Identifying microbes in non-sterile facilities

Published: 3-May-2012

Assuring compliance to the regulatory guidelines for microbial monitoring during the production of pharmaceuticals and nutraceuticals requires a thorough understanding of the regulatory environment and the different methodologies available for microbial identification and differential strain typing

You need to be a subscriber to read this article.
Click here to find out more.

Microbial identification is key for non-sterile facilities producing pharmaceuticals or nutraceuticals. Accugenix reviews modern identification methods for use in environmental monitoring programmes

For manufacturers of pharmaceuticals, dietary supplements or nutraceuticals, the presence of bacteria, filamentous fungi and yeasts are usually a cause for concern. However, a well-designed environmental monitoring (EM) programme should detect the presence of these micro-organisms before product contamination occurs.

When bacterial or fungal isolates are recovered from a production facility, it is important to be able to identify accurately the organism to the species, and possibly strain level, to track the potential origin of the contamination, mitigate it and avoid delays in product release or to complete investigations.

Accurate microbial identification requires significant and continuous process refinements, familiarity and expertise in interpreting data, consistent qualified methods of analysis and timely updates to organism libraries. Regardless of the method used, this information is only as reliable, or accurate, as the reference libraries used. To identify correctly a large percentage of the unknown isolates in EM programmes, the libraries must contain all species relevant to the manufacturing environment involved.1

Additionally, it is critical to update libraries continually and incorporate new entries to stay current with an ever-evolving microbial world, where taxonomic changes and novel species are described daily.

According to FDA2 and FAO3 guidelines, microbial monitoring of critical components, areas and personnel should include routine identification of isolates to the species level. Broadly speaking, three methods are currently used in commercial settings: genotypic, proteotypic and phenotypic (see Figure 1).


The FDA states that: ‘Genotypic methods have been shown to be more accurate and precise than traditional biochemical and phenotypic techniques. These methods are especially valuable for investigations into failures (e.g. sterility test; media fill contamination)’. 4 Genotypic identification methods involve sequencing different regions of the micro-organism’s ribosomal RNA genes (16S, ITS or D2), resulting in identification to the species or occasionally the sub-species level.

An emerging proteotypic technology in commercial bacterial identification for EM programmes is matrix-assisted laser desorption/ ionisation-time of flight (or MALDI-TOF) mass spectrometry, which uses a ribosomal protein spectral fingerprint to identify bacteria to the species level. Phenotypic technologies use either biochemical reactions or cell-surface component compositions to identify a micro-organism.

Each technology has advantages and disadvantages. The key is to balance the cost effectiveness of the technologies and to maintain a high level of accurate and reproducible identifications, while adhering to manufacturing guidelines and maximising the control over a production environment.

For identification of unknown microbial organisms, DNA sequencing rapidly provides data that are more accurate, robust and reproducible than relying solely on visual phenotypic characteristics. This is because the sequence-based result is not dependent on age and health of an organism, growth conditions or ancillary testing. In fact, samples can be viable or nonviable cultures or simply genomic DNA material collected from the microbe.

genotypic identification

Genotypic methods use comparative sequencing of the ribosomal RNA (rRNA) region. The use of ribosomal DNA sequences for microbial taxonomic classification has been in practice for decades because the technology is inherently stable and thus allows for reproducible data for classification as well as for identification.

In all living things, the ribosome contains different sized rRNAs, which are transcribed from ribosomal operons in the organism’s genome. The rRNAs fold into elaborate 3-D structures and are incorporated into an intricate protein-RNA complex that is critical for the ribosomal function and cell survival (see Figure 4).


Fig. 4: 16S ribosomal RNA structure for Escherichia coli

Fig. 4: 16S ribosomal RNA structure for Escherichia coli

Bacterial isolates are identified using the 16S rDNA region, because it is universally distributed among bacteria and contains species-specific variable regions. The identification of fungi, especially filamentous fungi, has historically been a very difficult task. Due to the amount of experience and time required to identify accurately filamentous fungi to the species level, it has been common practice to settle with either identifying these organisms to the genus level, or in some cases, simply identify them as ‘moulds’. Genetic approaches to fungal identification provide a more acceptable, reliable and rapid identification to the species level.

The internal transcribed spacer (ITS2) region is the ribosomal DNA region that is sequenced by fungal taxonomists, because it has a higher degree of variation between closely related species than other rDNA regions in fungi.5

To perform bacterial or fungal genotypic identifications, the target regions of the rRNA genes are amplified by PCR and sequenced (see Figure 3). The sequence data are then analysed and aligned in order of increasing genetic distance to relevant sequences against a library database to achieve an identity match. Since interpretation of the data and the database against which the sequence is compared are both essential parts of the identification process, it is important not to neglect the method of data analysis and library coverage when choosing an identification system or provider.


