| | Semiconductor quantum dots for multiplexed bio-detection on solid-state microarraysAccepted 17 April 2009. published online 21 May 2009. Abstract Understanding cellular systems requires identification and analysis of their multiple components and determination of how they act together and are regulated. Microarray technology is one of the few tools that is able to solve such problems. It is based on high-throughput recognition of a target to the probe and has the potential to simultaneously measure the presence of numerous molecules in multiplexed tests, all contained in a small drop of test fluid. Microarrays allow the parallel analysis of genomic or proteomic content in healthy versus disease-affected or altered tissues or cells. The signal read-out from the microarrays is done with organic dyes which often suffer of photobleaching, low brightness and background fluorescence. Recent data show that the use of fluorescent nanocrystals named “quantum dots” (QDs) allows to push these limits away. QDs are sufficiently bright to be detected as individual particles, extremely resistant against photobleaching and provide unique possibilities for multiplexing, thus supplying the microarray technology with a novel read-out option enabling the sensitivity of detection to reach the single-molecule level. This paper reviews QDs applications to microarray-based detection and demonstrates how the combination of microarray and QDs technologies may increase sensitivity and highly parallel capacities of multiplexed microarrays. Such a combination should provide the breakthrough results in drug discovery, cancer diagnosis and establish new therapeutic approaches through the identification of binding target molecules and better understanding of cell signalling pathways. 1. Introduction  Applications of DNA or protein microarrays are high-throughput technologies which allow the study on the genomic or proteomic contents, respectively [1], [2]. Although these two approaches provide information concerning completely different molecules, they are very supplementary in interpretation and both relate to high-content techniques which allow the simultaneous analysis of multiple genes, proteins or low-molecular-weight molecules in one shot. This capability enables very rapid comparative analysis of several physiological states, i.e. healthy, altered or diseased [3]. In general, the main principle of all microarray technologies is based on detection of the specific binding of a target to the probe. Here, DNA binding in DNA microarray is due to homology of the single strand nucleic acid sequences between the probe and the target. The detection of such binding event is normally performed by the reading of the signal from the label incorporated in the target [4]. For protein microarrays the target is a protein which binds to the probe. The detection of the binding is done here with target-specific antibodies coupled with the label molecule [5]. Usually the label molecules used in DNA and protein microarrays to detect binding events are fluorescent organic dyes or fluorophores. Although fluorophores provide a quite sensitive, safe and cheap detection system, enabling multiplexed stains essential for the high-throughput capacity of microarrays, they suffer from several limitations. Organic dyes are sensitive to photobleaching and they are often not bright enough to permit quantitative specific signal detection over the background fluorescence [6]. Moreover, their fluorescence spectra are not symmetric and each fluorophore is characterized by its specific optimal wavelength of excitation, which limits their multiplexing capabilities. All these disadvantages can be avoided by the use of a novel class of fluorophores, highly fluorescent semiconductor nanocrystals (NCs) quantum dots (QDs). QDs are ultra-resistant to photobleaching, extremely bright and their fluorescence colour can be tuned by slight variation of their diameter on the nanoscale [6]. Furthermore, QDs of different diameters (colours) can be excited with the same excitation wavelength and demonstrate ideally symmetrical fluorescent spectra, thus providing unique possibilities for multiplexing without overlapping of signals from different labels [7]. The integration of QDs in microarray technologies should certainly increase sensitivity and specificity of analysis, reduce sample quantity and even presumes the possibility of single-molecule detection. This last point is a critical one when studies of rare events are presumed or when low quantitative differences between the analysed probes which need to be detected [8]. Finally, the high brightness and photostability of QDs may solve one of the crucial problems of microarray technologies, i.e. quantitative detection of low specific fluorophore signal over background fluorescence. We will start this review from the short description of the general principles of DNA and protein microarray technologies, followed by an overview of the existing labelling procedures and labels in use in existing microarray read-out systems. We will further analyse the advantages and limits of the optical and physico-chemical properties of QDs when compared with organic dyes, as well as the technical requirements which would enable realisation of QDs advantages in the microarray read-out systems. And finally, the most promising applications of QDs as labels in multiplexed microarrays and the overview of published QDs applications to microarray technologies will be discussed. This review is focused on the solid-phase microarray technologies. Readers interested in the QDs applications to the liquid-phase arrays based on optically encoded microbeads are addressed to the recently published reviews [9], [10], [11]. 2. Multiplexed microarray high-throughput technologies Understanding complex cellular systems requires first the identification and analysis of its components and then to determine how they interact together and are regulated. Traditionally, scientists study one gene or one molecule at once. In contrast to this traditional approach, some high-throughput methods have been developed in the last decade to optimize the study of large number of molecules: DNA and protein microarrays. The development of microarray technologies has been phenomenal in the past few years. It has become a standard tool in many research laboratories. The reason for this popularity is that microarrays have revolutionized the approach to biological research thanks to their multiplexed staining capacities. For example, instead of working on a gene-by-gene basis, scientists can now study tens of thousands of genes at once with DNA microarrays. Applications of microarrays in detection and diagnostics are quantitative, and have minimal reagent and sample consumption due to miniaturization. They can be done on solid-state chips as support. The parallel processing power of microarrays depends essentially on the number of different probes spotted on the array. More probes per array means more information. However, between DNA and protein microarray, the number of probes differs significantly. Indeed, as examples, Affimetrix has developed a DNA microarray support for 400,000 oligonucleotide probes on a 1.6 cm2 glass surface [12], ready for hybridization and nucleic acid content analysis, and more recently, some DNA microarrays were made with 2.1 million micro-scale probes DNA microarray (NimbleGen, Roche). For protein microarrays, the number of immobilized proteins, peptides or antigens on the chips is not so large. As an example, the MagyArray® microarray possesses 624 immobilized antibodies (208 different antibodies in triplicate). Despite this difference, these two kinds of microarrays are useful and provide different but complementary informations. DNA microarrays are a powerful approach for the analysis of global transcriptional response. They are a suitable approach to study changes in gene expression and regulation [13], [14], and responses to changing environment conditions [15]. Effects of diseases [16], drug treatment [17], virulence and evolution of pathogens [18] or chemotherapy medications [4] can be detected by these tools allowing insights into the dynamics of the genome. Protein microarrays provide post-translational information and protein activity levels through phosphorylation states. Functional protein microarrays are used in the field of drug target identification/validation [19], studies of protein interactions [20], [21] and signalling pathways [22], [23], biochemical activity and immune responses [24], [25]. Protein microarrays also promise diagnosis capacities [5], [26], [27], [28]. The evolution of gene regulation during environmental changes can be studied by both DNA and protein microarrays. The two approaches are more or less complementary and can lead to a full understanding of regulation steps and events. DNA microarrays provide direct information on the gene regulation by methylation level but also on the transcription regulation level (positive or negative) through the transcription factor, DNA cluttering, miRNA (micro RNA) or other transcription regulation events. Protein microarrays allow studying both the level of protein expression and protein activity through their post-translational modifications (phosphorylation, lipidation, glycosylation). This leads to the establishment and best understanding of diseases deregulation signalling pathways and thus facilitates the creation of new therapy approaches based on different effectors and targets. 