FLOW CYTOMETRY

Several studies have shown the applicability of using ‘‘viability markers’’ and flow cytometry for the rapid enumeration of microorganisms in pharma- ceutical grade water [26–28]. The viability maker most commonly used is based upon the reaction of bacteria with the ChemChrome B (CB) dye. Sample preparation consists in filtering the sample through a 0.45-Am membrane followed by cell labeling and laser scanning (Fig. 1). The dye, a fluorescein-type ester, is converted to a fluorescent product, a free fluorescein derivative, by intracellular esterase activity after being taken up by microbial cells previously captured by membrane filtration (Fig. 2). Microbial cells with an intact cell membrane only retain the fluorescein derivative. The bacteria are then enumerated by using a laser scanning instrument, which has been shown to be sensitive down to one cell in a sample within 90 min, and dem- onstrated a substantially wider linear range than the conventional hetero- trophic plate count method. Similar results have been found by fluorescent staining using 4V-6-diamine-2-phenylindole (DAPI), membrane filtration with

FIGURE 1 Sample preparation for flow cytometry analysis. Courtesy of AES- Chemunex.

FIGURE2 Determination of viability by flow cytometry analysis. Courtesy of AES-Chemunex.

tryptic soy agar (TSA) and R2A as growth media, and flow cytometry. An ion-exchange system, reverse osmosis system, and purified water in a hot loop have been sampled and processed. Fluorescence microscopy analysis of water samples using DAPI has resulted in higher microbial counts because DAPI stained all cells containing DNA including dead cells. Of the two growth media used for membrane filtration, R2A has shown higher microbial numbers than TSA because of the longer incubation time. However, flow cytometry has generally demonstrated a cell recovery closer to R2A. Rapid and accurate enumeration of labeled microorganisms is completed within 90 min. Bacterial numbers obtained by the laser scanning instrument appear to be higher than standard plate counts by an order of magnitude. Analysis of tap water, purified water, and water for injection (WFI) at several pharma- ceutical sites has also shown that flow cytometry is equivalent to the con- ventional membrane filtration method. Recovery studies in pure cultures demonstrate a good correlation between methods, with a coefficient of cor- relation of >0.97 for all organisms tested (vegetative bacteria, spores, yeast, and mold). However, none of the studies reported the multiple processing of water samples. Furthermore, the assay does not provide accurate quantita- tion when samples exhibit more than 104cells/membrane. The scanning of the filters is interrupted due to the agglomeration of cells resulting in a high fluorescence background. Nevertheless, because of recent modifications to the instrument, a higher accuracy can be achieved with 105cells/membrane for bacteria and 104cells/membrane for yeast and mold [29].

Additional studies have recently been performed on the macrolide an- tibiotic, Spiramycin, using solid phase cytometry [30]. Artificially contami- nated samples of the antibiotic have been analyzed. The solid phase cytometry has been found to detect all microbes regardless of their sensitivity to the bacteriostatic activity of the drug. With the conventional heterotrophic plate method run in parallel, complete recovery has been only obtained for Spir- amycin-resistant organisms. The spiked microorganisms that were sensitive to the antibiotic have remained inhibited or stressed by the action of the Spiramycin and do not grow on the plate but are detected by flow cytometry. These results further indicate the inadequacy of standard methods to recover injured microorganisms.

Bioburden of in-process samples of recombinant mammalian cell cul- tures have also been performed using flow cytometry [31,32]. Instead of the 7-day incubation time required for standard bioburden testing, analyses are completed within 4 hr. The assay is sensitive enough to detect from 5 to 15 CFU/mL after 4 hr. The advantage of rapid analysis of in-process samples is that bioburden results are known before a batch is pooled or processed. In some cases, microbial contamination has been found after the batches are polled and processed resulting in huge financial losses. However, to optimize

the detection of bacteria from a background of mammalian cells, different sample preparation procedures and modification of the original protocol are needed. Residual fluorescence appears to be a problem when detection limits go down to 1 cell/filter.

Another current application of flow cytometry to pharmaceutical quality control is the enumeration of biological indicators (BIs) [33]. BIs are used for the validation of sterilization cycles in pharmaceutical environments. Once the BIs are exposed to the sterilizing agent, the level of lethality must be determined. Conventional enumeration testing of BIs is based on the stan- dard plate count of serial dilutions. Because sample incubation is required for growth of visible colonies, results are obtained after 2 or more days. Fur- thermore, results might vary for different types of BIs based on media and culture conditions. Flow cytometry analysis has demonstrated that spore trips showed interference from paper, counts were lower than plate counts. Modifications of the sample preparation prior to flow cytometry analysis demonstrated that enumeration of BIs is faster, e.g., 2–4 hr, and that results were equivalent to standard plate counts. The advantages of using a rapid method to analyze BIs are a significant reduction in sterilizer holding time, cycle development time, and better understanding of lethality and sterility assurance.

6. IMPEDANCE

When microorganisms grow in enrichment media as a result of microbial metabolism, some of the substrates are converted into highly charged end products. These substrates are generally uncharged or weakly charged but are transformed during microbial growth. Because of their nature, the end products increase the conductivity of the media causing a decrease in im- pedance. Impedance is the resistance to flow of an alternating current as it passes through a conducting material.

Impedance detection time (Td) is when the resistance to the flow of an

alternating current indicates the growth of a particular microorganism as a result of changes in the growth media. Several studies have shown the ap- plicability of direct impedance for detecting microbial activity in pharma- ceutical products. Because impedance is a growth-dependent technology, a medium must be chosen that will support the growth of microorganisms and also to be optimized for electrical signal. Substrates for this kind of media will be uncharged or weakly charged—such as glucose that, when converted to lactic acid, will increase the conductivity of the media. However, a current modification called indirect impedance monitors microbial metabolism by measuring the production of carbon dioxide. The carbon dioxide removed from the growth media results in a decrease in conductivity. The use of in-

direct impedance allows the use of media that might not generate an optimal electrical response by using the direct method.

A good correlation between direct impedance detection time (Td) and

total colony counts has been obtained for untreated suspensions of S. aureus ATCC 6538, C. albicans ATCC 10231, A. niger ATCC 16404, and P. aeru- ginosaATCC 9027 in phosphate-buffered saline (PBS) [34]. Similar results have been found with suspensions of test microorganisms treated for varying contact periods with selected concentrations of antimicrobial agents. The only difference found is that the detection time for treated cells is extended. The assay is sensitive enough to detect bacteria, yeast, and mold.

Impedance has been compared to the direct epifluorescence technique (DEFT-MEM) and ATP bioluminescence (ATP-B) for detecting microbial contamination in cells exposed to different antimicrobial agents [35]. ATP-B, impedance, and DEFT-MEM have shown a strong correlation between the rapid method response and total colony counts for bacteria and yeast. However, for mold, impedance has been the only rapid method that showed a strong correlation between colony counts and the rapid method. When chlorhexidine-treated suspensions of S. aureus ATCC 6538 and C. albicans ATCC 10231 have been analyzed by impedance a good dose–response curve was obtained. Different results have been found with ATP-B and DEFT- MEM methods, which underestimate the kill by the order of 1–6 logs. Im- pedance application to pharmaceutical screening requires the development of growth curves for different microorganisms. Furthermore, the systems available do not provide high throughput.