Enlargement of the WHO international repository for platelet transfusion-relevant bacteria reference strains

BACKGROUND AND OBJECTIVES

Interventions to prevent and detect bacterial contamination of platelet concentrates (PCs) have reduced, but not eliminated the sepsis risk. Standardized bacterial strains are needed to validate detection and pathogen reduction technologies in PCs. Following the establishment of the First International Reference Repository of Platelet Transfusion-Relevant Bacterial Reference Strains (the 'repository'), the World Health Organization (WHO) Expert Committee on Biological Standardisation (ECBS) endorsed further repository expansion.

MATERIALS AND METHODS

Sixteen bacterial strains, including the four repository strains, were distributed from the Paul-Ehrlich-Institut (PEI) to 14 laboratories in 10 countries for enumeration, identification and growth measurement on days 2, 4 and 7 after low spiking levels [10-25 colony-forming units (CFU)/PC bag]. Spore-forming (Bacillus cereusPEI-B-P-07-S, Bacillus thuringiensisPEI-B-P-57-S), Gram-negative (Enterobacter cloacaePEI-B-P-43, Morganella morganiiPEI-B-P-74, PEI-B-P-91, Proteus mirabilisPEI-B-P-55, Pseudomonas fluorescensPEI-B-P-77, Salmonella choleraesuisPEI-B-P-78, Serratia marcescensPEI-B-P-56) and Gram-positive (Staphylococcus aureusPEI-B-P-63, Streptococcus dysgalactiaePEI-B-P-71, Streptococcus bovisPEI-B-P-61) strains were evaluated.

RESULTS

Bacterial viability was conserved after transport to the participating laboratories with one exception (M. morganiiPEI-B-P-74). All other strains showed moderate-to-excellent growth. Bacillus cereus, B. thuringiensis, E. coli, K. pneumoniae, P. fluorescens, S. marcescens, S. aureus and S. dysgalactiae grew to >10⁶ CFU/ml by day 2. Enterobacter cloacae, P. mirabilis, S. epidermidis, S. bovis and S. pyogenes achieved >10⁶ CFU/ml at day 4. Growth of S. choleraesuis was lower and highly variable.

CONCLUSION

The WHO ECBS approved all bacterial strains (except M. morganiiPEI-B-P-74 and S. choleraesuisPEI-B-P-78) for repository enlargement. The strains were stable, suitable for spiking with low CFU numbers, and proliferation was independent of the PC donor.

Efficiency of riboflavin and ultraviolet light treatment against high levels of biofilm-derived Staphylococcus epidermidis in buffy coat platelet concentrates

BACKGROUND AND OBJECTIVES

Staphylococcus epidermidis forms surface-attached aggregates (biofilms) in platelet concentrates (PCs), which are linked to missed detection during PC screening. This study was aimed at evaluating the efficacy of riboflavin-UV treatment to inactivate S. epidermidis biofilms in buffy coat (BC) PCs.

MATERIALS AND METHODS

Biofilm and non-biofilm cells from S. epidermidis ST-10002 and S. epidermidis AZ-66 were individually inoculated into whole blood (WB) units (~10⁶ colony-forming units (CFU)/ml) (N = 4-5). One spiked and three unspiked WB units were processed to produce a BC-PC pool. Riboflavin was added to the pool which was then split into two bags: one for UV treatment and the second was untreated. Bacterial counts were determined before and after treatment. In vitro PC quality was assessed by flow cytometry and dynamic light scattering.

RESULTS

Bacterial counts were reduced during BC-PC production from ~10⁶ CFU/ml in WB to 10³ -10⁴ CFU/ml in PCs (P < 0·0001). Riboflavin-UV treatment resulted in significantly higher reduction of S. epidermidis AZ-66 than strain ST-10002 (≥3·5 log reduction and 2·6-2·8 log reduction, respectively, P < 0·0001). Remaining bacteria post-treatment were able to proliferate in PCs. No differences in S. epidermidis inactivation were observed in PCs produced from WB inoculated with biofilm or non-biofilm cells (P > 0·05). Platelet activation was enhanced in PCs produced with WB inoculated with biofilms compared to non-biofilm cells (P < 0·05).