2.1. DNA microarrays The DNA microarray technology is the parallel hybridization of a mixture of labelled nucleic acids usually called targets, with thousands of individual nucleic acid species called probes (or features), that can be identified by their spatial position in a single experiment. Whereas the probes are immobilized on the chips, the targets are labelled and then incubated in solution onto the chips for hybridization [29] (Fig. 1A). Historically, the targets were radioactively labelled [30], thereby allowing only one-channel experiment (usually called one-colour experiment now). One-channel experiments suffer from experimental variations between each single DNA microarray. This problem was solved by using fluorescent labels. Indeed, they allow two-channel experiments (or two-colour experiments) in which two mRNA populations are labelled with different fluorescent dyes and are simultaneously assayed on the same chip. DNA microarrays can be made using cDNA (complementary DNA) [31], [32], miRNA [33] or genomic DNA [34]. An amplification step can be performed to minimize sample consumption. During this step, fluorescent nucleotides can be randomly incorporated into the resulting amplified target probe. This procedure is called body-labelling. Alternatively, the end-labelling allows the addition of a fluorescent label only at the end of the target probe. Of the two strategies, the body-labelling method is more commonly used because it provides a much higher fluorescence signal for the same target sample. For body-labelling, there are two main strategies for label incorporation in cDNA by reverse transcription of RNA. The direct labelling incorporates label-modified nucleotide triphosphates like dUTP or dCTP. The nucleotide triphosphate selected to support the label is provided at a lower concentration to counterbalance the amount of label-modified nucleotide triphosphates. This allows the insertion of dye into the resulting cDNA. If a two-colour experiment is performed, the labelling step of cDNA of the two different conditions with the dyes is done separately. Then the free label-modified nucleotide triphosphates are removed and the assay can be done. This method has a critical limit: the number of dyes incorporated is different between dye/probe due to the difference in size of each label molecules. In practice, the smaller label is preferentially incorporated by the reverse polymerase. This experimental bias must be corrected by a normalization step to obtain relevant biological data. However, this normalization step needs some calibration studies for each new label used for two-colour DNA microarray experiments. The indirect labelling is a good alternative method to avoid the effects of size on the incorporation rate. In this method, the nucleotide triphosphate incorporated is modified by target molecules such as biotin, aminoallyl or digoxigenin. In this case, the two preparations are labelled separately but using the same molecule thereby avoiding size biases. Furthermore, these molecules are often smaller than labels thus the incorporation rate of this modified nucleotide triphosphate is higher. Then, the fluorescent labels coupled to the correspondent ligand molecule are separately assessed on the probes for a fluorescent labelling of nucleotide triphosphates. Genomic DNA is usually labelled by direct incorporation of fluorescent labelled nucleotides by nick translation or by random priming with the Klenow fragment of DNA polymerase. Genomic DNA is used for comparative genomic studies as a reference target in normalization or for slide quality control. 2.2. Protein microarrays 2.2.1. Types of protein microarrays Two types of protein microarray analyses are currently available: analytical and functional microarrays. Analytical microarrays, also called quantitative microarrays are used to profile a complex mixture of proteins in order to measure the protein expression level of a specific protein in the mixture. Typically antibodies are coated onto the chip followed by incubation with the protein sample. The captured molecules are then detected using labelled antibodies specific of the same protein. This type of assay is particularly suitable to compare molecular profiles (e.g. protein content and protein phosphorylation state) between healthy and diseased or stressed cells and tissues [35]. For sensitive detection of protein alterations, the so-called reverse phase microarray can be performed. In this type of analytical microarray, a lysate protein sample is arrayed on the solid-state chip and therefore antibodies are used to label the protein of interest on the chip [36]. Functional microarrays differ from analytical microarrays in that they are composed of full-length functional proteins or protein domains. These microarrays are used to study the biochemical activities of an entire proteome in a single experiment and protein interactions with proteins, DNA, RNA, small molecules or phospholipids [37], [38]. For protein microarray assays, the target can be labelled according to one of these two approaches: the direct labelling of the target or the use of a specific antibody. The second approach is most often chosen because it does not require a modification of the target (Fig. 1B) which makes possible to detect proteins directly from patient samples. Another possibility is to indirectly label the sample with a small molecule such as biotin, thus enabling the later recognition of this protein through streptavidin. 2.2.2. Solid supports and surface chemistry for arraying proteins One of the key steps of the development of protein arrays is the choice of a solid support considering three key requirements: an optimal binding condition with a uniform coating, high-throughput manufacturing and non-denaturing environment [39]. This choice is the first step in the development of protein arrays [40], [41]. A number of different slide surfaces can be used for protein chips. These supports must satisfy such strict requirements as immobilizing the protein on the chip, maintaining the conformation and the functionality of this protein, achieving maximum binding capacity, providing good quality spots, low background, simplicity of manipulation and compatibility with detection systems. The direct and random immobilization of proteins on solid phase limits their number, often causes their denaturation and decreases of interaction specificity. It is also important to consider that either a random or a uniform orientation of proteins is desired. The oriented immobilization of target proteins enables a considerably greater quantity, and more functional capture proteins deposited [42]. Glass slides are currently the most common support for biochips because of their easy handling, low cost, low variability, low fluorescent background and greater durability. Glass is an inert and mechanically stable support, which requires a chemical coating to make it functional. The chemistry used to immobilize target proteins on glass slides is based on either non-covalent binding of proteins or covalent bonding between amino acids or carbohydrates attached to the proteins and functional groups of chemical agents previously fixed on the slide. This modified glass surfaces can be aldehyde, epoxy, poly-l-lysine, silane or nickel coated slides [43]. Another technique that can increase the efficiency of protein immobilization consists in covering the glass slide with a layer of a polymer, such as polyacrylamide [44], agarose or gelatine which provides a porous structure. A film of gel, which is 70–95% water, ensures that the three-dimensional structure of the immobilized proteins is maintained and their accessibility is favoured without them having to be oriented. Moreover, the low background of fluorescence of this type of support leads to a greater sensitivity of detection. The nitrocellulose membrane is also proved to be an appropriate support for the immobilization of different proteins and to be easy to manipulate [45]. Attachment of nitrocellulose membranes to glass slides combines the advantages of microarrays with the protein binding capacity and the long-term stability of printed proteins. Although hydrophobic interactions are generally considered to be responsible for the immobilization of proteins, the actual forces involved in their binding are not yet known. Due to its microporous surface, the nitrocellulose is able to retain a greater number of capture proteins than the planar surfaces such as a glass slide. Moreover, its porosity contains, under appropriate conditions, aqueous microspaces that allow proteins to maintain their active configuration [46]. In addition, the nitrocellulose membranes FAST-type slides, for example, have a higher signal-to-noise ratio which notably improves the detection by fluorescence. Proteins are usually purified before being deposited. However, porous supports, like nitrocellulose or hydrogel, are equally compatible with crude cell extracts containing non-purified over expressed proteins [42]. An alternative method is the use of plastic slides or microplates. Maxisorp (Nunc) slides are primarily intended for the immobilization of biomolecules by electrostatic forces and ionic bonds. A chemically modified surface of conventional microplate wells can also be used to fabricate protein microarrays [37]. The benefit of microwells is the ability to perform experiments in an aqueous environment while preventing cross contamination. 2.2.3. Capture molecules Availability of a large number of specific capture molecules showing a high affinity to their targets is a prerequisite for the identification and quantification of proteins using an array-based approaches. Microarrays are generated by immobilizing specific capture molecules on a slide, allowing the analysis of complex samples. Capture molecules can be proteins to analyse specific protein–protein interactions. Affinity tag surfaces can be used for the uniform orientation of proteins on the chip surface. One popular slide choice is the nickel coated slide for use with HisX6 tagged proteins [47]. It is also possible to immobilize probes via C-terminal affinity tags such as C-myc upon pre-arrayed monoclonal anti-tag antibodies [42]. Finally a very popular method consists in capturing biotinylated proteins on streptavidin-coated slides. Synthetic oligonucleotides or peptides, named aptamers, selected from random combinatorial synthetic libraries for high affinity and selectivity for their target molecules, are also useful molecules for the design of capture microarrays [48]. Enzymatic processing of immobilized substrates on microarray can also be monitored in a microarray format [49]. Immobilization of receptor ligands (oligosaccharides, hormones …) can be used for receptor binding studies. However, and so far, the most popular capture molecules for arrays are still antibodies. Although monoclonal antibodies have become key research tool in a various fields of biology, they have some disadvantages in terms of generation, cost, and overall applications. To bypass these problems, antibody fragments are now used as capture molecules. The development of Fab (Fragment antigen binding), scFv (single chain variable Fragment) [50], affibodies [51] and VHHs (variable domain of heavy chain antibody) [52] for antibody arrays leads to easier fabrication processes. VHHs have many advantages for biotechnological applications such as their high microbial production and their ability to recognize antigenic sites that are normally not recognized by conventional antibodies such as enzyme active sites. Furthermore, contrary to conventional antibodies or others antibody fragments, VHHs have been shown to remain functional after incubation at high temperatures [53]. This high stability is mainly attributed to their efficient refolding after chemical or thermal denaturation [54]. Considering stability, selectivity and high affinity to ligand as well as conserved structure and conferring similar physical and chemical properties, antibody arrays remain the method of choice and VHH arrays might lead to a high potential alternative. 