Random array of micro-spheres for the analysis of nucleic acids

[0001] This invention concerns biological microarray technology in general. In particular, it concerns a nuclei acid microarray system comprising random distributed nucleic acid probe modified micro-spheres coated on a substrate that contained no designated sites prior to coating. The micro-spheres contain optical barcodes generated from one or more colorants associated with the micro-spheres.

[0002] Ever since it was invented in the early 1990s (Science, 251, 767-773, 1991), high-density arrays formed by spatially addressable synthesis of bioactive probes on a 2-dimensional solid support has greatly enhanced and simplified the process of biological research and development. The key to current microarray technology is deposition of a bioactive agent at a single spot on a microchip in a spatially addressable manner.

[0003] Current technologies have used various approaches to fabricate microarrays. For example, U.S. Pat. Nos. 5,412,087, and 5,489,678 demonstrate the use of a photolithographic process for making peptide and DNA microarrays. The patent teaches the use of photolabile protecting groups to prepare peptide and DNA microarrays through successive cycles of deprotecting a defined spot on a 1 cm1 cm chip by photolithography, then flooding the entire surface with an activated amino acid or DNA base. Repetition of this process allows construction of a peptide or DNA microarray with thousands of arbitrarily different peptides or oligonucleotide sequences at different spots on the array. This method is expensive. An inkjet approach is being used by others (e.g., U.S. Pat. Nos. 6,079,283; 6,083,762; and 6,094,966) to fabricate spatially addressable arrays, but this technique also suffers from high manufacturing cost in addition to the relatively large spot size of 40 to 100 m. Because the number of bioactive probes to be placed on a single chip usually runs anywhere from 1000 to 100000 probes, the spatial addressing method is intrinsically expensive regardless of how the chip is manufactured. An alternative approach to the spatially addressable method is the concept of using fluorescent dye-incorporated polymeric beads to produce biological multiplexed arrays. U.S. Pat. No. 5,981,180 discloses a method of using color coded beads in conjunction with flow cytometry to perform multiplexed biological assay. Micro-spheres conjugated with DNA or monoclonal antibody probes on their surfaces were dyed internally with various ratios of two distinct fluorescence dyes. Hundreds of spectrally addressed micro-spheres were allowed to react with a biological sample and the liquid array was analyzed by passing a single micro-sphere through a flow cytometry cell to decode sample information. U.S. Pat. No. 6,023,540 and 6,266,459 discloses the use of fiber-optic bundles with pre-etched microwells at distal ends to assemble dye loaded micro-spheres. The surface of each spectrally addressed micro-sphere was attached with a unique bioactive agent and thousands of micro-spheres carrying different bioactive probes combined to form beads array on pre-etched microwells of fiber optical bundles. More recently, a novel optically encoded micro-sphere approach was accomplished by using different sized zinc sulfide-capped cadmium selenide nanocrystals incorporated into micro-spheres (Nature Biotech. 19, 631-635, (2001)). Given the narrow band width demonstrated by these nanocrystals, this approach significantly expands the spectral bar coding capacity in micro-spheres.

[0004] Even though the spectrally addressed micro-sphere approach does provide an advantage in terms of its simplicity over the old fashioned spatially addressable approach in microarray making, there are still needs in the art to make the manufacture of biological microarrays less difficult and less expensive and to provide nucleic acid identification systems that are accurate, less complex and less expensive.

[0005] According to the present invention, there is provided a solution to the problems of the prior art.

[0006] According to a feature of the present invention, there is provided a method of identifying nucleic acid samples comprising: providing a mircoarray including a substrate coated with a composition including a population of micro-spheres dispersed in a fluid containing a gelling agent or a precursor to a gelling agent and immobilized at random positions on the substrate, at least one sub-population of said population micro-spheres containing an optical barcode generated from at least one colorant associated with the micro-spheres and including a nucleic acid probe sequence; contacting said array with a target nucleic acid sequence; and detecting the color barcode of said sub-population of micro-spheres due to the interaction of said probe nucleic acid sequence and said target nucleic acid sequence.

[0007] The invention has the following advantages.

[0008] 1. Analysis of nucleic acids is simpler, less expensive, and more cost effective.

[0009] b 2. Nucleic acid identification systems are accurate, less complex, and less expensive.

[0010] FIG. 1 is a diagram of a coating device.

[0011] FIGS. 2A and 2B show micrographs, obtained under high magnification, of coatings with (FIG. 2A) and without (FIG. 2B) a gelling agent.

[0012] FIGS. 3A and 3B show micrographs, obtained under low magnification, of coatings with (FIG. 3A) and without (FIG. 3B) a gelling agent.

