- Email: [email protected]
Microarray-based detection and typing of foot-and-mouth disease virus
The Veterinary Journal The Veterinary Journal 172 (2006) 473–481 www.elsevier.com/locate/tvjl
Microarray-based detection and typing of foot-and-mouth disease virus Mohit K. Baxi
a,*
, Shailja Baxi a, Alfonso Clavijo b, Kimberly M. Burton a, Dirk Deregt
a
a
b
Virology Section, Lethbridge Laboratory (Animal Diseases Research Institute), Canadian Food Inspection Agency, P.O. Box 640, Lethbridge, Alberta, Canada T1J 3Z4 National Center for Foreign Animal Diseases, Canadian Food Inspection Agency, Suite T2300, 1015 Arlington Street, Winnipeg, Manitoba, Canada R3E 3M4
Abstract Foot-and-mouth disease virus (FMDV) is the most economically important veterinary pathogen because of its highly infectious nature and the devastating effects the virus has on the livestock industry. Rapid diagnostic methods are needed for detection and typing of FMDV serotypes and differentiation from other viruses causing vesicular diseases. We developed a microarray-based test that uses a FMD DNA chip containing 155 oligonucleotide probes, 35–45 base pair (bp) long, virus-common and serotype-specific, designed from the VP3-VP1-2A region of the genome. A set of two forward primers and one reverse primer were also designed to allow amplification of approximately 1100 bp of target sequences from this region. The amplified target was labelled with AlexaFluor 546 dye and applied to the FMD DNA chip. A total of 23 different FMDV strains representing all seven serotypes were detected and typed by the FMD DNA chip. Microarray technology offers a unique capability to identify multiple pathogens in a single chip. Crown Copyright 2005 Published by Elsevier Ltd. All rights reserved. Keywords: Foot-and-mouth disease; FMD DNA chip; Microarray; Diagnosis
1. Introduction Foot-and-mouth disease (FMD) is a severe, clinically acute, vesicular disease of cloven-hoofed animals including domesticated ruminants, swine, and more than 70 wildlife species (Coetzer et al., 1994). Foot-and-mouth disease virus (FMDV), a member of the Apthovirus genus of the family Picornaviridae, is a single-stranded RNA virus with a plus-sense genome of about 8500 nucleotides. Seven different serotypes of the virus exist: A, O, C, Asia 1 and South African Territories 1 (SAT 1), SAT 2 and SAT 3. A large number of subtypes have *
Corresponding author. Tel.: +1 403 382 5589; fax: +1 403 381 1202. E-mail address: [email protected] (M.K. Baxi).
also evolved resulting in a high degree of genetic variability within each serotype (Domingo et al., 2003; Knowles and Samuel, 2003). FMDV-infected animals shed large amounts of virus which spread to other animals by direct contact, inhalation of virus-infected aerosols, or ingestion of contaminated material, including water and feed. Cattle that recover from the disease may still shed the virus for a considerable period of time and be the source of infection for other animals. FMD outbreaks usually result in very large economic losses. For instance, the cost of the 2001 FMD outbreak in the UK amounted to £6 billion (€9 bn) (Thompson et al., 2002). FMDV is the most infectious virus known and is considered the most dangerous foreign animal disease agent that could be used for agrobioterrorism. Therefore, it is important
1090-0233/$ - see front matter Crown Copyright 2005 Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.tvjl.2005.07.007
474
M.K. Baxi et al. / The Veterinary Journal 172 (2006) 473–481
to develop rapid, sensitive, and specific diagnostic methods which provide immediate and detailed information should an outbreak/attack occur. Routine methods for detection of FMDV are virus isolation in cell culture and ELISA (Ferris and Dawson, 1988). Although these tests are highly sensitive, they are also time consuming and labour intensive. In recent years, various PCR-based methods have been developed for detection of viral RNA from clinical samples. These assays employ universal (common) primers (AmaralDoel et al., 1993; Reid et al., 1998) and serotype-specific primers for detection and typing (Callens and De Clercq, 1997; Vangrysperre and De Clercq, 1996). Other PCR-based methods developed for FMDV are: PCRELISA (Alexandersen et al., 2000) and antigen capture RT-PCR (Suryanarayana et al., 1999). More recently, real-time RT-PCR assays have been developed by various groups (Moonen et al., 2003; Oleksiewicz et al., 2001; Rasmussen et al., 2003; Reid et al., 2002; Reid et al., 2003). PCR-based assays are rapid, sensitive, and specific. However, a limitation of these assays is the inability to simultaneously amplify in a multiplex fashion all serotypes or detect many viral pathogens, which may be critical, especially in the early stages of an outbreak or attack with an unknown infectious agent. To address the limitations of existing detection methods to type FMDV, we developed an oligonucleotide microarray ‘‘FMD DNA Chip’’. The chip, which contains 155 oligonucleotide probes can simultaneously detect and type all seven FMDV serotypes. To increase the overall robustness of the assay, each serotype is identified by several probes. Besides its speed and sensitivity, advantages of a chip-based test are its ability to screen for multiple pathogens simultaneously and to readily cope with the evolution of new variant strains.
2. Material and methods 2.1. Virus and cell culture Various FMDV strains were used to infect primary lamb kidney (LK) cells cultured in EagleÕs medium (Alpha modification), containing 1% L-glutamine, 10% fetal bovine serum, 10,000 U/mL penicillin, 10,000 lg/mL streptomycin, and 10,000 U/mL nystatin (0.1%). Viruses were harvested 16 h post-infection and titrated in LK cells. 2.2. Probe design and slide printing Complete viral sequences for FMDV strains were obtained from the NCBI GenBank and stored in a database. The compiled database comprised of complete
and partial FMD viral sequences belonging to serotypes A (48 strains), O (49 strains), C (9 strains), Asia 1 (11 strains), SAT 1 (12 strains), SAT 2 (15 strains) and SAT 3 (5 strains). To generate FMDV-common and serotype-specific probes, the program OligoArray 2.0 (Rouillard et al., 2003), available at the Canadian Bioinformatics Resource Web Site (http://cbr-rbc.nrccnrc.gc.ca), was used. This relies on two programs, BLAST (ftp://ftp.ncbi.nih/gov/blast) and MFOLD (www.bioinfo.rpi.edu/~zukerm/rna/mfold-3.1.html) to generate highly specific probes. Prior to running OligoArray 2.0, all complete FMDV sequences were saved in a FMD database in the FASTA format. To ensure the specificity of the probe for its target, the program computes the input sequence against the FMDV database by using the BLAST program for sequence similarity and uniqueness. Specific probes were selected against a set of parameters: probe length (35–50 bp), Tm (80–85), GC content (40–50%), and those governing secondary structure formation and self- and crosshybridization. Probes that met the appropriate criteria were selected and used in the development of FMD DNA chip. A list of probes used in this study is available as Appendix 1 found at http://www.inspection.gc.ca/ english/sci/tech/publish/baxi2005e.shtml. Oligonucleotides were synthesized commercially (Invitrogen Inc.) and were suspended in 50% DMSO at a concentration of 10 nmol/lL. They were printed in duplicate on poly-l-lysine coated slides by SMP3 microspotting pins using the VersArray ChipWriter printer (Bio-Rad Laboratories). The average size of each spot was 100 lm and the distance between spots was 400 lm. Printed slides were stored at room temperature until used. 2.3. RNA extraction Total RNA was extracted from FMDV-infected cells in TriPure reagent (Roche Diagnostics Canada) according to the manufacturerÕs protocol. The RNA pellet was suspended in 25 lL of DEPC-treated water and stored at 80 C. 2.4. PCR amplification of FMDV serotypes All available FMDV genomic sequences were aligned using the program T-coffee (Poirot et al., 2004). A set of FMDV-common primers were designed to amplify the VP3-VP1-2A region of the genome (Fig. 1a). To amplify this region for serotypes A, O, C, and Asia 1, the primer set VP3-Fwd1 and 2A-Rev was designed (Fig. 1b). For serotypes SAT 1, 2, and 3 the primer set VP3-Fwd2 and 2A-Rev was designed. One-step RT-PCR was standardized for the amplification of the VP3-VP1-2A region using the SuperScript III one-step RT-PCR Kit (Invitrogen). The RT-PCR
M.K. Baxi et al. / The Veterinary Journal 172 (2006) 473–481
475
Probes). Briefly, the PCR products were extracted with equal volumes of phenol–chloroform–isoamylalcohol and precipitated with 2 vol of absolute ethanol. The final pellet was dissolved in 20 lL of Alexa Fluor labelling buffer. DNA was denatured at 95 C for 5 min and then put on ice. Alexa Fluor 546 dye was added at 5 lL per 1 lg of DNA and incubated at 80 C for 15 min. The labelling reaction was stopped by placing on ice. Unlabelled dye was removed using DNA clean up columns (Zymo Research Corp). Finally, the labelled DNA was eluted in 20 lL of elution buffer and stored at 20 C until used. The relative efficiency of the labelling reaction was determined as follows: pmol of Dye in labelled sample ¼ ðA546 =104000Þ ðDFÞ ðzLÞ ðwÞ 1012 ; where A546 is the absorbance at 546, DF is the dilution factor, zL is the volume of sample after purification and w is the optical path in cuvette. Labelling efficiency ¼ pmol of Dye in labelled sample =lg of labelled sample. The labelling efficiency was found to be 40–80:1. Fig. 1. PCR amplification of VP3-VP1-2A region. (a) Schematic representation of the FMDV genome and the location of genes for structural proteins (VP4, VP2, VP3 and VP1) is highlighted. The location of forward and reverse primers used to amplify the VP3-VP12A region is shown. (b) List of PCR primers used for amplification. (c) PCR amplification of FMDV, serotypes A (lane 1), O (lane 2), C (lane 3) and Asia 1 (lane 4) generated a product of 1073 bp, whereas amplification of serotypes SAT 1 (lane 5), SAT 2 (lane 6) and SAT 3 (lane 7) generated a product of 1109 bp. 1 kb DNA ladder (lane M) was used for sizing of DNA fragments. PCR products were analyzed in a 1% agarose gel and stained with ethidium bromide.
