Novel Surveillance Network for Norovirus Gastroenteritis Outbreaks, United States

TOC Summary: The launch of CaliciNet in March 2009 was a milestone.

N oroviruses are the primary cause of epidemic viral gastroenteritis and the leading cause of foodborne outbreaks in the United States (1)(2)(3). Although the course of disease is in most cases self-limiting, young, elderly, and immunocompromised persons are at risk for complications caused by severe vomiting and diarrhea (4)(5)(6)(7)(8). In addition to the clinical impact of norovirus disease, the economic effects in lost wages, time, and intervention procedures (e.g., clean-up costs and recalls) can be signifi cant (9)(10)(11). Although norovirus outbreaks occur year-round, they are more common during the winter months (12)(13)(14).
Noroviruses are genetically classifi ed into 5 genogroups, GI-GV, with GI and GII strains responsible for most human disease (2,15). GII viruses can be further divided into at least 19 genotypes, of which GII.4 is responsible for >85% of outbreaks (14,16), although other genotypes and viruses continue to circulate and cause sporadic disease in children (17)(18)(19). Over the past 15 years, new GII.4 variants have been identifi ed; several have been associated with a global increase in the number of outbreaks (15). The last pandemic GII.4 variant, GII.4 2006b or GII.4 Minerva, was identifi ed in late 2005/early 2006 and has been the predominant outbreak strain in the United States since then. The successive displacement of GII.4 variants suggests that population immunity is driving the evolution of GII.4 viruses (20,21), and the emergence of a new variant will cause an increase in the number of outbreaks in an immunologically naive population.
It is not fully understood why some GII.4 variants become pandemic whereas others do not. The combination of novel antigenic sites in protruding regions of the capsid (centered around amino acids 295 and 396) and the change or expansion of a susceptible population may be responsible for the emergence of pandemic variants (20,22). The latter theory has been supported by the discovery that different norovirus strains may have different histo-blood group antigen (HBGA) binding patterns and that nonsecretors are not susceptible to infection with certain genotypes or variants (23). Most mutations between genotypes and variants occur in the P2 region of the major capsid viral protein (VP), VP1, which contains the HBGA binding sites.
Since 2008, all 50 states have had the laboratory capacity for norovirus testing; the Centers for Disease Control and Prevention (CDC) National Calicivirus Laboratory (NCL) provides laboratory support to states that do not have in-house capacity for norovirus strain typing. Recent studies on the molecular epidemiology of norovirus in the US have been based on specimens from a subset of outbreaks that were submitted to CDC (13,24,25). To enhance and harmonize norovirus outbreak surveillance, CDC and its state partners have developed a national norovirus outbreak surveillance network, CaliciNet. CaliciNet was developed to improve standardized typing of norovirus outbreaks, assist in linking geographically different clusters of norovirus illness, allow rapid classifi cation and identifi cation of new norovirus strains, and establish a comprehensive strain surveillance network in the United States. In this article, we describe the CaliciNet network and report fi rst-year results, including the identifi cation of a new GII.4 norovirus variant.

CaliciNet
CaliciNet is a novel electronic laboratory surveillance network of local and state public health laboratories in the United States, coordinated by CDC. CaliciNet participants perform molecular typing of norovirus strains by using standardized laboratory protocols for reverse transcription PCR (RT-PCR) followed by DNA sequence analysis of the amplicons. A customized CaliciNet database developed in Bionumerics version 5.1 (Applied Maths, Austin, TX, USA) includes norovirus sequence and basic epidemiologic information (Table 1), which are submitted electronically via a secure connection to the CaliciNet server at CDC. Both epidemiologic and sequence data can then be used to help link multistate outbreaks to a common source (e.g., contaminated food). To ensure high-quality data entry, submissions to the CaliciNet server are performed by certifi ed laboratory personnel of the participating state or local health laboratories, and fi nal quality assurance/quality control is performed at CDC.
CaliciNet certifi cation for participants is a 2-step process that involves evaluation of data entry and analysis of sequences and a laboratory panel test. Each laboratory must pass an annual profi ciency test. The laboratory certifi cation and profi ciency test consists of analyzing a panel of fecal samples by real-time RT-PCR and conventional RT-PCR analysis followed by bidirectional sequencing as described below. Certifi ed participants are then authorized to upload norovirus outbreak data consisting of >2 samples per outbreak to the national CaliciNet database (Table 1). GII.4 sequences with >2% and 3% difference in region C or D, respectively, and >10% difference with all other noroviruses are further analyzed at CDC by amplifi cation of the VP1 or P2 region.