[0013] FIGS. 4A and 4B show application of Poisson statistics for a random distribution to coating prepared with a gelling agent.

[0014] FIGS. 5A and 5B show application of Poisson statistics for a random distribution to coating prepared without a gelling agent.

[0015] FIG. 6 is a schematic diagram illustrating the gelatin coated microsphere based nucleic acid detection system comprising instrumentation and appropriate control units.

[0016] FIG. 7 is a workflow chart showing the process of how to use the microarray system to analyze an unknown nucleic acid sample.

[0017] FIGS. 8A and 8B show the microarray images captured respectively under white light illumination and in a dark field without any illumination after exposure to a solution containing three target nucleic acid sequences.

[0018] FIGS. 9A and 9B show the microarray images captured respectively under white light illumination and in a dark field without any illumination after exposure to a solution containing one target nucleic acid sequence.

[0019] FIGS. 9, 10A, 10B, and C schematically show the optical design principle respectively used to analyze coated color beads signals in adsorption, fluorescence and chemiluminescence detection systems.

[0020] As used herein, the term sol-to-gel transition or gelation means a process by which fluid solutions or suspensions of particles form continuous three-dimensional networks that exhibit no steady state flow. This can occur in polymers by polymerization in the presence of polyfunctional monomers, by covalent cross-linking of a dissolved polymer that possesses reactive side chains and by secondary bonding, for example, hydrogen bonding, between polymer molecules in solution. Polymers such as gelatin exhibit thermal gelation that is of the latter type. The process of gelation or setting is characterized by a discontinuous rise in viscosity. (See, P. I. Rose, The Theory of the Photographic Process, 4^th Edition, T. H. James ed. pages 51 to 67).

[0021] As used herein, the term random distribution means a spatial distribution of elements showing no preference or bias. Randomness can be measured in terms of compliance with that which is expected by a Poisson distribution.

[0022] The present invention teaches a composition and a method for making a random array of micro-spheres, also referred to as beads, on a substrate. The distribution or pattern of the micro-spheres on the substrate is entirely random and the micro-spheres are not attracted or held to sites that are pre-marked or predetermined on the substrate as in other methods previously disclosed. In the present invention, the micro-spheres are immobilized randomly when the gelling agent in which they are carried undergoes a sol-to-gel transition.

[0023] The invention discloses a polymeric latex bead based random microarray system for nucleic acid analysis with each bead in the array having a distinct signature that would distinguish the bead. Such a signature may be based on color, shape or size of the bead. For signatures based on color, the color may be derived from mixing three dyes representing the primary colors R, G, B to create thousands of distinguishable beads with distinct color addresses (unique RGB values, e.g. R0, G204, B153). One color, two colors, four or more colors may also be used. Colors other than RGB may also be used, such as cyan, magenta, and yellow. The beads can be made with sites on their surface that are active, meaning that at such sites physical or chemical interaction can occur between the bead and other molecules or compounds. Such compounds may be organic or inorganic. Usually, the molecule or compound is organicnucleic acid, protein or fragments thereof, are examples. To the surface of each color coded bead may be attached a pre-synthesized oligonucleotide, a monoclonal antibody, or any other biological agents. Therefore, each color address can correspond to a specific bioactive probe. These beads may be mixed in equal amounts, and the random microarray fabricated by coating the mixed beads in a single or multilayer format.

[0024] Although micro-spheres or particles having a substantially curvilinear shape are preferred because of ease of preparation, particles of other shape such as ellipsoidal or cubic particles may also be employed. Suitable methods for preparing the particles are emulsion polymerization as described in Emulsion Polymerization by I. Piirma, Academic Press, New York (1982) or by limited coalescence as described by T. H. Whitesides and D. S. Ross in J. Colloid Interface Science, vol. 169, pages 48-59, (1985). The particular polymer employed to make the particles or micro-spheres is a water immiscible synthetic polymer that may be colored. The preferred polymer is any amorphous water immiscible polymer. Examples of polymer types that are useful are polystyrene, poly(methyl methacrylate) or poly(butyl acrylate). Copolymers such as a copolymer of styrene and butyl acrylate may also be used. Polystyrene polymers are conveniently used. The formed microsphere is colored using an insoluble colorant that is a pigment or dye that is not dissolved during coating or subsequent treatment. Suitable dyes may be oil-soluble in nature. It is preferred that the dyes are non-fluorescent when incorporated in the micro-spheres.

[0025] The micro-spheres are desirably formed to have a mean diameter in the range of 1 to 50 microns; more preferably in the range of 3 to 30 microns and most preferably in the range of 5 to 20 microns. It is preferred that the concentration of micro-spheres in the coating is in the range of 100 to a million per cm , more preferably 1000 to 200,000 per cm² and most preferably 10,000 to 100,000 per cm.