mixture contained 2· reaction mix (3.2 mM MgSO4, 0.4 mM of each dNTPs), 10 lM of each primer, and a mixture of Superscript III reverse transcriptase and Platinum Taq DNA polymerase enzymes. PCR amplification was performed in an ABI 9700 thermal cycler (Applied Biosystems Inc.) using the following parameters: 50 C for 30 min for reverse transcription; 94 C for 2 min, followed by 40 cycles of 30 s at 94 C, 1 min at 50 C and 1 min at 68 C; and a final extension time for 5 min at 68 C. The resulting PCR products were analyzed by electrophoresis in a 1% agarose gel. 2.5. Synthesis of fluorescent labelled product To generate labelled products, symmetric and asymmetric RT-PCR were performed. The conditions used for asymmetric RT-PCR were the same as those for regular (symmetric) PCR except the forward and reverse primers were added in 1:10 molar concentration. The RT-PCR products were labelled using the ULYSIS Alexa Fluor 546 nucleic acid labelling kit (Molecular
2.6. Microarray hybridization Before using the printed slides for microarray experiments, they were rehydrated with 0.5· SSC to allow the oligonucleotides to spread evenly throughout the spot. Following rehydration, the oligonucleotides were UVcross-linked onto the glass slides at 150 mJ. Poly-l-lysine arrays require that exposed amines be blocked prior to hybridization to prevent non-specific binding of target sequences. To block exposed amines, arrays were incubated in pre-hybridization solution (5· SSC, 0.1% SDS, 0.1% BSA) at 50 C for 30 min. Hybridization of the AlexaFluor-labelled target sequences (or target DNA) with oligonucleotide probes on the FMD DNA chip was performed as previously described with some modifications (Eisen and Brown, 1999). Briefly, 30 pmol of AlexaFluor-546 labelled target DNA was added to the hybridization mixture (30% formamide, 3· SSC, 0.1% SDS and 0.1 ng/mL of salmon sperm DNA). The mixture was preheated at 95 C for 2 min to denature the dsDNA and placed on ice. The hybridization mixture was applied to the array, and the array was covered with a lifter cover slip to prevent probe drying. The slides were placed in the hybridization chamber (Tele-Chem International, Inc.) and incubated in a water bath at 42 C overnight. Following hybridization, slides were washed at room temperature with 0.5· SSC containing 0.1% SDS for 2 min followed by a second wash with 0.05· SSC for 1 min. The array slides were dried by spinning at 600 rpm for 5 min and were then scanned.
476
M.K. Baxi et al. / The Veterinary Journal 172 (2006) 473–481
2.7. Data visualisation and analysis Hybridized microarrays were imaged using the Axon 4000B scanner (Axon Inc.) and the images were analyzed using the GenePix Pro software (Axon Inc.). Microarray data was represented as median pixel intensity at a wavelength of 532 nm with background subtracted (F532median B532median), and spots which had a fluorescent intensity greater than 2 SD (cut-off value) were considered positive.
3. Results 3.1. Design of the FMD DNA chip To develop the FMD DNA chip, oligonucleotide probes were designed from the VP3-VP1-2A region of the genome, as this region has only 60–70% homology between the seven FMDV serotypes which allows the generation of serotype-specific probes. There are also areas of high homology in this region which allows the design of FMDV-common chip probes and PCR primers. A total of 172 oligonucleotide probes of 35–45 bp were designed, of which 155 were selected to be printed on the FMD DNA chip (Table 1). Of these, 23 are specific to serotype A, 28 to serotype O, 15 each to serotypes C and Asia 1, 22 to serotype SAT 1, 34 to
serotype SAT 2, and 12 to serotype SAT 3. In addition, the chip contains six FMDV-common probes. As there is a high genetic variability within each serotype, probes were designed specifically to a number of subtypes or topotypes (geographically restricted subtypes) within each serotype (Knowles and Samuel, 2003). Serotype A is considered to be the most diverse group, both antigenically and genetically. Nineteen probes were designed to the commonly prevalent serotype A subtypes: Europe–South America (Euro–SA; 12 probes), Asia (5 probes) and Africa (2 probes). Similarly, 26 subtype-specific probes were designed to serotype O: Middle East–South America (ME–SA/Pan-Asia strains; 16 probes) and Cathay (10 probes). Thirteen probes designed against serotype C covered subtypes C1 (5 probes) and C3/C5 (8 probes), and 13 probes designed against Asia 1 subtypes covered genotype I (10 probes) and genotype II (3 probes). SAT 1, 2, and 3 serotypes have been shown to have a low level of nucleotide conservation, thereby making it necessary to generate probes that are specific to various topotypes within each serotype. Four topotype-specific probes were designed against SAT 1 topotypes: I (3 probes) and II (1 probe). A panel of 22 topotype-specific probes were also designed for serotype SAT 2 as shown in Table 1. Thus, a large number of specific oligonucleotide probes were designed to cover the genetically distinct groups of FMDV.
Table 1 Summary of oligonucleotide probes designed from the VP3-VP1-2A region Serotype
OligoÕs designed
OligoÕs selected
1
A
25
23
2
O
31
28
3
C
16
15
4
Asia 1
16
15
5
SAT1
28
22
6
SAT2
38
34
7 8
SAT3 FMD-Common Total
18 6 172
12 6 155
Subtypes Common Euro-SA Asia Africa Common Pan-Asia Cathy Common C1 C3/C5 Common Genotype I Genotype II Common Topotype I Topotype II Common Topotype A Topotype B Topotype C Topotype D Topotype E Topotype F Topotype G Common
4 12 5 2 2 16 10 2 5 8 2 10 3 18 3 1 12 3 1 4 4 3 4 3 12
M.K. Baxi et al. / The Veterinary Journal 172 (2006) 473–481
3.2. PCR amplification of target sequences Since there is a high degree of genetic dissimilarity in the VP3-VP1-2A region between the serotype A, O, C and Asia 1 group and the serotype SAT 1, 2, and 3 group, two forward primers (VP3-Fwd1 and VP3Fwd2) and one common reverse primer (2A-Rev) were designed to amplify target sequences of both groups. After PCR amplification, a product of 1073 bp was generated for serotypes A, O, C, and Asia 1 and a 1109 bp product was generated for serotypes SAT 1, 2, and 3 as expected (Fig. 1c). No products were generated following PCR amplification of negative controls (BVDV RNA and water). Two different PCR-based methods were used to amplify target RNA sequences: symmetric PCR to generate dsDNA PCR products and asymmetric PCR to generate ssDNA PCR products. The products were then labelled with Alexa-fluor 546 dye and applied on the FMD DNA chip. As shown in Fig. 2, the labelled asymmetric PCR product generated fluorescent intensities that were much higher than those achieved with the symmetric PCR product. The number of target-probe reactions detected positive were also greater using the asymmetric PCR product as compared to the symmetric PCR product. The abundance of ssDNA produced by asymmetric PCR provides a high amount of complimentary labelled target DNA available to be bound to the oligonucleo-
Fig. 2. Fluorescent intensities of microarray spots on the FMD DNA chip following hybridization with: (a) symmetric and (b) asymmetric PCR products generated from FMDV strain A-24 Cruzeiro. The variable fluorescent intensities are visualised as green spots on scanned images at wavelength 532 nm. (c) Comparison of fluorescent intensities (y-axis) obtained after hybridization with symmetric and asymmetric PCR products with various FMDV-specific probes (x-axis) on the FMD DNA chip. The fluorescent intensities were quantified using GenePix Pro software.
