Highly parallel microbial diagnostics using oligonucleotide microarrays

Clinica Chimica Acta 363 (2006) 106 – 119 www.elsevier.com/locate/clinchim

Review

Highly parallel microbial diagnostics using oligonucleotide microarrays Alexander Loy a,1, Levente Bodrossy b,* a

b

Department of Microbial Ecology, University of Vienna, A-1090 Vienna, Austria Department of Bioresources/Microbiology, ARC Seibersdorf research GmbH, A-2444 Seibersdorf, Austria Received 3 April 2005; accepted 5 May 2005 Available online 26 August 2005

Abstract Oligonucleotide microarrays are highly parallel hybridization platforms, allowing rapid and simultaneous identification of many different microorganisms and viruses in a single assay. In the past few years, researchers have been confronted with a dramatic increase in the number of studies reporting development and/or improvement of oligonucleotide microarrays for microbial diagnostics, but use of the technology in routine diagnostics is still constrained by a variety of factors. Careful development of microarray essentials (such as oligonucleotide probes, protocols for target preparation and hybridization, etc.) combined with extensive performance testing are thus mandatory requirements for the maturation of diagnostic microarrays from fancy technological gimmicks to robust and routinely applicable tools. D 2005 Elsevier B.V. All rights reserved. Keywords: Microbial diagnostics; Microarray; Hybridization; Oligonucleotide; Marker gene

Contents 1. 2. 3.

Basic concept. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Microarray hybridization formats. . . . . . . . . . . . . . . . . . . . . . Development and analytical performance . . . . . . . . . . . . . . . . . 3.1. Resolution: choice of marker genes and probe lengths . . . . . . . 3.2. Development and optimisation of microarray probe sets: specificity, 3.3. Further selected strategies to increase specificity and/or sensitivity . 4. Data analysis and quantification potential . . . . . . . . . . . . . . . . . 5. Diagnostic applications . . . . . . . . . . . . . . . . . . . . . . . . . . . 6. Conclusions and outlook . . . . . . . . . . . . . . . . . . . . . . . . . . Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1. Basic concept DNA microarray (microchip, biochip, gene chip) technology allows the parallel analysis of highly complex gene * Corresponding author. Tel.: +43 50550 3548; fax: +43 50550 3444. E-mail addresses: [email protected] (A. Loy), [email protected] (L. Bodrossy). URLs: www.microbioal-ecology.net (A. Loy), www.arcs.ac.at/u/ub/ microbiology (L. Bodrossy). 1 Tel.: +43 1 4277 54207; fax: +43 1 4277 54389. 0009-8981/$ - see front matter D 2005 Elsevier B.V. All rights reserved. doi:10.1016/j.cccn.2005.05.041

. . . . . . . . . . . . . . . . . . . . . . . . sensitivity, . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . and uniformity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

106 110 110 110 111 114 114 114 115 116 116

mixtures in a single assay and thus symbolises, as no other method, the (post-)genomic era of high-throughput science. While microarrays initially emerged as tools for genomewide expression analysis and are nowadays routinely used for this purpose, they are also increasingly being developed for diagnostic applications [1]. Microbial diagnostic microarrays (MDMs) consist of nucleic acid probe sets, with each probe being specific for a given strain, subspecies, species, genus or higher taxon [1]. A characteristic MDM experiment is depicted in Fig. 1. MDMs fall into two distinct

A. Loy, L. Bodrossy / Clinica Chimica Acta 363 (2006) 106 – 119

107

Fig. 1. Typical steps of a diagnostic microarray experiment.

categories according to their intended use. Environmental MDMs [2,3] are primarily applied in environmental and industrial microbiology to obtain a picture of the structure of the microbial community being analysed. Requirements for this class of MDMs are the parallel detection of many microorganisms at the level of species, genus or a higher

taxon and the potential for some level of quantification. Detection/identification MDMs, typically applied in clinical, veterinary, food and biodefense microbiology [4,5] must usually enable the reliable detection and/or identification at the species/subspecies/strain level of one or a few microbes out of many that may be present in a sample. It

108

Table 1 Selected applications of oligonucleotide microarrays for microbial diagnostics 1)

Target organisms

Marker gene(s)

Highest phylogenetic resolution

L

Planar glass slide

Entamoeba histolytica, E. dispar, Giardia lamblia, Cryptosporidium parvum Cryptosporidium Bacteria

Various

Species, subtypes

20 – 30

hsp70 16S rRNA

Isolates Higher level bacterial taxa Species

15 20 50

23S rRNA

Species/ subspecies

16S rRNA cyt, rpoN, gyrB, toxR, ureC, dly, vapA, fatA, A. plassal fur, glyA, cdtABC, ceuBC, fliY 16S rRNA, 16S – 23S intergenic spacer, Campylobacter-specific genes 16S rRNA 16S and 23S rRNA gyrB

Planar glass slide Affymetrix Planar glass slide

Planar glass slide Planar glass slide Planar glass slide

3)

4)

Planar glass slide Planar glass slide

4)

Planar glass slide Planar glass slide Planar glass slide

5)

Planar glass slide Planar glass slide Planar glass slide

Planar Planar Planar Planar

glass glass glass glass

slide slide slide slide

6)

3D-surface Planar glass slide

4)

Bacteria involved in nitrification, denitrification, nitrogen fixation, methane oxidation and sulfite reduction Bacteria causing abortion and sterility in mares Bacterial fish pathogens Bacterial fish pathogens Campylobacter jejuni, C. coli, C. lari, C. upsaliensis Campylobacter spp.

Cyanobacteria Enterococcus Escherichia coli, Shigella, Salmonella Selected taxa of marine bacterioplankton Listeria Listeria spp., Campylobacter spp., Staphylococcus aureus, Clostridium perfringens Marine bacterioplankton Methanotrophs Mycobacterium spp. Rifampin-resistant Mycobacterium tuberculosis Rifampin-resistant Mycobacterium tuberculosis Pathogenic Vibrio spp.

Sample type

Reference

C/F

–

–

[79]

68 31179

E E

– 1

– Air filtrate

[94] [17]

763

E

1

Marine sediment

[47]

24 – 32

32

C

21

Cervical swabs

[21]

Species Subspecies

22 – 31 25

18 9

C/F E/F

– –

– –

[18] [95]

Species

17 – 35

74

C

16 + 6

[76]

Species

27 – 35

5

C/F

10 + 65

Isolates and mixed cultures Chicken cloacal swabs

Above genus Species Species

20 – 29 41 15 – 19

19 18 10

E C/F C

1 2 –

Lake water Milk –

[98] [99] [27]

16S rRNA

Higher taxa

15 – 20

21

E

1

Sea water

[58]

iap, hly, inlB, plcA, plcB, clpE Various

Species Species

17 – 33 17 – 35

132 178

C B/F

– –

– –

[36] [100]

16S rRNA pmoA gyrB rpoB

Species Species/ subspecies Species Strain

15 – 20 17 – 27 13 – 15 15 – 16

21 61 28 18

E E C C

1 >100 40 –

Sea water Landfill cover soil Human sputum –

[101] [1,39] [102] [103]

rpoB

Strain

15 – 23

43

C

31

Clinical samples

[90]

vvh, viuB, ompU, toxR, tcpI, hlyA, tlh, tdh, trh, etc.

