Introduction
Last year marked the 25th anniversary of the first description of the Polymerase Chain Reaction (PCR), a process used to make a large number of copies of a specific segment of the genetic material (DNA) in a relatively short period of time. Kary Mullis, a DNA chemist at Cetus Corp. in Emeryville, California, invented the PCR while running some research projects on oligonucleotide synthesis [1]. At the time, the existing methods to copy oligonucleotide polymers were fastidious and timeconsuming. The new PCR process constituted a major breakthrough because it solved the problem of how to produce multiple copies of any particular piece of DNA using a relatively simple, economical, and reliable procedure. The technique allowed the generation of million of copies of a particular DNA target from a single copy of DNA in just a few hours. It was recognized almost immediately as a revolutionary tool applicable to research and applied projects in molecular genetics and biotechnology, respectively. Not surprisingly, the Royal Swedish Academy of Sciences awarded one half of the 1993 Nobel Prize in Chemistry to Mullis for his invention of the PCR method [2]. As expressed in the 1993 Academy’s press release, Kary Mullis’ discovery had “hastened the rapid development of genetic engineering” and “greatly stimulated biochemical” and genetics “research and opened the way for new applications in medicine and biotechnology”.
The PCR has been applied to a myriad of research projects since then, encompassing practically every single aspect of molecular biology studies. PCR technology has also been extensively and increasingly used in microbiology, and numerous PCR-based assays and instrumentation were developed for and are currently used by a number of industries and fields, from bio-defense, cosmetics, and forensics, to beverages and food, and clinical diagnostics. Yet, the application of this powerful technology in the pharmaceutical arena has been elusive. Why? This paper attempts to summarize some of the main obstacles PCR, and in more general terms nucleic acid amplification-based technology (NAAT) had, and still is encountering in the pharmaceutical quality control (QC) microbiology area, analyzing the differences between the QC goals of this industry compared to other industries where PCR assays are already established and accepted as a firm alternative microbiological method.
Nucleic Acid Amplification-based Technologies (NAAT)
PCR and Real Time qPCR: Since the original description of PCR as a method to amplify DNA, numerous variations of the technology have been described. One of them is real-time quantitative PCR, more commonly known as real time qPCR [3]. This technique is potentially applicable to microbial enumeration in Quality Control (QC) microbiology. With real time qPCR, the exponential phase of PCR is monitored continuously during 30-50 cycles of amplification, allowing to estimating the initial amount of target DNA. Moreover, since the amount of DNA detected by qPCR is directly proportional to the number of microbial cells present originally in a particular sample, a correlation curve between DNA molecules and number of cells can be made, letting the result to be expressed as number of microorganisms per unit of mass or volume of the original sample. Unless the microorganism present in the sample is known, and a correlation curve for that culture already exists, the result is actually an estimate as there are variations in the amount of DNA per cell between different genera and even within one species depending on the metabolic state of the cells at the time of the analysis.
Though initial nucleic acid amplification based tests were developed using the well known polymerase chain reaction as a template, development of other techniques followed.
LCR and SDA. The ligase chain reaction (LCR) also detects DNA and is based on the ligation of microorganism-specific oligonucleotides probes which serve as a copy of the original target sequence and which are adjacent to each other. Both methods PCR and LCR, were shown to have high sensitivity (95 to 99%) in detecting chlamydial and M. tuberculosis-specific nucleic acids in clinical samples [4-5]. An alternative technique successfully applied to the diagnostic of Chlamydia trachomatis and Neisseria gonorrhoeae amplified DNA using an isothermal Strand Displacement Amplification (SDA) method coupled with fluorescent energy transfer (ET) measurement to detect amplified product as it is produced during the reaction [6]. The SDA assay consists of 3 steps: 1) The target double stranded DNA sample is heat denatured in the presence of primers and reagents, 2) restriction enzyme HincII and exonuclease-deficient large fragment of E. coli DNA polymerase (Klenow fragment) are added, and 3) the sample is incubated at 37oC. More than 107-fold amplification of a genomic sequence can be achieved in 2-3 hours of isothermal incubation.
One of the concerns raised by the use of NAAT in general, is the possibility of given false negative results by failure to detect a specific target sequence (see Table 1). The interesting work of Gilpin et al [7] describes a case of smear-positive pulmonary tuberculosis patient that had a false-negative result using the LCR technique. It was shown to be due to a deletion of the target region in the 38-kDa protein antigen B gene of the Mycobacterium tuberculosis strain found in this particular patient [7]. Further development of this type of test should include the use of sequences that are essential for survival as a target to detect the specific microbe under investigation (Table 1).