Outbreaks
All outbreaks submitted to CaliciNet and the NCL from October 2009 through March 2010 were genotyped by region D analysis (26). To verify GII.4 New Orleans variants, a subset of outbreaks from CaliciNet participating laboratories and 2 specimens from each outbreak received at the NCL from October 2009 through May 2010 were analyzed by using the P2 region as described below.

Viral RNA Extraction
Viral RNA was extracted from clarifi ed 10% fecal suspensions in phosphate-buffered saline with the MagMax-96 Viral RNA Isolation Kit (Ambion, Foster City, CA, USA) on an automated KingFisher magnetic particle processor (Thermo Fisher Scientifi c, Pittsburgh, PA, USA) according to the manufacturer's instructions and eluted into 100 μL of elution buffer (10 mmol/L Tris pH 8.0 and 1 mmol/L EDTA). Extracted RNA was stored at -80°C until further use.

Real-time RT-PCR
Viral RNA was tested for GI and GII noroviruses in a duplex format by using the AgPath-ID One-Step RT-PCR Kit (Applied Biosystems, Foster City, CA, USA) on a 7500 Realtime PCR platform (Applied Biosystems). The fi nal reaction mix of 25 μL consisted of 400 nmol/L of each oligonucleotide primer, Cog1F, Cog1R, Cog2F, and Cog2R, and 200 nmol/L of each TaqMan Probe Ring 2 (27) and Ring 1C (28) ( Table 2). Cycling conditions included reverse transcription for 10 min at 45°C and denaturation for 10 min at 95°C, followed by 40 cycles of 15 s at 95°C and 1 min at 60°C.

Region D RT-PCR
The QIAGEN One-Step RT-PCR Kit (QIAGEN, Valencia, CA, USA) was used for region D amplifi cation in a 25-μL reaction volume. RNAse Inhibitor (Applied

P2 Region Amplifi cation
The P2 region was amplifi ed by using the SuperScript III One-Step RT-PCR with Platinum Taq High Fidelity Kit (Invitrogen, Carlsbad, CA, USA). The fi nal reaction volume of 25 μL consisted of 4 μmol/L of EVP2F and EVP2R (Table 2). RT-PCR conditions included reverse transcription at 55°C for 30 min and denaturation at 94°C for 2 min, followed by 40 cycles of PCR at 94°C for 15 s, 55°C for 30 s, 68°C for 1 min, and a fi nal extension step of 68°C for 5 min.

Amplifi cation and Cloning of GII.4 New Orleans
Novel GII.4 New Orleans sequences were identifi ed by region D sequence analysis and further analyzed by amplifi cation of complete open reading frame 2. Extracted RNA from fecal samples underwent cDNA synthesis with a TVN-L1 primer (29) ( Table 2) for 60 min at 50°C by using the Superscript III cDNA synthesis kit (Invitrogen). The reaction mixture was purifi ed by using the DNA Clean and Concentrator-5 (Zymo Research, Orange, CA, USA). The cDNA was amplifi ed by using oligonucleotides (0.5 μmol/L each) L1 and Cog2F (Table 2), using the Phusion PCR Kit with the addition of 3% dimethyl sulfoxide (Finnzymes, Woburn, MA, USA). PCR conditions included denaturation at 98°C for 30 s followed by 40 cycles of 98°C for 10 s, 48°C for 30 s, and 72°C for 1.5 min. A fi nal elongation step was run at 72°C for 10 min.
PCR products of ≈2.5 kb were gel purifi ed and cloned by using a TOPO-TA Cloning Kit (Invitrogen). Five clones of each strain were fully sequenced bidirectionally and their respective consensus sequences were submitted to GenBank. The accession no. for GII.4 New Orleans is GU445325.

DNA Sequencing
All amplicons were purifi ed with the QIAquick Gel Extraction or PCR Purifi cation Kits (QIAGEN) and sequenced by using the BigDye Terminator Kit version 1.1 (Applied Biosystems). Sequence reactions were cleaned up by using the BigDye Xterminator Kit (Applied Biosystems) and analyzed on a 3130XL Automated Sequencer (Applied Biosystems).

