Microarray technology: the future of blood testing? .fr

RNA-derived) would also be desirable, as serious and some- times fatal reactions owing to bacterial contamination of blood components continue to occur [38].
481KB taille 10 téléchargements 343 vues
VOX010.fm Page 1 Wednesday, February 28, 2001 4:18 PM

Vox Sanguinis (2001) 80, 1–11 © 2001 Blackwell Science

REVIEW

Microarray technology: the future of blood testing? Blackwell Science, Ltd

J. Petrik Scottish National Blood Transfusion Service, Edinburgh, UK

Received: 15 July 2000, revised 19 September 2000, accepted 10 October 2000

The increasing pace of development in molecular biological techniques during the last 10–15 years has had a direct effect on mass testing and diagnostic applications, including blood screening. Nucleic acid amplification techniques (NAT), usually based on the polymerase chain reaction (PCR), have been successfully applied to blood grouping and implemented recently in screening of blood donations for hepatitis C virus (HCV). The majority of microarray technologies involve an amplification step, yet the main benefits of this technology come from simultaneous analysis of thousands of analytes. Microarrays were developed to utilize the huge amount of information provided by genome projects, but they have clear potential in mass screening and diagnostics. The application of microarray technology may revolutionize blood testing, providing for the first time the prospect of an integrated platform for comprehensive donor and donation testing, replacing multiple individual assays. Design features of a blood-testing chip and various technologies with potential application in this field are discussed in this review.

Introduction Until the 1990s, blood screening, typing and diagnostics (blood testing for the purpose of this review) depended entirely on serological (antigen/antibody) techniques. In the past 7–10 years, nucleic acid amplification techniques (NAT), predominantly polymerase chain reaction (PCR)-based techniques [1] have been applied to blood group, human leucocyte antigen (HLA) type and platelet-specific antigen determination, as in many cases they can provide more precise results. Mandatory PCR screening of plasma pools for hepatitis C virus (HCV) RNA is being extended as a criterion for release of red cells and platelets. Implementing PCR as a routine screening technique represented a major technological challenge and was greeted with profound scepticism by many blood transfusion professionals; however, it took a considerably shorter time than envisaged to implement it. It is fair to say that the early implementation was driven by manufacturers rather than by blood transfusion Correspondence: Juraj Petrik, Scottish National Blood Transfusion Service, NAT Reference Lab, LCMV, Royal (Dick) Veterinary College, Summerhall, Edinburgh EH9 1QH Tel.: 0131-650-7841 Fax: 0131-650-7965 E-mail: [email protected]

services, and the preparedness varied considerably. Nevertheless, at least two lessons should be learnt from this exercise. First, if there is a new technology available that is superior in some respect (faster, more reliable, efficient, reproducible, precise or sensitive) to current techniques, it is probable that it will sooner or later be implemented, even if the initial cost–benefit analysis may not be favourable. As a rule, the cost decreases dramatically with the increase in use of the technology. Second, blood services should be ready for new technologies applicable to blood screening and the best way to achieve this is to be involved in their development. Perhaps the most important consideration is that the current PCR assays, while technically demanding for routine use, represent additional, complementary and, in some cases, replacement assays, but they do not change the fundamental operational aspects of blood testing and management. In contrast, microarray technology has the potential to radically change many aspects of the blood supply chain. A biochip capable of combining microbiology tests with blood typing and platelet assays would offer the opportunity for reengineering transfusion centre and blood bank operations.

Microarray technology The concept of an array as an ordered collection of molecules is not entirely new. Southern blot [2] is a classical example,

1

VOX010.fm Page 2 Wednesday, February 28, 2001 4:18 PM

2 J. Petrik

especially in the form of dot-blot or gridded library. Microarray has larger density (> 100 probe elements/cm2) than array, and development of the current miniaturized version of a high-throughput microarray required a number of technological improvements. One of the crucial changes was the replacement of porous hybridization membranes with non-porous glass or silicone support, allowing the use of much smaller (nl to pl) reaction volumes, driving up the reaction kinetics and dramatically reducing reaction times. Fluorescent dyes, which cause a high background on membrane supports, are the method of choice for sample labelling for microarray experiments on non-porous supports, providing an option for simultaneous use of two or more fluorophores and signal quantification. Thousands to hundreds of thousands of reactions can be performed in miniature, simultaneously under identical conditions, providing a solution to continuously increasing throughput demands from large genomic, pharmacogenetic and proteomic programmes. The term microarray is mostly perceived as for DNA rather than for protein microarrays. This reflects first the logic of genome research, from the identification of the sequence information in large genomic projects to functional characterization of the products of this information and their interactions, as provided by proteomics. Second, the rules governing base pairing of two complementary nucleic acid strands are rather simpler than those for more complex protein–protein interactions. Most of the work to date has been performed on DNA microarrays, reflecting the availability of sequence information required for systematic DNA microarray construction. However, numerous proof-of-principle experiments have been carried out in antibody-based assays for protein and small-molecule analytes [3]. Tables 1 and 2 list various formats of DNA and protein microarrays, respectively. The inclusion in either of the two tables is, however, rather arbitrary for some technologies (marked by a), which are in fact applicable to nucleic acids, proteins and sometimes to other types of molecules.