RT-PCR. The conversion of RNA into a complementary copy of DNA (cDNA) by means of the enzyme Reverse Transcriptase (RT) followed by a PCR is one of most powerful techniques (RT-PCR) to detect RNA molecules from virtually any type of living cell. As with PCR, the RT-PCR can also be analyzed in real time and used as either a qualitative or a quantitative tool [3, 8]. A deoxyribonuclease (DNase) pretreatment has been shown very effective to eliminate DNA contamination when applied prior to RT-PCR analysis, resolving one of the concerns related to this technique (Table 1).
NASBA and TMA. Nucleic Acid Sequence Based Amplification (NASBA) and Transcription Mediated Amplification (TMA) are very similar transcription-based, amplification methods that amplify RNA from a RNA target. The RNA amplification process by NASBA was modeled on the enzymes essential for retrovirus replication, including ribonuclease H, reverse transcriptase and a DNA-dependent RNA polymerase to produce first cDNA and then copies of the original RNA target [9]. These products then function as templates for a series of transcription and reverse transcription reactions that are self sustained, amplifying the RNA target millions of time in a relatively short time. NASBA depends on selective primer-template recognition to drive a cyclical, exponential amplification of the target sequence. It has been extensively used for the detection of viruses and specific bacterial pathogens in clinical and food samples [10]. Real-Time TMA is very similar to the NASBA technique, but instead of using an independent RNaseH activity, the reaction has been optimized in order to activate the RNAse H activities of the other enzymes present in the assay [11]. Similarly as in the NASBA technique, each of the newly synthesized RNA amplicon re-enters the TMA process and serves as a template for a new round of replication leading to an exponential expansion of the RNA amplicon. Unlike reverse transcriptase PCR, both NASBA and TMA are isothermal (approx 41oC); therefore, there is no need for thermal cycling. The product in both NASBA and TMA is an antisense single stranded RNA, so the high temperature denaturation of double stranded DNA characteristic of RT-PCR is not required. Unlike RT-PCR, NASBA and TMA are able to selectively amplify RNA sequences in a DNA background. In NASBA and TMA compared with RT-PCR, there is an exponential rather than a binary increase in amplified product, resulting, in theory, in much greater amplification. However, in practice most of the reports in the literature show that all three techniques have equivalent sensitivity for bacterial cells, in the order of 10-100 colony forming units (CFU).
By now, all the precedent techniques have been applied extensively in molecular biology, genetics, clinical research, and related applications. At the industrial level, PCR and many other NAAT are extensively used in genetic engineering and biotechnology, and are also increasingly applied in the detection of specified bio-weapons (bio-defense), forensics, archeology, clinical microbiology (molecular diagnostics and identification) for the detection of specific pathogens, and as quality control rapid tools in the cosmetics, beverages, and food industries.
NAAT in the Pharmaceutical Industry
In spite of all these applications in a wide range of fields and industries, NAATs have not been readily adopted yet in the microbiology quality control (QC) area of the pharmaceutical industry. Moreover, why has not there been a single regulatory submission related to this RMM technology thus far? The following discussion will try to provide an explanation to this, by addressing some of the main technical challenges and concerns that the rapid NAAT still present for microbiological applications in the pharmaceutical industry (Table 1). The discussion will not address all of these concerns in detail, neither other concerns of a broader nature, comprising Rapid Microbiological Methods in general, that have been previously addressed by others [12].
Table 1 summarizes some of the main challenges and concerns regarding applicability of the NAA technologies in the pharmaceutical industry. Some of these points are discussed below.
Universality
The microbiology QC of pharmaceutical products and manufacturing environments require in most of the main assays (e.g., sterility testing) the ability to detect all possible microorganisms potentially contaminating any particular sample. This is a requirement that drastically differs from those found in other industries. For example in the food and beverage field many of the required assays are developed to demonstrate that certain microbial pathogens are not present in the samples tested. Interestingly, a large number of the NAA technologies and kits have been primarily developed for applications in these industries. Why? According to a study from Strategic Consulting Inc (SCI), Food processing companies perform half of all of the microbiology testing in the global industrial market, and it is anticipated that the market for food microbiology testing will be worth $2.4bn by 2013 [13]. The surge in food micro testing is being driven by the key factors of an increase in food production, increased regulation, growing demand from retailers as well as the industry’s food safety priorities. All these trends are helping to accelerate the conversion of traditional microbiological testing methods to more rapid versions despite their higher cost per test.