Phylogenetic Analysis
VP1 or P2 sequences were aligned by using MEGA4 software (30). Maximum-likelihood phylogenetic analysis of VP1 amino acids were run in PhyML version 3.0 (www. atgc-montpellier.fr/phyml/binaries.php) by using the LG amino acids replacement matrix (31). The initial tree was the best of 5 random trees, and branches were supported by 100 bootstrap replicates. Branches with bootstrap support <60 were collapsed. , and GII.4 New Orleans) were also included in the analysis. Maximum-likelihood scores were generated for all models available in jModelTest by using the Akaike information criterion or the Bayesian information criterion (32) to select the best nucleotide replacement matrix for phylogenetic analysis. A transitional model with rate variation between sites, TIM3+G, had the best maximum-likelihood scores. The custom model was then run in PhyML version 3.0 with the approximate-likelihood ratio test calculated for branch support (33).

Results
As of February 2011, public health laboratories in 20 states have been CaliciNet certifi ed ( Figure 1); these states represent 53% of the US population (34).  Because region C and region D analyses are not able to differentiate genetically closely related GII.4 variants, we further analyzed the P2 region of 20 GII.4 New Orleans region D-positive strains from 4 CaliciNet states and sequenced the complete VP1 gene from a representative outbreak (Figure 3). Compared with recent GII.4 variants, GII.4 New Orleans had several amino acid substitutions, which were located near protruding regions (aa 294 and 396) and HBGA interaction sites (aa 339-341) (Figure 4).
In addition to the 20 strains from the CaliciNetcertifi ed laboratories, the P2 region from 75 GII.4 outbreaks submitted by 12 non-CaliciNet states was sequenced ( Figure 1). Of these, 72 (96%) outbreaks had P2 sequences with <2% nt difference compared with the prototype GII.4 New Orleans strain. Sequences from 3 outbreaks were closely related to a GII.4 variant fi rst detected in Australia in 2008 (Figure 3).

Discussion
The  when GII.4 Minerva (GII.4 2006b) was identifi ed, another GII.4 variant was co-circulating (21,37). CaliciNet uses the same software as the US bacterial enteric pathogen surveillance network (PulseNet) (38), but it is customized with plug-ins to add CaliciNet-specifi c functionality. CaliciNet uses sequence data, whereas PulseNet is based on pulsed-fi eld gel electrophoresis restriction digestion patterns of bacterial enteric pathogens. Current typing regions of CaliciNet target small regions of the norovirus genome, which makes it diffi cult to discern closely related norovirus strains, although the implications to human health may be signifi cant. Our data and data from other studies (39) demonstrated that P2 region analysis enables more sensitive identifi cation of new GII.4 variant strains compared with currently used CaliciNet regions. Use of these analyses would increase the sensitivity of outbreak surveillance to track strains that are part of a single outbreak and likely to have a common source. Hence, P2 is under consideration to be included in CaliciNet.
Like CaliciNet, the Foodborne Viruses in Europe network (FBVE) uses a central database to which users can submit norovirus sequences (40). Compared with the FBVE network, CaliciNet focuses primarily on noroviruses, is not web-based, and is based on a secured network connection to CaliciNet servers at CDC where the states log on as clients, enabling them to upload, view, and query outbreak data submitted by other states. CaliciNet also organizes training workshops and sends standardized protocols and annual profi ciency panels to its members. The benefi t of the FBVE network is that it can be more easily expanded to include laboratories outside its network, whereas to date CaliciNet allows only participants from state and local health laboratories in the US to participate.
The success of CaliciNet in linking multistate outbreaks to a common source (e.g., contaminated food) will depend on joint efforts of state and local epidemiologists to rapidly identify the likely common source and on CaliciNet laboratories for the timely upload of outbreak sequences to the national CaliciNet database. Although CaliciNet has selected region D as its preferred sequence region, a region C and soon a P2 region sequence database will be maintained to enable exchange of information with other norovirus surveillance networks. Because the region D assay targets a genetically highly heterogeneous region of VP1, the performance of this assay will be closely monitored over time, and necessary changes will be implemented to improve assay sensitivity and specifi city. Future CaliciNet expansion will include other gastroenteritis viruses, such as sapovirus and astrovirus, as well as add capability for CaliciNet members to submit fecal samples from patients involved in norovirus-negative outbreaks to CDC for further testing, including novel pathogen discovery sequencing technologies (18