Application of Real Time RT-PCR to Detect Avian Influenza Virus David L. Suarez D.V.M., Ph.D. Research Leader Exotic and Emerging Avian Viral Disease Research Unit Agricultural Research Service United States Department of Agriculture
Diagnosis of Avian Influenza Virus • Clinical signs – Low pathogenic avian influenza – Highly Pathogenic avian influenza
• Diagnostic Tests – Antibody detection • AGID, HI • ELISA
– Virus detection • • • • •
Virus isolation Directigen Realtime RT-PCR (RRT-PCR) Traditional RT-PCR NASBA
– Other Diagnostic Tests
Influenza A Virus Negative sense RNA Single stranded Segmented
Neuraminidase Hemagglutinin
M2
PB1 PB2 PA HA NP NA MA NS
M1
16 Hemagglutinin subtypes 9 Neuraminidase Subtypes
Clinical Signs • Highly Pathogenic Avian Influenza – Symptoms depend on species – Systemic lesions with high mortality
• Low Pathogenic Avian Influenza – – – – – –
Can be subclinical Increased daily mortality Respiratory disease symptoms Drops in egg production Increased condemnation at slaughter Other symptoms can be strain dependant
Serology • Detection of type and subtype specific antibodies to previous influenza infection (AGID, ELISA, and HI) • Typically have antibody response 7-10 days after bird is infected • Birds stop shedding virus soon after antibodies are detectable • Widely used for routine surveillance for LPAI viruses because of cost and birds stay seropositive for months • Serology less valuable for HPAI
Direct Detection of Virus “Critical Component of Outbreak Control” • • • • • •
Virus Isolation Real-time RT-PCR Traditional RT-PCR NASBA Antigen Capture Elisa tests Other tests
Virus Isolation in SPF embryonated chicken eggs • Sensitive • Necessary for viral characterization • Embryonated chicken eggs are perishable and supply may be limited • Requires days to weeks for results • Concerns about cross contamination • Requires high levels of biosecurity
Real-time RT-PCR (RRT-PCR) • One-Step RT-PCR test was developed for typing and subtyping of influenza viruses using fluorescent Taqman probes • Sensitivity similar to virus isolation • The test doesn’t require running the PCR product on a gel and the probe confirms specificity • The complete test, including the RNA isolation step, can be completed in less than three hours • Requires expensive equipment, but can be done faster and cheaper than conventional
Positive
Negative
Traditional RT-PCR • Perform RT-PCR in a one-step or two-step format • Analyze DNA product on an ethidium bromide stained agarose gel • Sensitivity can be similar to virus isolation • Working with large amounts of amplified DNA creates a cross contamination risk! • Numerous recommendations to reduce contamination • Requires less expensive equipment, but greater manpower to perform test
NASBA Technology • Nucleic acid sequence based amplification (NASBA) • Isothermal amplification with several enzymes • NASBA-ECL similar sensitivity to virus isolation • Commercial kit-requires additional equipment • Not quantitative • Expensive
Antigen Capture ELISA Tests • Antigen capture test-most produced for human use only • Type A test only- does not subtype • Lower sensitivity than virus isolation or RRT-PCR • Low level of false positives from clean samples • Test is easy to use and requires no additional equipment • Not always available in large quantities • Directigen most commonly used in U.S. ($1520/test)
Why RRT-PCR for AIV? • Advantages
– Rapid test results (3 hours for rush samples) – Reduced cross contamination – Reagent Cost • RNA extraction $2-4 • RRT-PCR $4
– Scalable- large numbers of samples may be processed – Can type and subtype AIVs – Viable virus not necessary