The precedent helps to understand why most of the kits and platforms available in the market are better fitted to detect specific microbial species rather than to cover a more universal approach for microbial contaminant detection, which is, as mentioned above, the main requirement in the pharmaceutical industry. The list of PCR and NAAT kits commercially available for the detection of specific microorganisms for the QC of food, beverages, and cosmetics and also applicable in bio-defense, and clinical microbiology is enormous. The following are just some examples of commercially available PCR kits for the detection of specific microorganisms:
• Bacillus anthracis
• Beer-spoilage bacteria & ID for 5 species
• Campylobacter
• Coliforms
• E. coli O157:H7
• Enterobacteriaceae
• Enterobacter sakazakii
• Listeria
• Salmonella
• Shigella
• Pseudomonas aeruginosa
• Staphylococcus aureus
• Vibrio
• Yeast & Mold
• vancomycin resistance genes (vanA/vanB) in Enterococcus spp
• Mycobacterium tuberculosis, avium, intracellulari
• Chlamydia trachomatis and/or Neisseria gonorrhoeae
• Cronobacter sakazakiii
• Salmonella enterica
Detection of microbial contaminants on a universal basis is much more complicated. There are a few commercially available platforms that target rRNA using universal sequences and real time RT-PCR [8] or TMA [11]. Though studies are usually available from the vendors covering the detection of microbial species listed in the pharmacopoeia, more extensive validation studies are still needed, including a large number of both common and rare pharmaceutical manufacturing environmental isolates to define the universality of these technologies in detecting a large variety of microbial contaminants and the equivalency to the traditional microbiological methods.
The list of validation parameters needed for the implementation of a RMM have been well described in a Parenteral Drug Association report [14], and two pharmacopoeia chapters (USP chapter 1233 and Ph Eur 5.1.6). One of these validation parameters, specificity, is defined as the ability of an alternative method to detect a range of microorganisms that may be present in the test article. This parameter is arguably one of the most difficult to validate when using NAAT, particularly in those cases where an enrichment phase is not used, and therefore, the assay does not require growth as an indicator of microbial presence. There are some concerns related to this topic, such as:
• Are the primers “universal” enough to detect all of the potential microbial contaminants currently detected by the traditional method?
• Similarly, is the nucleic acid detection assay (e.g., PCR, RTPCR, NASBA, TMA, etc), “universally” applicable?
• What about the cell harvesting, cell lysing, and nucleic acid preparation steps? Are they applicable to a broad range of microbial species, including “compendial” strains, and a long and diverse list of environmental isolates, and rare cultures?
Limit of Detection (LOD)
The LOD of the traditional growth assay (agar plate or liquid medium) is very good (1 CFU, under the conditions of the test) and well accepted by the pharmaceutical industry and the regulatory agencies. But, a LOD of 1 CFU is still a challenge for some of the NAAT, particularly, those relying on RNA and universal primers. The LOD for these technologies (e.g., RT-PCR, TMA, NASBA) varies between 10- 100 CFU depending on the bacterial species under analysis. Similar as with other areas of concerns related to NAAT, further development in instrumentation, reaction optimization, and other protocol improvements are still needed to overcome this shortcoming.
Non-Culturable and Viable Microorganisms
The ability to discriminate metabolically active from dormant cells, and viable from dead cells constitute probably the most important concerns related to the use of NAAT in pharmaceutical microbiology.
DNA has been shown to be highly persistent after cell death, therefore, leading to false positives if detected from dead bacterial cells present in microbiological samples. In addition, DNA is also going to be detected from cells that entered a dormant state as a result of an exposure to an unfavorable environment. These cultures that were partially damaged or stressed by the harsh environment, and are unable to grow under the conditions of the traditional microbiology, are known as viable but non-culturable (VBNC). Despite these concerns, DNA is undoubtedly an excellent marker for the detection and quantification of specific bacterial pathogens because its robustness after extraction and purification helps to avoid false negative results. The latter advantage has prompted a myriad of efforts to overcome the concerns discussed above [15]. Among the techniques developed for that purpose, pretreatment of samples with Ethidium MonoAzide (EMA) and Propidium MonoAzide (PMA) has been shown to be an effective tool to distinguish viable from damaged cells in some bacterial species [16-17]. EMA and PMA are DNA intercalating dyes that enter bacteria with damaged cell membranes. The dyes can then be covalently linked to DNA by a photoactivation step preventing PCR amplification. However, the method is currently limited to the detection of specific pathogens. The use of these reagents on a more general basis it will require a significant validation effort, because its effectiveness is known to vary among related cultures, and it has not been demonstrated yet to be universally applicable to different microbial genera.