477
tide probes, thus generating high fluorescent intensities. In the case of symmetric PCR, probable self-reannealing of the labelled dsDNA product decreases the number of target sequences that can bind to oligonucleotide probes resulting in lower fluorescent intensities. Thus, asymmetric PCR was chosen for the FMD DNA chip assay. 3.3. Serotype specificity of probes in the FMD DNA chip To determine the ability of the FMD DNA chip to accurately type FMDV serotypes, RNA was extracted from 23 different strains representing all seven FMDV serotype and amplified by asymmetric PCR. To achieve efficient hybridization and high serotype specificity for the probes a number of conditions were compared and optimized which included hybridization temperature (35, 42, 55, and 75 C) and formamide concentration (30% and 50%) in the hybridization mixture. The best results were achieved with a hybridization temperature of 42 C and 30% formamide in the hybridization mixture (data not shown). The results of hybridization assays for each of the 23 different FMDV strains are presented in Table 2 and representative results of signal intensities obtained for seven strains are shown in Fig. 3. Each FMDV strain was tested at least twice on the FMD DNA chip containing duplicate probes. All strains were detected by the six FMD-common probes. For serotype-specific probes, it was observed that individual probes were differentially reactive with strains within a serotype. Seven serotype A strains were tested against a panel of serotype A-specific probes (n = 23) and all were detected by the panel. Due to the high genetic variability within serotype A, some strains were detected by a greater number of probes (24 Cruzeio, n = 11; Arg 87, n = 11) as compared to other strains (22 Iran/99, n = 2). All four serotype O strains were detected by a variable number of serotype O-specific probes. Similarly, strains for serotype C (n = 2), Asia 1 (n = 2), SAT 1 (n = 2), SAT 2 (n = 4), and SAT 3 (n = 2) were detected by serotype-specific probes. Non-specific binding of target sequences to non-homologous probes in the assay was not observed, indicating that the FMD DNA chip was highly specific as well as sensitive. Also printed on the FMD DNA chip, as controls, were 10 specific probes for type 1 and type 2 bovine viral diarrhoea virus (BVDV). As expected, none of the target sequences from the 23 FMDV strains was reactive with these probes. 3.4. Sensitivity of the FMD DNA chip RNA was extracted from 10-fold serial dilutions of the C1 Noville FMDV strain and the VP3-VP1-2A region was amplified by asymmetric PCR. The labelled target sequences for each dilution were hybridized to
478
M.K. Baxi et al. / The Veterinary Journal 172 (2006) 473–481
Table 2 Detection and typing of FMDV strains with the FMD DNA chip
1 2 3 4 5 6 7 8 9 10
Virus/serotypes
Total number of probes
Serotype A 24 Cruzeio
Iran 1/96
Arg 87
Arg 2/2001
Col/85
22 Iran/99
22 Iraq 24/64
Serotype A Serotype O Serotype C Serotype Asia 1 Serotype SAT1 Serotype SAT2 Serotype SAT3 FMD-Common BVDVI BVDII
23 28 15 15 22 34 12 06 05 05
11/23 – – – – – – 6/6 – –
5/23 – – – – – – 6/6 – –
11/23 – – – – – – 6/6 – –
6/23 – – – – – – 6/6 – –
4/23 – – – – – – 6/6 – –
2/23 – – – – – – 6/6 – –
9/23 – – – – – – 6/6 – –
Virus/serotypes
1 2 3 4 5 6 7 8 9 10
Serotype A Serotype O Serotype C Serotype Asia 1 Serotype SAT1 Serotype SAT2 Serotype SAT3 FMD-Common BVDVI BVDII
Total number of probes
Serotype O UKG 11/2001
1 Mansia
Taw 10/97
1 BF s/860
C1 Noville
C3 Resende
Pak 1/54
1 Shamir
23 28 15 15 22 34 12 06 05 05
– 11/28 – – – – – 6/6 – –
– 9/28 – – – – – 6/6 – –
– 8/28 – – – – – 6/6 – –
– 3/28 – – – – – 6/6 – –
– – 10/15 – – – – 6/6 – –
– – 12/15 – – – – 6/6 – –
– – – 10/15 – – – 6/6 – –
– – – 4/15 – – – 6/6 – –
Virus/serotypes
1 2 3 4 5 6 7 8 9 10
Serotype A Serotype O Serotype C Serotype Asia 1 Serotype SAT1 Serotype SAT2 Serotype SAT3 FMD-Common BVDVI BVDII
Serotype C
Serotype Asia 1
Total number of probes
Serotype SAT 1
Serotype SAT 2
Ken 4/98
Bot 1/68
Zim 5/81
Swa 1/69
Sau 1/2000
Zim 10/91
Bec 1/65
Zim 4/81
23 28 15 15 22 34 12 06 05 05
– – – – 6/22 – – 6/6 – –
– – – – 8/22 – – 6/6 – –
– – – – – 3/34 – 6/6 – –
– – – – – 2/34 – 6/6 – –
– – – – – 3/34 – 6/6 – –
– – – – – 4/34 – 6/6 – –
– – – – – – 7/12 6/6 – –
– – – – – – 3/12 6/6 – –
the FMD DNA chip to determine sensitivity. The lower limit of detection for the assay was 7.5 TCID50/mL with FMDV-common probes.
4. Discussion Microarray technology offers greater screening capabilities for pathogen detection and an attractive alternative to existing diagnostic methods. Depending on the availability of appropriate probe sets, microarrays permit the identification of several thousand microorganisms at the species, subspecies, or subtype level in a single assay. Although PCR-based assays are sensitive, specific, and rapid, they are not able to detect many pathogens simultaneously (Clewley, 2004). The co-application of PCR and microarray technology, PCR for its sensitivity and microarray for its high-throughput abil-
Serotype SAT 3
ity, results in a powerful diagnostic assay. Recently, a number of diagnostic microarray chips have been developed: a multiple viral pathogen microarray (Wang et al., 2002), a multiple-pathogen microarray for bioterrorism agents (Wilson et al., 2002), and microarrays for detection of enteroviruses (Shih et al., 2003), orthopoxviruses (Laassri et al., 2003), retroviruses (Seifarth et al., 2003) and certain plant viruses (Boonham et al., 2003; Lee et al., 2003). Most genetic-based methods for detection of FMDV target highly conserved regions of the genome such as the 5 0 UTR (Reid et al., 2002, 2003) and the non-structural 3D gene (Moonen et al., 2003). Similarly, serological methods are based on the detection of antibodies against highly conserved non-structural proteins (3ABC, 3D) (Clavijo et al., 2004; Mackay et al., 2001; Sorensen et al., 1998, 2005). To generate serotype-specific probes for microarray typing, highly conserved regions are not
M.K. Baxi et al. / The Veterinary Journal 172 (2006) 473–481
479
Serotype O - UKG 11/2001
Serotype A - 24 Cruzeio 70000 70000 60000 60000 50000 50000 40000
40000
+ 2SD
MFI
MFI
+ 2SD
30000
30000
20000
20000
10000
10000
0
0
0
20
40
60
100
80
120
140
0
160
20
40
Serotype C - C1 Noville
60
80
100
120
140
160
Serotype Asia1-Pak 1/54
60000
70000
50000
60000
50000
40000
MFI
MFI
40000
+ 2SD
30000
+ 2SD 30000
20000
20000 10000
a
10000
0
0
0
20
40
60
80
100
120
160
140
0
20
Serotype SAT1 - Bot 1/68
40
60
80
100
120
140
160
Serotype SAT2 - Zim 10/91 60000
70000
60000
50000
50000 40000
+ 2SD
MFI
MFI
40000
30000
30000
+ 2SD 20000
20000 10000
10000
0
0
0
20
40
60
80
100
120
140
160
0
20
40
60
80
100
120
140
160
Serotype SAT3 - Bec 1/65 50000
40000
30000
MFI
+ 2SD 20000
10000
b
0
0
20
40
60
80
100
120
140
160
Fig. 3. (a,b) Microarray results for the FMD DNA chip after hybridization with 7 FMDV strains (from Table 2) representing each serotype. The median fluorescent intensities (MFI) are plotted against FMDV oligonucleotide probes (1–155) and BVDV control probes (156–165). The mean MFI value for all probes (dotted line) and the positive ±2 SD cut-off value (dark line) are shown. Positive probes are highlighted in each graph; FMDVcommon ( ) and serotype-specific: A ( ), O ( ), C ( ), Asia 1 ( ), SAT 1 ( ), SAT 2 ( ) and SAT 3 ( ).
suitable as these regions do not allow sufficient differentiation. Thus, the less conserved VP3-VP1-2A region was chosen for probe design in this study. Previously, Alexandersen et al. (2000) chose the less conserved VP1 gene for generating specific probes and successfully typed a number of strains using a novel PCR-ELISA system. Because of the high heterogeneity within each serotype, we found
it was not possible to generate single probes to detect all members of an individual serotype. Thus, panels of serotype-specific probes were created. As expected because of genetic heterogeneity, individual probes varied in their ability to bind target sequences. The test developed in this study detected all strains and amplified target sequences showed an absence of
480
M.K. Baxi et al. / The Veterinary Journal 172 (2006) 473–481
binding to non-homologous probes (100% sensitivity and serotype-specificity). The analytical sensitivity of the test was comparable to PCR assays. However, since the assay was developed against a limited number of strains, further validation with additional FMDV strains and testing of biological samples is needed to better define the usefulness of the test. We have demonstrated the potential of an oligonucleotide microarray for detection and typing of FMDV strains. The assay, from the time of RNA extraction to serotype determination, can be performed in a single day. It would be difficult to accomplish the same task (serotype determination) in this time frame using conventional PCR and sequencing. Real-time PCR has the advantage of serotyping in less time, but seven assays (one for each serotype) are needed per sample. Panels of serotype-specific probes ensure a high redundancy of results in the microarray assay, thereby increasing confidence in serotype detection which is lacking in PCR-based assays. Redundancy is critical for viruses such as FMDV which mutate rapidly and lose their diagnostic markers. While conventional detection methods are limited in their ability to cope with rapidly mutating pathogens, the microarray allows the addition of a virtually unlimited number of oligonucleotide probes for identification of such infectious agents. Another potential advantage of a microarray chip is its potential use for subtype determination. The chip in this study contained some subtype-specific probes. However, additional subtype probes and refinement of the assay could extend its usefulness for phylogenetic and epidemiological studies. Such typing assays have been reported for influenza viruses (Li et al., 2001) and rotaviruses (Chizhikov et al., 2002). Because of the rapid spread of a virus such as FMDV, it is critical to identify quickly the causative agent of an outbreak. The FMD DNA chip is a significant step in this direction. Its use may provide rapid information for tracking the disease and for developing strategies of containment. It may also provide serotype information for vaccine selection during an outbreak. Our long term objectives are to further refine the FMD DNA chip and to develop a viral chip for all animal vesicular diseases (for example: vesicular stomatitis, swine vesicular disease, and vesicular exanthema) so that rapid differential diagnosis and response can be initiated when confronted with an unknown vesicular disease.