2.2.4. Applications of protein microarrays Miniaturized and paralleled immunoassays are generally interesting for all diagnostic applications, where several parameters in an individual sample have to be determined simultaneously from a limited amount of material. In diagnostics, protein and antibody microarrays are applied for the detection of antigens and antibodies in blood samples as well as in the profiling of sera to discover new disease markers. The development of fully validated biomarkers into a clinical diagnostic test involves two steps: the first phase involves multiplexed immunoassays development for validated biomarkers and the next stage is in vitro diagnostic development [50]. In the area of cancer research, a recombinant scFv antibody microarray has been used to classify female, post menopausal, age-matched metastatic breast cancer patients versus healthy controls based on differential serum protein profiling [55]. This array of 129 scFv allowed profiling at picomolar range analytes while consuming only microliters of clinical samples. Other protein microarrays, generated for the diagnostic of autoimmune diseases such as lupus erythematosus [55] and allergy tests have already been made and proved the potential of this technique. Microarrays have enormous potential to become robust and reliable diagnostic assays. Although first experiments were made to apply protein and antibody arrays in the field of diagnostics, the major application area is still basic proteome research. The focus of basic proteome research is to target identification and discovery, functional analysis of proteins and cellular expression profiling. Protein phosphorylation is an especially important regulator of many processes inside cells. For instance, constitutive Notch activation is required for the proliferation of a subgroup of T-cell acute lymphoblastic leukemia. Downstream pathways that transmit pro-oncogenic signals are not well characterized. To explore these issues, a reverse phase protein microarray has been used to profile the phosphorylation state of 108 distinct epitopes on 82 signalling proteins in a panel of 13 human T-cell leukemia lines [23]. The authors found that gamma-secretase inhibitor treatment suppressed the phosphorylation of multiple signalling proteins in the mTOR pathway in a Notch-specific manner, indicating that the mTOR pathway receives activating signals from Notch. This discovery might lead to a novel therapeutic approach for Notch-dependent cancers. Another important aspect is the studies of membrane proteins. A membrane protein array has been developed for the analysis of the ligand binding properties of receptors [56]. These microarrays consisted of immobilized G protein coupled receptors that were capable of demonstrating the specific binding of their respective targets. Membrane microarrays allow multiplexed screening models of the cell surface and should also enable highly parallel studies of fundamental processes such as multivalent interactions and cell–cell communication. Most biological processes are driven at some level by enzymes, and the states of those processes are determined by the activity level of the enzymes. An antibody microarray method was developed to permit isolation, detection and identification of the functional state of probe-labelled enzymes in a complex proteome in a single step [57]. This technology will facilitate the characterization of new markers and targets for the diagnosis and treatment of human diseases. The wide variety of different applications in which protein and antibody microarrays are employed reflects the versatility of the technology. Although several limitations currently hinder their use, it can be expected that advances such as protein production on the chip will facilitate the generation of protein arrays. Recently, high-density self-assembling protein microarrays have been developed based on the nucleic acid programmable protein array (NAPPA) concept. This approach allows the display of thousands of proteins that are freshly produced and captured in situ from immobilized cDNA templates using cell-free lysates [58]. The authors arrayed up to 1000 unique human cDNAs and obtained high yields of protein expression and capture. This next generation NAPPA method allows making fresh proteins in situ to produce high-content protein microarray. This will promote the application of protein arrays for qualitative measurements, such as interaction or modification screening. Protein and antibody microarrays still require modifications and optimization to overcome several limitations such as sensitivity and cross reactivity. Until then, the benefits of microarrays such as low analyte consumption should help to facilitate the analysis of complex samples in small-scale applications. 2.2.5. Nanoarrays Using the previously described approaches, microarrays, often of less than 1cm2, have been fabricated with about 1000 proteins, at a density <2000probes/cm2. In some cases, arrays composed of about 10000 proteins have been manufactured. Low to medium density antibody microarrays (<1000 probes) and low- to high-density protein arrays (<10000 protein) have already been used for a variety of applications. However, a complex proteome is composed of >100000 proteins, requiring the use of high-density array (>10000probes/array) [59]. For this purpose, nanotechnologies must be implemented to generate miniaturized array at a high density. Different nanopattern techniques are available to create nanoarrays, such as microfluidics technology [60], dip-pen nanolithography [61] and microcontact printing. The choice of design will be dependent on several factors, such as nanopatterning techniques, properties of the substrate and the probes, as well as compatibility with the detection method. Detection methods for nanoarrays are either based on fluorescent label-dependent or label-free detection. Several types of labelling reagent are available and recently non-organic fluorophores such as metallic (Au), magnetic or silica nanoparticles [62] and QDs (see below) have been developed to provide multiple possibilities for highly sensitive detection under different conditions. For label-free detection methods, different methodologies have been developed such as atomic force microscopy, microcantilever [63], surface plasmon resonance [64], and nanowires, as biologically gated transistors, transducing molecular binding events into real-time electrical signals [65]. An example of microcantilever is a study of Backmann et al. [66] where they reported a microcantilever-based immunosensor operated in static deflection mode with a performance comparable with surface plasmon resonance technique, using scFv antibody fragments as receptor molecules. It has been demonstrated that molecular adsorption also results in measurable mechanical forces [66]. In this study, the differential deflection signal revealed specific antigen binding and was proportional to the antigen concentration in solution. Using small and oriented scFv fragments as receptor molecules they increased the sensitivity of microcantilever to 1nM. Detecting biomolecular interactions by measuring nanomechanical forces offers an exciting opportunity for the development of highly sensitive, miniature and label-free biological sensors. 3. Choices of the labelling strategies Although the detection on the microarrays can be performed without any label, traditionally the detection of targets is done with the use of several labels. The most commonly used approaches employ radioactive labels, immunoperoxydase reaction and fluorescent labels including lanthanide chelates and organic dyes. Contrary to fluorescent organic dyes, immunoperoxydase labelling does not suffer of tissue autofluorescence and the peroxydase is durable and does not quench. Thus the stain is durable and can be reviewed at a later time. However immunoperoxydase labelling presents some disadvantages in that realistically only one staining can be done on one sample. Additionally, multistep enzyme-based immunoperoxydase stain is inherently not stochiometric and is dependent on experimental conditions such as pH, temperature, concentration, which makes it not suitable for quantitative analysis. Because the number of patient samples are usually limited (generally four needle biopsies for patient) immunoperoxydase staining may not be efficiently used for microarrays-based quantitative detections and diagnosis. Lanthanide chelates is an important and special group of luminescent molecules. These labels contain an organic dyes (commonly used compounds are EDTA, beta-diketonates, DTPA, etc.), which serves as a sensitizers to absorb the excitation light and to transfer this energy to the lanthanide ions which are usually trivalent cations [67], [68]. Consequently, lanthanide chelates exhibit broad excitation spectra and narrow emission spectra. The lanthanide chelates may be coupled to a wide variety of compounds to create specific labels, probes, diagnostic and/or therapeutic reagents. However, despite the excellent spectral properties of lanthanide chelates, their photochemical stability is limited. Fluorescence labelling is sensitive, safe, quantitative, allows multicolour staining but suffers from photobleaching of organic dyes and from the contribution of background autofluorescence generated by immobilized proteins and support. Although fluorescence is a sensitive detection technique, traditional organic dyes exhibit many limitations, especially when the multiplexed detections need to be performed. These limitations can be reconsidered with the use of inorganic fluorescent nanoparticles named quantum dots [7]. 