Isotopically-encoded nanoparticles for multimodal high-order multiplexed detection and imaging

Abstract

A system of barcoding isotopically encoded particles in combination with elemental analyses and imaging that includes a particulate matrix, at least one isotope label contained in the particulate matrix, where the isotope label operates as i) an elemental identifier, ii) a mass identifier, or iii) an elemental identifier and a mass identifier, where the matrix operates as multi-digit particulate barcodes, at least i) a mass-based imager, ii) an elemental analyzer, iii) or the mass-based imager and the elemental analyzer, and a debarcoding algorithm and an automated machine learning analysis algorithm programmed on a computer to computational extract the multi-digit particulate barcodes for quantification of spatial nanotag distributions in ion beam imaging areas.

Claims

1) A system of barcoding isotopically encoded particles in combination with elemental analyses and imaging, compromising: a) a particulate matrix; b) at least one isotope label contained in said particulate matrix, wherein said isotope label operates as i) an elemental identifier, ii) a mass identifier, or iii) said elemental identifier and said mass identifier, wherein said matrix operates as multi-digit particulate barcodes; c) at least i) a mass-based imager, ii) an elemental analyzer, iii) or said mass-based imager and said elemental analyzer; and d) a debarcoding algorithm and an automated machine learning analysis algorithm programmed on a computer to computational extract said multi-digit particulate barcodes for quantification of spatial nanotag distributions in ion beam imaging areas.

2) The system according to claim 1, wherein said particulate matrix is selected from the group consisting of a metal(loid) chalcogen, a metalloid oxide, silica, titanium oxide, tantalum oxide, a soft nanoparticle, a liposome, a micelle, and a lipid nanoparticle.

3) The system according to claim 1, wherein said multi-digit nanoparticle-based barcodes comprise a combinatorial incorporation of an isotope into said silica nanoparticle matrix.

4) The system according to claim 3, wherein said isotopes are selected from the group consisting of halogen, chalcogen, pnictogen, metal isotopes, .sup.2H, .sup.15N, .sup.19F, .sup.79/81Br, and .sup.127I.

5) The system according to claim 4, wherein said isotopically enriched molecular scaffold for said .sup.2H comprises N-ethyl-d5-maleimide.

6) The system according to claim 4, wherein said isotopically enriched molecular scaffold for said .sup.15N comprises L-arginine-.sup.15N.sub.4.

7) The system according to claim 4, wherein said isotopically enriched molecular scaffold for said .sup.19F comprises trimethoxy(3,3,3-trifluoropropyl)-silane.

8) The system according to claim 4, wherein said isotopically enriched molecular scaffold for said .sup.79/81Br comprises eosin-maleimide.

9) The system according to claim 4, wherein said isotopically enriched molecular scaffold for said .sup.127I comprises L-thyroxine.

10) The system according to claim 1, wherein a modified Stöber reaction is used to produce said silica nanoparticles having diameters in a range of 90 nm to 110 nm, wherein said modified Stöber reaction comprises a mixture of 100-nm silica nanoparticles comprising 0.7% (v/v) NH.sub.3, 4% (v/v) of the silica precursor tetraethyl orthosilicate, and 0.31% (v/v) 3-mercaptopropyltrimethoxysilane (MPTMS) in 91% (v/v) aqueous isopropanol.

11) The system according to claim 1, wherein N-ethyl-d5-maleimide and eosin-maleimide were reacted with 3-mercaptopropyltrimethoxysilane (MPTMS) in dimethylsulfoxide (DMSO) under ambient conditions before said metalloid oxide silica nanoparticle were synthesized, wherein L-thyroxine was conjugated to said MPTMS using a heterobifunctional linker succinimidyl 4-N-maleimidomethyl-1-cyclohexane-carboxylate (SMCC) in a 1:1.1:1 ratio in DMSO for uniform and covalent incorporation within said silica nanoparticle matrix.

12) The system according to claim 1, wherein trimethoxy(3,3,3-trifluoropropyl)-silane and silane appended N-ethyl-d5-maleimide, eosin-maleimide and L-thyroxine are mixed into a Stöber reaction mixture in any (.sup.2H:.sup.19F:.sup.79/81Br:.sup.127I) isotope ratio to yield isotopically encoded silica nanotags.