[0026] The attachment of bioactive agents to the surface of chemically functionalized micro-spheres can be performed according to the published procedures in the art (Bangs Laboratories, Inc, Technote 205). Some commonly used chemical functional groups include, but not limited to, carboxyl, amino, hydroxyl, hydrazide, amide, chloromethyl, epoxy, aldehyde, etc. Examples of bioactive agents include, but are not limited to, oligonucleotides, DNA and DNA fragments, PNAs, peptides, antibodies, enzymes, proteins, and synthetic molecules having biological activities.

[0027] Coating methods are broadly described by Edward Cohen and Edgar B. Gutoff in Chapter 1 of Modern Coating And Drying Technology, (Interfacial Engineering Series; v.1), (1992), VCH Publishers Inc., New York, N.Y. For a single layer format, suitable coating methods may include dip coating, rod coating, knife coating, blade coating, air knife coating, gravure coating, forward and reverse roll coating, and slot and extrusion coating.

[0028] Successful analytical imaging (resolution) of colored beads bearing other fluorescent or luminescent reporters is a complicated but feasible task. The preferred imaging system must be sufficiently flexible to execute all image capture tasks without the movement of the prepared sample array. Based upon its demonstrated merits, the full-frame capture of the array with a CCD camera is preferred. Carefully designed array illumination is critical to both color and fluorescent probe resolution. Considering the additional requirements imposed by luminescent resolution, the minimum optical system requirements are satisfied with a CCD sensor of at least 12-bits monochrome (<30 electrons read noise), 1 million pixels, 40% quantum efficiency, cooled (dark current <0.01 electron/sec), utilizing an f/3 lens and a magnification of approximately one. A sensor pixel approximately the size of a bead is most efficient; larger pixels compromise the spatial resolution of beads and smaller pixels compromise the signal resolution demanded for both color and reporter measurement. In theory, and in limited practice, the above requirements enable the analysis of approximately 50,000 randomly dispersed beads per field of capture. Clearly, software analysis of vast amounts of data must be implemented to manage the throughput of array images. The software analysis must provide a measure of bead reporting, localize and qualify a bead, calculate its color address, and possibly estimate its surface area in a timely manner.

[0029] A carefully designed illumination system includes well-managed epi-illumination (array illumination for the same side as the CCD camera). The illumination must be conditioned/controlled monochromatic light, sufficiently monochromatic to elicit appropriate color analysis and fluorescent excitation, and sufficiently intense to assure reasonably short exposure times (minutes). Illumination must be limited to an intensity that minimizes deleterious photolysis or heat and that assures only single-photon fluorescent processes, simplifying the effort to minimize undesirable fluorescent artifacts from solid supports and various biological materials essential to the arrays. The full-frame image-capture capability of the appropriately implemented CCD camera is uniquely qualified to adhere to the requirements imposed by feasible illumination systems, but must be assisted by precise optical filters to analyze light for the purpose of discerning color and fluorescence. Appropriate interference filters are readily designed upon specifying color dye and fluorescent reporter characteristics. Further, the specified CCD camera can accommodate the capture and analysis of luminescent reporters in about a 10 minute time period per field of capture. A tethered desktop computer appropriately linked to hardware can feasibly implement all analyses within a time period commensurate with the minutes required to gather all relevant array data.

[0030] While the above hardware specifications may feasibly manage the analysis of random bead arrays, the analysis is feasible only if the supporting materials are optically qualified. Optical requirements of the array support include appropriate cleanliness (artifact free, e.g., dust, debris, digs, scratches and inclusions) and physical/optical requirements that minimize fluorescent quenching, preclude fluorescent supports or colored dyes, and present a spatial noise frequency and amplitude that minimally interfere with the array analysis. In addition, the appropriate optical coupling of the array to the imaging system must accommodate the requirements of both illumination and capture.

[0031] That the entire process is feasible may be assumed on the basis that common photographic film is coated and processed to produce images that resolve millions of colors at a spatial resolution of a few micrometers. By comparison, the task of analyzing the random-bead microarray seems modest in that only a thousand colors of about 10-micrometer resolution is required. However, the bit-depth/sensitivity required, the essential delicate biochemistries and the control of fluorescence present many challenges, and much of the resolution common to photographic film will be sacrificed for the sake of meeting the developmental challenges implicit in array analysis.