DNA microarrays Oligonucleotides or products of DNA amplification (usually of the PCR) can be deposited on the solid support in several ways (Table 1). In a mechanical microspotting method pioneered by the Stanford group [4,5] a robot (arrayer) was used to spot cDNAs onto chemically treated glass slides by direct surface contact between the slide and a set of metallic printing pens. Each pen can deposit 0·25–1 nl, creating spots of 100 –150 µm in diameter with 200–250 µm between the centre of each spot [6]. Contact printing is very flexible but sometimes produces irregular spots and slide-to-slide variation. An alternative, non-contact method of depositing cDNAs or presynthesized oligonucleotides is the use of inkjet technology. Bubble jet variation developed by Canon (Tokyo,

Japan) [7] has the potential to reduce spot size to 25–30 µm, making it possible to fit ≈ 100 000 gene spots onto a single 4-cm2 chip [8]. Oligonucleotides can also be manufactured on-chip by adapting the photolithographic techniques used by the semiconductor industry, a method developed by Fodor et al. and marketed by Affymetrix (Santa Clara, CA) [9]. Such an approach reduces chip-to-chip variability and eliminates individual clone amplification, storage and handling, but the cost of masks for each step of synthesis is high. The current upper limit for microarray density is ≈ 250 000 features/cm2. A potentially cheaper, maskless version of oligonucleotide array fabrication has been described recently [10]. The sample, rather than the probe, is labelled. This is in contrast to Southern blot where labelled probe is in vast excess to the target in order to reach saturation hybridization (100% coverage). On the other hand, fluorescently labelled mRNA or cDNA is hybridized only to a proportion (< 1%) of probe molecules on microarray and the signal intensity at each position can be determined by use of a confocal microscope or with a charge-coupled device (CCD) camera [11]. One of the most frequent applications of DNA microarrays is the comparison of expression profiles of labelled RNA from cells or tissues differing in a well-defined manner, such as by exposure to drugs, temperature, cytokines, etc. [12]. Expression of complex viral genomes can also be studied successfully [13]. Changes in the expression of particular genes can be linked to the action of particular stimuli, to a particular physiological stage, metabolic or signalling pathway, ultimately suggesting the function of the gene product [14–17]. Another major application involves studies of DNA mutations focusing on single nucleotide polymorphism (SNP). A population-based knowledge of SNPs might be able to explain their contribution to human disease as well as individual differences in, for example, therapeutic efficiency of various drugs [18,19].

Protein microarrays Sequence information, even in combination with the expression profiling data at RNA level, may not be able to provide a comprehensive picture of protein functions. In many cases, protein-expression data [20] and protein–protein interactions will need to be determined, especially when multiple proteins and other molecules interact within a complex. Similarly in diagnostics, there are protein assays that cannot simply be replaced by genomic testing, in many cases because of the variability in genetic background of a tested marker. Until recently, most of the protein arrays used porous materials, although sometimes at reasonably high density, as described by Lueking et al. for the human fetal brain cDNA expression library [21], or experimenting with novel deposition devices such as the inkjet printer [22]. © 2001 Blackwell Science Ltd. Vox Sanguinis (2001) 80, 1–11

VOX010.fm Page 3 Wednesday, February 28, 2001 4:18 PM

© 2001 Blackwell Science Ltd. Vox Sanguinis (2001) 80, 1– 11

Table 1 Various formats of DNA chip Chip/company

Matrix

Probe

Deposition

Array size

Detection

Sensitivity

Applications

Reference

GeneChip Affymetrix

Silicon wafers

Oligonucleotides 20–25 mers

On-chip photolithographic synthesis

Up to 400 000/1·6 cm2

Fluorescence

1 : 100 000

Expression profiling SNPs

[9]

Stanford group-type array

Glass

cDNA; prefabricated oligonucleotides

Mechanical spotting

10 000/cm2

Fluorescence

1 : 100 000

Expression profiling; SNPs

[4], [5], [17]

MAGIChip

Polyacrylamide patches on a glass slide

Oligonucleotides 40-mers; DNA up to 1 kb; (proteins up to 400 kDa)a

Pipetting device

20–30 000/cm2

Fluorescence

NS

Expression profiling; DNA/RNA–protein interactions; (enzyme assays, immunoasays)a

[24], [25]

Randomly ordered bead arrays Illumina

Etched microwells of fibreoptic imaging bundle

Oligonucleotides; (various molecules compatible with bead immobilization)a

Random distribution of microparticles with analytes (or cells)

250 000 microwells of fibreoptic bundle/1 mm2

Fluorescence; epifluorescence microscope coupled to a CCD detector

NS

SNPs (diagnostics;a chemical monitoringa toxicologya)

[48]

Nano Chip cartridge Nanogen

Silicon wafers, electroactive spots

Oligonucleotides 20-mers; (various charged molecules)a

Electric field-directed immobilization; hictin– streptavidin

25–400/chip

Fluorescence

< 500 labelled molecules

Short tandem repeat identification; SNPs (diagnostics;a cell separation;a cell lysisa)

[43], [54], [55]

Spectro CHIP Sequenom

Single crystal silicon

Oligonucleotidesa

Piezoelectric pipette

250 locations/chip

MALDI-TOF mass spectrometry

0·2 fmol

SNPs; STRs; sequencing

[49], [50], [51]

Applicable to proteins and other molecules. CCD, charge-coupled dHevice; MALDI–TOF, matrix-assisted laser desorption ionization – time-of-flight; NS, not specified; SNP, single nucleotide polymorphism.

Microarrays in blood testing 3

a

© 2001 Blackwell Science Ltd. Vox Sanguinis (2001) 80, 1–11

Chip/company

Matrix

Probe

Deposition

Array size

Detection

Sensitivity

Applications

Reference

German Human Genome Project (RZPD)

PVDF membrane

Bacterial lysate; Ni-NTA purified proteins

Mechanical spotting or inkjet printer

4800 (25 × 75 mm); 27 648 (222 × 222 mm)

Fluorescence; CCD camera

250 amol (10 pg)

Gene expression; antibody screening; receptor–ligand interactions

[21]

Genometrix

Glass wells in planar microarray plates

Purified proteins (IgG in the role of antigens in this case)

36 capillary print head

96 × 144 in 96-well plate format

Fluorescence scanning; CCD imager

13·4 ng/ml (340 pg/well in 25 µl)

High-throughput multiplex ELISAs

[27]

