ABSTRACT
Human noroviruses are the leading cause of epidemic and sporadic gastroenteritis across all age groups. Although the disease is usually self-limiting, in the United States norovirus gastroenteritis causes an estimated 56,000 to 71,000 hospitalizations and 570 to 800 deaths each year. This minireview describes the latest data on laboratory methods (molecular, immunological) for norovirus detection, including real-time reverse transcription-quantitative PCR (RT-qPCR) and commercially available immunological assays as well as the latest FDA-cleared multi-gastrointestinal-pathogen platforms. In addition, an overview is provided on the latest nomenclature and molecular epidemiology of human noroviruses.
INTRODUCTION
Norovirus is a good example of a pathogen where improved diagnostics has increased its recognition from a relatively unknown virus before the mid-1990s to the leading cause of epidemic and sporadic gastroenteritis in people of all ages worldwide (1, 2). The majority of norovirus outbreaks occur in health care settings (including long-term care facilities and hospitals), where the virus is predominantly spread from person to person. In addition, noroviruses have also been identified in over 58% of the reported foodborne outbreaks in which an etiologic agent was determined (3). In the most recent disease burden estimates in the United States (US), norovirus causes 570 to 800 deaths, 56,000 to 71,000 hospitalizations, 400,000 emergency room visits, and 1.7 to 1.9 million outpatient visits annually (4). In pediatric populations in industrialized countries where a rotavirus vaccine has been introduced, noroviruses are rapidly replacing rotavirus as the most common cause of medically attended acute gastroenteritis (2, 5). After an incubation period of 12 to 48 h, norovirus illness is characterized by projectile vomiting, nonbloody diarrhea, nausea, abdominal cramps, and low-grade fever. Some persons might experience only vomiting or diarrhea. In healthy individuals, the duration of symptoms is usually not longer than 48 h, and the disease is self-limiting in most patients. However, young children and the elderly are at increased risk for more-severe and prolonged illness leading to hospitalization, while the disease is increasingly recognized as an important cause of chronic gastroenteritis for immunocompromised patients (6). In countries that belong to temperate latitudes, most infections occur in the fall and winter and at least 70% of outbreaks are reported in semiclosed communities such as long-term-care facilities, schools, hospitals, and cruise ships. Noroviruses can infect humans via multiple routes, including the oral route, transmitted through contact with fecal matter or aerosolized vomitus from infected people, as well as contaminated surfaces, food, or water. Upon infection, noroviruses replicate in cells in the upper small intestinal tract (duodenum and upper jejunum), leading to both epithelial barrier and secretory pathway dysfunction. T cells are likely required for virus clearance from the intestine, and, as was reported in a case study of an immunocompromised patient, after more than 1 year of chronic norovirus diarrhea, increasing levels of T cells were associated with resolution of symptoms (7). Outside the human host, the virus is environmentally stable and has an estimated 50% human infectious dose (HID50) ranging from 18 to 1,015 genome equivalents, although a recent study estimated that the HID50 is more similar to those of other RNA viruses (1,320 to 2,800 particles) (8). This article reviews methods based on antigen and molecular detection of human noroviruses. Although other molecular methods such as those employing isothermal amplification (nucleic acid sequence-based amplification [NASBA], loop-mediated isothermal amplification [LAMP]) and microarrays have been described, this review focuses on immunological and reverse transcription-PCR (RT-PCR)-based molecular methods.
Noroviruses are a group of nonenveloped single-stranded positive-sense RNA viruses classified in the family Caliciviridae. The virus particles are 27 to 40 nm in diameter, and the genome is 7.5 to 7.7 kb in length and, except for murine norovirus, contains 3 open reading frames (ORF1, ORF2, and ORF3). ORF1 encodes a polyprotein that is posttranslationally cleaved into seven nonstructural mature proteins (NS1 to NS7) that are involved in viral replication. ORF2 encodes the major structural protein (VP1) of approximately 60,000 Da, and ORF3 encodes a minor structural protein (VP2). The viral capsid contains 90 dimers of VP1 and a few copies of VP2. X-ray crystallographic structure studies using Norwalk virus-like particles have revealed that the VP1 has a shell (S) and a protruding (P) domain (9). The S domain surrounds the viral RNA, and the P domain, which is linked to the S domain through a flexible hinge, corresponds to the C-terminal part of the VP1. The P domain is further divided into P1 and the highly variable P2 subdomain which contains the putative neutralization sites and interacts with histoblood group antigens (HBGAs). VP2 is located interior to the virus particle and is most likely involved in capsid assembly and genome encapsidation (10).
