VIRUS IDENTIFICATION: A CRITICAL FIRST STEP IN THE DEVELOPMENT OF ANTIVIRAL THERAPY Biswendu B. Goswami, Ph.D. United States Food and Drug Administration (Retired) Current Address: 7624 Epsilon Drive, Derwood, MD 20855 Tel:301-990-6940; Abstract: The necessity for rapid methods of virus identification arose from the fact that human diseases caused by food-borne viruses range from infectious hepatitis to acute gastroenteritis. The viral causative agents run the gamut from simple single stranded RNA viruses (hepatitis A or infectious hepatitis), to more complex double stranded RNA viruses such as rotavirus and large single stranded RNA viruses such as corona virus. To complicate matters further, each of these virus species have dozens of genotypes ( each genome differing from the prototype by a few to hundreds of nucleotides, depending on the size of the genome. The number of human infections reported each year has made these viruses a major public health problem in the USA in recent years. Food-borne viruses have been associated with high morbidity and some mortality. Currently, other than disinfection guidelines, there are no effective control measures available for public health agencies to intervene in a meaningful way to reduce the incidence of food-borne infections. Viruses can be detected in both stool and vomitus, and the usual route of spread is via person to person. With the availability of more sensitive molecular diagnostics, increasingly outbreaks are reported where the initial source of the infection is contaminated food or water. For FDA (Food and Drug Administration), CDC (Centers for Disease Control) and ORA (Office of Regulatory Affairs), detection of the virus in clinical samples and contaminated food or water is essential to develop meaningful policies for control and prevention. Keywords: Virus Identification; Microarray; Sequence independent target synthesis 2 Introduction: Procedures for virus identification is going through a revolution due to two recent developments. The orthodox methods for virus identification by cell culture followed by secondary confirmation by immunological or Polymerase Chain Reaction (PCR) techniques were too slow and labor intensive. PCR based methods have the added disadvantages they are extremely sensitive to impurities in the test material, and have very narrow specificity, requiring individual tests for each virus, and confirmation by sequencing is frequently a necessity to distinguish viral genotypes and sub-genotypes in a species.. Microarray hybridization based techniques for the identification of known mutations in specific areas of viral genomes initially were successfully performed by synthesizing oligodeoxy nucleotide probes based on the sequence of wild-type and mutated viral genomes. The probes (individually synthesized) were then immobilized on solid supports by hand spotting, and later by automated programmable applicators that could handle several hundred probes [1] - [9]. An alternative to the traditional dye termination method of sequencing nucleic acids based on hybridization to oligonucleotide probes on solid support was developed in the 1990s [10]- [13].. However, the true potential for identification of viruses on a massive scale based on sequence of viral genomes did not take off until the development of in situ probe synthesis and immobilization techniques by [11] [14] ;[15]. These newer photolithographic techniques allowed for the synthesis and simultaneous immobilization of hundreds of thousands of probes in a single microarray allowing re-sequencing on a large scale by hybridization to identify virus genotypes. Wong et al., [16]Genome Res. 2007 or 2004?) used such an 3 array incorporating more than 383,000 probes to sequence the entire 29.7 kb genome of the SARS (Severe Acute Respiratory Syndrome-CoV (coronavirus). Since then, microarrays have been developed that can incorporate over two million probes , This development opened up the possibility of identifying multiple virus species and genotypes in a single experiment. However, to realize this potential, it is necessary to have a target synthesis method that is independent of the sequence of the viruses present in the test material. The new microarray based method developed in my laboratory at Food and Drug Administration (FDA), avoids PCR amplification with virus specific primers normally used for target synthesis in microarray experiments. However, it must be realized that PCR increases the number of specic targets by millions of folds, that makes very sensitive virus detection possible, As can be imagined, such sensitive detection is not possible without a means of target or signal amplification. This disadvantage in sensitivity is somewhat mitigated by the ability to amplify the signal produced by hybridization using biotin labeling of targets follwed by immunological detection of the biotin by Cy3 labeled streptavidin due to its strong affinity to biotin. A further signal amplification is possible by a second round of reaction with streptavidin foloowed by reaction with Cy3 labeled strptavidin. [17 ]. Detection and identification of any viral sequence is thus possible within this limitations, as,long as it’s sequence is represented in the array [17; 18;19]. Sensitivity may be further improved using modifications such as anti-sense RNA (aRNA) amplification. Considering the number of viral diseases extant in the world today, only a few of them are controllable by prophylactic use of vaccines. The ease with which viruses 4 develop resistance to vaccines is another major concern. Emergence of resistance viruses is difficult to detect by orthodox methods of cell culture, or even by conventional PCR. As an example , we can cite the experience in many third world countries , which at one time were considered free of poliovirus, but are now experiencing reemergence of this scourge, mainly due to the use of attenuated virus strains for vaccination, which then circulate within the population mainly through fecal contamination [5; 6]. This recirculation allows the virus to undergo multiple rounds of replication raising the possibility of mutations in the genome. A single or a handful of mutations in the genome gives rise to a resistance strain. It is not difficult to imagine a situation when other viruses now effectively controlled by vaccination will develop resistance to existing vaccines. The flu virus is a prime example. New Flu vaccines are needed to be developed almost every year due to development of resistance due to genome mutation []. An even more dangerous and potentially disastrous trend is developing in the United States where a false correlation between Thermosal , a mercury based preservative used in vaccines and Autism has been widely publicized. As a result, many parents now believe that it is preferable not to vaccinate the children to prevent development of Autism, and leave them exposed to risks of diseases such as polio and mumps. The faulty initial report has since been rescinded, but a potentially very dangerous movement has taken root. Food-borne viruses have been recognized as a major challenge to public health for a number of years. Historically, paralytic poliomyelitis transmitted mainly by unpasteurized milk was an eye opener to public health agencies including FDA to the connection between unsafe foods and human health. The introduction of pasteurization 5 and subsequently the development of polio vaccines, and a national program for polio vaccination have made polio a thing of the