Apparatuses and methods involving detection of binding and interactions of molecules with target proteins in live cells

Abstract

In certain examples, methods and apparatuses are directed to small molecules (e.g., drug-like molecules) binding to a target protein in live cells. In a specific example, a method includes: processing a laser beam, from a quantum cascade laser (QCL), carrying light in a wavelength range that overlaps a mid-IR range; using optics, in response to the laser beam, to process the light in a set of related beams along respective paths, including a reference path and a sample path, respectively towards a reference target and a sample target that includes living cells: and detecting, via a vibrational spectroscopy detector, small molecules binding to a target protein in live cells. The detection is carried out by collecting incident light, relative to the reference target and from the sample target, carried along respective incident paths and by discerning (e.g., measuring) vibrational probe frequency shifts in the sample target.

Description

STFD.469PCT (S24-491) 1APPARATUSES AND METHODS INVOLVING DETECTION OF BINDING AND INTERACTIONS OF MOLECULES WITH TARGET PROTEINS IN LIVE CELLSFEDERALLY-SPONSORED RESEARCH AND DEVELOPMENT

[0001] This invention was made with Government support under contract 1656518 awarded by the National Science Foundation and under contract GM118044 awarded by the National Institutes of Health. The Government has certain rights in the invention.BACKGROUND

[0002] Aspects of the present disclosure are related generally to the field of small molecule spectroscopy, as may be exemplified in connection with transmission-based infrared (IR) spectroscopic equipment.

[0003] In exemplary contexts, detection and quantification of drug binding is one of the most sought-after applications of analytical chemistry' to human health and comprises a multi- billion-dollar industry. While many drug binding techniques rely upon separation methods or highly perturbative techniques that are carried out in vitro, there is a decisive advantage to being able to directly probe drug protein interactions in cellulo. In cellulo assays account for factors that may influence binding, such as metabolic degradation, trafficking, or off-target competition. However, often cell-based assays for drug discovery rely on indirect physiological measures of downstream cellular functions or toxicity’, which give no direct information on molecular drug-protein interactions.

[0004] There have been various attempts to detect binding interactions directly in living cells. Such efforts have included, for example, use of vibrational probes to serve as minimally perturbative reporters of their cellular environment, by observing frequency shifts that are physically interpretable in terms of environmental electrostatics molecular conformation, or hydrogen bonding. Many common vibrational probes like nitriles (C=N) are intrinsic to drugs themselves or can be engineered selectively onto the target protein scaffold via amber suppression and can thus serve as effective reporters of drug-protein interactions (e.g., in contexts such as proteins, lipid membranes, surfaces, microdroplets and metal-organic frameworks). Unfortunately, efforts to advance use of such vibrational probes to detect interactions in and around living cells have been problematic due to insufficient instrumentation sensitivity, which has been unable to distinguish adequately the intrinsically weak cellular-based vibrational signal due to interference from the cellular background,STFD.469PCT (S24-491) 2 instrument noise, and, oftentimes present biocompatibility issues. Consequently, such efforts which endeavor to use vibrational probes suffer from poor detectivity of weak signals.

[0005] Fluorescence-based measurements, typically fluorescence-based assays, are the primary methods for observing binding interactions directly in cells. Fluorescence-based assays are advantageous in that they are versatile and provide relatively sensitive detection for quantifying a wide range of analytes by monitoring changes in fluorescence signals such as intensity, lifetime, and / or frequency spectrum. Unfortunately, fluorescence-based measurements are disadvantageous in that they have limited applicability for non-fluorescent drug systems, and / or they require labeling by perturbative fluorescent labels. Further, such fluorescent labels often require the attachment of bulky dye molecules, which are comparable in size to the drug itself and can interfere with the normal function, localization, and interactions of the target proteins of the living cells, thereby leading to possibly experimental artifacts and inaccurate data interpretation.

[0006] These and other matters have presented challenges to efficiencies of such detection and / or quantification for a variety of applications.STFD.469PCT (S24-491) 3SUMMARY OF VARIOUS ASPECTS AND EXAMPLES

[0007] Various examples / embodiments presented by the present disclosure are directed to issues such as those addressed above and / or others which may become apparent from the following disclosure. For example, in certain examples, aspects of the present disclosure are directed to methods and apparatuses that directly probe small molecule protein interactions in cellulo, and in certain specific applications or implementations, thereby overcoming limitations of certain fluorescence-based techniques.

[0008] In certain other examples, methods and optical-circuitry-based structures are directed to small molecules (aka “drug-like molecules”) binding to a target protein in live cells. In a specific example, a method includes: processing a laser beam, from a quantum cascade laser (QCL), carrying light in a wavelength range that overlaps a mid-IR range; using optics, in response to the laser beam, to process the light in a set of related beams along respective paths, including a reference path and a sample path, respectively towards a reference target and a sample target that includes living cells; and detecting, via a vibrational spectroscopy detector, small molecules binding to a target protein in live cells. The detection is carried out by collecting incident light, relative to the reference target and from the sample target, carried along respective incident paths and by discerning (e.g., measuring) vibrational probe frequency shifts in the sample target.

