Global proteomic screening of random bead arrays using mass spectrometry imaging
11846634 · 2023-12-19
Assignee
Inventors
Cpc classification
G01N33/6845
PHYSICS
International classification
Abstract
Methods for proteomic screening on random protein-bead arrays by mass spec is described. Photocleavable mass tags are utilized to code a protein library (bait molecules) displayed on beads randomly arrayed in an array substrate. A library of probes (prey) can be mixed with the protein-bead array to query the array. Because mass spec can detect multiple mass tags, it is possible to rapidly identify all of the interactions resulting from this mixing.
Claims
1. A method of mass spectrometric imaging of mass-tags, comprising: a. providing compounds comprising mass-tags linked to a photocleavable linker attached to a probe, said probe bound to a surface, b. illuminating said mass-tags with light so as to photocleave at least a portion of said mass-tags from said surface, wherein said photocleaved mass-tags have a terminal amine group, c. applying a matrix to said surface after step b, and d. performing matrix-assisted laser desorption ionization mass spectrometric imaging of said photocleaved mass-tags on said surface after step c.
2. The method of claim 1, wherein said matrix compound is selected from the group consisting of alpha-cyano-4-hydroxycinnamic acid (CHCA) and 3,5-Dimethoxy-4-hydroxycinnamic acid (sinapinic acid).
3. The method of claim 1, wherein said probe is selected from the group consisting of proteins, antibodies and nucleic acids.
4. A method of mass spectrometric imaging of mass-tags, comprising: a. providing compounds comprising mass-tags comprising a linker, at least a portion of said linker being photocleavable, said linker attached to a probe, said probe attached to a surface, b. illuminating said mass-tags with light so as to photocleave at least a portion of said mass-tags from said surface, wherein said photocleaved mass-tags have a terminal amine group, c. applying a matrix to said surface after step b, and d. performing matrix-assisted laser desorption ionization mass spectrometric imaging of said photocleaved mass-tags on said surface after step c.
5. The method of claim 4, wherein said matrix compound is selected from the group consisting of alpha-cyano-4-hydroxycinnamic acid (CHCA) and 3,5-Dimethoxy-4-hydroxycinnamic acid (sinapinic acid).
6. The method of claim 4, wherein said probe is selected from the group consisting of proteins, antibodies and nucleic acids.
7. A method of mass spectrometric imaging of mass-tags, comprising: a. providing compounds comprising mass-tags, each of said mass-tags attached to a probe, said mass-tags comprising a photocleavable linker, said probe bound to a surface, b. illuminating said mass-tags with light so as to photocleave at least a portion of said mass-tags from said surface, wherein said photocleaved mass-tags have a terminal amine group, c. applying a matrix to said surface after step b, and d. performing matrix-assisted laser desorption ionization mass spectrometric imaging of said photocleaved mass-tags on said surface after step c.
8. A method of mass spectrometric imaging of mass-tags, comprising: a. providing compounds comprising mass-tags linked to a photocleavable linker, said mass-tags comprising a polymer, said photocleavable linker attached to a probe, said probe bound to a surface, b. illuminating said mass-tags with light so as to photocleave at least a portion of said mass-tags from said surface, wherein said photocleaved mass-tags have a terminal amine group, c. applying a matrix to said surface after step b, and d. performing matrix-assisted laser desorption ionization mass spectrometric imaging of said photocleaved mass-tags on said surface after step c.
9. The method of claim 8, wherein said polymer comprises a peptide.
10. The method of claim 8, wherein said polymer comprises an oligonucleotide.
Description
DETAILED DESCRIPTION OF THE INVENTION
(1) Example of Approach:
(2) A simplified flow diagram for one embodiment of the present invent designed for discovery of autoimmune biomarkers is shown in
(3) 1. Sample:
(4) For the discovery process, the sample consists of blood sera obtained from patients with a known autoimmune disease or no known disease (control). Auto antibodies that are typically present in such autoimmune disease patients constitute the prey molecules in this invention
(5) 2. Probe Proteomic Bead-Library with Sample:
(6) The sera are mixed with a proteome bead library prepared using recombinant or cell-free protein expression methods. The full library will encompass 15,000 unique human proteins and 20 bead replicates for each protein in the library (total 300,000 beads). These proteins constitute the bait molecules in this invention and are also potential autoantigens or biomarkers that can be utilized in diagnostic tests for autoantigen diseases. In some applications, after mixing the proteomic bead-library with sample, the beads are randomly arrayed into a pico-wall plate suitable for both fluorescent and MALDI-MS imaging.
(7) 3. Selection:
(8) Positive beads that interacted with autoantibodies present in the sera are selected for decoding. A variety of methods are provided for selection. In one preferred method, beads are first scanned for fluorescence which originates from fluorophores that are bound directly or indirectly to the autoimmune antibodies through binding agents. In one embodiment, the fluorescence label is bound directly to a secondary antibody which is specific for the class of antibodies in the sera (e.g. human IgG). In this case, the secondary antibody serves as the binding agent In all cases, autoantibodies which bind to beads will result in fluorescence emitted from the beads where prey molecules interact that is detected by conventional fluorescent scanners. In a second embodiment, the beads which are positive are identified using flow-cytometry often and physical separated from other beads (FACS). In a third method, the beads which are positive are physically separated using magnetic beads.
(9) In another embodiment, the positive beads are selected using MALDI-MS bead imaging wherein the mass-tag is attached to the prey molecule, either directly or indirectly through a binding agent. For example, the prey molecule could be an autoantibody directed at a specific autoantigen on an individual bead. In some cases the mass tag is attached via a binding agent such as a secondary antibody. In these cases, positive hits are identified by detection of the specific mass tag associated with the prey molecule.
(10) It is also to be understood that in the present invention different mass tags can be attached directed or indirectly to different prey molecules in order to distinguish which prey molecules interact with which bait molecules. For example, in the case of studying protein-protein interactions, a library of proteins can be coded by attaching different mass tags to each type of protein in the library. Once a protein in the library (prey molecule) interacts with a specific bait molecule on a specific bead, its identity can be determined because of its attachment to a specific mass tag which is different from other mass tags used to code prey molecules and specific bead types.
(11) In another example, different mass tags are used to code autoantibodies originating from different samples. In this embodiment, each sample which may containing autoantibodies are mixed with a suspension of beads and then the autoantibodies allowed to interact with the bait molecules residing on beads. Mass tags specific for each sample are then attached to autoantibodies (prey molecules) through a binding agent which in this case is a secondary antibody. The various bead suspensions exposed individually to different samples are then mixed together and an bead array formed which is measured using MALDI-MS-Imaging. The beads which have bound prey molecules can then be identified along with the sample which the prey molecules originated since each prey molecule (autoantibody) has a unique mass tag.
(12) 4. Decoding:
(13) The beads selected in the previous step as positive are decoded to determine the identity of the protein residing on the bead. This is accomplished by measuring the mass of the photocleavable mass tag present on the selected bead by MALDI-MS. A variety of other methods are also described in this invention to determine the identity of photocleavable tag including RT-PCR and DNA hybridization.
(14) 5. Validation:
(15) Using a new patient cohort, each individual biomarker determined in steps 1-4 is clinically validated by testing for its diagnostic sensitivity and specificity using an independent method such as two-epitope solid-phase T.sup.2-ELISA™ assay based on cell-free protein synthesis technology.
(16) Bead Library Construction
(17) Several methods can be used to construct the proteomic-bead library. One method termed Parallel Preparation, consists of cell-free expression in a separate reaction of each protein in the library from a specially designed plasmid containing the target gene. This is followed by binding of the expressed protein to beads and pooling of all beads in a bead-library. A second method described in US Patent Application Nos. 20090264298, 20090270278, 20090286286, 20100062451 and 20100075374, termed here as the Batch Preparation method, involves only two reactions to create the entire library and relies on advanced methods utilizing single-molecule solid-phase emulsion PCR and self-assembling cell-free expression.
(18) We described below several examples of steps used in bead library construction in the present invention:
(19) Plasmid Construction:
(20) As illustrated in
(21) Cell-Free Protein Transcription & Translation:
(22) Customized plasmid constructs as described above will be used to express the required protein for the library in a cell-free (in vitro) expression system. Until recently, relatively high costs and low yields have discouraged the use of cell-free protein expression systems in the field of proteomics. However, recent dramatic improvements in this field hold great promise for solving many of the problems that have previously limited the use of this approach. Advantages and improvements include: On-Demand Expression: Express specific proteins, on-demand, typically in <1 hr, even in eukaryotic (e.g. mammalian) systems using a single facile reaction (e.g. Promega's batch mode rabbit reticulocyte or insect cell coupled transcription/translation system). High Yield: New “continuous exchange” cell-free (CECF) expression systems capable of ˜1 mg/mL yields (e.g. Roche's Wheat Germ CECF now sold by 5 Prime, Inc.). In these systems, the expression reaction is separated from a feeding buffer by a semi-permeable membrane. The feeding buffer provides a continuous supply of small molecule precursors while absorbing (by diffusion) inhibitory byproducts. The devices (
(23) Since the proteomic-library will consist of a wide-variety of different proteins it is essential that cell-free reactions are designed to be compatible with expression of these different types of proteins. However, there is abundant evidence that such a diversity of proteins can be universally expressed if the proper system is chosen. For example, recent studies reported in Nature Methods [Goshima, Kawamura et al. (2008) Nat Methods 5: 1011-7] using wheat germ expression system has been shown to be capable of expressing at least 13,346 proteins from the proteome. The expression system was a variant of the aforementioned high yield CECF system, except that as an alternative to the 2-chambered devices, the so-called “bilayer” method was used to overlay the feeding buffer directly onto the expression reaction [Sawasaki, Hasegawa, Tsuchimochi, Kamura, Ogasawara, Kuroita and Endo (2002) FEBS Lett 514: 102-5]. Recently, synthetic nanolipoparticles (NLP) [Cappuccio, Blanchette et al. (2008) Mol Cell Proteomics 7: 2246-53; Katzen, Fletcher et al. (2008) J Proteome Res 7: 3535-42; Cappuccio, Hinz et al. (2009) Methods Mol Biol 498: 273-96], small discoidal membranes mimicking the native membrane protein environment, have been used for cell-free expression of properly folded and functional membrane proteins.
(24) In one example of the prototype library construction used in this invention, we have used the Promega rabbit reticulocyte cell-free expression system, known for its ability to produce functional, properly folded and even post-translationally modified proteins, including multi-pass membrane proteins, especially considering it is a mammalian system, and as a cell lysate, native chaperones are not removed [(Gibbs, Zouzias and Freedberg (1985) Biochim Biophys Acta 824: 247-55; Hirose, Kim, Miyazaki, Park and Murakami (1985) J Biol Chem 260: 16400-5; Vorburger, Kitten and Nigg (1989) Embo J 8: 4007-13; Pensiero, Dveksler, Cardelliohio, Jiang, Ella, Dieffenbach and Holmes (1992) J Virol 66: 4028-39; Middleton and Bulleid (1993) Biochem 1296 (Pt 2): 511-7; Popov, Tam, Li and Reithmeier (1997) J Biol Chem 272: 18325-32; Lyford and Rosenberg (1999) J Biol Chem 274: 25675-81]. Examples of two high yield systems which can be used in conjunction with library construction are: i) insect cell coupled transcription/translation systems (Promega) and ii) wheat-germ continuous flow system (Roche; now sold by 5 Prime Inc.).
(25) One-Step Capture of Protein on Bead:
(26) As illustrated in Figure, 29, in one embodiment of the present invention each expressed protein in the library is purified and attached to the bead through a simple one-step process. This is accomplished through the common HSV-epitope (C-Tag) incorporated into each protein which binds to an anti-HSV antibody present on the 34-micron agarose beads. We have measured a 75% capture efficiency from cell-free expression lysates with this antibody system [Lim and Rothschild (2008) Anal Biochem 383: 103-115]. It should also be noted that we have shown this attachment method to be extremely stable. For instance, after overnight incubation of a mixed population of protein-beads at room temperature with vigorous shaking, no significant drop in signal and no bead-to-bead cross-contamination was observed.
(27) As an alternative to the above approach, tRNA mediated engineering technology such as described in various patents issued to AmberGen Inc. [e.g. U.S. Pat. Nos. 5,643,722; 5,922,858; 6,210,941; 6,303,337; 6,306,628; 6,344,320; 6,358,689; 6,566,070; 6,596,481; 6,875,592; 6,949,341; 7,169,558; 7,252,932; 7,288,372 and 7,524,941] can be used to co-translationally incorporate biotin labels for attachment to (strept)avidin beads [Lim and Rothschild (2008) Anal Biochem 383:103-115], thereby exploiting the unparalleled affinity of this interaction (Kd=10.sup.−15, roughly 6-orders of magnitude better than an average antibody).
(28) An important feature of the present invention when agarose beads are used is that it exploits the extremely high protein binding capacity of the porous 6% cross-linked 3-dimensional matrix of the 34 micron agarose beads. According to the manufacturer's specifications (GB Life Sciences), the chemical surface activation (primary amine-reactive NHS groups) of the uncoated agarose beads is 10 moles/mL of packed bead volume, which would correspond to 50,000 pg/bead of antibody attachment capacity. However, in practice, due to steric factors, the much larger antibody cannot be loaded to such levels. The typical antibody binding capacity of such beads is ˜10 mg/mL of beads, or ˜300 pg or 2 fmoles/bead (1 mL 30 million beads). This still compares favorably to conventional 2-dimensional planar proteome microarrays such as Invitrogen's ProtoArrays®, with published densities of 1 pg/spot maximum (100 micron spots) [Zhu, Bilgin et al. (2001) Science 293: 2101-5]. Furthermore, this should easily facilitate our targeted intended loading of 10 pg (0.2 fmoles) per bead of expressed protein on average, in addition to the loading of the peptide mass-tags. Note that because an antibody is larger (150 kDa) than a typical protein (calculations on 50 kDa), 30 pg/bead (0.2 fmoles) of antibody is targeted.
(29) Another important feature of the bead-based capture is the ability to easily normalize for protein-protein variances in expression yield (see
(30) Automation of Bead-Library Fabrication:
(31) An important feature of the bead-library fabrication is the ability to fully automate the process. This ability stems from relatively few steps involved in the overall preparation of the protein-beads. In essence, because of the cell-free expression and in situ protein capture, the fabrication of the bead library is only a series of steps involving mixing of reagents and washing beads. In particular, each step can be carried out in parallel on standard 96-well plates using our in-house robotic plate handling systems. For example, transfer of the genes from the ORFeome library to the pVSV-DBST expression vectors essentially involves a single recombination step using the Gateway™ recombination system. In fact, high throughput automated Gateway™ recombination has been done with 93% efficiency without the need for bacterial plating and/or colony picking at any stage [Janney, Roby, Getbehead, Bell, Daniels and Chesnut 2009; Aguiar, LaBaer et al. (2004) Genome Res 14: 2076-82]. Similarly, expression of the protein in a high yield coupled translation/transcription reaction requires just one step which involves mixing reaction components and vector in a tube followed by incubation at a controlled temperature. The purification of the protein from the reaction system and binding to beads also involves a single-step process. Since our library encompasses ˜15,000 proteins, this will require methods that involve robotically processing approximately 150×96-well plates for each of the 5-steps (note we anticipate that the ORFeome will expand to this number from its currently available 12,000 clones). This number of plates can be easily accommodated using industry standard multi-plate stacking devices which are compatible with our robotic systems.
(32) Alternative Batch Library Production with Advanced Single-Molecule Technologies:
(33) As an alternative to the Parallel Preparation method of fabricating the proteome bead-libraries in the present invention, an alternative method described in US Patent Application Nos. 20090264298, 20090270278, 20090286286, 20100062451 and 20100075374 is potentially more efficient and cost effective. This fully multiplexed Batch Preparation method uses a single-tube reaction to first produce an entire bead sorted library of in vitro expressible DNA (BS-LIVE-DNA). For this, universal PCR primer beads (vector targeted) are emulsified in oil with the expressible ORFeome library as template such that single beads are compartmentalized in aqueous droplets (millions per reaction) containing single copies of template. The emulsion PCR reaction thus clones and amplifies single DNA molecules onto beads, and is normally used for genetic assays [Dressman, Yan, Traverso, Kinzler and Vogelstein (2003) Proc Natl Acad Sci USA 100: 8817-22]. After loading the common capture antibody onto the resultant BS-LIVE-DNA beads (via a biotin-(strept)avidin bridge), the bead population is subjected to a single-tube self-assembling cell-free expression reaction [Nord, Uhlen and Nygren (2003) J Biotechnol 106: 1-13; Ramachandran, Hainsworth et al. (2004) Science 305: 86-90] to convert BS-LIVE-DNA to a bead sorted library of in vitro expressed proteins (BS-LIVE-PRO). In this reaction, proteins are simultaneously translated and captured back onto the parent DNA beads from which they were made, by way of the capture antibody directed against a common epitope tag in each protein. Although not required per se, methods to reduce mRNA/protein escape from its parent DNA encoded bead can be employed, such as expression in an emulsion or in the pico-well plates to reduce diffusion.
(34) MALDI-MS Measurement of Individual Bead in an Array
(35) An important feature of this invention is the demonstrated ability of MALDI-MS to rapidly measure a variety of molecules residing on individual beads as part of an array of beads. In one preferred embodiment of this invention this ability allows decoding of individual beads that have been identified as displaying positive interactions between bait molecules residing on bead and prey molecules which interact with said bait molecules. In several example described in this invention, we have demonstrated that MALDI-MS is capable of rapidly decode PC-Mass-Tags on beads, including their utilization to detection the interaction of autoantibodies with autoantigens residing on individual beads. In general this ability to measure molecular interactions an individual beads using MALDI-MS, offers many advantages not limited to just mass-tag decoding but also for direct identification of proteins and other biomolecules residing on the bead surface or indirectly attached to the bead as described below.