past. Recently however, other food-borne viral illnesses have emerged to threaten the health and safety of a public that have become increasingly complacent due to the increased bacterial surveillance and implementation of Good Agricultural Practices (GAP) throughout United States and Europe. Fortunately, it is much easier to detect bacteria in foods. Bacteria grow in simple media, which can then be identified by multiple methods [17, 20]. Despite this, frequent localized epidemics of E.coli and Salmonella from contaminated foods such as meat and vegetables are frequently reported. Virus detection in foods is much more difficult, because unlike bacteria they are obligatory intracellular parasites, and can only be grown with difficulty in cultured animal cell lines [1,2,3,5,7,9,16]. The time factor is of critical importance, and detection can only be achieved by extremely sensitive molecular methods such as PCR. More recently, attempts are being made to use the microarray techniques for virus detection [7; 9 21; 22; 23; 24; 25; 26; 27; 28]. Unfortunately, both types of methods are prone to false negative results, particularly when applied to food matrices. Laboratories are reluctant to invest in equipment and trained personnel because the success rate is so low. The success rate of virus detection is fairly high in body fluids and feces, because these sources have relatively high levels of virus contamination [5; 6; 16; 21; 22; 29; 30; 31. The success rate at FDA and other laboratories in virus detection in foods is very low as noted above, because we had to deal with foods, where contamination levels are naturally low. Moreover, the Centers for Disease Control (CDC) have the added advantage of hindsight [32]. They test specimens only from patients with overt signs of disease. But once a disease outbreak has occurred in a 6 population, control measures by the CDC are totally inadequate. Vaccines available are effective only prophylactically to prevent further spread of the disease among the non-infected population. Not knowing what food caused the disease outbreak, FDA is unable to advise the population to avoid certain types of food except in very general terms. Recommendations concerning the cleaning of food prior to cooking and consumption, also is hardly effective.. Most food-borne viruses have RNA genomes, with minimal viral functions necessary for virus propagation. As a result, development of chemical therapeutic agents has been very slow, with only a handful of agents such as Ribavirin [33] and a few other nucleoside analogs such as Xyloadenosine (XyloA) showing potential to be effective against certain viruses [34; 35; 36]. Interferon induced 2’-5’ linked oligoadenosine (2-5A) and its analogs have not been adequately tested [37; 38; 39]. Additionaly, food-borne RNA viruses are notoriously resistant to inactivation by physical or chemical agents [40; 41 ]. Chief among them in developed countries is norovirus (NV) and hepatitis A virus (HAV).. It is now a recognized fact that NV is the causal agent of most food-borne viral infections in the U.S., while other viruses such as hepatitis A virus (HAV), enteroviruses A and B (HEA and HEB) and rotavirus are also contributors. These illnesses can occur following consumption of contaminated food, person-to-person contact, or food contaminated by infected food handlers, or by contact with water contaminated by treated or untreated sewage [42]. A number of factors have compromised FDA’s ability to respond to this new threat by instituting adequate regulatory framework. Projects implemented by FDA have been top-heavy with much more emphasis on sophisticated methods of virus detection, while 7 the problem resides in the isolation of low levels of virus from large quantities of suspected foods in a sufficiently pure form to be effectively detected by these ultrasensitive techniques [. We will address this gap in our ability to provide an underpinning of laboratory science by investigating new detection technologies that will allow FDA and other regulatory agencies such as CDC to respond to food-borne virus infections by identifying the viral agent, determining the source, and the relatedness and extent of such outbreaks, when discussing specific results [24;27] Materials and Methods The paradigm of virus detection technologies currently available is inadequate in several respects. The technology is mainly based on polymerase chain reaction (PCR) based amplification of viral nucleic acids. Since most food-borne viruses have an RNA genome, a process called reverse transcription (RT) to first transcribe viral RNA to cDNA is appended to the PCR procedure (called RT-PCR). This technology is two decades old [43] , although the application of this technology for virus detection in foods is more recent [24;27; 28;29;30; 31]. Several important modifications to the basic RT-PCR have been instituted, such as multiplex PCR [30], competitive PCR [32], real time PCR [23] ,which have increased the sensitivity of virus detection and reduced the time needed to obtain results. The major disadvantage of (real-time) RT-PCR for detection of an unknown virus is the specificity of most of these assays. For example, during an outbreak investigation samples will be screened for an etiologic agent that best matches the clinical symptoms when evaluable (norovirus and rotavirus for acute gastroenteritis and HAV for hepatitis) which may delay reporting if an unexpected etiologic agent is causing the outbreak [33].. 8 To develop new high density oligonucleotide microarray methods with broad specificity for the sensitive detection and typing of foodborne viruses, these methods will have to be able to i) detect which virus is present, ii) the genogroup and genotype and if needed sub-genotype of the virus. This information could then be used to alert the regulatory agencies on the emergence of new virus strains which is currently not possible without sequencing multiple PCR products. Such detection processes will not be dependent on PCR amplification. Therefore, no prior knowledge of the nature of the virus or its genome sequence will be necessary. The same analyte will be used in a single experiment to obtain all the information needed to identify multiple viruses in the same sample. The high density oligonucleotide microarray based procedure has been used in combination with PCR amplification to study the evolution of the SARS coronavirus ([ 16], and bacterial genomes [20;17] as well as for the detection of mutations in the poliovirus genome ([6] . We have identified different types of HAV and HEA and HEB (more commonly known as coxsackieviruses A and B) viruses as well as studied the evolution of the wild type HAV genome HM175 to a cytopathic variant HM175/18f using a high density (> than 13,000 probes) microarray to detect the nucleotide changes accompanying this phenotypic change [18]. Advances in microarray technology have allowed the identification of genetic variability over very long stretches of DNA in bacterial genomes [17; 20]. These newly developed high density microarrays contain thousands to hundreds of thousands of oligonucleotide probes, instead of a few dozen as used for detection of mutations in poliovirus genome during evolution from a vaccine strain to paralytic poliomyelitis strain [ 6 ; 5] in a single array thereby expanding the power of identification. 