[0032] One of the feasible system set up for nucleic acid analysis based on the description above was schematically shown in FIG. 6 with the inventive color beads based nucleic acid microarray and the necessary hardware components. The design principle for such a system is described in detail in Example 7. However, one skilled in the art can modified the system to produce other variations similar to what is described in this invention. As shown coated microarray A was placed on the surface of closed box H containing light source B and light deflector D. The surface I contacting A can be made of transparent glass or other plastic materials. E is an optical lens to focus a light signal from A onto full frame electronic detector F (e.g., CCD, MOS), and the signal was analyzed by computer G. In practice, the color beads barcode information can be captured by illuminating the microarray with a conditioned/controlled monochromatic light B. If the reporter signal from A is due to fluorescence, the fluorescence can be excited using selected monochromatic light from B with cover C closed. If the reporter signal from A is chemiluminescence, the microarray A will be in contact with a chemiluminescence substrate solution and the reporter signal can be captured in the dark with cover C closed.

[0033] FIG. 7 shows schematically a workflow chart on how to use the described nucleic acid analysis system. In step 1, box 100 fluorescently/chemiluminescently labeled nucleic acid sample is hybridized to the beads based random microarray. This step requires good physical contact of the coated microarray with the nucleic acid sample by either placing a layer of sample solution on the coated side of the microarray or dipping the microarray into the sample solution. The non-specific bound nucleic acid will be removed in step 2 box 110 through multiple washing of the microarray in buffer solution. In step 3, box 120, the fluorescence/chemiluminescence signals which resulted from the hybridization of the unknown nucleic acid sample with probe sequences on the surface of the coated micro-spheres are analyzed by an imaging system, as shown in FIG. 6. The recorded fluorescence/chemiluminescence image designated as IMAGE1 stored in computer G. In step 4 box 130, a bright field illumination condition is used to capture the color beads image to obtain the optical signature/barcode information of the coated microarray. This image will be designated as IMAGE2 and stored in computer H. In the final step, both IMAGE1 and IMAGE2 can be automatically analyzed using an image processing algorithm to identify the unknown nucleic acid sequences by comparing IMAGE1 with IMAGE2.

[0034] This example illustrates the influence of the polymer on gelation or sol-to-gel transition of a formulation containing colored micro-spheres or beads.

[0035] Twenty four grams of a 4% aqueous suspension of polystyrene beads prepared by emulsion polymerization and having a mean size of 9.5 micrometers was combined with 0.48 grams of poly(vinyl alcohol)(75% hydrolyzed, molecular weight 2000).

[0036] A suspension of magenta colored beads was prepared by first dissolving 0.084 grams of Dye 1 in 0.08 grams of toluene and 7.92 grams of acetone. From the above suspension of polystyrene beads containing poly(vinyl alcohol) an amount of 8.16 grams was then added slowly (drop-wise) to this solution of the dyes while stirring to prepare a suspension of colored beads. The suspension of colored beads was then filtered using a porous cotton filter, poured into a dialysis bag (12,000 to 14,000 molecular weight cut off) and washed with distilled water for one hour. After washing, the suspension of colored beads was filtered again using a porous cotton filter. The concentration of magenta colored beads in the suspension after this final step was 1.2%.

[0037] Suspensions of cyan and orange colored beads were prepared in a similar manner using Dye 2 and Dye 3 (Sudan Orange 220 from BASF corporation) respectively in place of Dye 1. The concentration of colored beads in these suspensions were 1.6% and 1.45% respectively.

[0038] Formulations for coating were prepared in the following manner:

[0039] Formulation 1 (Invention)

[0040] This was prepared by combining 4.0 grams of the suspension of cyan colored beads, 5.33 grams of the magenta colored beads and 4.41 grams of the orange colored beads with 13.91 grams of a 11.5% solution of lime processed ossein gelatin in water, 3.2 grams of coating aid (6.8% solution of Triton X 200E in water) and 49.15 grams of distilled water.

[0041] Formulation 2 (Control)

[0042] A second formulation for coating was prepared by combining the same amounts cyan, magenta and orange colored beads with 7.27 grams of poly(vinyl alcohol) (GH23 from Nippon Gohsei) and 55.79 grams of water. The amount of coating aid was the same as in formulation 1. The amount of poly(vinyl alcohol) was chosen to match the viscosity of formulation 1 at the higher temperatures (See, below).

[0043] Both samples (formulation 1 and formulation 2) were equilibrated in a 50 C. water bath for 30 minutes and then analyzed using a Rheometric Fluid Rheometer. Viscosities were determined as a function of temperature by the dynamic oscillatory technique as the samples were cooled at the rate of one degree centigrade per minute.

