Current methods for the detection of gastroenteric pathogens are insensitive, slow, and laborious. Testing directed at specific organisms misses important infections. Systematically testing all fecal samples using multiplex PCR for common viral, bacterial and parasitic pathogens allows laboratories to increase diagnostic yield, improve workflow, reduce waste and turn-around times.

by Dr Gary McAuliffe

Introduction
Within a single diagnostic laboratory, multiple methods are used to detect gastroenteric pathogens. Selective agar plates differentiate bacterial pathogens, whereas immunoassays are used to detect viruses and parasites such as Giardia lamblia and Cryptosporidium spp.. Laboratories perform microscopy with special stains for the detection of Entamoeba histolytica and Dientamoeba fragilis. PCR has generally been restricted to the detection of norovirus, but many studies have demonstrated its potential for the detection of other enteric pathogens.

Multiplex PCR (M-PCR) panels have been shown to enhance detection of gastroenteric organisms. These panels combine several enteric pathogen targets in one or more PCR reaction vessel(s). O’ Leary et al. demonstrated 100% sensitivity compared with culture for the detection of four bacterial pathogens [1]. Wolffs et al. used a panel targeting seven viral pathogens and detected a pathogen in 97% of samples compared with 49% by conventional methods [2]. Stark et al. showed 100% sensitivity and specificity of a M-PCR panel targeting four parasites, which also reliably differentiates E. histolytica from non-pathogenic E. dispar and E. moshkovskii [3]. de Boer et al. successfully replaced bacterial culture at their institution with a molecular screening approach targeting four bacteria and G. lamblia [4].

Several commercial M-PCR fecal panels are available [Table 1]. These may be directed against parasites, viruses or bacteria, or contain targets from all three groups of organisms, allowing systematic testing of stool samples for all common gastrointestinal pathogens.   

In the author’s study, 1758 samples from community and hospital patients were tested using the Fast-Track Diagnostics bacterial, viral and parasite M-PCR panels. Pathogens were detected in 30% of samples by this systematic M-PCR approach compared with 18% using conventional testing as directed by clinician request [5][Table 2].

Advantages of a systematic M-PCR approach
Studies have demonstrated enhanced detection of entero-hemorrhagic E. coli (EHEC), Clostridium difficile, G. lamblia, E. histolytica, norovirus, adenovirus and rotavirus by PCR compared with immunoassays and other conventional assays [2–4, 6, 7]. Bacterial PCR has generally performed comparably with culture, with the exception of Salmonella spp., where reduced detection by PCR has been demonstrated in several, but not all, studies [4, 5, 8]. An advantage of increased sensitivity is that smaller amounts of feces are required for testing, and multiple samples are not required for the detection of common parasites. Clinicians should be aware that organisms such as adenovirus and norovirus can be shed in feces for several weeks following infection and that PCR detects low levels of these organisms which may not be clinically relevant. Selective bacterial media lack specificity, requiring time and further testing to discount commensal organisms. E. histolytica cysts cannot be reliably differentiated from other members of the Entamoeba complex by microscopy and staining. M-PCR generally overcomes these issues, though specificity depends on the target sequence chosen. For adenovirus, some panels target the hexon gene which does not differentiate between enteric and non-enteric serotypes, whereas others target sequences specific to enteric sub-types.

Currently laboratories restrict their testing to a limited range of pathogens for which they have sensitive and affordable assays available. A number of organisms that laboratories do not commonly test for, such as astrovirus, sapovirus, Vibrio spp., and non-O157 EHEC, can be missed. M-PCR allows a wider range of organisms to be targeted. In the author’s study astrovirus was found to be the second most common cause of gastroenteritis, and non-O157 EHECs were detected in sixteen samples by targeting the stx genes. Our laboratories did not have assays for these organisms prior to the study. Laboratories may design their own panels to reflect locally and internationally important pathogens, or buy commercial panels which reflect these requirements. Rare or imported infections not targeted by the panels will not be detected, and laboratories need to decide when additional tests are required. Up to five targets may be tested in a single real-time PCR reaction vessel; therefore increasing the number of targets reduces the number of samples on a given PCR amplification run. The xTAG system (Luminex Corp.) overcomes this limitation by post-amplification analysis using microspheres with up to 100 differing spectra. Eleven targets are currently included in the xTAG gastroenteritis panel within a single PCR reaction vessel [9]. TaqMan array cards (Life technologies) also overcome this restriction by allowing simultaneous real-time PCR in 384 PCR reaction wells. This platform allows up to eight samples to be run in parallel [10].