9 Isolation and labeling of Viral RNA. Viral RNA from HAV and CXKV strains were purified from 140 μl of virus cell culture supernatant with a QIAamp viral RNA mini kit (Qiagen, CA) according to manufacturer’s instructions. For norovirus samples, a 10% (wt/vol) stool suspension was prepared with phosphate-buffered saline (PBS) and clarified by centrifugation at 3,000 x g for 20 minutes. RNA was extracted from 140 μl to 280 μl of the supernatant with QIAamp viral RNA mini kit. The RNA samples were eluted in 50 μl of elution buffer and stored at -80 ºC until used. Reverse transcription was performed for one cycle (42 °C, 60 min) using random hexamer primers (Invitrogen, CA) followed by 15 min at 70 °C. This is essentially the initial step in current RT-PCR protocols. Reverse transcriptase requires a primer DNA or RNA) with a free 3’OH group hybridized to a template (DNA or RNA) to initiate cDNA synthesis. Viral RNAs that do not have a poly (A) sequence at their 3’end (such as rotavirus genomic RNA) cannot be primed by oligo (dT)15 and are primed by random hexamers. Resulting cDNA was fragmented with DNase I (Invitrogen, CA) at 37 °C for 1 min. Fragmented cDNA was labeled with Terminal Transferase (Invitrogen, CA) in the presence of biotin-11-ddATP (PerkinElmer, MA) at 37 °C for 4 hrs so that the fragmentwd cDNA is extended by a single nucleotide, which does not change its hybridization characteristis to the probes in the array.. The biotin labeled synthesized targets are then hybridized to the probes on the microarray. Following a series of washes, the hybridized targets are reacted with Cy3 labeled streptavidin. After further washes to remove unreacted Cy3 coupled strptavidin, hybridized targets are detected by Cy3 fluorescence by scanning the array. 10 A typical result is shown in Fig.1... Computer analysis of the signal by software developed for this purpose is used to identify the virus species, genogroup and the strain within the genogroup. Microarray Hybridization, Scanning, and Data Analysis. Microarray hybridization was performed using the Affymetrix protocol. Briefly, Biotinylated cDNA in the presence of Affymetrix hybridization buffer was hybridized to the microarray chip in a total volume of 120 μl. Before application to the array, the samples were heated to 98 °C for 1 min, cooled at 45 °C for another 5 min, and centrifuged at 12,000 x g for 5 min. The microarray chip was then incubated at 45 °C for 16 hrs in a hybridization oven. Following hybridization, the wash and stain procedures were carried out by the Fluidics station (Affymetrix, CA). All arrays were imaged by using Affymetrix microarray scanner at a resolution of 10 μm per pixel. Signal intensity of the hybridization was extracted by using Affymetrix power tools, and the subsequent data analysis was performed using MS Excel. For each viral genome represented on the array, the average signal intensity for all the probes within that genome was determined. The average intensity is first determined as described by Jackson et al. (16), and Ayodeji et al. (17). Each average genome intensity was then normalized by the average intensity of all the probes represented on the array. To minimize effects of nonspecific hybridization, an empirical cutoff value of 3 was considered as a threshold value for a positive signal. Microarray hybridization data were then converted to color visualization schemes in which hybridization signal intensity is reflected by the color scale of vertical strips. Microarray design: 11 All microarrays used in this study were customer ordered to be fabricated by Affymetrix Inc. (Santa Clara, CA) using a maskless array synthesis technology (Singh-Gasson et al., 1999; Nuwaysir et al., 2002. The tiling microarray contained 25-mer oligonucleotide probes which were designed to detect common food-borne viruses including hepatitis A virus, norovirus, coxsackievirus, rotavirus, hepatitis E virus, and astrovirus as well as saporovirus. All viral genetic sequence data were obtained from database of viral genomes from Genbank. The chips were composed of overlapping 25-mer oligonucleotides which covered 3,000 bp sequences from 5’ end viral genomes of each virus. Each sequence of interest was tiled with the 25-mer probes with 2-bp spacing. The microarray representation of common foodborne viruses are listed in Tables. 1 to 6 along with specific hybridization signal intensities that reflect relatedness of virus strains within the same species.. The design of the microarray follows closely the protocol we have recently described ([39]Ayodeji et al. 2009; Chen at al. [41}). However, instead of 13,000 probes, these arrays each contain 100,000 probes covering every complete and partial sequences available in the database for NoV, HAV, HEA, HEB and RV. We also developed probes covering the entire sequence of several individual viral genome sequences. . Probes are designed to have a 2 base overlap i.e., each probe starts at the 2nd nucleotide of every sequence, thereby covering a 7.5kb genome with about 1500 probes. 7.5 kb is the average size of the RNA genome of a common food-borne virus except rotavirus that has a segmented double stranded RNA genome, where each segment RNA is 1.5 to 2kbp long. Alternatively, the 5 base overlap between consecutive probes may be reduced to increase the number of probes to cover the same sequence Since the capacity of the 12 microarrays (up to 2 million probes) far exceeds the number of probes in our current array design (100,000), we could incorporate probes covering more viruses, although not all virus groups were tested for this study. However, these served the purpose of negative controls to show that false positives were not seen. For target synthesis for these experiments, we have investigated whether viral targets can be obtained in sufficient quantities in a sequence non-specific manner to hybridize to a high density microarray containing greater than 100,000 probes representing all the known sequences (partial or complete) of the major foodborne viruses (NV, HAV, HEA, HEB and rotaviruses) and identify the viral strain (or strains) present in a sample. RESULTS Tables1to 6 show the lists the selected enteric viruses and the number of probes for each strain of virus that were synthesized directly on the FDA_EVIR microarray by photolithography whereby the total number of probes on the array is 91542. Selected regions of several enteric viruses’ genomes were tiled at two nucleotide spacing. Each probe is 25 nucleotides long and there is a 23 base-pair overlap between consecutive probes for the same virus genotype; therefore, the complete array covers 183,084 nucleotides of viral genomic sequence. The probe numbers we used for each strain of HAV for the sequence coverage of several HAV strains belonging to HAV genogroups I to III are shown in Table 1.. But no attempt was made to include all the partial sequences that exist in the database., although we included some of the VP1-2 and VP3 regions, that are reported to contain maximum sequence conservation. 13 Table 1. Tiling microarray representation of common foodborne hepatitis A viruses Table 1. Tiling microarray representation of common foodborne hepatitis A viruses

Genus Genotype/Genogroup Number of probes

Hepatitis A virus IA 4500

                   IB                    4500
                   IIA                   1500
                   IIB                   1500
                   IIIA                  1500
                   IIIB                  1500
                   VP1-2                 5371 
                   VP3                   4526