Acknowledgements We thank Dr. Paul Kitching for his helpful advice. This research was supported through funding from the Canadian Food Inspection Agency and the Chemical,
Biological, Radiological and Nuclear Research and Technology Initiative (CRTI) Project No. 0196RD.
References Alexandersen, S., Forsyth, M.A., Reid, S.M., Belsham, G.J., 2000. Development of reverse transcription-PCR (oligonucleotide probing) enzyme-linked immunosorbent assays for diagnosis and preliminary typing of foot-and-mouth disease: a new system using simple and aqueous-phase hybridization. Journal of Clinical Microbiology 38, 4604–4613. Amaral-Doel, C.M., Owen, N.E., Ferris, N.P., Kitching, R.P., Doel, T.R., 1993. Detection of foot-and-mouth disease viral sequences in clinical specimens and ethyleneimine-inactivated preparations by the polymerase chain reaction. Vaccine 11, 415–421. Boonham, N., Walsh, K., Smith, P., Madagan, K., Graham, I., Barker, I., 2003. Detection of potato viruses using microarray technology: towards a generic method for plant viral disease diagnosis. Journal of Virological Methods 108, 181– 187. Callens, M., De Clercq, K., 1997. Differentiation of the seven serotypes of foot-and-mouth disease virus by reverse transcriptase polymerase chain reaction. Journal of Virological Methods 67, 35–44. Chizhikov, V., Wagner, M., Ivshina, A., Hoshino, Y., Kapikian, A.Z., Chumakov, K., 2002. Detection and genotyping of human group A rotaviruses by oligonucleotide microarray hybridization. Journal of Clinical Microbiology 40, 2398–2407. Clavijo, A., Zhou, E.M., Hole, K., Galic, B., Kitching, P., 2004. Development and use of a biotinylated 3ABC recombinant protein in a solid-phase competitive ELISA for the detection of antibodies against foot-and-mouth disease virus. Journal of Virological Methods 20, 217–227. Clewley, J.P., 2004. A role for arrays in clinical virology: fact or fiction? Journal of Clinical Virology 29, 2–12. Coetzer, J.A.W., Thomson, G.R., Tustin, R.C., Kriek, N.P.J., 1994. Foot-and-mouth disease. In: Coetzer, J.A.W., Thomson, G.R., Tustin, R.C., Kriek, N.P.J. (Eds.), Infectious Disease of Livestock with Special Reference to Southern Africa. Oxford University Press, Cape Town, SA, pp. 825–852. Domingo, E., Escarmis, C., Baranowski, E., Ruiz-Jarabo, C.M., Carrillo, E., Nunez, J.I., Sobrino, F., 2003. Evolution of foot-andmouth disease virus. Virus Research 91, 47–63. Eisen, M.B., Brown, P.O., 1999. DNA arrays for analysis of gene expression. Methods in Enzymology 303, 179–205. Ferris, N.P., Dawson, M., 1988. Routine application of enzyme-linked immunosorbent assay in comparison with complement fixation for the diagnosis of foot-and-mouth and swine vesicular diseases. Veterinary Microbiology 16, 201–209. Knowles, N.J., Samuel, A.R., 2003. Molecular epidemiology of footand-mouth disease virus. Virus Research 91, 65–80. Laassri, M., Chizhikov, V., Mikheev, M., Shchelkunov, S., Chumakov, K., 2003. Detection and discrimination of orthopoxviruses using microarrays of immobilized oligonucleotides. Journal of Virological Methods 112, 67–78. Lee, G.P., Min, B.E., Kim, C.S., Choi, S.H., Harn, C.H., Kim, S.U., Ryu, K.H., 2003. Plant virus cDNA chip hybridization for detection and differentiation of four cucurbit-infecting Tobamoviruses. Journal of Virological Methods 110, 19–24. Li, J., Chen, S., Evans, D.H., 2001. Typing and subtyping influenza virus using DNA microarrays and multiplex reverse transcriptase PCR. Journal of Clinical Microbiology 39, 696–704. Mackay, D.K., Bulut, A.N., Rendle, T., Davidson, F., Ferris, N.P., 2001. A solid-phase competition ELISA for measuring antibody to
M.K. Baxi et al. / The Veterinary Journal 172 (2006) 473–481 foot-and-mouth disease virus. Journal of Virological Methods 97, 33–48. Moonen, P., Boonstra, J., van der Honing, R.H., Leendertse, C.B., Jacobs, L., Dekker, A., 2003. Validation of a LightCycler-based reverse transcription polymerase chain reaction for the detection of foot-and-mouth disease virus. Journal of Virological Methods 113, 35–41. Oleksiewicz, M.B., Donaldson, A.I., Alexandersen, S., 2001. Development of a novel real-time RT-PCR assay for quantitation of foot-and-mouth disease virus in diverse porcine tissues. Journal of Virological Methods 92, 23–35. Poirot, O., Suhre, K., Abergel, C., OÕToole, E., Notredame, C., 2004. 3DCoffee@igs: a web server for combining sequences and structures into a multiple sequence alignment. Nucleic Acids Research 32, W37–W40. Rasmussen, T.B., Uttenthal, A., de Stricker, K., Belak, S., Storgaard, T., 2003. Development of a novel quantitative real-time RT-PCR assay for the simultaneous detection of all serotypes of foot-and-mouth disease virus. Archives of Virology 148, 2005– 2021. Reid, S.M., Ferris, N.P., Hutchings, G.H., Zhang, Z., Belsham, G.J., Alexandersen, S., 2002. Detection of all seven serotypes of footand-mouth disease virus by real-time, fluorogenic reverse transcription polymerase chain reaction assay. Journal of Virological Methods 105, 67–80. Reid, S.M., Forsyth, M.A., Hutchings, G.H., Ferris, N.P., 1998. Comparison of reverse transcription polymerase chain reaction, enzyme linked immunosorbent assay and virus isolation for the routine diagnosis of foot-and-mouth disease. Journal of Virological Methods 70, 213–217. Reid, S.M., Grierson, S.S., Ferris, N.P., Hutchings, G.H., Alexandersen, S., 2003. Evaluation of automated RT-PCR to accelerate the laboratory diagnosis of foot-and-mouth disease virus. Journal of Virological Methods 107, 129–139. Rouillard, J.M., Zuker, M., Gulari, E., 2003. OligoArray 2.0: design of oligonucleotide probes for DNA microarrays using a thermodynamic approach. Nucleic Acids Research 31, 3057–3062. Seifarth, W., Spiess, B., Zeilfelder, U., Speth, C., Hehlmann, R., LeibMosch, C., 2003. Assessment of retroviral activity using a universal retrovirus chip. Journal of Virological Methods 112, 79–91.
481
Shih, S.R., Wang, Y.W., Chen, G.W., Chang, L.Y., Lin, T.Y., Tseng, M.C., Chiang, C., Tsao, K.C., Huang, C.G., Shio, M.R., Tai, J.H., Wang, S.H., Kuo, R.L., Liu, W.T., 2003. Serotype-specific detection of enterovirus 71 in clinical specimens by DNA microchip array. Journal of Virological Methods 111, 55–60. Sorensen, K.J., de Stricker, K., Dyrting, K.C., Grazioli, S., Haas, B., 2005. Differentiation of foot-and-mouth disease virus infected animals from vaccinated animals using a blocking ELISA based on baculovirus expressed FMDV 3ABC antigen and a 3ABC monoclonal antibody. Archives of Virology 150, 805–814. Sorensen, K.J., Madsen, K.G., Madsen, E.S., Salt, J.S., Nqindi, J., Mackay, D.K., 1998. Differentiation of infection from vaccination in foot-and-mouth disease by the detection of antibodies to the non-structural proteins 3D, 3AB and 3ABC in ELISA using antigens expressed in baculovirus. Archives of Virology 143, 1461– 1476. Suryanarayana, V., Madanamohan, B., Bist, P., Natarajan, C., Tratschin, J.D., 1999. Serotyping of foot-and-mouth disease virus by antigen capture reverse transcriptase/polymerase chain reaction. Journal of Virological Methods 80, 45–52. Thompson, D., Muriel, P., Russell, D., Osborne, P., Bromley, A., Rowland, M., Creigh-Tyte, S., Brown, C., 2002. Economic costs of the foot and mouth disease outbreak in the United Kingdom in 2001. Revue Scientifique et Technique (International Office of Epizootics) 21, 675–687. Vangrysperre, W., De Clercq, K., 1996. Rapid and sensitive polymerase chain reaction based detection and typing of foot-and-mouth disease virus in clinical samples and cell culture isolates, combined with a simultaneous differentiation with other genomically and/or symptomatically related viruses. Archives of Virology 141, 331– 344. Wang, D., Coscoy, L., Zylberberg, M., Avila, P.C., Boushey, H.A., Ganem, D., DeRisi, J.L., 2002. Microarray-based detection and genotyping of viral pathogens. Proceedings of the National Academy of Science of the United States of America 99, 15687– 15692. Wilson, W.J., Strout, C.L., DeSantis, T.Z., Stilwell, J.L., Carrano, A.V., Andersen, G.L., 2002. Sequence-specific identification of 18 pathogenic microorganisms using microarray technology. Molecular and Cellular Probes 16, 119–127.
