Multiplexed analyte screening in biological spatial assays

Abstract

Disclosed herein are various example systems, methods, and compositions, for extremely high-resolution in situ spatial analysis of a sample, with particular applicability to biological materials, by strategically multiplexing a. plurality of different imaging techniques. For instance, certain inputs can be configured to elicit multiple different types of output signals from the sample simultaneously, each of which encodes a different attribute of the sample. These various outputs, once detected, decoded, and assembled, provide for a more- robust analysis than their individual techniques, i.e., if performed sequentially. As just one illustrative example, the infrared light involved in photothermal-infrared (PI) microscopy can also be used, to simultaneously trigger fluorescence-encoded infrared (FEIR) excitation within the sample. Pegging the excitation signals to specific biomarker targets can drastically enhance the combined 3D imagery' that results. Additional techniques may be integrated (i.e., higher-order multiplexing) to resolve the imagery even further.

Description

MULTIPLEXED ANALYTE SCREENING IN BIOLOGICAL SPATIAL ASSAYSTECHNICAL FIELD

[0001] The present disclosure relates to vibrational spectro-microscopy for in situ chemical analysis of biological samples.BACKGROUND

[0002] Highly multiplexed imaging of molecular biomarker targets within individual cells can provide a comprehensive understanding of fundamental biology by deciphering the interactions of various molecular pathways simultaneously. Multiplexed measurements are thus attracting considerable interest from researchers in microbial ecology, synthetic biology, biomanufacturing, immunology, oncology, cell and gene therapy, drug discovery, tissue engineering, patient stratification, and personalized medicine. To facilitate the identification, localization, and quantification of numerous molecular biomarkers including critical cellular parameters related to the genome, epigenome, transcriptome, proteome and metabolome requires specialized instrumentation, specifically designed for rapid, high-resolution imaging with the ability to differentiate molecular species with sufficient signal -to-noise ratio (SNR).

[0003] The comprehensive expression analysis of deoxyribose nucleic acid (DNA), ribonucleic acid (RNA), peptides, proteins, metabolites, and secondary metabolites in cells and their surrounding microenvironments including in tissues is critical for understanding mechanisms of biology in natural environments, controlled environments, fermenters, bioreactors, laboratories, and clinical settings. Of particular importance is the in situ spatial analysis of these biomarkers to evaluate their dynamics, structural relationships, and interactions in heterogeneous and multicellular samples without disrupting their native spatial organization. While traditional optical microscopy provides morphological information it does not capture functional or molecular information. Analytical techniques that can provide a broad view of biomarker distribution and overall composition are poised to solve major questions surrounding various complex biological systems. Despite extensive efforts to develop such suitable analytical methods, conventional in situ spatial biology assays including fluorescence in situ hybridization (FISH) and immunofluorescence (IF) remain constrained in the number of biomarkers that can be probed in each round of labeling due to spectral overlap in the emission profiles of the fluorescent probes used in the assays. While fluorescent probes can reach single molecule sensitivity, they can significantly alter biological activities, especially when applied to small molecules, and are subject tophotobleaching and phototoxicity effects. To facilitate greater multiplexing requires a cyclic labeling procedure in which samples are fluorescently labeled, imaged, and washed, such that a new round of labels can be applied to progressively build an information rich, multichannel image. Multiple rounds of labeling can compromise sample integrity and the long processing times limits throughput, requiring the use of automated image acquisition hardware and fluidic handling systems.

[0004] At present, there is an ongoing need to facilitate probe multiplexing such that a greater number of probes can be deployed in each labeling cycle while providing a method to identify and quantify the presence of each probe. Concurrently, methods for improving spatial resolutions within a biological sample beyond the diffraction-limit is essential to achieve better localization and quantification of both labeled and unlabeled (endogenous) molecular biomarker targets. Existing super-resolution microscopy technologies will benefit from integration with vibrational spectro-microscopy to achieve high-multiplexed, superresolution microscopy.SUMMARY OF THE DISCLOSURE

[0005] The present disclosure is directed to a set of novel imaging systems, devices, associated techniques and methods, compositions, and products of manufacture for high- resolution, high-multiplex spatial mapping of molecular biomarker targets within biological samples.

[0006] One such method includes, as a high-level example: (1) exciting, by a narrow- spectral-linewidth infrared (IR) light source, specifically in the mid-infrared (MIR) spectral range of the electromagnetic spectrum, molecular vibrational resonances within a biological sample; (2) capturing images while illuminating the biological sample with another light source having a shorter wavelength than the MIR light source; and (3) resolving, based on the captured images, a set of spatial features within the biological sample, including a morphology and an organization of biomarker targets within the biological sample. In some implementations, the excited vibrational resonances belong to molecules that are part of endogenous biomarker targets native to the biological sample. Additionally or alternatively, the biomarker targets may be labeled with one or more probes encoded by functional group(s) that exhibit distinct molecular vibrational resonance characteristics, referred to herein as “molecular probes.” These molecular probes are defined by their characteristics, numbers, orders, positions, patterns, configurations, and / or interactions relative to the one or more biomarker targets. In examples described below, microscopic analysis of such probes isachieved through measurements of the vibrational photothermal effect, referred to herein as “Photothermal-Infrared” (PI) microscopy.

[0007] In some examples, the biomarker targets are labeled with one or more probes having luminescent, light-emitting moieties with defined spectral absorption / emission characteristics that are in communication with a functional group with distinct molecular vibrational resonance characteristics. In the context of this disclosure, the term “moiety” refers to a fluorescent molecule that typically, though not necessarily, contains conjugated double bonds such as in an aromatic structure, as will be readily understood by those of ordinary skill in the art. These luminescent probes are activated when the photon energy from the MIR light source matches the vibrational resonance of the functional group, and, the photon energy from a second light source with a shorter wavelength is sufficient to excite an electronic transition in the luminescent probe from the intermediate vibrational state to the excited electronic state in a process referred to herein as fluorescence-encoded infrared (FEIR) excitation. After a characteristic lifetime in the excited state, these luminescent probes, referred to herein as FEIR probes, emit a fluorescence photon and return to the ground energy state. In some examples, the FEIR probes are defined by their characteristics, numbers, orders, positions, patterns, configurations, and interactions relative to the one or more biomarker targets.

[0008] In some examples, the biological sample is stained or labeled with a plurality of probes of one kind, or a plurality of different types of probes, either simultaneously or sequentially. In some examples, the labeling is performed in multiple rounds, or cycles, whereby the method includes placing the biological sample in a compartment that allows fluid flow and facilitates sample-labeling and / or sample-washing processes. In some such examples, the compartment includes a system of microfluidic channels to automate the application of various buffers, reagents, labels, and other sample treatments. In some examples, the probe labeling and imaging cycles are performed in a barcoding manner such that detection errors are reduced to improve accuracy in biomarker assignment even during high biomarker multiplexing. In some other examples, the compartment enables control of environmental parameters, such as temperature and atmospheric composition.

[0009] The present disclosure further describes a set of “super-resolution” methods for spatially determining, visualizing, quantifying, or localizing the presence of biomarker targets beyond the standard diffraction limit, while simultaneously capturing chemical information by exciting molecular vibrational resonances based on IR-absorption spectroscopy. These super-resolution microscopy (SRM) techniques allow biomarker targets to be localized withthree-dimensional (3D) spatial resolutions of about 200 nanometers (nm), about 100 nm, about 50 nm, or even less. In some implementations, the excited vibrational resonances belong to molecules that are part of unlabeled, endogenous biomarker targets native to the biological sample. Optionally, the excited vibrational resonances may arise from the presence of a molecular probe used to label one or more biomarker targets within the sample. Optionally, the probe may be a FEIR probe and may additionally contain a luminescent, light-emitting moiety with distinct or defined spectral absorption, emission, and / or luminescence lifetime characteristics.

[0010] Additionally, the present disclosure provides for certain compositions, and products of manufacture that include, for instance: (1) a plurality of primary-target-molecule probes, each primary-target-molecule probe comprising: (a) a biorecognition motif with a complementary region which can selectively bind to a specific portion or region of the molecular biomarker target in the sample, (b) an optional functional group with a distinct molecular vibrational resonance comprising a unique absorption feature for direct labeling, (c) an optional light-emitting moiety in communication with the functional group, and (d) an optional extension element or a “read-out” or “adapter” element that can selectively bind to a specific portion or region of a secondary probe for indirect labeling; (2) a plurality of secondary probes, each comprising: (a) a region that binds specifically to the corresponding extension element on the primary probe, and (b) a functional group with a distinct molecular vibrational resonance comprising a unique absorption feature, and (c) an optional lightemitting moiety in communication with the functional group.

[0011] Optionally, each primary and / or secondary probe further includes a light-emitting moiety with a unique luminescence spectrum and / or lifetime characteristic. In some examples, at least one light-emitting moiety includes a fluorophore. In some examples, one or more of the primary-target-molecule probes includes an oligonucleotide. In some examples, one or more of the primary-target-molecule probes includes an antibody or antibody -bindingfragment thereof.

[0012] The present disclosure further provides for compositions that include: (1) a first set of probes configured to bind directly to one or more target molecules; (2) an optional second set of probes conjugated to, or configured to bind to the first set of probes; and (3) agents configured for sample preparation such as fixation, permeabilization, hybridization, hydration, blocking, washing, buffering, and / or mounting.BRIEF DESCRIPTION OF THE DRAWINGS

[0013] FIG. 1 is a conceptual diagram of an example biological sample that may be spatially profiled using the techniques of the present disclosure.

[0014] FIG. 2 is a simplified conceptual diagram of an example biological-sample- analysis system that may be configured for multiplex analyte screening in in situ spatial biology assays, in accordance with the techniques of the present disclosure.

[0015] FIG. 3 is a simplified conceptual diagram illustrating some example optical configurations for the imager and detector of FIG. 2 that may be implemented while acquiring analyte measurements from a biological sample.

[0016] FIG. 4A is a simplified conceptual diagram illustrating an example of hyperspectral imaging of a biological sample in relation to corresponding regions of an IR- absorption spectrum.