Three HAV virus strains were used to validate the microarray. They included HM-175/18f, HAS-15, and PA21, which represented genotype IB, IA, and IIIA, respectively. Labeled cDNA from each tested viruses was hybridized to the custom tiling microarrays individually. Table 2 demonstrate hybridization pattern of HM-175/18f strain. Table 2. Positive nucleotide sequence identity from hybridization profile of HAV HM-175/18f strain. Accession Signal Intensity Ratio Strain/Genotype AF268396 29.3 HAF-203/IB HM17518f 17.9 IB HM175wt 16.8 IB DQ646426 16.4 IB AF386888 13.5 HS-21/11/00 / IB MBB 8.9 IB AF386864 8.4 HS-11/07/00 / IB M16632 7.8 HM-175/7MK-5/IB M59810 7.7 HM-175/24A/IB AY226607 6.3 IA GMBwt 4.2 IA 14 AF386850 4.0 IA AF386861 3.9 SA-16/04/99 /IB LU38wt 3.8 IA AB258596 3.7 HAJ94-3/IA AB258576 3.7 IA AF386875 3.2 IA AF314208 3.2 IA AH3 3.0 IA ________________________________________________________________________ In the sample of HM-175/18f, which belonged to genotype IB, strong hybridization to the IB genotype-specific oligoprobes was observed (Table. 2). Although various degree of cross-hybridization to genotype IA oligoprobes was also observed, the strongest signal was identified in genotype IB oligoprobes. Genotype IB oligoprobes yielded at least seven times above the local background, ranging from 29 to 7.7 times. The average ratio of hybridization signal of genotype IB oligoprobes was 13.1 while IA was 3.9 (Table. 2). Table 3 depicted hybridization pattern of HAS-15 strain. In this sample, cross-hybridization to genotype IB and IIIA was observed. However, genotype IA-specific probes still yielded the highest signal intensity of 23.3 (Table.3). Table 3. Positive nucleotide sequence identity from hybridization profile of HAV HAS-15 strain. Accession Signal Intensity Ratio Strain/Genotype AJ505572 23.3 IT-SCH-00/IA AF386888 11.9 HS-21/11/00 IA AF386858 11.2 SA-11/02/98 IA X04200 9.9 IA AB258557 8.2 HAJ90-3/IA X15463 7.0 IA 15 GBMwt 5.9 IA LU38wt 4.9 IA AB258579 4.2 HAJ94-6/IA AH3 4.2 IA AF386863 4.1 IA AY974170 4.0 M2/IA? AY294047 3.9 IT-MAR-02/IB EF406357 3.8 H2/IA AB258596 3.7 HAJ99-2/IA M34085 3.7 PA21/IIIA K02990 3.7 IA AB259814 3.6 MNA06-490 AF386856 3.5 SA-16/01/98/IA AB258670 3.4 HAJ04-3/IA AB258658 3.4 HAJ96-2/IA AY226607 3.3 Kular-1982-2101/IA NOR21 3.2 IIIA AF386864 3.2 SA-11/07/00/IB AB258662 3.1 HAJ99-2/IA AJ505562 3.1 IT-BON-00/IA AJ296172 3.0 NOR-18/IIIA Hybridization pattern observed from RNA isolated from PA21 strain showed that this strain belongs to genotype IIIA. As predicted, Viral RNA from PA21 strain hybridized dominantly to oligoprobes derived from genotype IIIA genomic sequences (Table. 4), but it also cross-hybridized to spots from genotype IIIB and IA sequences. Table 4. Positive nucleotide sequence identity from hybridization profile of HAV PA21 strain Accession Signal Intensity Ratio Strain/Genotype AJ299464 10.2 NOR-21/IIIA M34085 8.5 IIIA 16 AY226607 7.3 Kular-1982-2101/IA NOR21 7.0 IIIA AJ296172 6.7 IIIA AJ968416 6.7 IIIA AJ299462 6.1 NOR-19/IIIA AY644337 5.4 HMH/III AB245928 5.3 HA-JNG04-90/IIIA AJ299463 5.1 NOR-20/IIIA M66695 4.6 ? AB258539 4.5 HAJ84-1/IIIB AY226600 3.9 Moscow-1999-10/IIIA AB258573 3.6 HAJ93-3/IIIA AB258637 3.4 HAJ93-1/IA AJ968417 3.3 Eastbourne-2/IIIA DQ010413 3.2 GBS2003/IIIA HAJNG0690F 3.0 IIIA In all three tested hepatitis A virus strains, no cross-hybridization to other virus genus such as norovirus, rotavirus, and coxsackievirus, et al was observed. Although cross-hybridization within different genotype strains in HAV genus occurred, this was expected due to less divergence among genotypes of HAV. Stronger hybridization signal was obtained between specific genotype viral RNA and array elements derived from the same genotypic virus. Based on the hybridization pattern and the sequence information of the probes, it is clearly possible to detect and type HAV. Our results indicate that the microarray hybridization technique can be applied to the identification of different viruses present in a sample and detect single nucleotide polymorphisms (SNP) to identify closely related viral strains belonging to the same species [35](Ayodeji et al. 2009). However, the current protocols for SARS virus ([16]Wong et al., 2004), poliovirus ([5]Cherkasova et al., 2003; [6]Laassri et al., 2005), 17 and rotavirus ([38]Honma et al., 2007; [3]Chizhikov, et al., 2002), still depend on RT-PCR for target synthesis to be used in hybridization experiments. Using virus-specific RT-PCR assays requires prior knowledge of the virus to be expected in the sample hence potentially missing other viral disease agents that may be present in the sample but go undetected. Since current microarray allow of the use of up to 2 million probes in a single array, the potential for the detection and identification of many viral disease agents in a single experiment is not fully realized. In our laboratory, we have investigated whether viral targets can be obtained in sufficient quantities in a sequence non-specific manner to hybridize to a high density microarray containing greater than 100,000 probes representing all the known sequences (partial or complete) of the major foodborne viruses (NV, HAV, HEA, HEB and rotaviruses) and identify the viral strain (or strains) present in a sample. Synthesis of cDNA in a non-sequence specific manner was accomplished by reverse transcription of viral RNA using a mixture of random hexamers or olig(dT)15 to prime cDNA synthesis using viral RNA as template by AMV reverse transcriptase. This is essentially the initial step in current RT-PCR protocols. Reverse transcriptase requires a primer DNA or RNA) with a free 3’OH group hybridized to a template (DNA or RNA) to initiate cDNA synthesis. Viral RNAs that do not have a poly (A) sequence at their 3’end (such as rotavirus genomic RNA) cannot be primed by oligo (dT)15 and are primed by random hexamers. We then label the cDNA (see below for details). The synthesized targets are then hybridized to the probes on the microarray and the array scanned for hybridization signal (see Fig.1). Computer analysis of the signal by software developed for this 18 purpose is used to identify the virus species, genogroup and the strain within the genogroup. Simaltaneous Detection of Mulitple HAV Genotypes Detection of Noroviruses. To access the applicability of the tiling microarray for norovirus identification in a clinical setting, we examined three stool specimens isolated from infected patients. The II IIIAIA IBIIIB Fig. 1 Hybridization result of three HAV strains representing 3 subgenotypes: HM175/18f (IB), HAS15 (IA) and PA21 (IIIA). Vertical strip represents virus nucleotide sequence (strain). Detection of strips are grouped by virus genotype of origin. Hybridization signal intensity above threshold is reflected by the color scale of the strip. Black indicates signal intensity at background level. In Table 5, data is shown on the microarray analysis of samples including noro#1, noroGI and #3263..They all belonged to genogroup I. As anticipated, hybridization to genogroup I-specific norovirus-derived oligoprobes was detected in all three virus samples. No cross-hybridization to noro genogroup II oligoprobes was observed. In sample of GI and #3263, strong hybridization to genogroup I specific sequence L07418 was obtained. In the meantime, hybridization to genogroup I specific sequence DQ340089 was also observed, whose signal was slightly lower than cut-off value.. Viral RNA isolated from sample #1 was hybridized to a number of genogroup I-specific oligoprobes without cross-hybridization to genogroup II-specific