4. Optical and physico-chemical properties of quantum dots: advantages and limits Quantum dots are semiconductor nanocrystals generally composed of elements from groups II–IV or III–V of the chemical periodic table, such as CdSe, CdTe and InAs. So-called core/shell QDs are usually composed by a fluorescent semiconductor core (CdSe, for example) coated with a shell of another semiconductor with a larger bandgap (CdS or ZnS) [69], [70]. The presence of the shell is almost essential today because its introduction increases significantly the chemical and fluorescent stabilities of QDs. The most widespread used QDs are composed of CdSe core covered with a ZnS shell [71]. This composition has proved to be the best choice in terms of QDs brightness and photo- and mechanical stability so far. The absorption and emission spectra of QDs depend on their chemical composition and it is actually possible to prepare QDs with the wavelengths of the fluorescence emission covering all visible and nearly-infra-red (NIR) regions of optical spectrum. Moreover, the size and shape of QDs can be controlled precisely by the time, temperature and ligand molecules used during their synthesis. Importantly, QDs of the same chemical composition absorb light in the UV and visible region of the optical spectrum but emit the fluorescence in the very narrow and ideally symmetric fluorescence emission bands of the wavelengths determined by their diameter. This possibility to excite QDs of different diameters (fluorescent colours) with a light of the same wavelength opens unique possibilities for multiplexing. Additionally, the surface chemistry of as-synthesized QDs permits the application of different approaches to QDs functionalisation and tagging with capture and other biomolecules thus allowing various biomedical applications (Fig. 1) [72], [73], [74]. The organic dyes currently in use for biomolecule labelling suffer from several limitations. We will discuss some of them now and the relative advantages for use QDs to overcome these limitations (Table 1). One of the key parameters for detection sensitivity – the brightness of the label – is determined by the product of the label's quantum yield by the extinction coefficient. Due to the fact that the QDs have similar quantum yields (around 70%) but much bigger extinction coefficients compared with the organic dyes (Table 2), their brightness is shown to be 20–100 times higher. This makes possible to achieve the QDs detection limit down to single-QD level and to ensure their ultrasensitive detection over the autofluorescence of cells, tissues or microarray supports. This high brightness also enables to decrease the quantity of fluorescent labels at a minimal quantity in order to limit the non-specific binding which occurs upon the use of highly concentrated QD conjugates applied for revelation on microarrays [75]. Moreover, it is worth to mention that QDs are 50–1000 times more stable against photobleaching than the best organic dyes [76]. This property enables to burn the autofluorescence by just illumination of the microarrays with the light of high intensity before the acquisition. The high stability of quantum dots also allows a retrospective reviewing of the experiments. | a Only single parameter may be measured. |
The organic dyes are not realistically suitable for multiplexed staining with more than four different labels. Indeed, the fact that they present a narrow absorption spectrum and a broad red-tailed emission spectrum makes difficult to avoid overlapping of absorption/emission spectra between the different fluorescent dyes. Moreover, when different organic dyes are used, practically the same number of excitation wavelengths should be applied thus making experimental set-ups complex and expensive. For these reasons, organic fluorochromes are until now not too suitable for high-throughput multiplexed microarrays. In a marked contrast, QDs have a broadband absorption spectrum and narrow symmetric emission spectrum (the emission peak width of QDs in solution is typically about 20–25nm at room temperature [77]) and QD of any size can be excited with the excitation light of the same wavelength from UV to blue-violet region of the optical spectrum [75] (Fig. 1D). These characteristics make them particularly interesting for multiplexed analysis because they allow the excitation of all quantum dot populations with the same exciting light using such sources as diodes, lasers or lamps, and to collect non overlapping specific emission peaks just with different filters. Moreover, although most of autofluorescence from cells and tissues is excitable by UV light, the large Stokes shifts (the shift between excitation and emission wavelengths) of QDs provide a mean to easily separate the QDs fluorescence from the background autofluorescence even in the case of the use of UV light for sample fluorescence excitation. Fig. 2 provides an example of multiplexed recognition of four prostate cancer biomarkers using QDs as labels and Table 1 summarizes relative comparison of properties of QDs, classical organic dyes and immunoenzyme assays. Some of optical characteristics of quantum dots and organic dyes are also compared in Table 2. Although all these promise of multiplexing, labelling with QDs still needs several improvements. The QD anatomy is not perfect yet. The size and stability of QDs strongly depend on the core/shell components chosen and on their functionalisation strategy. The “perfect” QD must be fully stable in aqueous solutions at a wide range of pH and ionic strengths but it must also be as tiny as possible. Today, one may cover QDs by an additional organic shell to increase significantly their stability but such stabilization procedure increases their size up to 20nm or even more. Future work should find the compromise between the QDs stability and size in order to develop the smallest nanoparticles which will be sufficiently stable in crude biological fluids and tissues. Additionally, the QDs fluorescence randomly blinks, this point may be particularly important for single-molecule detection and accurate fluorescence intensity quantification. Surface defects in the QD crystal structure act as the temporary “traps” for the electron or hole, preventing their radiative recombination (which leads to the QD fluorescence emission). The alternation of trapping and entrapping events results in QD blinking being visible at the single-molecule level [78], [79]. This blinking can be greatly decreased in ambient and biologically relevant conditions by QDs surface passivation, thus making them ideal labels detectable on the single-molecule level for a variety of applications [80]. 5. Use of quantum dots in microarray applications 5.1. Technical requirements for quantum dots applications The use of the quantum dot labelling in microarray analysis is relatively new, so, for an accurate multiplexed analysis, the quantitative comparison of traditional, based on the use of organic dyes, and newly developing QDs-based approaches needs to be realized. Some recent studies have already demonstrated the advantages and key shortcomings for QDs application to microarray technologies thus establishing the basic technical requirements. For DNA microarrays the best choice for labelling with QDs seems to be the body-labelling approach which allows a uniform incorporation of the labels between the targets. Moreover, as QDs are bigger than organic dyes, the complexes of QDs with modified nucleotides seem to be too bulky for correct uptake by polymerases and thus their insertion rate is both low and not uniformed. For this reason, an indirect labelling methodology of body-labelling technique, with the use of a tiny intermediate molecule, is the most suitable. In a very recent study, Vora et al. [81] raised the need of comparative analysis of the different labelling techniques and necessity of exact determination of the size limit for correct labelling of the probe. Although the authors have succeeded in comparing different labelling techniques applied to the DNA microarrays [81], they do not provide an overview of systematic studies in standardized conditions. The procedures of application of different labels to various detection systems have been discussed thus leading to experimental biases for comparison of detection sensitivity and differences of label repartition for each labelling technique, due to the labels differential brightness and size. For protein microarrays, it is much more difficult to establish the best labelling technology employing QDs. As the optical properties of QDs in solution ultimately depend on the nature and the number of ligands present on their surface, these parameters need to be carefully controlled during QDs solubilisation and conjugation steps. Additionally, the functionality of the surface ligands should also be controlled in order to improve additionally the QD stability in solution. Proceeding in this way, Geho et al. [36] have used the multifunctional polyethyleneglycol (PEG)-based polymers to solubilise and functionalise the QDs in order to avoid unspecific binding of them when applied to a reverse phase protein microarray. It should also be underlined that the QDs conjugation step is not perfect yet. Indeed, Pathak et al. have recently demonstrated that the number of functional antibodies conjugated with QDs, i.e. those antibodies which are sterically available for functional binding to target protein, is much lower than expected. They have compared the number of functional antibodies in the context of the two general models for protein labelling with QDs: (1) direct coupling of functionalised QDs with the monoclonal antibodies using standard bioconjugation techniques, or (2) employing the biotin–streptavidin binding of streptavidin-coated QDs with the biotinylated antibodies. Pathak et al. estimated that only very rare antibodies rest to be functional after their direct bioconjugation with the QDs whereas the streptavidin-coated QDs coupled to biotinylated antibodies contain from 10 to 20 times higher number of functional antibodies [82]. They also demonstrated that currently available commercial antibody–QD bioconjugates possess only 0.076±0.014 functional antibodies per QD. Although the use of the streptavidin–biotin system involves an additional step in the probe-labelling procedure and also an additional increase in the size of the formed complexes, thus presuming a potential decrease of the ability to bind the targeted protein by cluttering in a sterically restricted analysis like solid-state microarrays, such labelled complexes can be easily detected in protein microarrays [5]. The fluorescence intensity of detected signal was found to be nearly 30 times higher when streptavidin–biotin interacting system was used. The study of Pathak et al. revealed that the number of functional antibodies in the assays employing QDs should be carefully controlled. These results revealed a major problem of actually available QD–antibody conjugates and demonstrated a clear preference of the streptavidin–biotin system over the direct coupling of antibodies with the QDs. Another important requirement for the use of QD-adapted microarrays is an optimized read-out fluorescence detection system. Development of such system is certainly a critical element for sensitive and accurate detection of emissions from QDs of different fluorescence colours (sizes) in a multiplexed analysis. Nowadays, development of quantitative QD-based microarrays is still limited by the absence of conventional instrumentations optimized for multiplexed QD detection. For example, existing Red–Green–Blue (RGB) cameras cannot (i)distinguish pure yellow colour from a colocalized mixture of green and red colours, (ii)remove contribution of tissue autofluorescence, (iii)resolve spectrally overlapping fluorescent components, and (iv)provide a numerical read-out [83]. The actual disadvantages of largely widespread applied detection systems can be solved with the use of a hyperspectral imaging approach providing not only information about the intensity level of fluorescence signal but also complete spectral informations on the sample. Traditional fluorescence detection systems confine the user to work with a limited number of fixed chosen wavelengths requiring the knowledge of sample's excitation and emission spectra whereas a hyperspectral imaging provides a series of images acquired at many different wavelengths. This possibility allows the fine differentiation between two nearly similar but different emitting wavelengths. For example, truly yellow emission may be easily differentiated of green and red colocalization colours as well as the spatial distribution of QDs emitting at only slightly different wavelengths could be differentiated. These possibilities make the hyperspectral technology really promising for multiplexed analysis in microarray applications (Fig. 2). 