University of Bologna

Cellulose filters

Horseradish peroxidase

Modified inkjet printer

900 (200-µm spots); 36 (1000-µm spots)

Chemiluminescence; CCD

0·7–2·5 pg

NA

[22]

LabMAP System Luminex

Polystyrene microspheres

Purified proteins (various molecules compatible with bead immobilization)a

Covalent attachment

100 different analytes in one tube

FACScan (two laser, one file system)

< 10 pg

Immunoassays (cytokines, allergy testing, DNA tissue typing)a

[47]

ProteinChip Ciphergen

Various treated surfaces

Lysates or purified proteinsa

Affinity capture on chemically or biochemically (e.g. antibody) treated surfaces

Eight per strip; 96-well format

SELDI mass spectrometry

1 fmol

Protein profiles; diagnostics; identification of new proteins

[52], [53]

a

Applicable to DNA and other molecules. CCD, charge-coupled device; ELISA, enzyme-linked immunosorbent assay; FACScan; fluorescence-activated cell sorter; IgG, immunoglobulin G; Ni-NTA, nickel-nitriotriacetic acid; PVDF, polyvinyl dine fluoride; surface-enhanced laser desorption/ionization

VOX010.fm Page 4 Wednesday, February 28, 2001 4:18 PM

4 J. Petrik

Table 2 Various formats of protein chip

VOX010.fm Page 5 Wednesday, February 28, 2001 4:18 PM

Microarrays in blood testing 5

Ge described an array for protein–protein, protein–RNA, protein–DNA and protein–ligand interactions, prepared by spotting native purified proteins onto a nitrocellulose membrane [23]. Similarly flexible is microarray composed of the polyacrylamide gel pads separated by hydrophobic areas on a glass surface. The gel can be loaded with oligonucleotides or DNA, but also with proteins of up to 400 kDa (MW) [24,25]. Enzyme-linked immunosorbent assays (ELISAs) have been extremely successful screening and diagnostic tools in the previous two to three decades and it has been shown that they can be modified to miniaturized multiplex assays [26]. In a recent publication, Mendoza et al. (Genometrix, The Woodlands, TX) described a new-generation high-throughput ELISA biochip on an optically flat glass plate with 96 wells, each containing four identical 36-element arrays [27]. This chip can be successfully used for a multiplex ELISA assay, although any cross-reactivity has yet to be carefully evaluated.

Mandatory testing of blood donations Blood and blood components for transfusion must be microbiologically safe and correctly typed. This is reflected in the list of required tests. The level of current microbiological safety of blood supply would be difficult to contemplate before 1970, when the only testing available was that for syphilis antibody [28]. Most of the current microbiology screening assays are serological assays for the presence of antibodies (HCV; human immunodeficiency virus [HIV]-1 and -2; syphilis; and in some countries also human T-cell lymphotrophic virus [HTLV]-I and -II and hepatitis B core) or antigen (hepatitis B surface antigen [HBsAg], and in some countries HIV p24). PCR testing for the presence of HCV RNA on plasma pools was introduced in the USA and Europe in 1999. This is expected to be extended for the release of cellular components already required in Germany. Other candidates for NAT testing include HIV and, at least for a subset of donations/products, parvovirus B19. ABO and RhD blood grouping and screening for red-cell antibodies are required for all donations. In addition, screening is carried out for irregular red-cell antibodies. Additional tests are required, e.g. to provide cytomegalovirus (CMV)negative components.

Design features of microarray for blood testing The goals of microarray applications in diagnostics blood testing are somewhat different from large-scale expression profiling or SNP studies. Instead of arrays containing the maximum possible number of genes or their products, donation testing focuses on selected, known targets. Consequently, the chip density is several orders of magnitude lower. On the © 2001 Blackwell Science Ltd. Vox Sanguinis (2001) 80, 1– 11

other hand, the sensitivity and specificity of a blood-testing chip is of utmost importance, in line with actively pursued zero risk policy in blood transfusion. This may require a more sophisticated chip containing additional features. When contemplating a designer blood-testing chip, there are a number of key design aspects that have to be considered: (1) What is the preferred layout: a chip with one or a few targets for a large number of donations/samples versus one or two chips covering all targets for a donation/sample? (2) Will the chip consist exclusively of arrayed specific probes or will it include (or otherwise be linked to) the nucleic acid amplification device/compartment? (3) Is it possible to replace all current protein-based assays by nucleic acid-based assays or are we going to have two or more different chips? Alternatively, is it feasible to construct a combined protein–nucleic acid chip? (4) What is the optimal number of targets and what are the criteria for their selection? (5) What type of sample preparation and/or enrichment will be required? (6) What are the sensitivity requirements as compared to current assays?

Preferred layout of the chip Constructing arrays that target one or a few assays would in fact just extend the present testing approaches, although it could substantially increase the throughput and minimize reagent consumption. It would improve, but not necessarily change, the way we are conducting blood testing at present. The alternative method, represented by one or two chips covering all required assays per donation or blood sample, may result in substantial change in current operational practices (reduction in the number of separate assays performed and equipment required; reduction and, in some cases, elimination of repeat, extended or confirmatory assays; reduction in the time required to complete all assays, unified data acquisition and handling, etc.).

