To employ partial 16S rDNA PCR and automated sequencing technique to identify non-culturable causal agents of bacterial meningitis, 73 peripheral blood samples and 413 culture-negative and eight culture-positive CSF clinical specimens from patients with suspected acute meningitis were examined for the presence of bacterial genomic DNA employing broad range 16S rDNA PCR followed by sequencing of the amplicons. In blood samples, 63/73 specimens were PCR positive (86.3%) and after direct sequencing of the PCR amplicons, only 12.7% (8/63) gave clear sequencing results and 55/63 (87.3%) were mixed with more than one organism detected. The mixed PCR amplicons were separated by using PAGE and mixed amplicons from 29/55 (52.7%) specimens were successfully identified through sequencing. Of the CSF samples, 8/8 culture-positive samples were also PCR positive and 45/413 (10.9%) of culture-negative gave a strong PCR signal and 88/413 (21.3%) specimens yielded a weak PCR signal. The remaining 280 culture-negative specimens were also PCR negative. Nested PCR was set up for the 88 weak positive samples and yielded 72/88 (81.8%) strong positives, with the remainder failing to amplify 133/413 (32.2%) culture-negative samples were PCR positive. In this study, the most common bacteria identified from blood specimens were Neisseria meningitidis, 13/63 (20.6%); Streptococcus spp, 5/63 (7.9%); Acinetobacter spp and Pseudomonas spp 4/63 (6.3%). From culture-negative CSF, the pattern was different in that Staphylococcus spp (13/58, 22.4%), Neisseria meningitidis (9/58, 15.52%) and Pseudomonas spp (8/58, 14.79%), were the most frequent. Overall, 16S rRNA broad-range PCR combined with direct DNA sequencing is a valuable molecular tool to aid with the detection as well as identification of non-culturable aetiological agents of acute bacterial meningitis and can augment cultural methods in the diagnosis of central nervous system infections in patients who have been treated with antibiotics. However, this study demonstrates that contamination is an important complication of the molecular assay, which should be attempted to be eliminated through careful laboratory controls. Hence there should be careful interpretation of any molecular finding, in tandem with other laboratory findings, such as culture, immunological and biochemical markers, and the clinical scenario of the patient.

Edible seaweed (Palmaria palmata) has traditionally been consumed rawafter harvesting from coastal shorelines around Northern Ireland.To date, there have been no reports examining the microbiology of this material, hence itwas the aimof this study to detect the diversity of themicro£ora found on readytoeat produce. Conventionalmicrobiological analyses of the product failed to detect any gastrointestinal pathogens and gave a mean total count of 1•3×105 cfug-1. 16S rRNA sequencing of culturable bacteria identified the Sanguibacter/Oerskovia/Cellulomonas complex, the Clavibacter/Frigoribacterium/urtobacterium complex, Enterobacter agglomerans, Erwinia herbicola, Flavobacterium spp., Micrococcus lylae, Microbacterium spp., Corynebacterium spp., and Dietza maris. A comparison of growth of isolated environmental organisms was performed to ascertain the most appropriate artificial culture media to employ for their culture in vitro. Tryptone soya broth with yeast extract (TSBYE) and brainfiheart infusion brothwith yeast extract (BHIYE) may be employed as suitable basal broth media for the laboratory culture of these organisms.This is the ¢rst preliminary report on the microbial diversity of edible seaweed and demonstrated the presence of several halophilic genera and species in fresh ready-to-eat edible seaweed from Northern Ireland. Although no gastrointestinal pathogens were cultured from this material, a larger study requiring examination of seasonal e¡ects, quality of marine water and e¡ect of drying on faecal pathogens, is required to support a functional HACCP-based approach to ensuring safety of this product.

Summary Universal or 'broad-range' eubacterial polymerase chain reaction (PCR) was performed on 53 isolates from environmental water-associated sites in a haematology unit (N=22), and the outer surfaces of cleaning lotion containers sited throughout a tertiary referral hospital (N=31). 16 S rDNA PCR was performed using two sets of universal primers, including the novel reverse primer, XB4, to generate a composite amplicon of 1068 bp, which was sequenced to obtain each isolate's identity. Sequence analysis was able to identify 51 isolates. Most (75% from the haematology unit and 81% from cleaner containers) were Gram-positive. Nine different genera were identified from the haematology unit and 13 from the cleaning lotion containers. This study provides the first reports of Terrabacter spp. And Brachybacterium paraconglomeratum isolated from a hospital environment. As unusual and difficult-to-identify environmental organisms are unlikely to be clinically significant, and molecular identification is costly and labourintensive, we recommend that molecular methods are only used as an adjunct to first-line phenotypic identification schemes where a definitive identification is required. Where molecular identification methods are justified, partial 16 S rDNA PCR and sequencing employing the novel universal primer XB4, is a valuable and reliable technique.

During October 1999 to November 2000, 98 blood culture specimens from the same number of febrile episodes originating from 49 patients with haematological malignancies were examined for the presence of eubacteria and fungi based on 16S rRNA gene and the 5.8, 18 and 28S rRNA combined with in vitro PCR amplification and sequencing, in addition to conventional blood culture laboratory techniques. Nineteen of the samples were associated with positive blood cultures. Eubacterial (16S rRNA) PCR detected bacterial DNA in 26 febrile episodes, i.e. in an additional seven febrile episodes than blood-culture alone. The species identified by partial 16S Rrna gene sequencing were as follows Staphylococcus spp (n=6), Staphylococcus epidermidis (n=5), Acinetobacter spp (n=5), Escherichia coli (n=2), Enterobacter agglomerans (n=2), Campylobacter spp (n=1), Citrobacter spp (n=1), Corynebacterium spp (n=1), Enterobacter faecium (n=1), Ralstonia spp (n=1), Acidovorax spp. (n=1) and Stenotrophomonas maltophilia (n=1). Gram-positive bacteria were found in 12/27 (44.6%) and Gram-negative bacteria were found in 15/27 (55.6%). After optimization of a PCR-based fungal detection method, none of the febrile episodes were shown to be attributable to fungi. The results of this study suggest that fungi are not common causal agents of febrile episodes in patients with a haematological malignancy at this centre and that molecular techniques can augment cultural methods in the diagnosis of causal agents of bacteraemia in patients, so that appropriate antibiotic regimens may be commenced in patients with culture-negative episodes of infection.