Except for murine strains, noroviruses cannot be cultivated in vitro, which prevents their classification into distinct serotypes. Therefore, they are genetically classified into 6 established genogroups (GI to GVI) (11), while tentative genogroup VII is proposed in this paper (12) (Fig. 1). GI and GII viruses are responsible for the majority of disease in humans, whereas GIV viruses are rarely detected as the cause of epidemic or sporadic disease. Based on the most recent phylogenetic analysis, GII.15 viruses may need to be reclassified as a separate genogroup, but this would need consensus approval from the international norovirus working group (13). Each genogroup is based on phylogenetic clustering of the complete VP1 amino acid sequence and is further divided into genotypes (13, 14) (Fig. 1). To date, nine capsid genotypes have been recognized in GI and 22 in GII, and three genotypes of GII (GII.11, GII.18, and GII.19) have been uniquely detected in swine. GIV viruses consist of 2 genotypes, of which GIV.1 has been detected in humans and GIV.2 in feline and canine species (15). GII viruses are most frequently detected (89%), whereas GI viruses, which include virus of the GI.1 prototype Norwalk virus strain, cause approximately 11% of the outbreaks (16).
Classification of noroviruses into 7 genogroups (GI to GVII) based on amino acid sequence diversity in the complete VP1 capsid protein. To build the phylogenetic tree, capsid sequences from 105 strains representing the spatial and temporal sequence diversity of noroviruses from diverse geographic regions across the world were selected. Viruses belonging to GI, GII, and GIV infect humans, except GII.11, GII.18, and GII.19 viruses, which infect porcine species, and GIV.2 viruses, which infect canine species. GII.15 viruses, which have been detected only in humans, form a tentative new genogroup (dotted circle). GIII viruses infect cows and sheep, GIV.2 infects canines, GV.1 and GV.2 infect mice and rats, respectively, and GVI and GVII infect canine species. GII.4 viruses (arrow) are responsible for the majority of norovirus infections worldwide. The scale bar reflects the number of amino acid substitutions per site.
Despite the extensive genetic diversity among noroviruses, viruses from a single genotype, GII.4, are responsible for the majority of the norovirus outbreaks worldwide (17). Due to epochal evolution, novel pandemic GII.4 variants have emerged every 2 to 3 years since the mid-1990s, replacing previous predominant GII.4 strains but not other endemic strains (17). These global GII.4 variant strains include the GII.4 US95/96 strain in 1995, GII.4 Farmington Hills in 2002, GII.4 Hunter in 2004, GII.4 Den Haag in 2006, GII.4 New Orleans in 2009, and GII.4 Sydney in 2012. These new GII.4 variants are often, but not always, associated with an increase in the number of outbreaks (18). In the United States, GII.4 Sydney has continued to cause the majority of the norovirus outbreaks during the 2013-to-2014 season (unpublished data).
Several mechanisms that could explain the evolution of GII.4 viruses have been proposed, including the host herd immunity that drives antigenic variation in the hypervariable P2 domain of VP1. This domain of the viral capsid binds HBGAs, which serve as cell attachment factors for noroviruses (19). Expression of HBGAs on cell surfaces is affected by the ABO, Secretor, and Lewis genotypes. Because GII.4 viruses can bind a wider range of HBGAs than other genotypes, they are able to infect a larger susceptible population. Another mechanism which may explain the emergence of new variants is homologous recombination, with most breakpoints identified in the ORF1-ORF2 junction region. Intergenotype and intragenotype recombination is also widespread, suggesting that both escape from herd immunity and recombination are important factors that drive the emergence of novel GII.4 viruses (20).