Composition of the chip Current nucleic acid-based tests depend on an amplification step almost always provided by the PCR. The sensitivity of current detection methods (radioactive, chemiluminescent, fluorescent) is not sufficient for the direct detection of viral genome in plasma/serum or to determine blood group and platelet-specific antigens without prior amplification of the cellular genomic DNA. It is possible that some viral targets could be enriched by using a special sample preparation step (Fig. 1) in order to omit the amplification step. Similarly, nucleic acid-based blood-group typing and platelet antigen determination could theoretically be performed on cellular genomic DNA rather than on the amplified (e.g. PCR) product. However, the complexity of the whole genome analysis is ≈ 30-fold compared with the expression profiling, which deals

VOX010.fm Page 6 Wednesday, February 28, 2001 4:18 PM

6 J. Petrik

Fig. 1 Design options for a blood-testing chip.

with transcribed sequences only. In a recent publication, Winzeler et al. [29] compared the genomic DNA of two yeast strains without prior amplification. However, yeast genome is haploid and 250 times smaller than the haploid human genome. In order to be able to use unamplified genomic DNA, the sensitivity of detection would have to be increased by two or more orders of magnitude, especially when considering present problems in distinguishing some heterozygous alleles in diploid genomes, even after including an amplification step. One option is to build a combined chip consisting of an amplification compartment and an array-detection compartment. On-chip PCR has been studied and shown to be feasible [30–32]. In such a case the question of efficient and balanced multiplexing will be extremely important. A considerable number of sequences can be co-amplified by carefully designed primers. Adding general sequence at their 5′ ends proved beneficial and allowed the majority of 46 loci amplified simultaneously to be scored correctly [33].

Construction of the chip There is little doubt that the introduction of NAT to detect blood-borne viruses can shorten the window period and improve blood safety [34]. It would seem logical to start

thinking about replacing, for example, serological assay for HCV (which has a prolonged window period) with singledonation reverse transcription (RT)–PCR. However, a proportion (10 – 20%) of HCV antibody-positive individuals are RNA negative in serum. They have presumably recovered from infection but this has not been proven comprehensively. In addition, the HCV RNA titres may fluctuate considerably (several orders of magnitude within a couple of weeks), occasionally falling below the detection point. It is unrealistic at this time to replace serological assay by PCR. Routine blood typing is still performed serologically, with genotyping playing a supporting role, especially in situations where it is difficult or dangerous to obtain sufficient quantities of cells. The genetic basis of most of the blood group systems is known (see below) and molecular methods are easily applicable where alleles differ by a point mutation. However, in some cases identical phenotypic manifestation may be caused by genetically diverse mechanisms in different populations: while the D-negative phenotype in most Caucasians is a result of complete RHD gene deletion, Dnegative individuals in a minority of Caucasians, Blacks and Japanese have intact genes and the negative phenotype is a result of a different molecular mechanism [35]. Therefore, the results of a gene-amplification assay could be misleading.

© 2001 Blackwell Science Ltd. Vox Sanguinis (2001) 80, 1–11

VOX010.fm Page 7 Wednesday, February 28, 2001 4:18 PM

Microarrays in blood testing 7

Ethnic differences are also important for phenotypic manifestations of the blood group O. The amplification of a region in a particular gene does not necessarily mean that the gene is complete or that the protein is expressed. However, even if the protein is expressed, minor modification in the protein or even in another interacting protein can affect assembly of the antigen (Kell, RhD, RhCE) and/or the copy number on a cell surface [36,37]. Yet another problem is posed by protein-only agents such as new variant Creutzfeldt-Jakob disease (nvCJD). A protein chip able to distinguish normal and diseased prion forms would be extremely valuable. Considering the examples mentioned above, it does not seem realistic to contemplate the use exclusively of nucleic acid-based chips for blood diagnostic testing in the foreseeable future. A protein microarray composed of antibodies, antigens or both will probably be required to complement nucleic acid microarray for full characterization of donations or blood samples. Although reaction conditions for nucleic acid hybridization and antigen–antibody reactions differ, it is possible that a combined protein–nucleic acid microarray could be produced, especially if positions in an array can be individually addressed and reaction conditions controlled. Nanogen’s electronic array (see below) may be an example of a microarray with such potential.

Targets and selection criteria Although human leucocyte antigen (HLA) genotyping is well developed and suited for microarray detection, it will probably be the subject of a separate testing procedure and is therefore not discussed in this review. Consequently, the major candidate targets for inclusion on a designer blood-testing chip are blood-borne pathogens, antigens and antibodies of blood-group systems, platelet-specific antigens and perhaps granulocyte antigens. It should be taken into account that for some variable pathogens, such as HIV or HCV, it may be safer to use more than one probe. Given that there are greater than 240 bloodgroup antigens and additional platelet and granulocyte antigens, initial selection will probably be dictated by the clinical significance and frequency of particular targets. On the other hand, if the incremental cost of additional probes is sufficiently low, it may be more economical to have a complete screening microarray (still representing the same assay) rather than having to perform additional tests on a more specialized microarray. Table 3 lists candidate targets for a blood-testing chip. Apart from current and prospective viral targets, the inclusion of group-specific bacterial probes (usually ribosomal RNA-derived) would also be desirable, as serious and sometimes fatal reactions owing to bacterial contamination of blood components continue to occur [38]. Parasitic infections are often restricted to particular geographical areas but © 2001 Blackwell Science Ltd. Vox Sanguinis (2001) 80, 1– 11

increasing global travel and migration may make their inclusion necessary. Finally, if transfusion-transmission was proved, the screening for prion agents, such as nvCJD, would be of highest priority. The genetic background of the majority of more than 240 antigens in 25 blood-group systems is known [39,40] and gradual transition to nucleic acid-based testing is expected as a result of more precise analysis. There is still a danger of erroneous genotyping with some phenotypic manifestations resulting from different molecular mechanisms, as discussed above. Combination of a DNA-typing chip with a proteinbased chip for blood-group typing is therefore probable. Single amino acid replacement is responsible for allelic differences of the majority of the platelet-specific alloantigens (human platelet antigens [HPA] ). HPA-1–5 genotyping is being performed correctly by an increasing number of laboratories [41] and is generally suitable for microarray analysis. This is true also for the granulocyte antigens NA1 and NA2. The number of probes for blood-group typing, plateletspecific and granulocyte antigens, combined with a set of probes for microbiological screening, would not at present exceed 300–400, even considering redundant probes. This is rather a small number for microarray but could be reduced, if necessary, based on clinical significance and frequency of the target. In addition, only a proportion of targets require both nucleic acid and protein assay to be carried out in parallel so that the approximate number shown will be split, at least initially, between nucleic acid-based and protein-based chips. Such a number is close to the number of positions available at present on even the most specialized types of chips, such as Nanogen’s electronic array or Ciphergen’s protein array (see Tables 1 and 2).