13) The system according to claim 1, wherein trimethoxy(3,3,3-trifluoropropyl)-silane and silane appended N-ethyl-d5-maleimide, eosin-maleimide and L-thyroxine are mixed into Q Stöber reaction mixture in any 1:1:1:1 (.sup.2H:.sup.19F:.sup.79/81Br:.sup.127I) isotope ratio to yield isotopically encoded silica nanotags.

14) The system according to claim 1, wherein ionic metal isotopes are combinatorially mixed into a dispersion of silica nanoparticles to generate metal-based isotopically encoded silica nanotags.

15) The system according to claim 1, wherein said mass-based imaging platform is selected from the group consisting of multiplexed ion-beam imaging, and mass cytometry.

16) The system according to claim 1, wherein said elemental analysis platform is selected from the group consisting of X-ray fluorescence, energy dispersive X-ray spectroscopy, and laser induced breakdown spectroscopy.

17) The system according to claim 1, wherein a mixture of said isotopically encoded nanotags are applied to a substrate gold-coated silicon substrate for use in a multi-ion beam image.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

[0017] FIGS. 1A-1C show (1A) A mixture of isotopically encoded nanotags on a gold-coated silicon substrate is raster-scanned using a cesium ion beam. Next, secondary elemental ions are analyzed using SIMS and spatially deconvoluted using debarcoding algorithms to provide quantitative information on the spatial distribution of the individual nanotags. (1B) A modified Stöber reaction that involves the addition of isotopically labeled silanes in the presence of tetraethyl orthosilicate (TEOS) and NH.sub.4OH in an aqueous isopropanol (IPA) solution was used to synthesize 100-nm isotopically encoded isotopically encoded silica nanoparticles. The four-digit barcodes are based on labeling of silica nanoparticles with .sup.2H, .sup.19F, .sup.79/81Br, .sup.127I, or combinations thereof. (1C) Molecular structures of the isotopically labeled silanes. .sup.2H—, .sup.79/81Br—, and .sup.127I-containing molecules were appended to the thiol-containing MPTMS either directly via straightforward maleimide chemistry in the case of the .sup.2H— and .sup.79/81Br-containing scaffolds or using the heterobifunctional linker SMCC for the .sup.127I-containing molecular scaffold L-thyroxine, according to the current invention.

[0018] FIG. 1D shows isotopically-encoded soft nanotags, where a liposome can be isotopically encoded by using isotopically-tagged lipids (see example for a deuterium-labeled lipid). Furthermore, the aqueous compartment (center) can be loaded with isotonic solutions of ((non-)radioactive)/(non-)metal ions as well as reporters (fluorophores, iron oxides) etc. or combinations thereof, according to the current invention.

[0019] FIG. 1E shows soft nanoparticles, where micelles are labeled with (hypertonic) sodium fluoride or potassium iodide solutions encapsulated within the micellar matrix and analyzed using MIBI.

[0020] FIG. 1F shows Hydrodynamic diameters of isotopically encoded silica nanotags. The hydrodynamic diameters of each preparation of nanotags was measured in water using nanoparticle tracking analysis, according to the current invention.

[0021] FIG. 2A shows MIBI of isotopically encoded nanotags, where isotopically encoded nanotags were imaged using MIBI. The MIBI data are generated by raster scanning of the samples with a cesium ion beam followed by secondary ion mass spectrometry. All silica-based nanotags were imaged in the .sup.28Si mass channel. MIBI data are presented for mass channels of .sup.2H, .sup.19F, .sup.79/81Br, and .sup.127I. The four-digit nanotags were assigned based on signal in the mass channel (‘1’) or no signal above background (‘0’). The secondary electron (SE) image reflects total electrons ejected from the sample. Scale bars, 1 μm, according to the current invention.

[0022] FIG. 2B shows intraparticle isotope distribution, where sequential cesium ion beam scans of the .sup.19F-labeled nanotags (0100) over 50-slices (Z, 5-10 nm step size) show uniform three-dimensional distribution of .sup.19F throughout the nanotag matrix. Scale bars, 500 nm, according to the current invention.