CONCLUSION

Riboflavin-UV treatment was similarly efficacious in PCs produced from WB inoculated with S. epidermidis biofilm or non-biofilm cells. Levels of biofilm-derived S. epidermidis ≥10³ CFU/ml were not completely inactivated; however, further testing is necessary with lower (real-life) bacterial levels.

Bacterial survival and distribution during buffy coat platelet production

BACKGROUND AND OBJECTIVES

At Canadian Blood Services, buffy coat (BC) platelet concentrates (BC-PCs) show a generally lower bacterial contamination rate than apheresis PCs. This study investigated whether the PC production method contributes to this observation.

MATERIALS AND METHODS

Whole blood (WB) inoculated with eight bacterial strains was processed using the BC method. Bacteria were enumerated throughout BC-PC production and subsequent PC storage. Endotoxin production and bacterial adhesion to PC bags were evaluated during PC storage. PC quality was monitored by CD62P expression (flow cytometry) and changes in dynamic light scattering (ThromboLUX^® ).

RESULTS

During overnight WB hold, Staphylococcus epidermidis titres remained unchanged, commercial Escherichia coli and Klebsiella pneumoniae were eliminated and the remaining organisms proliferated to high concentrations. Through BC-PC production, bacteria segregated preferentially towards the cellular fractions compared to plasma (P < 0·05). During PC storage, most bacteria adhered to the PC bags and Gram negatives produced clinically significant endotoxin levels. Changes in CD62P expression or ThromboLUX scoring did not consistently reflect bacterial contamination in BC-PCs.

CONCLUSION

WB hold during BC-PC production does not have a broad-spectrum bactericidal effect, and therefore, other factors contribute to low rates of contamination in BC-PCs.

Evaluation of a universal point-of-issue assay for bacterial detection in buffy coat platelet components

Bacterial contamination of platelet concentrates poses a major post-transfusion infectious risk. This study was aimed at evaluating the efficacy of the BacTx(®) assay (Immunetics Inc.) for bacterial detection in leucocyte-reduced buffy coat platelet pools and for its sensitivity in detecting clinical isolates, including bacteria that form surface-attached aggregates (biofilm positives). Platelet pools were inoculated at bacterial concentrations of 0·8-13 CFU/ml. The BacTx(®) assay detected all species at concentrations ≥10(3)  CFU/ml within 20-69 h of platelet incubation. Detection of slow-growing and biofilm-forming strains was delayed in comparison with the other strains. This assay could be used as a point-of-issue method to increase the safety of the platelet supply.

Establishment of the first international repository for transfusion-relevant bacteria reference strains: ISBT working party transfusion-transmitted infectious diseases (WP-TTID), subgroup on bacteria

BACKGROUND

Bacterial contamination of platelet concentrates (PCs) still remains a significant problem in transfusion with potential important clinical consequences, including death. The International Society of Blood Transfusion Working Party on Transfusion-Transmitted Infectious Diseases, Subgroup on Bacteria, organised an international study on Transfusion-Relevant Bacteria References to be used as a tool for development, validation and comparison of both bacterial screening and pathogen reduction methods.

MATERIAL AND METHODS

Four Bacteria References (Staphylococcus epidermidis PEI-B-06, Streptococcus pyogenes PEI-B-20, Klebsiella pneumoniae PEI-B-08 and Escherichia coli PEI-B-19) were selected regarding their ability to proliferate to high counts in PCs and distributed anonymised to 14 laboratories in 10 countries for identification, enumeration and bacterial proliferation in PCs after low spiking (0·3 and 0·03 CFU/ml), to simulate contamination occurring during blood donation.