Sample preparation and/or enrichment Sample preparation may prove extremely important for subsequent multitarget analysis, especially as targets will come from both plasma (some pathogens) and cellular components (DNA for genotyping). There are promising developments in on-chip cell separation [42] as well as electronic release of nucleic acids from concentrated bacteria with subsequent hybridization [43]. Development of a fully integrated DNA analysis system capable of cell separation, nucleic acid processing and hybridization, and perhaps including on-chip amplification, is technically feasible and can be achieved in the near future. Alternatively, the reaction kinetics considerably improved by miniaturization, in combination with a sample preparation chip capable of some form of target enrichment, may eventually make the amplification step obsolete (Fig. 1).

Sensitivity requirements With greater than 1 × 106 of A and B sites per red cell, sensitivity is not a problem for ABO blood grouping. Estimated

VOX010.fm Page 8 Wednesday, February 28, 2001 4:18 PM

8 J. Petrik

Table 3 Potential targets for a blood testing chip/chips

Targets Virus HBV HIV-1, -2 HCV HTLV-I, -II HAV Parvovirus B19 HGV (GBV-C)b TTV b CMV HHV-8 Prion nvCJD Bacteria Treponema pallidum (Syphilis) SHOT-identified bacteria Parasites Plasmodium species (Malaria) Babesia microti (Babesiosis) Trypanosoma cruzi (Chagas’ disease) Toxoplasma gondii (Toxoplasmosis) Red cell antigens and antibodies ABO RhD Other 23 systems Irregular red cell antibodies Additional antigens on blood cells Platelet-specific Granulocyte

Transfusion transmitted

Mandatory testing

+ + + + + (rarely) + + + + ? ?

+ + + +a – – – – (+) (subset) – –

+ +

+ –

+ + + +

– – – –

NA NA NA NA

+ + – +

NA NA

– –

a

Not in the UK. No disease association. CMV, cytomegalovirus; GBV-C, GB virus-C; HAV, hepatitis A virus; HBV, hepatitis B virus; HCV, hepatitis C virus; HGV, hepatitis G virus; HHV, human herpes virus; HIV, human immunodeficiency virus; HTLV, human T-cell lymphocytotrophic virus; NA, not applicable; nvCJD, new variant Creutzfeldt-Jakob disease; SHOT, serious hazards of transfusion; TTV, transfusion-transmitted virus.

b

numbers for C, c, D, E and e are lower, between 450 and 85 000 per cell, depending on the antigen and phenotype of the cell [44], with RhD between 10 and 40 000. Agglutination end-point titres of anti-D immunoglobulin M (IgM) antibodies were between 7 and 26 ng/ml [45], which for a 180 kDa protein corresponds to 39–144 fmol/ml. Blood-borne pathogens can be present at a wide concentration range. The detection limit of the most sensitive ELISAbased assays for viral antigens (HBsAg, HCV core) is between 1 and 5 pg/ml, 40 – 200 fmol/ml for a 25 kDa protein. Genomic detection by PCR is several orders of magnitude more sensitive still, detecting dozens of RNA copies/ml.

The sensitivity of confocal scanners is around 0·5 molecules of fluor/µm2 (between 10 and 34·5 molecules of Cy3 or Cy5 dyes/µm2) or 0·15 amol of end-labelled oligonucleotide [46]. The actual sensitivity of various microarray systems is shown in Tables 1 and 2, and is usually of fmol to subfmol in range. This is of comparable or higher sensitivity to the currently used serological blood assays. However, some targets could still be missed if non-amplification systems were used exclusively. It appears that the sensitivity of current nonamplification systems would have to increase by two to three orders of magnitude, either through target enrichment (sample concentration, extraction, capture … ), increase in detection sensitivity, or a combination of both, in order to approach the sensitivity of NAT systems.

Microarray technologies potentially applicable to blood screening The format of microarrays may differ considerably from two-dimensional high-density oligonucleotide microarrays (Affymetrix) or cDNA microarrays (Incyte, Palo Alto, CA; Molecular Dynamics, Sunnyvale, CA) produced on glass or silicon support (see Tables 1 and 2). Flexibility, rather than density, is important for diagnostic microarray. The flexibility can be provided by the ability to use more than one type of molecule as a probe or the potential for combining various steps of the procedure (sample preparation, enrichment/ amplification, specific reaction, detection) on the same chip or on a linked set of chips. The identity of the probe in two-dimensional microarray is given by its position. A different approach must be adopted for microparticle-based microarrays. Several companies use various types of microparticles, with the advantage of extensive knowledge of their applications in separation procedures, affinity sample preparation and diagnostics, as well as known chemistries for probe immobilization. Importantly, the probe can be nucleic acid, protein, carbohydrate or small molecule, depending on the aim of the procedure. Luminex (Austin, TX) developed a homogeneous single tube system where up to 100 analytes can be analysed simultaneously using so-called ‘two-colour suspension array’. Polystyrene microparticles are internally labelled by two spectrally distinct fluorochromes at 10 different quantities each, producing 100 combinations. A third fluorochrome coupled to a reporter molecule is used for reaction detection. Microparticles pass in single file through two separate laser beams, identifying at the same time the particle (and corresponding ligand) and the extent of the reaction, as shown for simultaneous cytokine detection [47]. Another microparticle-based but conceptually different microarray is being developed by Illumina (San Diego, CA) using fibreoptic technology pioneered by Walt and colleagues at Tufts University [48]. Each bundle can contain tens to hundreds of thousands of © 2001 Blackwell Science Ltd. Vox Sanguinis (2001) 80, 1–11