[0009] In another specific example, an apparatus comprises: a QCL, a set of optics, and a vibrational spectroscopy detector. The QCL is to direct a laser beam carrying light in a wavelength range that overlaps a mid-IR range. The set of optics is to respond to the laser beam, to process the light in a set of related beams along respective paths. The respective paths include a reference path and a sample path through which the light is directed respectively towards a reference target and towards a sample target, the latter of which includes living cells. The vibrational spectroscopy detector is to detect small molecules binding to a target protein in the living cells by collecting incident light, relative to the reference target and from the sample target, carried along respective incident paths and by discerning (e.g.. measuring) vibrational probe (e.g., nitrile) frequency shifts in the sample target.

[0001] In certain other examples which may also build on the above-discussed aspects, methods and apparatuses may include implementing a laser beam source such that the laser beam carries the light in a wavelength range approximately from 2000 cm’1to 2300 cm4, and the set of optics and the vibrational spectroscopy detector are cooperatively configured to set a 250 micron sample pathlength for the sample target, or otherwise cooperatively configuredSTFD.469PCT (S24-491) 4 to set a sample pathlength for the sample target that is adequate to measure small molecules binding to the target protein in the living cells. Further, such aspects may include detecting, via the vibrational spectroscopy detector (e.g., as a balanced, thermoelectrically cooled Mercury Cadmium Telluride detector), changes in vibrational frequency of the chemical probes in response to binding of the proteins to an externally -added drug-like molecule. [001 1 j Additional specific examples, also in accordance with the present disclosure, relate to the above methodology and / or apparatuses and also may build on the abovediscussed aspects. As a few specific examples in this regard: the optics further includes no more than three neutral density filters optically configured between the QCL and a beam splitter, as part of the optics, to separate the laser beam from the QCL into the set of related beams; and the sample target may include vibrational probes embedded in proteins within the living cells, and / or with the vibrational spectroscopy detector being configured to detect at least some of the proteins that are tagged, for example, by nitrile diatoms. Also, the set of optics can be configured to include: one or more neutral density filters to balance beams from among the set of separated beams; an optical diffuser as a polarization scrambler to depolarize an incident laser beam; and / or CaF2 focusing optics to pass the incident light, from the reference target and from the sample target, before the incident light reaches the vibrational spectroscopy.

[0012] In other specific examples, aspects of the present disclosure are directed to use of a vibrational spectroscopy detector. As such, the vibrational spectroscopy detector may be configured to detect: vibrational probes of frequency 2000-2300 cm ', such as nitrile vibrational probes embedded in proteins within the living cells, in response to spectral shifts anywhere within the same 2000-2300 cm1range, and observed in proof-of-concept experiments to shift within various ranges, for example, by 0.3 cm as one of the smaller shifts and up to 15 cm as one of the larger shifts; molecular vibrational signatures embedded in proteins within the living cells; changes in vibrational frequency of the chemical probes in response to binding of the proteins to an externally-added drug-like molecule; and / or covalent drug binding in a cellular environment.

[0013] In yet more specific aspects, the above-characterized Mercury Cadmium Telluride detector structure are directed to detecting changes in vibrational frequency of the chemical probes in response to covalent drug binding of the proteins to an externally -added drug-like molecule. When considering covalent drug binding to a target protein, measurements in live cells are important to account for significant factors such as crossing of the membrane, off- target binding, or degradation pathways.STFD.469PCT (S24-491) 5

[0014] In yet another specific example, the present disclosure is directed to: a QCL- based transmission infrared (IR) source to provide balanced detection and to enhance sensitivity to nitrile vibrational probes embedded in proteins within live cells; and detecting small-molecule covalent binding in a cellular environment with spectral shifts demonstrated in an approximate range (e.g.. from 0.3 cm to 15 cm1as in specific experimentation, or significantly larger frequency shifts), therein enabling the ability of a QCL-based spectrometer to monitor sensitivity' in drug-protein interactions.

[0015] In yet further aspects of the present disclosure, exemplary instrumentation in a detection system includes: a QCL-based spectrometer configured to operate with transmission IR detection for vibrational probes in the wavelength range of 2000-2300 cm’1(e.g. characteristic of nitrile, azide, carbon-deuterium bonds, isonitriles, and alkyne carboncarbon stretches), thereby allowing two-fold or more (e.g., up to five-fold) longer pathlengths to be used for aqueous samples with spectra measured in the range from 2000 cm’1to 2300 cm’1, increasing signal-to-noise for most oscillators within this spectral region accordingly (e.g., by five-fold compared to conventional FTIR instruments). For example, when measuring benzonitrile in water, the limit of detection (LOD) drops from 80 pM with a Bruker Vertex 70 FTIR to just 18 pM, according to examples of the present disclosure, for the same number of scans. Building on such further aspects of the present disclosure and in addition to substantial gain in sensitivity, such instrumentation with a double-beam optical format enables: an ability to probe a different spectral region relative to known QCL-based spectrometers; use of much longer pathlengths than previously possible (e.g., surpassing the limitations seen in the 1500-1700 cm’1range for aqueous samples); and an enhancement in sensitivity rendering effective detection of chemically labeled proteins in complex cellular environments which is especially useful for high-precision molecular analysis.