RESULTS

Bacteria References were correctly identified in 98% of all 52 identifications. S. pyogenes and E. coli grew in PCs in 11 out of 12 laboratories, and K. pneumoniae and S. epidermidis replicated in all participating laboratories. The results of bacterial counts were very consistent between laboratories: the 95% confidence intervals were for S. epidermidis: 1·19-1·32 × 10(7) CFU/ml, S. pyogenes: 0·58-0·69 × 10(7) CFU/ml, K. pneumoniae: 18·71-20·26 × 10(7) CFU/ml and E. coli: 1·78-2·10 × 10(7) CFU/ml.

CONCLUSION

The study was undertaken as a proof of principle with the aim to demonstrate (i) the quality, stability and suitability of the bacterial strains for low-titre spiking of blood components, (ii) the property of donor-independent proliferation in PCs, and (iii) their suitability for worldwide shipping of deep frozen, blinded pathogenic bacteria. These aims were successfully fulfilled. The WHO Expert Committee Biological Standardisation has approved the adoption of these four bacteria strains as the first Repository for Transfusion-Relevant Bacteria Reference Strains and, additionally, endorsed as a project the addition of six further bacteria strain preparations suitable for control of platelet contamination as the next step of enlargement of the repository.

Unexplained agglutination of stored red blood cells in Alsever's solution caused by the gram-negative bacterium Serratia liquefaciens

Alsever's solution has been used for decades as a preservative solution for storage of RBCs. From October 2005 to January 2006, unexplained hemagglutination of approximately 10 to 20 percent of RBCs stored for several days in a modified version of Alsever's solution was noticed in quality control testing at the Canadian Blood Services Serology Laboratory. An investigation, including microbial testing, was initiated to determine the cause of the unexplained hemagglutination. The gram-negative bacterium Serratia liquefaciens was isolated from supernatant solutions of agglutinated RBCs. Further characterization of this strain revealed that it has the ability to form biofilms; presents high levels of resistance to chloramphenicol, neomycin, and gentamicin; and causes mannose-sensitive hemagglutination. The source of S. liquefaciens contamination in RBC supernatants was not found. However, this bacterium has not been isolated since January 2006 after enhanced cleaning practices were implemented in the serology laboratory where the RBCs are stored. This biofilm-forming, antibiotic-resistant S. liquefaciens strain could be directly linked to the unexplained hemagglutination observed in stored RBCs.

N terminus determinants of MinC from Neisseria gonorrhoeae mediate interaction with FtsZ but do not affect interaction with MinD or homodimerization

While bacterial cell division has been widely studied in rod-shaped bacteria, the mechanism of cell division in round (coccal) bacteria remains largely enigmatic. In the present study, interaction between the cell division inhibitor MinC from Neisseria gonorrhoeae (MinC(Ng)) and the gonococcal cell division proteins MinD(Ng) and FtsZ(Ng) are demonstrated. Protein truncation and site-directed mutagenic approaches determined which N-terminal residues were essential for cell division inhibition by MinC(Ng) using cell morphology as an indicator of protein functionality. Truncation from or mutation at the 13th amino acid of the N terminus of MinC(Ng) resulted in loss of protein function. Bioinformatic analyses predicted that point mutations of L35P and L68P would affect the alpha-helical conformation of the protein and we experimentally showed that these mutations alter the functionality of MinC(Ng). The bacterial two-hybrid system showed that interaction of MinC(Ng) with FtsZ(Ng) is abrogated upon truncation of 13 N-terminal residues while MinC(Ng)-MinD(Ng) interaction or MinC(Ng) homodimerization is unaffected. These data confirm interactions among gonococcal cell division proteins and determine the necessity of the 13th amino acid for MinC(Ng) function.