Microarray-based detection and typing of foot-and-mouth disease virus Mohit K. Baxi
a,*
, Shailja Baxi a, Alfonso Clavijo b, Kimberly M. Burton a, Dirk Deregt
a
a
b
Virology Section, Lethbridge Laboratory (Animal Diseases Research Institute), Canadian Food Inspection Agency, P.O. Box 640, Lethbridge, Alberta, Canada T1J 3Z4 National Center for Foreign Animal Diseases, Canadian Food Inspection Agency, Suite T2300, 1015 Arlington Street, Winnipeg, Manitoba, Canada R3E 3M4
Abstract Foot-and-mouth disease virus (FMDV) is the most economically important veterinary pathogen because of its highly infectious nature and the devastating effects the virus has on the livestock industry. Rapid diagnostic methods are needed for detection and typing of FMDV serotypes and differentiation from other viruses causing vesicular diseases. We developed a microarray-based test that uses a FMD DNA chip containing 155 oligonucleotide probes, 35–45 base pair (bp) long, virus-common and serotype-specific, designed from the VP3-VP1-2A region of the genome. A set of two forward primers and one reverse primer were also designed to allow amplification of approximately 1100 bp of target sequences from this region. The amplified target was labelled with AlexaFluor 546 dye and applied to the FMD DNA chip. A total of 23 different FMDV strains representing all seven serotypes were detected and typed by the FMD DNA chip. Microarray technology offers a unique capability to identify multiple pathogens in a single chip. Crown Copyright 2005 Published by Elsevier Ltd. All rights reserved. Keywords: Foot-and-mouth disease; FMD DNA chip; Microarray; Diagnosis
1. Introduction Foot-and-mouth disease (FMD) is a severe, clinically acute, vesicular disease of cloven-hoofed animals including domesticated ruminants, swine, and more than 70 wildlife species (Coetzer et al., 1994). Foot-and-mouth disease virus (FMDV), a member of the Apthovirus genus of the family Picornaviridae, is a single-stranded RNA virus with a plus-sense genome of about 8500 nucleotides. Seven different serotypes of the virus exist: A, O, C, Asia 1 and South African Territories 1 (SAT 1), SAT 2 and SAT 3. A large number of subtypes have *
Corresponding author. Tel.: +1 403 382 5589; fax: +1 403 381 1202. E-mail address: [email protected] (M.K. Baxi).
also evolved resulting in a high degree of genetic variability within each serotype (Domingo et al., 2003; Knowles and Samuel, 2003). FMDV-infected animals shed large amounts of virus which spread to other animals by direct contact, inhalation of virus-infected aerosols, or ingestion of contaminated material, including water and feed. Cattle that recover from the disease may still shed the virus for a considerable period of time and be the source of infection for other animals. FMD outbreaks usually result in very large economic losses. For instance, the cost of the 2001 FMD outbreak in the UK amounted to £6 billion (€9 bn) (Thompson et al., 2002). FMDV is the most infectious virus known and is considered the most dangerous foreign animal disease agent that could be used for agrobioterrorism. Therefore, it is important
1090-0233/$ - see front matter Crown Copyright 2005 Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.tvjl.2005.07.007
474
M.K. Baxi et al. / The Veterinary Journal 172 (2006) 473–481
to develop rapid, sensitive, and specific diagnostic methods which provide immediate and detailed information should an outbreak/attack occur. Routine methods for detection of FMDV are virus isolation in cell culture and ELISA (Ferris and Dawson, 1988). Although these tests are highly sensitive, they are also time consuming and labour intensive. In recent years, various PCR-based methods have been developed for detection of viral RNA from clinical samples. These assays employ universal (common) primers (AmaralDoel et al., 1993; Reid et al., 1998) and serotype-specific primers for detection and typing (Callens and De Clercq, 1997; Vangrysperre and De Clercq, 1996). Other PCR-based methods developed for FMDV are: PCRELISA (Alexandersen et al., 2000) and antigen capture RT-PCR (Suryanarayana et al., 1999). More recently, real-time RT-PCR assays have been developed by various groups (Moonen et al., 2003; Oleksiewicz et al., 2001; Rasmussen et al., 2003; Reid et al., 2002; Reid et al., 2003). PCR-based assays are rapid, sensitive, and specific. However, a limitation of these assays is the inability to simultaneously amplify in a multiplex fashion all serotypes or detect many viral pathogens, which may be critical, especially in the early stages of an outbreak or attack with an unknown infectious agent. To address the limitations of existing detection methods to type FMDV, we developed an oligonucleotide microarray ‘‘FMD DNA Chip’’. The chip, which contains 155 oligonucleotide probes can simultaneously detect and type all seven FMDV serotypes. To increase the overall robustness of the assay, each serotype is identified by several probes. Besides its speed and sensitivity, advantages of a chip-based test are its ability to screen for multiple pathogens simultaneously and to readily cope with the evolution of new variant strains.
2. Material and methods 2.1. Virus and cell culture Various FMDV strains were used to infect primary lamb kidney (LK) cells cultured in EagleÕs medium (Alpha modification), containing 1% L-glutamine, 10% fetal bovine serum, 10,000 U/mL penicillin, 10,000 lg/mL streptomycin, and 10,000 U/mL nystatin (0.1%). Viruses were harvested 16 h post-infection and titrated in LK cells. 2.2. Probe design and slide printing Complete viral sequences for FMDV strains were obtained from the NCBI GenBank and stored in a database. The compiled database comprised of complete
and partial FMD viral sequences belonging to serotypes A (48 strains), O (49 strains), C (9 strains), Asia 1 (11 strains), SAT 1 (12 strains), SAT 2 (15 strains) and SAT 3 (5 strains). To generate FMDV-common and serotype-specific probes, the program OligoArray 2.0 (Rouillard et al., 2003), available at the Canadian Bioinformatics Resource Web Site (http://cbr-rbc.nrccnrc.gc.ca), was used. This relies on two programs, BLAST (ftp://ftp.ncbi.nih/gov/blast) and MFOLD (www.bioinfo.rpi.edu/~zukerm/rna/mfold-3.1.html) to generate highly specific probes. Prior to running OligoArray 2.0, all complete FMDV sequences were saved in a FMD database in the FASTA format. To ensure the specificity of the probe for its target, the program computes the input sequence against the FMDV database by using the BLAST program for sequence similarity and uniqueness. Specific probes were selected against a set of parameters: probe length (35–50 bp), Tm (80–85), GC content (40–50%), and those governing secondary structure formation and self- and crosshybridization. Probes that met the appropriate criteria were selected and used in the development of FMD DNA chip. A list of probes used in this study is available as Appendix 1 found at http://www.inspection.gc.ca/ english/sci/tech/publish/baxi2005e.shtml. Oligonucleotides were synthesized commercially (Invitrogen Inc.) and were suspended in 50% DMSO at a concentration of 10 nmol/lL. They were printed in duplicate on poly-l-lysine coated slides by SMP3 microspotting pins using the VersArray ChipWriter printer (Bio-Rad Laboratories). The average size of each spot was 100 lm and the distance between spots was 400 lm. Printed slides were stored at room temperature until used. 2.3. RNA extraction Total RNA was extracted from FMDV-infected cells in TriPure reagent (Roche Diagnostics Canada) according to the manufacturerÕs protocol. The RNA pellet was suspended in 25 lL of DEPC-treated water and stored at 80 C. 2.4. PCR amplification of FMDV serotypes All available FMDV genomic sequences were aligned using the program T-coffee (Poirot et al., 2004). A set of FMDV-common primers were designed to amplify the VP3-VP1-2A region of the genome (Fig. 1a). To amplify this region for serotypes A, O, C, and Asia 1, the primer set VP3-Fwd1 and 2A-Rev was designed (Fig. 1b). For serotypes SAT 1, 2, and 3 the primer set VP3-Fwd2 and 2A-Rev was designed. One-step RT-PCR was standardized for the amplification of the VP3-VP1-2A region using the SuperScript III one-step RT-PCR Kit (Invitrogen). The RT-PCR
M.K. Baxi et al. / The Veterinary Journal 172 (2006) 473–481
475
Probes). Briefly, the PCR products were extracted with equal volumes of phenol–chloroform–isoamylalcohol and precipitated with 2 vol of absolute ethanol. The final pellet was dissolved in 20 lL of Alexa Fluor labelling buffer. DNA was denatured at 95 C for 5 min and then put on ice. Alexa Fluor 546 dye was added at 5 lL per 1 lg of DNA and incubated at 80 C for 15 min. The labelling reaction was stopped by placing on ice. Unlabelled dye was removed using DNA clean up columns (Zymo Research Corp). Finally, the labelled DNA was eluted in 20 lL of elution buffer and stored at 20 C until used. The relative efficiency of the labelling reaction was determined as follows: pmol of Dye in labelled sample ¼ ðA546 =104000Þ ðDFÞ ðzLÞ ðwÞ 1012 ; where A546 is the absorbance at 546, DF is the dilution factor, zL is the volume of sample after purification and w is the optical path in cuvette. Labelling efficiency ¼ pmol of Dye in labelled sample =lg of labelled sample. The labelling efficiency was found to be 40–80:1. Fig. 1. PCR amplification of VP3-VP1-2A region. (a) Schematic representation of the FMDV genome and the location of genes for structural proteins (VP4, VP2, VP3 and VP1) is highlighted. The location of forward and reverse primers used to amplify the VP3-VP12A region is shown. (b) List of PCR primers used for amplification. (c) PCR amplification of FMDV, serotypes A (lane 1), O (lane 2), C (lane 3) and Asia 1 (lane 4) generated a product of 1073 bp, whereas amplification of serotypes SAT 1 (lane 5), SAT 2 (lane 6) and SAT 3 (lane 7) generated a product of 1109 bp. 1 kb DNA ladder (lane M) was used for sizing of DNA fragments. PCR products were analyzed in a 1% agarose gel and stained with ethidium bromide.