Development and Evaluation of an RRT-PCR Test for AIV • • • •
Test Design Bench Validation Field Validation Transfer of technology to other laboratories • Proficiency testing • Performance during an outbreak • Future goals
Critical Control Points For Successful RRT-PCR • RNA extraction procedure – Efficient and reproducible extraction procedure – Works with a variety of samples – High throughput and cost are issues
• RT-PCR Amplification reagents – Commercial kits preferred – Sensitivity, cost and ease of use are issues
• PCR primers and probes – Provides specificity – Affects sensitivity
Real-time PCR chemistries • Sequence specific – – – – –
Taqman/ Dual-labeled probe/ Hydrolysis probe Fluorescence resonance energy transfer (FRET) Molecular Beacons Lux primers Scorpion probes
• Non-sequence specific – dsDNA binding dyes (SYBR green)
Hydrolysis/Taqman probes Reporter Primer 1
Quencher
Taq
Primer 2 Reporter
Quencher Taq
RRT-PCR for Avian Influenza A Two Test Procedure • Type A influenza Test – Detects any Type A influenza virus – Detects conserved Matrix gene – Works with samples from any species – Used as screening test – Sensitivity similar to virus isolation
• Subtype specific Test – Different test for every subtype – Provides confirmation of Matrix (MA) test – Less sensitive than MA – Available H5, H6, H7 and H9 tests – More affected by strain variation
Sample
Negative No further testing
AIV
RNA extraction
Type A flu RRT-PCR
Negative
Positive HA subtype RRT-PCR Positive
Virus Isolation
Report to NVSL Virus isolation
Bench Validation Procedures • Develop specific primers and probes for pathogen based on available sequence • Optimize for Mg, cycling parameters, probe/primer concentrations • Determine sensitivity and limit of detection on laboratory samples • Determine specificity with a panel of characterized viruses • Compare samples from experimentally infected animals to performance standard
Real-time RT-PCR for AIV Sensitivity • Compared to Egg Infectious Dose – Type A influenza – 10-1 EID50 – H5 and H7 subtype- 101 EID50
• Determined with in vitro transcribed RNA – Type A influenza - 103 copies – H5 and H7 subtypes– 104 copies
Specificity Panel Matrix
H5
H7
Isolate
Matrix
H5
Duck/NJ/7717-70/95 H1N1
+
-
-
Chicken/Netherlands/03
+
-
-
Mallard/NY/6750/78 H2N2
+
-
-
Turkey/Ontario/6118/67 H8N4
+
-
-
Env/NY/19019-6-98 H3N8
+
-
-
Chicken/NJ/1220/97 H9N2
+
-
-
Duck/Victoria/9211-18-1400/92 H3N8
+
-
-
Chicken/Korea/96006/96 H9N2
+
-
-
Duck/Alberta/286/78 H4N8
+
-
-
Chicken/Germany/N/49 H10N7
+
-
-
Chicken/Puebla/8629-602/94 H5N2
+
+
-
Turkey/VA/31409/91 H10N7
+
-
-
Chicken/MA/11801/86 H5N2
+
+
-
Chicken/NJ/15906-6/96 H11N1
+
-
-
Avian/NY/31588-2/00 H5N2
+
+
-
Duck/England/56 H11N1
+
-
-
Chicken/NJ/17169/93 H5N2
+
+
-
Duck/LA/188B/87 H12N5
+
-
-
Chicken/Hong Kong/220/97 H5N1
+
+
-
Gull/MD/704/77 H13N6
+
-
-
Duck/Malaysia/97 H5N3
+
+
-
Mallard/Gurjev/263/82 H14N5
+
-
-
Chicken/NY/14677-13/98 H6N2
+
-
-
Shearwater/W.Australia/2576/79 H15N6
+
-
-
Turkey/PA/7975/97 H7N2
+
-
+
Aichi/68 H3N2
+
-
-
Chicken/PA/13552-1/98 H7N2
+
-
+
Equine/KY/211/87 H3N8
+
-
-
Quail/AR/16309/94 H7N3
+
-
+
Swine/MN/9088/99 H3N2
+
-
-
Chicken/NY/8030-2/96 H7N2
+
-
+
Swine/IN/1726/89 H1N1
+
-
-
Isolate
Subtype
Subtype
H7
Initial Comparison of Tests • Virus isolation and RRT-PCR were compared during the summer of 2001 on 1550 samples from the LBMs of NY and NJ • H7N2 was commonly isolated from birds in the markets • Good correlation of tests at market level • Virus isolation appeared to be more sensitive than RRT-PCR although both tests appeared to miss positive samples