[0017] FIG. 4B is a closeup view of the cell “silent region” portion of the IR-absorption spectral plot of FIG. 4A, illustrating an example set of distinct vibrational resonances that enable superior probe multiplexing.

[0018] FIG. 5A is a simplified conceptual diagram of a molecular biomarker target within the biological sample of FIG. 1, illustrating some example bonding configurations between the target, primary, and secondary probes when labeling the target for imaging via the microscopy techniques described herein.

[0019] FIG. 5B is a conceptual diagram illustrating an example technique for multiplex imaging of a biological sample via primary and secondary labeling of target biomarkers within the sample.

[0020] FIG. 6A is a Jablonski energy diagram illustrating an example of linear fluorescence excitation and emission.

[0021] FIG. 6B is a Jablonski energy diagram illustrating an example of fluorescence- encoded infrared (FEIR) excitation and emission.

[0022] FIG. 7A is a graph illustrating an example absorption and emission spectrum of a fluorescence-encoded infrared (FEIR) probe with a small electronic transition energy gap.

[0023] FIG. 7B is a graph illustrating an example absorption and emission spectrum of a fluorescence-encoded infrared (FEIR) probe with a large electronic transition energy gap.

[0024] FIG. 7C is a phasor plot showing an example of fluorescence-lifetime data.

[0025] FIGS. 8A & 8B are conceptual diagrams illustrating how the super-resolution microscopy (SRM) techniques of this disclosure can be used to reduce optical crowding when imaging a biological sample.

[0026] FIG. 9 is a simplified block diagram of an example composition that may be assembled in order to facilitate implementation of the microscopy techniques of this disclosure.

[0027] FIG. 10 is a flowchart illustrating an example technique for applying biomarker labels for in situ profiling of a biological sample.

[0028] FIG. 11 is a conceptual block diagram illustrating an example computing device for the system of FIG. 2, which may be used to implement any or all of the processor-based techniques referenced herein.DETAILED DESCRIPTION

[0029] The present disclosure is directed to various example systems, devices, techniques, methods (including computer-implemented methods), compositions, and products of manufacture commonly providing for high-resolution, in situ, three-dimensional (3D) mapping (or “spatial profiling”) of certain types of analytes, and in particular, in biological samples. For purposes of illustration (and not limitation) FIG. 1 depicts a conceptual diagram of one such biological sample 100, which may be, or may include: one or more biological cells 102A, 102B (collectively, “cells 102”); cell cultures; biological tissues; or even biological organisms (including single-celled and / or multi-cellular organisms), depending on the particular implementation of the teachings provided herein.

[0030] More specifically, the techniques of this disclosure enable greater insights into the composition of the biological sample 100 to be ascertained through strategic, high- multiplexed imaging of certain molecular biomarker (or “biomolecule”) targets 104 within the biological sample 100, including at super-resolutions. Such biomarkers 104 may be, or may include, for instance: deoxyribonucleic acid (DNA); ribonucleic acid (RNA), including messenger RNA (mRNA); proteins, including antigens; and / or metabolites, including glycans, lipids, and other small molecules. As elaborated further below, through the systematic selection and integration of vibrational spectro-microscopy techniques, the present disclosure presents a number of unique opportunities for substantial technical advancements in research and investigations into the presence, distributions, and concentrations of molecular species in biological systems.

[0031] Modem biological, clinical, and biomanufacturing applications demand rapid, high-resolution identification and quantification of numerous molecular biomarker targets 104 associated with fields of study from across the “multi omics” continuum, including genomics, epigenomics, transcriptomics, proteomics, metabolomics, and ultimately,phenomics. In particular, analytical tools that can non-invasively identify, localize, and quantify chemical composition within cells 102 will rapidly accelerate research in many important fields such as microbial ecology, synthetic biology, biomanufacturing, biopharmaceuticals, immunology, oncology, cell and gene therapy, drug discovery, rare cell discovery, tissue engineering, patient stratification, and personalized medicine. Of critical importance is the in situ spatial analysis of molecular biomarker targets 104 such as DNA; RNA, including mRNA; proteins, including enzymes, biologies, antibodies, hormones, and (poly)peptides; primary and secondary metabolite elements, including fatty acids, amino acids, lipids, glycans, carbohydrates, and other small molecules, which may be found, for instance: within cells 102 (i.e., “intracellular”), on the surface of cells, or outside of cells (i.e., “extracellular”), in the microenvironment of cell cultures, tissues, organs, and throughout organisms. For instance, determining the locations, quantities, structural relationships, and interactions of many biomarker targets 104 through a high-multiplex assay provides for an information-rich dataset that may illuminate new insights when characterizing cell type, cell phenotype, and active metabolic pathways as influenced by gene-environment interactions.

[0032] Accordingly, the techniques of this disclosure are of unique applicability to a cellbased biological sample 100, which can be, or can include cell(s) 102, cell fragments (e.g., platelets), adherent cell cultures, suspension cell cultures, spheroids, organoids, three- dimensional (3D) cell cultures, cells in a (microfluidic) flow cell compartment, tissues, tissue sections, and organisms, any of which can originate or derive from any species of organism. As the microscopy imaging techniques described below are tailored for non-destructive, in situ analysis, the samples 100 may even include living subjects and / or specimens. In some examples, the samples 100 are fixed and preserved, such as with a formalin-fixed, paraffin- embedded (FFPE) tissue section. Additionally or alternatively, the sample 100 can be from a biopsy, such as a fresh frozen (FF) tissue sample or autologous cells. Additionally or alternatively, the sample 100 may contain cells 102 grown in a natural environment or cultivated in a controlled environment such as a fermenter or bioreactor. Additionally or alternatively, the sample 100 may contain multiple cell types (e.g., 102A & 102B), cell phenotypes, cell species such as co-cultures or xenografts, or any combination of the aforementioned sample types.

[0033] Cell(s) 102 of the biological sample 100 may be or may include, for example: a primary cell, a cloned cell, a cancer cell, a tumor a cell, a reproductive cell, an immune cell (for example lymphocyte, granulocyte, macrophage, monocyte, neutrophil, neuroglia, or dendritic cell), a neural cell, an engineered cell using gene editing, a cell with recombinantDNA, a therapeutic cell, a stem cell, an (induced) pluripotent stem cell, a progenitor cell, an adult cell (e.g., fibroblast), a eukaryotic cell, a prokaryotic cell, an autologous cell, an allogenic cell, a mammalian cell, an animal cell, a Chinese hamster ovary (CHO) cell, a plant cell, a bacterial cell, a fungal cell (e.g., yeast), an archaeal cell, a virus, a cell treated with a viral vector, a pharmaceutical, a nanoparticle, a micelle, a liposome, an antibody-drug conjugate, an antibody-oligonucleotide conjugate, or any other drug delivery system, and / or a mixture of any two or more of the aforementioned cell types.

[0034] The techniques disclosed herein can further be used to investigate, evaluate, and / or “assay” biological molecules 104, as well as their associated functions, events, dynamics (through longitudinal imaging studies), and / or processes including, for example: cell metabolism, cell state / status (e.g., division, proliferation, differentiation, apoptosis, phenotype, etc.), cell migration, molecular interactions (e.g., protein-protein interactions, protein-glycan interactions, protein-nucleic acid interactions, receptor-ligand binding), transcription, translation, modification, biomolecule uptake, activity, conformation, orientation, and mobility, nuclear structures and organizations, chromosome dynamics, gene expression and activation in a temporal and spatial fashion, transcript or protein abundance, post-translational modifications, molecular folding such as in proteins, endocytosis, exocytosis, and / or any other biological function.

[0035] Previously established in situ spatial-biology technologies used to quantitatively detect and visualize molecular biomarker targets 104 include, for example, immunohistochemistry (IHC) and immunofluorescence (IF) for identifying protein (e.g., antigen) targets, and fluorescence in situ hybridization (FISH) for identifying DNA and / or RNA nucleotide sequences. In general, fluorescence-based measurements are largely constrained by the relatively small number of targets 104 that can be labeled and uniquely identified at any one time (e.g., typically around 4 to 6 targets), due to the broad, overlapping spectral-emission profiles of conventional fluorescence probes used in widefield and confocal microscopes. This makes it generally challenging to characterize the presence of many different biomarker targets 104 in a high-multiplex assay with high specificity, and timeconsuming to do so through multiple rounds (or “cycles”) of labeling, washing, imaging, and removing / stripping away fluorescent probes from the sample 100. Simultaneous monitoring of a greater number of biomarker targets 104 than what is currently available with fluorescence will provide a more holistic view of the cell 102, enabling superior differentiation when evaluating attributes of cell types 102A / 102B and cell phenotypes such as morphology and chemical composition. Furthermore, standard fluorescence microscopyoften suffers from autofluorescence backgrounds, photobleaching, phototoxicity, challenges with labeling small molecules, and is limited in its ability to resolve spatial features beyond the diffraction limit, typically around 250 nanometers (nm) for high-numerical-aperture (NA) objective lenses, without using sophisticated super-resolution techniques.

[0036] The advent of new vibrational spectro-microscopy techniques with increased sensitivity and signal-to-noise ratio (SNR) has enabled real-time structural and functional measurements of biological samples 100, including cells 102 and tissues. These innovations provide molecular contrast by probing spectroscopic signatures, which serve as a “fingerprint” to differentiate cell types 102A / 102B and phenotype by probing their molecular composition. Vibrational imaging can be used to investigate endogenous molecular species native to the sample 100 as well as molecular probes used to label specific molecular biomarker targets 104, similar to fluorescent probes.

[0037] Two over-arching families of techniques for exciting fundamental vibrational resonance modes are Raman scattering and infrared (IR) absorption. Of these two types of light-matter interactions, the IR-induced transition using mid-infrared (MIR) radiation2.5 pm to 25 pm) is by far the strongest, characterized by IR-absorption cross-sections (e.g., ~10'22cm2sr_1) that are on the order of 100-million times larger than the corresponding spontaneous Raman scattering cross-sections. IR-absorption occurs when an IR-active molecule, defined by its permanent dipole moment, absorbs a photon that matches the energy of the vibrational resonance through a direct dipole-allowed transition to an excited vibrational state.