[0023] FIGS. 3A-3H show combinatorial nanotag barcoding with four isotope mass channels (3A-3B) Nanotags were imaged on a gold-coated silicon substrate. Shown are (3A) secondary electron beam image (SE) and b) merged MIBI image from mass channels corresponding to .sup.28Si, .sup.19F, .sup.79/81Br, and .sup.127I isotopes. Numbered arrows indicate nanotags with different barcodes. Scale bars, 2 μm. (3C) Higher magnification images of nanotags numbered in panels a and b. As expected, all but one barcode (i.e., 011) were detected. Scale bars, 100 nm. (3D) Histogram displaying the quantification per barcode of the 99 isotopically encoded nanotags detected in the field of view. (3E) Number of 001 and 101 nanotags detected using manual counting and machine learning. (3F) Principal component (PC) analysis of distributions of nanotags with barcodes color-coded as indicated. (3G) t-SNE plot of nanotag subtypes provide accurate classification, according to the current invention.

[0024] FIG. 4 shows additional mass labels used for isotope encoding of nanotags. Molecular structures of the scaffolds that were used to produce covalently incorporated .sup.15N-enriched nanotags, covalently incorporated natural-abundance Se nanotags (abundances of isotopes .sup.76Se, .sup.77Se, .sup.78Se, and .sup.80Se are 9.2%, 7.6%, 23.7%, 49.8%, respectively), and non-covalently incorporated, natural abundance Te nanotags (abundances of .sup.122Te, .sup.124Te, .sup.125Te, .sup.128Te and .sup.130Te isotopes are 2.6%, 4.7%, 7.1%, 18.8%, 31.7%, and 34.1%, respectively), according to the current invention.

[0025] FIG. 5 shows evaluation of .sup.15N as an isotope label. Indicated channels for an .sup.15N, .sup.19F, .sup.79/81Br and .sup.127I-encoded nanotag with 1111 barcode demonstrated that all isotopes signal overlapped between the channels, according to the current invention.

[0026] FIG. 6 shows MIBI results, where the top row shows MIBI results for nanotags that were covalently encoded with natural abundant Se. Since .sup.80Se is the most abundant isotope (˜49.8%) we selected the 80-Da mass channel to image .sup.80Se. Signal in .sup.28Si and .sup.80Se overlap indicating that the selenium signal is associated with the silica nanoparticles. No significant signal was detected in the 130-Da mass channel, where .sup.130Te would be detected. The bottom row shows the MIBI results for nanotags that were non-covalently labeled with Te. Te was imaged in the 130-Da channel as .sup.130Te is the most abundant isotope (at 34%), according to the current invention.

[0027] FIGS. 7A-7B show stoichiometric labeling with distinct concentration levels of .sup.81Br isotopes. (7A) Nanotag images for 1× (top row) and 2x (bottom row) concentrations of .sup.81Br encoded in nanotags (see Table 51 for clarification on how different molar ratios of the isotopes were incorporated). Scale bars, 1 μm. (7B) Bar plot of the median distribution of .sup.81Br/.sup.28Si ratio for Br-1× (n=9 nanotags) and Br-2× (n=29 nanotags). Individual nanotags were digitally segmented, isolated, and quantified for total ion signal in .sup.81Br and .sup.28Si signals per nanotag. Student's t-test, *** p<6.45×10.sup.−5, according to the current invention.

[0028] FIGS. 8A-8B show stochiometric barcoding with ratios of .sup.19F to .sup.79/81Br isotopes. (8A) The top row shows MIBI images of a nanotag mixture of .sup.19F, .sup.79/81Br, and .sup.127I at a ratio of 1:1:1, and the bottom images are of a mixture at a ratio of 0.5:2:1. Scale bars, 1 μm. (8B) Ratio of indicated isotopes to .sup.28Si for the 1:1:1 mixture (n=8 nanotags) and for the 0.5:2:1 mixture (n=58 nanotags), according to the current invention.