VOX010.fm Page 9 Wednesday, February 28, 2001 4:18 PM

Microarrays in blood testing 9

optic fibres in an area of 1 mm2. Miniature wells able to accommodate just one bead (or cell) are formed at the end of each fibre by hydrofluoric acid etching, resulting in an optical fibre bundle microarray. Particles with various ligands are distributed randomly (using 20-fold redundancy to ensure inclusion of each type of particle) and identified subsequently. For oligonucleotide microarrays, the position identification is via a series of quick hybridization steps. The reading of the reaction with labelled sample occurs at the other end of the bundle or array of bundles. This technology has the potential for extremely high throughput but can be scaled down for diagnostic purposes. Only a few years ago, large-scale diagnostic use of mass spectrometry-based techniques would not have been considered feasible. However, the progress achieved, especially by Sequenom (San Diego, CA), in miniaturization, automatization and throughput of their matrix-assisted laser desorption ionization – time-of-flight (MALDI–TOF)-based technology has changed the situation dramatically. In order to miniaturize the arrays, Sequenom focused on single crystal silicon, reducing the well size to nanoliters and reducing the sample to the same size as the spot illuminated by the laser [49]. The daily throughput of Sequenom’s mass array is ≈ 10 000 samples. Initially focused on DNA sequencing [50], the current main applications include mutation/SNP analyses, although this technology is applicable to medical diagnostics, identity testing, etc. [51]. Potential diagnostic applications will be dependent largely on the speed and complexity of the sample preparation step, which may be more demanding compared with alternative techniques. While Sequenom focuses on DNA analysis, Ciphergen (Palo Alto, CA) uses surface-enhanced laser desorption/ ionization (SELDI) technology [52] for separation, detection and analysis of proteins with a detection limit of 1 fmol of protein and an ability to produce ‘phenotypic fingerprints’ directly from patient samples. Until recently this technology was mainly research orientated but the ability to affinitycapture proteins on ProteinChip array prior to analysis makes the diagnostic applications realistic. SELDI has been used successfully to detect and distinguish amyloid beta variants [53]. One of the advantages of mass spectrometry-based systems is that although they focus currently on DNA or protein, respectively, they are in principle able to deal simultaneously with various types of molecules. In addition, these are the only microarray platforms providing direct verification of the identity of the analysed molecule. Yet another approach is represented by the electronic array mentioned above and developed by scientists at Nanogen (San Diego, CA). Twenty-five to 400 test site arrays represent one of the most versatile systems available. Test sites can be addressed individually, allowing users to concentrate and immobilize probes of their choice and regulate the reaction by applying controlled electric fields: the hybridization can be © 2001 Blackwell Science Ltd. Vox Sanguinis (2001) 80, 1– 11

significantly accelerated both for single nucleotide differences [54] and for short tandem repeats [55]. In addition, multiplexing is not restricted to identical types of molecules. As mentioned earlier, the chip can also be used for cell concentration/separation and electronic release of nucleic acid [42,43]. In order to make the chip even more flexible, on-chip amplification is being developed using strand displacement amplification through collaboration with Becton-Dickinson. This is not a complete list of all options available. There are, of course other approaches and platforms being developed, as expected in such a vibrant field. Those most suitable for particular applications will undoubtedly emerge through the extensive testing and technology adjustment.

Conclusion Rather than continuing the trend of an ever-increasing number of assays resulting in a considerable increase in the cost of blood testing in pursuit of zero-risk policies, microarrays permit a conceptually new approach. They provide an integrated testing platform where even with an increased number of assay targets there is a dramatic decrease in the number of individual assay procedures. At the same time, the multiple sets of equipment required for various individual assays can be replaced by one or two assay platforms. Although the initial costs of microarray systems may be high, the microarray blood-testing technique should become more economical as a result of reduced reagent consumption, reduction (and in some cases even elimination) of repeat, extended or confirmatory assays, reduction in the time required to complete the assays and, from simpler ways of results processing, data acquisition and handling. Current subsets of stocks resulting from testing for an additional parameter (e.g. CMV) could be eliminated. Considering the results of serious hazards of transfusion (SHOT) reports [56], there is at present disproportionate attention paid to various aspects of blood testing, such as the viral safety of blood supply as compared to blood-group incompatibilities or bacterial contamination. Microarrays would provide a more balanced approach to blood testing as probes for various types of targets can be included in the same assay. The inclusion of probes required should not be a limiting factor, as the incremental cost of additional probes in an array is small. We should, however, try to avoid testing just for the sake of testing because there is a tool to do it. Attempts to provide, for example, super-clean blood, devoid of all known agents, could backfire, especially when some of the agents, such as transfusion-transmitted (TT) virus are unlikely to be harmful in view of their ubiquity [57]. There are numerous outstanding problems to be solved. Issues such as balanced multiplex amplification of various target sequences, stability of complex membrane antigens

VOX010.fm Page 10 Wednesday, February 28, 2001 4:18 PM

10 J. Petrik

on a chip or an increase in the detection sensitivity, require further detailed studies. However, the application of microarray technologies in blood testing may be rewarded by making blood supply even safer, more efficient, and providing, at the same time, the potential to simplify and accelerate the operational aspects of blood banking.

Acknowledgements I thank Drs G. J. M. Alexander, D. B. L. McClleland and C. W. Prowse for their support during the preparation of this manuscript.