[0038] Infrared-based spectroscopy, either alone or in combination with additional microscopy techniques (as detailed further below), may be implemented via a suitable imaging system like the one shown in the simplified conceptual diagram of FIG. 2. In general, imaging system 206 is used to generate spatial imagery 208 and other data from a biological sample 100, and includes at least: a first electromagnetic (EM) radiation source 210; a second EM radiation source 212; an optical imager 214 (or “microscope 214”) that includes a stage 216 configured to retain the sample 100, and internal optical component s) 218 (e.g., lenses, mirrors, etc.); a photodetector 220, and a computing device 222 configured to process, generate, analyze, and or display the corresponding imagery 208 and other data.

[0039] IR spectroscopy techniques are useful for analyzing and identifying samples 100, and when implemented in a microscope 214, can be used to produce imagery 208 with spectroscopic contrast. The most common type of IR microscope 214 is the conventionalFourier-transform infrared (FTIR) microscope, which covers molecular absorption features over a broad spectral range in the mid-infrared portion of the EM spectrum. Despite the advantages of FTIR imaging, the technique suffers from practical hurdles, including (but not limited to): (1) significantly lower image resolutions compared to visible-light microscopy due to the long wavelengths of IR radiation 224 used to resonantly excite nuclear vibrational motion within the sample 100; (2) the inability to perform three-dimensional (3D) optical sectioning in thicker samples 100; and (3) distorted imagery 208 due to thermal noise and / or relatively low pixel densities than other detectors, e.g., those that operate in the visible and near-IR spectral regimes.

[0040] Fortunately, recent developments in photothermal-infrared (PI) microscopy have broken the existing IR-diffraction limit to provide sub-micrometer (pm) resolutions over the entire MIR range by encoding the long-wavelength MIR photons 224 (also referred to herein as MIR radiation or MIR illumination field) onto shorter-wavelength photons 226 in the ultraviolet (UV), visible, and near-IR spectral regimes, referred to herein as “probe” photons or probe radiation / probe illumination field 226, providing a compelling technology for sub- cellular, high-throughput chemical imaging for spatial biology.

[0041] In general, the relevant Pl-microscopy techniques include: (1) capturing a first image of the sample 100 (a “hot frame”) by using a standard or custom-built optical microscope 214 while the sample 100 is simultaneously illuminated by a pulsed, MIR- spectrum IR beam 224 from the first EM source 210 and a pulsed, visible-spectrum probe beam 226 from the second EM source 212; (2) capturing a second image of the sample 100 (a “cold frame”) in which the MIR source 210 (but not the probe source 212) is blocked or remains inactivated; and (3) resolving the difference between the hot frame and the cold frame to generate the desired imagery 208. This difference is a result of thermal lensing, i.e., a change in local refractive index and thermal expansion during the hot frame which lasts only a few microseconds ( is) - this is the photothermal signal. The absorbed energy from the MIR photon 224 is converted to kinetic energy through nuclear motion within a molecule of the sample 100, thereby populating a higher energy vibrational state.

[0042] This energy is then dispersed to the surrounding environment through phonons during the relaxation process back to ground state, causing a localized temperature increase and corresponding thermal-lensing effect. Despite using long-wavelength radiation to resonantly excite vibrational motion in IR-absorption spectroscopy (and thereby, in PI microscopy), after the IR photon 224 is absorbed and the phonon relaxation occurs, the induced temperature gradient remains within the local environment and can be probed atmuch higher resolutions, limited only by the diffraction limit of the imaging system 206 using a visible-spectrum -to-near-IR spectrum (or even UV-spectrum) probe source 212. Additional intricacies, techniques, and applications of the photothermal effect can be found in commonly assigned U.S. Patent Application No. 18 / 407,576, filed January 09, 2024, and titled “PIXELATED PHASE MASK FOR CHEMICAL IMAGING, ” the entire contents of which are hereby incorporated by reference.

[0043] The photothermal effect can be induced by a suitable IR source 210, such as a compact and cost-effective pulsed Quantum Cascade Laser (QCL), although newer high- power sources including tunable, MIR light sources 210 based on single-step parametric down conversion in an optical parametric oscillator (OPO) may also be used. These IR sources 210 are narrowband (i.e., spectral linewidth < 10 cm'1, and preferably < 5 cm'1) in order to discretely target individual spectral features that often have full-width-half-max (FWHM) linewidths around 10-15 cm'1, although some vibrational modes, such as the hydroxyl stretching mode, can be many times wider.

[0044] Pulsed MIR light sources 210 typically operate at a repetition rate between about 0.1 kilohertz (kHz) and about 10 megahertz (MHz), and have pulse durations ranging from about 0.1 nanosecond (ns) to about 1,000 ns. Such IR-illumination sources 210 are ideally tunable across at least a portion of the MIR spectral range and are available by commercial suppliers including: Thorlabs, Inc., of Newton, New Jersey; Block Engineering, of Southborough, Massachusetts; and DRS Daylight Solutions, of San Diego, California.

[0045] The photothermal effect typically warms the local milieu within the sample 100 by a few degrees Kelvin for a few microseconds before thermalizing and returning to the ambient temperature. The corresponding change to the refractive index for typical biological specimens 100 is approximately 10'4per degree. As an illustrative example, assuming the sample 100 has a refractive index of 1.38 (comparable to that of water, at about 1.33), then such a minor change in refractive index typically requires numerous frame averages to reach an appreciable signal-to-noise ratio. One example method to optimize the optical path for improved contrast and SNR includes deploying pupil engineering for dark-field imaging. As described by Cheng et al. in International Patent Application Publication No. WO2022 / 204525A1, the entire contents of which are hereby incorporated by reference, this approach can boost the SNR six -fold.

[0046] Another method is to change the optical setup to be sensitive to optical phase (“phase contrast”) rather than intensity, e.g., as described in “Molecular contrast on phasecontrast microscope,” by K. Toda, M. Tamamitsu, Y. Nagashima, etal., the contents ofwhich are hereby incorporated by reference. Phase-based methods are sensitive to optical path length (OPL), which is the product of the distance the light travels and the refractive index, and provide contrast to otherwise largely transparent samples 100 through wave interference, i.e., without requiring the use of stains or dyes. Quantitative phase imaging (QPI) measurements can be conducted with exceptional sensitivity, e.g., routinely achieving nanometer (nm) scale sensitivity to the OPL while maintaining diffraction-limited resolution. Combining the chemical specificity of PI microscopy with the sensitivity of QPI has potential to enable routine access to lower chemical concentration detection limits in the submicromolar regime for many different types of biomarker targets 104 and samples 100.Although there are several variations of QPI setups, the most fruitful approaches benefit from a common-path geometry for mechanical stability, white-light illumination 226 from, for example, lamps or light emitting diodes (LEDs) to eliminate speckle noise and improve resolution, and single-frame image acquisition for high-throughput imaging.

[0047] An example of a suitable light source 212 to probe the photothermal effect is a spectrally broad (i.e., “white”) microscopy-grade LED with short pulse durations. Given that the transient photothermal response lasts only a few microseconds (ps), preferred probe sources have a “rise” time up to their peak intensity of less than about 5 ps, and a similarly short “fall” time back to zero intensity. Advantageously, such sources 212 require only modest optical powers (e.g., less than about 5 milliwatts (mW) for typical “thin” samples) given the near-unity quantum efficiencies of modem back-illuminated imaging-sensor arrays. Such probe sources 212 are commercially available from optical suppliers such as Thorlabs, Inc., of Newton, New Jersey; and Edmund Optics, of Barrington, New Jersey.

[0048] Detector 220 as shown includes an array of sensors, although point-like detectors may be used for beam-scanning implementations, and is configured to generate and output a corresponding “raw” electrical signal 230, having a charge or voltage amplitude that is directly proportional to the optical intensity on the respective sensor element. Detector 220 outputs raw electrical signal(s) 230 to a digitizer module of a computing device 222. The digitizer module is configured to convert raw electrical signal(s) 230 into corresponding bits of raw data indicative of the amplitude of raw electrical signal 230. The digitizer module is further configured to encode the data into a computer-readable medium (or, “memory”) of computing device 222, such as Random-Access Memory (RAM) and / or Read-Only Memory (ROM). Processing circuitry (or “processor(s)”) of computing device 222 is / are configured to retrieve data from memory, and process the data in order to generate imagery 208, e.g., in the form of chemical imagery and / or IR-absorption-spectra imagery. The processor transmits theimagery 208 to a visual display (e.g., a display screen of computer 222) or other suitable output device for viewing by a user. In some examples (but not all examples), the digitizer, memory, processor(s), and / or display are integral components of the same computing device 222, such as a common personal computer (PC), laptop, smartphone, server, or the like, as detailed further below with respect to FIG. 11.

[0049] In accordance with the techniques of this disclosure, aspects of IR-absorption- based microscopy techniques (particularly PI microscopy) can be combined with various types of optical microscopy techniques in order to capture images of sample 100 with complementary contrast mechanisms (commonly referred to as “multimodal imaging”). The resulting images can then be combined or otherwise analyzed together. In particularly notable examples (but not all examples) of this concept detailed further below, biomarker targets 104 within sample 100 can be labeled with (i) non-fluorescent molecular probes and (ii) fluorescent FEIR probes, simultaneously. In such cases, the same IR radiation beam 224, in unison with probe radiation 226, elicits two different types of measurable responses: (1) photothermal excitation, which is “imprinted” onto the probe radiation field 226 and captured by detector 220; and (2) fluorescence signals 232 of which are spectrally separated e.g., by a dichroic mirror and captured by a detector 220. Regions within the sample that elicit a photothermal signal but not a fluorescence signal, or vice versa, can be differentiated from one another, supporting enhanced probe multiplexing.