Figure 3: DNA sequencing chromatogram

Figure 3: DNA sequencing chromatogram

Systems can vary from fully automated to manual interpretation of the data. Manual data analysis and alignment are extremely repeatable and result in the ability to correct for standard sequencing anomalies and sequence variations that can cause the data to appear to be mixed or of poor quality. Poor quality data are usually truncated in fully automated analysis programmes, which can lead to faulty interpretation of the data and incorrect identification of EM isolates or of bioburden samples taken from final product.

proteotypic identification

In proteomics (a field devoted to studying the full set of proteins encoded by the genome) ‘proteotypic’ describes a peptide sequence that serves to identify a particular protein. In microbial identification, ‘proteotypic’ is used to indicate a protein spectrum that serves to identify a particular micro-organism. For industries that routinely need to identify micro-organisms, the ideal technology is one that is accurate, reproducible, fast and inexpensive.


Mass spectrometry has demonstrated increased accuracy and reproducibility over phenotypic systems

MALDI-TOF mass spectrometry exhibits all these traits. This proteotypic method has demonstrated increased accuracy and reproducibility over phenotypic systems.7 The analysis of a bacterial sample results in a unique protein spectral fingerprint that is then compared with a validated database for identification. The spectrum is composed of the ribosomal proteins of the bacteria, which are translated from DNA but are not subject to the expression variability seen in phenotypic methods. These ribosomal protein molecules are constitutively expressed and structural, so they are consistently present in the cell in very high levels.

MALDI-TOF analyses a small amount of bacterial sample from a fresh culture that is suspended in a solution containing a matrix and then dried to a target plate. The plate is placed in the vacuum source chamber and irradiated with a pulsed laser beam. The ionisation process utilises the matrix to absorb laser energy while protecting the proteins and transferring ions to the intact molecules. The proteins are released from the matrix and enter the gaseous state as charged molecules. The ions, accelerated by electrical fields, are then separated based on the time it takes them to travel a specified distance – time of flight (TOF).

The lower the mass of the ion, the faster it will reach the detector. The protein spectrum or fingerprint that emerges from this process is then compared with a library of spectra from known bacteria. For effective identification, the library needs to contain relevant organisms found in manufacturing environments and must be continually updated to include novel organisms and taxonomic changes.

For situations where full sequence data are not required, such as for routine monitoring, but when a repeatable, accurate, cost-effective identification is still required, the MALDI-TOF technology is a good alternative when supported by a robust, relevant library for EM programmes. However, because the technology is in its infancy for identifications of micro-organisms found in a production environment with respect to library database development, a polyphasic approach using the proven method of rDNA sequencing is recommended when an identification cannot be made by MALDI-TOF.

phenotypic identification

There are multiple systems that use phenotypic characterisation for microbial identification. These systems have been in use for decades and differentiate between organisms based on the results of biochemical tests, such as sugar fermentation, or physiological properties, such as salt or pH tolerance. One phenotypic method provides microbial identification based on cellular fatty acids that are extracted and methylated. The resultant methyl esters are separated by gas chromatography and their patterns are compared against a database.

All of these phenotypic systems require a healthy organism, and depending on the technology used, may or may not need to be cultured on specific media and require ancillary testing, such as Gram stain, to achieve identification. The systems tend to be easy to use and have high throughput and are primarily used in clinical settings. However, micro-organisms isolated from manufacturing environments on compendial media will probably be physiologically stressed and may not fully express their phenotypic or biochemical characteristics, resulting in erroneous identification.

For many organisms, traditional phenotypic identification is problematic. Identification can be dependent on media and temperature used to grow the organism and can lead to subjective interpretation of the test results and a higher rate of inaccurate identifications. Not all strains within a given species consistently exhibit a particular characteristic, thereby limiting phenotypic identification methods. Additionally, library databases used in support of phenotypic identification are often limited and geared towards clinical isolates.


Phenotypic reactions still have a role in a microbial quality programme

Despite their major shortcomings, these methods still have a role in a microbial quality programme. The results from the phenotypic reactions can help determine the biochemical activity of an organism on the product and the resulting stability of the product in the presence of that organism. Understanding the nutrient requirements of an organism can provide insight into controlling or eliminating the organism in the environment or preserving the product.

In conclusion, current available methods of identification range from genotypic and proteotypic to phenotypic, with 16S and ITS2 sequencing being the gold standards for bacterial and fungal identification, respectively, especially when combined with reference quality interpretation methods and curated libraries focused on organisms relevant to the dietary supplements and other manufacturing industries.


Figure 2: MALDI-TOF ribosomal protein spectrum

Figure 2: MALDI-TOF ribosomal protein spectrum

The introduction of the proteotypic MALDI-TOF-based method of identification provides an additional highly accurate, fast and inexpensive option for routine monitoring programmes. When identifications are based on phenotypic characteristics, such as with biochemical analysis, the methods are more error prone, variable and subjective.