DIAGNOSTIC METHODS
Although norovirus can be detected in rectal swabs and vomitus, whole-stool samples are the preferred clinical specimen for the detection of norovirus because they contain a higher quantity of virus. Until the cloning and sequencing of the Norwalk virus genome in 1990 (21) followed by the development and application of the first RT-PCR assays for norovirus, electron microscopy (EM) was the only method to detect the virus. Initially named Norwalk-like viruses or small-round structured viruses, based on their morphology in EM, this group of viruses is now officially known as noroviruses, with Norwalk virus as its prototype. Although EM can also visualize other established gastroenteritis viruses such as rotaviruses, adenoviruses, astroviruses, and sapoviruses, the method is costly and insensitive and therefore not widely available in diagnostic microbiology laboratories.
Because the rapid spread of norovirus is a major public health issue, rapid laboratory diagnosis is essential to assist implementation of appropriate control measures to reduce the spread of the virus and the magnitude of outbreaks. Hence, a simple rapid norovirus test would be an attractive alternative to more technically demanding assays such as enzyme immunoassays (EIAs) and reverse transcriptase PCR (Table 1). Immunochromatographic (ICG) lateral flow assays do not require specialized laboratory equipment and are designed for rapid (15-min) testing of individual samples. In a recent evaluation of 4 norovirus ICG tests, using a comprehensive panel of a wide variety of norovirus genotypes, the specificity of all tests was 100%. However, the overall sensitivity ranged from 35% to 52% and was strongly genogroup dependent, as the sensitivities ranged from 17% to 52% for GI strains to 59% to 78% for the predominant GII.4 viruses (22). These results were significantly lower than the sensitivities reported by other investigators as well as by the different manufacturers of the ICG kits, suggesting that robust evaluation of norovirus test requires validation with a norovirus stool panel that includes a wide variety of different GI and GII genotypes (22).
Overview laboratory assays for detection of norovirus
The development of a broadly reactive EIA for noroviruses has been challenging because of the number of antigenically distinct humans norovirus genotypes (n = 29) and the antigenic drift of certain strains (e.g., GII.4) over time. Although genogroup-specific monoclonal antibodies have been described, most commercial kits, including the IDEIA Norovirus EIA (Oxoid, Hampshire, United Kingdom), SRSV (II)-AD (Denka Seiken Co. Ltd., Tokyo, Japan), and RIDASCREEN (r-Biopharm AG, Darmstadt, Germany), include combinations of several cross-reactive monoclonal and polyclonal antibodies. The sensitivity of these kits is typically <70%, while the specificity is usually >90%, depending on the diagnostic goal (outbreak or sporadic cases), the number of samples tested per outbreak, and the time after the onset of symptoms that clinical samples were collected. The general scientific consensus is that EIA may be useful for rapid screening of multiple fecal samples collected during an outbreak of acute gastroenteritis for norovirus but, because of the low sensitivity, caution should be exercised in interpreting test results from sporadic cases (23).
In the mid-1990s, the first conventional or endpoint RT-PCR assays were developed, targeting a relatively conserved small region of the RNA polymerase (POL) gene in ORF1 (region A). With the increasing number of sequences that became available during those early years, those assays were rapidly replaced by second-generation assays that proved to be more broadly reactive and able to detect the majority of the circulating norovirus strains. One of those early assays is, in a slightly modified format, still being used successfully for the detection and typing of noroviruses (24). Increased specificity and sensitivity are accomplished by the use of real-time RT-quantitative PCR assays (RT-qPCR) that do not require agarose gel analysis and subsequent confirmation and, in most protocols, use fluorescently labeled oligonucleotide probes. One-step RT-qPCR assays, in which both reverse transcription and cDNA amplification are performed in a single reaction, require less sample handling and therefore decrease the risk of cross-contamination, making them a preferred format in clinical laboratories. Because only one small region of the norovirus genome is sufficiently conserved for the development of genogroup-specific oligonucleotide primers and probes (25), most of the reported norovirus RT-qPCR assays target this ORF1-ORF2 junction region (26, 27). And although no commercial stand-alone norovirus RT-qPCR assay has yet been FDA cleared, in recent years these assays have become the gold standard for the rapid and sensitive detection of norovirus in clinical (stool, vomitus, serum) samples as well as in food, water, and environmental samples. Increasingly, RT-qPCR assays include an internal extraction control to reduce false-negative results and are able to simultaneously detect GI and GII strains (28) or GI, GII, and GIV strains (29). In addition to the high analytical sensitivity, RT-qPCR assays can also be used to determine the amount of nucleic acid in a sample in a semiquantitative way as a proxy to determine the viral load. Patients with higher viral loads have been reported to excrete the virus longer, and data from several studies suggest that GII viruses (i.e., GII.4) are shed in larger amounts than GI viruses (30).