[0029] FIGS. 9A-9D show machine-learning analysis of nanotag images. (9A) Absolute number of nanotags detected using manual counting and machine learning for nanotag mixtures of .sup.19F, .sup.79/81Br, and .sup.127I at ratios of 1:1:1 (n=8) and 0.5:2:1 (n=58). The machine learning pipeline was based on support vector machine (SVM)-based training and prediction. (9B-9D) Classification results of SVM for b).sup.79/81Br/.sup.28Si vs. .sup.19F/.sup.28Si, c) .sup.127I/.sup.28Si vs. .sup.19F/.sup.28Si, and d) .sup.127I/.sup.28Si vs. .sup.79/81Br/.sup.28Si. The corresponding 1:1:1 and 0.5:2:1 barcoded nanotags are shown, which were separated by a Gaussian kernel on the decision surface. Individual nanotags segmented, digitally quantified for total ion signal per channel as a feature vector. Nanotags were automatically classified based on ratiometric isotope barcoding scheme. The selected fields denote classification region for each nanotag, where the outlines indicate the expected region for 1:1:1-ratio barcoded nanotags, and the 0.5:2:1-ratio barcoded nanotags, according to the current invention.

DETAILED DESCRIPTION

[0030] The current invention provides a system of barcoding isotopically encoded particles in combination with elemental analyses and imaging that includes a particulate matrix, at least one isotope label contained in the particulate matrix, where the isotope label operates as i) an elemental identifier, ii) a mass identifier, or iii) an elemental identifier and a mass identifier, where the matrix operates as multi-digit particulate barcodes, at least i) a mass-based imager, ii) an elemental analyzer, iii) or the mass-based imager and the elemental analyzer, and a debarcoding algorithm and an automated machine learning analysis algorithm programmed on a computer to computational extract the multi-digit particulate barcodes for quantification of spatial nanotag distributions in ion beam imaging areas.

[0031] According to one aspect of the invention, the particulate matrix includes a metal(loid) chalcogen such as a metalloid oxide (i.e., silica), a metal oxide (i.e., titanium oxide, tantalum oxide, etc.), a soft nanoparticle, a liposome, a micelle, or a lipid nanoparticle, or a combination thereof.

[0032] According to embodiments of the invention, the mass-based imaging platform includes multiplexed ion-beam imaging, or mass cytometry. In further embodiments of the invention, the elemental analysis platform includes X-ray fluorescence, energy dispersive X-ray spectroscopy, and laser induced breakdown spectroscopy.

[0033] One embodiment includes a nano-barcoding platform for multiplexed ion beam imaging (MIBI) using secondary ion beam spectrometry that utilizes fabricated isotopically encoded nanotags. In one embodiment, the invention uses combinatorial isotope distributions in 100-nm-sized nanotags to expand the labeling palette to overcome the spectral bounds of mass channels. In an exemplary embodiment, a four-digit (i.e., 0001 to 1111) barcoding scheme is provided to detect 16 variants of .sup.2H, .sup.19F, .sup.79/81Br and .sup.127I elemental barcode sets that are encoded in silica nanoparticle matrices. A computational debarcoding method and an automated machine learning analysis approach are provided to extract barcodes for accurate quantification of spatial nanotag distributions in large ion beam imaging areas up to 0.6 mm.sup.2. Isotopically encoded nanotags should boost the performance of mass imaging platforms such as MIBI and other elemental-based bioimaging approaches.

[0034] According to the current invention, a nano-barcoding platform that is based on metal and metalloid oxide nanoparticles is provided. The method relies on combinatorial incorporation of halogen, chalcogen, and pnictogen isotopes of low biological abundance (i.e., .sup.2H, .sup.15N, .sup.19F, .sup.79/81Br, and .sup.127I) into a silica nanoparticle matrix to produce isotopically encoded nanotags (FIG. 1A). Here, the isotopically enriched molecular scaffold for the .sup.2H comprises N-ethyl-d5-maleimide. Further, the isotopically enriched molecular scaffold for the .sup.15N comprises L-arginine-.sup.15N.sub.4. Additionally, the isotopically enriched molecular scaffold for the .sup.19F comprises trimethoxy(3,3,3-trifluoropropyl)-silane. Still further, the isotopically enriched molecular scaffold for the .sup.79/81Br comprises eosin-maleimide. In addition, the isotopically enriched molecular scaffold for the .sup.127I comprises L-thyroxine. The metalloid oxide silica is provided as the matrix for the nanoparticle-based barcodes, because silica precursors and methods for synthesis of silica nanoparticles of controlled sizes are available and silica surface modifications to enable antibody conjugation are understood. A modified Stöber reaction is used to produce silica nanoparticles with diameters of in a range of 90 nm to 110 nm. According to one embodiment, a reaction mixture for the synthesis of 100-nm silica nanoparticles contains 0.7% (v/v) NH.sub.3, 4% (v/v) of the silica precursor tetraethyl orthosilicate, and 0.31% (v/v) 3-mercaptopropyltrimethoxysilane (MPTMS) in 91% (v/v) aqueous isopropanol. After a reaction time of 30 minutes under ambient conditions, the 100-nm silica nanoparticles were collected using centrifugation (5 min at 10,000 g) and washed with 100% ethanol to afford 100-nm silica nanoparticles.