[0016] The above discussion is not intended to describe each aspect, embodiment or every implementation of the present disclosure. The figures and detailed description that follow also exemplify various embodiments.STFD.469PCT (S24-491) 6BRIEF DESCRIPTION OF FIGURES

[0017] Various example embodiments, including experimental examples, may be more completely understood in consideration of the following detailed description in connection with the accompanying drawings, each in accordance with the present disclosure (unless otherwise stated), in which:

[0018] FIGs. 1A, IB and 1C are respective graphs showing detection limits of a conventional spectrometer (FIG. 1A) and of a spectrometer according to the present disclosure (FIG. IB), and a block diagram of an example spectrometer according to the present disclosure (FIG. 1C);

[0019] FIG. 2 is a schematic showing a strategy, according to the present disclosure, associated with labeling a target protein (photoactive yellow protein, or PYP) with nitrile vibrational probes, for purposes of detecting covalent binding of a drug-like molecule using vibrational spectroscopy;

[0020] FIGs. 3A and 3B are depictions to show incorporation of chromophore p- coumaric acid (pCA) into PYP by nucleophilic attack at C69, and the resulting observed spectral shifts using the disclosed apparatus;[0021 j FIG. 4 is a graph showing spectral power distribution in an experimental set up according to the present disclosure;

[0022] FIG. 5 a graph show ing a detector output after being processed using the block diagram of FIG. 1C;

[0023] FIGs. 6A and 6B are respective graphs, for detected and reference outputs, showing calibration of MCT detector output to incident power using nonlinear curve fitting;

[0024] FIG. 7 is a graph showing recalibration of original laser scan frequencies determined by rime from initial scan trigger, based on periodic w avelength triggers from the QCL;

[0025] FIGs. 8A, 8B, 8C, 8D and 8E are respective graphs showing spectra and concentration-absorbance correlations (FIGs. 8A and 8B) from a conventional Bruker Vertex 70 Fourier Transform Infrared (FTIR) spectrometer and a spectrometer according to the present disclosure, with FIGs. 8C and 8D showing respective correlation plots betw een concentration and absorbance for the associated figures of FIGs. 8A and 8B;

[0026] FIG. 9 is a graph showing plots of baseline IR spectra recorded by an exemplary QCL spectrometer according to the present disclosure, using 16 scans versus a conventional FTIR approach;STFD.469PCT (S24-491) 7

[0027] FIG. 10 is a graph showing root mean square (RMS) noise for a blank (balanced) absorption spectrum through 250-pm pathlength water generally decreased with increasing number of scans;

[0028] FIG. 11 is a graph showing Benzonitrile detection on a ThermoFisher Nicolet iS50 FTIR instrument;

[0029] FIG. 12 is a graph showing further experimental data with spectra plotted for 10 rnM 1 -Hexyne in water after 16 scans on either the Bruker Vertex 70 FTIR using a 50 pm pathlength, or on the QCL using a 250 pm pathlength, which shows better sensitivity of the QCL in this frequency range;

[0030] FIG. 13 is a graph showing yet further experimental data with spectra of 1 mM 4- azidophenylalanine in water after 16 scans on either the Bruker Vertex 70 FTIR (here using 50 and 25 pm spacers to reduce etalon fringes), or on the QCL using a 250 pm pathlength, also showing better sensitivity of the QCL in this frequency range; and

[0031] FIG. I4A and 14B are respective graphs showing spectra of N- Cyclohexylformamide, singly deuterated at the carbonyl, at concentrations of 1 mM and 5 mM in water to compare sensitivity of (A) the Bruker Vertex 70 FTIR at a 50 pm pathlength, and (B) the QCL spectrometer at a 250 pm pathlength, with 16 scans used on both instruments.

[0032] While various embodiments discussed herein are amenable to modifications and alternative forms, aspects thereof have been shown by way of example in the drawings and will be described in detail. It should be understood, however, that the intention is not to limit the disclosure to the particular embodiments described. On the contrary, the intention is to cover all modifications, equivalents, and alternatives falling within the scope of the disclosure including aspects defined in the claims. In addition, the term “example” as used throughout this application is only by way of illustration, and not limitation.STFD.469PCT (S24-491) 8DETAILED DESCRIPTION

[0033] Aspects of the present disclosure are believed to be applicable to a variety of different types of apparatuses, systems and methods involving devices characterized at least in part by use of transmission-based infrared (IR) spectroscopic equipment and / or approaches involving a non-perturbative vibrational spectroscopy -based assay to directly probe small molecule protein interactions in cellulo and demonstrated in the context of experimental / more-detailed (e.g., proof-of-concept) examples. For example, while many of these examples involve vibrational probes, such as nitrile, in the wavelength range of 2000-2300 cm'1, other types of vibrational probes may be used (e.g., azide, carbon-deuterium bonds, isonitriles, and alkyne carbon-carbon stretches). Accordingly, while the present disclosure is not necessarily limited to such exemplary aspects, an understanding of specific examples in the following description may be understood from the following discussion which may be used such specific contexts.