oligoprobes. Detailed 19 nucleotide sequence identity with positive signal is listed in Table.5. Table 6 lists the the different NV strains and their identified genogroups as available at the time of the experiments. We have also listed other food borne pathogens which were not tested in this study, but act as negative controls against false positive results. Table 5. Positive nucleotide sequence identity from hybridization profile of norovirus #1 . Accession Signal Intensity Ratio Genogroup AB262150 6.8 GI AB186101 5.3 GI AB058537 4.4 GI AB058538 4.1 GI AJ865501 4.1 GI AB112097 3.4 GI AB262171 3.3 GI AB264151 3.2 GI AB112092 3.1 GI AB058536 3.1 GI Table 6. Genogroups of NV and some other viruses whose sequences were included in the microarray along with the number of probes synthesized and immobilized. ______________________________________________________________________________ Norovirus Genogroup I 11468 Genogroup II 18736 Rotavirus Genotype IA 6620 IB 1770 IC 1650 IVA 4660 IVB 1128 IVC 1113 Hepatitis E virus - 1500 Astrovirus - 1500 Sapotovirus - 1500 20 ________________________________________________________________________ Detection of Noroviruses. To access the applicability of the tiling microarray for norovirus identification in a clinical setting, we examined three stool specimens isolated from infected patients. The samples including noro#1, noroGI and #3263 have been genotyped via quantitative RT-PCR. They all belonged to genogroup I. As anticipated, hybridization to genogroup I-specific norovirus-derived oligoprobes was detected in all three virus samples. No cross-hybridization to noro genogroup II oligoprobes was observed. In sample of GI and #3263, strong hybridization to genogroup I specific sequence L07418 was obtained. In the meantime, hybridization to genogroup I specific sequence DQ340089 was also observed, whose signal was slightly lower than cut-off value. (Fig.2, 3). Viral RNA isolated from sample #1 was hybridized to a number of genogroup I-specific oligoprobes without cross-hybridization to genogroup II-specific oligoprobes (Fig. 4). Detailed nucleotide sequence identity with positive signal was listed in Table.5. SaitamaU17 (GII)Fig 2. Detection of NV sample Poo-186,which has been genotyped independently as genogroupII (GII) via quantitative RT-PCR. 21 Fig 3. Detection of norovirusGI. The insert shows the close-up view of positive detection of HAV-derived sequence.L07418 (GI)DQ340089 (GI) Fig 4. Detection of NV sample #1,which has been genotyped independently as genogroupI (GI) via RT-PCR. The insert shows the close-up view of positively hybridized sequences derived from norovirusGI strains. Simultaneous Detection and Typing of CXKV, HAV and Noroviruses 22 To evaluate if the microarray can detect different virus genus simultaneously, we examined CXKV3, HAV HM-175/18f and norovirus #3263. A mixture of cDNA derived from the three viruses was hybridized to the microarray chip. As shown in Fig. 5, CXKV B3 displayed a clear-cut hybridization pattern distinguishable from other coxsackieviruses, making it possible to determine the virus serotype; Strong hybridization to the genotype IB oligoprobes was observed in HAV HM-175/18f although cross-hybridization to genotype IA oligoprobes also occurred (Table.6). In all positive hybridization, the average signal intensity ratio of IB-specific oligoprobes was 8.3 while IA was 4.2, which indicated that the array elements are capable of detecting and typing HAV. Good hybridization was observed between #3263 and genogroup I oligoprobes derived from noro1 L07418 sequence with signal intensity ratio of 3.9 (Table.6). Fig 5. Simultaneous detection and typing of CXKV, HAV and Norovious. A mixed cDNAderived from CXKB3, HAV HM-175/18f and norovirus#3263 was hybridized to the array. The insert shows close-up view of positively hybridized sequences derived from each virus. 23 DISCUSSION The long term objective is to use this technique to identify any viral agent present in the sample without investigator bias. Thus the proposed technological change will reveal a true picture of the viral disease agent and not just the confirmation of a notion based on vague symptoms of a disease. This is particularly important because many of the enteroviruses often show similar disease symptoms. It will also enable us to determine whether multiple strains or virus species are involved in a disease outbreak. Although we deem it unnecessary, the outcome of a microarray detection experiment can be independently verified using the more conventional PCR based amplification and sequencing or a second round of hybridization to a more narrowly focused less complex microarray with fewer probe features. We believe that the strength of the microarray technique is in the redundancy of probes used to interrogate a particular sequence making the technique essentially error-free. Variations in the probe sequences in the form of mutated probe oligonucleotides can be included in the array design to further increase the fidelity of virus identification and confidence in the result. Noroviruses (NoV) and to a lesser extent HAV and human rotavirus (RV) are the most common foodborne viruses. Othe picornaviruses such as human enteric viruses A and B have at least the potential to be transmitted via person to person contact, through infected food handlers and contaminated foods. Current epidemiological studies indicate that NoV is responsible for the vast majority of food borne outbreaks in developed countries. The time frame of gastrointestinal disease caused by NoV is very unique due to the rapid onset of symptoms and easier to spot and a correlation to specific food consumption easier to establish that other viruses such as HEA or HEB or RV. The 24 second most common foodborne virus HAV has a rather long incubation period before disease symptoms become apparent. Epidemiological correlation to any specific food consumption thus is rather difficult to establish. The detection of food-borne viruses in clinical samples (mainly stool samples from outbreak investigations) is extremely difficult using cell culture due to the inability to grow NV and HAV in cultured cells. Hence molecular methods such as real-time RT-PCR are currently used for the detection of foodborne viruses in clinical samples. The current hypothesis developed from this need to broaden the scope of detection without sacrificing the sensitivity and rapidity of detection and the need to obtain more information from an experiment that is currently possible with RT-PCR. The strategy behind this proposal is to investigate whether food-borne virus detection and identification based on high density microarray hybridization currently being developed at Center for Food Safety and Applied Nutrition (CFSAN) can be utilized for routine virus detection in FDA and other agencies such as CDC and Office of Regulatory Research (ORA) for the rapid detection and unambiguous identification (genotyping) of these viral disease agents from diverse source materials. Due to time limitation, we will concentrate our efforts on the detection of virus from stool samples from outbreak investigations available from CDC. Currently, outbreak investigations are geared to detecting and broad genogroup identification of the most common food borne virus NV. Due to the relative low frequency of other foodborne viral diseases, these viruses are usually not targeted. However, RNA virus populations are highly heterogeneous even within a particular genogroup, and genogroup identification using PCR primers based on consensus sequences are unable to identify the particular strain within the genogroup. 