Species

30 – 32

13

C/F

30

Oyster

[82]

88

11)

Field

2)

No. of samples analysed

nirS, nirK, amoA, nifH, pmoA, dsrAB

No. of probes

[96,97]

A. Loy, L. Bodrossy / Clinica Chimica Acta 363 (2006) 106 – 119

Platform

Planar glass slide

7)

gyrA

Genotype

19

42

C

–

–

[104]

Planar glass slide

Quinolone-resistant Escherichia coli Respiratory bacterial pathogens

gyrB, parE

Species

20 – 24

27

C

94

[105]

Planar glass slide 3D-surface Planar glass slide Planar glass slide

Rhodocyclales Some Bacillus spp. Some Bacillus spp. Sulphate-reducing prokaryotes

16S rRNA 16S rRNA 16S – 23S intergenic spacer 16S rRNA

Species Species Species Species

18 15 – 23 18 – 22 18

79 30 42 132

E B/C B/C C/E

1 – – 6+8

Thermophilic anaerobic Archaea and Bacteria Major potential biowarfare agents (bacteria, viruses, eukaryotes) Acute respiratory disease-associated Adenoviruses Human group A rotaviruses Human papillomaviruses Human papillomaviruses Human papillomaviruses

16S rRNA

Genus

17 – 20

17

E

3

Various

Species

20

B/C

2

Middle ear fluid, throat swabs Activated sludge – – Hypersaline cyanobacterial lake mat; periodontal tooth pocket, fen soil Oil reservoir formation waters Air filtrate

E1A, fibre, hexon

Serotypes

60 – 72

36

C

19

Throat swabs, nasal wash samples

[80]

VP7 E1 Various Various

Genotype Types Types Types

18 – 26 20 14 – 28 30

50 51 10 27

C C C C

– 130 100 73

[108] [109] [110] [111]

Various Various (a.o. hemagglutinin and neuraminidase genes) C23L/B29R gene, ORF 62 (varicella-zoster) crmB Various

Subtype Subtypes

17 – 29 45 – 65

476 29

C C

– –

– Clinical samples Clinical samples Clinical samples (tonsillar cancer) – –

Types

13 – 21

57

B/C

–

–

[114]

Species Serotype

12 – 16 70

15 1600

C C

– 6

– Clinical samples

[115] [50]

7)

3D-surface Affymetrix

8)

Planar Planar Planar Planar

glass glass glass glass

slide slide slide slide

5)

Planar glass slide 3D-surface 9)

Influenza viruses Influenza viruses

Planar glass slide

Orthopoxviruses

3D-surface Planar glass slide 1)

10)

Orthopoxviruses Viruses

Length of probe [nt]. B: biodefense; C: clinical microbiology; F: food microbiology; E: environmental microbiology. 3) On-Chip PCR. 4) Multiplex PCR. 6) DNA ligation detection reaction. 5) SOLAC-short oligo ligation assay on chip. 8) Perfect match and single mismatch probes used. 7) 3 – 10 diagnostic regions were identified per microorganism and each one of them was covered by 100 – 300 probes. 9) Flow-Thru Chip. 10) 70mer discriminatory oligonucleotides were designed based on all published viral genome sequences. 11) Probes comprised of 5 specific nucleotides (participating in specific ligation) + 10 – 11 spacer T residues. 2)

[77] [107]

[112] [113]

A. Loy, L. Bodrossy / Clinica Chimica Acta 363 (2006) 106 – 119

Planar glass slide

53660

[56] [64] [22] [15,106]

109

110

A. Loy, L. Bodrossy / Clinica Chimica Acta 363 (2006) 106 – 119

should be noted, however, that there are also medical microbiology contexts where (as described for environmental MDMs) the microbial community structure has to be addressed, e.g. the composition of the human gastrointestinal tract microbiota. Although recent studies (Table 1) undoubtedly demonstrate that MDMs are valuable tools for identification of microorganisms and viruses in a highly parallel fashion, their use for routine diagnostics is still hampered by a lack of standardisation (regarding factors such as probes, target genes, hybridization platforms, protocols, and data analysis) and insufficient evaluation of newly developed MDMs. In this review, we summarise recent progress in the MDM field. Many (seemingly conflicting) parameters and steps have to be integrated during microarray design and application in order to fulfil the potential of MDMs as high-throughput screening tools for routine diagnostic purposes. We thus highlight crucial points related to development and evaluation of MDMs. The key to a validated MDM, and consequently to reliable results, is rigorous in silico and in vitro performance testing. Furthermore, we exemplarily show recent technological advances, which should further improve the use of MDMs in the near future.

2. Microarray hybridization formats Although many diverse microarray platforms have become available over the past few years [6], only a limited number of solid supports are currently used for MDMs. For Affymetrix microarrays (www.affymetrix.com), oligonucleotide probes are synthesised directly onto the microarray surface by employing specific masks and the photolithography method. This approach enables a very high probe density (well over 100,000 probes per microarray). The high price, low flexibility and lack of a suitably high number of validated oligonucleotide probes currently limit the wide application of Affymetrix GeneChips in microbial diagnostics. In contrast, the NimbleGen technology (www.nimblegen.com) uses digital micromirrors instead of physical masks to guide on-chip probe synthesis, allowing a more flexible design of custom-made high-density microarrays. Specific, three-dimensional microarray formats, such as the gel-pad platform (www.biochip.ru/en/) [7,8] or the PamGene system (www.pamgene.com) [9,10], are coupled with appropriate hybridization and detection devices that offer the option to record hybridization and dissociation events in real-time. Melting curves for all probes on a microarray can thus be rapidly established, making the development of validated probe sets significantly easier [11,12]. However, these systems are currently available only in a few laboratories. The pioneer microarray format and still the most widely used miniaturised solid support for the covalent immobili-

sation of probes are planar 1  3 in. glass slides. Oligonucleotides are in most cases tethered via their 5V ends to reactive groups on the coating layer of the glass surface. The establishment of microarray core facilities (including microarray spotting and detection devices) in many laboratories and the general utility, flexibility, and moderate price of planar glass microarrays are mainly responsible for the success of this standard format.

3. Development and analytical performance 3.1. Resolution: choice of marker genes and probe lengths Two main parameters affect the resolution of a diagnostic microarray assay: (i) the degree of conservation of the marker gene and (ii) the length of the oligonucleotide used as a probe to target it [13]. The most widely employed target molecule for the detection and phylogenetic analysis of microorganisms is the small-subunit ribosomal RNA (SSU rRNA) and its gene [7,14 – 18]. The popularity of the SSU rRNA is reflected in the existence of large and regularly updated sequence [19, 20] and probe databases (http://www.microbial-ecology.net/ probebase/) [16] for this target molecule. The main limitation of using the SSU rRNA (gene) as a marker in microarray assays is that resolution below the species level is generally not possible owing to high overall sequence conservation. Because differentiation of strains is often essential in clinical diagnostics in order to initiate appropriate treatment of an infection, less-conserved target molecules are needed. Potential probe targets which offer strain-level resolution include (i) the large-subunit ribosomal RNA (LSU rRNA) [21], (ii) the SSU-LSU rRNA intergenic spacer region [22], (iii) house-keeping genes (e.g. rpoB [23 – 25], gyrA [26], gyrB [27], recA [28], tuf [29,30], groEL [31], atpD [30], ompA, gapA, pgi [32], tmRNA [33]), (iv) virulence genes [34 – 37], (v) antibiotic resistance genes [9,38], (vi) functional genes encoding enzymes responsible for specific metabolic traits [1,39 – 42], etc. (for a detailed list, please refer to www.arcs.ac.at/u/ub/microbiology). However, individual sequence databases for these alternative markers, if they exist at all, currently contain considerably fewer entries than the SSU rRNA databases, constraining the development and evaluation of encompassing probe sets for microarrays [43]. The hybridization properties of long oligonucleotide probes (typically 50 –100mer), which show pronounced hysteresis (higher temperatures for dissociation than for association) are fundamentally different from those of short probes (typically 15– 30mer) [44,45]. While in principle short oligonucleotides allow the discrimination of single nucleotide differences under optimal conditions (see below for further details), this does not hold true for long oligonucleotides. Their threshold for differentiation is approximately 75 – 87% sequence similarity [41,46,47].