Conserved glycines in the C terminus of MinC proteins are implicated in their functionality as cell division inhibitors

Alignment of 36 MinC sequences revealed four completely conserved C-terminal glycines. As MinC inhibits cytokinesis in Neisseria gonorrhoeae and Escherichia coli, the functional importance of these glycines in N. gonorrhoeae MinC (MinC(Ng)) and E. coli MinC (MinC(Ec)) was investigated through amino acid substitution by using site-directed mutagenesis. Each mutant was evaluated for its ability to arrest cell division and to interact with itself and MinD. In contrast to overexpression of wild-type MinC, overexpression of mutant proteins in E. coli did not induce filamentation, indicating that they lost functionality. Yeast two-hybrid studies showed that MinC(Ec) interacts with itself and MinD(Ec); however, no interactions involving MinC(Ng) were detected. Therefore, a recombinant MinC protein, with the N terminus of MinC(Ec) and the C terminus of MinC(Ng), was designed to test for a MinC(Ng)-MinD(Ng) interaction. Each MinC mutant interacted with either MinC or MinD but not both, indicating the specificity of glycine residues for particular protein-protein interactions. Each glycine was mapped on the C-terminal surfaces (A, B, and C) of the solved Thermotoga maritima MinC structure. We found that MinC(Ec) G161, residing in close proximity to the A surface, is involved in homodimerization, which is essential for MinC function. Glycines corresponding to MinC(Ec) G135, G154, and G171, located within or adjacent to the B-C surface junction, are critical for MinC-MinD interactions. Circular dichroism revealed no gross structural perturbations of the mutant proteins, although the contribution of glycines to protein flexibility and stability cannot be discounted. Using molecular modeling, we propose that exposed conserved MinC glycines interact with exposed residues of the alpha-7 helix of MinD.

Expression of Neisseria gonorrhoeae cell division genes ftsZ, ftsE and minD is influenced by environmental conditions

The activity of the promoter regions of the cell division genes ftsZ, ftsE, minC, minD and minE from Neisseria gonorrhoeae (Ng) was studied under different environmental conditions using lacZ translational fusions. The promoters of the minNg genes have not been previously determined and we identified promoter regions upstream of each gene (minCp, minDp and minEp). We determined that minDp had the strongest activity. Expression of the promoter regions of ftSZ(Ng) and ftsE(Ng), which we had previously identified, as well as minD(Ng), were then studied under conditions reflecting the environment of the genitourinary tract. These conditions included anaerobiosis, presence of isoleucine or urea (3 mM and 400 mM, respectively) and acidity of pH 6. Both beta-galactosidase expression and northern blot analysis indicated that all three genes were upregulated under anaerobiosis. The addition of isoleucine as well as media at pH 6 did not have any significant effects on the promoter activity of these genes while the presence of urea significantly decreased ftsZ(Ng) promoter activity. The expression of the minD(Ng) promoter region was analyzed during different growth phases and shown to follow the growth behavior of the culture. By contrast, the ftSZ(Ng) promoter activity continued to rise after the onset of the stationary phase. When gonococcal ftsZ promoter 1, (Pz1) was altered by site-directed mutagenesis, a significant decrease in the expression of ftsZ(Ng) was observed under both aerobic and anaerobic conditions. These data infer that gonococci regulate their cell division in response to different environments.

Gonococcal MinD affects cell division in Neisseria gonorrhoeae and Escherichia coli and exhibits a novel self-interaction

The Min proteins are involved in determining cell division sites in bacteria and have been studied extensively in rod-shaped bacteria. We have recently shown that the gram-negative coccus Neisseria gonorrhoeae contains a min operon, and the present study investigates the role of minD from this operon. A gonococcal minD insertional mutant, CJSD1, was constructed and exhibited both grossly abnormal cell division and morphology as well as altered cell viability. Western blot analysis verified the absence of MinD from N. gonorrhoeae (MinD(Ng)) in this mutant. Hence, MinD(Ng) is required for maintaining proper cell division and growth in N. gonorrhoeae. Immunoblotting of soluble and insoluble gonococcal cell fractions revealed that MinD(Ng) is both cytosolic and associated with the insoluble membrane fraction. The joint overexpression of MinC(Ng) and MinD(Ng) from a shuttle vector resulted in a significant enlargement of gonococcal cells, while cells transformed with plasmids encoding either MinC(Ng) or MinD(Ng) alone did not display noticeable morphological changes. These studies suggest that MinD(Ng) is involved in inhibiting gonococcal cell division, likely in conjunction with MinC(Ng). The alignment of MinD sequences from various bacteria showed that the proteins are highly conserved and share several regions of identity, including a conserved ATP-binding cassette. The overexpression of MinD(Ng) in wild-type Escherichia coli led to cell filamentation, while overexpression in an E. coli minD mutant restored a wild-type morphology to the majority of cells; therefore, gonococcal MinD is functional across species. Yeast two-hybrid studies and gel-filtration and sedimentation equilibrium analyses of purified His-tagged MinD(Ng) revealed a novel MinD(Ng) self-interaction. We have also shown by yeast two-hybrid analysis that MinD from E. coli interacts with itself and with MinD(Ng). These results indicate that MinD(Ng) is required for maintaining proper cell division and growth in N. gonorrhoeae and suggests that the self-interaction of MinD may be important for cell division site selection across species.