Messenger RNA is much more labile than DNA and the presence of microbial mRNA is indicative of ongoing bacterial transcriptional activity, an indication of microbial viability. However, mRNAs sequences are in most of the cases not universally represented in the microbial world, and low copy number in many transcripts could affect significantly sensitivity of any specific assay. Some studies have shown that some transcripts could be detected from dead cells with one method (NASBA or RT-PCR) whereas the other one did not detect it [18].
Ribosomal RNAs are recognized as better targets. Most of the rRNA sequences are universally represented in all bacterial species, and the fact that the number of copies per bacterial cell amounts to several thousands makes any detection assay much more sensitive. However, rRNAs have intrinsically long stretches of secondary structure that confer increased stability. Therefore, depending on the targeted molecule and region, differentiating between live and dead cells could be potentially difficult. A solution to this problem is still awaiting elucidation.
Background DNA
False positives due to the presence of genomic DNA carried over from the RNA preparation is a main technical concern in RT-PCR assays. Though a DNase step during the RNA purification step can be used to prevent this problem, a careful validation analysis is still highly recommended to prove that DNA contamination is absent under routine assay conditions. In addition, many of the enzymes used in NAAT are contaminated with minute amounts of bacterial DNA originated in the recombinant cultures (e.g., E. coli) used to produce these enzymes. This tiny amount of contaminant can contribute significantly to the background of the PCR, particularly if universal primers are used, diminishing the sensitivity of the assay at a low count range.
Cross Contamination and Carry Over
False Positives: The high sensitivity of nucleic acid amplification tests makes these assays susceptible to contamination that can cause false-positive results. One of the major hurdles to overcome in achieving full automation of nucleic acid amplification tests is to prevent even minor nucleic acid contaminations in the instrumentation which can result in false positives. The two major types of contamination are specimen cross-contamination and carryover contamination.
Specific laboratory protocols must be followed, including a special training program dedicated to minimize contamination and/ or amplicon carry-over. These include separate work areas for reagent preparation, sample extraction, and nucleic acid amplification with dedicated equipment and supplies for each area, practices to minimize aerosolization, unidirectional workflow, and the use of internal amplification standard controls in each test. To minimize variability and some of above mentioned sources of contamination, reactions should be set up using automated platforms located inside a filter sterilized air laminar flow chamber.
False Negatives: Nucleic acid amplification assays are theoretically capable of detecting as little an amount as one organism in a sample. However, in practice, this sensitivity is rarely achieved. Sample inhibition can hinder the amplification reaction and result in false-negative results. Sample inhibition occurs when factors in some samples inhibit the enzyme activities needed for the amplification reaction. Validation protocols should include special sections to deal with both false positive and negative results.
Conclusion
Challenges and concerns for the application of NAAT as rapid methods to replace some of the traditional microbiological QC tests of the pharmaceutical industry are still valid. Some of these points were summarized in Table 1. It is the hope of the author that this summary will serve as a guideline for future technical refinements in order to overcome these obstacles. Further improvements in enzymology, chemistry and reaction optimization, instrumentation, and a more comprehensive validation work are still needed to attain a broad applicability of NAAT in the QC microbiology labs of the pharmaceutical industry.
References
1. Mullis, K.B. and Faloona, F.A. “Specific synthesis of DNA in vitro via a polymerase-catalyzed chain reaction” (1987) Methods Enzymol. 155:335-50.
2. www.nobelprize.org Details on the Nobel prize for the PCR discovery.
3. Bustin, S.A. “Quantification of mRNA using real-time reverse transcription PCR (RT-PCR): trends and problems” (2002) J. Molec. Endocrinology 29:23-39
4. Black, C. M., Marrazzo, J., Johnson, R. E., Hook, E. W. 3rd., Jones, R. B., Green, T. A., Schachter, J., Stamm, W. E. , Bolan, G., St Louis, M. E. & Martin, D. H. “Head-to-head multicenter comparison of DNA probe and nucleic acid amplification tests for Chlamydia trachomatis infection in women performed with an improved reference standard” (2002) Journal of Clinical Microbiology 40, 3757 - 3763.