References 1 Saiki RK, Scharf S, Faloona FA, Mullis KB, Horn GT, Erlich HA, Arnheim N: Enzymatic amplification of β-globin genomic sequences and restriction site analysis for diagnosis of sickle cell anaemia. Science 1985; 230:1350 –1354 2 Southern EM: Detection of specific sequences among DNA fragments separated by gel electrophoresis. J Mol Biol 1975; 98:503 – 517 3 Ekins R, Chu FW: Microarrays: their origins and applications. TIBTECH 1999; 17:217–218 4 Schena M, Shalon D, Davis RW, Brown PO: Quantitative monitoring of gene expression patterns with a complementary DNA microarray. Science 1995; 270:467– 470 5 Shalon D, Smith SJ, Brown PO: A DNA micro-array system for analyzing complex DNA samples using two-colour fluorescent probe hybridization. Genome Res 1996; 6:639 – 645 6 Cheung VG, Morley M, Aguilar F, Massimi A, Kucherlapati R, Childs G: Making and reading microarrays. Nat Genet 1999; 21:15 –19 7 Okamoto T, Suzuki T, Yamamoto N: Microarray fabrication with covalent attachment of DNA using bubble jet technology. Nat Biotechnol 2000; 18:438 – 441 8 Harris TM, Massimi A, Childs G: Injecting new ideas into microarray printing. Nature Biotechnol 2000; 18:384 –385 9 Fodor SPA, Read JL, Pirrung MC, Stryer L, Lu AT, Solas D: Light-directed, spatially addressable parallel chemical synthesis. Science 1991; 251:767–773 10 Singh-Gasson S, Green RD, Yue Y, Nelson C, Blattner F, Sussman MR, Cerrina F: Maskless fabrication of light-directed oligonucleotide microarrays using a digital micromirror array. Nat Biotechnol 1999; 17:974 – 978 11 Winzeler EA, Schena M, Davis RW: Fluorescence-based expression monitoring using microarrays. Methods Enzymol 1999; 306:3 –19 12 Ramsay G: DNA chips: state-of-the art. Nat Biotechnol 1998; 16:40 – 44 13 Chambers J, Angulo A, Amaratunga D, Guo H, Jiang Y, Wan JS, Bittner A, Frueh K, Jackson MR, Peterson PA, Erlander MG, Ghazal P: DNA microarrays of the complex human cytomegalovirus genome: profiling kinetic class with drug sensitivity of viral gene expression. J Virol 1999; 73:5757– 5766 14 Zhang L, Zhou W, Velculescu VE, Kern SE, Hruban RH, Hamilton SR, Vogelstein B, Kinzler KW: Gene expression profiles in normal and cancer cells. Science 1997; 276:1268 –1272

15 Whitney M, Rockenstein E, Cantin G, Knapp T, Zlokarnik G, Sanders P, Durick K, Craig FF, Negulescu PA: A genome-wide functional assay of signal transduction in living mammalian cells. Nat Biotechnol 1996; 16:1329 –1333 16 Zong Q, Schummer M, Hood L, Morris DR: Messenger RNA translation state: the second dimension of high-throughput expression screening. Proc Natl Acad Sci USA 1999; 96:10632–10636 17 Brown PO, Botstein D: Exploring the new world of the genome with DNA microarrays. Nat Genet 1999; 21:33–37 18 Linder MW, Prough RA, Valdes R Jr: Pharmacogenetics: a laboratory tool for optimizing therapeutic efficiency. Clin Chem 1997; 43:254 –266 19 Hacia J: Resequencing and mutational analysis using oligonucleotide microarrays. Nat Genet 1999; 21:42–47 20 Halzimanikatiz V, Lee KH: Dynamical analysis of gene networks requires both mRNA and protein expression information. Metab Eng 1999; 1:275–281 21 Lueking A, Horn M, Eickhoff H, Bussow K, Lehrach H, Walter G: Protein microarrays for gene expression and antibody screening. Anal Biochem 1999; 270:103 –111 22 Roda A, Guardigli M, Russo C, Pasini P, Baraldini M: Protein microdeposition using a conventional ink-jet printer. Biotechniques 2000; 28:492– 496 23 Ge H: UPA, a universal protein array system for quantitative detection of protein–protein, protein–DNA, protein–RNA and protein–ligand interactions. Nucl Acids Res 2000; 28:e3 24 Guschin D, Ershov G, Zaslavsky A, Gemmell A, Shick V, Proudnikov D, Arenkov P, Mirzabekov A: Manual manufacturing of oligonucleotide, DNA and protein microchips. Anal Biochem 1997; 250:203–211 25 Arenkov P, Kukhtin A, Gemmell A, Voloshchuk S, Chupeeva V, Mirzabekov A: Protein microchips: use for immunoassay and enzymatic reactions Anal Biochem 2000; 278:123–131 26 Ekins R, Chu FW: Multianalyte microspot immunoassay – microanalytical ‘compact disk’ of the future. Clin Chem 1991; 37:1955–1967 27 Mendoza LG, McQuary P, Mongan A, Gangadharan R, Brignac S, Eggers M: High-throughput microarray-based enzyme-linked immunosorbent assay (ELISA). Biotechniques 1999; 27:778–788 28 Barbara JAJ: Microbiological safety of blood transfusion. Vox Sang 1998; 74:11–13 29 Winzeler E, Richards DR, Conway AR, Goldstein AL, Kalman S, McCullogh MJ, McCusker JH, Stevens DA, Wodicka L, Lockhart DJ, Davis RW: Direct allelic variation scanning of the yeast genome. Science 1998; 281:1194–1197 30 Schoffner MA, Cheng J, Hvichia GE, Kricka LJ, Wilding P: Chip PCR: I. Surface passivation of microfabricated slicon-glass chips for PCR. Nucl Acids Res 1996; 24:375–379 31 Cheng J, Schoffner MA, Hvichia GE, Kricka LJ, Wilding P: Chip PCR: II. Investigation of different PCR amplification systems in microfabricated silicon-glass chips. Nucl Acids Res 1996; 24:380 – 385 32 Kopp MU, deMello AJ, Manz A: Chemical amplification: continuousflow PCR on a chip. Science 1998; 280:1046–1048 33 Wang DG, Fan J-B, Siao C-J, Berno A, Young P, Sapolsky R, Ghandour G, Perkins N, Winchester E, Spencer J: Large-scale identification, mapping and genotyping of single-nucleotide polymorphisms in the human genome. Science 1998; 280:1077–1082