Fecal samples are usually tested for a limited range of viruses, bacteria, or parasites dependent upon clinician request and laboratory algorithms. Studies have shown that important pathogens such as G. lamblia and EHEC are missed using this approach. Testing for enteric viruses is not widely employed outside hospitals despite their prevalence. The M-PCR approach tests every sample for every target in the panels. It is less reliant on clinicians’ knowledge of the organisms that the patient has likely been exposed to, or those that may be causing the patients clinical syndrome. This systematic approach accounted for the majority of the increased diagnostic yield in the author’s study.

In our laboratory it can take 3–4 days for the identification of Salmonella spp. by culture. Time to generation of results is significantly reduced with the M-PCR approach.  In the author’s study, samples collected were batch tested the following day, giving results for eleven pathogens simultaneously within 24 hours of collection. Hands-on time is also reduced. The preparation and reading of a trichrome stain can take up to 40 minutes by an experienced operator, whereas the hands-on time for testing a sample by PCR is a quarter of this.

The workflow created by the systematic M-PCR approach fits well with current laboratory systems; testing is performed in a single pathway rather than by several laboratory departments. Conventional fecal testing employs multiple diagnostic kits and selective media, some of which come with significant cost, short expiry times, and extensive quality control requirements. In our laboratory stool samples for bacterial culture are inoculated onto seven agar plates which need to be stored, processed, and incubated, generating a significant amount of waste.  O’ Leary et al. reported that M-PCR significantly reduced this wastage [1]. M-PCR does not obviate the need for bacterial culture as it does not offer an antibiogram, but it allows focused testing of samples found to be positive for these bacterial targets.

Pathogens such as Shigella spp, G. lamblia and D. fragilis are labile in stool. Delays may occur in transportation, inoculation or examination which can compromise yield. For M-PCR less initial processing is required. Specimens are inoculated into tubes containing Stool Transport and Recovery buffer (STAR) (Roche Diagnostics) which binds inhibitors, and stabilizes nucleic acids for later processing. PCR is also able to detect organisms rendered non-viable by inadequate transport, or the use of antibiotics. Feces contains high levels of bilirubin and bile salts, which can lead to inhibition of PCR amplification. Inoculating samples into STAR buffer on arrival, and using extraction methods such as the EasyMag platform (Biomerieux) can help overcome this issue. In the author’s study 1.7% of samples exhibited inhibition, but all gave adequate internal control amplification following dilution and repeat testing.

Cost is a major factor preventing systematic testing of fecal samples by conventional techniques in diagnostic laboratories. The costs of PCR are reducing relative to conventional tests, and batch testing of samples using the systematic M-PCR approach is becoming a viable option. In the author’s study testing a sample for bacteria, viruses and parasites was significantly cheaper by M-PCR than using equivalent conventional tests [NZ$152 (£75) versus NZ$280 (£140)]. Laboratories have successfully reported replacing their conventional methods with a molecular screening approach for bacteria [1, 4] or viruses [2]. Where the cost of replacing all traditional diagnostics with systematic testing may currently remain restrictive, laboratories may choose to replace testing for organism groups, e.g. viruses, and institute systematic testing at a later date.

Conclusion
Current conventional methods for the detection of enteric pathogens are labour intensive, insensitive, slow, and applied piecemeal to submitted fecal samples. Testing stool samples using multiplex PCR panels which simultaneously detect all common bacteria, viruses and parasites increases the detection of gastroenteric pathogens. This approach improves turn-around time, workflow, reduces labour and waste. Costs are reducing, making systematic M-PCR testing an attractive alternative to currently used techniques.