[0034] In certain examples, methods and apparatuses are directed to small molecules (e.g., drug-like molecules) binding to a target protein in live cells. In a specific example, a method includes: processing a laser beam, from a quantum cascade laser (QCL), carrying light in a wavelength range that overlaps a mid-IR range; using optics, in response to the laser beam, to process the light in a set of related beams along respective paths, including a reference path and a sample path, respectively towards a reference target and a sample target that includes living cells; and detecting, via a vibrational spectroscopy detector, small molecules binding to a target protein in live cells. The detection is carried out by collecting incident light, relative to the reference target and from the sample target, carried along respective incident paths and by discerning (e.g., measuring) nitrile frequency shifts in the sample target.

[0035] Accordingly, in the following description various specific details are set forth to describe specific examples presented herein. It should be apparent to one skilled in the art, however, that one or more other examples and / or variations of these examples may be practiced without all the specific details given below. In other instances, well known features have not been described in detail so as not to obscure the description of the examples herein. For ease of illustration, the same connotation and / or reference numerals may be used in different diagrams to refer to the same elements or additional instances of the same element. Also, although aspects and features may in some cases be described in individual figures, it will be appreciated that features from one figure or embodiment can be combined with features of another figure or embodiment even though the combination is not explicitlySTFD.469PCT (S24-491) 9 shown or explicitly described as a combination. Further (and especially in connection with the following discussion and the discussion presented in the Appendices of the underlying U.S. Provisional Application (Serial No. 63 / 735,202 filed on December 17, 2024, STFD.469P1 / S24-491. to which priority is claimed), unless otherwise indicated ranges (of any, and all metrics) are merely exemplary of “approximate ranges’" wherein this term may be understood to vary the bound (s) of the range (e.g., using improved and / or degraded material- or circuit-based design parameters) by a degree of anywhere from 10-to-20 percent, and, in the context of comparison to an improvement over a previously -reported effort, by a degree of improvement of at least 10 percent.

[0036] Consistent with the above aspects, such a manufactured device or method of such manufacture may involve aspects presented and claimed in the above-referenced U.S. Provisional Application (Serial No. 63 / 735,202. To the extent permitted, such subject matter is incorporated by reference in its entirety generally and to the extent that further aspects and examples (such as experimental and / more-detailed embodiments) may be useful to supplement and / or clarify.

[0037] As noted above, certain exemplary aspects of the present disclosure involve methodology and apparatuses, which may also build on the above-discussed aspects, may include implementing a laser beam source such that the laser beam carries the light in a wavelength range approximately from 2000 cm’1to 2300 cm’1, and the set of optics and the vibrational spectroscopy detector are cooperatively configured to set a sample pathlength for the sample target that is adequate to measure small molecules binding to the target protein in the living cells (e.g., a 250 micron sample pathlength for the sample target). Further, such aspects may include detecting, via the vibrational spectroscopy detector (e.g., as a balanced thermoelectrically cooled Mercury' Cadmium Telluride detector), changes in vibrational frequency of the chemical probes in response to binding of the proteins to an externally- added drug-like molecule.

[0038] Specific aspects of the present disclosure concern vibrational spectroscopy of live cells by using QCL-based detection of drug binding to target proteins in live cells, for example, as improvements over known approaches for such drug-like molecule interactions. While many cell-based assays examine top-down phenotypic physiological responses of cells to drugs, such as cell proliferation, toxicity, production of markers, activation of specific signaling pathways, downstream alterations and changes in morphology, such previous assays are rarely direct molecular readouts of the drug-protein interaction. Hence, they' are generally non-quantitative, rarely giving information on binding constants to specific proteinsSTFD.469PCT (S24-491) 10- which is believed to be provided in specific example embodiments of the present disclosure. Moreover, the mechanism of action for such drugs is not even necessarily clear, as multiple mechanisms of action could result in similar physiological consequences. Further, cellular physiological readouts may not necessarily be intuitive or simple to measure for several drug targets that do not cause significant, noticeable effects on the cell, making these kinds of assays difficult to generalize.

[0039] According to specific examples of the present disclosure, by using QCL-based detection of drug binding to target proteins in live cells as an improvement over current cellbased assays, limit of detection (LOD) experiments with benzonitrile (an analogue for cyanophenylalanine in proteins) show that between the FTIR and this QCL spectrometer with balanced detection approach, the apparatus provides a five-fold better sensitivity for detection of the nitrile vibration when compared to the top-of-the-line Bruker Vertex 70 FTIR spectrometer. LOD is determined by the concentration of benzonitrile required for the nitrile absorbance peak to be three times larger than the root-mean squared (RMS) noise over the range of 2160-2260 cm'1(including where the 2235 cm'1benzonitrile peak would be). With the flux provided by the globar source used in a Bruker Vertex 70 FTIR, pathlengths greater than 50 pm cannot be used due to a strong absorbance feature arising from water’s libration mode, which is centered at 2100 cm'1. Pathlengths larger than 50 pm in the FTIR lead to near-complete blockage of light in this spectral range, thereby limiting the FTIR’s use for pathlengths beyond 50 pm in this window . In contrast, with the high photon flux from a QCL source, this limitation is bypassed with longer pathlengths of up to 250 pm being achieved. For the FTIR, the maximum reasonable pathlength that can therefore be used for aqueous solutions is 50 pm, which gives an LOD for benzonitrile of 80 pM. Using the QCL spectrometer, the pathlength can be increased to 250 pm, resulting in a benzonitrile LOD of 16 pM. Both theoretical LODs are verified experimentally by observing the benzonitrile peak becoming discernible from the background near these concentrations. Hence, the QCL-based instrument gives substantially better sensitivity for nitriles compared to conventional FTIR instruments.