A significant number of patients excrete the virus 3 weeks after clinical symptoms have disappeared, and noroviruses are also frequently detected in fecal samples from asymptomatic patients, in particular, in children under the age of 5. Hence, virus detection by RT-qPCR does not always correlate with clinical norovirus disease but assessment of a possible difference in viral load in samples from clinical and asymptomatic cases may be a helpful tool to assess a causal relationship with clinical symptoms. In a study in the United Kingdom, higher viral loads were found in norovirus-positive cases than in asymptomatic controls and a clinically significant cutoff value of 31 cycles for all ages resulted in a sensitivity of 72% (31). However, in other studies, including studies in low-income countries, norovirus was as commonly detected in stools from cases with moderate to severe diarrhea as in those from healthy controls and was present in similar viral loads (reference 32 and unpublished data). This makes interpretation of positive RT-qPCR results, particularly those determined from samples with low viral loads (high threshold cycle [CT] values), a challenge. Additional data from studies of considerable sample size are required to determine robust CT cutoff values to interpret norovirus RT-qPCR results. Such cutoff values may depend on variables such as the sample collection date, PCR platform, reagent or kit used, and study population (e.g., outbreak versus sporadic samples). Alternatively, data from outbreak studies in which multiple samples have been collected from norovirus-positive patients after they have become asymptomatic may help in establishing a clinically relevant cutoff value.
In recent years, several different multi-gastrointestinal-pathogen diagnostic platforms have been developed for the simultaneous detection of pathogenic enteric viruses, bacteria, and parasites (Table 1). The xTAG GPP (Luminex Corporation, Toronto, Canada), FilmArray GI Panel (BioFire Diagnostics Inc., Salt Lake City, UT, USA), and Verigene Enteric Pathogens Test (EP) (Nanosphere, Northbrook, IL, USA) platforms have recently been FDA cleared and currently provide the most comprehensive commercial multiplex molecular diagnostic tests available for gastroenteritis diagnosis. The FDA-cleared version of the xTAG GPP platform simultaneously detects and identifies norovirus GI and GII, rotavirus group A, 7 bacterial species, and 2 parasite species (33), while the FilmArray GI Panel detects 23 enteric pathogens, including norovirus GI and GII, rotavirus group A, group F adenovirus, sapovirus, and astrovirus, 14 bacterial species, and 4 parasite species. The Verigene EP assay detects 5 bacterial species, 2 Shiga toxins, rotavirus, and norovirus. The Biofire and Luminex platforms are able to distinguish between GI and GII noroviruses. However, there are significant differences between these tests, including workflow and throughput differences (Table 1). The xTAG GPP can complete testing of 24 samples within 5 h, but this does not include preparation and extraction of the samples. In contrast, the FilmArray and Verigene systems have a turnaround time from unprocessed sample to results of 2 h, with minimal hands-on time. The drawback of the FilmArray and Verigene systems in the current formats is their low throughput, as only a single sample can be processed on the instrument at one time, which may not be an issue for clinical laboratories but limits the overall utility of the test in laboratories with moderate to high throughput. Another challenge of these multipathogen systems is that of data interpretation, specifically with high numbers of mixed infections and the lack of quantitative data to determine which pathogen is responsible for the gastrointestinal disease (34).
