PCR-based technique for the detection of bacteria in semen and urine

Microorganisms associated with genito-urinary tract infections are often difficult to detect due to limitations associated with culture techniques. We have applied PCR-based detection of clinical isolates to complex sample matrices. Clinical isolates of Chlamydia trachomatis, Escherichia coli, Proteus vulgaris and Pseudomonas aeruginosa were grown, diluted and used to spike human urine and human semen. The urine and semen samples were centrifuged and the bacterial pellet was kept. A lysis buffer containing proteinase K, 8-methoxypsoralen and lauryl alcohol polyether was exposed to UV light to remove the bacterial DNA in the proteinase K, and was added to the bacterial pellet. After digestion, the proteinase K was destroyed and the lysates were subjected to 35 cycles of 16S rDNA amplification using a hot-starts technique and two primer pairs specific for eubacterial 16S rDNA: 8FPL-806R and 515FPL-13B. After amplification, the amplicons were cloned and sequenced to confirm amplification of the bacteria used to spiked the samples. We were able to detect as few as 10(5) bacteria per mi of urine or semen. Thirty unspiked semen samples were tested by PCR, and 17 were positive, including 10 samples negative by routine culture. The amplicons were cloned and sequenced for four PCR-positive/culture-negative semen samples: the 16S rDNA sequences obtained were mainly from strict anaerobes, including Prevotella spp. and Peptostreptococcus spp. Some 16S rDNA sequences were obtained that did not match any other 16S rDNA sequences available in various nucleic acid databases. Half of 14 unspiked urine sample tested were positive by PCR, including four samples negative by routine culture. Amplicons were cloned and sequenced for one urine sample showing only Lactobacillus spp. by routine culture, and sequences from Bacteroides ureolyticus, Clostridium spp., Corynebacterium urealyticum, Peptostreptococcus spp., and Lactobacillus acidophilus were found. These methods have great promise for the rapid detection of viable, but non-culturable bacteria in semen and urine. We are currently applying this technique for the detection of bacteria associated with idiopathic inflammatory conditions.