[0035] In one aspect of the invention, isotopically encoded silica nanotags are provided by linking silane-appended isotopically enriched molecular scaffolds to yield four-digit barcodes (FIG. 1B). N-ethyl-d5-maleimide, trimethoxy(3,3,3-trifluoropropyl)-silane, eosin-maleimide, and L-thyroxine, are provided as the molecular scaffolds for .sup.2H, .sup.19F, .sup.79/81Br, and .sup.127I, respectively. Due to the low natural abundance of .sup.2H, a .sup.2H-enriched scaffold is used. The other scaffolds were based on isotopes of high natural abundance: Natural abundances of .sup.19F, .sup.79/81Br, and .sup.127I are 100%, 51/49%, and 100%, respectively. To enable uniform, covalent incorporation within the silica nanoparticle matrix, N-ethyl-d5-maleimide and eosin-maleimide were first reacted overnight with MPTMS in dimethylsulfoxide (DMSO) under ambient conditions and then silica nanoparticles were synthesized. L-thyroxine was conjugated to MPTMS using the heterobifunctional linker succinimidyl 4-N-maleimidomethyl-1-cyclohexane-carboxylate (SMCC) in a 1:1.1:1 ratio in DMSO (FIG. 1C). Silane-appended eosin-maleimide and L-thyroxine derivatives were used without further purification. The concentrations of silane-appended isotopically enriched scaffolds were normalized based on the isotope abundance in the scaffold and were added independently or as a combination as a fraction of the 0.31% of the MPTMS volume fraction (Table 1). As shown using nanoparticle tracking analysis, the average hydrodynamic diameter of the nanotags was 105.5±8.1 nm with a coefficient of variance of 7.6% between different nano-barcodes (FIG. 1D).

TABLE-US-00001 TABLE 1 Isotopic encoding of silica nanotags based on normalized addition of the individual isotope scaffolds. # Isotope Barcode .sup.2H.sub.5 .sup.19F.sub.3 .sup.79/81Br.sub.4 .sup.127I.sub.4 MPTMS total 0 — 0000 0.31% 0.31% 1 H 1000 0.060% 0.25% 0.31% 2 F 0100 0.10% 0.21% 0.31% 8 Br 0010 0.075% 0.24% 0.31% 4 I 0001 0.075% 0.24% 0.31% 5 H, F 1100 0.060% 0.10% 0.15% 0.31% 6 H, Br 1010 0.060% 0.075% 0.18% 0.31% 7 H, I 1001 0.060% 0.075% 0.18% 0.31% 8 H, F, Br 1110 0.060% 0.10% 0.075% 0.075%  0.31% 9 H, F, I 1101 0.060% 0.10% 0.075% 0.075% 0.31% 10 H, F, Br, I 1111 0.060% 0.10% 0.075% 0.075% — 0.31% (1:1:1:1) 11 H, F, Br, I 2142 0.060% 0.05% 0.150% 0.075% — 0.33% (1:0.5:2:1) *The addition of the isotope scaffolds was normalized based on the molar isotope ratio between the scaffolds. For instance, silane-.sup.19F.sub.3 incorporate 3 moles of .sup.19F per 1 mole of scaffold and thus 4/3 × the molar amount of silane-.sup.19F.sub.3 relative to eosin-maleimide, which contains 4 moles of Br per 1 mole of scaffold, was added in the reaction mixture.

[0036] FIG. 1E shows soft nanoparticles, where micelles are labeled with (hypertonic) sodium fluoride or potassium iodide solutions encapsulated within the micellar matrix and analyzed using MIBI. FIG. 1F shows Hydrodynamic diameters of isotopically encoded silica nanotags. The hydrodynamic diameters of each preparation of nanotags was measured in water using nanoparticle tracking analysis, according to the current invention.