5.2. State-of-the-art of quantum dots application in microarray technologies One of the first steps required for the establishment of QD-based multiplexed microarrays is to have a specific and quantitatively measurable signal. A recently published study demonstrated that QD-based protein microarrays can specifically detect one among six different cytokines at a picomolar concentration with the use of streptavidin-tagged QDs [5]. This detection technology was proved to be sensitive enough to detect the presence of antigens in solution relevant to biomarker concentration in physiological range. However, the sensitivity was found to be not sufficient enough for all tested cytokines. The detection limit for all cytokines ranged from 10 to 100pg/mL, concentrations which are not sufficiently low to detect necessary biomarkers in the healthy or early stage cancer specimens. Shingyoji et al. have demonstrated that the protein detection with QD conjugates can be achieved with 1pmol sensitivity – a value comparable with the sensitivity of standard commercial assays based on horseradish peroxidase (HRP)-catalyzed diaminobenzidine (DAB) chromogenesis [84]. The authors have designed a reverse protein microarray platform based on QDs detection that can perform accurate and reproducible quantification of protein concentrations in a crude cell lysate background. It is important to note that many detection systems in use nowadays are able to detect QD blinking, an effect which is characteristic only for the signals from single nanoparticles and not for QD ensembles or aggregates. QDs are thus sufficiently bright and existing detection systems are sufficiently sensitive to detect QDs at a single-particle level. This fact supposes that the key parameter which needs to be improved in order to fully explore the potential of QDs in detection of low abundant biomarkers on the single-molecule level is an efficient protein labelling preserving the functionality of tagged antibodies. When this goal will be achieved, the very early stage cancer diagnosis with QD conjugates will start to be realistically deliverable. In 2005, Geho et al. already made reverse phase microarrays employing streptavidin-coated QDs as labels and hyperspectral imaging as detection system demonstrating that only 1.5nL of lysate sample on a nitrocellulose slide may be detected with the QD conjugates [36]. The data revealed two critical parameters required for effective use of QDs in accurate and multiplexed microarrays: the target labelling method and the choice of detection system used. Geho et al. have also shown that steptavidin-coated QDs bind equally well to protein spots arrayed with specific biotinylated antibodies or non-biotinylated antibodies. They have succeeded to eliminate this non-specific binding with the use of PEG-coated steptavidin-tagged QDs and they have detected and characterized QDs’ fluorescence both in time- and concentration-dependent manners [36]. Moreover, Geho et al. have detected fluorescence of QD-labels with sensitivity that is at least comparable to standard commercial assays based on horseradish peroxides analysed DAB chromogenesis. An important advantage for future biomedical use of QDs-based microarrays in diagnostics for detection of rare biomarkers/proteins/nucleotide transcripts is that QDs were sufficiently bright for a very long time to be detected in the standard microarray format on the nitrocellulose slide. Geho et al. have demonstrated that QD labelling can be reviewed at least 6 months after the staining experiment with just a slight change in emission spectra during this time. Although only one QD population (colour) was tested in their work, Geho et al. have also demonstrated that QD labelling in protein microarrays can be effectively detected with hyperspectral imaging system. Further studies are required to prove the possibility to recognize simultaneously several QDs populations in order to validate multiplexed microarray detection approaches based on the hyperspectral imaging analysis particularly suitable for detection of several biomarkers in single-shot diagnostics. In one of the most promising examples of QD applications to cancer diagnosis with low quantity of patient samples, Ghazani et al. have detected three biomarkers (cytokeratine, epidermal cadherin and epidermal grown factor receptor (EGFR)) in protein microarrays specifically labelled with three different QDs (565, 605 and 655nm, respectively) using FFPE (formalin-fixed paraffin-embedded) lung carcinoma xenografts [85]. Due to the known fact that the EGFR expression level varies between the different FFPE, the authors validated their detection system through comparison of EGFR expression levels detected by QD fluorescence and by quantitative reverse transcriptase polymerase chain reaction (RT-PCR). Determined with QD fluorescence technique expression levels were found to have a strong statistical correlation reaching the correlation coefficient 0.90 with the results of RT-PCR control. The authors further compared the fluorescence intensity detected for these three biomarkers (labelled by respective QDs) in single-parameter and multiplexed conditions. They observed a decrease of QD565 and QD605 fluorescence intensities accompanied by an increase of QD655 fluorescence intensity in multiplexed assays. This fact revealed that the multiplexed conditions provoked a crosstalk between the different QDs of different colours thus leading to changes of their relative fluorescence intensities. The authors supposed an appearance of the FRET (Förster Resonance Energy Transfer) effect between the different populations of quantum dots. This problem is supposed to be important to solve for an establishment of the multiplexed cancer diagnostics employing microarray technology and QDs of different colours. 5.3. Future directions Existing methodologies for QD labelling and QD fluorescence detection in DNA and protein microarrays are not perfect yet. In order to provide a most discriminatory QD fluorescence detection, the hyperspectral imaging should be used to enable the spectral differentiation and quantification of labels fluorescence (Fig. 2). A standard hyperspectral imaging set-up includes a band-pass controlling device (such as acousto-optic tuneable filters or diffraction grating) and a scientific-grade monochrome CCD camera. Despite the acousto-optic tuneable filters allow to electronically change both the wavelength and the bandwidth, their low quality imaging hindered their use during several years. Recently, this major problem was solved by using both innovative transducer design and long interaction length crystals. Among the different hyperspectral imaging systems, the acousto-optic tuneable filter system seems to be the most suitable through the high speed of wavelength change (∼50μs), the high quality of fluorescence detection and the relatively fast treatment of acquired data. Although additional comparative work is still needed to evaluate real potential of such systems against traditional fluorescence detection systems, the combination of the unique QD optical characteristics and hyperspectral imaging will certainly enhance their respective advantages and capabilities. Another powerful technology for detection with QDs is the surface plasmon enhanced fluorescence spectroscopy [86]. Robelek et al. have successfully measured the spectrally resolved surface plasmon enhanced fluorescence signals derived from the probes of hybridized QD–DNA conjugates on the surface of the solid-state microarray chip [86]. Although the use of streptavidin-tagged QDs instead of the QD conjugates with monoclonal antibodies provides bigger amount of functional antibodies on the surface of QDs [78], their number is still too low. This parameter should be definitively improved for the future because it is critical for sensitivity and specificity of detection on the protein microarrays. Indeed, it is very important to be able to understand if the absence of label signal means the absence of detecting protein or is determined by the non-sufficient quantity of functional antibodies tagged with the quantum dot. Although the streptavidin-tagged QDs were shown to be efficient enough for labelling biotinylated antibodies [36], their application may be limited since the size of streptavidin-tagged QD conjugates is too big. Therefore, the size of detecting QD-conjugates may be decreased if the functionality of the ligands directly tagged with the QD can be improved. This may be done by development of optimized conjugation protocols preserving the ligand function and also providing the conjugates where the ligands are oriented in a manner to be fully exposed to interaction with the targets. Ghazani et al. have demonstrated that the multiplexed assays employing QDs of different colours may suffer from FRET effect [85]. This major problem for multiplexed assays should be addressed by careful selection of quantum dot fluorescence colours and variations of the capture molecules bound to their surface. Indeed, quantum yield and extinction of QDs (and thus their brightness) are size- and composition-dependents. Another potential