Future of microarray techniques for study of viral diseases Guy - - PDF document

“Future of microarray techniques for study

• f viral diseases”

Guy Vernet

Importance of viruses genetic variability

• Viruses target different host species, organs
• r cells,
• Polymorphisms in non-structural genes

impact viral replication and clinical outcome

• f infection,
• Polymorphisms can also affect response to

drug or immune treatment and vaccination efficacy.

Detection of viral polymorphisms has several interests

• Pathogens identification:

– sequences specific for genus, family, species, types, subtypes or even individual viral strains can be used for identification purposes in the diagnosis of patients disease or in epidemiological surveys, – for known and unknown pathogens

• Typing and detection of variants of clinical interest:

– the detection of single variants at key genome positions in genotyping assays can be used as a signature for virulence, transmissibility, resistance to treatment and for molecular epidemiology surveys

• Gene expression:

– the analysis of host genes expression changes during infection- is another aspect where genome information can be the basis of laboratory tools especially useful for studies of the mechanisms of viral diseases and their treatment.

Microarrays as laboratory tools for research and clinic

• Numerous examples of the use of microarrays have been

described in the last 15 years,

• Most of them are based on oligonucleotides hybridization but

microarrays using antigens or antibodies have been described,

• Both DNA and protein-microarrays are useful tools in the three

domains:

– pathogens identification, – typing and variants detection, – gen expression analysis.

Principle

DNA/RNA extraction Multiplex amplification /labelling Hybridization Reading, data mining

Chips versus beads microarrays

Probes and proteins can be coated on:

• glass slides (= chips)
• on « beads »

Only chips can currently provide high-density (thousands of probes).

Spotting versus in-situ synthesis of probes on chips

• « Supplier mode »
• Shorter probes
• Cost of platform (laser)
• High-density
• Result robustness

(redundancy)

• Robust manufacturing

process

In-situ synthesis

• Low-density
• Complexity of

manufacturing

• Lack of robustness
• Custom Chips
• Low-cost manufacturing

and reading equipment (ccd)

• Long probes

Spotting Drawbacks Advantages

Pathogens identification

The use of microarrays allows the simultaneous detection of numerous pathogens in a « syndrome-based » approach

• Affymetrix 2L-tilling
• 40588 probes

• The assay can efficiently detect enteroviruses,

flaviviruses and herpesviruses

• It can differentiate virus species among flaviviruses

and herpesviruses (including closely sequence-related HSV1 and 2)

• Enteroviruses can be typed in VP1

Recent examples of syndrome-based chips

• Ayodeji M et al., 2009. W ater-borne diseases.
• 13,000 elements,
• hepatitis A virus, human coxsackieviruses A and B, genogroups I

and II of Norovirus, and hum an rotavirus,

• tested on strains only,

« The results obtained with HAV and CV indicated that the hybridization profile thus generated can be used to identify closely related viral strains »

• Shi J et al., 2009. Encephalitis.
• herpes simplex virus type 1 and type 2, varicella-zoster virus,

Epstein-Barr virus, cytomegalovirus, and human herpes virus 6,

• probes in highly conserved regions of the DNA polymerase gene,
• tested on 290 CSF specimens,

« Compared with the results of TaqMan PCR, the sensitivity of PCR- microarray technology was 91.7% , the specificity was 100% »

Respiratory Viruses Panel (Luminex/Abbott) Luminex's xMAP technology

Microarrays are promising technologies but multiplex RT-PCR may be sufficient

Multiplex real-time PCR with multi-channel machines

1 tube 5-6 tubes real-time

FTD, Luxembourg Concordance: 94.1 % Currently, because of its automation on the Infiniti analyzer, the microarray assay described herein is the most adaptable system for clinical laboratories.

• Sequence database containing 228,638 viral sequences.
• 3 distinct genomic target regions for every family or genus of

vertebrate virus (1 conserved, 2 more variable).

• 29,495 probes (60-mer oligonucleotides):

– 9,477 probes for vertebrate viruses (1,710 species), – 11,479 16S rRNA bacterial probes, – 1,120 18S rRNA fungal probes – 848 18S rRNA parasite probes

• Random amplification, specific second step PCR + labeling:

– Reverse transcription with a random octamer linked to a specific primer sequence, – After digestion with RNase H, amplification using the above primer and a primer targeting the specific primer sequence, – Labeling in a subsequent PCR with the specific primer sequence linked to a capture sequence for 3 DNA dendrimers containing >300 fluorescent reporter molecules

Virus discovery: GreeneChip

• Pr. Ian Lipkin, Columbia University, NYC

Future of microarrays for pathogens identification

• Microarrays are well adapted for large panels of

highly variable pathogens like viruses,

• Multiplex amplification is a bottleneck:

– sensitivity is lower than single amplification, – sequence variabilty of viruses is well addressed at the detection level but multiple primers may be needed for a single virus.