[0037] Next, gold-coated silicon substrates are prepared with a 200-nm thick gold layer coated on a 20-nm titanium adhesion layer (substrate dimensions: 7×7 mm, Silicon Valley Microelectronics) using the Innotec E-beam metal evaporation system. Gold was selected because the mass of gold is 197 Da (100% natural abundance), which should not interfere with MIBI of the silica nanoparticles. A dispersion of isotopically encoded silica nanoparticles in ethanol (2-5 μL) was placed on the gold-coated silicon substrate and air-dried overnight before ion beam imaging.

[0038] To validate the incorporation of the individual isotopes into the silica nanoparticle matrix, MIBI was performed on nanotags without isotope encoding or with single or all isotopes. Bare silica nanoparticles had positive signal only in the .sup.28Si mass channel, and no significant background was observed in the mass channels corresponding to .sup.2H, .sup.19F, .sup.79/81Br, or .sup.127I isotope labels. Since the barcoding system is based on combinatorial encoding with these four isotopes (FIG. 1B), the associated four-digit barcode for bare silica corresponds to 0000. Nanotags that were encoded with a single isotope had signal only in the mass channel of the respective isotope (FIG. 2A). The isotope distributions were analyzed within the silica nanoparticle matrix and it was found that the isotope-label for .sup.19F-encoded nanotags closely matched .sup.28Si indicating that the labels are uniformly distributed within the silica nanoparticle matrix (FIG. 2B). MIBI of the nanotags incorporating all selected isotopes demonstrated that all mass-channels recorded positive signals, corresponding to a 1111 barcode (FIG. 2). Of note, the high mass-spectral separation that is needed for the .sup.2H mass label identification resulted in lower sensitivity of .sup.2H detection, relative to sensitivities of the halogen isotopes, which had smaller aperture settings. Due to the dynamic range difference (˜10-fold) between the .sup.2H and halogen mass channels, the .sup.2H mass channel was excluded for quantification purposes.

[0039] To demonstrate the utility of the barcoding strategy, an exemplary isotopically encoded nanotag mixture was prepared containing the three-digit (.sup.19F, .sup.79/81Br, .sup.127I) barcode combinations 000, 100, 010, 001, 110, 101, and 111; this is all possible combinations except for 011 because of it was not included in the prepared nanotag mixture. 5 μl of this mixture was deposited on the gold-coated silicon substrate. A large raster scan (512×512 pixels) was performed with data from 10 scans collected at a scanning speed of 5 minutes per scan using a NanoSIMS device.

[0040] The secondary electron image showed that most nanotags were isolated (FIG. 3A), which proved ideal for digital quantification. In the merged MIBI image 99 nanotags were within the field of view when the .sup.28Si signal was used as a nanotag identifier (FIG. 3B). The barcode of each nanotag was extracted based on .sup.19F, r, .sup.79/81Br, and .sup.127I mass channel signals (FIG. 3C). Detected were 15, 21, 16, 14, 12, 7, and 14 counts for the three-digit barcodes (.sup.19F, .sup.79/81Br, .sup.127I) 000, 001, 010, 100, 101, 110, and 111, respectively (FIG. 3D).

[0041] Barcode assignment was then automated by an unsupervised machine learning algorithm. Each isotope channel was treated as a feature vector that was used in the training and prediction. A mathematical basis for support vector machine (SVM) was used to deconvolve the barcoded nanotags. Direct digital and manual quantification and machine learning-based prediction of nanotag barcodes (e.g., 001 debarcoded from 101) provided good agreement in the bar plot (FIG. 3E). Classified barcode sets were then visualized on the principal component analysis and t-Distributed Stochastic Neighbor Embedding (t-SNE) plots (FIGS. 3F-3G), demonstrating the accurate debarcoding of the presented highly multiplexed nanotag results.

