Microorganisms have been used as producers of primary and secondary metabolites as exemplified by amino acids and antibiotics. For these purposes, microorganisms possessing desirable abilities have been traditionally screened. However, we are recently able to employ another approach utilizing (meta) genome databases. By comparative genomics, we revealed alternative pathways for menaquinone and peptidoglycan biosynthesis and developed new screening strategies for antibiotics specifically acting on pathogens possessing these new pathways. We have also been mining biosynthetic genes of secondary metabolites and characterized them in detail. In this review, I briefly summarize these results.

Children aged 6–18 years with a medical history of hematopoietic neoplasm and other immunocompromising conditions are at high risk for invasive pneumococcal disease (IPD). The Advisory Committee on Immunization Practices in the United States has therefore recommended that these children be immunized routinely with 13-valent pneumococcal conjugate vaccine (PCV13). Little is known, however, about the immunity of these children to pneumococci of PCV13 serotypes, except for heptavalent pneumococcal conjugate vaccine (PCV7) serotypes. Serum samples were therefore collected from Japanese children aged 5–18 years being followed up for hematological neoplasms or solid tumors, and concentrations of specific antibodies against Streptococcus pneumoniae PCV13 (non-PCV7) serotypes were measured. None of these children had received PCV13. Against all six serotypes, the immunoglobulin G (IgG) levels of the study patients were lower than those were of age-matched healthy controls. Children who received chemotherapy within 6 months prior to sample collection, had lower serotype-specific IgG levels than those who did not. These findings indicate that many immunocompromised children not recommended for routine vaccination owing to their age group did not have PCV13 serotype-specific IgG levels that could prevent IPD. These results suggest that the vaccination with PCV13 is necessary for children at high risk of IPD.

β-D-Glucan (BDG) is a fungal cell wall component and a useful marker in diagnosing invasive fungal infections. However, false-positive BDG test results occasionally limit its utility. We recently treated a patient whose plasma BDG levels were elevated during administration of penicillin G potassium despite the absence of fungal infections. We therefore investigated the BDG contents of different penicillin G potassium products available in Japan. BDG levels were measured using two different assays. An injectable penicillin G potassium product, the only product approved for clinical use in Japan, was found to contain significant amounts of BDG in all batches and lots tested. According to the Fungitec G test MK II “Nissui” (cutoff value, 20 pg/mL) and β-glucan test Wako (cut-off value, 11 pg/mL) assays, BDG levels in 10,000 units/mL solutions of this agent (5 lots) ranged from 79.4 to 192.9 pg/mL and 16.9 to 34.5 pg/mL, respectively. The two other penicillin G potassium products tested, which are used for cell culture, did not contain significant amounts of BDG. The amounts of BDG contained in the injectable product are probably sufficient to cause false-positive test results in patients receiving clinical doses (e.g., 24 million units per day). Further clinical studies are needed to confirm this result.

Tazobactam/piperacillin (Zosyn^TM), it was approved sepsis, pneumonia, pyelonephritis and complicated cystitis as the indications in July, 2008. Postmarketing surveillance of this agent was conducted to grasp the frequency of the adverse drug reactions, detection of the unknown adverse drug reactions and the factors to influence the development of the adverse drug reactions, and the efficacies in three years from January, 2009. For this surveillance of safety and efficacy, 3,626 and 3,181 patients were analyzed respectively from 3,684 patients enrolled in 558 study sites of the whole country.

The adverse drug reactions were reported in 283 (317 events) of 3,626 safety analyzed patients, and its incidence rate was 7.80%, including diarrhoea (2.43%), liver function abnormal (1.74%), liver disorder (0.77%) rash (0.50%), renal impairment (0.30%), alanine aminotransferase increased (0.25%) and blood creatinine increased (0.25%) and white blood cell count decreased (0.22%).

Twenty-four serious adverse drug reactions were reported in 22 patients, including platelet count decreased, pseudomembranous colitis in three patients, respectively, and diarrhoea, liver disorder, renal failure, acute renal failure in two patients, respectively.

As the factors to influence on the development of the adverse drug reactions, sex, age, hepatic dysfunction before administration, history of drug allergy and combination therapy were extracted by the multivariate variable decrease method using the logistic-regression analysis.

The response rate in 3,181 efficacy evaluable patients was 86.0%. The response rates by the infections were 83.8% in pneumonia, 93.3% in pyelonephritis, 95.8% in complicated cystitis and 83.2% in sepsis, respectively. Bacteriological efficacy of Zosyn^TM was 84.0% in 425 patients, in which the causative bacteria were identified.

From these observations, Zosyn^TM is expected as a useful empiric therapeutic antimicrobial agent for the treatment of various infectious diseases as in the practical guidelines.