[0040] While these experimental results are demonstrated in proof of concept for proteins in bacterial cells, this assay approach is applicable to mammalian cells as well. Furthermore, while such proof-of-concept experiments involve detecting binding of covalent inhibitors to target proteins for the purpose of clear verification by mass spectrometry following cell lysis and purification, this assay approach does not require the molecular interaction to be covalent to observe frequency shifts of the nearby nitrile groups, whether onSTFD.469PCT (S24-491) 11 the protein or the binding molecule of interest (e.g., specific aspects of the present disclosure are useful in applications when nitriles are already in the drug itself (and many drugs do already contain nitriles or other useful vibrational probes), rather than needing to be incorporated through Amber Suppression into the protein). This follows because nitriles inherently report on changes to their immediate protein environment; thus granting an ability to detect classes of molecular interactions beyond the normal scope of mass spectrometry experiments, as well as in a more native cellular state prior to cell rupturing.[0041 j Unmet need exists at this stage of drug discovery because most cell-based assays currently used are either indirect (measuring cellular physiological responses, which provide no direct information on molecular interactions with the target protein) or rarely direct fluorescence-based detection, requiring fluorescent labels that are often challenging to incorporate into robust assays. Fluorophores can also introduce steric hindrance, potentially impairing the interaction between the drug and protein. According to such examples of the present disclosure, the disclosed infrared spectroscopy-based assay addresses this gap by providing direct, quantitative insights into molecular interactions without the need for fluorescent labels. This will enhance the efficiency of drug screening prior to animal testing, saving time and reducing costs in the drug development process. Ultimately, such aspects of the present disclosure facilitate the development of more effective and beneficial pharmaceuticals.

[0042] It is also noted that small molecule protein interactions are ubiquitous in the development of recombinant bacteria that are rewired to produce engineered enzymes important to the food and agricultural industry. In other examples, the potential to modify proteins based on such interactions, using the principle of directed evolution, is utilized based on this assay approach by selecting for better binding interactions or larger vibrational frequency shifts. Thus even beyond the pharmaceutical industry', the agricultural industry' which routinely uses recombinant bacteria for designer enzy mes at large scales could also find uses for this assay.

[0043] Consistent with the above discussion, certain aspects of the present disclosure involve experimentation (experimental efforts) to extend the capabilities of QCU-based vibrational spectroscopy of proteins into a mid-IR frequency regime, with balanced detection for monitoring genetically encoded nitrile vibrational probes in live cells as molecular reporters of protein-small molecule binding. These experimental efforts demonstrate a significant sensitivity' improvement, up to five-fold, over conventional FTIR spectroscopy for detection of aromatic nitriles due to longer pathlengths for transmission measurementsSTFD.469PCT (S24-491) 12 accessible to the QCL setup. Then the QCL spectrometer is used to detect small molecule binding to proteins, which is often a key objective in drug discovery, using nitrile frequency shifts as direct indicators of fine protein structural perturbations caused by binding events in live cells. High-level polarizable molecular dynamics simulations provide a physical rationale for the observed experimental results.

[0044] Certain exemplary findings presented herein in accordance with such experimental efforts of the present disclosure validate the capacity and advantages of a QCL- based assay approach to detect subtle structural perturbations resulting from covalent bond formation at reactive cysteine residues, which is a common motif in many modem covalent drugs. While the experimental efforts and related current proof-of-concept studies of the present disclosure were conducted in bacterial cells, bacterial proteins account for a significant fraction of therapeutic targets of interest in modem drug discover}'. Furthermore, with emerging techniques in amber suppression, this assay approach is applicable to mammalian systems as well, where non-canonical amino acids can be incorporated efficiently. Further, the advantage in covalent drug uptake by mammalian cells, which do not have cell walls (as opposed to the E. coli used in such experimental efforts), is beneficial compared to the challenges of pCA which must be chemically activated immediately prior to use.

[0045] Additionally, while the vibrational probes were installed on the protein target in such experimental efforts, other example implementations of the present disclosure may use a similar assay which also take advantage of vibrational reporters located on the dmg itself — where a large portion of commercially available drugs already contain nitriles, alkynes, azides, or similar vibrational probes. This will come at the expense of measuring selectivity of a drug for a particular protein, as labels are not specific to a target protein, but will open avenues to looking at the overall fate of the bulk of the drugs added to the cells.