• Automation, integration, cost reduction, robustness:

– beads-based assays are more easily automated and integrated,

• Competition to microarrays by multiplex real-time

PCR for known viruses and by next-generation sequencing for virus discovery

Typing and variants detection

Keys for treatment tailoring and monitoring and for epidemiology

Hepatitis B microarray: amplification strategy

S

PreS2

C sp tp

PreS1

sp rt X S 1 AS 1 1.7 kb S 3 S 2 AS 2 1.5 kb

Nested-PCR rt / S

for low viral loads

S 1’ AS 1’ 0.7 kb

Hepatitis B: Affymetrix chip

• HBV sequences alignment for design : > 1400 sequences
• Spots size 11µm
• 520 000 probes per chip
• Probes length : 20 mers
• Re-sequencing using a set of 4 probes for each nucleotide

Hepatitis B microarray: mutations detection and genotyping

Positions (a.a.) to be detected All mutants (Nt) to be detected Role tp sp 2 10 rt 66 197 preS1 3 6 preS2 6 16 S 98 339 BCP 33 112 preC 14 55 C 21 115 X 55 144 Cancer Total

298 994

Resistance Vaccine escape Clinical outcome

• Genotyping (8A to H): based on 812 informative positions along the genome

Quasi-species analysis Genotyping performances

Mutations detection

rt 169 WT rt 204 WT n=164 (98.8 %) I T L n=164 (82.3%) M I V S WT I 161 1 WT M 104 2 1 3 T I 1 10 L 1 V 2 12 21 1 S 3 1 3 rt 173 WT rt 236 WT n=164 (96.3%) V L G n=123 (48.0%) N T Y WT V 150 1 1 WT N 57 L 3 8 1 MUT T 2 G Y Ind 58 6 rt 180 WT rt 250 WT MUT n=164 (81,7%) L M P n=114 (94.7%) M V WT L 104 1 WT M 107 M 2 28 MUT V 5 C Ind 1 1 14 13 2 rt 181 WT n=164 (81.7%) A T G V WT A 124 1 T 2 G V 2 12 20 2 1 rt 184 WT n=164 (93.9%) T S A G N I WT T 152 1 S 1 1 2 A G 1 N I 1 2 3 rt 194 WT MUT n=164 (100%) A T WT A 164 MUT T rt 202 WT n=163 (95.1%) S I G WT S 155 I G 5 3 SEQUENCING SEQUENCING Ind Ind MUT MUT Ind MUT Ind Ind MUT Ind MUT M IC R O A R R A Y M IC R O A R R A Y MUT MUT Ind MUT Ind MUT Ind Ind Ind MUT Ind MUT Ind MUT Ind MUT Ind MUT Ind Ind MUT Ind

Influenza typing: a few recent examples

• Li X et al., 2009,Taizhou Affynigen Biotechnologies, Inc.,Shanghai,

China

• a microarray with 46 short virus-specific oligonucleotides for detecting

influenza A virus of 5 subtypes: H1N1, H1N2, H3N2, H5N1, and H9N2. « When tested with 225 clinical samples, 20 were detected to be positive using

• ur microarray-based assay, whereas only 10 were positive by the

conventional culture method »

• Huang Y et al., 2009, Nanogen, Inc., San Diego, California
• an electronic microarray to simultaneously type influenza A and B viruses and

to distinguish influenza A virus subtypes H1N1 and H3N2 from the potentially pandemic avian virus subtype H5N1. « In a study of 146 human clinical specimens that had previously been shown to be positive for influenza virus or another respiratory virus, the assay showed a clinical sensitivity of 96% and a clinical specificity of 100% »

• Gall A et al., 2009, Friedrich-Loeffler-Institut, Riems, Germany
• a low-density microarray for the detection and typing of avian influenza virus

(AIV). One probe for the conserved matrix gene and 97 probes targeting the HA0 cleavage-site region. « For validation, a panel of 92 influenza A viruses which ncluded 43 reference strains representing all 16 HA subtypes was used. All reference strains were correctly typed with respect to their HA subtypes and pathotypes, including HPAIV H5N1/ Asia »