[0050] The components of imaging system 206 can be arranged into any suitable illumination-and-collection geometry (or “configuration”), two of which (epi -detection and transmission detection) are illustrated in the simplified conceptual diagram of FIG. 3. For instance, imaging system 206 can be configured for beam-scanning and / or widefield imaging, in which IR radiation 224 and probe radiation 226 illuminate sample 100 in oblique, counterpropagating, or collinear geometries. In some embodiments, PI microscopy can be deployed in confocal beam-scanning or spinning-disk geometries that provide an improved pointspread function (PSF) for optical sectioning during 3D imaging. In other embodiments, widefield-imaging geometries are used, which can offer video-rate imaging and up to 100* higher throughput relative to beam-scanning geometries without compromise to SNR. Such experimental setups may deploy “oblique” illumination, in which the MIR radiation 224 is loosely focused onto the sample 100 from one lateral side. Specifically, the MIR radiation 224 is focused onto an area of the sample 100 that is roughly 1.5 x the field-of-view (FOV) to provide uniform PI contrast and to reduce the “vignetting” effect, i.e., a measurably lower contrast around the outer periphery of the FOV compared to the center.

[0051] In some beam-scanning or widefield imaging setups, the IR light source 210 illuminates the sample 100 in a counterpropagating illumination geometry relative to the probe beam 226. In still other setups, the IR and probe sources 210 / 212 illuminate the sample 100 in a collinear illumination geometry. This requires use of a microscope objective lens 218 that can accommodate radiation over a very broad spectral range (e.g., from UV to MIR), such as reflective objective lenses including the Schwarzschild-Cassegrain or the non- concentric Hanninen-type objective lens. As shown in FIG. 3, in some experimental setups, the images are captured using a microscope 214 that is upright, or inverted, via epi-detection (320A / 328A / 332A). In others, the images are captured in transmission mode (320B / 328B / 332B).

[0052] FIG. 4A is a conceptual diagram illustrating an example technique for generating multiplex, imagery 208 of a biological sample 100. At a high level, generating imagery 208 involves capturing and then splicing-together or combining (i.e., multiplexing) multiple discrete sets of image data 408A-408C corresponding to different spectral regions of an IR- absorption spectrum 434A, More specifically, FIG. 4A illustrates an example set of representative absorption features 436A-436C from across the MIR spectral range, known as multispectral imagery, and how they correspond to spatially resolved imagery 208 with chemical specificity.

[0053] In particular, FIG. 4A depicts an IR-absorption spectrum 434A that distinguishes between three regions: (1) the “high-frequency fingerprint” region 438A, (2) the “cell-silent” region 438B, and (3) the “fingerprint” region 438C. For typical unlabeled biological specimens 100, regions 438A and 438C are likely to provide the most-useful IR-absorption features, whereas molecular and FEIR probes (shown and described with respect to FIGS. 5 A & 5B) are likely (but not limited to) to include distinct absorption features within the cell- silent region 438B (see FIG. 4B).

[0054] Tuning the wavelength of IR source 210 (FIG. 2) to discrete wavelengths across the IR-absorption spectrum 434A provides for the generation of discrete sets of image data 408A-408C where signal is generated through the absorption of IR photons by molecules in the sample. As shown by the individual images 408A-408C, regions within the field-of-view (FOV) that do not contain molecules that absorb at a particular IR wavelength will generate no signal and remain dark, while regions with lots of absorbing molecules or molecules with strong absorption features will have a large signal, corresponding to high brightness within image data 408. While targeting a discrete set of absorption features 436 allows for rapid identification, localization, and quantification of specific biomarker targets 104, a continuoussweep across the entire tunable range, known as hyperspectral, allows for the complete IR- absorption spectrum 434A to be generated for each pixel within the FOV. Combining the discrete sets of image data 408A-408C produces the final imagery 208.

[0055] Despite extensive efforts, it is generally difficult to decipher a large number of unlabeled (endogenous), chemically similar biomarker targets 104 composed of the same molecular constituents through vibrational imaging due to overlapping vibrational resonances. However, thanks to the narrowband nature of molecular vibrational modes, palettes of molecular probes can be used to selectively bind to biomarker targets 104 and provide additional contrast with greatly improved multiplexing relative to fluorescent probes. These molecular probes benefit from sharp, mutually resolvable absorption peaks 436 used to simultaneously quantify numerous molecular biomarker targets 104, and are not susceptible to photobleaching. Another advantage of molecular probes is that they can have relatively low molecular weights (i.e., <200 Da), allowing improved permeability through cell and organelle membranes within the sample 100. Even further, these molecular probes are compatible with labeling relatively small biomolecules 104 with negligible perturbation to their native biological function, which is difficult when using bulkier fluorophores.

[0056] Ideal molecular probes have functional groups with a large IR-absorption crosssection, narrow linewidth (10-15 cm'1), and vibrational resonances in the cell-silent region 438B (1900-2600 cm'1), a region that is usually spectrally silent in biological systems, thus allowing for the detection of such vibrational probes with a high SNR and specificity.

[0057] For instance, FIG. 4B depicts an example IR-absorption spectrum 434B focused on the cell-silent region 438B of FIG. 4A. Cell-silent spectrum 434B contains a representation of four distinct resonant vibrational modes 440A-440D suitable for multiplexing using molecular and / or FEIR probes. The narrow spectral linewidth of molecular vibrational modes and the ability to shift their resonant frequency by various means including, but not limited to, varying stable isotopes, perturbating the local environment by changing the surrounding chemical groups, or increasing the conjugation length or number of functional groups in the molecular probe, allow for an expanded set of biomarker targets 104 that can be uniquely labeled simultaneously. The wavenumber corresponding to the resonance energy for the representative vibrational modes 440 is provided below their respective peaks.

[0058] Molecular probes that contain functional groups with resonant vibrational modes 440 present in the cell-silent region 438B include, but are not limited to: deuterated hydroxyl, methyl, and methylene modes, as well as thiols, azides, alkynes, nitriles, isonitriles, cyanates,isocyanates, thiocyanates, and isothiocyanates. These molecular probes may further leverage stable isotopes within the functional group to increase probe multiplexing. Heavier isotopologues feature “red-shifted” or lower-energy vibrational resonances, which are used to separate otherwise-overlapping spectral features, allowing further differentiation between functional groups. Examples include, but are not limited to:13C,15N,18O, and34S.

[0059] FIG. 5A is a simplified conceptual diagram of an example molecular biomarker target 104, e.g., within the biological sample 100 of FIG. 1. In particular, FIG. 5A illustrates various example bonding configurations between the biomarker target 104 and one or more probes used for labeling the target 104 for imaging, i.e., when implementing the microscopy techniques described herein. As referenced above, molecular probes 542 have functional groups with distinct vibrational resonances that can be used to generate spectrally differentiated signals 440 (FIG. 4B) through the Pl-microscopy contrast mechanism. In order to use them to identify specific biomarker targets 104, the probes 542 require a target-specific binding region 544 that can conjugate or bind to complementary regions 554 of the particular biomarker target 104. The molecular probes 542 that bind directly to the biomarker target 104 are referred to as “primary” probes 542A and are suitable for direct labeling; molecular probes 542 that instead have a binding region designed to bind to the primary probe 542A are referred to as “secondary” probes 542B, and are suitable for indirect labeling. Primary probe 542A may or may not have a distinct functional group 546 when indirect labeling is used as the functional group 546 will be part of the secondary probe 542B (see bottom of FIG 5A).

[0060] Upon labeling, each given biomarker target 104 can bind to one or more primary probes 542A, each primary probe being equipped with the same or with different functional groups 546. The approach used to bind to the biomarker target 104 will depend on the nature of the target. In some cases, the primary probe 542A will contain oligonucleotides with sequences that specifically hybridize to nucleic acid targets 104. In other cases, the primary probes 542A may contain regions 544 (or “motifs”) configured for biorecognition of a specific target 104, including, but not limited to: antibodies, antigen-binding fragments, and / or their derivatives; nucleic acid or peptide aptamers; proteins; conjugate ligands; carbohydrates; glycans; lipids; synthetic binders; or any combination thereof.

[0061] In still other cases, the primary probe 542A includes antibody-oligonucleotide conjugates (AOCs), domains that allow binding of secondary probes 542B, or cause an enzymatic or a branched event 558 for amplification purposes (i.e., to add more secondary probes 542B and increase the detectable signal). In some instances, the primary probe(s) 542A are covalently bound to the target biomarker 104. In other examples, the primaryprobe(s) 542A are noncovalently complexed with the target biomarker 104. Those skilled in the art can also modify the biomarker target 104 and / or primary probe 542A with a chemical moiety with which a second chemistry (e.g., “click” chemistry) can be conducted to tether various primary probes to target 104 or other probes.

[0062] In addition to having a functional group having a distinct IR-absorption signature 546 and a region for binding to the target 104 (primary probe) or to a primary probe (secondary probe), a molecular probe 542 (whether primary or secondary) may contain a luminescent, light-emitting moiety 550 in communication with the IR-active functional group 546 through vibronic coupling 560, e.g., via Franck-Condon activity. Through an excitation process referred to herein as “fluorescence-encoded infrared” (FEIR) (depicted below in FIG. 6B), these probes, referred to herein as “FEIR” probes 552, absorb both a resonant IR photon 224 and a short- wavelength probe photon 226 so the FEIR probe 552 enters an excited electronic state. After a characteristic lifetime in the excited state, the FEIR probe 552 returns to the ground state by emitting a fluorescence photon 232 (FIG. 2). FEIR probes 552 are distinct from molecular probes 542 in that they are observed by detecting the emitted fluorescence photons 232 rather than through the PI microscopy contrast mechanism, although both leverage the distinct, narrowband vibrational resonances of the IR-active functional groups 546. FEIR probes 552, therefore, advantageously combine the high- multiplex attributes of molecular probes 542 with the near-single-molecule sensitivity of standard fluorescent probes (e.g., as shown in FIG. 6A). Minimum requirements for a “good” FEIR probe 552 include high fluorescence brightness, large IR-absorption cross-section of the functional group 546, and strong vibronic coupling 560 between the functional group 546 and the electronic transition.