[0046] In yet other example implementations of the present disclosure, this assayapproach can be harmonized with directed evolution to optimize molecular interactions in cells. For instance, in connection with the present disclosure it has also been discovered that subculturing pCA-incubated cells in fresh media after the assay allowed for the regeneration of a new batch of cells available for further mutagenesis or future purification. Sensitive QCL spectroscopy enables the observation of vibrational frequency shifts in live cells, offering opportunities to study molecular interactions in their native environments. These abilities prove especially useful for understanding functions and properties of membrane andSTFD.469PCT (S24-491) 13 intrinsically disordered proteins, and more applications in which the native cellular environment is considered as being important.

[0047] To model covalent drug-binding events, the experimental efforts employ an example based on photoactive yellow protein (PYP) from Halorhodospira halophila. Apo- PYP. as expressed in cells, forms a covalent complex with the chromophore, p-Coumaric acid (pC A), a process that is analogous to covalent drug binding. PYP is a highly soluble globular protein from the PAS domain superfamily. While not a drug target in the conventional sense, it shares chemical characteristics with common covalent drug targets, especially those with reactive cysteines. Moreover, PYP as part of an optogenetic system is useful in measuring interactions with its chromophore and other binding partners in living cells. While a FRET- based assay using a fusion of a blue fluorescent protein and po-PYP has previously enabled an in cellulo binding study, it is noted that a vibrational spectroscopy approach without the use of a fusion protein construct is much more generalizable including to non-photoactive compounds involved in other systems. According to these experimental efforts of the present disclosure, the spectroscopic analyses of the nitrile frequency shifts using this assay, supported by detailed structural studies including X-ray cry stallography and high-level AMOEBA molecular dynamics (MD) simulations, indicate subtle changes in hydrogen bonding interactions involving the embedded nitriles within the protein upon drug incorporation. These findings underscore the utility of this assay for measuring drug binding in live cells at the molecular level using QCL-based IR spectroscopy frequency shifts.

[0048] In connection with these experimental efforts, a QCL-spectrometer system was designed for transmission IR measurements, as a specific experimental (and / or proof-of- concept) example. The design includes a double-beam infrared spectrometer, employing an external -cavity quantum cascade laser (EC-QCL) as the mid-IR radiation source to enable spectral acquisition within the 2000-2300 cm1range (FIGs. 4-5 infra). The instrument utilizes a balanced detection scheme, where the incident laser beam is split into reference and sample paths, and subsequently focused onto a thermoelectrically cooled HgCdTe (MCT) detector. The MCT detector is commercially available from VIGO Photonics (Poland), and a basic instrument used to configure the external-cavity7quantum cascade laser (EC-QCL) as the mid-IR radiation source may use a MIRcat-QT-2000 (commercially available from Daylight Solutions, San Diego, CA, USA), with a M2048-P tunable module that is cooled to 19 C using a closed-loop water chiller.

[0049] FIGs. 1 A, IB depict various structural and advantageous aspects of the present disclosure. FIG. 1A is a graph showing detection limits of a conventional spectrometerSTFD.469PCT (S24-491) 14(consistent with such previously -published work). In contrast to the limits depicted by the graph of FIG. 1 A, FIG. IB is a graph showing improved detection limits of a QCL- spectrometer according to the present disclosure.

[0050] FIG. 1C shows an optical (block) diagram of such a system including the QCL- spectrometer used in connection with the plots in FIG. IB. The optical diagram of FIG. 1C is calibrated (e.g., with calibration information discussed with FIGs. 6A-6B and 7) for use with a high-speed SR865A lock-in amplifier (e.g., available at Stanford Research Systems, Sunnyvale, CA, USA), which may also form part of the system. Custom MATLAB routines may be used for data acquisition and signal processing, including phase correction. Savitsky- Golay filtering, and Fourier filtering. Absorbance measurements may be derived from the ratio of beam intensities taken from reference and balanced channel voltage outputs, effectively mitigating fluctuations (e.g., primarily driven by the QCL source) using the balanced detection scheme. Two identical liquid sample cells equipped with CaF2 windows and Teflon spacers (50-250 pm; Pike Technologies, Fitchburg, WI, USA) were mounted on machined sample holders, serving as the sample and blank sample cells. The double-beam QCL spectrometer, as depicted in the optical block diagram of FIG. 1C, shows an optional lock-in amplifier, on the output side of the MCT detector, with lock-in detection for balanced absorption measurement.[0051 Other optional aspects include various optics (or optical elements) such as various neutral density (ND) Filters, a beam splitter (e.g., providing a 50-50 split) for implementations in which the laser source does not already provide such a split beam, diffusers, and focusing optics directing the split light beams to the balanced MCT detector. For example, infrared-compatible neutral density (ND) filters (e.g., Andover Optics Cat. No. 150FNIR-25; OD. = 1.5, 100FNIR-25; OD. = 1.0, and 030FNIR-12.5; OD. = 0.3) are optionally used to equalize the intensity of the two beams or bring the overall pow er intensity into the dynamic range of the instrument (only when aqueous samples are not present that attenuate the light due to the libration and bending combination mode of water centered at ~2100cm‘1). CaF2 diffusers were optionally included (Edmund Optics Cat. No. 19-734) to scramble polarization prior to detection. The two beams pass through sample and reference demountable liquid IR cells (Pike Technologies, Madison, WI, USA) designed for standard FTIR spectrometers (2 x 3” plate), which allowed comparison of the same sample across different IR spectrometers compatible with this cell used in the study. This cell features a needle plate with Luer-Lok fittings for convenient syringe filling with ~75 pL sample. The window^ dimensions are 32 x 3 mm, providing a 13 mm clear aperture. An O-ring seal modelSTFD.469PCT (S24-491) 15 was utilized, employing two small O-rings around the window filling holes instead of a flat gasket. A pathlength of 250 pm for each cell was used by combining two circular Teflon spacers of 200 pm and 50 pm each. Cell holders are machined aluminum blocks that hold the demountable liquid cells in place using screws.