(36) Several examples of the ability of MALDI-MS to image individual beads that were performed using an ABI 4800 Plus MALDI-TOF-TOF mass spectrometer (see Examples below) although such measurements are not limited to this particular model of MALDI-TOF instrument. Typically, scanning is done using the ABI 4800 software in the positive ion reflector mode with internal calibration using 50-200 laser pulses per sample spot, which results in measurement times of ˜0.25-1 s per bead. Image acquisition and analysis is performed using public domain software (www.maldi-msi.org; 4000 Series and BioMap software respectively).
(37) Fluorescent Selection of Positive Beads
(38) Fluorescent Scanning of Beads in Pico-Well Plate:
(39) In one embodiment of the present invent designed for biomarker discovery involving autoimmune disease, hits corresponding to proteins that interact with autoimmune antibodies in the sample serum (i.e. putative biomarkers) are selected by detection of fluorescent spots on a conventional proteomic microarray. In contrast, the present invention relies on identification of positive beads. This can be accomplished in several ways such as but not limited to fluorescent scanning of the beads residing in a pico-well plates after exposure to the test sera. The positive beads identified can then be decoded using MALDI-MS bead scanning and positive hits further confirmed by detection of mass-tags attached to the read-out antibody.
(40) Fluorescent Labeling of the Probe (Prey) Molecule:
(41) There are a multitude of fluorescent based methods commonly used to identify the interaction of a bait molecule with probe (prey) molecule. One common method is based on using a fluorescently-labeled antibody directed against the probe molecule which is applied either before or after the bait-prey interaction occurs. In the case of autoimmune antibodies residing in patient sera for example which are themselves the prey molecules, a fluorescently labeled secondary antibody is used which is directed against the autoimmune antibody. Alternatively, a fluorescently labeled streptavidin molecule can be used directed at a biotinylated antibody which is in turn directed against the probe molecule such as an autoimmune antibody. In all of these cases the antibody serves as a binding agent coupling the prey molecule with a fluorescent molecule. The binding agent might also comprise a group residing on the prey molecule which facilitates a covalent linkage with the fluorescent label such as through an activated or reactive chemical group. For example, the epsilon amino group of lysines residing on proteins can be chemically labelled with appropriate fluorophores which are commercially available. It is to be understood that labeling of the probe molecules is not limited to a single fluorescent molecule. In some embodiments it is advantageous to use different fluorescent labels for different prey molecules that might be coupled to the prey molecule through different binding agents such as antibodies or reactive chemical groups. It is also to be understody that probe molecules can also comprise a coding agent such as a photocleavable mass tag.
(42) As listed below, there are several key advantages to this approach over conventional microarrays including:
(43) More Reliable Hit-Identification Due to Increased Replicates:
(44) The ability to scan in a few minutes a large bead library of beads such as Y-million beads arrayed on the pico-well substrate provides more reliable hit-identification than conventional microarrays. In comparison to the typical <20,000 spots with 2-replicates used on the conventional proteomic microarrays, the current invention can scan significantly more replicates (˜20) providing better statistics for identifying weak protein-antibody interactions.
(45) False-Positive Rejection:
(46) Since in one embodiment of the invention the probe antibody is doubly labeled with not only a fluorescent label but a photocleavable mass-tag designed for MALDI-MS bead imaging read-out, this assures that any false positive beads selected during the sort step will be rejected during decoding. See FIG. 7 (Experimental Examples) for one embodiment of this redundant design.
(47) Multi-Dimensional Hit Identification:
(48) The use of multiple fluorescent labels with different wavelengths of excitation provides more robust ability to separate different positive hits. For example, the probe antibody can be coded with different fluorescent dyes in order to differentially label different samples scanned. Such a capability is not possible using conventional microarrays since the sera is applied directly to chip instead of to beads in separate reaction vessels prior to application to pico-well slide.
(49) Non-Dry Conditions:
(50) Conventional proteome microarray fabrication involves several steps which subject the proteins to non-physiological conditions. For example, printing followed by drying can lead to protein denaturation. Microarrays are also stored and shipped dry for long periods of time which can lead to further protein alterations. In addition, antibody interactions occur on two-dimensional protein spots printed on an array surface which could alter the ability of the antibody to freely interact with all proteins in the spot In contrast, the beads used in the present invention can be kept fully hydrated so the proteins are never exposed to drying. Furthermore, the antibody-antigen interaction and hence selection of positive hits occurs in a fully controllable aqueous environment selected to promote native protein conformation.
(51) In-Line Quantification of Protein Per Bead
(52) Variability of protein content on each bead can be easily accounted for in hit selection by using AmberGen's proprietary FluoroTag™ technology (
(53) Fluorescence-MALDI Image Synchronization for Bead Selection
(54) An important feature of one embodiment of this invention is the ability to synchronize the fluorescent scan with the MALDI-MS measurements on the bead array thereby reducing the number of beads which need to be scanned by MALDI-MS. In this embodiment, beads which exhibit an interaction between the bait molecule residing on the bead and prey molecule is measured first using conventional fluorescent scanning and detection methods to determine if a fluorescently labeled prey molecule has interacted with the bait molecule on a particular bead in the array. The position of the positive hits detected in this scan are then used to direct the MALDI-MS spectrometer to the position of these positive beads in order to determine the mass of the mass tags residing on individual bead and thus the identity of the protein. It is to be understood that this approach is not limited to fluorescent measurements but other detectable properties such absorption, luminescence, Raman, infrared could be used to detect positive interactions of bait and prey molecules and the positions determined used to guide measurements of the bead array for subsequent MALDI-MS measurements and decoding.
(55) Typically, fluorescent detection is performed with a conventional fluorescent scanner which can assay hundreds of thousands of beads in a single scan in a few minutes at high resolution (typically 3-5 microns). The arraying of the beads on a substrate with uniform spaced wells of proper depth allows only one bead per well such as for the case of the pico-well plates designed for 34 micron diameter beads described above thus preventing the overlap of beads and subsequent ambiguous measurements. Once the positive-hits are identified, the mass spectrometer can be directed to measure the mass tags residing on those particular beads identified by fluorescence, thereby avoiding the measurement of mass tags on every bead residing in the array.
(56) As described, elsewhere in this invention prey molecules can also be preferentially attached to mass-tags, either directly or through binding agents, thus providing an separate means to detect and confirm positive hits (i.e. interaction of prey and bait molecules on a bead) and also to determine the identify of one or more prey molecules which might exist in a mixture or solution and may interact with the bait molecule residing on an individual bead.
(57) One embodiment of this invention utilizes beads which are both mass-tagged and fluorescently labeled to aid in guiding the MALDI-MASS spectrometer to measure more accurately to the position of beads with positive hits identified by fluorescence. In this embodiment, the x-y position of the so-called synchronization beads is first determined in the fluorescent scan. The synchronization beads comprise a fluorescent label and a mass tag. The fluorescent label used for synchronization beads has properties different from those fluorescent labels used to label prey molecules (e.g. a distinguishable emission spectrum). In addition, the mass tag(s) used for synchronization beads has a different mass then other mass tags used to identify beads with bait molecules or prey molecules. Once the x-y coordinates of these synchronization beads is determined from the fluorescent image of the bead array, the information can be used to guide the MALDI-MS instrument to find the same beads in the MALDI-MS generated image.
(58) Confirmation that the MALDI-MS system is measuring the correct bead in the array is provided by measuring the corresponding mass tag residing on the synchronization beads. Importantly, once the synchronization beads are identified in both the fluorescent image and MALDI image, the position of positive beads identified by the fluorescent scan can be more accurately located in the MALDI-MS scan and mass tags measured to determine the identity of the bait and prey molecules residing an the bead. Increased accuracy for identification of positive hits can be accomplished by increasing the number of synchronization beads randomly incorporated into the array, thus providing more local coordinate information to determine precise location of nearby positive-hits. This method is especially useful for MALDI-MS systems which do not intrinsically have high resolution scanning capability (e.g. MALDI laser beam diameter which exceeds the size of the bead diameter) or in cases where the positional accuracy of the wells incorporated into the pico-well plates vary in position compared to exact 2-D periodic lattice.
(59) It is to be understood that in addition to using fluorescently labeled synchronization beads described above other detectable properties can be used including absorbance in the visible, infrared or UV, Raman scattering and even magnetic properties. In each case, a scan of this property provides the MALDI-MS system with coordinates that aid in scanning of the beads. As one example, of how this feature might be implemented on commercially available instruments, the Bruker Ultraflextreme MALDI-MS spectrometer is equipped with software that allows one to synchronize features obtained by scanning the sample on an external system based on fluorescence or absorption properties with a visual image obtained using its high resolution camera incorporated into the machine.
(60) In one preferred embodiment, the detection of fluorescent or other properties such a light absorption of the beads is measured directly on the MALDI-MS system in order to identify positive hits (positive interactions between prey and bait) and then used this information to determine which beads are measured directly by MALDIMS.
(61) For example, the high resolution capability of the camera and imaging system which is incorporated into the Bruker Ultraflextreme MALDI-MS instrument allows detection of beads of less than 20 microns that have been colored with a light absorbing chromogenic dye. Since a variety of chromogenic based agents have been developed to label positive interactions between bait and prey molecules such as the use of antibodies conjugated to horse radish peroxidase (HRP). A similar labeling method can be used to directly label positive bait and prey interactions on individual beads which can be detected directly in the MALDI-MS instrument such as the Bruker Ultraflextreme. Those skilled in the art of MALDI-MS instrumentation will also recognize it is possible to incorporate fluorescent detection so that fluorescent labeling methods can be utilized to identify bait-prey interactions on individual beads.
(62) In addition to pico-wells formed from fiber optic bundles, a photolithographic method of well fabrication can be utilized to increase accuracy of the position of each well. In addition, the ability of fluorescent scanners to detect multiple wavelengths will enable marker beads to be utilized that will allow more accurate registration of the fluorescence and MALDI images to be made.
(63) Additional Methods of Selecting Positive Bait-Prey Interactions on Beads for MALDI-MS Decoding.
(64) An alternative (or complement) to direct selection of positive hits (beads) by fluorescence imaging prior to MALDI-MS decoding is the use of physical methods to separate positive beads from negative beads. One such approach is the use of fluorescence assisted cell-sorting (FACS). A second method is based on a magnetic bead sorting techniques.
(65) Similar to conventional magnetic cell-separation techniques, the protein-beads of Bead-GPS™ can be pre-isolated by small (1 micron) magnetic particles prior to fluorescence imaging and/or MALDI-MS decoding. Moreover, magnetic particle manipulation is particularly amenable to automation, for example, as achieved by Bio-Rad's (Hercules, CA) BioPlex multiplex immunoassay system.
(66) Another embodiment of this invention involves the isolation of the 34 micron agarose beads using fluorescence activated cell sorting (FACS). Importantly, this method is high throughput (can process millions of beads in a few minutes) and has the ability for greater reproducibility and specificity than the magnetic method, since beads can be analyzed by multiple parameters on a bead-by-bead basis. As shown in
(67) Although selection either by imaging or physical separation of positive hits (beads) prior to MALDI-MS decoding is the preferred method, it should be noted that it is not required. As demonstrated, mass-tags can be used alone for both bead identification and autoantibody readout (see examples). In this case, since hits are not pro-imaged by fluorescence, the entire library is imaged by MALDI-MS in the pico-well plates. Importantly, the newer generation of faster scanning MALDI-MS instruments can do this In a relatively short amount of time.
(68) Mass-Tag for Bead and Prey Decoding
(69) Basic Concept:
(70) Once beads have been sorted or selected for positive “hits” on the basis of fluorescence scanning as described above, the next step in one preferred embodiment of this invention is decoding the beads in order to identify the bait molecules which are bound to them. A similar process is also used for identifying prey molecules which interact with the bait molecules (see below). It is also to be understood that the use of fluorescence scanning in some embodiments is not necessary in cases where each bead in the bead array is canned individually. In this case, positive hits can be identified using decoding methods described in this invention.
(71) One preferred embodiment of the invention which entails a method of detecting the interaction of prey molecules with bait molecules, comprises: a) providing a mixture comprising first and second beads, said first bead comprising a first mass tag and a first bait molecule, said second bead comprising a second mass tag and a second bait molecule, wherein said first and second bait molecules and said first and second mass tags are different; b) making an array with said beads; c) contacting said first and second bait molecules with a solution comprising a prey molecule, wherein said prey molecule comprises a mass tag; and d) subjecting said array to MALDI mass spec analysis under conditions wherein binding of said prey molecules to a bait molecule is detected.
(72) A second embodiment of the invention involves reversing steps b) and c) listed above so that the contacting of said and first and second bait molecules with a solution comprising a prey molecule occurs before making an array of said beads.
(73) It is to be understood that the bait and prey molecules can consist of a large variety of different biomolecules or biologically active molecules such as proteins, antibodies or potential drug compounds. For example, bait or prey molecules includes but are not limited to proteins, polypeptides, nucleic acid molecules, lipids, carbohydrates, biologically active drug compounds, hormones, antigens, antibodies and combinations of these molecules. The aforementioned bait and/or prey molecules can also be labeled using standard fluorescent labeling reagents comprising one or more fluorophores (see examples) in order to identify positive bait/prey interactions and facilitate fluorescent image synchronization with MALDI-MS imaging as described elsewhere in this invention.
(74) In one example, bait molecules consist of various proteins selected from a protein library such as can be expressed as described in this invention using the commercially available 12,000-member Open Reading Frame (ORF) template library (ORFeome), whereas prey molecules consist of autoantibodies present in a patient's blood that are associate with autoantigens underlying an autoimmune disease such as primary biliary cirrhosis or lupus. Similarly, antigens might be tumor specific (tumor associated antigens (TAAs) related to a particular cancer tumor and antibodies freely circulating in blood may be formed in response to these TAAs and used as biomarkers for detection of the cancer or to predict the course of the cancer (prognostic). Detecting and identifying the interaction of specific antigens with specific antibodies using the present invention provides critical information in designing diagnostic tests, prognostic tests and therapeutic methods related to these specific autoimmune diseases and for specific cancers.
(75) In another example, bait molecules consist of various proteins selected from a protein library such as expressed from the commercially available 12,000-member Open Reading Frame (ORF) template library (ORFeome), whereas prey molecules consist of small molecules that have been selected using standard screening methods well known in the pharmaceutical industry as potential drug compounds. Alternatively the prey molecules might be part a small compound library used to screen for possible drug candidates or drug targets. Detecting and identifying the interaction of a library of small compounds and library of proteins simultaneously using the present invention provides critical information for the pharmaceutical industry in identifying potentially useful drugs, drug targets and also to identify side-effects of drugs.
(76) In another example, both bait and prey molecules consist of various molecules selected from a protein library such as expressed from the commercially available 12,000-member Open Reading Frame (ORF) template library (ORFeome) proteins. In this case, the detection and Identification of specific protein-protein interactions provides important information for elucidating various cellular pathways and the role that specific proteins play in active cellular process and in disease. In addition, this information can lead to the discovery of new biomarkers for diagnosis and new drug targets to treat specific diseases involving these cellular pathways.
(77) In another example, bait molecules consist of various proteins selected from a protein library such as expressed from the commercially available 12,000-member Open Reading Frame (ORF) template library (ORFeome) which have been treated with a biologically active molecule which produces a chemical or structural change in particular proteins or polypeptides in the library, whereas prey molecules consist of molecules which are detect or probe chemical and structural changes in the bait molecules. As one example, the protein bait molecules is treated with a specific kinase which causes phosphorylation of specific Tyr, Ser or Thr residues present in the sequence of specific proteins or polypeptides. Once the proteins are treated with this kinase, antibodies that are specific for phosphorylated Tyr, Ser or Thr which constitute the prey molecules are allowed to interact with said protein bait molecules. Detecting and identifying the interaction of specific antibodies with the phosphorylated proteins provides important information about kinase substrate specificity and can identify new drug targets and drugs to treat specific diseases.
(78) In another example, bait molecules consist of various antibodies selected for their specific interaction with a target analytes such as a particular proteins or other molecules whose concentration in a sample is to be determined, whereas prey molecules consist of various antibodies selected for their specific interactions with the same set of target analytes. In this case, detecting and identifying the interaction of bait and prey via mutual interaction with the analyte provides important information about the analytes presence in the sample and its concentration. Such a “sandwich” configuration of antibodies directed against a target analyte in a mixture to be measured is commonly used in sandwich ELISAs and is well known to those in the field of biochemistry and molecular biology. In this case, using the methods provided in this invention, the concentration of thousands of analytes in a sample can be simultaneously measured. In addition, using methods described for coding bait and prey with mass tags, the expression level of thousands of analytes in multiple samples can be determined.
(79) It is to be understood in addition that the mass tags used to code bait or prey molecules can consist of a wide variety of molecules and their isotope labeled variants including but not limited to polypeptides, oligonucleotides, linear block co-polymers, branched polymers and small molecules such as those part of a small compound library used to probe drug targets.
(80) In one embodiment, the method of decoding is based on the use of mass-tags and more preferably photocleavable mass-tags which remain covalently attached to the bead or attached via a binding agent until exposed to light (see below). In one preferred embodiment of this invention, the mass-tags are modified polypeptides whose sequence has been chosen so that its mass is unique (i.e. differs from every other mass tag used in the library). In a second embodiment, the mass-tags are isotopically labeled molecules with the same structure but different masses. In a third embodiment, the mass tags consist of a different polymers than a polypeptide such as an oligonucleotide. In a forth example, a 2,2′-(ethylenedioxy)-bis-(ethylamine) is used as the basic building block for constructing the mass tag. It is to be understood that in this invention there are a variety of molecules which would serve as mass tags and it is not limited to one class of molecule or polymer.