[0063] In addition to having a distinct IR-active functional group 546, FEIR probes 552 can benefit from having distinct absorption and / or fluorescence-emission spectral profiles for even higher assay multiplexing. Fluorescent probes optimized for distinct spectral characteristics ranging from the UV through the near-IR are currently used in multiplexed assays, however, they often do not include a distinct, vibronically coupled functional group 546 as in FEIR probe 552. Commercially available fluorescent probes with a light emitting moiety 550 that are suitable for conversion to a FEIR probe by including a distinct IR-active functional group 546 through a vibrionic coupling 560 include, but are not limited to, the BODIPY series, the Alexa series, the ATTO series, the cyanine series, DAPI, Coumarin, Fluorescein, Bora-Fluorescein, Rhodamine, Texas Red, Luciferin, organic dyes, biological fluorophores, quantum dots, nanocrystals, and / or fluorescent proteins such as greenfluorescent protein (GFP). In some examples, fluorescence donor — acceptor pairs are used where the donor is a FEIR probe, as in Forster resonance energy transfer (FRET). Those skilled in the art can also employ (photo)activatable, (photo)switchable, (photo)cleavable probes, or photobleaching to modulate or alter properties of the FEIR probe 552 using means of external stimulus (for example heat, UV light) or changing local environment (for example pH, temperature) to infer higher multiplexing capabilities despite overlapping spectral emission profiles.

[0064] In some examples, biomarker targets 104 are labeled by primary probes 542A so as to create a branch-like structure 558 that clusters out from each target 104. Each branch cluster 558 may comprise a probe 542 / 552 that can elicit a signal (i.e., through PI microscopy or FEIR fluorescence) or an “extension” component 548 that can be used to further extend out the branch 558 for subsequent attachment to additional probes 542 / 552. Each branch cluster 558 may be labeled with many of the same type of probe to enhance the SNR, or a unique combination of probes 542 / 552 that generates a more-diverse probe signature, which enhances multiplexing capabilities for improved detection and mapping of many biomarker targets 104.

[0065] FIG. 5B is a conceptual diagram illustrating an example technique for multiplex imaging of a biological sample 100, via primary and secondary probe labeling of target biomarkers 104 within the sample 100. First, the technique includes collecting a sample 100, e.g., from a biological specimen or subject. In FIG. 5B, sample 100 is depicted as a cluster of cells 102, which may be either alive or fixed. In implementations in which probes 542 / 552 are not used (i.e., label-free, indicated by arrow 556A) sample 100 is positioned directly onto stage 216 of optical imager 214 to conduct label-free chemical imaging measurements. In implementations utilizing molecular probes 542 and / or FEIR probes 552 (indicated by arrow 556B), sample 100 is then processed or treated for the addition of a set of primary probe(s) 542A. Each primary probe 542A has a region 544 (FIG. 5A) that can selectively bind or adhere to a specific, complementary region 554 on a desired target 104 within sample 100.

[0066] The technique of FIG. 5B further depicts the application of secondary probe(s) 542B to sample 100 (indicated by arrow 556C) that contain molecular probes 542 and / or FEIR probes 552 and are used for indirect labeling. In some examples, primary probe 542A may be directly labeled by a secondary mol ecular / FEIR probe 542B (see bottom -right of FIG. 5 A). Additionally or alternatively, primary probe 542A may include an extension element 548 that can couple to additional probes during subsequent labeling steps, e.g., to facilitate branching 558 and signal-amplification. In the example of FIG. 5B, only two labeling stepsare depicted; however those of skill in the art will recognize that additional (e.g., “tertiary”) steps and beyond may be included when performing this technique.

[0067] After secondary probes 542B are bound to their target and indirect labeling is complete, the technique of FIG. 5B proceeds (arrow 556D) to using optical imager 214 to perform imaging measurements on sample 100. Label-free imaging 556A may be performed to visualize unlabeled biomarker targets as well. In some instances, the biomarker target 104 may be labeled with multiple primary (and optionally secondary) probes with the same molecular / FEIR probe 542 / 552 to boost signal, e.g., as indicated for target 104A. Alternatively or additionally, a biomarker target 104B may be labeled with multiple primary (and optionally secondary) probes with different molecular / FEIR probes for combinatorial barcoding e.g., as indicated for target 104B. In additional embodiments not shown in FIG. 5B, branching elements 558 (FIG. 5A) from a single primary probe 542A or secondary probe 542B may facilitate binding of several secondary or tertiary probes, respectively, to boost signal or for combinatorial barcoding, which can be used to differentiate multiple different biomarker targets 104 in a high-degree multiplex assay.

[0068] Additional relevant techniques include adding, mixing, and / or incubating primary probes 542A, secondary probes 542B, and / or tertiary probes to the samples 100 for staining, labeling, binding, or hybridization and their related steps or processes. For instance, in some examples, the sample 100 to be analyzed is placed on a simple microscope slide, or in a dish, multi -well plate, or microtiter plate positioned on top of a microscope sample stage 216, where staining occurs prior to imaging. Additionally or alternatively, the sample 100 may be placed in a flow-cell compartment on top the sample stage 216, whereupon a network of fluidic and / or microfluidic channels are used for sample-processing steps (for example: capture, conjugation, digestion, washing) in addition to sample-labeling. In such cases, the flow compartment comprises a plurality of automated fluidic and / or microfluidic networks generally known as a microfluidic system or chip-cytometry. In some exemplary embodiments, the flow-cell compartment is used to house sample 100 and facilitate automated fluid exchange while applying primary probes 542A and optionally secondary probes 542B to the sample 100. The flow compartment may further be used to wash, strip, or remove probes 542A / 542B from sample 100 before applying another cycle of probes.

[0069] Additionally or alternatively, the compartment or chamber on the microscope sample stage 216 may be environmentally controlled, e.g., to maintain a consistent ambient temperature during imaging or incubate a sample 100 that includes live or living cells, tissues, or organisms. Regulated environmental parameters include but are not limited totemperature (heating or cooling), pH, glucose, salinity, and atmospheric conditions including oxygen and carbon-dioxide concentrations.

[0070] Additionally or alternatively, biomarker targets 104 may be labeled or barcoded combinatorically. As a generalized example, various biomarker targets 104 may be labeled with similar probes 542 / 552, but as long as they differ in at least one component, a different signature can be detected and used to differentiate biomarker species from each other. Since labeling targets 104 in this way scales up multiplexed detection of targets 104 combinatorically, a small increase in the number of possible probes 542 / 552 to use (or not use) for labeling will significantly increase multiplexing capabilities. For instance, for the given combinatorial equation, C(n, r) = n! / [(n-r)! r!], where C is the total number of combinations possible, n is the number of probes 542 / 552 available for use, r is the number of probes 542 / 552 actually used to label each target 104, and ! is the factorial function, if 16 unique probes 542 / 552 are available for use and only 3 probes 542 / 552 are used in any combination to label each target 104, there are 560 different possible combinations of probe signatures for which to label a biomarker target 104. This is a drastic increase compared to labeling each biomarker target 104 with a single probe 542 / 552, which only allows detection up to 16 different biomarker targets 104. To decode and identify the target 104 of interest, a codebook or index library can be utilized.

[0071] The techniques of this disclosure further include an important class of labeling methods, in which samples 104 are labeled sequentially or serially to enable higher-level multiplexing. Here, a set or panel of probes (whether molecular 542, FEIR 552, or both) are added to the sample 100 and allowed to label their biomarker target 104. The sample 100 is imaged, and the probes 542 / 552 are then washed, deactivated, or removed from the target 104 via any suitable chemical reagent or physical process, such as UV or heat exposure. This permits a subsequent round, or cycle, of sample-labeling to be completed. In the subsequent round, the same biomarker targets 104 may be labeled again, either with a probe 542 / 552 having the same molecular / FEIR signal as the first round, or a different signal. The sample 100 is again imaged and the probes are stripped, removed from the sample 100, allowing for the next cycle of labeling to begin. Through this approach, a unique “barcode” is generated for the sample 100, and an unlimited number of biomarker targets 104 can be positively identified, provided that sufficient rounds of labeling, imaging, and stripping can be performed without damaging the sample 100. More generally, any type of probe interactions that create a distinct, detectable signature may be implemented to represent a unique label. For instance, such signatures may be based on intensity, phase, absorption / emission spectra,emission lifetime, or any biophotonics-based property (e.g., blinking). Notably, while each individual labeling technique may appear sufficiently unique, multiple different techniques can also be implemented in combination, enabling even-higher multiplexing capabilities.

[0072] To accommodate such a diverse set of labeling possibilities for each target 104, a codebook may be used as a legend or an index to track a significant number of biomarker targets 104 for post-analytical identification, quantification, and spatial validation. Each measured and distinct probe signature can thus be decoded for identification. In some embodiments, this codebook may be a simple library that matches each specific probe 542 / 552 or set or combination of probes 542 / 552 to a specific biomarker target 104 of interest on a one-to-one basis. In other embodiments, this codebook may be a library that encodes each biomarker target 104 of interest using multiple probes 542 / 552 with redundancy or degeneracy. The codebook may also employ an error-correction statistical mechanism where, for example, even if a target 104 should be labeled with five labels 542 / 552, simultaneously or sequentially, and two of those labels 542 / 552 fail to appear, it can still assign a “probability factor” to that target 104 as being correctly identified as the particular target of interest. Another error-correction mechanism that the codebook can apply can be sequential- labeling error correction where, for example, if certain probes 542 / 552 show up at certain rounds but not others, there is a likelihood it is still that particular biomarker target 104 of interest with a probability factor assigned to it.

[0073] FIG. 6A is a Jablonski energy diagram illustrating a standard, linear fluorescence excitation process 662A. In linear fluorescence, a single absorbed photon 664 is absorbed by a fluorescent molecule or probe. If the photon 664 has sufficient energy, it will drive the molecule / probe into an excited state, often from the electronic ground state (denoted So) to the first electronic excited state (denoted Si). The molecule / probe can dissipate energy through non-radiative processes that occur on the picosecond (ps) timescale, such as vibrational deactivation 665, where the energy is transferred to other vibrational modes as kinetic energy, or internal conversion (not shown) as the molecule relaxes toward the lower energy levels within the excited state.