5. O’Connor, T., Sheehan, S., Cryan, B., Brennan, N., and Bredin, C. “The ligase chain reaction as a primary screening tool for the detection of culture positive tuberculosis” (2000) Thorax 55(11): 955–957. 6. Walker, G.T., M. S. Fraiser, J. L. Schram, M. C. Little, J. G. Nadeau, and D. P. Malinowski. “Strand displacement amplification--an isothermal, in vitro DNA amplification technique” (1992) Nucleic Acids Res. 20(7):1691-1696.
7. Gilpin, C.M., Dawson, D.J., O’Kane. G., Armstrong, J.G., and Coulter, C. “Failure of Commercial Ligase Chain Reaction To Detect Mycobacterium tuberculosis DNA in Sputum Samples from a Patient with Smear-Positive Pulmonary Tuberculosis Due to a Deletion of the Target Region” (2002) J Clin Microbiol., 40(6): 2305–2307.
8. Weis, J.H., S.S. Tan, B.K. Martin, and C.T. Wittwer. “Detection of rare mRNAs via quantitative RT-PCR“ (1992) Trends in Genetics 8:263-264.
9. Fox, J.D., C. Moore, and D. Westmoreland. “Real time NASBA” (2009) in Chapter 12, from Real-Time PCR: Current Technology and Applications (Eds J. Logan, K. Edwards, and N. Saunders), Caister Academic Press, Norwich, UK.
10. Cook, N. “The use of NASBA for the detection of microbial pathogens in food and environmental samples. (2003) J. Microbial. Methods, 53(2):165-174.
11. Stary, A., E. Schuh, M. Kerschbaumer, B. Götz, and Helen Lee. “Performance of Transcription-Mediated Amplification and Ligase Chain Reaction Assays for Detection of Chlamydial Infection in Urogenital Samples Obtained by Invasive and Noninvasive Methods” (1998) J. Clin. Microbiol. 36(9):2666-2670.
12. Middleton, A.M. “Rapid Microbiological Methods: Are the needs of the pharmaceutical industry really being met?” (2007) American Pharm Rev., 10(6):108-113.
13. www.foodproductiondaily.com (14/Nov/2008). Report of the SCI study
14. Parenteral Drug Association (PDA) Technical Report. Evaluation, validation and implementation of new microbiological testing methods. PDA J Pharm Sci & Technol 2000; 54(3), suppl. TR33:2-9.
15. Keer, J.T. and Birch L. Molecular Methods for the assessment of bacterial viability. J. Microbiol. Methods, 2003. 53(2):175-183.
16. Cawthorn, D.M., and Witthuhn, R.C. “Selective PCR detection of viable Enterobacter sakazakii cells utilizing propidium monazide or ethidium bromide monoazide” (2008) J. Appl. Microbiol. 105(4):1178-1185.
17. Lee, J.L., and Levin, R.E. “A comparative study of the ability of EMA and PMA to distinguish viable from heat killed mixed bacterial flora from fish fillets” (2009) J. Microbiol. Methods 76(1):93-96.
18. Birch, L., Dawson, C. E., Cornett, J. H. & Keer, J. T. “A comparison of nucleic acid amplification techniques for the assessment of bacterial viability”. (2001) Letters in Applied Microbiology 33, 296 – 301.
To contact the author please, email him directly at: [email protected]
Claudio Denoya, Ph.D., is a Research Fellow and Group Leader of the Microbiological Technical Assessment group at the Parenteral Center of Emphasis, Pfizer Global R&D, Groton, CT, working on the implementation of rapid microbiological methods (RMM). He is a member of the Pfizer Global RMM Application Development Team. He is also an Adjunct Professor at the Department of Molecular and Cell Biology, Univ. of Connecticut. Claudio led the Streptomyces genetic engineering group that delivered the recombinant culture producer of the antiparasite product Dectomax. He also held several other positions at Pfizer, such as (1) Leader of the Bioprocess Recombinant DNA group achieving expression of therapeutic proteins in bacteria and large-scale manufacturing of DNA for gene therapy; (2) Leader of the Eukaryotic Transfection & Expression group accomplishing the development of several recombinant mammalian cell lines currently used for scale up production of biologics in support of advanced clinical studies; and (3) Leader of the Biologics & Microbiological Testing group in Supply Chain-Analytical Control. He has received several recognitions, including the Pfizer Global R&D Achievement Award, two Supply Chain Recognition Awards, and the National Hispanic Corporate Achiever Award. He has more than 60 patents, book chapters, and journal articles.