[0042] In addition to .sup.2H, .sup.19F, .sup.79/81Br, and .sup.127I mass-labels, we also explored the incorporation of .sup.15N-enriched and natural abundance .sup.76/77/78/80Se— and .sup.122/124/125/128/130Te— containing scaffolds (FIG. 4). As a molecular scaffold for .sup.15N, arginine-.sup.15N.sub.4 was selected, which was conjugated to MPTMS using SMCC to enable covalent incorporation of the .sup.15N-scaffold into the silica matrix of the nanotags. The .sup.15N-.sup.19F-.sup.79/81Br-.sup.127I barcode (1111) was distinguished from the .sup.19F-.sup.79/81Br-.sup.127I barcode (0111) scheme (FIG. 5). For Se— and Te— incorporation, meso-chloro-substituted selenopyrylium and telluropyrylium scaffolds were used. The meso-chloro-substituted selenopyrylium was conjugated with MPTMS by reacting it overnight in DMSO at 72° C. in a 1:1 molar ratio to enable covalent incorporation into the nanotag matrix. The telluropyrylium scaffold was used without appending to silane. Se— and Te-encoded nanotags were prepared and imaged by MIBI demonstrating that .sup.80Se was detectable and overlapped with specific .sup.28Si host matrices, with no spillover into the .sup.130Te mass channel (FIG. 6). Similarly, specific signals were observed for .sup.130Te with minimal ion signal in the .sup.80Se mass channel (FIG. 6). Compared to .sup.80Se, which was silane-appended, non-covalent incorporation of the Te-scaffold yielded lower signal under identical imaging conditions due to lower levels of incorporation of the tellurium scaffold into the silica nanoparticle matrix.

[0043] Since selenium and tellurium have mass overlap with bromine and iodine, respectively, neither was further explored for use in the isotopically encoded nanotags mixtures. In contrast, .sup.2H and .sup.15N do not have mass overlap with any of the halogens that were successfully incorporated in the isotopically encoded nanotags. Moreover, since the sensitivity of .sup.2H and .sup.15N mass labels is related to the aperture setting, the sensitivity of both isotopes could be improved by increasing the stochastic ratio of .sup.2H and .sup.15N relative to the halogens. Incorporation of .sup.2H and .sup.15N in addition to .sup.19F, .sup.79/81Br, .sup.127I into the current set of nanotags will enable the generation of 2.sup.5 or 64 distinct barcodes.

[0044] As an alternative method for expansion of the barcode library, we also explored the incorporation of different stoichiometric ratios of the halogen isotopes (Table 1, entry 11). Nanotags prepared with .sup.19F—.sup.79/81Br-.sup.127I in a 1:1:1 stoichiometric ratio were able to be separated from those prepared in a 0.5:2:1 ratio (FIGS. 7A-7B and FIGS. 8A-8B). This was achieved by measuring the intensity of signal of each isotope relative to the signal of .sup.28Si-silica nanoparticle matrix. Machine-learning debarcoding successfully classified barcodes based on the ratiometric analysis (FIGS. 9A-9D).

[0045] In summary, isotopically encoded nanotags were synthesized that combinatorially incorporate .sup.19F, .sup.79/81Br, and .sup.127I to generate a library of nanobarcodes for multiplexed analysis in nanoscopic applications using cesium ion beams. The nanotags were uniformly labeled with the isotopes. The ratios of different nanotags in mixtures were successfully determined automatically via digital analysis and a machine-learning framework. Since silica surface modification is straightforward, the nanotags can be conjugated to analyte capturing moieties such as aptamers, peptides, or antibodies to enable highly sensitive and multiplexed analyte detection or imaging.

[0046] The present invention has now been described in accordance with several exemplary embodiments, which are intended to be illustrative in all aspects, rather than restrictive. Thus, the present invention is capable of many variations in detailed implementation, which may be derived from the description contained herein by a person of ordinary skill in the art. For example, the nanobarcodes can be conjugated to antibodies to enable high-level multiplexed detection of analytes during mass imaging-based histopathology or mass-cytometry. In addition, the nanobarcodes can be used to study analyte-biomolecule interactions where a subset of nanotags is labeled with the analyte and the other subset with the biomolecule of interest. By analyzing the proximity of the different nanotags using automated counting and interparticle distance measurements binders and non-binders for a specific analyte/biomolecule can be identified in tandem. Moreover, nanobarcodes can be used as anticounterfeiting labels for art, money, or any other object to establish the authenticity using mass-label or non-destructive elemental analyses (i.e. XRF) approaches.

[0047] All such variations are considered to be within the scope and spirit of the present invention as defined by the following claims and their legal equivalents.