[0074] The fluorescence process then takes place by radiating emitted fluorescence photon 668, which occurs over a characteristic lifetime corresponding to the average time the molecule / probe stays in the excited state prior to radiative relaxation. This lifetime can be used to differentiate between different fluorescent molecules / probes, even if their spectral- emission profiles overlap, through fluorescence-lifetime imaging microscopy (FLIM). Emitted photon 668 generally has a longer wavelength and lower energy than absorbedphoton 664, known as a “Stokes shift,” and is due to the energy lost through vibrational deactivation 665 and other non-radiative relaxation processes.

[0075] For comparison, FIG. 6B is a Jablonski energy diagram for a fluorescence- encoded infrared (FEIR) excitation process 662B. FEIR excitation process 662B constitutes a doubly-resonant fluorescence excitation process, in which IR photon 670 (e.g., IR radiation 224 of FIG. 2) is tuned to match the resonance energy of a molecular vibrational mode in the FEIR probe 552. When IR photon 670 is absorbed, it drives the FEIR probe 552 into an intermediate vibrational state within the So ground state. This is distinct from “two-photon absorption” wherein the fluorescent probe does not populate a real intermediate energy state.

[0076] Next, a probe photon 672 (e.g., probe radiation 226 of FIG. 2), whose energy is sufficient to bridge the energy gap between the vibrational state and the Si excited state, is absorbed. The FEIR process 662B has a linear dependence on the intensities of the IR- illumination field 224 and probe-illumination field 226. The lifetime of the intermediate vibrational state typically lasts up to only a few picoseconds, so IR photon 670 and probe photon 672 effectively need to illuminate the sample 100 simultaneously for the FEIR excitation process 662B to occur. This is unlike molecular probes 542 used in PI microscopy, in which the thermal-lensing effect lasts a few microseconds, allowing the probe field 226 to continue to detect signals originating from the molecular probe 542 even after the pulsed IR field 224 is off. A second image may be captured with a time delay between IR-illumination field 224 and probe-illumination field 226 greater than about 20 ps to measure background signal (for example autofluorescence) not originating from the FEIR probes 552.

[0077] Once in the electronic excited state, the FEIR probe 552 dissipates energy through vibrational deactivation 674 while approaching the bottom of the electronic excited state, before radiating emitted FEIR photon 676 (e.g., fluorescent light 232 of FIG. 2). Of note, the energy of emitted FEIR photon 676 is not necessarily lower than optical frequency photon 672. If non-radiative energy losses 674 total less than the energy of IR photon 670, then the emitted FEIR photon 676 will have a greater photon energy than probe photon 672; this is more likely to happen when the energy gap between So and Si is relatively small, as in the case of near-IR fluorescent probes, where IR photon 670 provides an appreciable portion of the excitation energy.

[0078] Once in the electronic excited state, the FEIR probe 552 may undergo non- radiative relaxation processes 674 such as internal conversion (IC) or vibrational deactivation, where energy is transferred to vibrational modes in the probe 552 as kinetic energy. The FEIR probe 552 eventually returns to the ground state after a characteristiclifetime in the excited state through a radiative relaxation process and the emitted FEIR fluorescence 676 is collected by the imaging components 218 (FIG. 2) and focused onto photodetector 220. As with standard fluorescence microscopy, FEIR probes 552 can be differentiated by their fluorescence lifetime, and the achievable spatial resolution using FEIR probes 552 is limited by the wavelength of the emitted photon 676 / 232 when using a diffraction-limited imaging system 214.

[0079] FIG. 7A features a spectral plot 700A for a near-IR FEIR probe 552, i.e., having a relatively small electronic energy gap. Spectral plot 700A includes absorption spectrum 710 and emission spectrum 720. As shown in FIG. 7A, probe photon 730 (e.g., probe radiation 226 of FIG. 2) is far from having sufficient energy to be absorbed by the FEIR probe 552, but adding the energy from a resonant IR photon (e.g., IR radiation 224 of FIG. 2) through the doubly resonant FEIR excitation process 662B of FIG. 6B, provides excitation energy 740 which is within absorption spectrum 710. After a characteristic lifetime, the near-IR FEIR probe 552 will undergo radiative relaxation and emit a fluorescent photon (e.g., fluorescent light 232 of FIG. 2) within emission spectrum 720, These fluorescent photons 676 / 232 are selectively collected for detection by using bandpass filter 750, which is designed to maximize the coverage of emission spectrum 720 while rejecting probe photon 730 / 226.

[0080] FIG. 7B features a spectral plot 700B for a UV / visible-spectrum FEIR probe, i.e., having a relatively large electronic energy gap. Spectral plot 700B includes absorption spectrum 712 and emission spectrum 722. In this case, probe photon 732 (226) is much closer to absorption spectrum 712, but still far enough away to avoid single-photon absorption, as in linear fluorescence (e.g., FIG. 6A). Since the energy of the resonant IR photon 224 is much less than probe photon 732 / 226, in this case, the spectral shift between probe photon 732 / 226 and excitation energy 742 is much narrower. Nonetheless, excitation energy 742 is within absorption spectrum 712, while probe photon 732 / 226 is not. After a characteristic lifetime, the UV / visible FEIR probe will undergo radiative relaxation and emit a fluorescent photon 676 / 232 within emission spectrum 722. These photons are selectively collected for detection by using bandpass filter 752, which is designed to maximize the coverage of emission spectrum 722 while rejecting probe photon 732 / 226. In this case, however, bandpass filter 752 covers wavelengths that are longer than probe 732 / 226, not shorter.

[0081] As a representative demonstration, in FIG. 7C, phasor plot 700C provides an analysis of fluorescence lifetime and is used to discriminate FEIR probes 552 based on the time they remain in the excited state before radiating back to the ground state. In this analysis, probes 552 with a single decay constant, typically between 0.2 ns and 20 ns, arepositioned along the periphery of a semi-circle, with shorter lifetimes located farther from the origin. This is a convenient way to identify FEIR probes 552 with different emission lifetimes even if they have overlapping spectral emission profiles, which is beneficial for very high- multiplex assays, especially when combined with the FEIR excitation process 662B (FIG. 6B). FLIM data may be visualized using (spectral) phasor plots to unmix multiple lifetime and spectral components in the same diffraction-limited region of the sample to ensure detection fidelity for targets 104, and to “decode” a plurality of FEIR probes 552.

[0082] FIGS. 8 A & 8B are conceptual diagrams illustrating a super-resolution microscopy (SRM) technique, which can be used to reduce optical crowding when imaging a biological sample 100 (FIG. 1), which enables superior localization of biomarker targets 104. Specifically, FIG. 8A shows example imagery 800A (e.g., imagery 208 of FIG. 2) from a triplex-biomarker (104A-104C) assay captured using a standard, diffraction-limited imaging system. Using standard imaging systems, regions within the field-of-view may become saturated with biomarker targets 104 in a high-multiplex assay. For instance, the same diffraction-limited pixel 810A includes two different biomarker targets 104A, 104B. This effect, also known as “optical crowding,” diminishes the ability to effectively and quantitatively identify individual biomarker targets 104.

[0083] By contrast, FIG. 8B shows the resulting imagery 800B from the same triplexbiomarker assay when captured using a super-resolution microscopy (SRM) imaging system. In this case, optical crowding has been alleviated with improved resolving power through SRM techniques. For instance, super-resolution pixel 810B includes just one biomarker target 104A, and no more than one biomarker target 104 is present in any other resolvable region of imagery 800B.

[0084] SRM techniques enable access to high-performance quantitative biology by allowing accurate measurement of the abundances, distributions, and nanoscale movements of molecular biomarker targets 104. These biomarker targets 104 can include unlabeled, endogenous (i.e., label-free) chemical species, as well as molecules labeled with molecular probes 542 and / or FEIR probes 552. Some SRM methods can achieve 200, 100, 50, and even sub-20-nm resolutions. However, as a practical matter, SRM datasets are much larger, requiring improved processing power and often take much longer to acquire, which limits imaging throughput.

[0085] “Super-resolution” encompasses a family of microscopy techniques that all provide access to spatial information beyond the diffraction limit, typically around 250 nm for visible light and high-NA objectives. Many SRM techniques were pioneered for use withfluorescence probes, and as such, are capable of being adapted for use with FEIR probes 552. Some of these optical techniques include, but are not limited to: stimulated-emission depletion (STED) microscopy, reversible saturable optical fluorescence transitions (RESOLFT), saturated structured illumination microscopy (SSIM), stochastic optical reconstruction microscopy (STORM), and photo-activated localization microscopy (PALM).

[0086] The techniques of this class leverage different aspects of fluorescent probes to achieve super-resolution, however they all traditionally rely on a single-photon-absorption event 664 (FIG. 6A) to excite the fluorescent probe to the Si excited state before detecting the fluorescence signal 668. Using FEIR probes 552, and deploying the two-step FEIR excitation process 662B (FIG. 6B) using IR and probe illumination sources 210 / 212 (FIG. 2), these techniques can be used to visualize biomarker targets 104 at super-resolutions in high- multiplex assays.

[0087] SRM techniques are also amenable for use with non-light-emitting probes 542 and provide spatial localization beyond the diffraction limit for a broader class of molecules including endogenous (label-free) biomarker targets 104 and exogenous molecular probes 542. These SRM techniques, including structured illumination microscopy (SIM) and expansion microscopy (ExM), are fully compatible with PI microscopy, FEIR microscopy, and can be used simultaneously. The strategic deployment of SIM and / or ExM with the teachings described herein is worth far more than the sum of its parts, as these techniques enable nanoscopic, 3D chemical mapping of both labeled and unlabeled biomarker targets 104 alike. They can also be employed to improve measurement accuracy when quantifying the total presence of biomarker targets 104 by increasing the number of resolvable features within the field-of-view (FOV). In exemplary embodiments, molecular / FEIR probes 542 / 552 are deployed using SIM and / or ExM to achieve super-resolution localization of biomarker targets 104 across the multi omic cascade, including DNA, RNA, proteins, and metabolites in a high-multiplex assay. Additionally or alternatively, sample 100 can remain label -free (e.g., arrow 556A in FIG. 5B) and still be analyzed through super-resolution chemical-imaging measurements by combining these SRM techniques with PI microscopy.