A. Loy, L. Bodrossy / Clinica Chimica Acta 363 (2006) 106 – 119

However, probes of increased length display orders of magnitude higher target binding capacities and hence the use of long oligonucleotides should generally improve the detection sensitivity of a microarray [48]. Typically, long oligonucleotide MDMs are used in combination with universal (not PCR based, low bias) amplification strategies or without any amplification [47,49,50]. In these approaches probes target various, non-conserved genes which are specific to the microorganisms targeted. Thus, the low differentiation power of long oligonucleotide probes is compensated for by the host specificity of the genes they are designed against. With or without universal amplification, the resulting target represents the entire gene pool present in the investigated sample, without any reduction in its complexity. Higher binding capacities of the long oligoprobes and higher target complexity result in similar relative detection thresholds to that of short oligoprobes in conjunction with PCR amplification (approximately 5%) [1,47,51]. There is no universal answer to the question of which probe-target combination is the best for a diagnostic microarray, as this will depend strongly on the intended application. In principle, the highly parallel nature of microarrays allows various probes of different length and targeting different genes to be applied simultaneously, although homogenous hybridization behaviour of complex probe/target combinations remains a problem (see below). 3.2. Development and optimisation of microarray probe sets: specificity, sensitivity, and uniformity A crucial and challenging first step in the development of oligonucleotide microarrays is the design of a suitable set of probes and this is thus presented in more detail in the following section. The following criteria set the quality standard for a microarray probe set. All probes on a microarray should (i) be highly specific for their target genes i.e. not crosshybridize with non-target sequences (specificity), (ii) bind efficiently to target sequences to allow the detection of low abundance targets in complex mixtures (sensitivity), and (iii) display a similar hybridization behaviour i.e. similar thermodynamic characteristics under the same experimental conditions (homogeneity, uniformity). Unfortunately, these ideals represent conflicting goals in practice, and thus efforts to fulfil these criteria need to be carefully balanced during probe design and experimental procedures. Design of microarray probes in silico usually entails the use of specific software tools in conjunction with an underlying sequence database. Numerous software tools have been developed and widely applied for the parallel design of oligonucleotide probes for whole genome microarrays [52,53], but not so for diagnostic microarrays. Although only one probe can be designed at a time, making the design of multiple probes laborious, the probe

111

software tool of the ARB program package [19] is commonly used for the design of 10 –100mer oligonucleotide probes for diverse hybridization formats [15,54,55] including diagnostic microarrays [56 – 58]. The first step in the design of a diagnostic oligonucleotide is selection of the target group, either arbitrarily or based on sequence analysis (e.g. a phylogenetically coherent group of organisms) (Fig. 2). Second, depending on user-defined settings, the ARB probe search algorithm identifies unique sequence stretches which could serve as probe target sites and subsequently returns a ranked list of potential oligonucleotides. Third, the suggested probes can be matched against all (usually aligned) sequences in the database. This probe match option of ARB is highly beneficial because the user has the opportunity to evaluate the specificity of a given probe in silico by checking for the number, types, and positions of mismatches to nontarget sequences. It is vital to stress that the value of this evaluation step and thus the quality of a developed probe set will strongly correlate with the completeness and up-todateness of the underlying sequence database. Regarding the position of mismatches to non-target organisms, a rule of thumb for short oligonucleotides is that mismatches located in the middle of the probe-binding site destabilize more strongly the formation of the probe-target duplex. In contrast, an even distribution of mismatches in the probebinding site is required to achieve optimal discrimination for long oligonucleotides [46,48]. Please note that single mismatches located at the terminal or next-to-terminal position can hardly be resolved in microarray hybridization [11,48,57]. In a further step, the local alignments of perfectly-matched and mismatched target sites can be used as input for other freely-available software such as CalcOligo (www.arcs.ac.at/u/ub/microbiology), Mfold [59], or HyTher [60], to calculate the theoretical thermodynamic properties of these potential hybridization events. However, these nearest-neighbour algorithms for the calculation of free energies (DG) and melting temperatures (T m) have been developed for hybridizations in solution involving known concentrations of probe and target molecules, factors which are not readily known or do not apply for hybridization of immobilized probes on microarrays. Data obtained in the laboratory must therefore accurately verify the use of these algorithms to predict potential cross-hybridization events of a microarray probe [1,15,39,61]. For single probes, it is in most cases possible to adjust experimental conditions in such a way that no crosshybridization occurs. However, the most widely used microarray hybridization formats only allow hybridization and/or washing at a single stringency, making it impossible to provide optimal hybridization conditions for all probes on a microarray. Thus, promiscuous binding of some probes to non-target sequences (false-positive signals) is a frequently encountered problem [1,15]. Nonetheless, several approaches still guarantee the reliability of identifica-

112

A. Loy, L. Bodrossy / Clinica Chimica Acta 363 (2006) 106 – 119

Suitable software

Creation/update of sequence database

ARB

Identification of target group e.g. by phylogenetic analysis

ARB

ARB PrimRose ARB PrimRose OligoCheck CalcOligo Mfold HyTher

Probe design Specificity, sensitivity, homogeneity Type, number, position of mismatch(es)

Secondary structures (∆G, Tm) 2

1

3 4

CalcOligo Mfold HyTher

Predicted hybridization behavior Pre-selected probe set

In silico

Preliminary microarray

Thorough experimental evaluation (Specificity, sensitivity, homogeneity) Elimination of “bad” probes

Re-design

Re-evaluation Refined probe set Fully evaluated final microarray

In vitro

Fig. 2. Schematic representation of the procedures involved in the development and evaluation of a microarray probe set. Types of molecular interactions of the probe and target molecules: 1, oligo-oligo dimer; 2, oligo hairpin; 3, secondary structure of the target; 4, oligo-target duplex. Free energies and melting temperatures of the different types of interactions of probe and target molecules can be calculated to estimate their influence on duplex yield.

tion by microarrays. One approach involves the design of multiple probes having identical specificities for the same target sequence or group of sequences. Therefore, all probes in a set of probes perfectly matching the target organism must show positive hybridization signals to minimize the risk of false-positive identification [34,62,63]. This multiple probe concept is further extendable by probes having hierarchical (nested) specificities for the target sequences (Fig. 3), which also enables the detection of novel members of known groups [54]. Nested

probes can be best exploited if rRNA or its gene serves as the target molecule [15,54,56,64]. In contrast, the highly variable third codon (‘‘wobble’’) position in protein-coding target genes hampers the design of probes having a broader specificity and spanning more distantly related sequences. In addition to, or instead of, the use of multiple probes for one marker gene, it is possible to use more than one marker gene of a target organism as a probe target during hybridization. This multiple probe-multiple target strategy additionally increases redundancy and hence the confi-

A. Loy, L. Bodrossy / Clinica Chimica Acta 363 (2006) 106 – 119

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28

A1 B1

A2

113 A3

A4

B2

C1

D1

E1

Phylogeny

Sequence alignment

Fig. 3. Illustration of the nested multiple probe concept. A schematic sequence alignment is displayed in dashed lines. Colored boxes show perfectly matched probe target sites. As an example, sequence number one is unambiguously rated as present if each probe in the set of probes E1 Y D1 Y C1 Y B1, B2 Y A1, A2, A3, perfectly matching this sequence at identical and hierarchical levels of specificity, shows a positive hybridization signal.

dence in a positive call [62,63]. Furthermore, careful selection of multiple marker genes may also allow genotyping of the detected microbes. Information on virulence factors, antibiotic resistance, etc., can thereby be obtained simultaneously in a single diagnostic assay. Another prominent way (also applied in the Affymetrix GeneChips protocols [17]) to ensure specificity is the inclusion of so-called mismatch-control probes on the microarray [65,66]. Comparison of signal intensities from perfectly matched and mismatched probes allows crosshybridization to be identified and its extent estimated. In summary, one set of experimental conditions is not suitable to ensure absolute specificity of all probes during simultaneous hybridization on a microarray. However, microarrays offer possibilities to compensate for lack of specificity on the single probe level by inclusion of a multiplicity of redundant probes, an option which is not available for traditional hybridization formats. Sensitivity is (i) commonly defined as the lowest absolute and/or relative abundance of the target group which is detectable, and (ii) dependent on the duplex yields of the individual probes on a microarray. To avoid the worst-case scenario, low sensitivity of a given probe leading to a false-negative result, a variety of parameters can be adjusted during the design of the actual microarray and the experimental set-up. An important strategy for increasing duplex yield is to reduce steric hindrance during microarray hybridization. Lifting the oligonucleotide probe physically away from the surface of the microarray by the introduction of simple spacer elements enhances its accessibility by several orders of magnitude [67,68]. Furthermore, the tendency of a probe and its target to form stable secondary self-structures must be minimized [1,69,70]. T m and DG of secondary structures can be calculated by using the software tools mentioned above but should be treated cautiously due to the limited knowledge on the thermodynamics of microarray hybridization (see

