *NO: Due to low frequency of the FA in the tested isolate it was excluded from final analysis

The frequency and mean percentages of the other fatty acid methyl esters (FAMEs) are shown in Table 1. The mean values were calculated by using FAMEs data for the tested isolates, regardless of the fact that every fatty acid was not detected in all of the isolates. All seven isolates produced two summed features, the summed features represent groups of two or three fatty acids which could not be separated by gas-liquid chromatography with MIDI system, the summed feature 3 (16:1 w7c/15iso 2OH and/or 15:0 ISO 2OH/16: 1w7c) and summed feature 5 (18:2 w6, 9c/18:0 ANTE and/or 18:0 ANTE/18:2 w6, 9c) were found in all the isolates.

The Cellular fatty acid Identification were made based on a strain's having cellular fatty acid composition and products of fermentation consistent with species of the genus Streptococcus agalactiae. Averaged fish isolate data plus ATCC 12386 gave rise to, using principle component analysis, a value of 5.32, (Table 1) whereas the same data with just fish isolates averaged 1.34. In analysis of data an averaged values of <10 suggests that the strains studied are members of the same taxon group. However, the lower the number, (such as found for fish isolates alone), the more the strains are considered as being like colonel. All strains appeared to be nearly identical to each other after CFA analysis using Library Generation System (LGS) software (MIDI) and were consistent to that of S. agalactiae ATCC 12386 (reference strain used in these studies) (Table 1).

To ensure 16S sequencing was performed on pure culture and to obtain classic bacteriological profiles, a number of accepted bacteriological tests (i.e., growth on differential media, biochemical tests, serological tests etc.) were executed (data not shown). The seven bacterial isolates were confirmed to be pure strains and were classified accordingly.

A significantly high degree of identity between the seven test isolates (100% similarity) with respect to 1,300 to 1,500 bp of 16 SrRNA gene sequence was found. The consensus sequence determined for the S. agalactiae 16S rRNA was examined for sequence homology with those of the prokaryotes by using National Centre for Biotechnology Information (NCBI), Gene Bank BLASTN function. The sequence alignment using BLASTN 2.2.1 for the comparison up to 1500 bp of the 16SrRNA gene sequence gave a high homology 100% to plus, plus strand with Streptococcus agalactiae strain A909 (ATCC 27591) (Accession number NC_007432.1) and S. agalactiae strain H36B (ATCC 12401; NCTC 8187). (Accession number AAJS01000007). The sequence alignment with sequence derived from validated S. agalactiae species (derived from GenBank) indicates the genus and species to be Streptococcus agalactiae (Table 2, Fig. 1).

Subsequently, an evolutionary distance matrix was then generated from the nucleotide sequences in the data set and a phylogenetic tree was constructed based on FastME algorithm using MEGA software. It clearly appears that all the seven isolates are affiliated with the species agalactiae. According to the phylogenetic tree, (Fig. 1) the species suis, equi, dysaglactiae, pyogenes, uberis, thermophilis and pneumoniae had a similarity of 90% or more with the test isolates. The most closely related species was S. suis and S. equi with similarity of 92 and 93 % to the test isolates (Table 2, Fig. 1)

Table 2:	Similarity percentage of the top twenty microorganisms with the highest similarity percentage to the test isolates 16S rRNA sequence

Conventional and molecular identification methods were used to investigate the isolates of the different organs from the different fish were S. agalactiae. The biochemical characterization of the seven isolates supporting the identity of the isolated bacteria as a β-hemolytic group B Streptococcus, this was also confirmed by other work done by Al-Marzouk et al. (2005) and Evans et al. (2002, 2008) due to high similarity between S. iniae, Lactococcus graviae, S. dificile, S. shiloi with S. agalactiae. We established the species confirmation using two other molecular methods namely 16S rRNA gene sequencing method and CFA composition analysis results.