mixture contained 2· reaction mix (3.2 mM MgSO4, 0.4 mM of each dNTPs), 10 lM of each primer, and a mixture of Superscript III reverse transcriptase and Platinum Taq DNA polymerase enzymes. PCR amplification was performed in an ABI 9700 thermal cycler (Applied Biosystems Inc.) using the following parameters: 50 C for 30 min for reverse transcription; 94 C for 2 min, followed by 40 cycles of 30 s at 94 C, 1 min at 50 C and 1 min at 68 C; and a final extension time for 5 min at 68 C. The resulting PCR products were analyzed by electrophoresis in a 1% agarose gel. 2.5. Synthesis of fluorescent labelled product To generate labelled products, symmetric and asymmetric RT-PCR were performed. The conditions used for asymmetric RT-PCR were the same as those for regular (symmetric) PCR except the forward and reverse primers were added in 1:10 molar concentration. The RT-PCR products were labelled using the ULYSIS Alexa Fluor 546 nucleic acid labelling kit (Molecular
2.6. Microarray hybridization Before using the printed slides for microarray experiments, they were rehydrated with 0.5· SSC to allow the oligonucleotides to spread evenly throughout the spot. Following rehydration, the oligonucleotides were UVcross-linked onto the glass slides at 150 mJ. Poly-l-lysine arrays require that exposed amines be blocked prior to hybridization to prevent non-specific binding of target sequences. To block exposed amines, arrays were incubated in pre-hybridization solution (5· SSC, 0.1% SDS, 0.1% BSA) at 50 C for 30 min. Hybridization of the AlexaFluor-labelled target sequences (or target DNA) with oligonucleotide probes on the FMD DNA chip was performed as previously described with some modifications (Eisen and Brown, 1999). Briefly, 30 pmol of AlexaFluor-546 labelled target DNA was added to the hybridization mixture (30% formamide, 3· SSC, 0.1% SDS and 0.1 ng/mL of salmon sperm DNA). The mixture was preheated at 95 C for 2 min to denature the dsDNA and placed on ice. The hybridization mixture was applied to the array, and the array was covered with a lifter cover slip to prevent probe drying. The slides were placed in the hybridization chamber (Tele-Chem International, Inc.) and incubated in a water bath at 42 C overnight. Following hybridization, slides were washed at room temperature with 0.5· SSC containing 0.1% SDS for 2 min followed by a second wash with 0.05· SSC for 1 min. The array slides were dried by spinning at 600 rpm for 5 min and were then scanned.
476
M.K. Baxi et al. / The Veterinary Journal 172 (2006) 473–481
2.7. Data visualisation and analysis Hybridized microarrays were imaged using the Axon 4000B scanner (Axon Inc.) and the images were analyzed using the GenePix Pro software (Axon Inc.). Microarray data was represented as median pixel intensity at a wavelength of 532 nm with background subtracted (F532median B532median), and spots which had a fluorescent intensity greater than 2 SD (cut-off value) were considered positive.
3. Results 3.1. Design of the FMD DNA chip To develop the FMD DNA chip, oligonucleotide probes were designed from the VP3-VP1-2A region of the genome, as this region has only 60–70% homology between the seven FMDV serotypes which allows the generation of serotype-specific probes. There are also areas of high homology in this region which allows the design of FMDV-common chip probes and PCR primers. A total of 172 oligonucleotide probes of 35–45 bp were designed, of which 155 were selected to be printed on the FMD DNA chip (Table 1). Of these, 23 are specific to serotype A, 28 to serotype O, 15 each to serotypes C and Asia 1, 22 to serotype SAT 1, 34 to
serotype SAT 2, and 12 to serotype SAT 3. In addition, the chip contains six FMDV-common probes. As there is a high genetic variability within each serotype, probes were designed specifically to a number of subtypes or topotypes (geographically restricted subtypes) within each serotype (Knowles and Samuel, 2003). Serotype A is considered to be the most diverse group, both antigenically and genetically. Nineteen probes were designed to the commonly prevalent serotype A subtypes: Europe–South America (Euro–SA; 12 probes), Asia (5 probes) and Africa (2 probes). Similarly, 26 subtype-specific probes were designed to serotype O: Middle East–South America (ME–SA/Pan-Asia strains; 16 probes) and Cathay (10 probes). Thirteen probes designed against serotype C covered subtypes C1 (5 probes) and C3/C5 (8 probes), and 13 probes designed against Asia 1 subtypes covered genotype I (10 probes) and genotype II (3 probes). SAT 1, 2, and 3 serotypes have been shown to have a low level of nucleotide conservation, thereby making it necessary to generate probes that are specific to various topotypes within each serotype. Four topotype-specific probes were designed against SAT 1 topotypes: I (3 probes) and II (1 probe). A panel of 22 topotype-specific probes were also designed for serotype SAT 2 as shown in Table 1. Thus, a large number of specific oligonucleotide probes were designed to cover the genetically distinct groups of FMDV.
Table 1 Summary of oligonucleotide probes designed from the VP3-VP1-2A region Serotype
OligoÕs designed
OligoÕs selected
1
A
25
23
2
O
31
28
3
C
16
15
4
Asia 1
16
15
5
SAT1
28
22
6
SAT2
38
34
7 8
SAT3 FMD-Common Total
18 6 172
12 6 155
Subtypes Common Euro-SA Asia Africa Common Pan-Asia Cathy Common C1 C3/C5 Common Genotype I Genotype II Common Topotype I Topotype II Common Topotype A Topotype B Topotype C Topotype D Topotype E Topotype F Topotype G Common
4 12 5 2 2 16 10 2 5 8 2 10 3 18 3 1 12 3 1 4 4 3 4 3 12
M.K. Baxi et al. / The Veterinary Journal 172 (2006) 473–481
3.2. PCR amplification of target sequences Since there is a high degree of genetic dissimilarity in the VP3-VP1-2A region between the serotype A, O, C and Asia 1 group and the serotype SAT 1, 2, and 3 group, two forward primers (VP3-Fwd1 and VP3Fwd2) and one common reverse primer (2A-Rev) were designed to amplify target sequences of both groups. After PCR amplification, a product of 1073 bp was generated for serotypes A, O, C, and Asia 1 and a 1109 bp product was generated for serotypes SAT 1, 2, and 3 as expected (Fig. 1c). No products were generated following PCR amplification of negative controls (BVDV RNA and water). Two different PCR-based methods were used to amplify target RNA sequences: symmetric PCR to generate dsDNA PCR products and asymmetric PCR to generate ssDNA PCR products. The products were then labelled with Alexa-fluor 546 dye and applied on the FMD DNA chip. As shown in Fig. 2, the labelled asymmetric PCR product generated fluorescent intensities that were much higher than those achieved with the symmetric PCR product. The number of target-probe reactions detected positive were also greater using the asymmetric PCR product as compared to the symmetric PCR product. The abundance of ssDNA produced by asymmetric PCR provides a high amount of complimentary labelled target DNA available to be bound to the oligonucleo-
Fig. 2. Fluorescent intensities of microarray spots on the FMD DNA chip following hybridization with: (a) symmetric and (b) asymmetric PCR products generated from FMDV strain A-24 Cruzeiro. The variable fluorescent intensities are visualised as green spots on scanned images at wavelength 532 nm. (c) Comparison of fluorescent intensities (y-axis) obtained after hybridization with symmetric and asymmetric PCR products with various FMDV-specific probes (x-axis) on the FMD DNA chip. The fluorescent intensities were quantified using GenePix Pro software.