above). If RNA is the target molecule, complexity of the target is typically reduced by physical or chemical fragmentation [1,45,71]. A popular method to create short DNA fragments of varying length is by random prime labelling [49]. Partial digestion with DNase I [49] or use of hydroxyl radical-producing reagents are other options to fragment DNA, the latter non-enzymatic treatment being more reproducible [72]. So-called helper or chaperone probes target a sequence region adjacent to a probe target site and have been demonstrated to increase target accessibility by relieving secondary structures [58,73]. However, the helper probe approach is less practical for high-density microarrays, as each probe on the microarray requires its own specific helper counterpart. Unique to formats involving the simultaneous hybridization of many probes is that the probes have to be finetuned for uniform thermodynamic behaviour. One way to achieve this is by using probes that are identical in length and adding tertiary amine salts such as tetramethylammonium chloride to the hybridization and/or washing buffers [15,56]. Thereby, differences in the G + C contents and thus duplex stabilities among the probes are attenuated [74,75]. Another strategy to equalize melting properties of different oligonucleotide probes is to manipulate their length [1,39]. One must realize that current in silico approaches for predicting the hybridization behaviour of microarray probes are limited due to the aforementioned reasons. At best, they will lead to a pre-filtered set of candidate probes whose true experimental performance will only be uncovered after extensive empirical testing (Fig. 2). In practice, a suitable set of test targets should contain at least one perfectly matched target for each probe on the microarray. After challenging the microarray’s performance by individual hybridizations with each test target, ‘‘bad’’ probes showing low sensitivity and/or specificity are removed or replaced. Subsequently, concentration series of targets

114

A. Loy, L. Bodrossy / Clinica Chimica Acta 363 (2006) 106 – 119

The most common way to enhance detectability of a target gene is the use of PCR [1,15,17,76 –78]. PCR is applied to focus the labelling to the target genes and thus to enrich them relative to non-targeted background DNA. In case of a single marker gene, highly selective PCR primers may achieve optimal enrichment of target sequences from organisms of interest [56]. However, a high number of selective PCR primers is needed to cover all organisms targeted by a microarray. Running multiple PCR reactions considerably increases the hands-on-time of the microarray assay and thus simultaneous target amplification via multiplex PCR is desirable [35,79 – 82]. Multiplex PCR has the inherent drawback that potential biases increase with increased complexity of the reactants (primers/target genes) [the number of successfully applied primer pairs within a multiplex PCR is normally low, up to 12 [83]]. Consequently, comparable to the actual microarray hybridization, the implementation of a pre-hybridization amplification step based on multiplex PCR requires extensive, careful testing and validation. A promising approach to increase the sensitivity of a microarray assay is tyramide signal amplification (TSA) [51]. Upon hybridization, this method relies on enzymatic amplification of the signal by employing horse radish peroxidase-mediated deposition of fluorochrome-labelled tyramides at the location of the probe. The ultimate specificity of microarray technology depends on the discrimination between a fully complementary target and a non-target differing in only a single nucleotide. Recently, various enzyme-assisted hybridization strategies (also used in single nucleotide polymorphism and resequencing assays [84,85]) have gained attention because of their promise in strongly discriminating single mismatches located near the 3V end of microarray probes [21,86 –90].

usually involve the hybridization of a single sample per microarray (single color experiment). Fluorescence signals for the individual probes are typically rated as present or absent (Fig. 1), a decision made depending on either simple visual inspection of the scanned microarray image or an arbitrary signal intensity threshold [1,15,21, 22,58,80]. Subsequently, complex hybridization patterns are in most cases still translated manually into inventory lists of organisms present in the analyzed samples. This is a tedious procedure which becomes more problematic with increasing numbers of probes per microarray and increasing numbers of samples to be analyzed, and thus constitutes a major bottleneck of microarray-based diagnostics. In an attempt to bridge the gap between data collection and analysis, a simple command line-based program was recently developed for diagnostic microarrays [15] [an advanced graphical user interface-based version, ChipAnalyser, is under development (Harald Meier, unpublished data)]. Although systematic and stochastic errors associated with the different steps during microarray fabrication and target preparation greatly hamper quantification, the fundamental potential of microarrays to provide quantitative data on the abundance of the detected target in a sample is widely acknowledged. Adopting the two color hybridization approach from transcriptome microarray analysis, relative abundance of targets in a sample (labelled in color one) can be measured by competitive hybridization with target mixtures of known concentrations (labelled in color two) on the microarray [1,92]. For single color hybridizations, a linear correlation between the signal intensity of a probe and the concentration of the respective target sequence has been observed for a certain range of target concentrations [41,47,93]. However, the slope of this linear relationship will vary among the different probes immobilized on a microarray because, as mentioned previously, different probes display different affinities to their targets. For a reliable quantitative interpretation of a microarray hybridization, a standard curve is thus needed for each probe on the microarray. This tedious evaluation procedure has not been accomplished for microarrays published to date and therefore the ability to correctly quantify across all probes on a microarray is yet to be demonstrated. A semiquantitative comparison of similar community structures is, however, possible, enabling the researcher to detect spatial and temporal changes in microbial community composition [39].

4. Data analysis and quantification potential

5. Diagnostic applications

For microarrays used in transcriptome analysis, signal intensity ratios indicating relative levels of expression are obtained upon competitive hybridization of a control sample versus the sample of interest (two color experiment) [91]. In contrast, diagnostic microarray experiments

MDMs are increasingly used in clinical, environmental and food microbiology as well as in biowarfare agent detection [43] (Table 1). While some clinical applications are limited to genotyping of isolated strains, there are more and more publications on cultivation-independent, rapid

perfectly matching those probes that have displayed the highest and lowest duplex yield should be hybridized to the microarray, giving an impression of the range of sensitivities achievable for the individual probes. In conclusion, the key to a reliable diagnostic microarray lies within its concerted evaluation integrating in silico predictions, adjustment of hybridization conditions and thorough testing in the laboratory. 3.3. Further selected strategies to increase specificity and/ or sensitivity

A. Loy, L. Bodrossy / Clinica Chimica Acta 363 (2006) 106 – 119

detection and identification (in some cases also genotyping) of pathogens from clinical specimens. Microarrays used merely for genotyping of isolates are outside the scope of our review and are thus not listed in Table 1.

6. Conclusions and outlook Technologies used for MDMs are rather diverse. PCR amplification, followed by conventional fluorescence labelling is used in most cases. Specificity is then defined by the hybridization step. Depending on the marker genes applied, current MDMs can provide resolution at various taxonomic levels down to species or even strain level. Ribosomal RNA genes are often used as marker genes, but less so in clinical microbiology due to their limited phylogenetic resolution. Alternative, higher resolution phylogenetic markers are thus beginning to be applied, with short oligonucleotide probes, used in a hierarchical and parallel manner, being the preferred choice. Reliable analysis of 10 – 30 samples can be achieved within 24 h (in most cases within a working day) by one researcher. The costs of consumables are approximately 30 – 100 EUR per analysis, depending on the exact methodology used. Although this may seem expensive, it must be remembered that a single analysis provides parallel detection and identification of tens to hundreds of microorganisms potentially present. In addition, basic genotyping information can be obtained as well as semi-quantitative information on the relative abundance of the detected microorganisms. The latter aspect is especially useful when comparing similar samples for changes in the microbial community structure (e.g. gut microbiota). Absolute quantification is troubled by biases associated with, for example, PCR amplification, nucleic acid recovery and different numbers of (marker) gene copies per genome. The sensitivity of MDMs is normally limited by the relative abundance of the microbial population within the targeted community, with reported detection limits being 1 – 5%. The hybridization potential of the probes is predicted based on the nearest neighbour model and a number of empirical microarray-specific factors. MDM technology lends itself to automation, DNA/RNA purification, amplification, labelling can all be carried out by existing laboratory robots. Moreover, there are proof-ofconcept experiments where all these tasks are carried out in a single lab-on-a-chip device [116]. The subsequent hybridization, wash and scan steps can also already be carried out by various instruments in an automated manner (for details please refer to http://genomicshome.com/). Automated result interpretation is primarily a bioinformatics challenge. In general, MDM technology is, analogous to the application of microarrays in gene expression profiling, a highthroughput screening tool with limited quantification potential. MDM results should thus be confirmed via established lower-throughput but more sensitive and/or