Affymetrix influenza typing chip

Position Mutation 118 R118K 119 E119A 119 E119D 119 E119V 119 E119G 122 I122L 122 I122V 151 D151E 152 R152K 198 D198N 224 R224K 227 E227D 274 H274Y 276 E276D 292 R292K 294 N294S 371 R371K Position Mutation 26 L26I 27 V27A 27 V27T 27 V27S 30 A30T 30 A30P 31 S31N 34 G34E 41 W41P Position Mutation 92 D92E Position Mutation 13 13P

NA M2 HA PB2 NS1 PB1 PA

Position Mutation

627 E627K 701 D701N Position Mutation 130 E130D 190 D190E 202 V202G 225 G225E 226 E226L 226 E226Q 228 T228S 228 T228G 242 I242V 243 P243S 545 V545A

Position Mutation

615 K615R

528.000 probes sense et antisense 20 mers 5 µM spots 8 different PCR- multiplex Detection of influenza A, B and C Typing of 16 HA and 9 NA in inf A

• Careful selection of probes: only

26 probes for HRV and 13 for HEV included in the RPM-Flu re- sequencing Affymetrix chip.

• All 34 HRV and 28 HEV strains were

detected at the species level

• 33/34 HRV and 16/28 HEV strains

were correctly serotyped

Future of microarrays for typing

• Low-density chips can address simple questions (a

few resistance mutations, HA/NA typing) but competition with strip-based techniques or small sequencers (Biotage/pyrosequencing),

• High-density chips can answer complex questions:

50-1000 mutations but:

– multiplex-PCR complexity, – competion with next-generation sequencing?

• Beads-based technologies like xMAP (Luminex) can

be used to combine antibody and probes capture for serotyping (example of pneumococci typing that could be extended to influenza)

Gene expression analysis

For research applications and markers discovery

Microarray analysis has been applied to determine gene expression patterns

– disease pathogenesis (de Jong et al.; Dodd et al.; Kendrick

et al.; Sengupta et al.),

– immune response to infection (Calvano et al.; Efron

et al.; Cliff et al.; Jenner et al.; Kawada et al.; Ockenhouse et al.; Ramillo et al.),

– immune response after immunization

(Blumerman et al.; Mueller et al.)

– vaccine-induced responses (Diaz-Mitoma et al.; Fuller

et al.; Giacalone et al.).

Upregulation of immune/defense response genes by HPV-16 L1 VLP, in particular interferon- induced genes was observed in PBMC collected prior to vaccination, with many of these genes being further induced following vaccination. Gene Expression Patterns Induced by HPV-16 L1 VLP in Leukocytes from Vaccine Recipients

Alfonso J. Garcнa-Piсeres*,†, Allan Hildesheim‡, Lori Dodd§, Troy J. Kemp*, Jun Yang¶, Brandie Fullmer¶, Clayton Harro∥, Douglas R. Lowy#, Richard A. Lempicki¶, and Ligia A. Pinto*,**

J Immunol. 2009 February 1; 182(3): 1706–1729.

• A DNA chip analysis to detect critical factors that

mediate fulminant pneumonia due to influenza virus and S. pneumoniae co-infection in mice.

• IntelliGeneTM Mouse Chip (Takara Shuzo, Kyoto):

600 mouse-related genes, 300 experessed sequence tags and 10 control genes. « Platelet activator factor-acetyl hydrolase is increased and PAF receptor expressed in co-infected mouse cells compared to cells infected by the virus or bacteria alone »

Future of microarrays for gene expression analysis: competition with next-generation sequencing

454 GS FLX (Roche) Illumina (Solexa) SOLIDiD (Applied)

From Ansorge, New Biotech., 2009

Analysis of gene expression by RNA sequencing

• Analysis of RNA transcripts in a biological sample,
• Short sequence tags, 20-35 bases,
• Number of copies of detected tags is correlated to

mRNA expression Advantages: less cross-hybridization and better quantitation, gene discovery, no need for prior sequence knowledge Drawbacks: sequence repeats, cloning step, high cost

Stem cell transcriptome profiling via massive-scale mRNA sequencing. Cloonan et al., 2008, Nature Methods A global view of gene activity and alternative splicing by deep sequencing of the human transcriptome. Sultan M et al., 2008, Science.

Microarrays are still more robust than next-generation sequencing

Willenbrock et al. RNA, 2009

Synthetic RNA Exiqon microarrays

Conclusion

• Microarrays generate large data sets which

are useful to address complex research and diagnostic questions like syndrome-based pathogen identification, fine virus typing or biomarkers discovery,

• They are challenged by simpler/less costly

techniques for diagnostic and next-generation sequencing for gene expression analysis,

• The difference will be on integration through

innovative microfluidic, automation, robustness, quality of data analysis software and cost-reduction.