Field Validation of Diagnostic Tests • Literature on validation primarily targeted to serologic tests • Goal of 1000 negative samples and 300 positive samples • Compare diagnostic sensitivity (Dsn) and diagnostic specificity (Dsp) with performance (“Gold”) standard • Ideally compare 3 different geographic regions
H7 Low Pathogenic Outbreak in Virginia, USA • March 2002, a low pathogenic H7N2 was diagnosed in turkeys in Virginia • Outbreak quickly spread and eventually 197 infected flocks were identified • Control was by eradication (stamping out) • 4.5 million turkeys and chickens were destroyed • Direct government costs for eradication was 65 million dollars and total cost was over 160 million dollars • First time a real-time RT-PCR test was used significantly in an animal disease outbreak
Results of VA Study • >3,600 samples tested between 4/30 – 5/15 – Directigen – Harrisonburg, VA – VI – NVSL (DVL) – RRT-PCR – NVSL (CVB-L) • USDA-ARS, SEPRL, Athens, GA (6), DVL (1), CVB-L (1) • Samples included dead bird surveillance and suspect samples • Evaluated at specimen and submission level
Results By Specimen PCR* Directigen VI +
+
+
+
+
–
+
–
+
+
–
–
–
+
+
–
+
–
– *MA+
– and H7+ only
+
Interpretation Total agreement (48/95 = 50.5%) VI missed (4/95 = 4.2%) D missed (19/95 = 20.0%) D & VI missed (13/95 = 13.7%) PCR missed (3/95 = 3.2%) D false positive (2/95 = 2.1%) PCR and D missed (6/95 = 6.3%)
Statistical Analysis By Specimen Paired Comparisons Tests
Dir/VI
PCR/Dir
# of Positive Specimens 88/80
60/80
88/60
Chi-Square Statistic
1.88
10.45
18.27
p-value
0.170
0.002
0.001
PCR/VI
Statistical Analysis By Specimen PCR/VI
Dir/VI
Sensitivity
88.2
67.1
Specificity
99.5
99.8
Total agreement on all samples (+, – ) = 98.7% n = 3,628
Conclusions • RRT-PCR was equal to or more sensitive than VI – Could replace VI without adversely affecting control program
• Directigen test was valuable because of rapid detection • Virus isolation used to confirm RRT-PCR and Directigen positives and characterize isolates • RRT-PCR was eventually used in the VA state lab to provide faster diagnosis
Conclusions: Sensitivity, Speed, and Cost Sensitivity
Speed
Cost
Virus Isolation
Best
Worst
Intermediate
Directigen
Worst
Best
Worst
RRT-PCR
Best
Intermediate Best
All 3 types of tests have a role in future disease outbreaks
Real-time PCR platforms • Many real-time PCR platforms available • Some optimization of test required • Platforms have different characteristics – – – –
Speed Capacity Optical channels (multiplex) Cost
Lab Equipment Logistics • Bio-safety cabinet space – Ideally 3 cabinets • 1. RNA extraction (full exhaust if Trizol is used) • 2. RNA Transfer to reaction tubes • 3. Clean reagents, master mix preparation (Cell culture hood)
– Two cabinets (more realistic) • 1. RNA extraction • 2. RNA transfer/ master mix preparation Between uses hoods need to be wiped out with disinfectant and gloves should be changed.
RNA extraction • Methods – Silica binding columns (ex. RNeasy, Qiagen) – Magnetic beads (ex. MagMax, Ambion) – Organic solvents (ex. Trizol, Invitrogen)
• Formats – – – –
Individual samples/ centrifuge Vacuum manifolds 96 well plates Robots/automated (96, 48, 32 sample)
Sample types and processing methods Species/ Sample Type
Processing Method
Notes
Tracheal or oropharyngeal swab
RNeasy or Ambion Magnetic bead RNA extraction, then RRT-PCR
Virus primarily replicates in the respiratory tract (LPAI)
Cloacal Swab
Ambion Magnetic Bead RNA extraction then RRT-PCR
Virus primarily replicates in the intestinal tract. RNA extraction method must be modified for cloacal samples
Any species
Tissue samples
For HPAI viruses high levels of virus may be in tissues.