115

truly quantitative alternative methods such as fluorescence in situ hybridization, quantitative PCR, or immunological detection techniques. Even though MDMs are already widely applied in most fields of microbiology, anticipated improvements (listed and discussed below) should overcome some of the aforementioned limitations. The prediction of the hybridization behaviour of the probes is still rudimentary and requires improvement. This can be achieved via input from bioinformatics and the establishment of datasets on array-specific effects such as immobilisation and steric hindrance. Large databases, comparable in size to that of the 16S rRNA databases, are critically needed for high resolution phylogenetic markers. Such markers will drastically improve the applicability of MDMs for clinical and food microbiology, epidemiology and related fields. If, in such databases, sequence information is linked to clinical traits (e.g. pathogenicity, host specificity, antibiotic resistance, geographic origin, etc.), MDM-based detection will serve the additional purpose of providing a prediction of these functions, at least on some of them, at a given level of certainty. Some enzyme-based labelling methods [86 – 88,117] have demonstrated the potential for an improved detection sensitivity (down to about 0.1% of the total community targeted), although at the cost of losing all quantification potential. Current hybridization platforms predominantly use fluorescence-based optical detection. Both the fluorescent label and the detection devices confer high costs to this technology, seriously limiting its spread across diagnostic laboratories. With a few exceptions, hybridization results are read after irreversible processing steps, thus the result reflects hybridization under a selected, suboptimal hybridization condition. Alternative detection platforms are thus being developed and elaborated which, if successful, could have two major advantages. One is reduced costs, the other is on-line detection of hybridization events. Electronic and mechanical detection methods [116,118 – 121] require cheaper or no labelling and simpler, cheaper readout devices. Label-free detection may also help in cutting the time-to-result, a critical issue in many medical applications. On-line detection of the hybridization event may be possible via a number of techniques [8 –10,116,118– 125]. Such an improvement will drastically increase the information obtainable from any hybridization event. Association and dissociation kinetics and maximum hybridization signal at the optimal hybridization stringency for each individual probe can all be obtained. By considering these extra information, the reliability of MDM-based diagnostics can be further improved. On-line hybridization detection is also another means of cutting the time-toresult. A label-free on-line hybridization detection approach could make it possible to obtain results within less than 2 h.

116

A. Loy, L. Bodrossy / Clinica Chimica Acta 363 (2006) 106 – 119

Multiple marker genes, providing reliable information on the phylogeny and identity of the detected microbe and its clinical traits, should be used on a single MDM in many cases to provide all the relevant information to the diagnostic laboratory. As discussed earlier, multiplex amplification is still a serious limitation to this approach. A true on-chip PCR where also the amplification primers are immobilised would enable unlimited multiplexing, solving this problem and opening new horizons in the field.

Acknowledgements Related work at Seibersdorf research (LB) was funded by the Fonds zur Fo¨rderung der wissenschaftlichen Forschung, Austria (P15044) and through the EU 5th Framework Quality of Life and Management of Living Resources Grant QLK-3 CT-2000-01528. Research of AL was funded by a Marie Curie Intra-European Fellowship within the 6th European Community Framework Programme. Funding from the EU towards COST Action 853 contributed by enhancing exchange of ideas. The authors thank Michael Taylor for critical revision of the manuscript. AL is greatly indebted to Michael Wagner for ongoing support. LB is indebted to Fodor Szilvia for her support and fresh, independent views and ideas. Helpful comments and discussions from many active members of the Yahoo discussion groups on microarrays and microbial diagnostics/microbial ecology (http://groups.yahoo.com/group/ microarray/ and http://groups.yahoo.com/group/MDME/) are acknowledged.

References [1] Bodrossy L, Stralis-Pavese N, Murrell JC, Radajewski S, Weilharter A, Sessitsch A. Development and validation of a diagnostic microbial microarray for methanotrophs. Environ Microbiol 2003;5: 566 – 82. [2] Zhou J. Microarrays for bacterial detection and microbial community analysis. Curr Opin Microbiol 2003;6:288 – 94. [3] Letowski J, Brousseau R, Masson L. DNA microarray applications in environmental microbiology. Anal Lett 2003;36:3165 – 84. [4] Stenger D, Andreadis J, Voraa G, Pancrazio J. Potential applications of DNA microarrays in biodefense-related diagnostics. Curr Opin Biotechnol 2002;13:208 – 12. [5] Clewley JP. A role for arrays in clinical virology: fact or fiction? J Clin Virol 2004;29:2 – 12. [6] Gershon D. Microarrays go mainstream. Nature Methods 2004;1: 263 – 70. [7] Guschin DY, Mobarry BK, Proudnikov D, Stahl DA, Rittmann BE, Mirzabekov AD. Oligonucleotide microchips as genosensors for determinative and environmental studies in microbiology. Appl Environ Microbiol 1997;63:2397 – 402. [8] Fotin AV, Drobyshev AL, Proudnikov DY, Perov AN, Mirzabekov AD. Parallel thermodynamic analysis of duplexes on oligodeoxyribonucleotide microchips. Nucleic Acids Res 1998;26:1515 – 21. [9] Anthony RM, Schuitema ARJ, Oskam L, Klatser PR. Direct detection of Staphylococcus aureus mRNA using a flow through microarray. J Microbiol Methods 2005;60:47 – 54.