solution to compensate the decrease of fluorescence detected for the QDs of smaller diameters (lower wavelengths) is the use of “quantum rods”, nanocrystals which are rod-shaped and their brightness depends of their length whereas their wavelength emission characteristics depend on their diameter and chemical composition. Finally the QDs-encoded polymeric beads can be also used to increase the distance between the quantum dots of different populations embedded within the polymeric matrices. Although the protocols of bead encoding with the QDs of different colours always need to be improved, Eastman et al. have demonstrated a breakthrough ultrasensitive multiplexed application of such QDs-nanobarcodes [87]. Although the FRET effect may be a problem for multiplexed detection with the QDs of different colours, it may be efficiently employed for detection of immune complex formation between the molecules labelled with QDs or organic dyes with carefully selected fluorescence colours. FRET technology involves the transfer of energy from a donor to an acceptor particle whenever the distance between the donor and the acceptor is smaller than the Förster radius (<10nm) and fluorescence emission of donor is strongly overlapping with the absorption of acceptor [88]. In such assays, QDs of different colours and organic dyes may be used as the donors and as the acceptors. Indeed the QDs emission spectra are much narrower and more symmetric than that for emission of organic dyes, making much easier to distinguish the emission of QD-donor from that of the acceptor. Several studies have shown the potential of QDs in the FRET approach for measuring the subtle changes in distances between the donor and acceptor labels [89], protein conformational changes [90] and analysing protein interactions [91], [92]. The use of quantum dots in FRET studies provides many advantages over organic fluorophores. The combination of quantum dots and organic dyes labels should result in a FRET-based immunoassay technique applicable to protein microarrays. The cancer research is a field which will greatly take advantage of QD-based labelling method in microarrays [93]. Indeed, microarrays are already developed as diagnostic approaches and the critical limit for efficient diagnosis is its sensitivity because an early detection is the key for cancer survival [28]. Up to now, almost all tumour diagnostics detect tumours at an advanced stage, sometimes too advanced to start an efficient therapy. An increase in sensitivity permits to push diagnosis limits, which have many consequences: cancers could be detected early thus therapy would be more effective and shorter and thus less restrictive and expensive for the patient. A best therapy also leads to an important decrease in the number of potential metastasis and thus the relapse likelihood. Furthermore, the paralleling capacities of microarrays allow a best understanding of cellular events required for tumour development in allowing the comparison of genes expression, genes regulation and proteins activity between metastatic and healthy tissues which might allow the discovery of some new drugs, biomarkers and potential drug targets. All these discoveries raise exciting opportunities for adapted and personalized cancer therapies based on the individual molecular profile of each patient and early diagnosis of cancers. However, the identification and selection of biomarkers that can be tested in a diagnostic assay is a critical point to make accurate diagnostics. Microarrays are a good approach to find new cancer biomarkers and to establish which one should be chosen for clinical use [94]. Microarrays with quantum dot labelling present interesting promises in such areas as rapid detection of pathogens [95] or discovery of natural common antibody targets following a pathogen invasion. This can lead to new therapy approaches for drug delivery and target. As an example, Li et al. used QDs on protein microarrays against 144 proteins to study the antibody response of plague patients [24]. This study permitted to provide information on common antibody targets, made fine patient groups through their antibody response and increased the understanding of the lack of immune responses to eradicate pathogen agents in order to improve therapy. 6. Concluding remarks During the last decades, the paralleling capacities of microarrays have already provided a global view of pathological states at a molecular level through better understanding of mechanisms of gene regulations, expressions and patient immune responses. These points are particularly important for an efficient therapy. However, organic dyes regularly used as labels in microarrays technologies are not easily suitable for multiplexed approaches. We have demonstrated in this review that the microarray analysis will be more efficient, sensitive and specific with the use of quantum dot labelling. Employment of quantum dots should ensure orders of magnitude improvement of the target detection limit, gain of accuracy and time, reduction of assay costs and breakthrough increase in multiplexing capabilities. Acknowledgements This work was supported by the French National Research Agency (Agence Nationale de Recherche – ANR) programs ‘Nanosciences and Nanotechnologies’ under the grant ANR-07-PNANO-051-01 and by the ANR program “Research and Innovations in Biotechnology” under the grant ANR-07-RIB-012-03. Partial supports from the Ligue Contre le Cancer, NATO SfP-983207, RFBR/CNRS (07-04-92164/PICS3868) and RFBR 07-04-01421 projects are also acknowledged. The study sponsors had no involvement in the study design, in the collection, analysis and interpretation of data, in the writing of the manuscript or in the decision to submit the manuscript for publication. Reviewers Dr. Andrey L. Rogach, Ludwig-Maximilians-Universität München, Lehrstuhl für Photonik und Optoelektronik, Amalienstrasse 54, D-80799 Munich, Germany. Dr. Yuri Volkov, Trinity College Dublin, Institute of Molecular Medicine, Department of Clinical Medicine, James's Street, Dublin, 8, Ireland. Prof. Bernard Desoize, Faculte de Pharmacie, IFR53 Biomolecules, 51 Rue Cognacq-Jay, F-51096 Reims, France. References [1]. [1]Ehrenreich A. DNA microarray technology for the microbiologist: an overview. Appl Microbiol Biotechnol. 2006;73:255–273. MEDLINE |
CrossRef
[2]. [2]Stoll D, Templin MF, Schrenk M, Traub PC, Vöhringer CF, Joos TO. Protein microarray technology. Front Biosci. 2002;7:c13–32.
CrossRef
[3]. [3]Gracey AY, Cossins AR. Application of microarray technology in environmental and comparative physiology. Annu Rev Physiol. 2003;65:231–259. MEDLINE |
CrossRef
[4]. [4]Park HG, Song JY, Park KH, Kim MH. Fluorescence-based assay formats and signal amplification strategies for DNA microarray analysis. Chem Engen Sci. 2006;61:954–965. [5]. [5]Zajac A, Song D, Qian W, Zhukov T. Protein microarrays and quantum dot probes for early cancer detection. Colloids Surf B Biointerfaces. 2007;58:309–314. MEDLINE |
CrossRef
[6]. [6]Resch-Genger U, Grabolle M, Cavaliere-Jaricot S, Nitschke R, Nann T. Quantum dots versus organic dyes as fluorescent labels. Nat Methods. 2008;5:763–775.
CrossRef
[7]. [7]Nabiev I, Sukhanova A, Artemyev M, Oleinikov V. Fluorescent colloidal particles as a detection tools in biotechnology systems. In: Elaissari A editors. Colloidal Nanoparticles in Biotechnology, pp. 376. Wiley & Sons Inc.; 2008;p. 133–168. [8]. [8]Sukhanova A, Nabiev I. Fluorescent nanocrystal quantum dots as medical diagnostic tools. Exp Opin Med Diagn. 2008;2:429–447. [9]. [9]Sukhanova A, Nabiev I. Liquid-phase chips based on the fluorescent nanocrystal-encoded microbeads for cancer diagnostics. Crit Rev Oncol Hematol. 2008;68:39–59. Abstract | Full Text |
Full-Text PDF (2120 KB)
|
CrossRef
[10]. [10]Ling MM, Ricks C, Lea P. Multiplexing molecular diagnostics and immunoassays using emerging microarray technologies. Expert Rev Mol Diagn. 2007;7:87–98.
CrossRef
[11]. [11]Blohm DH, Guiseppi-Elie A. New developments in microarray technology. Curr Opin Biotechnol. 2001;12:41–47. MEDLINE |
CrossRef
[12]. [12]McGall G, Labadie J, Brock P, Wallraff G, Nguyen T, Hinsberg W. Light-directed synthesis of high-density oligonucleotide arrays using semiconductor photoresists. Proc Natl Acad Sci U S A. 1996;93:13555–13560. MEDLINE |
CrossRef
[13]. [13]Hoek KS. DNA microarray analyses of melanoma gene expression: a decade in the mines. Pigment Cell Res. 2007;20:466–484.
CrossRef
[14]. [14]Duncan R. DNA microarray analysis of protozoan parasite gene expression: outcomes correlate with mechanisms of regulation. Trends Parasitol. 2004;20:211–215. MEDLINE |
CrossRef
[15]. [15]Yamakawa H, Hirose T, Kuroda M, Yamaguchi T. Comprehensive expression profiling of rice grain filling-related genes under high temperature using DNA microarray. Plant Physiol. 2007;144:258–277. MEDLINE |
CrossRef
[16]. [16]Wakamatsu E, Nakamura Y, Matsumoto I, et al. DNA microarray analysis of labial salivary glands of patients with Sjogren's syndrome. Ann Rheum Dis. 2007;66:844–845. MEDLINE |
CrossRef
[17]. [17]Feferman T, Aricha R, Menon R, Souroujon MC, Berrih-Aknin S, Fuchs S. DNA microarray in search of new drug targets for myasthenia gravis. Ann N Y Acad Sci. 2007;1107:111–117.
CrossRef
[18]. [18]Dorrell N, Hinchliffe SJ, Wren BW. Comparative phylogenomics of pathogenic bacteria by microarray analysis. Curr Opin Microbiol. 2005;8:620–626. MEDLINE |
CrossRef
[19]. [19]Tao SC, Chen CS, Zhu H. Applications of protein microarray technology. Comb Chem High Throughput Screen. 2007;10:706–718.
CrossRef
[20]. [20]Lee Y, Lee EK, Cho YW, et al. ProteoChip: a highly sensitive protein microarray prepared by a novel method of protein immobilization for application of protein–protein interaction studies. Proteomics. 2003;3:2289–2304. MEDLINE |
CrossRef
[21]. [21]Sasakura Y, Kanda K, Yoshimura-Suzuki T, Matsui T, Fukuzono S, Shimizu T. Investigation of the relationship between protein–protein interaction and catalytic activity of a heme-regulated phosphodiesterase from Escherichia coli (Ec DOS) by protein microarray. Biochemistry. 2005;44:9598–9605. [22]. [22]Zha H, Raffeld M, Charboneau L, et al. Similarities of prosurvival signals in Bcl-2-positive and Bcl-2-negative follicular lymphomas identified by reverse phase protein microarray. Lab Invest. 2004;84:235–244. MEDLINE |
CrossRef
[23]. [23]Chan SM, Weng AP, Tibshirani R, Aster JC, Utz PJ. Notch signals positively regulate activity of the mTOR pathway in T-cell acute lymphoblastic leukemia. Blood. 2007;110:278–286. MEDLINE |
CrossRef
[24]. [24]Li B, Zhou D, Wang Z, et al. Antibody profiling in plague patients by protein microarray. Microbes Infect. 2008;10:45–51.