References
1. O’Leary J, et al. Comparison of the EntericBio multiplex PCR system with routine culture for detection of bacterial enteric pathogens. J Clin Microbiol. 2009; 47: 3449–3453.
2. Wolffs PF, et al. Replacing traditional diagnostics of fecal viral pathogens by a comprehensive panel of real-time PCRs. J Clin Microbiol. 2011; 49: 1926–1931.
3. Stark D, et al. Evaluation of multiplex tandem real-time PCR for detection of Cryptosporidium spp., Dientamoeba fragilis, Entamoeba histolytica, and Giardia intestinalis in clinical stool samples. J Clin Microbiol. 2011; 49: 257–262.
4. de Boer RF, et al. Improved detection of five major gastrointestinal pathogens by use of a molecular screening approach. J Clin Microbiol. 2010; 48: 4140–4146.
5. McAuliffe GN, et al. Systematic application of multiplex PCR enhances the detection of bacteria, parasites, and viruses in stool samples. J Infect. 2013; 67(2): 122–129.
6. Costantini V, et al. Diagnostic accuracy and analytical sensitivity of IDEIA norovirus assay for routine screening of human norovirus. J Clin Microbiol. 2010; 48: 2770–2778.
7. Luna RA, et al. Rapid stool-based diagnosis of Clostridium difficile infection by real-time PCR in a children’s hospital. J Clin Microbiol. 2011; 49: 851–857.
8. Cunningham SA, et al. Three-hour molecular detection of Campylobacter, Salmonella, Yersinia, and Shigella species in feces with accuracy as high as that of culture. J Clin Microbiol. 2010; 48:2929–2933.
9. Coste JF, et al. Microbiological diagnosis of severe diarrhea in kidney transplant recipients by use of multiplex PCR assays. J Clin Microbiol. 2013; 51(6): 1841–1849.
10. Liu J, et al. A laboratory developed TaqMan array card for simultaneous detection of nineteen enteropathogens. J Clin Microbiol. 2013; 51(2): 472–480.

The author
Gary McAuliffe MBBS
Microbiology Department, LabPlus Laboratory, Auckland, New Zealand
E-mail: GMcAuliffe@adhb.govt.nz

Around 80% of people in Western countries experience low back pain at some point in their life; indeed during a single year up to half of the adult population will experience back pain. In the majority no single clear cause can be identified, and the condition is self limiting. However for the approximately 7% of patients who develop chronic low back pain, quality of life can be significantly impaired. Chronic low back pain also has a serious financial impact in terms of healthcare costs and lost working days; in most industrialized countries it is the most common reason for workplace absence.
Now researchers at the Spine Centre of Southern Denmark have published the results of two very interesting and potentially far-reaching studies. The first involved sixty-one patients who had been suffering from low back pain for more than six months. Lumbar disc herniation was confirmed by MRI and the patients underwent primary surgery for removal of nucleus material. When this material was cultured for micro-organisms, forty-six percent of patients had positive cultures, predominantly with the normally commensal anaerobic bacterium Propionibacterium acnes. A significantly higher number of patients with anaerobic infections developed new Modic changes (MC), MRI-visible bone edema associated with low back pain, in vertebrae adjacent to the previous disc herniation compared with patients with negative cultures or cultures positive for aerobic organisms.
The second study was a double blind randomized controlled trial involving 162 patients who had been suffering from low back pain for more than six months and with MC in a disc adjacent to a previous disc herniation. Patients received either placebo or the antibiotic Bioclavid for a hundred days and were followed up at the end of treatment and after one year; outcome measures included both pain and workplace absence. Improvement was highly significant in the group treated with the antibiotic.
The authors suggest that when the lumbar disc is herniated, the anaerobic bacteria penetrate it and precipitate an insidious infection and chronic low back pain. Although they stress that antiseptic techniques were rigorously followed when the nucleus material was removed, it is surely still necessary to find a method of demonstrating anaerobic infection in patients who have low back pain and relevant MC but who have not had surgery. If this could be done many desperate chronic lower back pain sufferers might finally be able to stop taking analgesics or visiting osteopaths, chiropractors and acupuncturists, and get relief from a course of antibiotics instead.