[10] Wu Y, de Kievit P, Vahlkamp L, et al. Quantitative assessment of a novel flow-through porous microarray for the rapid analysis of gene expression profiles. Nucleic Acids Res 2004;32:e123. [11] Urakawa H, Noble PA, El Fantroussi S, Kelly JJ, Stahl DA. Singlebase-pair discrimination of terminal mismatches by using oligonucleotide microarrays and neural network analyses. Appl Environ Microbiol 2002;68:235 – 44. [12] Urakawa H, El Fantroussi S, Smidt H, et al. Optimization of singlebase-pair mismatch discrimination in oligonucleotide microarrays. Appl Environ Microbiol 2003;69:2848 – 56. [13] Bodrossy L, Sessitsch A. Oligonucleotide microarrays in microbial diagnostics. Curr Opin Microbiol 2004;7:245 – 54. [14] Ludwig W, Schleifer KH. Phylogeny of bacteria beyond the 16S rRNA standard. ASM News 1999;65:1515 – 21. [15] Loy A, Lehner A, Lee N, et al. Oligonucleotide microarray for 16S rRNA gene-based detection of all recognized lineages of sulfatereducing prokaryotes in the environment. Appl Environ Microbiol 2002;68:5064 – 81. [16] Loy A, Horn M, Wagner M. probeBase: an online resource for rRNA-targeted oligonucleotide probes. Nucleic Acids Res 2003;31: 514 – 6. [17] Wilson KH, Wilson WJ, Radosevich JL, et al. High-density microarray of small-subunit ribosomal DNA probes. Appl Environ Microbiol 2002;68:2535 – 41. [18] Warsen AE, Krug MJ, LaFrentz S, Stanek DR, Loge FJ, Call DR. Simultaneous discrimination between 15 fish pathogens by using 16S ribosomal DNA PCR and DNA microarrays. Appl Environ Microbiol 2004;70:4216 – 21. [19] Ludwig W, Strunk O, Westram R, et al. ARB: a software environment for sequence data. Nucleic Acids Res 2004;32:1363 – 71. [20] Cole JR, Chai B, Marsh TL, et al. The ribosomal database project (RDP-II): previewing a new autoaligner that allows regular updates and the new prokaryotic taxonomy. Nucleic Acids Res 2003;31: 442 – 3. [21] Mitterer G, Huber M, Leidinger E, et al. Microarray-based identification of bacteria in clinical samples by solid-phase PCR amplification of 23S ribosomal DNA sequences. J Clin Microbiol 2004;42:1048 – 57. [22] Nubel U, Schmidt PM, Reiss E, Bier F, Beyer W, Naumann D. Oligonucleotide microarrays for identification of Bacillus anthracis based on intergenic transcribed spacers in ribosomal DNA. FEMS Microbiol Lett 2004;240:215 – 23. [23] Dahllof I, Baillie H, Kjelleberg S. rpoB-based microbial community analysis avoids limitations inherent in 16S rRNA gene intraspecies heterogeneity. Appl Environ Microbiol 2000;66:3376 – 80. [24] Drancourt M, Roux V, Fournier PE, Raoult D. rpoB gene sequencebased identification of aerobic Gram-positive cocci of the genera Streptococcus, Enterococcus, Gemella, Abiotrophia, and Granulicatella. J Clin Microbiol 2004;42:497 – 504. [25] Mollet C, Drancourt M, Raoult D. rpoB sequence analysis as a novel basis for bacterial identification. Mol Microbiol 1997;26: 1005 – 1011. [26] Brisse S, Verhoef J. Phylogenetic diversity of Klebsiella pneumoniae and Klebsiella oxytoca clinical isolates revealed by randomly amplified polymorphic DNA, gyrA and parC genes sequencing and automated ribotyping. Int J Syst Evol Microbiol 2001;51: 915 – 24. [27] Kakinuma K, Fukushima M, Kawaguchi R. Detection and identification of Escherichia coli, Shigella, and Salmonella by microarrays using the gyrB gene. Biotechnol Bioeng 2003;83:721 – 8. [28] Lloyd AT, Sharp PM. Evolution of the recA gene and the molecular phylogeny of bacteria. J Mol Evol 1993;37:399 – 407. [29] Baldauf SL, Palmer JD, Doolittle WF. The root of the universal tree and the origin of eukaryotes based on elongation factor phylogeny. Proc Natl Acad Sci U S A 1996;93:7749 – 54. [30] Ludwig W, Neumaier J, Klugbauer N, et al. Phylogenetic relationships of Bacteria based on comparative sequence analysis of

A. Loy, L. Bodrossy / Clinica Chimica Acta 363 (2006) 106 – 119

[31]

[32]

[33] [34]

[35]

[36]

[37]

[38]

[39]

[40]

[41]

[42]

[43]

[44]

[45]

[46]

[47]

[48]

[49]

elongation factor Tu and ATP-synthase beta-subunit genes. Antonie Van Leeuwenhoek 1993;64:285 – 305. Wong RSY, Chow AW. Identification of enteric pathogens by heat shock protein 60 kDa (HSP60) gene sequences. FEMS Microbiol Lett 2002;206:107 – 13. Wertz JE, Goldstone C, Gordon DM, Riley MA. A molecular phylogeny of enteric bacteria and implications for a bacterial species concept. J Evol Biol 2003;16:1236 – 48. Zwieb C, Gorodkin J, Knudsen B, Burks J, Wower J. tmRDB (tmRNA database). Nucleic Acids Res 2003;31:446 – 7. Sergeev N, Volokhov D, Chizhikov V, Rasooly A. Simultaneous analysis of multiple staphylococcal enterotoxin genes by an oligonucleotide microarray assay. J Clin Microbiol 2004;42: 2134 – 43. Chizhikov V, Rasooly A, Chumakov K, Levy DD. Microarray analysis of microbial virulence factors. Appl Environ Microbiol 2001;67:3258 – 63. Volokhov D, Rasooly A, Chumakov K, Chizhikov V. Identification of Listeria species by microarray-based assay. J Clin Microbiol 2002; 40:4720 – 8. Saunders NA, Underwood A, Kearns AM, Hallas G. A virulenceassociated gene microarray: a tool for investigation of the evolution and pathogenic potential of Staphylococcus aureus. Microbiology 2004;150:3763 – 71. Grimm V, Ezaki S, Susa M, Knabbe C, Schmid RD, Bachmann TT. Use of DNA microarrays for rapid genotyping of TEM betalactamases that confer resistance. J Clin Microbiol 2004;42: 3766 – 3774. Stralis-Pavese N, Sessitsch A, Weilharter A, et al. Optimisation of diagnostic microarray for application in analysing landfill methanotroph communities under different plant covers. Environ Microbiol 2004;6:347 – 63. Rhee SK, Liu X, Wu L, Chong SC, Wan X, Zhou J. Detection of genes involved in biodegradation and biotransformation in microbial communities by using 50-mer oligonucleotide microarrays. Appl Environ Microbiol 2004;70:4303 – 17. Taroncher-Oldenburg G, Griner EM, Francis CA, Ward BB. Oligonucleotide microarray for the study of functional gene diversity in the nitrogen cycle in the environment. Appl Environ Microbiol 2003;69:1159 – 71. Wu L, Thompson DK, Li G, Hurt RA, Tiedje JM, Zhou J. Development and evaluation of functional gene arrays for detection of selected genes in the environment. Appl Environ Microbiol 2001; 67:5780 – 90. Hashsham SA, Wick LM, Rouillard JM, Gulari E, Tiedje JM. Potential of DNA microarrays for developing parallel detection tools (PDTs) for microorganisms relevant to biodefense and related research needs. Biosens Bioelectron 2004;20:668 – 83. Wetmur JG. DNA probes: applications of the principles of nucleic acid hybridization. Crit Rev Biochem Mol Biol 1991;26: 227 – 259. Hughes TR, Mao M, Jones AR, et al. Expression profiling using microarrays fabricated by an ink-jet oligonucleotide synthesizer. Nat Biotechnol 2001;19:342 – 7. Kane MD, Jatkoe TA, Stumpf CR, Lu J, Thomas JD, Madore SJ. Assessment of the sensitivity and specificity of oligonucleotide (50mer) microarrays. Nucleic Acids Res 2000;28:4552 – 7. Tiquia SM, Wu L, Chong SC, et al. Evaluation of 50-mer oligonucleotide arrays for detecting microbial populations in environmental samples. BioTechniques 2004;36:664 – 5. Letowski J, Brousseau R, Masson L. Designing better probes: effect of probe size, mismatch position and number on hybridization in DNA oligonucleotide microarrays. J Microbiol Methods 2004;57: 269 – 78. Vora GJ, Meador CE, Stenger DA, Andreadis JD. Nucleic acid amplification strategies for DNA microarray-based pathogen detection. Appl Environ Microbiol 2004;70:3047 – 54.