Deletion of the cell-division inhibitor MinC results in lysis of Neisseria gonorrhoeae

The minCDE genes involved in division site selection in Neisseria gonorrhoeae were identified using raw data from the N. gonorrhoeae genome project and are part of a cluster of 27 genes. When gonococcal min genes were heterologously expressed as a cluster in Escherichia coli, minicells and filaments were produced, indicating that gonococcal min genes disrupted cell division in other genera. The insertional inactivation of the minC gene of N. gonorrhoeae CH811 resulted in a strain (CSRC1) with decreased viability and grossly abnormal cell division as observed by phase-contrast and electron microscopy analysis. Western blot analysis of N. gonorrhoeae CSRC1 confirmed that MinC(Ng) was not produced. Complementation of CSRC1 by integrating a minC-6xHis tag fusion at the proAB locus by homologous recombination restored viability and 1.9 times wild-type levels of MinC(Ng) expression. This slight increase of expression caused a small percentage of the complemented cells to divide aberrantly. This suggested that the 6xHis tag has partially affected the stability of MinC, or that the chromosomal position of minC is critical to its regulation. Comparison of MinC proteins from different bacteria showed a homologous region corresponding to residues 135-230 with five conserved amino acids. Overexpression of MinC(Ng) in wild-type E. coli cells induced filamentation and an E. coli minC mutant was successfully complemented with minC(Ng). Therefore, the evidence indicates that MinC from N. gonorrhoeae acts as a cell-division inhibitor and that its role is essential in maintaining proper division in cocci.

Organization and transcription of the division cell wall (dcw) cluster in Neisseria gonorrhoeae

A cluster of genes involved in cell division and cell wall (dcw) biosynthesis was identified in Neisseria gonorrhoeae using genomic analysis and through verification of gene order by polymerase chain reaction (PCR) analysis. The gonococcal dcw cluster consists of 17 genes, in the order 5'-mraZ-mraW-ftsI-murE-hyp1-murF- mraY-hyp2-murD-ftsW-murG-murC-ddl -ft sQ-ftsA-ftsZ-hyp3-3'. The gene organization of the dcw cluster of N. gonorrhoeae is more similar to that observed in Gram-negative rods such as Escherichia coli and Haemophilus influenzae than in Gram-positive bacteria. The cluster is characterized by several intergenic spaces. Compared with E. coli, two genes, ftsL and envA, are absent in the gonococcal dcw cluster and three hypothetical genes are novel to the cluster. The cluster is flanked by two transcriptional terminators consisting of paired neisserial uptake sequences and also includes four internal terminators, three of which are paired neisserial uptake sequences. We also found that a repeated sequence on the gonococcal genome, commonly called a Correia element, acts as the fourth transcriptional terminator. All termination sequences were shown to be fully functional by using reverse transcription PCR experiments. Transcriptional start sites upstream of ftsQ, ftsA and ftsZ were determined by primer extension and six promoters were identified; three promoters were located upstream of ftsZ in the intergenic space, two were upstream of ftsA within ftsQ and one was upstream of ftsQ within ddl. Some of these promoters were preferentially used under anaerobic conditions. The location of these promoters differed from those described in E. coli indicating dissimilar transcriptional regulation.