Current methods for detecting noroviruses (NoVs) have significant limitations in sensitivity and feasibility for use in remote locations. Our group recently identified phage-displayed peptides with specific binding to NoVs and sensitivity comparable to that of existing antibodies. These reagents can be easily optimized by mutagenesis and represent promising diagnostic tools.

by Amy M. Hurwitz, Prof. Robert L. Atmar and Prof. Timothy G. Palzkill

Norovirus infection and diagnosis
Each year, norovirus (NoV) infections cause approximately 267 million new cases of gastroenteritis and 200,000 deaths worldwide [1]. Infection spreads rapidly in areas of close human contact, such as cruise ships and hospitals, and is treated only by rehydration, as no antiviral therapy currently exists. An infectious dose estimated to be as low as 18 virions and high environmental stability contributed to classification of NoVs as a category B biodefense agent in the U.S. Therefore, rapid, accurate and highly sensitive diagnosis is important for outbreak recognition and control, and also to guide physicians in patient management. The potential health and economic consequences that may be ameliorated by early NoV detection have led to a high demand for optimized detection reagents that can be used to develop reliable diagnostic assays with minimal requirements for expensive, bulky equipment or technical training.

NoVs are divided into six different genogroups (GI–GVI) based on the amino acid sequence of the major capsid protein (VP1). These are organized further into more than 30 genotypes, and finally into numerous strains or variants [2]. The VP1 protein assembles to form an icosahedral shell with an inner shell (S) domain and outer protruding (P) domain. The P domain is on the virus surface and is the most accessible, while the S domain has the highest sequence conservation across different strains. Given the ability of NoVs to evolve rapidly to result in novel or recombinant strains, continual optimization of detection reagents may be necessary in order to recognize the majority of human-infecting strains. Strains classified into GI and GII are most relevant for human infections, and thus the focus for diagnostic assay development efforts.

Current diagnostic methods and their limitations
Methods used currently for the diagnosis of norovirus infection are far from ideal as they exhibit several limitations that hinder their use for individual patient diagnoses or in rural and developing locations. The gold standard for diagnosis is reverse transcriptase (RT)-PCR, which requires multiple sets of primers to detect about 90% of human-infecting strains [3]. This method has significant equipment and expertise requirements, which are often not available outside of large institutions. Further, the expense of running multiple samples and the need for timely instrument accessibility limit the feasibility of applying RT-PCR as point-of-care applications or for preventing the rapid spread of an outbreak.

Other existing methods include immune electron microscopy (IEM) and enzyme immunoassays. IEM was the first method described for identifying NoVs and was used originally to classify viruses based on structural appearance. This method has limited sensitivity, and also requires expensive equipment and skilled expertise. Enzyme immunoassays, developed after the discovery of type-specific antibody epitopes on the NoV capsid, detect viral particles in human stool samples [4]. This method offers increased specificity and has led to the development of commercially available ELISA and lateral flow assays.

Currently, the only FDA-approved antigen detection assay is an ELISA called RIDASCREEN® (3rd Generation) produced by R-Biopharm, which uses an antibody cocktail with specificity for GI and GII NoVs [Fig. 1]. Due to limitations in sensitivity, this assay is only approved for use during outbreaks and takes several hours to produce results. Several companies, including R-Biopharm, have developed rapid diagnostic assays that use lateral flow technology and have also demonstrated strong specificity for NoV GI and GII strains. However, these have similar limitations with sensitivity and thus are only recommended for preliminary screening to be confirmed by RT-PCR, and are distributed primarily outside of the United States [5]. Overall, there is a clear need for improved diagnostic methods to detect norovirus rapidly with strong specificity, high sensitivity, and with minimal equipment and expertise requirements.