TABLE 1
	Viscosity (centipoise)
Temperature ( C.)	Formulation 1	Formulation 2
45	2.0	2.0
40	2.2	2.2
35	2.8	2.9
30	3.4	3.5
25	5.1	4.3
20	2107	4.6
15	100,000	6.0
10	440,000	7.9

[0044] It may be seen from the above data that at the higher temperatures (above about 30 C.), both samples behave similarly in terms of the increase in viscosity upon cooling. However, below about 25 C. formulation 1 (invention) displays a much more dramatic increase in viscosity because of gelation. Such behavior is not exhibited by formulation 2 (control). The viscosity of formulation 1 increases by several orders of magnitude as the temperature is reduced from 25 C. to 10 C. The temperature of onset of gelation is estimated to be 21.8 C.

[0045] This example illustrates the incorporation of Cyan dye (Dye 2) into polymeric beads at different levels.

[0046] A suspension of Cyan colored beads was prepared by first dissolving 0.001 grams of dye 2 in 0.05 grams toluene and 4.95 grams acetone. 2.5 grams of a 4% aqueous suspension of polystyrene beads prepared by emulsion polymerization and having a mean size of 9.5 micrometers was then added slowly (drop-wise) to this solution of the dyes while stirring to prepare a suspension of 1% dye loaded cyan dye. The suspension of colored beads was then filtered using a porous cotton filter, poured into a dialysis bag (12,000 to 14,000 molecular weight cut off) and washed with distilled water for one hour. The concentration of cyan beads in the suspension after this final step was 0.78%.

[0047] Suspensions of the other 5 levels of cyan dye incorporated into the beads were prepared in a similar manner using 0.002 grams dye 2, 0.006 grams dye 2, 0.007 grams dye 2, 0.009 grams dye 2, and 0.01 grams dye 2.

[0048] For preparing the last two samples 2.5 grams of a 4% aqueous suspension of polystyrene beads prepared by emulsion polymerization and having a mean size of 9.5 micrometers was combined with 0.48 grams of poly(vinyl alcohol)(75% hydrolyzed, molecular weight 2000). This was used in place of the polystyrene beads alone.

[0049] This example illustrates the incorporation of other three dye 3 into polymeric beads.

[0050] A 4% aqueous suspension of 2.5 grams polystyrene beads prepared by emulsion polymerization and having a mean size of 9.5 micrometers was combined with 0.48 grams of poly(vinyl alcohol)(75% hydrolyzed, molecular weight 2000).

[0051] A suspension of Magenta colored beads was prepared by first dissolving 0.01 grams of dye 1 in 0.05 grams of toluene and 4.95 grams acetone. From the above suspension of polystyrene beads containing poly(vinyl alcohol) an amount of 5.00 grams was then added slowly (drop-wise) to this solution of the dyes while stirring to prepare a suspension of 5% dye loaded Magenta dye. The suspension of colored beads was then filtered using a porous cotton filter, poured into a dialysis bag (12,000 to 14,000 molecular weight cut off) and washed with distilled water for one hour. The concentration of cyan beads in the suspension after this final step was 2.59%.

[0052] Suspensions of the other 4 colors of dye incorporated into the beads were prepared in a similar manner using 0.05 grams of each of the following dyes: dye 2 (114FN-D89), dye 3 (BASF Sudan Orange 220), dye 4 (MM2500FAN), and dye 5 (BASF Yellow 075). Plain polystyrene beads were also used at 4% for the color white.

[0053] This example illustrates the influence of the gelling agent on the uniformity of a coating of colored micro-spheres.

[0054] Formulation 1 and formulation 2 from Example 1 were coated on a 0.18 mm thick substrate of polyethyleneterephthalate (PET) containing 6% by weight titanium dioxide using the coating device shown in FIG. 1. The formulations were introduced through a slot coating die 2, driven by a motor M 4, at a temperature of 45 C., on to a 12.7 cm wide web 6 moving at the rate of 3.7 m/min. The flow rate was adjusted to provide a level of 0.043 g/m² for each of the cyan, magenta and orange colored beads. The coatings were chill-set in a 2.4 m long chilling section 8 that was maintained at a temperature of 4 C. and 56.6% RH and then passed through a conditioning chamber 10 before being dried in a first 12 and then a second 14 drying section that were 9.8 m and 11.6 m in length respectively. The first drying section 12 was maintained at a temperature of 21 C. and 33.2% RH and the second 14 was maintained at a temperature of 37.8 C. and 18.6% RH.

[0055] At the end of this process it was found that an extremely uniform coating was obtained with formulation 1 (invention) whereas there were visible streaks in the case of formulation 2 (control). Examination of the coatings by optical microscopy at high magnification (FIG. 2) showed that in the invention, the beads were evenly spread over the surface (FIG. 2A) giving a uniform density of about 50,000 beads per cm² but the control at lower magnification showed that the macroscopic streaks were composed of lateral aggregates (FIG. 2B).