477
tide probes, thus generating high fluorescent intensities. In the case of symmetric PCR, probable self-reannealing of the labelled dsDNA product decreases the number of target sequences that can bind to oligonucleotide probes resulting in lower fluorescent intensities. Thus, asymmetric PCR was chosen for the FMD DNA chip assay. 3.3. Serotype specificity of probes in the FMD DNA chip To determine the ability of the FMD DNA chip to accurately type FMDV serotypes, RNA was extracted from 23 different strains representing all seven FMDV serotype and amplified by asymmetric PCR. To achieve efficient hybridization and high serotype specificity for the probes a number of conditions were compared and optimized which included hybridization temperature (35, 42, 55, and 75 C) and formamide concentration (30% and 50%) in the hybridization mixture. The best results were achieved with a hybridization temperature of 42 C and 30% formamide in the hybridization mixture (data not shown). The results of hybridization assays for each of the 23 different FMDV strains are presented in Table 2 and representative results of signal intensities obtained for seven strains are shown in Fig. 3. Each FMDV strain was tested at least twice on the FMD DNA chip containing duplicate probes. All strains were detected by the six FMD-common probes. For serotype-specific probes, it was observed that individual probes were differentially reactive with strains within a serotype. Seven serotype A strains were tested against a panel of serotype A-specific probes (n = 23) and all were detected by the panel. Due to the high genetic variability within serotype A, some strains were detected by a greater number of probes (24 Cruzeio, n = 11; Arg 87, n = 11) as compared to other strains (22 Iran/99, n = 2). All four serotype O strains were detected by a variable number of serotype O-specific probes. Similarly, strains for serotype C (n = 2), Asia 1 (n = 2), SAT 1 (n = 2), SAT 2 (n = 4), and SAT 3 (n = 2) were detected by serotype-specific probes. Non-specific binding of target sequences to non-homologous probes in the assay was not observed, indicating that the FMD DNA chip was highly specific as well as sensitive. Also printed on the FMD DNA chip, as controls, were 10 specific probes for type 1 and type 2 bovine viral diarrhoea virus (BVDV). As expected, none of the target sequences from the 23 FMDV strains was reactive with these probes. 3.4. Sensitivity of the FMD DNA chip RNA was extracted from 10-fold serial dilutions of the C1 Noville FMDV strain and the VP3-VP1-2A region was amplified by asymmetric PCR. The labelled target sequences for each dilution were hybridized to
478
M.K. Baxi et al. / The Veterinary Journal 172 (2006) 473–481
Table 2 Detection and typing of FMDV strains with the FMD DNA chip
1 2 3 4 5 6 7 8 9 10
Virus/serotypes
Total number of probes
Serotype A 24 Cruzeio
Iran 1/96
Arg 87
Arg 2/2001
Col/85
22 Iran/99
22 Iraq 24/64
Serotype A Serotype O Serotype C Serotype Asia 1 Serotype SAT1 Serotype SAT2 Serotype SAT3 FMD-Common BVDVI BVDII
23 28 15 15 22 34 12 06 05 05
11/23 – – – – – – 6/6 – –
5/23 – – – – – – 6/6 – –
11/23 – – – – – – 6/6 – –
6/23 – – – – – – 6/6 – –
4/23 – – – – – – 6/6 – –
2/23 – – – – – – 6/6 – –
9/23 – – – – – – 6/6 – –
Virus/serotypes
1 2 3 4 5 6 7 8 9 10
Serotype A Serotype O Serotype C Serotype Asia 1 Serotype SAT1 Serotype SAT2 Serotype SAT3 FMD-Common BVDVI BVDII
Total number of probes
Serotype O UKG 11/2001
1 Mansia
Taw 10/97
1 BF s/860
C1 Noville
C3 Resende
Pak 1/54
1 Shamir
23 28 15 15 22 34 12 06 05 05
– 11/28 – – – – – 6/6 – –
– 9/28 – – – – – 6/6 – –
– 8/28 – – – – – 6/6 – –
– 3/28 – – – – – 6/6 – –
– – 10/15 – – – – 6/6 – –
– – 12/15 – – – – 6/6 – –
– – – 10/15 – – – 6/6 – –
– – – 4/15 – – – 6/6 – –
Virus/serotypes
1 2 3 4 5 6 7 8 9 10
Serotype A Serotype O Serotype C Serotype Asia 1 Serotype SAT1 Serotype SAT2 Serotype SAT3 FMD-Common BVDVI BVDII
Serotype C
Serotype Asia 1
Total number of probes
Serotype SAT 1
Serotype SAT 2
Ken 4/98
Bot 1/68
Zim 5/81
Swa 1/69
Sau 1/2000
Zim 10/91
Bec 1/65
Zim 4/81
23 28 15 15 22 34 12 06 05 05
– – – – 6/22 – – 6/6 – –
– – – – 8/22 – – 6/6 – –
– – – – – 3/34 – 6/6 – –
– – – – – 2/34 – 6/6 – –
– – – – – 3/34 – 6/6 – –
– – – – – 4/34 – 6/6 – –
– – – – – – 7/12 6/6 – –
– – – – – – 3/12 6/6 – –
the FMD DNA chip to determine sensitivity. The lower limit of detection for the assay was 7.5 TCID50/mL with FMDV-common probes.
4. Discussion Microarray technology offers greater screening capabilities for pathogen detection and an attractive alternative to existing diagnostic methods. Depending on the availability of appropriate probe sets, microarrays permit the identification of several thousand microorganisms at the species, subspecies, or subtype level in a single assay. Although PCR-based assays are sensitive, specific, and rapid, they are not able to detect many pathogens simultaneously (Clewley, 2004). The co-application of PCR and microarray technology, PCR for its sensitivity and microarray for its high-throughput abil-
Serotype SAT 3
ity, results in a powerful diagnostic assay. Recently, a number of diagnostic microarray chips have been developed: a multiple viral pathogen microarray (Wang et al., 2002), a multiple-pathogen microarray for bioterrorism agents (Wilson et al., 2002), and microarrays for detection of enteroviruses (Shih et al., 2003), orthopoxviruses (Laassri et al., 2003), retroviruses (Seifarth et al., 2003) and certain plant viruses (Boonham et al., 2003; Lee et al., 2003). Most genetic-based methods for detection of FMDV target highly conserved regions of the genome such as the 5 0 UTR (Reid et al., 2002, 2003) and the non-structural 3D gene (Moonen et al., 2003). Similarly, serological methods are based on the detection of antibodies against highly conserved non-structural proteins (3ABC, 3D) (Clavijo et al., 2004; Mackay et al., 2001; Sorensen et al., 1998, 2005). To generate serotype-specific probes for microarray typing, highly conserved regions are not
M.K. Baxi et al. / The Veterinary Journal 172 (2006) 473–481
479
Serotype O - UKG 11/2001
Serotype A - 24 Cruzeio 70000 70000 60000 60000 50000 50000 40000
40000
+ 2SD
MFI
MFI
+ 2SD
30000
30000
20000
20000
10000
10000
0
0
0
20
40
60
100
80
120
140
0
160
20
40
Serotype C - C1 Noville
60
80
100
120
140
160
Serotype Asia1-Pak 1/54
60000
70000
50000
60000
50000
40000
MFI
MFI
40000
+ 2SD
30000
+ 2SD 30000
20000
20000 10000
a
10000
0
0
0
20
40
60
80
100
120
160
140
0
20
Serotype SAT1 - Bot 1/68
40
60
80
100
120
140
160
Serotype SAT2 - Zim 10/91 60000
70000
60000
50000
50000 40000
+ 2SD
MFI
MFI
40000
30000
30000
+ 2SD 20000
20000 10000
10000
0
0
0
20
40
60
80
100
120
140
160
0
20
40
60
80
100
120
140
160
Serotype SAT3 - Bec 1/65 50000
40000
30000
MFI
+ 2SD 20000
10000
b
0
0
20
40
60
80
100
120
140
160
Fig. 3. (a,b) Microarray results for the FMD DNA chip after hybridization with 7 FMDV strains (from Table 2) representing each serotype. The median fluorescent intensities (MFI) are plotted against FMDV oligonucleotide probes (1–155) and BVDV control probes (156–165). The mean MFI value for all probes (dotted line) and the positive ±2 SD cut-off value (dark line) are shown. Positive probes are highlighted in each graph; FMDVcommon ( ) and serotype-specific: A ( ), O ( ), C ( ), Asia 1 ( ), SAT 1 ( ), SAT 2 ( ) and SAT 3 ( ).
suitable as these regions do not allow sufficient differentiation. Thus, the less conserved VP3-VP1-2A region was chosen for probe design in this study. Previously, Alexandersen et al. (2000) chose the less conserved VP1 gene for generating specific probes and successfully typed a number of strains using a novel PCR-ELISA system. Because of the high heterogeneity within each serotype, we found
it was not possible to generate single probes to detect all members of an individual serotype. Thus, panels of serotype-specific probes were created. As expected because of genetic heterogeneity, individual probes varied in their ability to bind target sequences. The test developed in this study detected all strains and amplified target sequences showed an absence of
480
M.K. Baxi et al. / The Veterinary Journal 172 (2006) 473–481
binding to non-homologous probes (100% sensitivity and serotype-specificity). The analytical sensitivity of the test was comparable to PCR assays. However, since the assay was developed against a limited number of strains, further validation with additional FMDV strains and testing of biological samples is needed to better define the usefulness of the test. We have demonstrated the potential of an oligonucleotide microarray for detection and typing of FMDV strains. The assay, from the time of RNA extraction to serotype determination, can be performed in a single day. It would be difficult to accomplish the same task (serotype determination) in this time frame using conventional PCR and sequencing. Real-time PCR has the advantage of serotyping in less time, but seven assays (one for each serotype) are needed per sample. Panels of serotype-specific probes ensure a high redundancy of results in the microarray assay, thereby increasing confidence in serotype detection which is lacking in PCR-based assays. Redundancy is critical for viruses such as FMDV which mutate rapidly and lose their diagnostic markers. While conventional detection methods are limited in their ability to cope with rapidly mutating pathogens, the microarray allows the addition of a virtually unlimited number of oligonucleotide probes for identification of such infectious agents. Another potential advantage of a microarray chip is its potential use for subtype determination. The chip in this study contained some subtype-specific probes. However, additional subtype probes and refinement of the assay could extend its usefulness for phylogenetic and epidemiological studies. Such typing assays have been reported for influenza viruses (Li et al., 2001) and rotaviruses (Chizhikov et al., 2002). Because of the rapid spread of a virus such as FMDV, it is critical to identify quickly the causative agent of an outbreak. The FMD DNA chip is a significant step in this direction. Its use may provide rapid information for tracking the disease and for developing strategies of containment. It may also provide serotype information for vaccine selection during an outbreak. Our long term objectives are to further refine the FMD DNA chip and to develop a viral chip for all animal vesicular diseases (for example: vesicular stomatitis, swine vesicular disease, and vesicular exanthema) so that rapid differential diagnosis and response can be initiated when confronted with an unknown vesicular disease.