(81) In the case of polypeptides or modified polypeptides which serve as mass tags it is to be understood that a relatively small peptide (e.g. an octamer, N=8) can provide sufficient number of sequences to provide sufficient unique masses to satisfy even a large-library of 100,0000 different proteins (20N=˜25×109). In practice, the number of viable sequences depends on the mass resolution of the MALDI-MS instrument which is often better than 0.1 daltons in the mass range measured. In addition, any degeneracy in the molecular weight of the mass tags can be decoded using the ability of MALDI-MS to sequence small peptides (<5,000 MW), commonly know as MS-MS-TOF. Additional “fine-tuning” of masses can be accomplished by modification of the mass-tag such as the addition of fluorescent labels.
(82) As shown in
(83) A single mass tag of sufficient length or multiple mass tags can be used to code a bead set which is contacted with a solution of prey molecules from a set of multiple solutions in order to determine the presence of prey molecules in each solution. In this case, solution 1 containing a set of prey molecules is mixed with beads coded with mass tags that are unique for that set of bait molecules and solution 2 containing a different set of prey molecules is mixed with the beads coded with a different mass tags that uniquely coded that set of bait molecules. The beads from these two steps after contact with the respective solution 1 and 2 containing prey molecules are then mixed together and used to form a random array. The fact that different sets of mass tags are used to code the two sets of beads allows the prey from solution 1 which interact with bead set 1 and the prey from solution 2 which interact with bead set 2 to be uniquely determined. It will be understood by those familiar with the barcoding approach applied in genomic DNA sequencing that such an approach will allow prey interaction with bait to be determined uniquely for a large set of samples.
(84) Photocleavable Linkers for Mass Tags
(85) In some embodiments where mass-tags are attached to beads for identification of bait and/or directly to prey molecules, the mass tags do not need to be directly covalently linked to the bead surface or prey molecule but instead bound through a binding agent such as an antibody-polypeptide interaction (e.g. Experimental Example 3). However, this is non-ideal since stringent wash steps can result in partial removal of the tags (as observed during the course of some our experiments). One solution to this problem is covalently attached mass-tags which are photo-released upon exposure to UV-light. Alternatively, a near-covalent strength linkage between (strept)avidin and biotin (Kd=10.sup.−15) can be used in conjunction with a photocleavable linker (e.g. Experimental Example 5).
(86) AmberGen has developed a novel class of photocleavable linkers (PC-Linkers) useful in a variety of applications such as photocleavage assisted molecular purification, tRNA-mediated protein engineering, photo-activation of compounds, biomolecules and viruses as well as photocleavable mass-tagging for multiplexed assays [Olejnik, Krzymanska-Olejnik et al. (1996) Nucleic Acids Res 24: 361-6; Olejnik, Krzymanska-Olejnik et al. (1998) Nucleic Acids Res 26: 3572-6; Olejnik, Ludemann at al. (1999) Nucleic Acids Res 27: 4626-31]. In the case of mass-tagging of the proteomic bead-library, a short peptide with 7 or 8 amino acids is linked to the beads via a photocleavable linker. Note that previous experiments have demonstrated that AmberGen's PC-Linker is rapidly photocleaved with 95% efficiency is less than 10 minutes using a low-intensity commercial black-light [Olejnik, Ludemann at al (1999) Nucleic Acids Res 27: 4626-31].
(87) In one embodiment of the invention PC-Mass-Tags for protein identification are attached to beads in either one of 2 ways as illustrated in
(88) Ultra-High Affinity Biotin-(Strept)Avidin: Peptide mass-tags modified at the N-terminus with AmberGen's photocleavable biotin are attached to (strept)avidin coated beads (see
(89) Direct Covalent: Using the primary amine-reactive NHS chemistry on the uncoated agarose beads, peptide mass-tags bearing an N-terminal photocleavable primary amine moiety will be chemically attached simultaneously with the attachment of the protein capture element (e.g. capture antibody). This is highly analogous to AmberGen's phosphoramidite technology distributed through Glen Research Inc. for introducing a photocleavable primary amine at the 5′ end of DNA [OleJnik, Krzymanska-Olejnik eat al. (1998) Nucleic Acids Res 26: 3572-6]. For this method, peptide mass-tags lacking lysines (reactive primary amine on side chain), or where lysines are blocked on the ε-amine, will be used to avoid non-cleavable attachment to the NHS-activated agarose beads.
(90) A library of peptides pro-screened by mass spectrometry could be obtained by commercial synthesis from a variety of available vendors such as Mimotopes (Austria), Peptide 2.0 Inc. (Chantilly, VA) or GenScript Inc. (Piscataway, NJ) and used to create the mass-tags which will be photocleavably linked to the beads. High throughput peptide synthesis services are available from these vendors (e.g. soluble peptide arrays in 96-well plates) and peptides can be purchased with full HPLC and mass spectrometry quality controls. Conventional solid-phase chemical peptide synthesis begins at the C-terminus and ends at the N-terminus. The growing peptide is tethered to the solid-phase synthesis resin via its C-terminal carboxyl group, exposing its N-terminal amine (after deprotection) and allowing sequential attachment of another N-terminal blocked amino acid precursor (again followed by deprotection). Thus, the attachment of N-terminal modified PC-Biotin or PC-amine (amine protected) amino acid precursors at the final cycle of synthesis is a relatively strait forward process.
(91) We have calculated that due to the high analytical sensitivity of mass spectrometry (attomoles), even adding 10 fmoles per bead of mass-tags (10-mer), the aforementioned peptides with N-terminal PC-Linker modification and all quality control data will add only pennies (¢10) to the cost of an entire proteome-bead library.
(92) In addition to PC-Mass-Tags attached to the beads for identification purposes, Bead-GPS™ utilizes PC-Mass-Tags attached to the probes used to query the proteome library. In the case of autoantigen discovery, the PC-Mass-Tag is attached to the anti-human IgG secondary antibody used to detect the bound scrum autoantibody. In this case, only one species of unique mass-tag is required. This has already been demonstrated in Experimental Example 7 (
(93) Mass-Tag Decoding
(94) In general, a requirement of mass-tag decoding is that each mass-tag peak must correspond to the correct molecular weight predicted by the mass tag molecular structure such as for example a given polypeptide sequence plus any modifications or isotope labeling within the resolution of the spectrometer in the specific mass range (˜0.1 Da in 600-4,000 Da range).
(95) It is highly desirable that each mass-tag peak must have a signal-to-noise ratio of at least 50:1, although lower signal-to-noise Is sufficient for some applications. For comparison, the signal-to-noise ratio of single prototype mass-tags attached to beads incorporated into ordered arrays as described in the examples are routinely greater than 250:1 using a set of standard mass spectral parameters. Note that the signal-to-noise ratio in all experiments is determined using the ABI4800 software, which measures the integrated target peak intensity and ratios this to the integrated intensity of a nearby background region which exhibits no detectable peaks.
(96) Importantly, spectral resolution and mass accuracy of the ABI 4800 Plus MALDI-TOF-TOF analyzer is sufficient to unambiguously identify peptides separated by as little as 0.1 Da. However, one potential problem is the appearance of several peaks for each peptide in the mass spectra, which are separated by 1 Da (the “isotope envelope”), due to the presence of small amounts of mass-shifted C13 and N15 atoms in the protein sequence. In the case of two mass tags separated by only a few Daltons, the spectral overlap may affect the tag identification. This will be addressed by using, in real-time, a spectral processing routine called peak de-isotoping. The routine, which is built-in into the ABI 4800 data acquisition software, replaces multiple peaks in the isotope envelope with a single mono-isotopic peak (corresponding to the sequence containing only C12 and N14 atoms).
(97) MALDI-MS Imaging Software
(98) The MALD-MS imaging of individual beads described in this invention require software to analyze data and to identify mass-tags on individual beads. There are a variety of software packages available commercially for this purpose. As an example, we have utilized BioMap in the Experimental Examples, which is a powerful biomedical image analysis software package supporting various data types generated by optical, PET, CT and mass-spectrometry based imaging. The BioMap platform allows visualization and storage of large volumes of data including experiment-specific information such as scan ID, experimental protocol and sample history. It is also a flexible tool that can be easily modified to accommodate a specific requirement. It is contemplated that as part of this invention improved imaging of individual beads and mass-tags can be made that is designed for MALDI-MS bead-imaging workflow, such as automated co-registration of fluorescent and MALDI-MS scan images and identification of positive “hits” based on the detection of PC-Mass Tags.
(99) Automation
(100) In general, mass spectrometry and MALDI-MS in particular have proven to be highly amenable to high throughput applications in both clinical and basic research settings. For example, Sequenom Inc. has established MALDI-MS as an effective technique in the field of genotype profiling, and is providing diagnostic products in this area. As a second example of automation of mass spectrometry in clinical diagnostics, the Pediatrix Medical Group, the largest provider in the US for neonatal blood tests, uses tandem array mass spectrometry to detect metabolic disorders and has screened over 2 million babies using this method.
(101) In the case of this invention, many improvements are envisioned which can facilitate automation and high throughput biomarker discovery. For example, multiplexing can be achieved at several stages including during the preparation of the bead library and in bar-coding multiple sample. Importantly, the use of a highly automated mass spectrometer such as the ABI 4800 Plus MALDI-TOF MS or the more advanced ABI 5800 will also facilitate high throughput analysis at the MALDI-MS bead scanning stage. For example, this system uses advanced software designed for automated scanning of a two-dimensional area, data collection and spectral processing. The ability to automatically scan approximately 10,000 beads in the pico-well MALDI plate in one hour is possible. Use of the ABI 5800 should reduce this time to 1/10 (6 minutes). Furthermore, using one of the commercially available plate-loading robots will allow use of the instrument in the operator-free mode, 24 hours a day. As an example of automation levels achievable with MALDI, Sequenom, Inc. has introduced a MALDI-based system for SNP analysis which is capable of analyzing 100,000 genotypes per day.
(102) In Situ Mass-Fingerprinting of Proteome Bead-Arrays
(103) Rather that the addition of exogenous mass-tags, or any other tags, it is possible to utilize the bead-bound cell-free expressed human proteins themselves as identification codes. Analogous to mass spectrometry based on mass-fingerprinting used in classical proteomics, proteins are digested with protease (e.g. trypsin) and the resultant peptide “fingerprint” used for protein identification. If necessary, the peptides are further fragmented in the TOF/TOF tandem mass spectrometer and sequenced using the standard capabilities of today's instruments. We have explored this possibility using in situ trypsinization of protein-bead arrays in the pico-well plates. In the experiment shown in
(104) Application to Detection of Genetic Mutations Using MALDI-MS-Imaging, Mass Tags and Arrays of Beads
(105) One embodiment of this invention is directed towards the multiplex detection of mutations which might exist in multiple regions of single genes or multiple genes. In order to detect such mutations, a nascent protein or polypeptide (typically a portion of a gene product, wherein the portion is between 5 and 200 amino acids in length, and more commonly between 5 and 100 amino acids in length, and more preferably between 5 and around 60 amino acids in length-so that one can work in the size range that corresponds to optimal sensitivity on most mass spectrometry equipment) is (in one embodiment) first synthesized in a cell-free or cellular translation system from message RNA or DNA coding for the protein which may contain a possible mutation. The nascent protein or polypeptide is then separated from the cell-free or cellular translation system using an N-terminal (located at or close to the N-terminal end of the protein) which is designed to bind to a binding agent on the surface of a bead. For example, the C-terminal epitope can consist of an HSV sequence as discussed here and binding agent on the bead consists of an anti-HSV directed antibody. A C-terminal epitope is not used to avoid the case of chain truncating mutations which would eliminate this epitope.
(106) This process as described above can then be repeated to examine additional sequences in a given gene or multiple genes. In cases where genomic material is used this may be necessary in order to span whole exons or pieces of exons such as in the BRCA1 or BRCA2 gene which contains over 50 exons. Alternatively, different sequences in different genes may wish to be examined in the case for example of a tumor where multiple oncogenes may be suspect. Thus, using the methods described in this invention, a library of beads will be formed each one containing a sequences derived from different gene sequences and also containing unique mass tags coding that particular type of bead. In this case, the sequence interrogated on the individual bead may be a mixture of wild-type sequence and mutant sequences derived from the same region of the gene interrogated.
(107) The resulting isolated material (which may contain both wild-type and mutant peptide sequences) is then analyzed by mass spectrometry consisting of the measurement of individual beads which are part of a bead array. Detection of a peak in the mass spectrum with a mass correlating with the expected wild-type peptide indicate the wild-type peptide. Detection of a peak in the mass spectrum with a mass not correlating with the wild-type peptide indicates a mutation.
(108) It is important to note that in this example, the mass of the wild-type sequence and the resulting peak it produces from an individual bead serves the role of a mass-tag, e.g. it allows one to identify the bait species captured on the bead. The presence of a mutation is then identified by the additional peaks from the bead which do not correspond to the wild-type species. For example, a missense change of a wild-type sequence which corresponded to a codon shift from TAT to TCT would result in the substitution of a Tyr with a Ser and a subsequent mass shift of +176 daltons. It will be understood by those practiced in the art of mass spectrometry that advanced systems are able to distinguish much small shifts even below 1 dalton so almost all substitutions can be detected. Furthermore, MS-MS techniques allow sequencing of the peptides to resolve any ambiguity if the wild-type peptide is not unambiguously identified by its mass. In some embodiments it may be advantageous to also code the bead containing a particular polypeptide species using mass tags as described extensively in this invention. For example, in this embodiment the molecules capture with a binding agent on the bead surface consist of nascent proteins or polypeptides synthesized in a cell-free or cellular translation system from message RNA or DNA coding for the region which may contain a possible mutation. Furthermore, the nascent proteins produced in the cell free reaction could be added in separate reactions to a particular mass-tagged bead as described previously under parallel method or formed using the batch techniques described before.
(109) It is to be understood that the protein bead library used in this embodiment could be formed using the Parallel Synthesis methods described previously or the Batch Synthesis methods described previously. In the Parallel Synthesis methods each protein or polypeptide sequence is synthesized using PCR and cell-free synthesis in separate reactions and the resulting proteins added to individual suspensions of beads whereas in the Batch Preparation method, the entire library can be formed in two multiplex reactions. In either case, the overall library is randomly arrayed and the individual beads measured using mass spectrometry.
(110) Photocleavable DNA-Tags
(111) In addition to peptide based decoding of the proteome-bead library, we have developed an alternative method of coding individual beads based on the use of PC-DNA-Tags. Such tags are also based on proprietary photocleavage technology developed by AmberGen (see U.S. Pat. Nos. 5,948,624; 5,986,076; 6,589,736; 7,312,038; 7,339,045; 7,211,394; 6,057,096; 6,218,530; 7,057,031; 7,195,874; 7,485,427; 7,547,530) which offers convenient synthesis of DNA molecules with a 5′-modification consisting of a photocleavable linker such as PC-aminotag-phosphoramidites commercially distributed by Glen Research Inc (Sterling, VA). These tags can be directly linked to the activated agarose beads similar to the method used for PC-Mass-Tags and released upon exposure to near-UV light Once removed from individual positive beads, these PC-DNA-Tags can be rapidly decoded and quantified in bulk using a massively parallel PCR platform.
(112) One embodiment of this invention involves use of photocleavable DNA-tags to code and decode beads. As an example, solid-phase (bead) PCR with universal photocleavably attached primers, can be used to separately amplify various human ORF plasmid inserts on a 34 micron agarose beads; thus creating photocleavably tethered DNA amplicons (pure species on each bead). Several different DNA-bead species are then pooled at various ratios and then photocleaved.
(113) The photo-released DNA is then analyzed on a suitable instrument which can detect the DNA-tags such as a standard DNA hybridization chip (e.g. DNA microarray) or an RT-PCR device. In the case of DNA hybridization chips, many chips are available such as from Affymetrix which have probes for thousands of genes that can be used to detect the release of specific DNA sequences photoreleased from the beads. Commercial prototypes have also been introduced such as by WaferGen Inc. that can simultaneously analyze large numbers of such PC-DNA tags. In one example, a 5,000-member prototype RT-PCR chip was used containing probes to all members of the test bead library evaluated. As shown in
(114) Importantly, attachment of PC-DNA-tags is Ally compatible with the Parallel protein-bead library production methods described above. Photocleavage of the individual beads and collection of DNA-tags from positive beads can be easily accomplished almost simultaneous with fluorescent scanning (i.e. bead selection step) by using a modified fluorescence microarray scanner. In particular, a laser normally used for scanning the image can be replaced with a laser capable of photocleavage of DNA from individual beads such as a pulsed Nd-Yag laser with 355 nm output which are widely commercially available at low cost. It has been demonstrated by us that such lasers can photocleave >90% of the tags on a bead sample in less than a few seconds. Since commercial fluorescent scanners operating with multiple wavelengths and different lasers are designed to perform sequential scans maintaining image registration, software image identification of positive beads would allow the Nd-Yag laser to be switched on to expose only positive beads during a sequential registered scan. Scan resolution is normally 3-5μ allowing high precision for Nd-Yag laser beam to photocleave DNA-tags from 35μ beads located in the pico-well plate. Alternatively a scanner using a CCD imager along with a photocleaving laser can be readily used to selectively remove DNA-tags from specific beads identified as positive in the fluorescent scan. Photocleaved DNA-tags can be collected in a thin fluidic chamber overlaying the array for subsequent decoding. Importantly, selection of the positive hits is simplified for this approach since the imaging and photo-release are done simultaneously in the same instrument.
(115) Importantly, both the in situ trypsinization and PC-DNA based decoding approaches are also fully compatible with the Batch method of protein-library construction described earlier which involves single-molecule solid-phase emulsion PCR to create a bead-sorted library of expressible DNA in a single reaction, which is then converted to a bead-sorted library of in vitro expressed proteins in a single self-assembling cell-free protein synthesis reaction.