[0056] More detailed analysis showed that the distribution of beads in the coating of formulation 1 (invention) conformed to Poisson statistics which is characteristic of a random distribution (FIG. 4) whereas this was clearly not the case in the coating of formulation 2 (control) (FIG. 5).

[0057] Referring now to FIGS. 4 and 5 where coated colored beads are shown in selected microscope images, FIG. 4A corresponds to images of the invention (with gelling agent) and FIG. 5A corresponds to images of a control (without gelling agent). The images of colored beads are reduced to simple gray-scale and contrasted for clarity. To measure the spatial dispersion of the beads, they are analytically compared to the random dispersion expected if there is no spatial preference for a bead on the coating plane. The analysis is implemented by counting the number of beads that fall within the confines of each field of a grid placed on the image. A random dispersion of beads will conform to a Poisson distribution of numbers of beads among fields: P_n(x)xⁿe^x/n, where x is the average number of beads per field over all fields and n is the number of fields containing n beads. A perfect adherence to random can be expected only if an infinite number of beads are counted; however, counting several arbitrary sets of hundreds of beads can clearly distinguish a clustered or structured dispersion from a nearly-random dispersion of beads. An example analysis is shown by the graph in FIG. 4B (corresponding to the invention image in FIG. 4A) and the graph in FIG. 5B (corresponding to the control image in FIG. 5A). Clear bars indicate the number of fields expected to contain n beads for a Poisson distribution of the number of beads and fields; hatched bars show the expected error from counting; solid bars show the actual bead counts. As shown in the graphs, the extent to which the bead dispersion corresponding to the invention adheres to a random dispersion (a Poisson distribution) is well within the expected error. The control dispersion clearly departs from a random distribution of beads.

[0058] It is therefore clear that the method of the invention produces the desired well-resolved random distribution of micro-spheres.

[0059] This example illustrates the attachment of pre-synthesized single strand oligonucleotide probe to the surface of dye incorporated beads.

[0060] Three DNA oligonucleotide probe sequences and their complementary target sequences were used in this example as shown in Table 1.

Dyed Microsphere	Probe Sequence	Complementary Target Sequence
Dye 1	5-amino modification-	5-biotin modification-
	AGGTGAACGTGGATGAAGTT-3	AACTTCATCCACGTTCACCT-3
Dye 2	5-amino modification-	5-biotin modification-
	TGGGTTCTCTTGGCTGTTACTG-3	CAGTAACAGCCAAGAGAACCCA-3
Dye 3	5-amino modification-	5-biotin modification-
	AAGCTGTTAGCCGAGTAGCG-3	CGCTACTCGGCTAACAGCTT-3

[0061] One hundred microliters of Dye incorporated beads (4% w/v) was rinsed three times in acetate buffer (0.01 M, pH 5.0), and combined with one hundred microliters of 20 mM 2-(4-Dimethylcarbomoyl-pyridino)-ethane-1-sulfonate and ten percent of polyethyleneimine. The mixture was agitated at room temperature for one hour and rinsed three times with sodium boric buffer (0.05 M, pH 8.3). The beads were re-suspended in sodium boric buffer.

[0062] An oligonucleotide DNA probe with 5-amino-C6 modification was dissolved in one hundred microliters of sodium boric buffer to a final concentration of 40 nmol. A 20 microliters of cyanuric chloride in acetonitril was added to the DNA probe solution and the total volume was brought up to 250 microliter using sodium boric buffer. The solution was agitated at room temperature for one hour and then dialyzed against one liter of boric buffer at room temperature for three hours.

[0063] A 100 microliters of the dialyzed DNA solution was mixed with 200 microliters of beads suspension. The mixture was agitated at room temperature for one hour and rinsed three times with sodium phosphate buffer (0.01 M, pH 7.0). The modified beads were coated onto transparent plastic support according to formulation 1 as described in Example 2.

[0064] This example illustrates the hybridization and detection of target nucleic acid sequences to the gelatin coated microsphere based microarray.

[0065] An oligonucleotide DNA with 5-biotin labeling, which has complementary sequence to the DNA probe, was dissolved in a hybridization solution containing 0.9 M NaCl, 0.06 M NaH₂PO₄, 0.006 M EDTA, and 0.1% SDS, pH 7.6 (6 XSSPE-SDS) to a final concentration of 1M. The bead coated microarray was hybridized in the hybridization solution starting at 68 C. and slowly cooled down to room temperature. Following hybridization, the microarray was washed in 0.5 XSSPE-SDS for 15 minutes three times. The microarray was incubated in a solution containing streptavidin-horseradish peroxidase conjugate in a 0.01 M phosphate buffer 0.1 M NaCl pH 7.0 for 1 hour at room temperature. Following the incubation, the microarray was rinsed three times with the incubation buffer.