An EM programme should detect micro-organisms in a reproducible way so as to monitor effectively the state of control in the environment. Consistent methods will yield an identification history that allows for comprehensive data comparison and interpretations. DNA sequencing provides the most consistent and unambiguous data set that is reproducible from lab to lab and over time.

Genotypic and proteotypic identification are unlike phenotypic, which can be affected by differential gene expression resulting in variable characteristics. Additionally, the DNA sequence is stable and unchanging and is a tool for identification as well as a unique descriptor for the micro-organism that can be used for tracking and trending.

strain typing

In industrial settings, such as those for dietary supplement manufacture, strain typing, or characterisation of micro-organisms, is an important part of the EM programme. Of equal importance to the probiotics and nutraceutical industries is the confirmation of specific bacterial strains used for production.

While standard genotypic and proteotypic identification methods increase the ability to identify and track and trend micro-organisms accurately and consistently at the species-level, some common microbes cannot be resolved with these approaches alone. Furthermore, strain typing is the best resource in the case of a major excursion or sterility failure where characterisation to the strain level can be critical.

Sub-species level identification or strain typing of micro-organisms, as well as discrimination between closely related species, is a challenging goal but necessary since this analysis is very important in investigating a root cause. Some of the common approaches used to differentiate closely related strains compare organisms by considering their genotypic, phenotypic, serological, spatial or temporal characteristics.

While the combination of these traits can result in sub-species level identifications, the analysis of multiple characteristics increases the time, labour and expense needed to differentiate isolates, as well as increases the errors that can arise from qualitative and subjective analyses.

Two accepted methods are used to accurately and reproducibly differentiate closely related micro-organisms: ribotyping and single- and multi-locus sequence typing (S/MLST).

Ribotyping is an automated process for characterisation of bacterial isolates and assesses the genetic relationship between strains of the same species. This fragment-based technology utilises restriction enzymes (typically EcoRI and PvuII) to target and cut regions of the ribosomal RNA genes. The DNA fragments are resolved by gel electrophoresis and chemiluminescent probes are used to visualise the banding pattern, resulting in a unique fingerprint of the bacterium.

The pattern or fingerprint is compared to others in a database and assigned to a RiboGroup for characterisation. This fingerprint can be used for tracking at the sub-species or strain level. These patterns can also be generated from different samples at different times. A historical record of the operational environment can be generated and used to determine the level of similarity or differences between the historical isolates and the current isolates as part of an investigation to track the source of contamination.


Achieve increased resolution by looking at highly variable single or multiple loci

A more powerful method is emerging in the commercial setting that has increased discrimination at the strain level. This increased resolution is achieved by looking at highly variable single or multiple loci (regions) in the genome. By combining standard genotypic identification methods, such as 16S sequence-based analysis, with multi- or single-locus sequence typing (SLST), it is possible to resolve some of the most difficult organisms to trend and track, for example, in the dietary supplement sector.

SLST and MLST are well-established, highly accurate sequence-based methods that can be used to distinguish closely-related micro-organisms, providing data that resolves differences or similarities at the sub-species or strain level. Since the foundation of S/MLST is built upon DNA sequencing results, which can be easily catalogued and referenced, these techniques are highly reproducible, unambiguous and scalable. Given the reproducibility of S/MLST between experiments, and over time, these methods can be used to determine if isolates recovered from one area are the same as or different from another isolate – a trait that allows for high-resolution trending and tracking projects.

S/MLST methods involve sequencing one to 10 target genes outside of the standard regions sequenced for microbial identification (16S, ITS2 or D2). These target genes are known to harbour moderately to highly variable DNA sequences and are protein-coding or housekeeping genes that encode for the proteins necessary for the normal cellular functions of the bacteria. By using the gene sequence, as opposed to the gene product as in enzyme electrophoresis, more variation can be detected, resulting in more potential differences per locus.

The addition of multiple loci provides added variation to further differentiate between closely related strains. After sequencing, all the sequences from the multiple gene targets are concatenated (placed end-to-end), aligned and compared to sequences from other organisms.

This comparison enables the level of divergence/conservation between micro-organisms to be calculated and displayed with an evolutionary phylogenetic tree. The goal is to determine gene combinations that can give a high level of variability to differentiate to the strain or sub-species level.

The sub-species or strain level identification makes it possible to definitively track the source of contamination or to verify production strains with certainty.

trending and tracking

New regulatory guidelines for the manufacturing of dietary supplements, nutraceuticals and other products require testing to determine the microbiological quality of the raw materials, purity of the final product and establishment of a quality programme, assuring that the manufacturing process is sufficiently controlled to prevent contamination or adulteration of the final product and production environment.1,2,3,7

Thorough, accurate and reliable identification of the microbial population in the raw materials, final product and manufacturing environment allows for rapid and definitive resolution of sterility failures, alerts and other excursions.