[0052] Further, inclusion of ND-filters and diffusers for balancing, and CaF2 focusing optics are optionally used for increasing incident power at the detector elements, in order to increase signal-to-noise specific for the detection of molecular vibrations in the spectral region (2000-2300 cm'1). Note that the full power of the beam was used in conjunction with the 250-pm pathlength aqueous samples, with only attenuation of one beam relative to the other to equalize incident power on the detector. While the laser beam waist is < 2.5 mm, the beam is not focused prior to transmission through the sample cells, meaning the beam width is similar to but slightly less than that of the FTIR (maximum 6-mm aperture). The focusing optics, including of CaF2 plano-convex lenses (e.g., 25.4 mm diameter, 100 mm focal length, Newport Optics Cat. No. CAPX13) may be used to focus both the sample and reference beams after sample transmission into a custom-built thermoelectrically cooled balanced HgCdTe (MCT) detector (as may be supplied by VIGO Photonics, PVI-4TE series), and to detect the transmitted IR signal from both beams. Further modification of the setup by using additional focusing optics may be possible for additional optimization. Yet additional performance improvements may also arise from the specific detector modules used, as there is currently no single, unified head-to-head performance comparison published — either by manufacturers or third parties — between the VIGO PVI-4TE series and other commonly used liquid nitrogen-cooled MCT detectors, such as the Teledyne Judson models. However, detailed technical datasheets and catalogs are available for both to provide comparative performance metrics for HgCdTe detectors. These resources allow for an indirect but robust comparison of key parameters such as detectivity, responsivity, noise, and spectral range.

[0053] The EC-QCL type light source is modified to enable performance of certain analyses as described herein according to the present disclosure. The modifications include a different laser module to account for a mid-IR spectral window (2000-2300 cm'1) in which the instrument is equipped to make high-sensitivity measurements with uniquely long pathlengths (up to 250 pm when making measurements in aqueous media). The laser beam enters a box (e.g., housing) as in FIG. 1C, containing all optics and being purged with dry air (e.g., Altec Air CO2-PG28 Purge Gas Generator). The beam is split into a set of two or more beams. For example, the beam may be split as illustrated in FIG. 1C into a pair of beams by a 50:50 ThorLabs CaF2 Plate Beamsplitter (coating: 2-8 pm, Cat. No. BSW510). In thisSTFD.469PCT (S24-491) 16 example, mid-lR-enhanced gold mirrors (ThorLabs Cat. No. PF 10-03-M02) are used to define the optical path. CaF2 neutral density filters (Andover Optics Cat. No. 150FNIR-25, 100FNIR-25, and 030FNIR-12.5) are used when applicable to equalize the intensity of the two beams or bring the overall power intensity into the dynamic range of the instrument, with CaF2 diffusers (Edmund Optics Cat. No. 19-734) also implemented to scramble polarization prior to detection.

[0054] While FIG. 1C illustrates the laser source and the beamsplitter as being part of a singular system, in different examples there are alternate versions of such a system according to the present disclosure. In one example, the laser source is separate equipment from the remaining aspects (the optics and related aspects downstream from the laser source), wherein in use the laser source is integrated as depicted in FIG. 1C for operation. In another example, the laser source and the beamsplitter are separate equipment from the remaining aspects (the optics and related aspects downstream from the beamsplitter), wherein in use the laser source is integrated as depicted in FIG. 1C for operation. In each such example, it is appreciated that the various optics and much of the related downstream equipment, while preferred for many applications, is optional and used as may be helpful to enhance performance. As such, in operation such an apparatus in a more basic form includes: a QCL to direct a laser beam carrying light in a wavelength range that overlaps a mid-IR range; optics, responsive to the laser beam, to process the light in a set of related beams along respective paths, including a reference path and a sample path, respectively towards a reference target and a sample target that includes living cells; and a vibrational spectroscopy detector to detect small molecules binding to a target protein in the living cells by collecting incident light, relative to the reference target and from the sample target, carried along respective incident paths and by discerning (e.g., measuring) nitrile frequency shifts in the sample target.