CrossRef
[25]. [25]Xu R, Gan X, Fang Y, Zheng S, Dong Q. A simple, rapid, and sensitive integrated protein microarray for simultaneous detection of multiple antigens and antibodies of five human hepatitis viruses (HBV, HCV, HDV, HEV, and HGV). Anal Biochem. 2007;362:69–75. MEDLINE |
CrossRef
[26]. [26]Yuk CS, Lee HK, Kim HT, Choi YK, Lee BC, Chun BH, et al. Development and evaluation of a protein microarray chip for diagnosis of hepatitis C virus. Biotechnol Lett. 2004;26:1563–1568. MEDLINE |
CrossRef
[27]. [27]Zhu H, Hu S, Jona G, et al. Severe acute respiratory syndrome diagnostics using a coronavirus protein microarray. Proc Natl Acad Sci U S A. 2006;103:4011–4016. MEDLINE |
CrossRef
[28]. [28]Smith AM, Dave S, Nie S, True L, Gao X. Multicolor quantum dots for molecular diagnostics of cancer. Expert Rev Mol Diagn. 2006;6(March):231–244.
CrossRef
[29]. [29]Brown PO, Botstein D. Exploring the new world of the genome with DNA microarrays. Nat Genet. 1999;21:33–37. MEDLINE |
CrossRef
[30]. [30]Granjeaud S, Bertucci F, Jordan BR. Expression profiling: DNA arrays in many guises. Bioessays. 1999;21:781–790. MEDLINE |
CrossRef
[31]. [31]Saitou M, Yabuta Y, Kurimoto K. Single-cell cDNA high-density oligonucleotide microarray analysis: detection of individual cell types and properties in complex biological processes. Reprod Biomed Online. 2008;16:26–40. Abstract |
Full-Text PDF (1196 KB)
|
CrossRef
[32]. [32]Zekri AR, Hafez MM, Bahnassy AA, et al. Genetic profile of Egyptian hepatocellular-carcinoma associated with hepatitis C virus Genotype 4 by 15 K cDNA microarray: preliminary study. BMC Res Notes. 2008;1:106. [33]. [33]Liang RQ, Li W, Li Y, et al. An oligonucleotide microarray for microRNA expression analysis based on labeling RNA with quantum dot and nanogold probe. Nucleic Acids Res. 2005;33:e17.
CrossRef
[34]. [34]Williams BA, Gwirtz RM, Wold BJ. Genomic DNA as a cohybridization standard for mammalian microarray measurements. Nucleic Acids Res. 2004;32:e81.
CrossRef
[35]. [35]Kader HA, Tchernev VT, Satyaraj E, et al. Protein microarray analysis of disease activity in pediatric inflammatory bowel disease demonstrates elevated serum PLGF, IL-7, TGF-beta1, and IL-12p40 levels in Crohn's disease and ulcerative colitis patients in remission versus active disease. Am J Gastroenterol. 2005;100:414–423. MEDLINE |
CrossRef
[36]. [36]Geho D, Lahar N, Gurnani P, et al. Pegylated, steptavidin-conjugated quantum dots are effective detection elements for reverse-phase protein microarrays. Bioconjug Chem. 2005;16:559–566. MEDLINE |
CrossRef
[37]. [37]Hall DA, Ptacek J, Snyder M. Protein microarray technology. Mech Ageing Dev. 2007;128:161–167. MEDLINE |
CrossRef
[38]. [38]Zhu H, Bilgin M, Bangham R, et al. Global analysis of protein activities using proteome chips. Science. 2001;293:2101–2105. MEDLINE |
CrossRef
[39]. [39]Angenendt P, Glökler J, Murphy D, Lehrach H, Cahill DJ. Toward optimized antibody microarrays: a comparison of current microarray support materials. Anal Biochem. 2002;309:253–260. MEDLINE |
CrossRef
[40]. [40]Wingren C, Ingvarsson J, Dexlin L, Szul D, Borrebaeck CA. Design of recombinant antibody microarrays for complex proteome analysis: choice of sample labeling tag and solid support. Proteomics. 2007;7:3055–3065.
CrossRef
[41]. [41]Seurynck-Servoss SL, White AM, Baird CL, Rodland KD, Zangar RC. Evaluation of surface chemistries for antibody microarrays. Anal Biochem. 2007;371:105–115.
CrossRef
[42]. [42]Wingren C, Steinhauer C, Ingvarsson J, Persson E, Larsson K, Borrebaeck CA. Microarrays based on affinity-tagged single-chain Fv antibodies: sensitive detection of analyte in complexe proteomes. Proteomics. 2005;5:1281–1291. MEDLINE |
CrossRef
[43]. [43]Grainger DW, Greef CH, Gong P, Lochhead MJ. Current microarray surface chemistries. Methods Mol Biol. 2007;381:37–57. [44]. [44]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. MEDLINE |
CrossRef
[45]. [45]Mircean C, Shmulevich I, Cogdell D, et al. Robust estimation of protein expression ratios with lysate microarray technology. Bioinformatics. 2005;21:1935–1942. [46]. [46]Sakanyan V. High-throughput and multiplexed protein array technology: protein–DNA and protein–protein interactions. J Chromatogr B Analyt Technol Biomed Life Sci. 2005;815:77–95. MEDLINE |
CrossRef
[47]. [47]Wingren C, Borrebaeck CA. Progress in miniaturization of protein arrays-a step closer to high-density nanoarrays. Drug Discov Today. 2007;12:813–819.
CrossRef
[48]. [48]Lesaicherre ML, Uttamchandani M, Chen GY, Yao SQ. Developing site specific immobilization strategies of peptides in a microarray. Bioorg Med Chem Lett. 2002;12:2079–2083.
CrossRef
[49]. [49]Evans D, Johnson S, Laurenson S, Davies AG, Ferrigno PK, Wälti C. Electrical protein detection in cell lysates using high-density peptide-aptamer microarrays. J Biol. 2008;7:3. [50]. [50]Houseman BT, Huh JH, Kron SJ, Mrksich M. Peptide chips for the quantitative evaluation of protein kinase activity. Nat Biotechnol. 2002;20:270–274. MEDLINE |
CrossRef
[51]. [51]Carlsson A, Wingren C, Ingvarsson J, et al. Serum proteome profiling of metastatic breast cancer using recombinant antibody microarrays. Eur J Cancer. 2008;44:472–480. Abstract | Full Text |
Full-Text PDF (924 KB)
|
CrossRef
[52]. [52]Renberg B, Nordin J, Merca A, et al. Affibody molecules in protein capture microarrays: evaluation of multidomain ligands and different detection formats. J Proteome Res. 2007;6:171–179. MEDLINE |
CrossRef
[53]. [53]Harmsen MM, De Haard HJ. Properties, production, and applications of camelid single-domain antibody fragments. Appl Microbiol Technol. 2007;77:13–22. [54]. [54]Van der Linden RH, Frenken LG, de Geus B, et al. Comparison of physical chemical properties of llama VHH antibody fragments and mouse monoclonal antibodies. Biochim Biophys Acta. 1999;1431:37–46. MEDLINE [55]. [55]Kingsmore SF. Multiplexed protein measurement: technologies and applications of protein and antibody arrays. Nat Rev Drug Discov. 2006;5:310–320. MEDLINE [56]. [56]Robinson WH, DiGennaro C, Hueber W, et al. Autoantigen microarrays for multiplexed characterization of autoantibody responses. Nat Med. 2002;8:295–301. MEDLINE |
CrossRef
[57]. [57]Fang Y, Frutos AG, Lahiri J. Membrane protein microarrays. J Am Chem Soc. 2002;124:2394–2395. MEDLINE |
CrossRef
[58]. [58]Sieber SA, Mondala TS, Head SR, Cravatt BF. Microarray platform for profiling enzyme activities in complex proteomes. J Am Chem Soc. 2004;126:15640–15641. MEDLINE |
CrossRef
[59]. [59]Ramachandran N, Raphael JV, Hainsworth E, et al. Next-generation high-density self-assembling functional protein arrays. Nat Methods. 2008;5:535–538.
CrossRef
[60]. [60]Situma C, Hashimoto M, Soper SA. Merging microfluidics with microarray-based bioassays. Biomol Eng. 2006;23:213–231. MEDLINE |
CrossRef
[61]. [61]Lee KB, Park SJ, Mirkin CA, Smith JC, Mrksich M. Protein nanoarrays generated by dip-pen nanolithography. Science. 2002;295:1702–1705.
CrossRef
[62]. [62]Yan J, Estévez MJ, Smith JE, et al. Dye-doped nanoparticles for bioanalysis. Nanotoday. 2007;2:44–50. [63]. [63]Huber F, Hegner M, Gerber C, Güntherodt HJ, Lang HP. Label free analysis of transcription factors using microcantilever arrays. Biosens Bioelectron. 2006;21:1599–1605. MEDLINE |
CrossRef
[64]. [64]Klenkar G, Liedberg B. A microarray chip for label-free detection of narcotics. Anal Bioanal Chem. 2008;391:1679–1688.