Acknowledgements We thank Dr. Paul Kitching for his helpful advice. This research was supported through funding from the Canadian Food Inspection Agency and the Chemical,
Biological, Radiological and Nuclear Research and Technology Initiative (CRTI) Project No. 0196RD.
References Alexandersen, S., Forsyth, M.A., Reid, S.M., Belsham, G.J., 2000. Development of reverse transcription-PCR (oligonucleotide probing) enzyme-linked immunosorbent assays for diagnosis and preliminary typing of foot-and-mouth disease: a new system using simple and aqueous-phase hybridization. Journal of Clinical Microbiology 38, 4604–4613. Amaral-Doel, C.M., Owen, N.E., Ferris, N.P., Kitching, R.P., Doel, T.R., 1993. Detection of foot-and-mouth disease viral sequences in clinical specimens and ethyleneimine-inactivated preparations by the polymerase chain reaction. Vaccine 11, 415–421. Boonham, N., Walsh, K., Smith, P., Madagan, K., Graham, I., Barker, I., 2003. Detection of potato viruses using microarray technology: towards a generic method for plant viral disease diagnosis. Journal of Virological Methods 108, 181– 187. Callens, M., De Clercq, K., 1997. Differentiation of the seven serotypes of foot-and-mouth disease virus by reverse transcriptase polymerase chain reaction. Journal of Virological Methods 67, 35–44. Chizhikov, V., Wagner, M., Ivshina, A., Hoshino, Y., Kapikian, A.Z., Chumakov, K., 2002. Detection and genotyping of human group A rotaviruses by oligonucleotide microarray hybridization. Journal of Clinical Microbiology 40, 2398–2407. Clavijo, A., Zhou, E.M., Hole, K., Galic, B., Kitching, P., 2004. Development and use of a biotinylated 3ABC recombinant protein in a solid-phase competitive ELISA for the detection of antibodies against foot-and-mouth disease virus. Journal of Virological Methods 20, 217–227. Clewley, J.P., 2004. A role for arrays in clinical virology: fact or fiction? Journal of Clinical Virology 29, 2–12. Coetzer, J.A.W., Thomson, G.R., Tustin, R.C., Kriek, N.P.J., 1994. Foot-and-mouth disease. In: Coetzer, J.A.W., Thomson, G.R., Tustin, R.C., Kriek, N.P.J. (Eds.), Infectious Disease of Livestock with Special Reference to Southern Africa. Oxford University Press, Cape Town, SA, pp. 825–852. Domingo, E., Escarmis, C., Baranowski, E., Ruiz-Jarabo, C.M., Carrillo, E., Nunez, J.I., Sobrino, F., 2003. Evolution of foot-andmouth disease virus. Virus Research 91, 47–63. Eisen, M.B., Brown, P.O., 1999. DNA arrays for analysis of gene expression. Methods in Enzymology 303, 179–205. Ferris, N.P., Dawson, M., 1988. Routine application of enzyme-linked immunosorbent assay in comparison with complement fixation for the diagnosis of foot-and-mouth and swine vesicular diseases. Veterinary Microbiology 16, 201–209. Knowles, N.J., Samuel, A.R., 2003. Molecular epidemiology of footand-mouth disease virus. Virus Research 91, 65–80. Laassri, M., Chizhikov, V., Mikheev, M., Shchelkunov, S., Chumakov, K., 2003. Detection and discrimination of orthopoxviruses using microarrays of immobilized oligonucleotides. Journal of Virological Methods 112, 67–78. Lee, G.P., Min, B.E., Kim, C.S., Choi, S.H., Harn, C.H., Kim, S.U., Ryu, K.H., 2003. Plant virus cDNA chip hybridization for detection and differentiation of four cucurbit-infecting Tobamoviruses. Journal of Virological Methods 110, 19–24. Li, J., Chen, S., Evans, D.H., 2001. Typing and subtyping influenza virus using DNA microarrays and multiplex reverse transcriptase PCR. Journal of Clinical Microbiology 39, 696–704. Mackay, D.K., Bulut, A.N., Rendle, T., Davidson, F., Ferris, N.P., 2001. A solid-phase competition ELISA for measuring antibody to
M.K. Baxi et al. / The Veterinary Journal 172 (2006) 473–481 foot-and-mouth disease virus. Journal of Virological Methods 97, 33–48. Moonen, P., Boonstra, J., van der Honing, R.H., Leendertse, C.B., Jacobs, L., Dekker, A., 2003. Validation of a LightCycler-based reverse transcription polymerase chain reaction for the detection of foot-and-mouth disease virus. Journal of Virological Methods 113, 35–41. Oleksiewicz, M.B., Donaldson, A.I., Alexandersen, S., 2001. Development of a novel real-time RT-PCR assay for quantitation of foot-and-mouth disease virus in diverse porcine tissues. Journal of Virological Methods 92, 23–35. Poirot, O., Suhre, K., Abergel, C., OÕToole, E., Notredame, C., 2004. 3DCoffee@igs: a web server for combining sequences and structures into a multiple sequence alignment. Nucleic Acids Research 32, W37–W40. Rasmussen, T.B., Uttenthal, A., de Stricker, K., Belak, S., Storgaard, T., 2003. Development of a novel quantitative real-time RT-PCR assay for the simultaneous detection of all serotypes of foot-and-mouth disease virus. Archives of Virology 148, 2005– 2021. Reid, S.M., Ferris, N.P., Hutchings, G.H., Zhang, Z., Belsham, G.J., Alexandersen, S., 2002. Detection of all seven serotypes of footand-mouth disease virus by real-time, fluorogenic reverse transcription polymerase chain reaction assay. Journal of Virological Methods 105, 67–80. Reid, S.M., Forsyth, M.A., Hutchings, G.H., Ferris, N.P., 1998. Comparison of reverse transcription polymerase chain reaction, enzyme linked immunosorbent assay and virus isolation for the routine diagnosis of foot-and-mouth disease. Journal of Virological Methods 70, 213–217. Reid, S.M., Grierson, S.S., Ferris, N.P., Hutchings, G.H., Alexandersen, S., 2003. Evaluation of automated RT-PCR to accelerate the laboratory diagnosis of foot-and-mouth disease virus. Journal of Virological Methods 107, 129–139. Rouillard, J.M., Zuker, M., Gulari, E., 2003. OligoArray 2.0: design of oligonucleotide probes for DNA microarrays using a thermodynamic approach. Nucleic Acids Research 31, 3057–3062. Seifarth, W., Spiess, B., Zeilfelder, U., Speth, C., Hehlmann, R., LeibMosch, C., 2003. Assessment of retroviral activity using a universal retrovirus chip. Journal of Virological Methods 112, 79–91.
481
Shih, S.R., Wang, Y.W., Chen, G.W., Chang, L.Y., Lin, T.Y., Tseng, M.C., Chiang, C., Tsao, K.C., Huang, C.G., Shio, M.R., Tai, J.H., Wang, S.H., Kuo, R.L., Liu, W.T., 2003. Serotype-specific detection of enterovirus 71 in clinical specimens by DNA microchip array. Journal of Virological Methods 111, 55–60. Sorensen, K.J., de Stricker, K., Dyrting, K.C., Grazioli, S., Haas, B., 2005. Differentiation of foot-and-mouth disease virus infected animals from vaccinated animals using a blocking ELISA based on baculovirus expressed FMDV 3ABC antigen and a 3ABC monoclonal antibody. Archives of Virology 150, 805–814. Sorensen, K.J., Madsen, K.G., Madsen, E.S., Salt, J.S., Nqindi, J., Mackay, D.K., 1998. Differentiation of infection from vaccination in foot-and-mouth disease by the detection of antibodies to the non-structural proteins 3D, 3AB and 3ABC in ELISA using antigens expressed in baculovirus. Archives of Virology 143, 1461– 1476. Suryanarayana, V., Madanamohan, B., Bist, P., Natarajan, C., Tratschin, J.D., 1999. Serotyping of foot-and-mouth disease virus by antigen capture reverse transcriptase/polymerase chain reaction. Journal of Virological Methods 80, 45–52. Thompson, D., Muriel, P., Russell, D., Osborne, P., Bromley, A., Rowland, M., Creigh-Tyte, S., Brown, C., 2002. Economic costs of the foot and mouth disease outbreak in the United Kingdom in 2001. Revue Scientifique et Technique (International Office of Epizootics) 21, 675–687. Vangrysperre, W., De Clercq, K., 1996. Rapid and sensitive polymerase chain reaction based detection and typing of foot-and-mouth disease virus in clinical samples and cell culture isolates, combined with a simultaneous differentiation with other genomically and/or symptomatically related viruses. Archives of Virology 141, 331– 344. Wang, D., Coscoy, L., Zylberberg, M., Avila, P.C., Boushey, H.A., Ganem, D., DeRisi, J.L., 2002. Microarray-based detection and genotyping of viral pathogens. Proceedings of the National Academy of Science of the United States of America 99, 15687– 15692. Wilson, W.J., Strout, C.L., DeSantis, T.Z., Stilwell, J.L., Carrano, A.V., Andersen, G.L., 2002. Sequence-specific identification of 18 pathogenic microorganisms using microarray technology. Molecular and Cellular Probes 16, 119–127.