Novel diagnostic phage reagents
Recent studies in our laboratory have identified short, 12-mer peptide reagents with specific binding to the GI.1 NoV genotype [6]. The small size of these peptides displayed on phages offers the ability to access epitopes that may be buried in the capsid protein and not accessible to antibodies, and the potential for increased avidity through multiple linked peptide molecules. To identify peptides with specific binding to NoV, we used phage display technology to screen commercially available, large-scale libraries of randomized peptides that are fused to the gene III protein and expressed in five copies on one end of the phage. Rounds of biopanning were performed in which filamentous phage libraries were screened for phages displaying peptides that bind immobilized Norwalk (NV) GI.1 virus-like particles (VLPs). The phage libraries were added to VLPs and, after washing away non-binding phages, the phages displaying VLP-binding peptides were eluted with low pH [Fig. 2A]. Two to four subsequent rounds of biopanning using the resulting phage populations enriched for phages displaying peptides with the highest binding affinity for NV. DNA sequencing of individual phage clones recovered after multiple rounds of biopanning revealed three peptides, named NV-O-R5-3, NV-O-R5-6, and NV-N-R5-1, that occurred most commonly, and the phage clones displaying the peptides were further characterized for their NoV binding properties [6].

Phage-based ELISAs confirmed the binding specificity of phage-displayed peptides to NV VLPs. These affinity-binding assays used NV VLP captured by immobilized rabbit polyclonal anti-NV antibody in order to maintain the structural integrity of VLPs. Single phage clones were added to the captured VLPs and binding was detected using anti-M13 phage antibody that was conjugated to horseradish peroxidase to provide a signal for bound antibody [Fig. 2B]. Of the three peptide-displaying phage clones analysed, NV-N-R5-1 exhibited a dose-dependent response with decreasing NV VLP concentration and the highest sensitivity with a limit of detection at 1.56 ng NV VLP. Additional phage ELISAs indicated that NV-N-R5-1 binds to the P domain of the capsid protein, which extends the furthest out from the virus, and has comparable sensitivity for NV as existing antibodies used for diagnostics [6]. These results provide proof-of-concept and a strong lead reagent for developing novel phages displaying peptides as effective detection reagents for NoV. Further, the methods described establish a platform methodology for using phage display to identify antigen-specific binding reagents that may be applied to any pathogen with distinct surface epitopes.

Current status
To develop our lead phage-displayed peptide into a commercially viable tool, we are currently optimizing its binding affinity for other genogroups of NoV in order to broaden its diagnostic applications. Phage display technology provides a simple platform for constructing collections of new mutations in a lead peptide that can be used for additional rounds of biopanning to screen for variants with optimal affinity properties [Fig. 2C]. The three phage-displayed peptides discussed above share conserved amino acid sequence motifs that likely confer binding specificity for particular epitopes on the NV capsid protein. Directed evolution through mutagenesis of amino acids surrounding these consensus sequences can enable us to improve binding affinity to NV and alter binding specificities starting with the lead phage peptide, NV-N-R5-1. In particular, developing phage-displayed peptides with optimized binding affinity for the NoV GII.4 genotype, which accounts for >80% of NoV infections worldwide [1], and other GI and GII NoV genotypes will have the greatest relevance for diagnostic applications.

Future development of bacteriophage reagents
For decades, phages have been used to identify their target bacterial strains and species in order to diagnose the cause of infections by phage typing. More recent applications have begun to leverage synthetic biology and genomic engineering strategies to customize phage specificity and reporter signals to enable ‘near-real-time’ detection of a broader range of human pathogens [7]. Our recent work has established a methodology for the identification, characterization, and development of phage-based affinity reagents that may be applied to different pathogens and translated into diagnostic applications. The process outlined in Figure 2 demonstrates the progression from (A) identifying lead reagents against a target of interest, (B) characterizing binding affinity for the antigenic target, (C) optimizing leads through directed evolution or genomic engineering strategies, and finally (D) producing scalable quantities of reagent for commercial diagnostic applications. Zou and colleagues, for example, used a similar method to identify a phage-displayed peptide reagent with specific binding to transmittable gastroenteritis virus (TGEV) that also showed potential antiviral activity [8]. Several groups have also developed phage-based reagents to detect bacterial pathogens, such as Salmonella enterica and Escherichia coli [9, 10].
In summary, the use of phage-based reagents for microbial diagnostics offers many advantages in comparison to more commonly used detection reagents, such as antibodies. Phage display technology enables rapid identification and validation of candidate phage reagents with specificity for new or evolved pathogens through biopanning of commercial or custom made phage libraries (Fig. 2A, B). Phage manipulation through directed evolution facilitates development of reagents with optimized binding affinity and specificity to a target of interest (Fig. 2B). Finally, production of large quantities of phages is accomplished rapidly and inexpensively, as simple preparation methods can produce sufficient phage for hundreds of assays (Fig. 2D). As viral pathogens such as NoV continually evolve, the flexibility provided by phage-based reagents will be essential for developing next generation diagnostics for effective containment of outbreaks. A cocktail of phages, each of which binds to a specific target NoV genotype, may ultimately be the ideal strategy for producing an assay to detect the broadest possible range of NoVs without sacrificing specificity. Overall, phages have an enormous potential for use as detection reagents in clinical, agricultural, food, and environmental settings, and represent an underutilized resource for diagnostic development.