117

[50] Wang D, Coscoy L, Zylberberg M, et al. Microarray-based detection and genotyping of viral pathogens. Proc Natl Acad Sci U S A 2002; 99:15687 – 92. [51] Denef VJ, Park J, Rodrigues JL, Tsoi TV, Hashsham SA, Tiedje JM. Validation of a more sensitive method for using spotted oligonucleotide DNA microarrays for functional genomics studies on bacterial communities. Environ Microbiol 2003;5:933 – 43. [52] Rouillard JM, Zuker M, Gulari E. OligoArray 2.0: design of oligonucleotide probes for DNA microarrays using a thermodynamic approach. Nucleic Acids Res 2003;31:3057 – 62. [53] Gordon PM, Sensen CW. Osprey: a comprehensive tool employing novel methods for the design of oligonucleotides for DNA sequencing and microarrays. Nucleic Acids Res 2004;32:e133. [54] Behr T, Koob C, Schedl M, et al. A nested array of rRNA targeted probes for the detection and identification of enterococci by reverse hybridization. Syst Appl Microbiol 2000;23:563 – 72. [55] Wagner M, Horn M, Daims H. Fluorescence in situ hybridization for the identification and characterisation of prokaryotes. Curr Opin Microbiol 2003;6:302 – 9. [56] Loy A., Schulz C., Lu¨cker S., et al. 16S rRNA gene-based oligonucleotide microarray for environmental monitoring of the betaproteobacterial order ‘‘Rhodocyclales’’. Appl Environ Microbiol 2005;71(3):1373 – 86. [57] Bodrossy L. Diagnostic oligonucleotide microarrays for microbiology. In: Blalock E, editor. A beginner’s guide to microarrays, 1 ed. New York’ Kluwer Academic Publishers; 2003. p. 43 – 92. [58] Peplies J, Lau SC, Pernthaler J, Amann R, Glockner FO. Application and validation of DNA microarrays for the 16S rRNA-based analysis of marine bacterioplankton. Environ Microbiol 2004;6:638 – 45. [59] Zuker M. Mfold web server for nucleic acid folding and hybridization prediction. Nucleic Acids Res 2003;31:3406 – 15. [60] SantaLucia Jr J. A unified view of polymer, dumbbell, and oligonucleotide DNA nearest-neighbor thermodynamics. Proc Natl Acad Sci U S A 1998;95:1460 – 5. [61] Pfunder M, Frey JE. Dissociation analysis in PCR- and 1xSSCbuffer as a prerequisite for selection of 13mer microarray probe sets with uniform hybiridzation behavior. Mol Biotechnol 2005; 29(1):1 – 10. [62] Al Khaldi SF, Myers KM, Rasooly A, Chizhikov V. Genotyping of Clostridium perfringens toxins using multiple oligonucleotide microarray hybridization. Mol Cell Probes 2004;18:359 – 67. [63] Volokhov D, Pomerantsev A, Kivovich V, Rasooly A, Chizhikov V. Identification of Bacillus anthracis by multiprobe microarray hybridization. Diagn Microbiol Infect Dis 2004;49:163 – 71. [64] Liu WT, Mirzabekov AD, Stahl DA. Optimization of an oligonucleotide microchip for microbial identification studies: a nonequilibrium dissociation approach. Environ Microbiol 2001;3: 619 – 29. [65] Huang S, Wang Y, Chen P, Qian HR, Yeo A, Bemis K. Sum: a new way to incorporate mismatch probe measurements. Genomics 2004; 84:767 – 77. [66] Lee I, Dombkowski AA, Athey BD. Guidelines for incorporating non-perfectly matched oligonucleotides into target-specific hybridization probes for a DNA microarray. Nucleic Acids Res 2004; 32:681 – 90. [67] Shchepinov MS, Case-Green SC, Southern EM. Steric factors influencing hybridization of nucleic acids to oligonucleotide arrays. Nucleic Acids Res 1997;25:1155 – 61. [68] Anthony RM, Brown TJ, French GL. Rapid diagnosis of bacteremia by universal amplification of 23S ribosomal DNA followed by hybridization to an oligonucleotide array. J Clin Microbiol 2000; 38:781 – 8. [69] Matveeva OV, Shabalina SA, Nemtsov VA, Tsodikov AD, Gesteland RF, Atkins JF. Thermodynamic calculations and statistical correlations for oligo-probes design. Nucleic Acids Res 2003;31: 4211 – 7.

118

A. Loy, L. Bodrossy / Clinica Chimica Acta 363 (2006) 106 – 119

[70] Lane S, Evermann J, Loge F, Call DR. Amplicon secondary structure prevents target hybridization to oligonucleotide microarrays. Biosens Bioelectron 2004;20:728 – 35. [71] Southern EM. DNA fingerprinting by hybridization to oligonucleotide arrays. Electrophoresis 1995;16:1539 – 42. [72] Zhang XX, Kosier B, Priefer UB. Genetic diversity of indigenous Rhizobium leguminosarum bv. viciae isolates nodulating two different host plants during soil restoration with alfalfa. Mol Ecol 2001; 10:2297 – 305. [73] Chandler DP, Newton GJ, Small JA, Daly DS. Sequence versus structure for the direct detection of 16S rRNA on planar oligonucleotide microarrays. Appl Environ Microbiol 2003;69: 2950 – 8. [74] Maskos U, Southern EM. Parallel analysis of oligodeoxyribonucleotide (oligonucleotide) interactions. I. Analysis of factors influencing oligonucleotide duplex formation. Nucleic Acids Res 1992;20: 1675 – 8. [75] Jacobs KA, Rudersdorf R, Neill SD, Dougherty JP, Brown EL, Fritsch EF. The thermal stability of oligonucleotide duplexes is sequence independent in tetraalkylammonium salt solutions: application to identifying recombinant DNA clones. Nucleic Acids Res 1988;16:4637 – 50. [76] Volokhov D, Chizhikov V, Chumakov K, Rasooly A. Microarraybased identification of thermophilic Campylobacter jejuni, C. coli, C. lari, and C. upsaliensis. J Clin Microbiol 2003;41:4071 – 80. [77] Bonch-Osmolovskaya EA, Miroshnichenko ML, Lebedinsky AV, et al. Radioisotopic, culture-based, and oligonucleotide microchip analyses of thermophilic microbial communities in a continental high-temperature petroleum reservoir. Appl Environ Microbiol 2003; 69:6143 – 51. [78] Koizumi Y, Kelly JJ, Nakagawa T, et al. Parallel characterization of anaerobic toluene-and ethylbenzene-degrading microbial consortia by PCR-denaturing gradient gel electrophoresis, RNA-DNA membrane hybridization and DNA microarray technology. Appl Environ Microbiol 2002;68:3215 – 25. [79] Wang Z, Vora GJ, Stenger DA. Detection and genotyping of Entamoeba histolytica, Entamoeba dispar, Giardia lamblia, and Cryptosporidium parvum by oligonucleotide microarray. J Clin Microbiol 2004;42:3262 – 71. [80] Lin B, Vora GJ, Thach D, et al. Use of oligonucleotide microarrays for rapid detection and serotyping of acute respiratory diseaseassociated adenoviruses. J Clin Microbiol 2004;42:3232 – 9. [81] Boriskin YS, Rice PS, Stabler RA, et al. DNA microarrays for virus detection in cases of central nervous system infection. J Clin Microbiol 2004;42:5811 – 8. [82] Panicker G, Call DR, Krug MJ, Bej AK. Detection of pathogenic vibrio spp. in shellfish by using multiplex PCR and DNA microarrays. Appl Environ Microbiol 2004;70:7436 – 44. [83] Rudi K, Rud I, Holck A. A novel multiplex quantitative DNA array based PCR (MQDA-PCR) for quantification of transgenic maize in food and feed. Nucleic Acids Res 2003;31:e62. [84] Ericsson O, Sivertsson A, Lundeberg J, Ahmadian A. Microarraybased resequencing by apyrase-mediated allele-specific extension. Electrophoresis 2003;24:3330 – 8. [85] Lindroos K, Sigurdsson S, Johansson K, Ronnblom L, Syvanen AC. Multiplex SNP genotyping in pooled DNA samples by a four-color microarray system. Nucleic Acids Res 2002;30:e70. [86] Rudi K, Treimo J, Nissen H, Vegarud G. Protocols for 16S rDNA array analyses of microbial communities by sequence-specific labeling of DNA probes. Sci World J 2003;3:578 – 84. [87] Busti E, Bordoni R, Castiglioni B, et al. Bacterial discrimination by means of a universal array approach mediated by LDR (ligase detection reaction). BMC Microbiol 2002;2:27. [88] Gharizadeh B, Kaller M, Nyren P, et al. Viral and microbial genotyping by a combination of multiplex competitive hybridization and specific extension followed by hybridization to generic tag arrays. Nucleic Acids Res 2003;31:e146.