25 Analysis of outbreak strain identification is therefore incomplete, since a particular strain within a genogroup may be responsible for the outbreak being studied. In addition, the narrow focus on a particular species of virus ignores other viruses that are also food-borne and may be present in a particular outbreak, which could have provided valuable information as to the source o f the outbreak. Conclusion: We hypothesize that advanced microarray design that incorporates probes for multiple virus species and strains will be able to provide the needed information to pinpoint the exact strain of the virus responsible for the outbreak, and also provide information as to the presence of other viruses that may be associated with that outbreak. Statistical significance of array hybridization signals is computed using array package software where data points are calculated as signal intensity above normalized background (corrected to internal standards) and converted to average probe intensity (Jackson et al. 2007; Wong et al. 2004). Viral samples in the form of stool samples from outbreak investigations will be provided by CDC (NoV) and CBER (RV). The samples will have been typed by specific RT-PCR based methods prior to coding and then supplied to us for microarray experiments. Any initial negative hybridization will trigger a re-sampling and analysis due to the possibility of sampling error or a negative (no virus) sample. Sampling error rates and percent positive identification (following code breakage) will be used to determine statistical success rate of the microarray analysis. We realize that a sample with very low viral load may produce poor hybridization signal. In that case, the viral RNA will be amplified in a sequence non-specific manner using T7 RNA polymerase to preserve the original distribution. 26 The success of our microarray screening protocol depends on target synthesis that is not specific to any particular virus. In our preliminary experiments we have achieved target synthesis for our microarray hybridization experiments by eliminating the PCR step by relying on a highly efficient synthesis and labeling of target. Viral RNA was purified from the all virus containing stool samples with a QIAamp viral RNA minikit (Qiagen, CA), according to the manufacturer's instructions. RNA was reverse transcribed into cDNA using random hexamer primers. The resultant cDNA is fragmented by DNase 1 digestion and the mixture of the fragments is labeled with biotin-11-ddATP (PerkinElmer, MA) in the presence of Terminal Transferase (Invitrogen, CA). The use of ddATP assures that only a single nucleotide is added to the 3’end of each cDNA fragments and therefore the hybridization characteristics to the oligo-probes on the array is not altered. For signal amplification at the end of hybridization, the biotin in the target now hybridized to the probe is reacted with cy3 labeled streptavidin. The cy3 is detected by scanning and the scans analyzed for the detection of hybridization to specific probes on the array. An example of a typical result is shown in Fig.1. The applicability of this protocol in detecting a single species of virus is shown in Fig.2. The experiment detected genogroup I of NV, with the two peaks of average intensities identifying two different strains. Significantly, the microarray detected two different strains of group I NV in the same sample. The smaller peak has a 91% sequence identity to the main peak supporting our hypothesis (see above) that outbreak samples may contain multiple NV strains that can only be detected using microarray. The second hypothesis that we propose to test is that microarray hybridization can detect multiple viruses in the same experiment if the target is synthesized and labeled 27 using technology that are not dependent on sequence specific PCR amplification. An example is shown in Fig.5 from a recent experiment carried out in our laboratory. RNA samples from three different viruses HAV (genotype HM175), NoV (Group I) and HEB (strain CVB 3) were separately reverse transcribed and labeled as described above. The labeled cDNAs were mixed in equimolar quantities and hybridized. Analysis of hybridization data show that all three viruses were identified. In addition to the HAV and CV strains, the array experiment again identified two closely related strains of NoV as seen in Fig.2, that could not be distinguished by RT-PCR. Expected Outcomes: Experiments performed in different laboratories that have used microarray technology for virus detection and identification have concluded that the potential benefits of the use of microarrays in diagnostic settings is very high. The ability of the microarray for virus identification is attested by the proliferation of virus species that are being examined using this technology (Wong et al., 2004 for SARS coronavirus, Laasari et al. 2005, for Poliovirus, Honma et al. 2007, Chizikov et al. 2002, for rotavirus, Ryabinin et al. 2006, for orthopox virus, Deregt et al. 2006, for swine fever virus, Ayodeji et al. 2009, for hepatitis A and human enteric viruses A and B. The dependence on RT-PCR for target synthesis however has restricted the realization of the full potential of the microarray for multiple virus identification in a single experiment. We have shown by eliminating the PCR step during target synthesis that NoV, HAV and human enterovirus can be identified in the same analyte (Fig.3). Comparison of hybridization intensities from targets synthesized from two different strains of the same virus species can identify nucleotide changes in the virus genome occurring over time (Wong et al. 2004; Ayodeji et.al. 28 2009). As shown in Fig.3, perhaps the greatest advantage of this technique is to identify a virus strain in the absence of any prior knowledge of the sequence of the virus in the sample by eliminating the need for PCR. Eliminating the need for PCR does not reduce the sensitivity to the level that the virus can no longer be detected (Fig.3). Conceivably it is possible that such a situation will be encountered in samples with very low viral load. In such cases a sequence non-specific RNA amplification procedure may be used to increase the sensitivity. We have to yet test this hypothesis. The other advantage accruing to the agency will be the ability to identify emerging virus strains which are not detectable due to change of sequence at the PCR primer binding site on the genome, and the need for sequencing every PCR product. Currently, a vaccine against HAV, the second most prevalent foodborne disease is available for prophylactic use. However, a vaccine against NoV is not available to prevent NoV outbreaks. Even where a vaccine for a particular virus is available, vaccination programs to prevent viral disease outbreaks have to be initiated well in advance of a disease outbreak in order to allow time for the targeted population to develop immunity. Another drawback is that viruses frequently mutate to escape immunity as well as to become resistant to antivirals. The microarray data can be of value to provide indication to health officials if resistance strains are emerging and provide sufficient time to change the strategy for controlling an outbreak. END END 29 Acknowledgement: The author wishes to acknowledge the full financial support of the Food and Drug Administration for the work carried out while he was an employee of the Federal Government of the United States of America. References: [1] Verhoef LPB, Kroneman A, van Duynhoven Y, Boshuizen H, van Pelt W, Koopmans M. Selection tools for foodborne norovirus outbreaks. Emerg Infect Dis. 2009; 15: 31-38. [2]Foodborne viral infections. Innovations-report 2008 (August 4th). [3] Chizhikov V, Wagner M, Ivshina A, Hoshino Y, Kapikian AZ, Chumakov. Detection and genotyping of human group A rotaviruses by oligonucleotide microarray hybridization. J. Clin. Microbiol. 