EXPERIMENTAL
Example 1. Affinity Purification of Cell-Free Expressed Peptides onto an Agarose Bead Affinity Resin Followed by Mass Spectrometry Detection from Single Beads
(116) In this Example a test peptide corresponding to a segment of the APC gene, with an expected molecular weight of 6,203 Da, containing an N-terminal FLAG® epitope tag (Sigma-Aldrich, St. Louis, MO), was synthesized in a recombinant cell-free transcription/translation reaction according to the manufacturer's instructions (PureSystem; Post Genome Institute Co., LTD., Japan).
(117) The nascent peptide from the reaction was then purified on mouse anti-FLAG antibody-coated agarose affinity beads (˜75-150 micron diameter). For this, the crude cell-free expression mixtures were diluted with 50 μL of AB-T [100 mM ammonium bicarbonate with 0.1% Triton X-100 (v/v)]. Mouse anti-FLAG antibody coated agarose affinity beads were used in batch mode to purify the cell-free expressed peptides (EZview™ Red Anti-FLAG® M2 Affinity Gel, Sigma-Aldrich, St. Louis, MO). The diluted crude cell-free expression mixtures were combined directly with ˜1 μL beads in 0.5 mL polypropylene PCR tubes. The mixtures were then incubated for 20 minutes at room temperature with gentle mixing to keep the beads suspended. The beads were then spun down in a micro-centrifuge (˜16,000×g) and the fluid supernatant removed and discarded. The beads were then washed 2×10 min each in mass spectrometry grade water (MSG-Water), removing the fluid supernatant as before. After removing the final wash, beads were re-suspended in 50 mL MSG-Water and individual beads were selected from suspension by careful pipetting and deposited onto a stainless steel plate for matrix assisted laser desorption ionization time-of-flight mass spectrometry (MALDI-TOF or MALDI-MS). A small volume (0.2-0.5 μL) of MALDI-TOF matrix solution (20 mg/mL sinapinic acid matrix in 50% acetonitrile and 0.1% trifluoroacetic acid) was immediately applied directly on top of the beads. The droplet was then allowed to dry/crystallize under ambient conditions without disturbance. The size of the final spot was approximately 2 mm in diameter with the beads near the center of spot. Once completely dried, the spots were analyzed using a Voyager-DE MALDI-TOF mass spectrometer (Applied Biosystems; Foster City, CA). The MALDI-TOF spectra were acquired on the outer edge of the spot, inside the spot in the immediate vicinity of the beads and also directly from the beads.
(118) Results:
(119)
Example 2. Mass Spectrometry Readout and Mass-Imaging from Individually Resolved Beads
(120) In this Example, two test peptides with molecular weights of 3,483 Da and 3,287 Da, each containing an N-terminal FLAG® epitope tag (Sigma-Aldrich, St. Louis, MO), were separately synthesized in a recombinant cell-free transcription/translation reaction according to the manufacturer's instructions (PureSystem; Post Genome Institute Co., LTD., Japan). The nascent peptides from each reaction were then separately purified on mouse anti-FLAG antibody-coated agarose affinity beads (EZview™ Red Anti-FLAG® M2 Affinity Gel, Sigma-Aldrich, St. Louis, MO) (˜75-150 micron diameter). The beads were then mixed in a 9:1 ratio and manually deposited in a random pattern on a suitable substrate for MALDI-TOF mass spectrometry. The MALDI-TOF matrix (CHCA) was sprayed on in a thin and uniform film and allowed to dry. MALDI-TOF imaging of the surface was performed using an ABI 4800 Plus MALDI-TOF/TOF mass spectrometer (Applied Biosystems; Foster City, CA). The surface of the substrate was scanned with the instrument's laser, in the instrument's reflector mode, in the 1,500-4,000 m/z (mass/charge) spectral range and two images were constructed using spectral intensity at the m/z corresponding to the molecular weight of the peptides.
(121) Results:
(122) A two-color image overlay was created from the two mass-images of the beads that were constructed using the spectral intensity at the m/z corresponding to the molecular weight of the two test peptides (
Example 3. Mass Spectrometry Readout and Mass-Imaging from Individually Resolved 34 Micron Beads Deposited in Pico-Wed Plates: Mass-Tagging of Analyte-Bearing Beads for Identification
(123) In this Example, 34 micron agarose beads were conjugated to an antibody directed against the HSV epitope tag. The beads were then loaded with different recombinant proteins bearing this HSV epitope tag. The beads were additionally bound to one of three different peptide “mass tags” of unique mass, corresponding to the HSV peptide epitope itself conjugated on the N-terminus to different fluorophores. The beads were then deposited into a special pico-well plate and mass-imaged by scanning MALDITOF mass spectrometry. Separately, the presence of the recombinant proteins was detected on the same batch of beads by probing with a fluorescently labeled antibody directed against the common VSV-G epitope tag, also present in the recombinant proteins.
(124) Development of Pico-Well Plates for MALDI-TOF Bead Scanning
(125) In order to create mass images of beads or particles, it may be advantageous to randomly array the beads in a regular two-dimensional grid, similar to spots in a conventional microarray. This maximizes bead density, yet assures bead separation, and allows the MALDI-TOP instrument to efficiently and rapidly move from one bead to another, for optimal scan speed. Furthermore, in order to maximize bead resolution, it may also be advantageous to contain each bead in its own microscopic well. Finally, cross-platform imaging of the beads, such as by mass spectrometry and fluorescence, may be advantageous in certain embodiments of the technology.
(126) For this purpose, we developed a novel dual-use pico-well substrate suitable for both mass spectrometric and light based analyses. This substrate, whose overall dimensions are 75.0 mm long by 25.0 mm wide and 1.0 mm thick, is sliced from a fiber optic bundle (block of fused optic fibers) and fabricated by the etching of 44 micron diameter by 55 micron deep pico-wells (i.e. picoliter scale volume) at the ends of the optical fibers that are positioned 50 microns from center-to-center in a hexagonal ordered array (Incom Inc., Charlton, MA). The result is >0.5 million wells in the dimensions of a standard microarray or microscope slide (75×25×1 mm). The design allows deposition of only one bead per well, but maximum access of the MALDI-TOF laser beam to vaporize the matrix coating the bead and allow mass analysis. Since the array is fabricated from a fiber optic bundle, it also forms a face-plate for convenient measurement of light based signals, for example fluorescence or luminescence from each bead (each well) using direct-contact CCD cameras or by using conventional fluorescence microarray scanners.
(127) Preparation of Anti-HSV Antibody Coated 34 Micron Agarose Affinity Beads
(128) 34 Micron diameter agarose beads were conjugated to an anti-HSV tag antibody for later use in capturing peptides bearing this epitope tag. To do so, an anti-HSV monoclonal antibody (EMD Biosciences, Inc., San Diego, CA) was diluted to 0.5 mg/mL in a final buffer of 200 mM sodium bicarbonate and 200 mM NaCl (Binding Buffer). 6% cross-linked NHS-activated 34 micron agarose beads (NHS HP SpinTrap, GE Healthcare Life Sciences, Piscataway, NJ) were washed 4× in several bead volumes of ice cold 1 mM HCl. Beads were then reacted with the anti-HSV antibody solution at a ratio of 6 μg of antibody per each μL of actual bead volume for 1 hour with gentle mixing.
(129) Beads were then washed 1× briefly and 2×30 min with several bead volumes each of 200 mM glycine and 1 mM EDTA in Binding Buffer. Beads were then washed 2×5 min in Binding Buffer, 2×5 min in 10 mM Tris, 1 mM EDTA, pH 8 with 200 mM NaCl, and 1× briefly in 10 mM Tris, 1 mM EDTA, pH 8 with 50 mM NaCl. Beads were then prepared to a 20% (v/v) suspension in 10 mM Tris, 1 mM EDTA, pH 8 with 50 mM NaCl and stored at +4° C.
(130) Preparing Modified HSV Peptide Mass Tags of Different Mass
(131) Peptides of unique mass were prepared by chemical modification with different fluorophores of the HSV-Tag peptide (KQPELAPEDPED), which was purchased from Sigma-Aldrich (St. Louis, MO). To do so, the peptide was prepared to 5 mg/mL in 100 mM sodium bicarbonate and reacted overnight (with mixing) with equimolar amounts of the Cy3-NHS or Cy5-NHS activated (primary amine reactive) fluorescent dye labeling reagents (GE Healthcare Life Sciences, Piscataway, NJ). Peptides were used without further purification (MALDI-TOF analysis showed that the peptides were almost exclusively labeled at a ratio of 1 dye per peptide molecule). Because the NHS activated labeling reagents react only with primary amines, selective labeling of the N-terminal lysine is anticipated. The unlabeled as well as Cy3 and Cy5 labeled HSV peptides provided three species of unique mass tags, 1,368 Da, 2,048 Da and 2,074 Da respectively, which were used in subsequent steps.
(132) Binding of Recombinant Proteins and HSV Peptide Mass Tags to the Anti-HSV Agarose Affinity Beads
(133) Human p53 and human KLHL12 were expressed recombinantly in a cell-free reaction. Expression reactions were performed using a transcription/translation coupled rabbit reticulocyte lysate system (TNT® T7 Quick for PCR DNA; Promega, Madison, WI) according to the manufacturer's instructions. The expression plasmid used was a derivative of the pETBlue-2 vector (EMD Biosciences, Inc., San Diego, CA) containing a C-terminal polyhistidine (HHHHHH) and HSV epitope tag (QPELAPBDPED) as well as an N-terminal VSV-G epitope tag (YTDIEMNRLOK) flanking the Open Reading Frame (ORF) inserts of human p53 or human KLHL12. A parallel negative control expression reaction was performed lacking only the plasmid DNA.
(134) After the expression reaction, the nascent proteins were isolated onto the aforementioned anti-HSV antibody coated agarose affinity beads. Protein isolation onto the beads was performed in batch mode using 0.45 micron pore size, PVDPF membrane, micro-centrifuge Filtration Devices to facilitate manipulation of the small volumes of affinity beads and exchange the buffers (Ultrafree-MC Durapore Micro-Centrifuge Filtration Devices, 400 μL capacity; Millipore Billerica, Massachusetts). Filtration Devices were used unless otherwise stated. For each sample (1 Filtration Device per sample), 1 μL packed bead volume (˜30,000 beads) was washed briefly 4×400 μL with TBS [TBS=50 mM Tris (2-amino-2-(hydroxymethyl)-1,3-propanediol) pH 7.5 and 200 mM NaCl] and then 2×400 μL briefly with high purity Molecular Biology Grade Water (MB&Water). Washed bead pellets were then re-suspended at a ratio of 50 μL of crude expression reaction per μL packed bead volume (˜30,000 beads) and mixed for 30 min to capture the target nascent recombinant protein. Beads were then washed briefly 3×400 μL with TBS-T [TBS with 0.05% v/v Tween-20] and then 1×400 μL with Block Buffer [1% w/v BSA in TBS-T].
(135) Next the beads (which at this stage already contained the nascent recombinant proteins; 1 μL packed bead volume each sample) were additionally loaded with the aforementioned HSV peptide mass tags. 730 pmoles of the aforementioned HSV peptide mass tags was added to each of the three bead samples (blank, human p53 and human KLHL12 bead samples), one unique mass tag species per bead sample, in 200 μL of Block Buffer. This corresponded to ˜25 fmoles of HSV peptide mass tag added per bead. HSV peptide mass tags were allowed to bind for 30 min with mixing. Beads were then washed 3×400 μL briefly with TBS-T, 3×400 μL TBS and then 4×400 μL with Mass Spectrometry Grade Water (MS-Water).
(136) MALDI-TOF Mass Spectrometry Imaging of Beads
(137) As a result of the previous steps in this Example, the three mass-tagged bead species were as follows: Blank beads (no recombinant protein) coded with the unlabeled HSV peptide mass tag (1,368 Da), human p53 beads coded with the Cy3 labeled HSV peptide mass tag (2,048 Da) and human KLHL12 beads coded with the Cy5 labeled HSV peptide mass tag (2,074 Da). Next, the three bead species were pooled in equal amounts and the pooled bead population was then deposited into the aforementioned pico-well plates by brief centrifugation. MALDI-TOF Imaging (scanning) of the pico-well plate was performed essentially as described in Example 2, in order to detect the HSV peptide mass tags from individual beads.
(138) Verification of Bound Recombinant Protein
(139) In order to verify the presence of bound recombinant protein on the beads, an aliquot of the same batch of beads was saved alter loading the recombinant proteins but before loading the HSV peptide mass tags. These beads were probed with a fluorescently (Cy3) labeled anti-VSV-G tag antibody (Sigma-Aldrich, St Louis, MO) in order to detect this common N-terminal epitope tag present in all the nascent recombinant proteins. To do so, beads were probed with 200 μL of antibody diluted from the manufacturer supplied stock at 1/50 in Block Buffer. Probing was performed for 30 min with mixing. Using the aforementioned Filtration Devices, beads were then washed 4×400 μL briefly with TBS-T and then 2×400 μL with TBS.
(140) Beads were then embedded in a thin polyacrylamide film on a glass microscope slide for fluorescence imaging. The acrylamide mix was prepared by mixing 487 μL TBS, 113 μL of a 40% acrylamide and bis-acrylamide mixture (19:1 ratio; Bio-Rad Laboratories, Hercules, CA), 1 μL of 100% TEMED (Bio-Rad Laboratories, Hercules, CA) and 6 μL of a 10% (w/v) ammonium persulfate solution (prepared in water). This acrylamide mix was used to resuspend the washed bead pellet to form 1% (v/v) beads. Approximately 10-20 μL of the bead suspension was placed on a standard glass microscope slide, overlaid with an 18 mm round microscope cover glass and allowed to polymerize for approximately 10 min. The microscope slides were fluorescently imaged using a GenePix 4200 laser based microarray scanner (Molecular Devices, Sunnyvale, CA).
(141) Results:
(142) Two recombinant human proteins (p53 and KLHL), each having a common C-terminal HSV epitope tag, were expressed in a cell-free system and loaded/isolated onto anti-HSV antibody-coated 34 micron agarose beads. As a negative control, blank beads were also prepared in the same manner except only the expressible DNA was omitted from the cell-free protein synthesis reaction. After protein expression and bead-capture, each of the three bead species was additionally loaded with a unique HSV peptide mass tag having a molecular weight of 1,368, 2,048 or 2,074 Da (binding to beads again mediated by the anti-HSV antibody coating). The 3 bead species were then pooled and loaded into the aforementioned pico-well plates. For MALDI-TOF mass spectrometry, the matrix was applied as a thin and uniform film to the plate surface. The surface was then scanned in the MALDI-TOF mass spectrometry reflector mode in the 1,500-4,000 m/z spectral range. Three mass images were constructed using spectral intensity at the m/z corresponding to the molecular weight of the HSV peptide mass tags. Three distinct populations of beads are visible on the pico-well plate (
Example 4. Synchronization of Fluorescence Image and Mass Spectrometry Mass-Image of Individually Resolved Beads
(143) In this Example, 34 micron agarose beads coated with the anti-HSV tag antibody were loaded with cell-free expressed recombinant human p53 or recombinant human KLHL12, both of which contained the HSV epitope tag; performed exactly as described in Example 3. Also as in Example 3, a parallel set of blank beads was prepared in the same manner except only the expression DNA was omitted from the cell-free reaction used to synthesize the recombinant proteins.
(144) Beads were then loaded with different HSV peptide mass tags labeled on their N-terminus with different fluorophores in order to create unique masses; done exactly as described in Example 3.
(145) As a result of the previous steps in this Example, the three mass-tagged bead species were as follows: Blank beads (no recombinant protein) coded with the unlabeled HSV peptide mass tag (1,368 Da) human p53 beads coded with the Cy3 labeled HSV peptide mass tag (2,048 Da) and human KLHL12 beads coded with the Cy5 labeled HSV peptide mass tag (2,074 Da). The three bead species were pooled, the pooled beads deposited in the aforementioned pico-well plates and the beads then mass-imaged with MALDI-TOF mass spectrometry exactly as described in Example 3, in order to detect the mass tags on individual beads.
(146) After MALDI-TOF mass-imaging, the same area of the pico-wall plates was fluorescently imaged using a GenePix 4200 laser based microarray scanner (Molecular Devices, Sunnyvale, CA).
(147) In this Example, the mass-image of the Cy3 labeled HSV peptide mass tag (2,048 Da) was overlaid (synchronized) with the fluorescence (Cy3) image of the same HSV peptide mass tag (same region of pico-well plate). Results in
Example 5. Photocleavable Mass Tags—Mass Spectrometry Readout and Mass-Imaging from Individually Resolved Beads
(148) Preparation of NeutrAvidin Coated 34 Micron Agarose Affinity Beads
(149) Performed in the same manner as in Example 3 for the anti-HSV antibody coating of 34 micron agarose beads except in this case NeutrAvidin biotin binding protein (Invitrogen, Carlsbad, CA) was conjugated to the beads and loaded at a ratio 10 μg per μL of packed agarose bead volume (NeutrAvidin concentration at binding step was 2.5 μg/μL).
(150) Preparation of Photocleavable (PC) Biotin Labeled Peptide Mass Tags
(151) Performed in the same manner as in Example 3 for the N-terminal fluorescence labeling of the HSV peptide mass tags except that the target peptide was the VSV-G peptide (YTDIEMNRLGK) (Roche Applied Science, Indianapolis, IN) (1,340 Da) and instead of using NHS-activated (primary amine reactive) fluorescence dye labeling reagents, AmberGen's NHS-activated photocleavable (PC) biotin labeling reagent was used (AmberGen Incorporated, Watertown, MA) [Olejnik, Sonar, Krzymanska-Olejnik and Rothschild (1995) Proceedings of the National Academy of Science (USA) 92: 7590-7594; Pandori, Hobson, Olejnik, Krzymanska-Olejnik, Rothschild, Palmer, Phillips and Sano (2002) Chem Biol 9: 567-73].