[0088] SIM is a widefield optical technique that can achieve up to a 2x improvement in resolution over the diffraction-limit in each dimension, resulting in 3D voxels which are 8x smaller. It has been widely adopted in biological research as a practical balance between improving image resolution while retaining image throughput as only 9 raw frames are needed to reconstruct a 2D SIM image and 15 raw frames to reconstruct a 3D SIM image. SIM operates by capturing multiple images of a sample 100 using an illumination source witha spatially modulated intensity profile, typically generated using a diffraction grating or a spatial light modulator (SLM), with different illumination phases and orientations. This shifts high-frequency spatial components (in k-space) into the detectable bandpass frequencies of the microscope system 206 for detection (e g., by detector 220 in FIG. 2). Of note, only the visible-spectrum probe radiation 226 requires a spatially modulated intensity profile; the MIR source 210 can provide a uniform intensity across the FOV. When using SIM with FEIR probes 552, the combination of the two illumination sources drive the FEIR excitation process 662B resulting in FEIR fluorescence 676 / 232 that is detected by detector 220. For non-light emitting biomarker targets 104 or molecular probes 542, the spatially modulated probe radiation 226 is directly detected by detector 220. These raw frames are processed to reconstruct SIM images, and can be used to capture chemical information by reconstructing a “hot” SIM frame and a “cold” SIM frame, and calculating the difference as in PI microscopy. In some cases, the modulated probe radiation 226 may be configured to preserve phase sensitivity for quantitative phase imaging.

[0089] In another aspect, the sample 100 can be physically expanded and subjected to additional processing (e.g., capture, conjugation, digestion, washing, etc.) to further facilitate probe-binding and improved image quality, in accordance with techniques of expansion microscopy (ExM). ExM is a chemical approach designed to increase the resolving power of a microscope imager 214 by physically expanding the sample 100 in all three spatial dimensions. This may be achieved, for example, by using polyelectrolyte hydrogels that swell in response to changing pH, thereby expanding the sample by over 4x in each dimension. ExM protocols include pre-expansion and post-expansion labeling, and are carefully managed to ensure isotropic expansion of fixed samples. Since ExM is a chemical approach (as opposed to optical), it can be simultaneously deployed with complementary SRM techniques to reach spatial resolutions nearing those of electron microscopes. However, due to the expansion in the axial direction, capturing volumetric image sets may require use of an objective lens 218 with a longer working distance, which typically have a lower NA. Like SIM, ExM is fully compatible with samples 100 labeled with molecular probes 542, FEIR probes 552, and even label-free samples; however, as the sample 100 is fixed, probes 542 / 5522 are typically deployed.

[0090] In another respect, provided herein are certain compositions of matter and associated methodologies to facilitate labelling and imaging of molecular biomarker targets 104 within cells 102, cell cultures, tissues, organs, or organisms of a biological sample 100. These compositions use molecular probes 542 and / or FEIR probes 552, as described above,having distinct IR-absorption signatures from functional group 546, used for uniquely labeling one or more biomarker targets 104, including genomic, epigenomic, transcriptomic, proteomic, and / or metabolomic elements. For example, in nucleic acid detection using indirect labeling, one such composition includes: nucleic-acid target(s) 104; sets of primary probes 542A (e.g., oligonucleotides) with region 544 designed to bind or conjugate directly to the complementary ligand 554 of the target 104; optionally, a set of “readout” or “extension” domains 548; and, a set of secondary probes 542B with a distinct IR-absorption functional group 546 that conjugate either to the “readout” or “extension” domains 548 or directly to the primary probes 542A, and optionally a luminescent moiety 550 in communication with functional group 546 through vibronic coupling 560 to form an FEIR probe 552.

[0091] In another respect, provided herein are certain compositions of matter and associated methodologies to facilitate the detection of protein biomarker targets 104. Such compositions generally include: target protein(s) or epitopes 104; sets of primary probes 542A (e.g., antibodies); optionally, a simple amplification component; and optionally, a set of secondary probes 542B that bind to the primary probes 542A. The primary probes 542A feature a biorecognition motif including, but not limited to: nucleic acids, modified nucleic acids, peptides, proteins, enzymes, antibodies, antigen binding fragments, antibody - oligonucleotide conjugates (AOCs), carbohydrates, lipids, hapten, biotin, small molecules, or any combination thereof that can bind to a specific biomarker target 104. Generally, these primary probes 542A have a complementary region 544 (FIG. 5A) configured to selectively bind to a specific portion or region 554 of the biomarker target 104; and while they are often selected to bind to only one target 104, optionally, a primary probe 542A may have selective binding to multiple targets 104.

[0092] In some examples, a primary probe 542A is a molecular probe 542 and / or a FEIR probe 552 and used for direct labeling. In other cases, the primary probe 542A binds to secondary probe 542B or includes an extension element 548 that can bind to multiple secondary probes 542B for indirect labeling. Similarly, a set of tertiary probes may bind to already-bound secondary probes 542B, and so on. Each primary / secondary probe 542A / 542B may be labeled with a single functional group 546 and / or a luminescent moiety 550, or a set of same or different additional functional groups 546 and / or luminescent moieties 550 for combinatorial barcoding to detect multiple biomarker targets 104 in a high- order-multiplexed assay. After labeling, the decorated biomarker targets 104 are then imaged under a microscope 214.

[0093] As illustrated in FIG. 9, the present disclosure further provides for certain products of manufacture and compositions to facilitate the implementation of the various techniques described as described herein, i.e., for detecting one or more biomarker targets 104 within a biological sample 100 (FIG. 1). As one illustrative example, a useful composition 900 can be assembled that includes: sets of primary probes 542A configured to bind to specific biomarker targets 104; components for amplification 978; sets of secondary probes 542B; and in some cases, a set of tertiary probes 542C; wherein at least one set of probes 542 is configured to be a molecular probe 542 and / or a FEIR probe 552 by including a distinct functional group 546 and optionally a luminescent moiety 550.

[0094] In some cases, the composition may further include a set of written instructions 980 for practicing one or more of the techniques of this disclosure, in addition to various other relevant components, such as reagents 982 for sample fixation, permeabilization, hybridization, blocking, washing, buffering, mounting, expanding, etc. For example, in nucleic acid detection, the composition comprises target nucleic acid(s) 104, sets of primary probes 542A (often oligonucleotides), optionally, an amplification component 978, and optionally, a set of secondary probes 542B that stain the “readout” domains 548 of the primary probes 542A. In another example, for protein imaging, the composition comprises: target protein(s) 104 or epitopes; sets of primary probes 542A, for example, antibodies or antigen-binding fragments thereof (including Fab fragments or single-domain antibodies (sdAb), also known as “nanobodies”) or their derivatives; optionally, an amplification component 978; and optionally, a set of secondary probes 542B that bind to the primary probes 542A or products of a target-binding-mediated event or amplification. It should be understood that the combination of the above embodiments or multiple sets of components to label biomarker targets 104 using molecular probes 542 and / or FEIR probes 552 can be used together within a suitable composition for high content screening and multiplexed target detection.

[0095] FIG. 10 is a flowchart 1000 illustrating a novel approach for using molecular probes 542 and FEIR probes 552 with distinct, differentiable vibrational resonances 440 to label samples 100 and provide for high-multiplex microscopic imaging. At Step 1002, the labeling technique first includes providing a biological sample 100, such as collecting the sample from a biological specimen or subject. At Step 1004, the technique includes delivering a set of vibrationally encoded molecular / FEIR probes 542 / 552 to the sample, and allowing them to interact (e.g., bind, conjugate, etc.) with the sample.

[0096] At Step 1006, the technique includes distinguishing between two subsets of the vibrationally encoded probes 542 / 552, those that successfully interacted with biomarker targets 104 within the sample 100 and those that did not. The unattached probes 542 / 552 may be separated from the probes that successfully attached to their biomarker target 104 and washed away from the sample 100. At Step 1008, the technique includes proceeding to image the sample 100 (e.g., using any or all of the microscopy techniques described throughout this disclosure), in order to map the locations of biomarker targets 104 within the sample 100.

[0097] Finally, at Step 1010 of flowchart 1000, the technique provides for the removal of the probes 542 / 552 that did interact with targets 104 from the sample 100, allowing for a subsequent round or cycle of labeling to occur. In this manner, an unlimited number of biomarker targets 104 can, in principle, be localized.

[0098] FIG. 11 is a conceptual block diagram of an example computing system 1100 that may be used to implement one or more aspects of a sample-analysis system, as described herein. In particular, computing system 1100 is an example implementation of computing device 222 of system 206 of FIG. 2. Computing system 1100 can include volatile and nonvolatile memory 1110, such as random-access memory (RAM) 1102 and read-only memory (ROM) 1104, as well as one or more processing devices 1106 (e.g., central processing units (CPUs), graphical processing units (GPUs), and the like). Computing system 1100 can include various media devices 1108, such as a hard-disk module, an optical-disk module, and so forth. Computing system 1100 may perform any of all of the computer-based operations (e.g., data-processing and image-generation) described throughout this disclosure by processing device(s) 1106 executing instructions stored in memory 1110 (e.g., RAM 1102, ROM 1104, and the like).

[0099] More generally, instructions and other program information may be stored on any computer-readable medium 1110, including, but not limited to, static-memory storage devices, magnetic storage devices, and optical storage devices. The term “computer-readable medium” also encompasses plural storage devices. In all cases, computer-readable medium 1110 represents some form of physical and tangible entity. By way of example and not limitation, computer-readable medium 1110 may comprise storage media (e.g., RAM 1102, ROM 1104, EEPROM, Flash memory or other memory technology, CD-ROM, digital versatile disks (DVD) or other optical storage, etc.) and / or communications media (e.g., wired media such as wired network or direct-wired connection, and wireless media such as acoustic, RF, infrared, and other wireless media).