2002; 40: 2398-2407. [4] Chizhikov, H.S. et al., J. Clin Microbiol, 45, 2641-8 (2007) [5] Cherkasova E, Laassri M, Chizhikov V, Korotkova E, Dragunsky E, Agol VI, Chumakov K. Microarray analysis of evolution of RNA viruses: evidence of circulation of virulent highly divergent vaccine-derived polioviruses. Proc. Natl. Acad. Sci. USA 2003; 100: 9698-9403. [6] Laassri M, Dragunsky E, Enterline J, Eremeeva T, Ivanova O, Lottenbach K, Belshe R, Chumakov K. Genomic analysis of vaccine-derived poliovirus strains in stool specimens by combination of full-length PCR and oligonucleotide microarray hybridization. J. Clin Microbiol. 2005; 43: 2886-2894. [7] Ryabinin VA, Shundrin LA, Kostina EB, Laassri M., Chizhikov V, Shchelkunov SN, Chumakov K, Sinyakov AN. Microarray assay for detection and discrimination of orthopoxvirus species. J. Med. Virol. 2006; 78: 1325-1340. [8] Neverov, AA et al., J. Clin. Microbiol. 44, 3752 (2006) [9] Grinev, A et al 2008 J.Virol Methods West Nile virus single virus probed with 263 overlapping oligos tested 23 isolates [10] Drmanac, C., and Drmanac,S. 2001. Methods. Mol. Biol [11] Maskos, U., and Southern, M. 1992. NAR 30 [12] Southern, E.M., Maskos, U., and Elder, J.K. 1992. Genomics [13] Schena, M., et al., 1995. Science [14] Pease, A.C. et al., PNAS, 1994 [15] Nuwaysir, E.F. et al., Genome research, 2002 [16] Wong CW, Albert TJ, Vega VB, Norton JE, Cutler DJ, Richmond TA, Stanton LW, Liu ET, Miller LD. Tracking the evolution of the SARS coronavirus using high-throughput, high-density resequencing arrays. Genome Res. 2004; 14: 398-405. [17] Jackson SA, Mammel MK, Patel IR, Mays T, Albert TJ, LeClerc JE, Cebula TA. Interrogating genomic diversity of E.coli 0157:H7 using DNA tiling arrays. Forensic Sci. Internat. 2007; 168:183-199. [18] Ayodeji M, Kulka M, Jackson SA, Patel I, Mammel M, Cebula TA, Goswami BB. A microarray based approach for the identification of common foodborne viruses. The Open Virology Journal. 2009; 3: 7-20. [19] Chen, Haifeng, Mammel M., Kulka, M., Patel, I., Jackson,M., Goswami, B.B.2011. Detection and Identification of common food-borne virus with a tiling microarray. The Open Virology Journal. 5, 36-43. [20] Majtan T, Majtanova L, Timko J, Majtan V. Oligonucleotide microarray for molecular characterization and genotyping of Salmonella spp. strains. J. Antimicrob. Chemother. 2007; 60: 937-946. [21] Honma S, Chizhikov V, Santos N, Tatsumi M, Timenetsky MC, Linhares AC, Mascarenhas JD, Ushijina H, Armah GE, Gentcshy JR, Hoshino Y. Development and validation of DNA microarray for genotyping group A rotavirus VP4(P[4], P[6], P[8], P[9], and P[14]) and VP7 (G1 to G6, G8 to G10, and G12) genes. J. Clin. Microbiol. 2007; 45: 2641-48. [22] Deregt D, Gilbert SA, Dudas S, Pasick J, Baxi S, Burton KM, Baxi MK. A multiplex DNA suspension microarray for simultaneous detection and differentiation of classical swine fever and other pestiviruses. J. Virol. Methods 2006; 136: 17-23. [23] Jothikumar N, Lowther JA, Henshilwood K, Lees DN, Hill VR, Vinje J. Rapid and sensitive detection of noroviruses by using TaqMan-based one-step reverse transcription-PCR assays and application to naturally contaminated shellfish samples. Appl. Environ. Microbiol. 2005; 71: 1870-1875. [24] Goswami BB, Koch WH, Cebula TA. Detection of hepatitis A virus in Mercenaria mercenaria by coupled reverse transcription and polymerase chain reaction. Appl. Environ. Microbiol. 1993; 59: 2765-2770. 31 [25] Atmar RL, Metcalf TG, Neill FH, Estes MK. Detection of enteric viruses in oysters by using the polymerase chain reaction. Appl. Environ. Microbiol. 1993; 59: 631-635. [26] B. B. Goswami and M. Kulka in Food Consumption and Disease Risk: Consumer Pathogen Interactions, M. Potter, Ed., (CRC Press, Woodhead Publishing Ltd., Cambridge, England, 2006), pp 343-392. [27] Goswami, B.B., Kulka, M., Ngo, D., Istafanos, P., and Cebula, T.A, 2002. A polymerase chain reaction based method for the detection of hepatitis A virus in produce and shellfish. J. Food Prot. 65, 393-402. [28} FDA Science Forum (2006). Microarray Based Detection and Identiifcation of Foodborne Viruses. [29] 15. Vinje J, Vennema H, Maunula L, von Bonsdorff C-H, Hoehne M, Schreier E, Richards A, Green J, Brown D, Beard SS, Monroe SS, de Bruin E, Svensson L, Koopmans MPG. International collaborative study to compare reverse transcriptase PCR assays for detection and genotyping of noroviruses. J. Clin. Microbiol. 2003; 41: 1423-1433. [30] Elnifro EM, Ashshi AM, Cooper RJ, Klapper PE. Multiplex PCR: optimization and application in diagnostic virology. Clin. Microbiol Rev. 2000; 13:559-570. [31] Surveillance for foodborne disease outbreaks --- United States, 2007. Centers for Disease Control and Prevention (CDC). MMWR Morb Mortal Wkly Rep. 2010 59:973 (2010). [32] Goswami BB, Koch WH, Cebula TA. Competitor template RNA for detection and quantitation of hepatitis A virus by PCR. Biotechniques 1994; 16: 114-121 [33] Reevaluation of epidemiological criteria for identifying outbreaks of acute gastroenteritis due to norovirus: United States, 1998-2000. Turcios RM, Widdowson MA, Sulka AC, Mead PS, Glass RI. Clin Infect Dis. 42, 964 (2006) [33] Goswami, B.B., Borek, E., Sharma, O.K., Fujitaki, J. and Smith, R.A., 1979. The broad spectrum antiviral agent Ribavirin inhibits capping of mRNA. Biochem. Biophys. Res. Commun. 89, 830-836. 32 [34] Goswami, B.B., Gosselin, G., Imbach, J.L. and Sharma, O.K., 1984. Mechanism of inhibition of herpes virus growth by 2'-5' linked trimer of 9-β-D xylofuranosyladenine. Virology 137, 400-407. [35] Goswami, B.B., and Sharma, O.K., 1983. Inhibition of vaccinia virus growth and virus specific RNA synthesis by 3'-O-Methyl Adenosine and 3'-O-Methyl Guanosine. J. Virology 45: 1164-1167. [36] Goswami, B.B., Sharma, O.K. and Borek, E., 1983. An approach to inhibition of virus replication: inhibition of mRNA methylation. Recent Results in Cancer Res. Springer Press, Berlin-Heidelberg. 84: 275-282. [37] Goswami, B.B., Crea, R., Van boom, J.H. and Sharma, O.K., 1982. 2'-5' Linked oligo(adenylic acid) and its analogs: A new class of inhibitors of mRNA methylation. J. Biol. Chem. [38] Sharma, O.K., and Goswami, B.B., 1981. Inhibition of vaccinia mRNA methylation by 2'-5' linked oligo (adenylic acid) triphosphate (2'-5'A). Proc. Natl. Acad. Sci. USA 78: 2221-2224. [39] Sharma, O.K., Engels, J., Jager, A., Crea, R., Van Boom, J.H.and Goswami, B.B., 1983. 3'-O-methylated analogs of 2'-5'A as inhibitors of virus replication. FEBS Lett. 158: 298-300. [40] Bhattacharya, S.S., Kulka, M., Lampel, K.A., Cebula, T.A., and Goswami, B.B., 2004. Use of reverse transcription and PCR to discriminate between infectious and non-infectious hepatitis A virus. J. Virol. Methods. 116, 181-187. [41] Capsid Functions of Inactivated Human Picornaviruses and Feline Calicivirus Suphachai Nuanualsuwan and Dean O. Cliver*Appl. Environ. Microbiol. January 2003 vol. 69 no. 1 350-357. [42] 11. Atreya, CD. Major foodborne illness causing viruses and current status of vaccines against the diseases. Foodborne Path. Disease 2004; 1: 89-92. [43] Mullis KG, Faloona FA. Specific synthesis of DNA in vitro via a polymerase catalyzed chain reaction. Methods Enzymol. 