References
1. Donaldson EF, Lindesmith LC, Lobue AD, Baric RS. Norovirus pathogenesis: mechanisms of persistence and immune evasion in human populations. Immunological Reviews 2008; 225(1): 190–211.
2. Kroneman A, Vega E, Vennema H, Vinjé J, White P, Hansman G, Green K, Martella V, Katayama K, Koopmans M. Proposal for a unified norovirus nomenclature and genotyping. Archives of Virology 2013; doi:10.1007/s00705-013-1708-5.
3. Atmar RL, Estes MK. The epidemiologic and clinical importance of norovirus infection. Gastroenterology Clinics of North America 2006; 35(2): 275–290.
4. Parker TD, Kitamoto N, Tanaka T, Hutson AM, Estes, MK. Identification of Genogroup I and Genogroup II broadly reactive epitopes on the norovirus capsid. Journal of Virology 2005; 79(12): 7402–7409.
5. Ambert-Balay K, Pothier P. Evaluation of 4 immunochromatographic tests for rapid detection of norovirus in faecal samples. Journal of Clinical Virology 2013; 56(3): 194–198.
6. Rogers JD, Ajami NJ, Fryszczyn BG, Estes MK, Atmar RL, Palzkill TG. Identification and characterization of a peptide affinity reagent for detection of noroviruses in clinical samples. Journal of Clinical Microbiology 2013; 51(6): 1803–1808.
7. Lu TK, Bowers J, Koeris MS. Advancing bacteriophage-based microbial diagnostics with synthetic biology. Trends in Biotechnology 2013; 31(6): 325–327.
8. Zou H, Zarlenga DS, Sestak K, Suo S, Ren X. Transmissible gastroenteritis virus: Identification of M protein-binding peptide ligands with antiviral and diagnostic potential. Antiviral Research 99(3): 383–390.
9. Schofield DA, Sharp NJ, Westwater C. Phage-based platforms for the clinical detection of human bacterial pathogens. Bacteriophage 2012; 2(2): 105–283.
10. Galikowska E, Kunikowska D, Tokarska-Pietrzak E, Dziadziuszko H, Loś JM, Golec P, Węgrzyn G, Loś M. Specific detection of Salmonella enterica and Escherichia coli strains by using ELISA with bacteriophages as recognition agents. European Journal of Clinical microbiology & Infectious Diseases 2011; 30(9): 1067–1073.

The authors
Amy M. Hurwitz¹ BS, Robert L. Atmar^2,3 MD, Timothy G. Palzkill*^2,4 PhD
¹ Interdepartmental Graduate Program in Translational Biology and Molecular Medicine, Baylor College of Medicine, Houston, Texas, USA
² Department of Molecular Virology and Microbiology, Baylor College of Medicine, Houston, Texas, USA
³ Department of Medicine, Baylor College of Medicine, Houston, Texas, USA
⁴ Department of Pharmacology, Baylor College of Medicine, Houston, Texas, USA
 
*Corresponding author
E-mail: timothyp@bcm.edu