(152) Binding of PC-Biotin Peptide Mass Tags to NeutrAvidin Agarose Affinity Beads
(153) 250 pmoles of the aforementioned PC-Biotin labeled VSV-G peptide mass tag was added to 1.5 μL of packed NeutrAvidin agarose bead volume (˜45,000 beads) in 50 μL of Block Buffer (see Example 3 for buffer compositions). This corresponds to ˜5 fmoles of PC-Biotin VSV-G peptide mass tag added per bead. The PC-Biotin VSV-G peptide mass tag was allowed to bind for 30 min with mixing. Beads were then washed 3×400 μL briefly with TBS-T, 3×400 μL TBS and then 4×400 μL with Mass Spectrometry Grade Water (MSG-Water).
(154) Mass Tag Photocleavage and MALDI-TOF Mass Spectrometry Imaging of Beads
(155) Next, the beads loaded with the PC-Biotin VSV-G peptide mass tag were then deposited into the aforementioned pico-well plates (Example 3) by brief centrifugation. MALDI-TOF imaging (scanning) of the pico-well plate was performed essentially as described in Example 2, in order to detect the VSV-G peptide mass tags from individual beads, with the following exceptions: After bead deposition but before matrix coating and MALDI-TOF imaging, photo-release of the captured mass tag was achieved via illumination of the pico-well plates for 5 min with near-UV light (365 nm peak UV lamp, Blak-Ray Lamp, Model XX-15, UVP, Upland, CA) at a 5 am distance. The power output under these conditions was approximately 2.6 mW/cm.sup.2 at 360 nm, 1.0 mW/cm.sup.2 at 310 nm and 0.16 mW/cm.sup.2 at 250 nm. For the minus light negative control (−UV), a portion of the pico-well plate was masked using an opaque solid barrier.
(156) Results:
(157) As demonstrated by the mass-image in
Example 6. Photocleavable Mass Tags (for Bead Identification) Co-Loaded with “Bait” Molecules for Multiplex Bioassays: “Bait” Detection and Mass Spectrometry Readout from Beads
(158) One embodiment of mass spectrometry mass-imaging of beads or particles is to load onto the beads both a mass tag for bead identification and “bait” molecules or compounds for use in multiplex bioassays. In this Example, beads are co-loaded with both photocleavable (PC) peptide mass tags for identification and human recombinant proteins as “bait” compounds.
(159) Preparation of Dual Affinity Beads Coated with Both NeutrAvidin and the Anti-HSV Tag Capture Antibody
(160) Performed in the same manner as in Example 3 for the anti-HSV antibody coating of 34 micron agarose beads except in this case both the anti-HSV antibody and NeutrAvidin (Invitrogen, Carlsbad, CA) were conjugated to the same batch of beads. In this case, instead of 6 μg of anti-HSV antibody per μL packed agarose bead volume as done in Example 3, 4 μg of anti-HSV antibody and 2 μg of NeutrAvidin (6 μg total protein) was co-loaded per each μL of packed agarose bead volume.
(161) Fluorescent Labeling of Dual Affinity Beads for Bead-ELISA Assay
(162) The aforementioned dual affinity beads were directly labeled with fluorescence in order to enable normalization of total bead number per sample in downstream bead-ELISA assays (see later in this Example for Bead-ELISA). The beads were fluorescently labeled as follows: The aforementioned Filtration Devices (see Example 3) were used to manipulate the beads in the following procedures unless otherwise noted. Beads were washed 4× briefly with several bead bed volumes each of Conjugation Buffer (200 mM sodium bicarbonate and 200 mM NaCl). Beads were then prepared to a 25% v/v bead suspension in Conjugation Buffer. Beads were then labeled with 270 pmoles of the Alexa Fluor@ 594 SSE labeling reagent (Invitrogen, Carlsbad, CA) per each μL of packed bead volume (˜30,000 beads) for roughly 10 fmoles added labeling reagent per bead. Labeling reagent was added from a 27 mM stock in anhydrous DMSO. The labeling reaction was performed for 30 min with gentle mixing and protected from light. Beads were then washed 4× briefly in several bead bed volumes of quench buffer (100 mM glycine in TBS; see Example 3 for TBS) and then 2× briefly in several bead bed volumes of 0.1% BSA w/v in TBS. Beads were then prepared to a 10% v/v bead suspension in 0.1% BSA w/v in TBS and stored at +4° C. protected from light.
(163) As a quality control measure, 1 μL of packed bead volume in 100 μL of 0.1% BSA w/v in TBS was transferred to the wells of a 96-well opaque black flat bottom microtiter plate. Beads were allowed to settle by gravity for 5 min and the Alexa Fluor® 594 fluorescence read in a TECAN SpetraFluor Plus plate reader (recan Group Ltd, Männedorf, Switzerland) using a 560 nm excitation filter and 612 nm emissions filter. Using several replicate samplings, fluorescence signal was compared to beads lacking the AleaFluro® 594 (same bead amount), yielding an average signal-to-noise ratio of 11:1.
(164) Preparation of Photocleavable (PC) Biotin Labeled Peptide Mast Tags
(165) Performed in the same manner as in Example 5 except that the bradykinin peptide (Sigma-Aldrich, St. Louis, MO) (RPPGFSPFR) was used instead of the VSV-G peptide in Example 5.
(166) Binding of PC-Biotin Peptide Mass Tags to Dual Affinity Beads
(167) 112.5 pmoles of the aforementioned PC-Biotin labeled bradykinin peptide mass tag was added to 0.75 μL of pecked dual affinity agarose bead volume (˜22,500 beads) in 225 μL of Block Buffer (see Example 3 for buffer compositions). This corresponds to ˜5 fmoles of PC-Biotin bradykinin peptide mass tag per bead. The PC-Biotin bradykinin peptide mass tag was allowed to bind for 30 min with mixing. Beads were then washed 4×400 μL briefly with TBS-T (see Example 3 for buffer). As a negative control, a parallel batch of beads was processed in the same manner except that the PC-Biotin labeled bradykinin peptide mass tag was omitted.
(168) Binding of Recombinant Protein as “Bait” to Dual Affinity Beads
(169) Performed exactly as in Example 3 with the following exceptions: The aforementioned dual affinity beads, with and without the PC-Biotin labeled bradykinin peptide mass tag, were used for recombinant protein capture instead of the anti-HSV beads from Example 3. To create beads containing the “bait” (nascent recombinant protein), the dual affinity beads were loaded with cell-fee expressed recombinant human p53 protein (see Example 3). The p53 contained the HSV epitope tag for binding to the beads. Prior to capture on the beads, in this Example, the crude cell-free expression reactions was mixed with equal volume of 5% BSA w/v in TBS-T and pro-clarified by passing the solution through the aforementioned Filtration Devices (see Example 3). Also as in Example 3, a parallel set of blank beads was prepared in the same manner except only the expression DNA was omitted from the cell-free reaction used to synthesize the recombinant protein.
(170) After loading the protein to the beads, washing was also performed as in Example 3. These bead samples (bead suspensions) were than each split, whereby half of each sample was used for a Bead-ELISA for detection of the bead-bound p53 bait molecules and the other half used for MALDI-TOF mass spectrometry for detection of the photocleavable peptide mass tag. Both procedures are detailed below.
(171) Bead-ELISA for Detection of the Human Recombinant p53 “Bait” on the Beads
(172) Beads were manipulated with the aforementioned Filtration Devices unless otherwise noted. Beads were probed with 200 μL of a monoclonal anti-VSV-G horseradish peroaxidase (HRP) conjugated antibody (Clone P5D4, Sigma-Aldrich, St Louis, MO) to detect the bead-bound human recombinant pS3 “bait” with contained this N-terminal epitope tag. For probing, the manufacturer supplied antibody (˜1 mg/mL) was diluted 1/20,000 in Block Buffer (see Example 3 for buffer). Probing was performed for 30 min with gentle mixing. Beads were then washed briefly 4×400 μL with TBS-T and 2×400 μL with 0.1% BSA w/v in TBS. Each bead sample was then re-suspended in 100 μL of 0.1% BSA w/v in TBS and transferred to the wells of a 96-well opaque black flat bottom microtiter plate. Beads were allowed to settle by gravity for 5 min and the Alexa Fluor® 594 fluorescence read (bead normalization signal) in a TECAN SpetraFluor Plus plate reader (Tecan Group Ltd., Männedorf Switzerland) using a 560 nm excitation filter and 612 nm emissions filter.
(173) Next the anti-VSV-G HRP antibody was detected to measure the bead-bound “bait”, i.e. the recombinant human p53. Without further processing, to each well of the microtiter plate containing the beads, 100 μL of SuperSignal Pico Chemiluminescence ELISA substrate (Thermo-Fisher-Pierce, Rockford, IL) was added and mixed for 15 min on a plate shaker. Beads were again allowed to settle by gravity for 5 min and the signal read in the TECAN SpetraFluor Plus plate reader (Tecan Group Ltd., Männedorf Switzerland) using the instrument's luminescence mode.
(174) Mass Tag Photocleavage and MALDI-TOF Mass Spectrometry Analysis from Beads
(175) To detect the bead-bound PC-Biotin bradykinin peptide mass tag, an aliquot of the same batch of beads that were loaded with the mass tag as well as the recombinant human p53 (see earlier in this Example) was analyzed by mass spectrometry. This aliquot of beads was further washed 3×400 μL briefly with TBS-T, 3×400 μL TBS and then 4×400 μL with Mass Spectrometry Grade Water (MSG-Water) using the aforementioned Filtration Devices. The beads were suspended in a small volume of MSG-Water and photo-release of the bead-bound mass tag was achieved via illumination for 5 min with near-UV light (365 nm peak UV lamp, Blak-Ray Lamp, Model XX-15, UVP, Upland, CA) at a 5 cm distance. The power output under these conditions was approximately 2.6 mW/cm.sup.2 at 360 nm, 1.0 mW/cm.sup.2 at 310 ma and 0.16 mW/cm.sup.2 at 250 rm. The supernatant was then mixed with the MALDI-TOF matrix and analyzed under standard conditions (note that single bead mass-imaging was not done hero).
(176) Results:
(177)
(178) An aliquot of the same batch of beads containing the bound PC-Biotin bradykinin peptide mass tag and bound recombinant p53 “bait” protein was also analyzed by MALDI-TOF mass spectrometry, following photocleavage of the mass tag from the beads. In this Example, the photo-released mass tag from an entire population of beads was measured in bulk by conventional MALDI-TOF mass spectrometry, but other embodiments envisioned would involve mass-imaging of individually resolved beads similar to as done in Examples 5 and 7. Results in
Example 7. Photocleavable Mass-Tagged Probes for Mass Spectrometry Readout and Mass-Imaging from Individually Resolved Beads: Autoantibody Detection in Autoimmune Disease
(179) In this Example, specific probes were mass tagged with peptides and used to detect analytes bound to “bait” molecules present on beads. Detection was by using MALDI-TOF mass spectrometry mass-imaging. More specifically, the “bait” molecule on the beads in this Example is a recombinant human protein, acting as an autoantigen (i.e. self antigens targeted by serum autoantibodies in autoimmune disorders). The analyte in this Example is an autoantibody (specific human IgG) present in the serum of a patient having an autoimmune disorder (Primary Biliary Cirrhosis or PBC in this case); whereby the autoantibody is detected with a mass-tagged anti-[human IgG] secondary antibody probe.
(180) Binding of Recombinant Protein Autoantigen to Anti-HSV Agarose Affinity Beads
(181) Performed exactly as in Example 3. First, to create beads containing the “bait” (nascent recombinant protein autoantigen in this Example), 34 micron agarose beads coated with the anti-HSV tag antibody were loaded with cell-free expressed recombinant human KLHL12 protein (PBC autoantigen; see US provisional filling at USPTO application number 61248768). The KLHL12 contained the HSV epitope tag for binding to the beads. Also as in Example 3, a parallel set of blank beads was prepared in the same manner except only the expression DNA was omitted from the cell-free reaction used to synthesize the recombinant protein. After loading the protein to the beads, washing was also performed as in Example 3.
(182) Preparation of the VSV-G PC-Biotin Labeled Peptide Mass Tag
(183) Performed exactly as in Example 5.
(184) Preparation of Fluorescent Tetrameric NeutrAvidin Protein as a Bridge from Probe to Mass Tag
(185) A 5 mg/mL stock of tetrameric NeutrAvidin biotin binding protein (Invitrogen, Carlsbad, CA) was prepared in PBS (50 mM sodium phosphate, pH 7.5, 100 mM NaCl). The stock was then mixed with equal volume of 200 mM sodium bicarbonate 200 mM NaCl (no pH adjustment). 500 μL of this solution was labeled with the Cy3-NHS ester reagent (GE Healthcare Life Sciences, Piscataway, NJ) added from a 25 mM stock (stock in anhydrous DMSO) to yield a 10-fold molar excess of labeling reagent versus the NeutrAvidin. The reaction was carried out for 30 min with gentle mixing and protected from light Un-reacted labeling reagent was removed by passing the solution through an Illustra NAP-5 G-25 sepharose desalting column according to the manufacturer's instructions (GE Healthcare Life Sciences, Piscataway, NJ) versus a TBS buffer. Concentration of the purified and labeled NeutrAvidin was determined by measuring absorbance at 280 nm.
(186) Treatment of Autoantigen Beads with Human Autoimmune Serum and Subsequent Probing
(187) The aforementioned beads, prepared with and without the KLHL12 autoantigen and washed as described above, were then treated with either a known KLHL12 autoantibody-positive PBC autoimmune serum or with a known autoantibody-negative normal patient serum ProMedDx, LLC (Norton, MA). Both sera were previously confirmed KLHL12 autoantibody-positive or negative by analysis on commercial human proteome microarrays performed according the manufacturer's instructions (Human ProtoArray® 4.0, Invitrogen, Carlsbad, CA). Serum treatment, probing and bead washing steps were all performed in the aforementioned Filtration Devices (see Example 3) unless otherwise noted. Sera were diluted 1/1,000 in Block Buffer (see Example 3 for buffers unless otherwise noted) and 200 μL was used to treat 1 μL packed bead volume (˜30,000 beads) for each sample. Treatment was performed for 30 min with gentle mixing and the beads then washed 4×400 μL briefly with TBS-T. Beads were then probed with 200 μL of a non-cleavable biotin labeled mouse anti-[Human IgG] secondary antibody (Jackson ImmunoResearch Laboratories, Inc., West Grove, PA) diluted to 10 μg/mL (˜65 nM) in Block Buffer. Treatment was performed for 30 min with gentle mixing and the beads then washed 4×400 μL briefly with TBS-T. Beads were then probed with 200 μL of the aforementioned Cy3 labeled NeutrAvidin diluted to 4 μg/mL (˜65 nM) in Block Buffer. Treatment was performed for 30 min with gentle mixing and the beads then washed 4×400 μL briefly with TBS-T. Lastly, beads were then probed with 200 μL of the aforementioned PC-Biotin labeled VSV-G peptide mass tag diluted to 65 nM in Block Buffer. Treatment was performed for 30 min with gentle mixing and the beads were then washed 3×400 μL briefly with TBS-T, 3×400 μL TBS and then 4×400 μL with Mass Spectrometry Grade Water (MSG-Water).
(188) Mass Tag Photocleavage and MALDI-TOF Mass Spectrometry Imaging of Beads
(189) Next, this same batch of beads was then split and deposited into two separate pico-well plates (see Example 3 for plates) by brief centrifugation. One plate was used for MALDI-TOF mass spectrometry mass-imaging of individually resolved beads. Photocleavage of the PC-Biotin labeled VSV-G peptide mass tag from the beads in the plates and mass-imaging were performed as in Example 5, in order to detect bound autoantibody by virtue of the mass tag. The other plate was imaged fluorescently in a GenePix 4200 laser based microarray scanner (Molecular Devices, Sunnyvale, CA) to detect bound autoantibody by virtue of the fluorescently labeled NeutrAvidin.
(190) Results:
(191) The top panel of
(192) KLHL12 (autoantigen) beads and blank beads (minus antigen) were prepared and probed with the PBC autoimmune serum. A set of autoantigen beads was also probed with a normal serum as a negative control. The different bead samples were kept separate in this Example. In this experiment, bead-bound autoantibody was probed with a biotinylated (non-cleavable) anti-[human IgG] secondary antibody, followed by tetrameric NeutrAvidin as a bridge, and finally, the photocleavable (PC) biotin labeled VSV-G peptide mass tag. To enable fluorescence detection of the bead-bound autoantibody, the NeutrAvidin bridge was labeled with Cy3.
(193) The results in
Example 8. Application of the Protease Enzyme to the Bead Library Deposited on the Pico-Well Plates: Efficient Protein Digestion without the Loss of Spatial Resolution in MALDI-TOF Mass Spectrometry Mass-Imaging
(194) In this Example, a specific protein (recombinant human p53) bound to the beads, which were deposited inside wells of a pico-well glass slide, was detected using both MALDI-TOF mass spectrometry imaging (MSI) and fluorescent scanning. The Example shows the ability to apply both the enzyme-containing solution and the MALDI matrix required for MALDI Imaging in a manner that preserves the single-bead resolution of the array.
(195) Construction of the Probe Complex
(196) The 34 micron agarose beads were coated with the anti-HSV tag antibody. Cell-free produced recombinant human p53 protein containing an N-terminal VSV tag and a C-terminal HSV tag was purified on beads as described in Example 3. Subsequently, the protein was probed with a biotinylated anti-VSV antibody followed by incubation with a Cy3-labeled tetrameric NeutrAvidin and extensive wash to remove unbound fluorescent label.
