Monitoring the Quality of Water

Millipore Corporation’s Marilyn Romieux and Serge Ohresser look at the benefits of a rapid microbiology detection system.

Water is a key raw material utilized in the manufacturing of products within the pharmaceutical industry. The microorganisms found in pharmaceutical water systems are mainly stressed, slow-growing strains characterized by long incubation times before growth can be detected by traditional microbiology methods such as membrane filtration or pour plates. In addition, more than one opportunistic microorganism can be present in a sample. As a result, the incubation time must be adapted to the strain characterized by the organism with the slowest growth rate. The time it takes before contamination can be detected in water can cause delays in product release, and extend the storage time of products. Using a rapid microbial detection method allows manufacturers the ability to identify microbial contamination earlier, which provides them with better process control, product yield, and shortens time to market.

The Milliflex Rapid Microbiology Detection System from Millipore Corporation is an automated solution for the rapid detection and enumeration of microbial contamination in samples filtered from different locations throughout the manufacturing process.

Methods

To determine the incubation time necessary to detect waterborne microorganisms with the Milliflex Rapid system, an established protocol was followed. The incubation time required to grow the same microorganisms using the traditional membrane filtration method (control) with mixed esters of cellulose membrane funnels was also determined. The filtrations were done in replicates of five for both the traditional microbiology method and Milliflex Rapid system for each microorganism.

Results

Table 1 provides the incubation times for detectable growth, by organism, for the traditional microbiology method and the Milliflex Rapid system. The detection time is significantly reduced using the Milliflex Rapid system. Detection of growth is on average 4.5 times faster than traditional microbiology, and up to six times faster for the very slow growers tested (Methylobacterium mesophilicum ATCC 29983, stressed strain of Methylobacterium and a mix of various slow-growing strains).

The Milliflex Rapid system allows for overnight detection of the industrial-stressed microorganisms tested versus the six days of incubation required in order to detect these organisms using the traditional method.The mean recovery between the Milliflex Rapid system and the traditional microbiology method in these experiments was 92.7 percent, demonstrating the equivalence of the two methods.

Mixed populations

If contamination is present in industrial samples, more than one organism is typically present per sample. Mixed microbial populations require the incubation time to be adapted to the strain characterized by the slowest growth rate. This allows time for the colonies produced by the slower growing microorganisms to become visible for an accurate count.

Often a smaller colony can be obscured by a larger, faster-growing colony. This phenomenon is also observed when employing bioluminescence methods. A bioluminescence signal is more intense with larger, faster growing colonies, which contain greater amounts of ATP than smaller, slower growing colonies do. Whether using traditional or bioluminescence methods, this phenomenon can lead to inaccurate enumeration.

Ideally, detection systems must be able to properly detect and enumerate colonies of different sizes and growth rates.

Laboratory-prepared ATCC mixed population

A mix of ATCC strains, Burkholderia cepacia (ATCC 25416) and Staphyloccus epidermidis (ATCC 12228), was prepared. The mixed population was tested with the Milliflex Rapid system using TSA medium at 30°C. The incubation time was determined by the strain with the slowest growth rate, which is B. cepacia. B cepacia was detected in 16 hours as show in Table 1.

With 16 hrs of incubation, colonies of S. epidermidis were over-incubated, as its optimal incubation time is only nine hours. Both strains were analyzed separately, and as a mixed population using the traditional method and the Milliflex Rapid system.

The results are shown in Table 2. The images (2D and 3D) generated by the Milliflex Rapid system show that both types of microorganisms are properly detected and enumerated in pure cultures as well as in a mixed culture. The over-incubated colonies of S. epidermidis produce high 3D peaks, but they do not hide the smaller surrounding colonies of B. cepacia in a mixed culture. The narrowness of the base of the peaks corresponding to the fast-growing strain allows the detection and the accurate counting of the slower-growing strain.

With a recovery of 101.5 percent, the Milliflex Rapid system detects these two ATCC strains mixed in the same sample using a different incubation time in only 16 hours instead of two days using traditional microbiology methods.

Natural mixing of industrial slow-growing strains

Strains of Brevundimonas vesicularis and Mycobacterium sp. were obtained upstream of a Milli-RO® purification system, and detected using the membrane filtration technique. Growth was observed after incubation for 6 days on R2A medium at 30°C. This natural mixture of industrial slow-growing strains was tested with the Milliflex Rapid system. The incubation time was 24 hrs (Table 1), however, some strains grow faster than others.

Table 3 provides results for this sample grown on R2A medium at 30°C using the Milliflex Rapid system. These results were compared to traditional microbiology counts. The images show a mixture of colonies producing 3D peaks of very different heights, characteristic of a mixture of strains, in approximately 24 hours with a mean recovery of 106.2 percent, as opposed to six days as seen with traditional microbiology.

Environmental Isolate of Ralstonia pickettii in a mixture with ATCC Strains

A mix of an environmental isolate of R. pickettii and two ATCC strains, P. aeruginosa ATCC 9027 and E. coli ATCC 8739, was prepared. The mixed population was tested with the Milliflex Rapid system using R2A at 30°C. The incubation time was determined by the strain with the slowest growth rate, which is the environmental isolate of R. pickettii. This strain was detected in 11 hours, as shown in Table 1. With 11 hours of incubation, colonies of P. aeruginosa and especially E. coli are over-incubated, as their incubation times are nine hours and six hours respectively.

The strains were analyzed separately and as a mixed population using traditional microbiology and the Milliflex Rapid system (Table 4). The images (2D and 3D) generated by the Milliflex Rapid system show that after 11 hours of incubation, the system is able to provide an accurate counting of all types of spots/peaks in the mixed cultures, as the over-incubated colonies do not hide the smaller surrounding colonies of R. pickettii.

With a recovery of 78 percent, the Milliflex Rapid system detects this mix of environmental and ATCC strains in only 11 hours instead of two days using traditional microbiology. This demonstrates that the Milliflex Rapid system is able to test mixed populations of microorganisms exhibiting different growth rates.

Conclusions

The results of this study show that the Milliflex Rapid Microbiology Detection System reduces the time to detect water-stressed microorganisms by a factor of 4.5 compared to traditional microbiology, allowing an overnight detection of strains that would require up to six days of incubation using the traditional membrane filtration method. In addition, the data shows the Milliflex Rapid system allows for the accurate detection and counting of different types of strains when mixed together in the same sample.

The ability to quickly detect and enumerate microbial contamination in industrial water systems allows manufacturers to have better process control, higher product yields, and to reduce time to market. Knowledge of contaminants can be used to identify early risks during manufacturing and implement corrective actions. The product release can be accelerated and storage time decreased. This result is an efficient cost savings for the manufacturer and an improved quality assurance of the product.

The authors would like to acknowledge Gilles Aubut, Frank Panofen and Jamie Russo, of Millipore Corporation, for their contributions to this study.

Milliflex and Milli-RO are registered trademarks of Millipore Corporation. ATCC is a registered trademark of the American Type Culture Collection.