CrossRef
[65]. [65]Cheng MMC, Cuda G, Bunimovich YL, et al. Nanotechnologies for biomolecular detection and medical diagnostics. Curr Opin Chem Biol. 2006;10:11–19. MEDLINE |
CrossRef
[66]. [66]Backmann N, Zahnd C, Huber F, et al. A label-free immunosensor array using single-chain antibody fragments. Proc Natl Acad Sci U S A. 2005;102:14587–14592. MEDLINE |
CrossRef
[67]. [67]Yuan J, Sueda S, Somazawa R, Matsumoto K, Matsumoto K. Structure and luminescence properties of the tetradentate β-diketonate–europium(III) complexes. Chem Lett. 2003;32:492–493. [68]. [68]Wang F, Tan WB, Zhang Y, Fan X, Wang M. Luminescent nanomaterials for biological labelling. Nanotechnology. 2006;17:1–13. [69]. [69]Bruchez M, Moronne M, Gin P, Weiss S, Alivisatos AP. Semiconductor nanocrystals as fluorescent biological labels. Science. 1998;281(5385):2013–2016. MEDLINE |
CrossRef
[70]. [70]Chan WC, Nie S. Quantum dot bioconjugates for ultrasensitive nonisotopic detection. Science. 1998;25(281 (5385)):2016–2018. [71]. [71]Baranov AV, Rakovich YP, Donegan JF, et al. Effect of ZnS shell thickness on the phonon spectra in CdSe quantum dots. Phys Rev B. 2003;68:1653061–1653067. [72]. [72]Azzazy HME, Mansour MMH, Kazmierczak SC. From diagnostics to therapy: prospects of quantum dots. Clin Biochem. 2007;40:917–927.
CrossRef
[73]. [73]Shi C, Zhu Y, Cerwinka WH, et al. Quantum dots: emerging applications in urologic oncology. Urol Oncol. 2008;26(1):86–92. Abstract | Full Text |
Full-Text PDF (408 KB)
|
CrossRef
[74]. [74]Alivisatos AP, Gu W, Larabell C. Quantum dots as cellular probes. Annu Rev Biomed Eng. 2005;7:55–76. MEDLINE |
CrossRef
[75]. [75]Karlin-Neumann G, Sedova M, Falkowski M, Wang Z, Lin S, Jain M. Application of quantum dots to multicolor microarray experiments: four-color genotyping. Methods Mol Biol. 2007;374:239–251. MEDLINE [76]. [76]Sukhanova A, Devy J, Venteo L, et al. Biocompatible fluorescent nanocrystals for immunolabeling of membrane proteins and cells. Anal Biochem. 2004;324:60–67. MEDLINE |
CrossRef
[77]. [77]Zhong X, Feng Y, Knoll W, Han M. Alloyed ZnxCd1−xS nanocrystals with highly narrow luminescence spectral width. J Am Chem Soc. 2003;125:13559–13563. MEDLINE |
CrossRef
[78]. [78]Nirmal M, Dabbousi BO, Bawendi MG, et al. Fluorescence intermittency in single cadmium selenide nanocrystals. Nature. 1996;383:802–804.
CrossRef
[79]. [79]Empedocles SA, Neuhauser R, Shimizu K, Bawendi MG. Photoluminescence from single semiconductor nanostructures. Adv Mater. 1999;15(11):243–256. [80]. [80]Hohng S, Ha T. Near complete suppression of quantum dot blinking in ambient condition. J Am Chem Soc. 2004;126:1324–1325. MEDLINE |
CrossRef
[81]. [81]Vora GJ, Meador CE, Anderson GP, Taitt CR. Comparison of detection and signal amplification methods for DNA microarrays. Mol Cell Probes. 2008;22(5–6):294–300.
CrossRef
[82]. [82]Pathak S, Davidson MC, Silva GA. Characterization of the functional binding properties of antibody conjugated quantum dots. Nano Lett. 2007;7:1839–1845.
CrossRef
[83]. [83]True LD, Gao X. Quantum dots for molecular pathology: their time has arrived. J Mol Diagn. 2007;9:7–11. MEDLINE |
CrossRef
[84]. [84]Shingyoji M, Gerion D, Pinkel D, Gray JW, Chen F. Quantum dots-based reverse phase protein microarray. Talanta. 2005;67:472–478. [85]. [85]Ghazani AA, Lee JA, Klostranec J, et al. High throughput quantification of protein expression of cancer antigens in tissue microarray using quantum dot nanocrystals. Nano Lett. 2006;6:2881–2886. MEDLINE |
CrossRef
[86]. [86]Robelek R, Niu L, Schmid EL, Knoll W. Multiplexed hybridization detection of quantum dot-conjugated DNA sequences using surface plasmon enhanced fluorescence microscopy and spectrometry. Anal Chem. 2004;76:6160–6165. MEDLINE |
CrossRef
[87]. [87]Eastman PS, Ruan W, Doctolero M, et al. Qdot nanobarcodes for multiplexed gene expression analysis. Nano Lett. 2006;6:1059–1064. MEDLINE |
CrossRef
[88]. [88]Murata S, Herman P, Mochizuki K, et al. Spatial distribution analysis of AT- and GC-rich regions in nuclei using corrected fluorescence resonance energy transfer. J Histochem Cytochem. 2003;51:951–958. MEDLINE [89]. [89]Schwartz DE, Gong P, Shepard KL. Time-resolved Förster-resonance-energy-transfer DNA assay on an active CMOS microarray. Biosens Bioelectron. 2008;24:383–390.
CrossRef
[90]. [90]Schuler B, Eaton WA. Protein folding studied by single-molecule FRET. Curr Opin Struct Biol. 2008;18:16–26.
CrossRef
[91]. [91]Gertler A, Biener E, Ramanujan KV, Djiane J, Herman B. Fluorescence resonance energy transfer (FRET) microscopy in living cells as a novel tool for the study of cytokine action. J Dairy Res. 2005;72:14–19.
CrossRef
[92]. [92]Hallworth R, Currall B, Nichols MG, Wu X, Zuo J. Studying inner ear protein–protein interactions using FRET and FLIM. Brain Res. 2006;1091:122–131. MEDLINE |
CrossRef
[93]. [93]Jain KK. Applications of nanobiotechnology in clinical diagnostics. Clin Chem. 2007;53:2002–2009.
CrossRef
[94]. [94]Phan JH, Young AN, Wang MD. Selecting clinically-driven biomarkers for cancer nanotechnology. Conf Proc IEEE Eng Med Biol Soc. 2006;1:3317–3320. [95]. [95]Goldschmidt MC. The use of biosensor and microarray techniques in the rapid detection and identification of salmonellae. J AOAC Int. 2006;89:530–537. MEDLINE Gilles Rousserie graduated from Université de la Méditerranée in Marseille in 2007 and is preparing actually his Ph.D. Thesis under direction of Prof. Nabiev in the frames of research program developing introduction of the quantum dots technology within the technology of protein microarrays and immunological assays. Igor Nabiev pioneered, in the early 80s, together with the groups of Prof. Therese Cotton (US) and Dr. Eckhard Koglin (Germany), development of the new method of ultrasensitive analysis of biomolecules, surface-enhanced Raman scattering (SERS). He has started to develop, at the end of 90s, functional surface modification and biological applications of quantum dots as fluorescent labels for ultrasensitive detection and biomedical diagnosis. Prof. Nabiev actually coordinates National and International research programs developing nanotechnological tools for ultrasensitive non-isotopic detection and diagnosis with the solid-state and liquid-state chips. a EA n°3798 Détection et Approches Thérapeutiques Nanotechnologiques dans les Mécanismes Biologiques de Défense, Université de Reims Champagne-Ardenne, 51100 Reims, France b Laboratoire d’anticorps thérapeutiques et immunociblage, INSERM U624 Stress cellulaire, 13288 Marseille, France c Unité de Biotechnologie, Biocatalyse et Biorégulation, UMR6204, 44332 Nantes, France d Shemyakin-Ovchinnikov Institute of Bioorganic Chemistry, Russian Academy of Sciences, 117997 Moscow, Russia e CIC nanoGUNE Consolider Research Centre, E-20018 Donostia-San Sebastian, Spain  Corresponding author at: EA n°3798 Détection et Approches Thérapeutiques Nanotechnologiques dans les Mécanismes Biologiques de Défense, Université de Reims Champagne-Ardenne, 51100 Reims, France. Tel.: +33 326918127; fax: +33 326918127.
PII: S1040-8428(09)00083-3 doi:10.1016/j.critrevonc.2009.04.006 © 2009 Elsevier Ireland Ltd. All rights reserved. | |
|
FULL TEXT