(197) Trypsin Digest of the Protein
(198) The 34 micron beads containing the p53 probe complex were deposited inside wells of the pico-well glass slide. The bead density was approximately 1 bead per 20 wells. A dilute (25 μg/mL) aqueous solution of mass-spectrometry grade trypsin was applied to the surface of the slide in the form of a fine mist using a Pad (Midlothian, VA) LC® Sprint reusable nebulizer. Following trypsin application, the slide was incubated for 1 r at 37° C. to allow the digestion.
(199) Application of the MALDI Matrix
(200) For the purpose of MALDI imaging, the slides were coated with a thin layer of α-hydroxy cinnamic acid (CHCA) MALDI matrix A 16 mg/mL solution of MS-grade CHCA in 60% of pure acetonitrile and 40% of 0.1% trifluoroacetic acid (v/v) was delivered to the surface of the slide in the form of a fine mist using a Pari (Midlothian, VA) LC® Sprint reusable nebulizer.
(201) Fluorescence Scanning of the Bead Library following the Trypsin Digestion and MALDI Matrix Application
(202) The slides were scanned using a GenePix 4200 laser based microarray scanner (Molecular Devices, Sunnyvale, CA) at the 532 nm wavelength corresponding to the fluorescent signal of Cy3 (
(203) Mass-Spectrometry Readout of the Protein Digest
(204) MALDI MSI scanning of the surface of the same fluorescently imaged pico-well slide showed a series of peaks in the 600-3,200 Da mass range that can be assigned to proteolytic fragments of p53 produced by digestion of intact protein using trypsin. For example, the 890.4 Da peak (
(205) The above results demonstrate that digestion of protein on individual beads followed by application of MALDI matrix does not decrease the resolution of the bead array and individual beads can still be resolved even at 10 micron resolution of the fluorescence scanner. More generally, it is expected that application of other enzymes or compounds dissolved in either aqueous or organic solution to the bead array can be performed in a manner that preserves the resolution of the bead array.
Example 9. Measurement of Changes in the Protein Concentration Using a Combination of Protein Isotope Labeling, Proteolytic Digestion and MALDI-TOF Mass Spectrometry Mass-Imaging Analysis of Bead Microarrays
(206) This Example shows the ability to measure changes in the concentration of a specific protein (recombinant p53) obtained from two different sources using MALDI bead microarrays. This is useful, for example, when changes in the protein expression between two different cell types need to be measured for multiple proteins. The approach involves: (1) expressing proteins separately in the non-labeled and isotope-labeled medium that result in incorporation of the isotope label into the synthesized proteins; (2) combining the two samples and purifying the aforementioned proteins on affinity beads such as antibody beads; (3) arranging beads into the microarray, (4) performing on-bead proteolytic digestion and (5) measuring ratio of non-labeled versus isotope-labeled proteolytic fragments, which is indicative of the ratio of proteins in the starting mixture.
(207) Protein Isotope Labeling
(208) Recombinant human p53 was expressed in a cell-free translation reaction supplemented with non-labeled (natural abundance) amino acid mix. Separately, p53 was expressed in a reaction supplemented with .sup.13C.sub.6-Leu amino acid mix. Incorporation of an isotope labeled Leucine into the protein chain results in a mass shift of +6 Da per each Leucine residue.
(209) Affinity Purification
(210) After the cell-free translation, the protein mixtures were separately purified on anti-HSV antibody-coated 34 micron agarose beads. In a separate experiment, the protein mixtures were mixed in a 5:1 ratio (non-labeled vs labeled) before purification and subsequently purified on anti-HSV antibody-coated 34 micron agarose beads.
(211) Trypsin Digestion
(212) The bead mixtures were deposited on the MALDI plate and subject to trypsin digestion. In a separate experiment, the beads were deposited into the pico-well glass slides and treated with trypsin.
(213) MALDI-TOF Mass Spectra
(214) The trypsin digest spectra acquired from beads, each carrying a homogenous population of p53 protein (either labeled or non-labeled), reveal a series of peaks shifted by either +6 or +12 Da. For example (
(215) Protein Quantification Using Isotope Labeling and MALDI MS Detection
(216) Next, p53 was expressed in non-labeled and isotope labeled media and the translation reactions were mixed in a 5:1 ratio prior to binding to the beads to mimic different levels of protein expression. Trypsin digestion and MS analysis were performed as described previously. The mass spectra (
(217) Detection of Isotope-Shifted Peaks on Bead Microarrays
(218) In this example, the mixture of beads carrying either pure non-labeled or pure Isotope-labeled p53 proteins was deposited on the pico-well slides, so that each well contains no more than one bead. The on-bead trypsin digestion and MALDI matrix deposition were performed as described in Example 8. The slide was scanned using MALDI MSI and signals at 1,006 and 1,018 Da corresponding to the isotope-shifted p53 proteolytic fragments were detected. As seen in
Example 10. Photocleavable DNA Tags and Bead Decoding by Massively Parallel RT-PCR Chips
(219) One embodiment of this invention involves use of PC-DNA tags to code and decode beads. As an example, solid-phase (bead) PCR with universal photocleavably attached primers, was used to separately amplify various human open reading frame (ORF) plasmid inserts on a 34 micron agarose beads; thus creating photocleavably tethered DNA amplicons (pure species on each bead). Several different DNA-bead species were then pooled at various ratios and then photocleaved.
(220) The photo-released DNA can be analyzed on a suitable instrument which can detect the DNA tags, such as a standard DNA hybridization chip (e.g. DNA microarray), a massively parallel DNA sequencer or an RT-PCR device. In the case of DNA hybridization chips, many chips are available such as from AffyMetrix (Santa Clara, CA) which have probes for thousands of genes that can be used to detect the release of specific DNA sequences photo-released from the beads. In this Example, a commercial prototype massively parallel RT-PCR chip from WaferGen BioSystems Inc. (Fremont, CA) was used that can simultaneously analyze large numbers of such PC-DNA tags. In this Example, WaferGen's 5,000-member prototype RT-PCR chip was used containing probes to all members of the test bead library evaluated. As shown in
Example 11. Physical Pre-Selection of Beads for Decoding Using a Fluorescence Activated Cell-Sorting (FACS) Instrument
(221) We evaluated in the feasibility of pre-isolating 34 micron agarose beads using fluorescence activated cell-sorting (FACS). Pre-isolation of only the positive beads of interest (e.g. by virtue of bound fluorescent probes) greatly reduces the number of beads required to be decoded by MALDI-TOF mass spectrometry mass-imaging for example. Importantly, FACS is high throughput (can process millions of beads in a few minutes) and has the ability for greater reproducibility and specificity than magnetic particle based affinity isolation methods, since beads can be analyzed by multiple parameters on a bead-by-bead basis. In this Example, blank protein beads and beads containing a recombinant protein autoantigen for the autoimmune disease primary biliary cirrhosis (PBC) were separately prepared and probed with an appropriate autoantibody-positive human serum as detailed in Example 7. Bound autoantibody was detected with a fluorescently labeled secondary anti-[human IgG] antibody (fluorescein). Beads were then analyzed using a fluorescence activated cell sorting (FACS) instrument using a commercial service (BD FACS Vantage Cell Sorter, Cytometry Research LLC, San Diego, CA). As seen in
Example 12. Photocleavable Mass Tags for Bead Identification and Probe Readout in an Immune Response Profiling Scenario: MALDI-TOF Mass-Imaging of Individually Resolved Beads in an Array
(222) One embodiment of mass spectrometry mass-imaging of beads or particles is to load onto the beads both a mass tag for bead identification (“bead identification tag”) and “bait” molecules or compounds for use in multiplex bioassays. Furthermore, a probe (“prey”) used to treat (query) the beads can carry a different photocleavable mass tag for assay readout (“probe tag”). In this scenario, mass-imaging of the beads results in two mass tag signals from those beads on which the bait has bound its cognate probe, one mass tag for the bead identification tag and one for the probe tag. In this Example, human recombinant proteins act as the “bait” compounds. An immune response profiling application is shown here as an example, whereby one of the bait proteins is a known autoantigen and the other bait protein is a negative control (non-autoantigen). The beads are then treated with a human serum from an autoimmune patient known to have autoantibodies against said autoantigen. To detect bound autoantibody, the beads are then probed with an anti-[human IgG] secondary antibody which is ultimately detected with a unique photocleavable mass tag reporter (probe tag).
(223) Preparation of Dual Affinity Beads Coated with Both NeutrAvidin and the Anti-HSV Tag Capture Antibody
(224) Performed as in Example 6 (34 micron agarose beads).
(225) Preparation of Photocleavable (PC) Biotin Labeled Peptide Mass Tags
(226) Performed as in Example 5 except that the following peptides were labeled: Bradykinin (RPPGFSPFR) (Sigma-Aldrich, St. Louis, MO) was labeled for use as the probe tag and two custom peptides, obtained commercially from Sigma-Genosys (The Woodlands, TX), were labeled for use as the bead identification tags (Tag-3.1=MIGGAGGRIR and Tag-3.7=MIGGTGGRIR).
(227) Binding of PC-Biotin Peptide Mass Tags to Dual Affinity Beads
(228) Performed as in Example 6 except that 200 μL of PC-Biotin peptide mass tag solution at a concentration of 0.75 pmoles/μL (150 pmoles) was added to a 1 μL bead pellet volume (30,000 beads); for a final added amount of PC-Biotin peptide mass tag of 5 fmoles per bead. As in Example 6, this step is to capture the PC-Biotin peptide mass tags on the dual affinity beads by way of the NeutrAvidin coating on the beads (peptide capture efficiency not measured). In this Example, separate batches of beads were prepared that we loaded with either the Tag-3.1 or the Tag-3.7 mass tag, creating two pure populations of mass tag encoded beads. Ater capture of the PC-Biotin peptide mass tags on beads and washing as in Example 6, the beads were quenched with 1 mM d-biotin in TBS-T (200 μL per 1 μL bead pellet) (see Example 3 for buffer compositions). Quenching was performed for 30 min with mixing in the upper chamber of the Filtration Devices (see Example 3 for Filtration Devices and their usage). The 1 μL bead pellets were then washed (In the Filtration Device) 4×400 μL briefly with TBS-T.
(229) Binding of Recombinant Protein as “Bait” to Dual Affinity Beads
(230) Performed essentially as in Example 6 with the following exceptions: Beads with the Tag-3.1 identification tag were loaded with cell-free expressed recombinant human Smith protein, a well known autoantigen biomarker for systemic lupus erythematosus (SLE) [Mahler, Stinton and Fritzer (2005) Clin Diagn Lab Immunol 12: 107-13]. Here, the B isoform of Smith was used (SmB). Beads with the Tag-3.7 identification tag were loaded with cell-free expressed recombinant human GST A2 protein as a negative control (not a known autoantigen for SLE) (see Example 3 for cell-free protein expression). Treatment of the beads with the cell-free expression reactions for recombinant protein capture was performed as in Example 6, except that immediately following completion of the expression reactions (prior to mixing with beads), protease inhibitor was additionally added to the expression reactions to reduce the possibility of subsequent proteolytic degradation of the peptide mass tags on the beads (“Complete Mini Protease Inhibitor Cocktail Tablets” form Roche Applied Science, Indianapolis, IN, Catalog Number 11836153001; Stock Solution=1 mini-tablet in 1 mL of purified water, add 1/10 volume of Stock Solution to completed cell-free protein expression reactions). As in Example 6, the mechanism of capture of the cell-free expressed recombinant proteins on the dual affinity beads is by way of the anti-HSV antibody coating on the beads and the common C-terminal HSV epitope tag present in all expressed proteins.
(231) Finally, after loading the recombinant proteins onto the beads, washing was 4×400 μL briefly with TBS-T and 2×400 μL briefly with Block Buffer (see Example 3 for buffer compositions).
(232) Immune Response Profiling Using the PC-Mass-Tagged and Recombinant Protein Loaded Beads
(233) The two bead populations (separate), were each sequentially treated as follows: All bead manipulations and washes were performed in the Filtration Devices unless otherwise noted (see Example 3 for the Filtration Devices and their usage). See Example 3 for buffer compositions. Beads (1 μL pellet volumes) were first probed with a known SmB positive human serum from an SLE patient. Serum was diluted 1/1,000 in 5% BSA (w/v) in TBS-T and 100 μL used to treat the beads for 30 min with mixing. Beads were then washed 5×400 μL briefly with TBS-T. Beads were then probed with 200 μL of a non-cleavable biotin labeled mouse anti-[Human IgG] secondary antibody (Jackson ImmunoResearch Laboratories, Inc., West Grove, PA) diluted to 10 μg/mL (˜65 nM) in 5% BSA/TBS-T supplemented with 1 mM d-biotin. Treatment was performed for 30 min with gentle mixing and the beads then washed 4×400 μL briefly and 2×400 μL for 5 min each with TBS-T. Beads were further washed 4×400 μL briefly with 5% BSA/TBS-T. Beads were then probed with 200 μL of Cy3 labeled NeutrAvidin (see Example 7) diluted to 4 μg/mL (˜65 nM) in 5% BSA/TBS-T. Treatment was performed for 30 min with gentle mixing and the beads then washed 4×400 μL briefly with TBS-T. Lastly, beads were then probed with 200 μL of the aforementioned PC-Biotin labeled Bradykinin peptide mass tag (probe tag) diluted to 65 nM in 5% BSA/IBS-T. Treatment was performed for 30 min with gentle mixing and the beads were then washed 4×400 μL briefly with TBS-T, 4×400 μL TBS and then 4×400 μL with Mass Spectrometry Grade Water (MSG-Water).
(234) Mass Tag Photocleavage and MALDI-TOF Mass Spectrometry Imaging of Beads
(235) Performed in the pico-well plates as in Examples 5 & 7. Note that the two populations of bead species (SmB and GST) were pooled prior to deposition into the pico-well plates. In this Example, no florescence imaging was used.
(236) Confirmation of Mass-Imaging Results using ELISA
(237) Results of autoantibody detection on beads with the mass-imaging approach were confirmed by testing the same SLE serum against the same cell-free expressed recombinant proteins using a 96-well microtiter plate based ELISA. For this, AmberGen's T.sup.2-ELISA™ was used (see Description of Invention for overview of T.sup.2-ELISA™). The procedures were as follows:
(238) Cell-Free Protein Expression for ELISA
(239) See Example 3 for cell-free protein expression. Protein expression reactions contained the cognate plasmid DNA while blank expression reactions lacked only the plasmid DNA. Expression reactions were stopped by diluting I/O in TDB [1% BSA (w/v) and 0.1% (v/v) Triton X-100 in TBS-T (50 mM Tris, pH 7.5, 200 mM NaCl, 0.05% (v/v) Tween-20)].
(240) Enzyme-Linked Immunosorbent Assay (ELISA) for Autoantibody Detection
(241) Numc Brand 96-well Polysorp™ Microwell™ white opaque, flat bottom, untreated polystyrene microtiter plates (Numc Brand from Thermo-Fisher Scientific, Rochester, NY) were used for a sandwich type Enzyme-Linked Immunosorbent Assay (ELISA). Plates were coated with 0.5 μg/mL of a mouse monoclonal anti-HSV® tag capture antibody (EMD Biosciences, Inc., San Diego, CA) in sodium carbonate/bicarbonate pH 9.3 for 30 min with shaking (50 μL/well). Plates were then washed 6× in TBS-T (wells filled to maximum) on an ELx405 Select Robotic Plate Washer (BioTek, Winooski, VT). See Example 3 for TBS-T buffer composition. All plate washes were performed in this manner unless noted otherwise. Plates were then blocked for 30 min at 300 μL/well in 1% BSA (w/v) in TBS-T. The solution was removed from the plates and the aforementioned stopped (i.e. diluted) cell-free expression reactions (protein and blank reactions) were then added at 100 μL/well and shaken for 30 min to allow the nascent proteins to be captured by their common C-terminal HSV epitope tags. Plates were washed and the same SLE serum sample used for the bead assays earlier in this Example (diluted at 1/1,000 in 1% BSA (w/v) in TBS-T) was added at 100 μL/well and shaken for 30 min. The serum sample was run against wells of the proteins and walls of the cell-free expression blank. Additionally, one set of wells of protein and one set of wells of the call-free expression blank were designated for VSV-G epitope tag detection (common N-terminal tag in all expressed proteins), and therefore received plain 1% BSA (w/v) in TBS-T instead of diluted serum at this stage. To avoid contamination of the robotic plate washer with human serum, plates were subsequently washed 4× by manual addition of TBS-T (wells filled to maximum) followed by vacuum aspiration and than washed 6× in the robotic plate washer as described earlier in this Example. Wells designated for detection of the VSV-G epitope tag then received an anti-VSV-G horseradish peroxidase (HRP) labeled monoclonal antibody (Clone P5D4, Roche Applied Science, Indianapolis, IN) diluted 1/20,000 in 1% BSA/BS-T. Wells designated for detection of serum autoantibody received a mouse anti-[human IgG] HRP labeled monoclonal secondary antibody (minimum cross-reactivity with mouse immunoglobulin; Jackson ImmmoResearch Laboratories, Inc, West Grove, PA) diluted 1/20,000 in 1% BSA/TBS-T. Plates were shaken for 30 min. The solutions were than manually dumped from the plates by inversion followed by vigorous patting of the plates inverted on a dry paper towel to remove residual fluid. Plates were then washed in the robotic plate washer as described earlier in this Example. Chemiluminescence signal was generated by the addition of 50 μL/well of SuperSignal ELISA Pico Chemiluminescence Substrate (Pierce Brand from Thermo Fisher Scientific, Rockford, IL). Plates were developed by shaking for 15 min and then read on a LumiCount luminescence plate reader (is exposure, PMT of 650V, gain 1) (Packard/PerkinElmer Life and Analytical Sciences, Inc., Boston, MA).