[0066] The hybridization completed microarray was imaged with white light illumination using Olympus BH-2 microscope (Diagnostic Instruments, Inc. SPOT camera, CCD resolution of 13151033 pixels) to obtain the color beads barcode signature information, followed by a dark field chemiluminescence image capture by placing a thin layer of SuperSignal ELISA chemiluminescence substrate solution (purchased from PIERCE ENDOGEN) over the microarray. FIG. 8A, shows a microscopy view of the bead microarray under bright field illumination where three unknown probe sequences are detected in three bead sub-populations.

[0067] FIG. 8B shows a microscopic view of the bead microarray of FIG. 8A under dark field conditions.

[0068] FIG. 9A shows a microscopic view of the bead microarray under bright field illumination where one unknown probe sequence is detected in one bead sub-population.

[0069] FIG. 9B shows a microscopic view of the bead microarray of FIG. 9A under dark field conditions.

[0070] This example illustrates the preferred design principle to image and analyze coated color beads on a flat surface.

[0071] The most accommodating approach to obtain image data from a reporting flat object is full-frame image capture. The most important physical aspects of the image capture system is the balance between efficient light collection, optical resolution and effective illumination of the object, and the balanced system must accommodate luminescent, fluorescent and absorbance modes of optical reporting. Further, the system must do so without significant movement of the object of interest and the imaging data must support the necessary measures relating to subsequent analysis. The design solutions to many of these requirements are well known to practitioners of the art; however, the system design to accommodate all requirements into a single practical solution requires invention.

[0072] A rational approach to design is to adhere to the most important dictates of the modes of imaging/detection: luminescent, fluorescent and absorbance. The design principles are illustrated in FIGS. 10A, 10B and 10C.

[0073] FIG. 10A shows a lens 210 and object plane 212. FIG. 10B shows lens 210, object plane 212, and filter 214. FIG. 10C shows optical platen 216, aqueous layer 218, and object 220 on a mirror support.

[0074] The chief requirement of luminescent imaging/detection is efficient light detection, influenced by both collection and photometric integration. Efficient light collection requires a lens of large numerical aperture whose diameter is approximately the size of the field of view and positioned at a working distance that is close to the object. Practical solutions and cost lead to the solution illustrated in FIG. 10A, wherein an optical system specification of a lens 210 approximately f2 (lens diameter about half the working distance between lens 2210 and object plane 212) is represented. Photometric integration requires a well-implemented sensor having sufficient dynamic range to generate a signal/noise ratio appropriate to support analysis. In practice, an integrated, cooled CCD sensor of 12-bits resolution and a quantum efficiency of >40% that produces a signal/noise ratio >3000 suffices for contemporary chemiluminescent image capture and readily enables photometric analysis with a linear dynamic range of measure. An important practical limitation imposed by a lens with high numerical aperture is a small depth of focus that requires controlled flatness of the object plane. The plane of the object is consequently restricted in this system design by an optical platen (see FIG. 10C). Further, the reporting surface of the object is necessarily wet which offers the opportunity to optically couple the sample to the optical platen. Optically coupling the object to the system is essential to the intrascenic dynamic range of the measure (minimizing haze and other optical artifacts that detract from measuring bright and dim features within the same field of view).

[0075] The chief requirements of fluorescent imaging/detection are both efficient light detection and appropriate management of illumination. The design of efficient light detection discussed above (including optical coupling) delivers all of the essential capability to fluorescent imaging. Illumination management requires that excitation and emission light be discriminated. Design considerations for conditioning the wavelength of excitation light is discussed below (addressing absorption). Implicit to appropriate illumination management is the elimination of stray light, particularly that which emanates from materials that may fluoresce. Central to the design solution of this system is the minimization of excitation light from the productive path of emitted light (the acceptance cone of the lens system), as shown in FIG. 10A. This method of exciting fluorescence is traditionally recognized by those familiar with the art to be most effective; the method is generally called epi-illumination. Ideal epi-illumination of a flat surface (object plane) would involve illumination with parallel light at a 45 angle of incidence (shown as I in FIG. 10A). In practice, an angle of incidence substantially less than 45 increases the possibility that excitation light from a normal reflection will enter the acceptance cone of the lens; sources of normal reflection include stray light from illumination and features in the field of view that deviate from the object plane. An example of a design using a small angle of incidence is that which uses ring-light illumination. Another constraint is that the angle of incidence must be less than about 60, else an increasing amount of excitation light will not interact with the sample due to the total reflection caused by the critical angle of the platen (FIG. 10C). Any practical system of illumination will include light that is not parallel. It is practical, indeed usual, to illuminate with light that is divergent as illustrated in FIG. 10B, where the illumination is shown as the finer lines entering from both sides of the object, a common method of epi-illumination used to enhance illumination uniformity. The illustration depicts slightly divergent light (about 18 degrees, or approximately f11), much less divergent than many commonly used methods such fiber illumination. The epi-illumination shown in FIG. 10B is preferred, since the extreme angles of incidence (and reflectance shown as bold lines) minimize the possibility of stray light entering the acceptance cone of the lens or critically reflecting from a planar surface.