If no laboratory diagnosis can be performed (e.g., when no specimens are available for testing), norovirus infections can also be detected on the basis of the clinical and epidemiological profile, which has been used successfully to differentiate norovirus causing gastroenteritis from other causes of enteric disease. These Kaplan criteria (35) are based on (i) a mean duration of illness of between 12 and 60 h, (ii) a mean incubation period of 24 to 48 h, (iii) vomiting in >50% of patients, and (iv) the absence of bacterial pathogens detected in stool specimens. The criteria are highly specific (99%) and moderately sensitive (68%) for foodborne outbreaks but may not be valid for hospital outbreaks, where the duration of symptoms can be longer than 72 h.
GENOTYPING
Noroviruses are classified into genogroups and genotypes based on amino acid diversity in the complete VP1 protein, but, as recombination in the ORF1-ORF2 junction region is common and as some capsid genotypes seem to be more prone to recombination than others, a dual-nomenclature system has been proposed using both the RNA polymerase (POL) region in ORF1 and VP1 sequences (13) (Table 2). Currently, 9 genotypes in GI, 22 in GII, 2 in GIII, 2 in GIV, 1 in GV, 2 in GVI, and 1 in the tentative new GVII have been recognized on the basis of complete VP1 amino acids (Fig. 1). The nomenclature system includes information on genogroup, genotype, and, for GII.4 strains, variant type. For example, if both POL and capsid (CAP) sequences are known, the strain name should be written as follows: norovirus GII/Hu/US/2010/GII.P12-GII.12/HS206. When only CAP sequences are available, the strain should be written as follows: norovirus GII/Hu/AU/2012/GII.4 Sydney/Melbourne456.
Norovirus genogroups and genotypes as determined by phylogeny-based cluster analysis of capsid protein VP1, partial capsid nucleotide region (region C or region D), and RNA polymerase region (POL genotype)
Because sequencing of the complete VP1 gene is currently not a routine procedure, nucleotide sequences of relatively small regions of ORF1 (POL or region A) or ORF2 (CAP or regions C and D) of the norovirus genome are used to genotype strains. Region C assays are in general more robust because the lower (40°C) annealing temperature required for the region D assays increases the likelihood of nonspecific amplification and because region D is located in a more variable part of ORF2. As determined on the basis of nucleotide sequence diversity in region C and region D, several genotypes consist of up to 4 different subclusters (e.g., GI.3a to GI.3d); therefore, reference sequences representative of each subcluster are required for correct typing of these strains (Table 2). An online norovirus typing tool is available for both polymerase and capsid typing (36).
GII viruses, in particular, GII.4 viruses, are responsible for the majority of the norovirus outbreaks in people of all ages worldwide, whereas GI strains are more often detected in foodborne and waterborne outbreaks. For example, the GI.6 virus that emerged in 2012 was more often associated with foodborne disease outbreaks than the GII.4 viruses, which are strongly associated with person-to-person transmission and outbreaks in health care settings, resulting in an increased risk of more-severe disease outcomes such as hospitalization and death compared to other GI and GII viruses (37). GII.4 viruses have an epidemiology different those of other GI and GII genotypes. Since the mid-1990s, 7 different GII.4 variants have successively emerged every 2 to 3 years, replacing previous dominant variants, and most of them have produced global epidemics of gastroenteritis. The first reported GII.4 pandemic (caused by GII.4 US95_96) occurred in 1995, followed by the emergence of GII.4 Farmington Hills in 2002, GII.4 Hunter in 2004, GII.4 Yerseke and GII.4 Den Haag in 2006, GII.4 New Orleans in 2009, and GII.4 Sydney in 2012 (Fig. 2). Although media coverage often suggests otherwise, studies in the United States have shown that neither the emergence of GII.4 New Orleans in 2009 nor that of GII.4 Sydney in 2012 led to an increase in norovirus activity compared to previous years. These findings underscore the importance of conducting well-designed studies to better understand the contributions that individual genotypes may make to norovirus disease burden. Between 2009 and 2013, several non-GII.4 strains (GII.12, GII.1, GI.6) have emerged that cocirculated with the predominant GII.4 viruses and caused 11% to 15% of all outbreaks, but each strain did not circulate longer than one norovirus season (16). Genotype distribution in sporadic norovirus disease usually follows the same trends as in outbreaks (2), although rare genotypes are often reported in children under 5 years of age. Continuous norovirus outbreak surveillance through surveillance networks such as NoroNet and CaliciNet will be important to identify changing trends in genotype distribution and identify emerging new strains.