Data from EM sample sites should reflect operational considerations and should be proactively used to create tracking and trending reports on a routine basis, providing detailed analysis regarding the state of environmental control within the manufacturing area.


Figure 5: Banding patterns generated from automated ribotyping after digestion with EcoRI and PvuII

Figure 5: Banding patterns generated from automated ribotyping after digestion with EcoRI and PvuII

Any significant change in microbial flora should be considered in a review of the ongoing monitoring data and used in investigations of excursions to affect the mitigation process and root cause determination.6

Using the most accurate microbial identification methods while conducting gridding studies of the manufacturing plant and performing routine screenings allows reoccurring excursions of the same organism to be recognised and identified. Likewise, an increase in the number of micro-organisms recovered in certain areas of the facility may indicate breaches in the HVAC system or other sources of microbial contamination. A recognised shift in the types of organisms recovered from the area may be helpful in locating the source of contamination.

EM data can be compared against product bioburden data, and the results of the comparison can affect product development, process validation, production and an uninterrupted supply chain. Assurance of product quality through timely management and review of EM trending data can help to maintain control of the manufacturing process and facilities.

Reliable microbial identification systems are required to give consistent and accurate results for tracking and trending. Identification methods that provide inconsistent identifications or no identification for the same isolate are not useful for tracking isolates to their source nor for generating trending reports and can lead to misdirected remediation efforts. Additionally, if a health claim is made for a probiotic, the Latin name (i.e. genus, species and strain designation) of the probiotic micro-organism that is the subject of the claim should be declared.3 Thus, consistent strain typing is critical for the probiotics industry to assure that the exact strain of the micro-organism in production is established.

Genotypic methods for identification are based on the rDNA of an organism, which may exhibit but one nucleotide mutation every three million years. The DNA sequence is an unchanging descriptor of the organism and provides a superior tool for tracking and trending. By employing genotypic methods for an EM programme, sequences will be obtained for the microbial ecology, and if a name changes due to developments in taxonomy that reclassify an organism, the organism can still be tracked since the sequence will not have changed.

If a major excursion occurs, a sterility failure with a product on hold or production that ceases, a thorough characterisation of the microbial population in the manufacturing environment through strain typing is the preferred method for sourcing the contaminant. Investigative programmes that use only the genus and species name of an organism may not provide enough information to make definitive conclusions in an investigation.


Figure 6: Phylogenetic tree for P. acnes based on MLST combined gene targets

Figure 6: Phylogenetic tree for P. acnes based on MLST combined gene targets

The species name alone may not provide the appropriate evidence to develop a response to a contamination event if there is the potential for multiple strains as sources for that contamination. Many times it is necessary to differentiate microbial flora to the sub-species or strain level to definitively determine the source of contamination. Once a thorough genetic description is generated, it can contribute to determining a root cause, thus allowing remediation of the situation and determination of a corrective action plan in a timelier manner.

Assuring compliance to the regulatory guidelines for microbial monitoring during the production of dietary supplements and probiotics requires accurate identification of the organisms in the microbial environment of the production facility and in the components of the product. Accuracy of identification is key and dependent on the method used to generate and interpret the data as well as the database used as reference.

Genotypic methods are recognised as the gold standard for identification and strain typing, but the novel mass spectrometry-based technology provides a sound alternative for routine monitoring with accurate bacterial identification to the species level. When these consistent, qualified methods are combined with relevant, up-to-date, validated libraries, identification of organisms from the manufacturing environment to the species or strain level is unsurpassed.

By using dependable methods to obtain identifications, you can be certain that tracking the organisms in your environment, or the organisms that are your product, will be accurate and consistent and aid in documenting control of the production environment and leading to overall brand protection and consumer confidence.

References

1. Bacterial Library Listing and Comparison. White Paper, Accugenix (2011).

2. FDA Regulation 21 CFR Part 111 – Current Good Manufacturing Practice in Manufacturing, Packaging, Labeling, or Holding Operations for Dietary Supplements (2011).

3. Dietary Supplement Health and Education Act of 1994.

4. FAO/WHO Joint Working Group Meeting, Guidelines for the Evaluation of Probiotics in Food (2002).

5. FDA Regulation 21 CFR Part 211 – Current Good Manufacturing Practice for Finished Pharmaceuticals (2011).

6. ITS DNA Sequencing for Fungal Identification. White Paper, Accugenix (2010).

7. The AccuPRO-ID Solution from Accugenix, Inc. White Paper, Accugenix (2011).

This article is based on a technical note published by Accugenix.

You may also like