© 2001 Blackwell Science Ltd. Vox Sanguinis (2001) 80, 1–11

VOX010.fm Page 11 Wednesday, February 28, 2001 4:18 PM

Microarrays in blood testing 11

34 Busch MP, Stramer SL, Kleinman SH: Evolving applications of nucleic acid amplification assays for prevention of virus transmission by blood components and derivatives; in Garraty G (ed.): Applications of Molecular Biology to Blood Transfusion Medicine. Bethesda MD, American Association of Blood Banks, 1997:123–176 35 Avent ND: Molecular genetic methods: principles and feasibility in transfusion medicine. Vox Sang 1998; 74:275–284 36 Cartron JP, Bailly P, Le Van Kim C, Cherif-Zahar B, Matassi G, Bertrand O, Colin Y: Insights into the structure and function of membrane polypeptides carrying blood group antigens. Vox Sang 1998; 74:29 – 64 37 Reid ME, Rios M: Applications of molecular genotyping to immunohaematology. Br J Biomed Sci 1999; 56:145–152 38 Blajchman MA: Bacterial contamination and proliferation during the storage of cellular blood products. Vox Sang 1998; 74:155–159 39 Anstee DJ, Cartron J-P: Toward an understanding of the red cell surface; in Garratty G (ed.): Applications of Molecular Biology to Blood Transfusion Medicine. Bethesda MD, American Association of Blood Banks, 1997:17–49 40 Daniels GL, Anstee DJ, Cartron JP, Dahr W, Garratty G, Henry S, Jorgensen J: Terminology for red cell surface antigens. ISBT working party Oslo report. Vox Sang 1999; 77:52–57 41 Metcalfe P, Cavanagh G, Hurd C, Ouwehand WH: HPA genotyping by PCR-SSP: report of four exercises. Vox Sang 1999; 77:40 –43 42 Cheng J, Sheldon EL, Wu L, Heller MJ, O’Connell JP: Isolation of cultured cervical carcinoma cells mixed with peripheral blood cells on a bioelectronic chip. Anal Chem 1998; 70:2321– 2326 43 Cheng J, Sheldon EL, Wu L, Uribe A, Gerrue LO, Carrino J, Heller MJ, O’Connell JP: Preparation and hybridization analysis of DNA/ RNA from E. coli on microfabricated bioelectronic chips. Nat Biotechnol 1998; 16:541–546 44 Hughes-Jones NC: Quantitation and the Rh blood group system. Transfus Med 1991; 1:69–76 45 Hughes-Jones NC, Gorick BD, Beale D: Multiple epitopes recognised by human monoclonal IgM anti-D antibodies. Vox Sang 1990; 59:112–115 46 Bowtell DDL: Options available – from start to finish – for

© 2001 Blackwell Science Ltd. Vox Sanguinis (2001) 80, 1– 11

47

48

49

50

51

52

53

54

55

56

57

obtaining expression data by microarray. Nat Genet 1999; 21:25–32 Carson RT, Vignali DAA: Simultaneous quantitation of 15 cytokines using a multiplexed flow cytometric assay. J Immunol Methods 1999; 227:41–52 Michael KL, Taylor LC, Schultz SL, Walt DR: Randomly ordered addressable high-density optical sensor arrays. Anal Chem 1998; 70:1241–1248 Little DP, Cornish TJ, O’Donnell MJ, Braun A, Cotter RJ, Koster H: MALDI on a chip: analysis of low-femtomole to subfemtomole quantities of synthetic oligonucleotides and DNA diagnostic products dispensed by a piezoelectric pipet. Anal Chem 1997; 69:4540 – 4546 Koster H, Tang K, Fu D-J, Braun A, van den Boom D, Smith CL, Cotter RJ, Cantor CR: A strategy for rapid and efficient DNA sequencing by mass spectrometry. Nat Biotechnol 1996; 14:1123– 1128 Little DP, Braun A, O’Donnell MJ, Koster H: Mass spectrometry from miniaturized arrays for full comparative DNA analysis. Nat Med 1997; 3:1413–1416 Hutchens TW, Yip T-T: New desorption strategies for the mass spectrometric analysis of macromolecules. Rapid Commun Mass Spectrom 1993; 7:576–580 Davies H, Lomas L, Austen B: Profiling of amyloid beta peptide variants using SELDI ProteinChip arrays. Biotechniques 1999; 27:1258–1261 Gilles PN, Wu DJ, Foster CB, Dillon PJ, Chanock SJ: Single nucleotide polymorphic discrimination by an electronic dot blot assay on semiconductor microchips. Nat Biotechnol 1999; 17:365–370 Radtkey R, Feng L, Muralhidar M, Duhon M, Canter D, DiPierro D, Fallon S, Tu E, McElfresh K, Nerenberg M, Sosnowski R: Rapid, high fidelity analysis of simple sequence repeats on an electronically active DNA microchip. Nucl Acids Res 2000; 28:e17 Williamson LM, Lowe S, Love EM, Cohen H, Soldan K, McClleland DB, Skacel P, Barbara JA: Serious hazards of transfusion (SHOT) initiative: analysis of the first two annual reports. BMJ 1999; 319:16–19 Simmonds P, Prescott LE, Logue C, Davidson F, Thomas AE, Ludlam CA: TT virus – part of the normal human flora? J Infect Dis 1999; 180:1748–1750