[0076] A further important system design solution is embodied in FIG. 10C. The platen composition is a controlled optical surface, and as such may be constructed to minimize haze, autofluorescence and reflectance. Note that in this system design, the platen is the only material with which the excitation light has the opportunity to interact. Preferred composition of the platen is optical acrylic (UV transparent), whose sensor/illumination side is coated with an anti-reflective layer, and whose sample/object side is coated with an anti-abrasion layer (hard-coat). Other compositions (such as fused silica) are acceptable, but will manifest some performance decrement (transmit less light). As described above, the aqueous layer (ubiquitous to biological/biochemical systems) optically couples the sample surface to the platen and the object (the bead monolayer) is coated onto a flexible support having a mirror surface. The vast majority of the excitation light, which does not interact with the object beads, is simply reflected out of the system, away from the lens.

[0077] To summarize, in this preferred epi-illumination system for fluorescent excitation, the excitation light interacts only with a controlled optical component, a thin aqueous solution and a target bead before the light has a significant probability of entering the acceptance cone of the lens. Consequently, the lens system need only discriminate excitation from emitted light that emanates from the target bead. The discrimination of excitation from emitted light is commonly practiced in fluorescent imaging/detection, commonly using an interference filter (see filter 214-FIG. 10B). Such filters can be very costly and the capability of the filter to efficiently discriminate excitation and emitted light is often the chief limitation of detection sensitivity. In this preferred system design, the requirement of the filter to discriminate is reduced approximately 1000-fold compared to many fluorescent detection systems.

[0078] The chief requirements of absorption imaging/detection (generally called colorimetry) involve spectral resolution and photometry. The photometric precision of the system described above is more than 10-fold better than the basic system requirements for resolving thousands of colors. The application requirement is similar to resolving/deconvoluting R, G, B and combinations/levels thereof. The fundamental method of color resolution involves measurement of the relative absorption characteristics of an object with differing wavelengths of light; the relative differences are tabulated to functionally define categories of dye combinations and levels. The dyes chosen must not interfere with other reporting optical elements, such as fluorochromes. One way to minimize anticipated interference is to choose color dyes whose absorbance spectrum is higher than any excitation/emission spectrum of the reporting fluorochromes.

[0079] Given a sufficient photometric precision, the number of colors that may be resolved (functionally defined) will depend upon the management of illumination and detection of light. Two fundamentally different methods of light management for the purpose of color resolution are either spectrally resolving white light before or after the light interacts with an object. The above system design requires that both methods be invoked for the sake of fluorescent imaging. The preferred design solution that is suitable for both absorption and fluorescent modes of imaging detection is to manage the spectral resolution of the illumination. Therefore, the illumination is managed by a light source and monochrometer, having sufficient power and resolution to accommodate both modes of imaging. The solution is cost effective because the multi-filter design and implementation that would accommodate the excitation of multiple fluorescent probes and absorption of multiple colored dyes would easily exceed the cost of monochromator. Since the spectral excitation/absorption of the organic fluorochromes and dyes are relatively broad, a spectral resolution of only about 20 nm will suffice and the delivered illumination energy will be correspondingly higher than that associated with a high resolution monochromater (typically <2 nm). That the delivered energy will suffice for fluorescent imaging is assured by the sensitivity of the camera which is designed for luminescent imaging.

[0080] The invention has been described in detail with particular reference to certain preferred embodiments thereof, but it will be understood that variations and modifications can be effected within the spirit and scope of the invention.

PARTS LIST
2	coating die
4	motor
6	web
8	chilling section
10	conditioning chamber
12, 14	drying section
100	fluorescently/chemiluminescently labeled nucleic acid
	sample hybridized to the beads based random microarray
110	non-specific bound nucleic acid removed through multiple
	washing
120	fluorescently/chemiluminescently signals analyzed
130	bright field illumination condition used to capture color beads
	image to obtain optical signature/barcode information of
	coated microarray
140	IMAGE1 and IMAGE2 automatically analyzed using
	image processing algorithm to identify unknown nucleic
	acid sequences
210	lens
212	object plane
214	filter
216	optical platen
218	aqueous layer
220	object