[0100] Computing system 1100 includes an input / output (I / O) module 1112 for receiving various inputs (via input modules 1114), and for providing various outputs (via one or more output modules). One particular output module mechanism may be a presentation module 1116 and an associated GUI 1118. Computing system 1100 may also include one or more network interfaces 1120 for exchanging data with other devices via one or more communication conduits 1122. In some aspects, one or more communication buses 1124 communicatively couple the above-described components together.

[0101] Communication conduit(s) 1122 may be implemented in any manner, such as via a local-area network (LAN), a wide-area network (WAN, e.g., the Internet), and the like, or any suitable combination thereof. Communication conduit(s) 1122 may include any combination of hardwired links, wireless links, routers, gateway functionality, name servers, and the like, governed by any protocol or combination of protocols.

[0102] Alternatively, or in addition, any of the functions described herein may be performed, at least in part, by one or more hardware logic components, such as Field- Programmable Gate Arrays (FPGAs), Application-Specific Integrated Circuits (ASICs), Application-Specific Standard Products (ASSPs), Systems-on-a-Chip (SOCs), Complex Programmable Logic Devices (CPLDs), and the like.

[0103] In some example implementations of system 206, the image-processing and analysis functions may be deployed on a local PC, a cloud-based architecture, or a combination of both. For instance, raw data or partially processed data from detector 220 may be transmitted to an external (i.e., remote) cloud-based computer for processing, for instance, to generate imagery 208, which may then be transmitted back to the local device 222 / 1100 for display. In some such examples, the PC provides for a GUI 1118 that provides intuitive image representation functions. In additional embodiments, data can be analyzed using high bit-rate electronics such as CPUs, GPUs, and FPGAs, resulting in higher scalability and high parallelization processing.

[0104] In additional embodiments, image acquisition, image analysis, processing, and visualization is achieved via commercial, open source, or custom software packages. Examples include, but are not limited to, MATLAB, Lab VIEW, ImageJ, CellProfiler, and Napari. Additional code made be written to incorporate additional functionality in languages such as C, C++, R, Python, or any other programming language. They may export the data into any format of interest such as TIFF, JPEG, PNG, PDF, GIF, BMP, excel, doc, etc.

[0105] In some examples, image processing includes steps that require training, annotation, or supervision, such as manual cell membrane segmentation or otherclassification. This class of labeled data is treated as ground-truth and may be used to develop a trained model using algorithms such as neural networks used in machine learning (ML) and artificial intelligence (Al) image analysis applications. Incorporating such mathematical models into the analysis pipeline provides insight into attributes (morphological and molecular) that my otherwise be missed.

[0106] Notably, the high-multiplexing capabilities of molecular probes 542 and FEIR probes 552, along with the wide range of native (endogenous) spectroscopic features detectable through the microscopy techniques discussed herein, opens the possibility to acquire datasets of multispectral or hyperspectral hypercubes, including at multiple time intervals to track the multi omic dynamics of the biological sample 104 during longitudinal imaging studies. To accommodate these datasets, some embodiments include linear dimensional reduction techniques during image processing such as principal component analysis (PCA), which is an unsupervised approach to reduce complexity while minimizing information loss through spectral unmixing. Applying a pseudo-coloring, or look-up-table, to the resulting orthogonal vector components facilitates intuitive segmentation between features within the image with similar biomarker composition. This analysis may be used in conjunction with other image processing techniques such as deconvolution or statistical image analyses including (multivariate) analysis of variance or t-tests. Such quantitative imaging and bioinformatics may be useful in evaluating significance levels (e g., p-value and confidence intervals) and identifying cofounding variables which can introduce errors during hypothesis testing in target discovery, lead optimization, and clinical investigations. It may also be useful as a process analytical technology for inline and / or offline analyses during biomanufacturing.

[0107] The embodiments above are chosen, described and illustrated so that persons skilled in the art will be able to understand the invention and the manner and process of making and using it. The descriptions and the accompanying drawings should be interpreted in the illustrative and not the exhaustive or limited sense.

Claims

AMENDED CLAIMS received by the International Bureau on 17 June 2025 (17.06.2025) What is claimed is:

1. A method for spatially analyzing a biological sample comprising a plurality of biomarker targets, the method comprising: performing two or more rounds of imaging the biological sample, each round comprising: in situ labeling of the biological sample with a plurality of probes each configured to bind to one of the plurality of biomarker targets according to a respective combinatorial labeling configuration, wherein each probe comprises a functional group defined by a respective predetermined molecular-vibrationalresonance characteristic, at least one of which comprises a resonant vibrational frequency between about 400 cm'1and about 4000 cm'1; positioning the biological sample on a stage of an imaging device; directing a first radiation beam from a first radiation source onto a first area of the biological sample, wherein the first radiation beam comprises pulsed tunable infrared (IR) light having a wavelength of about 2.5 microns to about 25 microns; directing a second radiation beam from a second radiation source onto a second area of the biological sample, wherein the second radiation beam comprises ultraviolet (UV) light, visible-spectrum light, or near-IR light, and wherein the first and second areas of the biological sample at least partially overlap; receiving, by a photodetector from a lens of the imaging device, photons comprising: light derived from the second radiation beam after interacting with the biological sample; or luminescence emitted by one or more of the probes in response to absorbing radiation from both the first and second radiation beams; outputting, by the photodetector, an electrical signal indicating optical intensities of the photons; generating, by processing circuitry, imagery indicative of the biological sample based on the electrical signal; and deactivating or removing the plurality of probes from the biological sample for a subsequent round of imaging; determining, by the processing circuitry based on the imagery from each of the two or more rounds of imaging, spatial features for the plurality of biomarker targets, whereinthe spatial features define characteristics, properties, quantities, interactions, or structural organizations of the plurality of biomarker targets; and decoding, by the processing circuitry, probe signatures within the electrical signals from the two or more rounds to identify distinct biomarker species within the plurality of biomarker targets.

2. The method of claim 1, wherein at least one of the plurality of probes further comprises a luminescent, light-emitting moiety in communication with the respective functional group; and wherein the light-emitting moiety comprises a distinct or defined spectral- absorption, spectral-emission, or luminescence-lifetime characteristic.

3. The method of claim 1 or claim 2, wherein the plurality of biomarker targets comprises nucleotides, DNA, RNA, peptides, proteins, metabolites, or small molecules.

4. The method of claim 1, wherein, for each round of imaging: receiving the photons comprises: receiving a first set of photons derived from the second radiation beam after interacting with the biological sample while the biological sample is illuminated by the first radiation beam; and receiving a second set of photons derived from the second radiation beam after interacting with the biological sample while the biological sample is not illuminated by the first radiation beam; outputting the electrical signal comprises: outputting a first electrical signal indicative of the first set of photons; and outputting a second electrical signal indicative of the second set of photons; and generating the imagery comprises: generating first image data based on the first electrical signal; and generating second image data based on the second electrical signal; and resolving a difference between the first image data and the second image data, wherein the difference comprises the imagery, and wherein the imagery indicates a quantity and a spatial distribution of the plurality of probes.

5. The method of claim 1, wherein, during at least one of the rounds of imaging, at least one of the photons comprises the light derived from the second radiation beam after interacting with the biological sample; wherein the method further comprises determining, by the processing circuitry based on the electrical signal, quantitative phase information for the at least one photon; and wherein generating the imagery comprises generating, by the processing circuitry, the imagery based on the quantitative phase information.

6. The method of claim 1, wherein generating the imagery comprises generating two- dimensional (2-D) imagery, three-dimensional (3-D) imagery, longitudinal imagery, multi spectral imagery, or hyperspectral imagery.

7. The method of claim 1, wherein the biological sample comprises: bacteria, fungi, microalgae, plant, animal, or mammalian cells grown in 2-D or 3-D cultures; a biopsied, fresh-frozen, formalin-fixed paraffin-embedded (FFPE), or preserved tissue; or an organism.

8. The method of claim 1, wherein the plurality of probes comprises a protein, an antibody, an antibody fragment, a nucleic-acid, an oligonucleotide, an antibody - oligonucleotide conjugate, or a biomarker-target-specific ligand.

9. The method of claim 1, wherein the stage used to retain the biological sample comprises a compartment and microfluidic system that allows fluid flow for processing the sample during sequential rounds of imaging.

10. The method of claim 1, wherein decoding the probe signatures comprises executing, by the processing circuitry, a barcoding-error-reduction function.

11. The method of claim 1, further comprising reducing a signal overlap within the electrical signals using super-resolution microscopy (SRM).

12. The method of claim 1, wherein in situ labeling of the biological sample with the plurality of probes comprises: labeling the biological sample with a plurality of primary probes each comprising: a biorecognition motif; a first region configured to selectively bind to a complementary second region of a respective biomarker target within the biological sample; and an extension element; labeling the biological sample with a plurality of secondary probes each comprising: a complementary element specifically configured to bind to the extension element on one of the plurality of the primary probes; and the molecular functional group defined by the respective predetermined molecular-vibrational-resonance characteristic; and applying a reagent configured for sample fixation of the biological sample, permeabilization of the biological sample, hybridization of the biological sample, blocking the biological sample, washing the biological sample, buffering the biological sample, mounting the biological sample, or expanding the biological sample.

13. The method of claim 12, wherein one or more of the plurality of secondary probes further comprises a signal -amplification component.

14. The method of claim 12 or claim 13, wherein one or more of the plurality of secondary probes further comprises a light-emitting moiety vibronically coupled to the molecular functional group, wherein the light-emitting moiety is configured to generate a signal having a distinct or defined spectral-absorption, spectral-emission, or luminescencelifetime characteristic.

15. The method of claim 12, wherein one or more of the plurality of primary probes or the plurality of secondary probes comprises a protein, an antibody, an antibody fragment, a nucleic-acid, an oligonucleotide, an antibody-oligonucleotide conjugate, or a biomarker- target-specific ligand.

Images