1987; 155: 263-273. 33 Supplemental Material Microarray Hybridization, Scanning, and Data Analysis. Microarray hybridization was performed using the Affymetrix protocol. Briefly, Biotinylated cDNA in the presence of Affymetrix hybridization buffer was hybridized to the microarray chip in a total volume of 120 μl. Before application to the array, the samples were heated to 98 °C for 1 min, cooled at 45 °C for another 5 min, and centrifuged at 12,000 x g for 5 min. The microarray chip was then incubated at 45 °C for 16 hrs in a hybridization oven. Following hybridization, the wash and stain procedures were carried out by the Fluidics station (Affymetrix, CA). All arrays were imaged by using Affymetrix microarray scanner at a resolution of 10 μm per pixel. Signal intensity of the hybridization was extracted by using Affymetrix power tools, and the subsequent data analysis was performed using MS Excel. For each viral genome represented on the array, the average signal intensity for all the probes within that genome was determined. The average intensity is first determined as described by Jackson et al. (16), and Ayodeji et al. (17). Each average genome intensity was then normalized by the average intensity of all the probes represented on the array. To minimize effects of nonspecific hybridization, an empirical cutoff value of 3 was considered as a threshold value for a positive signal. Microarray hybridization data were then converted to color visualization schemes in which hybridization signal intensity is reflected by the color scale of vertical strips. Validation of Norovirus Microarray Genotyping. To confirm the microarray genotyping result of the norovirus sample #186 (NoV#186), we performed a specific PCR with a published primer set G1SKF (5’-CTGCCCGAATTYGTAAATGA-3’) and G1SKR (5’-CCAACCCARCCATTRTACA-3’); G2SKF (5’- 34 CNTGGGAGGGCGATCGCAA-3’) and G2SKR (5’-CCRCCNGCATRHCCRTTRTACAT-3’; Y=C/T; N=A/T/G/C; R=A/G; H=A/T/C). This pair of primer was used for the amplification of genogroup I and group II norovirus. (Kojima, S., Kageyama, T., Fukushi, S.,et.al. (2002). Genogroup-specific PCR primers for detection of nowalk-like viruses. J Virol Methods. 100: 107-114.) The PCR was carried out as described previously (Kojima, S. above). The amplicon was sequenced and its relationship to known NoV strains were determined by phylogentics analysis program ClustalX as described by Thompson et. al. (1997). (Thompson J.D., Gibson,T.J., Plewniak,F., Jeanmougin,F. and Higgins,D.G. (1997) The CLUSTAL_X windows interface: flexible strategies for multiple sequence alignment aided by quality analysis tools. Nucleic Acids Res., 25, 4876–4882.). A phylogentic tree from boot-strap analysis was generated by Neibour-Joining method. Other norovirus samples were genotyped by Dr. William Burkhardt (FDA, Dauphin Island, AL) and by Dr. Jan Vinje (CDC, Atlanta, GA) via RT-PCR. Hybridization Profile of coxsackieviruses A and B and different strains of HAV 35 CXKV A2Fig 1. Detection of CXKV A2. The insert shows the close-up view of specific detection of CXKV A2 in CXKV strains. Fig 2. Detection of CXKV B2. The insert shows the close-up view of specific detection of CXKV B2 in CXKV strains. 36 CXKV B3Fig 3. Detection of CXKV B3. The insert shows the close-up view of specific detection of CXKV B3 in CXKV strains. CXKV B4Fig 4. Detection of CXKV B4. The insert shows the close-up view of specific detection of CXKV B4 in CXKV strains. 37 Fig 5. Detection of HAV HAS-15. The insert shows the close-up view of positive detection of HAV-derived sequnce. Fig 6. Detection of HAV PA21. The insert shows the close-up view of positive detection of HAV-derived sequence. Statistical analysis: 38 Statistical significance of array hybridization signals is computed using array package software where data points are calculated as signal intensity above normalized background (corrected to internal standards) and converted to average probe intensity (Jackson et al. 2007; Wong et al. 2004). Viral samples in the form of stool samples from outbreak investigations will be provided by CDC (NoV) and CBER (RV). The samples will have been typed by specific RT-PCR based methods prior to coding and then supplied to us for microarray experiments. Any initial negative hybridization will trigger a re-sampling and analysis due to the possibility of sampling error or a negative (no virus) sample. Sampling error rates and percent positive identification (following code breakage) will be used to determine statistical success rate of the microarray analysis. We realize that a sample with very low viral load may produce poor hybridization signal. In that case, the viral RNA will be amplified in a sequence non-specific manner using T7 RNA polymerase to preserve the original distribution. Isolation and labeling of Viral RNA: The success of our microarray screening protocol depends on target synthesis that is not specific to any particular virus. In our preliminary experiments we have achieved target synthesis for our microarray hybridization experiments by eliminating the PCR step by relying on a highly efficient synthesis and labeling of target. Viral RNA was purified from the all virus containing stool samples with a QIAamp viral RNA minikit (Qiagen, CA), according to the manufacturer's instructions. RNA was reverse transcribed into cDNA using random hexamer primers. The resultant cDNA is fragmented by DNase 1 digestion and the mixture of the fragments is labeled with biotin-11-ddATP (PerkinElmer, MA) in the presence of Terminal Transferase (Invitrogen, CA). The use of 39 ddATP assures that only a single nucleotide is added to the 3’end of each cDNA fragments and therefore the hybridization characteristics to the oligo-probes on the array is not altered. For signal amplification at the end of hybridization, the biotin in the target now hybridized to the probe is reacted with cy3 labeled streptavidin. The cy3 is detected by scanning and the scans analyzed for the detection of hybridization to specific probes on the array. An example of a typical result is shown in Fig.1. The applicability of this protocol in detecting a single species of virus is shown in Fig.2. The experiment detected genogroup I of NV, with the two peaks of average intensities identifying two different strains. Significantly, the microarray detected two different strains of group I NV in the same sample. The smaller peak has a 91% sequence identity to the main peak supporting our hypothesis (see above) that outbreak samples may contain multiple NV strains that can only be detected using microarray. The second hypothesis that we propose to test is that microarray hybridization can detect multiple viruses in the same experiment if the target is synthesized and labeled using technology that are not dependent on sequence specific PCR amplification. An example is shown in Fig.3 from a recent experiment carried out in our laboratory. RNA samples from three different viruses HAV (genotype HM175), NoV (Group I) and HEB (strain CVB 3) were separately reverse transcribed and labeled as described above. The labeled cDNAs were mixed in equimolar quantities and hybridized. Analysis of hybridization data show that all three viruses were identified. In addition to the HAV and CV strains, the array experiment again identified two closely related strains of NoV as seen in Fig.2, that could not be distinguished by RT-PCR.