Environmental samples
(Swab)
Macerate with glass beads in trizol and then Magnetic beads Virus isolation to detect live virus
Gallinaceous Poultry (chickens, turkeys, quail)
Recommended Specimen
Waterfowlducks
RRT-PCR can detect inactivated virus, so may be inappropriate
National Animal Health Laboratory Network (NAHLN) • NAHLN laboratories are veterinary diagnostic laboratories (state, veterinary school, private) throughout the U.S. capable of testing for Foreign Animal Diseases • Goal was to provide NAHLN labs a validated RRT-PCR test for AIV, NDV, FMD and other select agents • Provide a rapid diagnostic response during an outbreak • Increase surge capacity • Continued federal (USDA/APHIS) oversight
Authorized Testing Laboratories • Person, not the laboratory, is authorized to perform the test • Person has to go through RRT-PCR training or have suitable experience • Must pass, on a yearly basis, a proficiency panel • Must provide data to USDA/APHIS to receive reimbursement for testing as part of surveillance programs
Proficiency Testing • Panel of 10-14 samples • Samples are whole virus inactivated by a phenol disinfectant • Includes negatives, strong and weak positives • Includes different HA subtypes • Successful tests require RNA extraction and amplification • Testing started by SEPRL, but transferred to NVSL/APHIS for AIV
Proficiency Panel Results Real-time PCR Instrument
Number of labs
Samples (data sets)
SD
CV
Correct sample ID’s (%)
SmartCycler
9
336 (24)
1.34
5.19
325 (96.7)
Light Cycler without BSA
1§
56 (4)
ND*
ND
49 (87.5)
Light Cycler with BSA
1
56 (4)
ND
ND
56 (100)
iCycler
1
28 (2)
ND
ND
28 (100)
ABI 7900
1
14 (1)
ND
ND
14 (100)
Total
12
490 (35)
ND
ND
472 (96.3)
Delmarva Outbreak 2004 • 1st flock identified by passive surveillance (clinical disease) in Harrington, DE on Feb 5th, 2004 • Presumptive diagnosis the following day by RRTPCR for H7 AI • Quarantines on farm and 2 mile buffer zone established per MOU the same day • Agreement in place to cover indemnification, depopulation, C+D, carcass disposal, etc. • Extensive surveillance in 2 mile quarantine and 6 mile buffer zone (11,728 samples in 10 weeks) • 2nd and 3rd infected flocks identified and depopulated quickly
H5N2 in Texas • Feb. 16 – increased mortality in a non-commercial broiler flock in Gonzales, Texas • Feb. 17 – diagnostic samples tested positive for H5 AIV at Texas Veterinary Medical Diagnostic Laboratory • Texas state officials arrived Feb. 17, placed flock under Hold Order, began to trace & test epidemiologic links • Two LBMs in Houston tested AI positive by RRT-PCR • Baseline testing began on all flocks within 8K and 16K-radius around the index farm • All flocks within 8K of index placed under Hold Order • Feb. 20 – NVSL confirmed virus as H5N2 • Feb. 21 – Index farm depopulated (6,608 birds)
Texas Outbreak Continued • Feb. 23 – NVSL determined H5N2 virus had amino acid sequence compatible with HPAI • Infected LBMs and their holding facilities were depopulated on Feb. 23 • All other Houston LBMs (3) depopulated as dangerous contacts • C&D of LBMs completed Feb. 29 • March 1 – NVSL reported results of chicken/embryo innoculation tests – no illness or deaths (pathogenicity index=0) • Surveillance-Over 2000 RRT-PCR tests and 3000 serologic tests performed in first month.
HPAI H5N2 – Gonzales, Texas
5-mile “infected zone” 39 non-commercial 5 commercial
10-mile “surveillance zone” 178 non-commercial 35 commercial
Keys to Rapid Control • Availability of sensitive and specific diagnostic test at a local or regional laboratory • Once a positive sample is identified initiate quarantine and confirm sample identity (virus isolation required) • Plans for control, including indemnity, carcass disposal, and movement controls, must be in place before outbreak occurs
Maintenance of Test • Sequence variation can cause test failure • Monitoring of new outbreaks to assure test performance is necessary • New commercial reagents should be evaluated for improved performance • Issues of application need to be evaluated
Evaluation of H5 Subtype RRTPCR Test for Asian H5N1 • H5 test was originally designed primarily for North American isolates • Could identify Asian H5N1 viruses with lower sensitivity • Sequence analysis of Asian isolates showed good conservation with reverse primer and probe, but 4 mismatches with forward primer • Redesigned H5 test to include forward primers optimized for both Asian and North American viruses – NA H5F TGACTATCCACAATACTCA – EA H5F TGACTACCCGCAGTATTCA
Future Developments • Dried down reagent beads – Includes Primer, probes, buffers, and internal control – Internal control should prevent false negatives – Multiplex test (requires multiple channels) – Longer shelf-life – Better reagent consistency among labs and tests – Simpler protocol
Thank you
Acknowledgements Suzanne DeBlois