(242) Results:
(243)
(244) MALDI-TOF mass-imaging of the pooled beads following their deposition into the pico-well plates was performed. Results in
(245) To validate these results, we used a conventional 96-well microtiter plate ELISA assay (T.sup.2-ELISA™) formatted in analogy to the bead assay. Cell-free expressed proteins were immobilized on the ELISA well surface by anti-HSV antibody-mediated capture. Following serum treatment, wells were probed with an enzyme-labeled (HRP) anti-[Human IgG] secondary antibody to detect bound autoantibody. As an additional control, the amount of captured protein was detected in separate wells with a reporter-labeled antibody to the common N-terminal VSV-G epitope tag in all expressed proteins. Results are shown in
Example 13. MALDI-TOF Mass-Imaging of 10 Unique Photocleavable Mass-Tag Encoded Bead Species in an Array: Mass-Imaging for Identification in Conjunction with Antibody Detection of a Bead-Bound Bait Protein
(246) In this Example, 10 distinct species of beads, each carrying unique photocleavable peptide mass tags, and 1 additionally containing a recombinant protein, were prepared and arrayed in the pico-well plates. The array of all 10 bead species, randomly distributed, was then mass-imaged using MALDI-TOF.
(247) All beads in this Example carried both a unique peptide mass tag for identification as well as a common capture antibody. The antibody serves to bind recombinant proteins that contain a common epitope tag. The captured recombinant proteins act as “bait” for specific probes (“prey”), such as a fluorescently labeled detector antibody. In this Example, 1 of the 10 bead species carried cell-free expressed recombinant p53 protein (captured by antibody) in addition to the peptide mass tag. The p53 beads were additionally detected using a fluorescent antibody directed against an epitope tag in the p53 protein.
(248) Preparation of Dual Affinity Beads Coated with Both NeutrAvidin and the Anti-HSV Tag Capture Antibody
(249) Performed as in Example 6 (34 micron agarose beads).
(250) Preparation of Photocleavable (PC) Biotin Labeled Peptide Mass Tags
(251) Performed as in Example 5 except that the following custom peptides obtained commercially from Sigma-Genosys (The Woodlands, TX) were used for labeling (peptide sequence in brackets):
(252) TABLE-US-00001 1. Tag-1.1 [QRPDVTR] 2. Tag-2.3 [DIEHNR] 3. Tag-2.8 [DIERNR] 4. Tag-3.1 [MIGGAGGRIR] 5. Tag-3.2 [MIGGEGGRIR] 6. Tag-3.4 [MIGGIGGRIR] 7. Tag-3.5 [MIGGSGGRIR] 8. Tag-3.6 [MIGGPGGRIR] 9. Tag-3.7 [MIGGTGGRIR] 10. Tag-3.8 [MIGGRGGRIR]
Binding of PC-Biotin Peptide Mass Tags to Dual Affinity Beads
(253) Performed as in Example 6 except that 200 μL of PC-Biotin peptide mass tag solution at a concentration of 0.75 pmoles/μL (150 pmoles) was added to a 1 μL bead pellet volume (30,000 beads); for a final added amount of PC-Biotin peptide mass tag of 5 fmoles per bead. As in Example 6, this step is to capture the PC-Biotin peptide mass tags on the dual affinity beads by way of the NeutrAvidin coating on the beads (peptide capture efficiency not measured). In this Example, 10 separate batches of beads were prepared, each batch uniquely loaded with 1 of the 10 aforementioned peptides, creating 10 pure populations of unique mass tag encoded beads. After capture of the PC-Biotin peptide mass tags on beads and washing as in Example 6, the beads were quenched with 1 mM d-biotin in TBS-T (200 μL per 1 μL bead pellet) (see Example 3 for buffer compositions). Quenching was performed for 30 min with mixing in the upper chamber of the Filtration Devices (see Example 3 for Filtration Devices and their usage). The 1 μL bead pellets were then washed (in the Filtration Device) 4×400 μL briefly with TBS-T.
(254) Binding of Recombinant Protein as “Bait” to Dual Affinity Beads
(255) Although in some embodiments all mass tagged bead species can additionally carry a “bait” molecule such as a protein, in this Example, with the exception of the Tag-3.7 beads, the other 9 mass tagged bead species were not loaded with cell-free expressed recombinant protein and thus not subjected to this portion of the procedure:
(256) Beads with the Tag-3.7 Identification tag were loaded with cell-free expressed recombinant human p53 protein essentially as in Example 6 with the following exceptions: Treatment of the beads with the cell-free expression reaction for recombinant p53 protein capture was performed as In Example 6, except that immediately following completion of the expression reaction (prior to mixing with beads), protease inhibitor was additionally added to the expression reaction to reduce the possibility of subsequent proteolytic degradation of the peptide mass tags on the beads (“Complete Mini Protease Inhibitor Cocktail Tablets” form Roche Applied Science, Indianapolis, IN, Catalog Number 11836153001; Stock Solution=1 mini-tablet in 1 mL of purified water; add 1/10 volume of Stock Solution to completed cell-free protein expression reactions). As in Example 6, the mechanism of capture of the cell-free expressed recombinant p53 protein on the dual affinity beads is by way of the anti-HSV antibody coating on the beads and the C-terminal HSV epitope tag present in the recombinant p53. Finally, after loading the recombinant p53 protein onto the Tag-3.7 encoded beads, washing was 4×400 μL briefly with TBS-T and 2×400 μL briefly with Block Buffer (see Example 3 for buffer compositions).
(257) As a negative control, a separate “blank” bead sample was prepared, corresponding to beads treated with a cell-free expression reaction that was performed lacking only the p53 DNA.
(258) Pooling Beads and Fluorescence Detection of p53 Beads
(259) All 10 bead species were then pooled in equal numbers. As a control, an aliquot of the pre p53 beads was also set aside. This created 3 bead populations (“samples”): The 10-species pool, the p53 beads and the blank beads (see above). Each of the 3 bead samples was then probed with a fluorescently labeled anti-VSV-Cy3 antibody to specifically detect the p53 beads, by way of the VSV epitope tag present in the recombinant p53. After fluorescence probing, an aliquot of the 10-species pooled sample was set aside for MALDI-TOF mass-imaging (next section). All remaining portions of the 3 bead samples were embedded in a polyacrylamide film on a microscope slide and fluorescently imaged in a microarray scanner. The fluorescence antibody probing, embedding and fluorescence imaging was performed as in Example 3 in the section headed “Verification of Bound Recombinant Proteins”.
(260) Mass Tag Photocleavage and MALDI-TOF Mass Spectrometry Imaging of Beads
(261) Performed in the pico-well plates as in Examples 5 & 7.
(262) Results:
(263) This Example successfully shows the ability to mass-image 10 distinct peptide-bead species (mass-tagged beads), using 34 micron beads arrayed in the pico-well plates (peptides affinity captured on beads with a photocleavable linker).
(264) To confirm that peptide mass tag imaging of beads is compatible with the presence of “bait” molecules on the same beads, such as a recombinant protein, an aliquot of the same beads was subjected to fluorescence imaging on a microscope slide. Note: These beads had also been probed with a florescent antibody directed specifically against an epitope tag in the p53 protein. Results are shown in
Example 14. Direct Chemical Linkage of Photocleavable Mass Tags to Bead Surfaces: Elimination of the Need for Affinity Based Linkages
(265) The previous Examples used affinity linkages to attach peptide mass tags to beads for MALDI-TOF mass-imaging and decoding, e.g. bead-bound antibodies used to capture epitope-containing peptide mass tags in Example 3, or bead-bound (strept)avidin used to capture photocleavable biotin labeled peptide mass tags in Example 13. However it is possible, and in fact desirable in some cases, to directly attach mass tags (peptides or otherwise) to beads using a direct chemical linker by way of a covalent bond, whereby the linker is photocleavable. This eliminates the need for an affinity capture agent (e.g. antibody or (strept)avidin)), which may interfere with some downstream applications. This Example will show one embodiment of this:
(266) In this Example, the compound shown in
(267) In this Example,
(268) Following preparation of the photocleavable mass tags as described above, experiments will be performed similar to Example 13, except that instead of attachment of photocleavable biotin labeled peptide mass tags to NeutrAvidin coated beads, the photocleavable amine modified peptide mass tags synthesized here will be directly attached to the NHS-activated beads (
(269) One expected benefit will be less background noise in assay readout (e.g. probe detection) and generally less interference with downstream assays due to the lack of a NeutrAvidin or (strept)avidin coating on the beads (e.g. less interference with assays involving in situ, on-bead protease digestion of bead-bound proteins such as in Example 9).
(270) In a related embodiment that will be evaluated, a compound similar to the “Photocleavable Amine Linker” shown in the upper left panel of
Example 15. Mass Spectrometry Readout and Mass-Imaging from Individually Resolved 34 Micron Beads in Metal Coated Pico-Well Plates
(271) In this Example, fluorescence imaging and MALDI-TOF mass-imaging of beads in plain and gold-coated pico-well plates was evaluated. Conductive surfaces are typically more ideal for MALDI-TOF as they avoid charge buildup.
(272) Pico-Well Plates
(273) Pico-well plates (Incom Inc., Charlton, MA) as described in Example 3 were either used as is, or coated with a thin layer of gold by the manufacturer (Incom Inc., Charlton, MA). Briefly, coating involves plasma cleaning of the plates, then sputtering on a thin coat of titanium to promote adhesion of gold to the glass, and then applying a 5 nm layer of gold on top of that. SEM (scanning electron microscopy), EDX (energy dispersive X-ray analysis) and AFM (atomic force microscopy) were used by the manufacturer to verify uniform coating of the plates.
(274) Preparation of Dual Affinity Beads Coated with Both NeutrAvidin and the Anti-HSV Tag Capture Antibody
(275) Performed as in Example 6
(276) Preparation of Photocleavable (PC) Biotin Labeled Peptide Mass Tags
(277) Performed in the same manner as in Example 5 except that the bradykinin peptide (Sigma-Aldrich, St. Louis, MO) (RPPGFSPFR) was used instead of the VSV-G peptide in Example 5.
(278) Binding of PC-Biotin Peptide Mass Tags to Dual Affinity Beads
(279) Performed as in Example 6 except that 200 μL of PC-Biotin peptide mass tag solution at a concentration of 0.75 pmoles/μL (150 pmoles) was added to a 1 μL bead pellet volume (30,000 beads); for a final added amount of PC-Biotin peptide mass tag of 5 fmoles per bead. As in Example 6, this step is to capture the PC-Biotin peptide mass tags on the dual affinity beads by way of the NeutrAvidin coating on the beads (peptide capture efficiency not measured).
(280) Binding of Recombinant Protein as “Bait” to Dual Affinity Beads
(281) Beads with the bradykinin mass-tag were loaded with cell-free expressed recombinant human p53 protein essentially as in Example 6 with the following exceptions: Treatment of the beads with the cell-free expression reaction for recombinant p53 protein capture was performed as in Example 6, except that immediately following completion of the expression reaction (prior to mixing with beads), protease inhibitor was additionally added to the expression reaction to reduce the possibility of subsequent proteolytic degradation of the peptide mass tags on the beads (“Complete Mini Protease Inhibitor Cocktail Tablets” form Roche Applied Science, Indianapolis, IN, Catalog Number 11836153001; Stock Solution=1 mini-tablet in 1 mL of purified water; add 1/10 volume of Stock Solution to completed cell-free protein expression reactions). As in Example 6, the mechanism of capture of the cell-free expressed recombinant p53 protein on the dual affinity beads is by way of the anti-HSV antibody coating on the beads and the C-terminal HSV epitope tag present in the recombinant p53. Finally, after loading the recombinant p53 protein onto the bradykinin encoded beads, washing was 4×400 μL briefly with TBS-T and 2×400 μL briefly with Block Buffer (see Example 3 for buffer compositions).
(282) Pooling Beads and Fluorescence Detection of p53 Beads
(283) Beads were then probed with a fluorescently labeled anti-VSV-Cy3 antibody to specifically detect the p53 beads, by way of the VSV epitope tag present in the recombinant p53. The fluorescence antibody probing was performed as in Example 3 in the section headed “Verification of Bound Recombinant Proteins”. For both the plain glass and gold-coated pico-well plates, imaging of fluorescence was performed through the bottom of the plates (i.e. thorough the fiber optics).
(284) In Situ Trypsinization for Gold Versus Glass Comparison
(285) In a second experiment, beads were created carrying recombinant proteins captured via the anti-HSV antibody similar to as earlier in this Example, but without any mass-tags and without any antibody probing steps. Beads were deposited into gold-coated and glass pico-well plates and In situ trypsin digestion was performed similar to as in Example 8. In this example human ACVR-2B was used as the recombinant protein.
(286) Mars Tag Photocleavage and MALDI-TOF Mass Spectometry Imaging of Bead
(287) Performed in the plain glass and gold-coated pico-well plates as in Examples 5 & 7. Photocleavage was not used for in situ trypsin digested samples.
(288) Results:
(289) Results in
(290) Results in
Example 16. MALDI-TOF Mass-Imaging of Multiple Unique Photocleavable Mass-Tag Encoded Bead Species in an Array: Synchronizing Mass-Image for Bead Identification with Fluorescence Antibody Detection of a Bead-Bound Bait Protein
(291) In this Example, 2 distinct species of beads, each carrying a unique photocleavable (PC) peptide mass tag, and 1 of the 2 additionally carrying bound recombinant p53 protein (“bait”), were prepared separately, pooled and then probed with a fluorescently labeled antibody (“prey”) to detect the recombinant p53 protein. A 3rd bead species was then spiked in as “Marker Beads”. These Marker Beads carried a unique photocleavable peptide mass tag for identification and the bead surface was covalently labeled with a fluorophore having different and distinguishable spectral properties than that used on the antibody probe. The pool of 3 bead species was randomly arrayed in the pico-well plates and then imaged by mass and by fluorescence.
(292) Preparation of NeutrAvidin Coated Beads
(293) NeutrAvidin coated 34 micron agarose beads were prepared as in Example 5. As described below, these NeutrAvidin beads were used to prepare both the “Dual Affinity Beads” and the “Marker Beads”.
(294) Preparation of Dual Affinity Beads Coated with NeutrAvidin Directly and Indirectly with the Anti-HSV Tag Capture Antibody
(295) The aforementioned NeutrAvidin beads were then loaded a biotin labeled anti-HSV tag polyclonal antibody as follows: Goat anti-HSV tag polyclonal antibody was purchased from Bethyl Laboratories (Montgomery, TX), provided at 1 mg/mL in PBS. 800 μL of this antibody solution (800 μg) was then mixed with 1/9.sup.th volume of 1M sodium bicarbonate. The antibody was biotin labeled by adding a 10-fold molar excess of EZ-Link-Sulfo-NHS-LC-Biotin (Thermo-Fisher-Pierce, Rockford, IL) and reacting for 30 min with gentle mixing. The reaction was quenched by adding 1/9.sup.th volume of 1M glycine for 15 min with gentle mixing. The labeled antibody was then purified on a PD MidiTrap G-25 desalting column against TBS (see Example 3 for buffer) and according to the manufacturer's instructions (GE Healthcare Life Sciences, Piscataway, NJ). The purified and biotin-labeled antibody was then diluted to 0.15 μg/μL in TBS-T (see Example 3 for buffer). Using this solution, the NeutrAvidin beads were coated at a ratio of 12 μg of the biotin labeled anti-HSV tag antibody per each 1 μL of actual bead pellet volume for 30 min with gentle mixing. Beads were then washed 4× briefly with an excess of TBS-T using the aforementioned Filtration Devices (see Example 3 for devices). Beads were stored as a 20% (v/v) suspension at +4° C.
(296) Note that because the antibody does not saturate all the biotin binding sites on the NeutrAvidin beads, it was therefore possible to additionally load PC-Biotin labeled peptide mass tags onto the beads (see later steps in this Example).
(297) Preparation of “Marker Beads” Coated with NeutrAvidin Directly and Labeled with Fluorescence
(298) Performed as in Example 6 except that the aforementioned NeutrAvidin beads (no bound antibody) were used for fluorescence labeling and the Cy5-NHS activated (primary amine reactive) fluorescent dye labeling reagent was used (GE Healthcare Life Sciences, Piscataway, NJ).
(299) Preparation of Photocleavable (PC) Biotin Labeled Peptide Mass Tags
(300) Performed as in Example 5 except that the following commercially available peptides were used for labeling with PC-Biotin (peptide sequence in brackets) (all peptides purchased for labeling were from AnaSpec, Fremont, CA, except for Bradykinin which was from Sigma-Aldrich, St. Louis, MO):
(301) TABLE-US-00002 1. Heparin-Binding Peptide V [Trp-Gln-Pro-Pro-Arg-Ala-Arg-Ile]; 1023 Da 2. [D-Phe7]-Bradykinin [Arg-Pro-Pro-Gly-Phe-Ser-D-Phe-Phe-Arg]; 1111 Da 3. Bradykinin [Arg-Pro-Pro-Gly-Phe-Ser-Pro-Phe-Arg]; 1060 Da
Binding of PC-Biotin Peptide Mass Tags to Dual Affinity Beads
(302) Performed as in Example 6 except that 150 μL of PC-Biotin peptide mass tag solution, at a concentration of 5 pmoles/μL (750 pmoles peptide mass tag), was added to a 1 μL bead pellet volume (30,000 beads); for a final added amount of biotin-labeled peptide mass tag of 25 fmoles per bead. As in Example 6, this step is to capture the PC-Biotin labeled peptide mass tags on the beads by way of the NeutrAvidin coating on the beads (peptide capture efficiency not measured). In this Example, 3 separate batches of beads were prepared, each batch uniquely loaded with 1 of the 3 aforementioned peptide species, creating 3 pure populations of unique mass tag encoded beads. The Bradykinin mass tag was loaded onto the aforementioned “Marker Beads” (NeutrAvidin beads with direct fluorescent label attached to beads) while all other mass tags were loaded onto the aforementioned “Dual Affinity Beads” (NeutrAvidin beads additionally containing a common capture antibody for tagged recombinant proteins).