[89] Lovmar L, Fock C, Espinoza F, Bucardo F, Syvanen AC, Bondeson K. Microarrays for genotyping human group a rotavirus by multiplex capture and type-specific primer extension. J Clin Microbiol 2003; 41:5153 – 8. [90] Mikhailovich V, Lapa S, Gryadunov D, et al. Identification of rifampicin-resistant Mycobacterium tuberculosis strains by hybridization, PCR, and ligase detection reaction on oligonucleotide microchips. J Clin Microbiol 2001;39:2531 – 40. [91] Schena M. Microarray analysis. Hoboken, Canada’ John Wiley and Sons, Inc.; 2003. [92] Rudi K, Treimo J, Moen B, Rud I, Vegarud G. Internal controls for normalizing DNA arrays. BioTechniques 33 (2002) 496, 498, 500. [93] Cho JC, Tiedje JM. Quantitative detection of microbial genes by using DNA microarrays. Appl Environ Microbiol 2002;68: 1425 – 30. [94] Straub TM, Daly DS, Wunshel S, Rochelle PA, DeLeon R, Chandler DP. Genotyping Cryptosporidium parvum with an hsp70 singlenucleotide polymorphism microarray. Appl Environ Microbiol 2002; 68:1817 – 26. [95] Gonzalez SF, Krug MJ, Nielsen ME, Santos Y, Call DR. Simultaneous detection of marine fish pathogens by using multiplex PCR and a DNA microarray. J Clin Microbiol 2004;42:1414 – 9. [96] Keramas G, Bang DD, Lund M, et al. Use of culture, PCR analysis, and DNA microarrays for detection of Campylobacter jejuni and Campylobacter coli from chicken feces. J Clin Microbiol 2004;42: 3985 – 91. [97] Keramas G, Bang DD, Lund M, et al. Development of a sensitive DNA microarray suitable for rapid detection of Campylobacter spp. Mol Cell Probes 2003;17:187 – 96. [98] Castiglioni B, Rizzi E, Frosini A, et al. Development of a universal microarray based on the ligation detection reaction and 16S rRNA gene polymorphism to target diversity of cyanobacteria. Appl Environ Microbiol 2004;70:7161 – 72. [99] Lehner A, Loy A, Behr T, Gaenge H, Schleifer K-H, Wagner M. Oligonucleotide microarray for simultaneous identification and differentiation of Enterococcus species in food samples. FEMS Microbiol Lett 2005;246(1):133 – 42. [100] Sergeev N, Distler M, Courtney S, et al. Multipathogen oligonucleotide microarray for environmental and biodefense applications. Biosens Bioelectron 2004;20:684 – 98. [101] Peplies J, Lau SCK, Pernthaler J, Amann R, Glockner FO. Application and validation of DNA microarrays for the 16S rRNAbased analysis of marine bacterioplankton. Appl Environ Microbiol 2005;71(5):2723 – 31. [102] Fukushima M, Kakinuma K, Hayashi H, Nagai H, Ito K, Kawaguchi R. Detection and identification of Mycobacterium species isolates by DNA microarray. J Clin Microbiol 2003;41:2605 – 15. [103] Deng JY, Zhang XE, Lu HB, et al. Multiplex detection of mutations in clinical isolates of rifampin-resistant Mycobacterium tuberculosis by short oligonucleotide ligation assay on DNA chips. J Clin Microbiol 2004;42:4850 – 2. [104] Yu X, Susa M, Knabbe C, Schmid RD, Bachmann TT. Development and validation of a diagnostic DNA microarray to detect quinoloneresistant Escherichia coli among clinical isolates. J Clin Microbiol 2004;42:4083 – 91. [105] Roth SB, Jalava J, Ruuskanen O, Ruohola A, Nikkari S. Use of an oligonucleotide array for laboratory diagnosis of bacteria responsible for acute upper respiratory infections. J Clin Microbiol 2004;42: 4268 – 74. [106] Loy A, Kusel K, Lehner A, Drake HL, Wagner M. Microarray and functional gene analyses of sulfate-reducing prokaryotes in lowsulfate, acidic fens reveal cooccurrence of recognized genera and novel lineages. Appl Environ Microbiol 2004;70:6998 – 7009. [107] Wilson WJ, Strout CL, DeSantis TZ, Stilwell JL, Carrano AV, Andersen GL. Sequence-specific identification of 18 pathogenic microorganisms using microarray technology. Mol Cell Probes 2002; 16:119 – 27.

A. Loy, L. Bodrossy / Clinica Chimica Acta 363 (2006) 106 – 119 [108] Chizhikov V, Wagner M, Ivshina A, Hoshino Y, Kapikian AZ, Chumakov K. Detection and genotyping of human group a rotaviruses by oligonucleotide microarray hybridization. J Clin Microbiol 2002;40:2398 – 407. [109] Klaassen CH, Prinsen CF, de Valk HA, Horrevorts AM, Jeunink MA, Thunnissen FB. DNA microarray format for detection and subtyping of human papillomavirus. J Clin Microbiol 2004;42: 2152 – 60. [110] Delrio-Lafreniere SA, Browning MK, McGlennen RC. Low-density addressable array for the detection and typing of the human papillomavirus. Diagn Microbiol Infect Dis 2004;48:23 – 31. [111] Oh TJ, Kim CJ, Woo SK, et al. Development and clinical evaluation of a highly sensitive DNA microarray for detection and genotyping of human papillomaviruses. J Clin Microbiol 2004;42: 3272 – 80. [112] Sengupta S, Onodera K, Lai A, Melcher U. Molecular detection and identification of influenza viruses by oligonucleotide microarray hybridization. J Clin Microbiol 2003;41:4542 – 50. [113] Kessler N, Ferraris O, Palmer K, Marsh W, Steel A. Use of the DNA flow-thru chip, a three-dimensional biochip, for typing and subtyping of influenza viruses. J Clin Microbiol 2004;42:2173 – 85. [114] Laassri M, Chizhikov V, Mikheev M, Shchelkunov S, Chumakov K. Detection and discrimination of orthopoxviruses using microarrays of immobilized oligonucleotides. J Virol Methods 2003;112:67 – 78. [115] Lapa S, Mikheev M, Shchelkunov S, et al. Species-level identification of orthopoxviruses with an oligonucleotide microchip. J Clin Microbiol 2002;40:753 – 7. [116] Liu RH, Yang J, Lenigk R, Bonanno J, Grodzinski P. Self-contained, fully integrated biochip for sample preparation, polymerase chain reaction amplification, and DNA microarray detection. Anal Chem 2004;76:1824 – 31.

119

[117] Baner J, Isaksson A, Waldenstrom E, Jarvius J, Landegren U, Nilsson M. Parallel gene analysis with allele-specific padlock probes and tag microarrays. Nucleic Acids Res 2003;31:e103. [118] McKendry R, Zhang J, Arntz Y, et al. Multiple label-free biodetection and quantitative DNA-binding assays on a nanomechanical cantilever array. Proc Natl Acad Sci U S A 2002;99:9783 – 8. [119] Gabig-Ciminska M, Los M, Holmgren A, et al. Detection of bacteriophage infection and prophage induction in bacterial cultures by means of electric DNA chips. Anal Biochem 2004;324:84 – 91. [120] Zimmermann K, Eiter T, Scheiflinger F. Consecutive analysis of bacterial PCR samples on a single electronic microarray. J Microbiol Methods 2003;55:471 – 4. [121] Westin L, Miller C, Vollmer D, et al. Antimicrobial resistance and bacterial identification utilizing a microelectronic chip array. J Clin Microbiol 2001;39:1097 – 104. [122] Meunier-Prest R, Raveau S, Finot E, Legay G, Cherkaoui-Malki M, Latruffe N. Direct measurement of the melting temperature of supported DNA by electrochemical method. Nucleic Acids Res 2003;31:e150. [123] Stimpson DI, Hoijer JV, Hsieh WT, et al. Real-time detection of DNA hybridization and melting on oligonucleotide arrays by using optical wave guides. Proc Natl Acad Sci U S A 1995;92:6379 – 83. [124] Nelson BP, Liles MR, Frederick KB, Corn RM, Goodman RM. Label-free detection of 16S ribosomal RNA hybridization on reusable DNA arrays using surface plasmon resonance imaging. Environ Microbiol 2002;4:735 – 43. [125] Podsadlowski V, Peter C, Cammann K, Brunen-Nieweler C, Borchers T. Rapid species identification in food samples with an evanescent-field-DNA-biosensor-chip. 2004. Statusseminar Chip Technologies: Microarrays, Hygiene and Health; Frankfurt a. Main, 26.-27.1.2004.