GII.4 norovirus variants with a global distribution and the first season in which they emerged. New GII.4 variants emerge approximately every 2 to 3 years and replace the previously predominant strains. They include US95_96 in 1995, Farmington Hills in 2002, Hunter in 2004, Yerseke in 2006, Den Haag in 2006, New Orleans in 2009, and Sydney in 2012.
FUTURE PERSPECTIVES
Noroviruses are the leading cause of epidemic and sporadic cases of acute gastroenteritis worldwide and a leading cause of foodborne disease. Therefore, rapid laboratory diagnosis is a critical tool to guide controlling norovirus outbreaks by choosing the most appropriate intervention and control practices such as enhanced cleaning and disinfection protocols, isolation, grouping patients based on symptoms, exclusion of symptomatic staff members or food handlers, or, ultimately, closing of units in hospitals (38). Over the last decade, significant progress has been made in the development of diagnostic methods for the routine detection of human noroviruses. RT-qPCR assays have become the gold standard for norovirus detection in most public health and research laboratories and are increasingly commercially available. Continued improvement of rapid, sensitive, and broadly reactive point-of-care assays, such as ICG assays, will be required to allow simple and reliable norovirus diagnosis where no laboratory facilities are available. Use of the multi-gastrointestinal-pathogen molecular platforms that are now available for the rapid detection of a suite of different enteric pathogens, including norovirus, in a single sample will become routine in many clinical laboratories over the next couple of years.
Recent advances in nucleic acid sequencing technologies, such as “next-generation” sequencing (NGS), have opened new perspectives for research and diagnostic applications because of the high speed and throughput of data generation. NGS has been applied for the discovery of novel viruses and the characterization of viral communities as well as whole-viral-genome sequencing and detection of the viral genome variability of RNA viruses. Although challenges remain, including difficulties in sample preparation and high cost, NGS is a potentially powerful method for the rapid identification, and characterization of any infectious agent, including norovirus, directly from stool could assist in infection control of outbreaks.
Genotyping of norovirus strains is important, as certain genotypes are more often associated with foodborne transmission whereas others (e.g., GII.4) have led to more-severe disease outcomes. Standardized genotyping as performed by surveillance networks such as CaliciNet and NoroNet will make it easier to identify new emerging strains or common-source outbreaks and provide useful information on the distribution of strains in different populations, which is important for norovirus vaccine formulations (39).
ACKNOWLEDGMENTS
Many thanks to my colleagues at CDC and the CaliciNet surveillance network for their continued efforts to improve our understanding of the public health burden of norovirus, to Everardo Vega for help with Fig. 1, and to Harry Vennema (RIVM) for sharing polymerase gene reference sequences used in NoroNet.
The findings and conclusions in this article are mine and do not necessarily represent the official position of the Centers for Disease Control and Prevention.
- Copyright © 2015, American Society for Microbiology. All Rights Reserved.
REFERENCES
Author Bios
Jan Vinjé, Ph.D., is head of the National Calicivirus Laboratory and Director of CaliciNet at the Centers for Disease Control and Prevention (CDC) in Atlanta, GA. Dr. Vinjé received his Ph.D. degree at the University of Utrecht, the Netherlands, in 1999. After receiving a postdoctoral fellowship and an appointment as research assistant professor at the University of North Carolina in Chapel Hill, he joined CDC in 2006. Over the past 10 years, he has served on several program advisory committees from several European research projects (FP6, FP7). He is serving as technical expert on the norovirus subcommittee of the National Advisory Committee on Microbiological Criteria for Foods and is a member of the International Committee on Taxonomy of Viruses study groups on Caliciviridae (chair as of 2014) and Astroviridae. He is currently a member of the editorial board of the Journal of Clinical Microbiology and associate editor of the journal Food and Environmental Virology, and he serves as an ad hoc reviewer for multiple high-impact journals. Dr. Vinjé has published over 100 peer-reviewed publications and several book chapters. His research interests include all aspects of viral gastrointestinal disease, including detection, characterization, and prevention and control of norovirus infections.