(303) Binding of Recombinant Protein as “Bait” to Dual Affinity Beads
(304) The following was performed on all bead species (kept separate) except the “Marker Beads” which were not subjected to the procedures in the paragraph below:
(305) Beads with the “Heparin—Binding Peptide V” identification tag were loaded with cell-free expressed recombinant human p53 protein essentially as in Example 6 with the following exceptions: Treatment of the beads with the cell-free expression reaction for recombinant p53 protein capture was performed as In Example 6, except that immediately following completion of the expression reaction (prior to mixing with beads), protease inhibitor was additionally added to the expression reaction to reduce the possibility of subsequent proteolytic degradation of the peptide mass tags on the beads (“Complete Mini Protease Inhibitor Cocktail Tablets” form Roche Applied Science, Indianapolis, IN, Catalog Number 11836153001; Stock Solution=1 mini-tablet in 1 mL of purified water, add 1/10 volume of Stock Solution to completed cell-free protein expression reactions). As in Example 6, the mechanism of capture of the cell-free expressed recombinant p53 protein on the Dual Affinity Beads is by way of the anti-HSV antibody coating on the beads and the C-terminal HSV epitope tag present in the recombinant p53. Finally, after loading the recombinant p53 protein onto the “Heparin—Binding Peptide V” encoded beads, washing was 4×400 μL briefly with TBS-T. Beads with the “[D—Phe7]—Bradykinin” identification tag were also subjected to the same above procedure, except that a blank cell-free expression reaction was used as a negative control instead of a p53 cell-free expression reaction. In this case, the cell-free expression reaction lacked the cognate expressible p53 DNA.
(306) Pooling Beads and Fluorescence Detection of p53 Beads
(307) All bead species, except “Marker Beads” which were added later, were than pooled in equal numbers (2 species). The pooled beads were then probed with a fluorescently labeled anti-p53 antibody to specifically detect the p53 containing beads. The fluorescently labeled anti-p53 antibody was prepared in the same manner as the fluorescently labeled NeutrAvidin described in Example 7, except that the anti-p53 antibody (clone BP53-12, Santa Cruz Biotechnology, Santa Cruz, CA) was used at 0.2 mg/mL in PBS and the Alexa Fluor® 594 SSE labeling reagent (Invitrogen, Carlsbad, CA) was used. The anti-p53 fluorescence antibody probing was performed similar to as in Example 3 in the section headed “Verification of Bound Recombinant Proteins”. After fluorescence probing, “Marker Beads” were spiked in at an approximate equal ratio to the other bead species.
(308) Mass Tag Photocleavage and MALDI-TOF Mass Spectrometry Imaging of Beads
(309) Performed in the pico-well plates as in Examples 5 & 7.
(310) Fluorescence Imaging of Pico-Well Plats
(311) After MALDI-TOP mass-imaging, the pico-well plates were fluorescently imaged using a GenePix 4200 laser based microarray scanner (Molecular Devices, Sunnyvale, CA). This is possible because the fluorescent antibody probe is not significantly depleted during the MALDI-TOF process. However, fluorescence imaging could be achieved either before or after MALDI-TOF mass-imaging with comparable results. Furthermore, pico-well plates could be imaged from the bottom (through fiber optics) or from the top, with comparable results, despite the presence of MALDI matrix crystals on the top surface of the plate. Finally, fluorescence imaging before matrix application and before MALDI-TOF was also possible, again with similar results.
(312) Results:
(313) The same region of the pico-well plate was imaged by both MALDI-TOF and fluorescence. Results are shown in
(314)
(315) Taken together, these data show strong concordance between the mass tag encoding the p53 beads in the MALDI-TOF mass image and the fluorescence image of the anti-p53 antibody probe, and that this concordance is non-random (due to lack of overlap from “Marker Beads” and negative control blank beads).
Example 17. MALDI-TOF Mass-Imaging of Multiple Unique Photocleavable and Non-Cleavable Mass-Tag Encoded Bead Species in an Array: Synchronizing Mass-Image for Bead Identification with Fluorescence Antibody Detection of a Bead-Bound Bait Protein
(316) In this Example, 10 distinct species of beads, each carrying a unique peptide mass tag, and 1 of the 10 additionally carrying bound recombinant p53 protein (“bait”), were prepared separately, pooled and then probed with a fluorescently labeled antibody (“prey”) to detect the recombinant p53 protein. An 11.sup.th bead species was then spiked in as “Marker Beads”. These Marker Beads carried a unique peptide mass tag for identification and the bead surface was covalently labeled with a fluorophore having different and distinguishable spectral properties than that used on the antibody probe. The pool of 11 bead species was randomly arrayed in the pico-well plates and then imaged by mass and by fluorescence.
(317) Finally, it should be noted that 10 of the 11 peptide mass-tags used were attached to the beads by a photocleavable biotin (including on p53 beads and Marker Beads), while the 11.sup.th was attached via a modified reduced-affinity non-cleavable biotin, in order to additionally demonstrate the possibility of using non-cleavable affinity-bound mass tags.
(318) Preparation of NeutrAvidin Coated Beads
(319) NeutrAvidin coated 34 micron agarose beads were prepared as in Example 5. As described below, these NeutrAvidin beads were used to prepare both the “Dual Affinity Beads” and the “Marker Beads”.
(320) Preparation of Dual Affinity Beads Coated with NeutrAvidin Directly and Indirectly with the Anti-HSV Tag Capture Antibody
(321) The aforementioned NeutrAvidin beads were then loaded a biotin labeled anti-HSV tag polyclonal antibody as follows: Goat anti-HSV tag polyclonal antibody was purchased from Bethyl Laboratories (Montgomery, TX), provided at 1 mg/mL in PBS. 800 μL of this antibody solution (800 μg) was then mixed with 1/9.sup.th volume of 1M sodium bicarbonate. The antibody was biotin labeled by adding a 10-fold molar excess of EZ-Link-Sulfo-NHS-LC-Biotin (Thermo-Fisher-Pierce, Rockford, IL) and reacting for 30 min with gentle mixing. The reaction was quenched by adding 1/9.sup.th volume of 1M glycine for 15 min with gentle mixing. The labeled antibody was then purified on a PD MidiTrap G-25 desalting column against TBS (see Example 3 for buffer) and according to the manufacturer's instructions (GB Healthcare Life Sciences, Piscataway, NJ). The purified and biotin-labeled antibody was then diluted to 0.15 μg/μL in TBS-T (see Example 3 for buffer). Using this solution, the NeutrAvidin beads were coated at a ratio of 12 μg of the biotin labeled anti-HSV tag antibody per each 1 μL of actual bead pellet volume for 30 min with gentle mixing. Beads were then washed 4× briefly with an excess of TBS-T using the aforementioned Filtration Devices (see Example 3 for devices). Beads were stored as a 20% (v/v) suspension at +4° C.
(322) Note that because the antibody does not saturate all the biotin binding sites on the NeutrAvidin beads, it was therefore possible to additionally load biotin labeled peptide mass tags onto the beads (see later steps in this Example).
(323) Preparation of “Marker Beads” Coated with NeutrAvidin Directly and Labeled with Fluorescence
(324) Performed as in Example 6 except that the aforementioned NeutrAvidin beads (no bound antibody) were used for fluorescence labeling and the Cy5-NHS activated (primary amine reactive) fluorescent dye labeling reagent was used (GE Healthcare Life Sciences, Piscataway, NJ).
(325) Preparation of Photocleavable (PC) and Non-Cleavable Biotin Labeled Peptide Mass Tags
(326) Performed as in Example 5 except that the following commercially available peptides were used for labeling (peptide sequence in brackets) (all peptides purchased for labeling were from AnaSpec, Fremont, CA, except for Bradykinin which was from Sigma-Aldrich, St. Louis, MO):
(327) TABLE-US-00003 1. Heparin-Binding Peptide V [Trp-Gln-Pro-Pro-Arg-Ala-Arg-Ile]; 1023 Da 2. Alpha-Bag Cell Peptide (1-9) [Ala-Pro-Arg-Leu-Arg-Phe-Tyr-Ser-Leu]; 1122 Da 3. [D-Phe7]-Bradykinin [Arg-Pro-Pro-Gly-Phe-Ser-D-Phe-Phe-Arg]; 1111 Da 4. Antioxidant Peptide B [Thr-Arg-Asn-Tyr-Tyr-Val-Arg-Ala-Val-Leu]; 1254 Da 5. Beta-Casomorphin (1-6), Bovine [Tyr-Pro-Phe-Pro-Gly-Pro]; 676 Da 6. [Leu8, Des-Arg9]-Bradykinin [Arg-Pro-Pro-Gly-Phe-Ser-Pro-Leu]; 870 Da 7. [Ile-Ser]-Bradykinin (T-Kinin) [Ile-Ser-Arg-Pro-Pro-Gly-Phe-Ser-Pro-Phe-Arg]; 1260 Da 8. [Des-Arg1]-Bradykinin [Pro-Pro-Gly-Phe-Ser-Pro-Phe-Arg]; 905 Da 9. LRRASLG [Leu-Arg-Arg-Ala-Ser-Leu-Gly]; 772 Da 10. Thrombin Receptor (42-48) Agonist, Human [Ser-Phe-Leu-Leu-Arg-Asn-Pro]; 846 Da 11. Bradykinin [Arg-Pro-Pro-Gly-Phe-Ser-Pro-Phe-Arg]; 1060 Da
(328) All above peptides were labeled on their N-terminus with the amine-reactive PC-Biotin-NHS reagent, as detailed in Example 5, except for “Alpha—Bag Cell Peptide (1-9)”, which was labeled with the amine-reactive DSB-X™ Biotin SSE labeling reagent (Invitrogen, Carlsbad, CA) using the same protocol as in Example 5. DSB-X™ Biotin is a modified biotin derivative that has reduced affinity for (strept)avidin, believed to be several orders of magnitude weaker, and therefore can be dissociated from (strept)avidin under comparatively mild conditions [Hirsch, Eslamizar, Filanoski, Malekzadeh, Haugland and Beechem (2002) Anal Biochem 308: 343-57]. While the native biotin moiety (ring structure) of the PC-Biotin labeled peptide mass tags is poorly dissociated from NeutrAvidin beads by the MALDI-TOF process (see Example 5), it was expected that the lower affinity binding of DSB-X™ Biotin would allow efficient dissociation of the mass tags via the denaturing MALDI-TOF matrix solution and/or by the energy introduced by the MALDI laser.
(329) Binding of PC-Biotin Peptide Mass Tags to Dual Affinity Beads
(330) Performed as in Example 6 except that 150 μL of PC-Biotin or DSB-X™ Biotin peptide mass tag solution, at a concentration of 5 pmoles/μL (750 pmoles peptide mass tag), was added to a 1 μL bead pellet volume (30,000 beads); for a final added amount of biotin-labeled peptide mass tag of 25 fmoles per bead. As in Example 6, this step is to capture the biotin-labeled peptide mass tags on the beads by way of the NeutrAvidin coating on the beads (peptide capture efficiency not measured). In this Example, 11 separate batches of beads were prepared, each batch uniquely loaded with 1 of the 11 aforementioned peptide species, creating 11 pure populations of unique mass tag encoded beads. The Bradykinin mass tag was loaded onto the aforementioned “Marker Beads” (NeutrAvidin beads with direct fluorescent label attached to beads) while all other mass tags were loaded onto the aforementioned “Dual Affinity Beads” (NeutrAvidin beads additionally containing a common capture antibody for tagged recombinant proteins).
(331) Binding of Recombinant Protein as “Bait” to Dual Affinity Beads
(332) Although in some embodiments all mass tagged bead species can additionally carry a “bait” molecule such as a protein, n this Example, with the exception of the “Alpha—Bag Cell Peptide (1-9)” and “Heparin—Binding Peptide V” mass tag encoded beads, the other 10 mass tagged bead species we not loaded with cell-free expressed recombinant protein and thus not subjected to this portion of the procedure:
(333) Beads with the “Alpha—Bag Cell Peptide (1-9)” identification tag were loaded with cell-free expressed recombinant human p53 protein essentially as in Example 6 with the following exceptions: Treatment of the beads with the cell-free expression reaction for recombinant p53 protein capture was performed as in Example 6, except that immediately following completion of the expression reaction (prior to mixing with beads), protease inhibitor was additionally added to the expression reaction to reduce the possibility of subsequent proteolytic degradation of the peptide mass tags on the beads (“Complete Mini Protease Inhibitor Cocktail Tablets” form Roche Applied Science, Indianapolis, IN, Catalog Number 11836153001; Stock Solution=1 mini-tablet in 1 mL of purified water; add 1/10 volume of Stock Solution to completed cell-free protein expression reactions). As in Example 6, the mechanism of capture of the cell-fee expressed recombinant p53 protein on the Dual Affinity Beads is by way of the anti-HSV antibody coating an the beads and the C-terminal HSV epitope tag present in the recombinant p53. Finally, after loading the recombinant p53 protein onto the “Alpha—Bag Cell Peptide (1-9)” encoded beads, washing was 4×400 μL briefly with TBS-T. Beads with the “Heparin—Binding Peptide V” identification tag were also subjected to the same above procedure, except that a blank cell-free expression reaction was used as a negative control instead of a p53 cell-free expression reaction. In this case, the cell-free expression reaction lacked the cognate expressible p53 DNA.
(334) Pooling Beads and Fluorescence Detection of p53 Beads
(335) All bead species, except “Marker Beads” which were added later, were then pooled in equal numbers (10 species). The pooled beads were then probed with a fluorescently labeled anti-p53 antibody to specifically detect the p53 containing beads. The fluorescently labeled anti-p53 antibody was prepared in the same manner as the fluorescently labeled NeutrAvidin described in Example 7, except that the anti-p53 antibody (clone BP53-12, Santa Cruz Biotechnology, Santa Cruz, CA) was used at 0.2 mg/mL in PBS and the Alexa Fluor® 594 SSB labeling reagent (Invitrogen, Carlsbad, CA) was used. The anti-p53 fluorescence antibody probing was performed similar to as in Example 3 in the section headed “Verification of Bound Recombinant Proteins”. After fluorescence probing, “Marker Beads” were spiked in at an approximate equal ratio to the other bead species.
(336) Mass Tag Photocleavage and MALDI-TOF Mass Spectrometry Imaging of Beads
(337) Performed in the pico-well plates as in Examples 5 & 7.
(338) Fluorescence Imaging of Pico-Well Plates
(339) Ater MALDI-TOF mass-imaging, the pico-well plates were fluorescently imaged using a GenePix 4200 laser based microarray scanner (Molecular Devices, Sunnyvale, CA). This is possible because the fluorescent antibody probe is not significantly depleted during the MALDI-TOF process. However, fluorescence imaging could be achieved either before or after MALDI-TOP mass-Imaging with comparable results. Furthermore, pico-well plates could be imaged from the bottom (through fiber optics) or from the top, with comparable results, despite the presence of MALDI matrix crystals on the top surface of the plate. Finally, fluorescence imaging before matrix application and before MALDI-TOF was also possible, again with similar results.
(340) Results:
(341) The same 4.8×2.8 mm region of the pico-well plate was imaged by both MALDI-TOF and fluorescence (13.44 mm; roughly 6,000 wells). Results are shown in
(342) In
(343) Data in
(344) Finally, comparison of florescence spot sizes with the MALDI-TOF images of both the “Marker Beads” and p53 beads suggests that the spatial resolution of the MALDI-TOF imaging approaches that of the fluorescence imaging (fluorescence scanned at 10 micron/pixel resolution).
Example 18. Colorimetric Detection of Beads in Pico-Well Plates
(345) In this Example, the ability to detect 34 micron diameter beads in the wells of pico-well plates is shown. In this case, the aforementioned 34 micron diameter agarose beads were coated with streptavidin-HRP and incubated in a precipitating chromogenic HRP substrate (TrueBlue Peroxidase Substrate; KPL Inc., Gaithersburg, Maryland). The now opaquely colored beads were deposited into the wells of commercially obtained pico-well plates (PicoTiter™ Plates designed for 454 Life Sciences GS FLX DNA Sequencer; Roche Applied Science, Indianapolis IN). The PicoTiter™ Plates are similar to the pico-well plates used in previous Examples, except that the slide (plate) thickness is 2 mm instead of 1 mm. Opaquely colored beads were imaged in the wells of the PicoTiter™ Plates using a visible light based ArrayIt SpotWare™ Flatbed Colorimetric Microarray Scanner (TeleChem International, Inc. ArrayIt Division, Sunnyvale, CA). Results shown in
(346) In some embodiments of the technology described in this patent, colorimetric detection of the beads could be used to visualize probe or “prey” binding to “bait” molecules on the beads. Alternatively, colorimetrically detected beads could serve as “landmarks” to direct the MALDI-TOF instrument to specific wells or coordinates in the array.
FIGURES AND TABLES
Description of the Figures for Experimental Examples
(347)
(348)
(349)
(350)
(351)
(352)
(353)
(354)
(355)
(356)
(357)
(358)
(359)
(360)
(361)
(362)
(363)
(364)
(365) Opaquely colored beads were deposited into the wells of pico-well plates and imaged using a visible light based colorimetric microarray scanner. In the presented grayscale image (2 fields of view shown), beads in the wells are identified by their dark color (dark spots), whereas empty wells appear as clear (white spots).
Description of the Figures for Specifications
(366)
(367)
(368)
(369)
(370)
(371)
(372)
(373)
(374)
(375)
(376)
(377)
(378)