Limited information was available on the epidemiology of Kuwaiti S. agalactiae isolates recovered from different fish species. To our knowledge, only three studies using DNA-based techniques have been carried out with a large collection of field isolates of fish and sewage origin in Kuwait (Qasem et al., 2008; Olivares-Fuster et al., 2008; Duremdez et al., 2004).

In the present study, at least seven isolates with an identical 16SrRNA and CFA pattern were found and sequenced. This fact might suggest that, in some isolates, there may be a common source of S. agalactiae in different infected fish from the same region (Welch, 1991). No sequence variation was observed for fish isolates originating from the different fish species, except for those from same species (Mullet). This result is consistent with a previous report of Olivares-Fuster et al. (2008), who analyzed S. agalactiae of fish origin by single standard conformation polymorphism (SSCP). They reported the Kuwaiti isolates as one genotype group while the other clustered of non-Kuwaiti isolates (USA, Brazil and Honduras) indicating correlation between geographical origin of genotypes was observed (Olivares-Fuster et al., 2008).

The traditional identification of bacteria on the basis of phenotypic characteristics is generally not accurate as identification based on genotypic methods. Comparison of the bacterial 16S rRNA gene sequence has emerged as a preferred genetic technique. The 16S rRNA gene sequence analysis can better identify poorly described, rarely isolated, or phenotypically aberrant strains (Woese et al., 1985). The 16S rRNA gene sequencing was conformed the species identity of the 7 fish isolates to be S. agalactiae. The sequencing results gave further support to the biochemical and PCR as to the identity of the isolates being S. agalactiae. Despite its accuracy, 16S rRNA gene sequence analysis lacks widespread use beyond the large and reference laboratories because of technical and cost considerations (Edwards et al., 1989).

The method used for the characterization was 16S rRNA sequencing, which is genomic approach to the classification of bacteria based upon the 16S ribosomal RNA gene (Lane et al., 1985). This gene is both ubiquitous and highly conserved among prokaryotic organisms, making it a useful primer site for gene amplification by PCR (Lin and Schwarz, 2003). Results emphasized the need to take a polyphasic approach when isolating and identifying bacterial isolates, particularly when they are to be used in future pathogenicity studies.

Fig 1:	Polygenetic tree based on Fast Minimum Evolution (FastME) after alignment of 1500 bp from 16 S rRNA sequences of the Kuwaiti isolates bacteria from diseased fish, with maximum sequence difference of 0.1 created using MEGA V 4.0 software

The present study genetically characterized seven bacterial isolates derived from infected fish and compared these results against bacteriological and metabolic identification techniques to assess the need for a polyphasic approach when identifying bacteria from environmental and fish samples. As bacteriological and metabolic data showed discrepancies across different phenotypic analyses (Al-Marzouk et al., 2005), present results suggest the need to incorporate 16S rRNA analysis to identify the strains.

In conclusion the bacterial isolates of fish are identical or similar with high homology; this might suggest that the bacteria that infected the Mullet originated from the same source. This is the first work from fish kill in Kuwait using CFA technique to identify and characterize Streptococcus agalactiae from Kuwait. It is interesting that, by phylogenetic analysis of the 16S rRNA gene sequence, S. agalactiae; S. suis, S. uberis and S. equi were most closely related (92 to 93% identity) with each other but at times are difficult to differentiate even after using a polyphasic approach, for phenotypically aberrant strains (Evans et al., 2008). Also, the conclusion derived from CFA analysis were consistent with those derived from morphological, cultured and physical characteristic (Welch, 1991). We conclude that the utility of CFA testing as an in vitro procedure that can help to verify or to refine the identity of species of viable bacteria present in environmental samples.

ACKNOWLEDGMENTS

We thank Mr Schder C. Solman and Mss Salwa A. Al-Mouqati for there excellent technical assistance. And we also would like to thank Kuwait Institute for Scientific Research for partial funding of this work (FB019K).