[0055] In each of the above example, the beam as split into separate beams (for example, into two beams) pass through sample and reference liquid IR cells (Pike Technologies Demountable liquid cells) using CaF2 windows and 250 pm Teflon spacers. Cell holders are machined aluminum blocks that hold the demountable liquid cells in place using screws. Note that the inclusion of the ND-filters and diffusers for balancing, and CaF2 focusing optics to increase incident power at the detector elements, are components put into place relative to previously-described instrumentation to increase signal-to-noise specific for the detection of molecular vibrations in the spectral region (2000-2300 cm'1). See, e.g., Akhgar, C. K., Ramer, G., Zbik, M., Trajnerowicz, A., Pawluczyk, J., Schwaighofer, A., & Lendl, B. (2020), The Next Generation ofIR Spectroscopy: EC-QCL-Based Mid-IR Transmission SpectroscopySTFD.469PCT (S24-491) 17 of Proteins with Balanced Detection, Anal. Chem., 92(14), 9901-9907. Also in such experimental efforts, focusing optics, consisting of CaF2 plano-convex lenses (25.4 mm diameter, 100 mm focal length, Newport Optics Cat. No. CAPX13), are used to focus both the sample and reference beams into a custom-built thermoelectrically cooled balanced HgCdTe (MCT) detector, supplied by VIGO Photonics, to detect the transmitted IR signal from both beams.

[0056] Consistent with the above example scheme using an MCT-type detection system, such a VIGO-provided detector can be implemented using three output channels; namely , signal (S), reference (R). and balanced (B = S - R) channels. Channel readout voltages from S, R and B are digitized by a high-speed lock-in amplifier (e.g., available at Stanford Research Systems SR865A 4 MHz dual-phase lock-in amplifier) whose reference signal is dictated by the laser pulse rate (1 MHz at 30% duty cycle). The lock-in amplifier sends capture buffers of S, R, and B covering laser scans from 2320-2000 cm'1(as one example of a range in which such laser scanning may be carried out) to a host computer for analysis with an in-house MATLAB script that controls both data capture and laser operation. Digitized signal packets are processed with phase correction from the lock-in detector, Savitsky -Golay filtering, removal of anomalous spectra using similarity indices, scan averaging, and Fourier filtering (Alcaraz et al., Anal. Chem. 2015, 87, 6980-6987). Additionally, to increase sensitivity and make use of the full flux of photons from the laser (and full dynamic range of the detector), the nonlinear detector response of voltage was calibrated in terms of laser power externally using a ThorLabs power meter (PM100D with S302C Thermal Power Sensor for 0. 19 - 25 pm), generating a cubic fit function used to convert measured detector voltage to beam incident power. Such aspects are additional (e.g., beyond that provided by the Akhgar and Lendl type instrumentation) and aimed to provide maximum possible photon flux and pathlength usage within the 2320 cm'1to 2000 cm'1spectral region. Moreover, to reduce heating effects, a 2 sec. dark time was provided between scans, as an alteration from the Lendl setup Peltier cooling apparatus. Absorbance of a sample analyte relative to a reference blank is determined using a balanced detection scheme, which involves multiple measurements across different channels. Both a reference channel measurement of power versus laser frequency, and a balanced channel measurement of power difference versus laser frequency are measured when reference blanks are loaded into both liquid cells. These are followed by loading the sample liquid cell with the analyte sample of interest and making a second balanced channel measurement of power difference versus frequency. This method ensures that background noise from the laser source is significantly reduced and that theSTFD.469PCT (S24-491) 18 measured absorbance is consistent with the Beer-Lambert law, which relates absorbance to the concentration of the analyte and the pathlength of the sample.

[0057] Utilizing benzonitrile as a model nitrile probe in water in the pM range, the experimental efforts compared the limit of detection for the spectrometer against a conventional Br ker Vertex 70 FTIR spectrometer (see. e.g., FIGs. 8A through FIG. 12). The experimental efforts also documented (discussed infra) the increased sensitivity of the QCL for other useful vibrational probes including alkyne (C=C). azides (N=N+=N), and carbondeuterium (C-D) stretches (FIGs. 13-15). A notable challenge for IR spectra of aqueous samples in this spectral window is restricted transmission due to water's combination mode centered at approximately 2100 cm1Consequently, this absorption imposes a limit on the maximum spacer length that can be used in the Broker Vertex 70, which is restricted to ~50 pm and leads to a limit of detection (LOD) for benzonitrile of ~80 pM for the nitrile stretch centered at -2235 cm . In contrast and according to the present disclosure, the balanced QCL-based detection (limited by the same phenomena) using 250 pm spacers significantly enhances the experimental LOD, achieving a value of 18 pM. This QCL-based method therefore provides an approximately five-fold increase in sensitivity for IR measurements within the nitrile-stretch region, outperforming the conventional FTIR spectrometer and enabling detection of low concentrations of nitriles in aqueous environments, such as proteins or nitrile-labeled molecules directly within cells.

[0058] This outperformance is apparent from a comparison of FIG. 1A and FIG. IB. For example, FIG. 1 A and FIG. IB are useful to show a comparison of detection limits for benzonitrile using, as in FIG. 1A, a conventional Broker Vertex-70 FTIR spectrometer and, as in FIG. IB, a QCL-based spectrometer. The conventional FTIR approach, limited by aqueous sample transmission with 50 pm spacers, achieves a limit of detection (LOD) of 80 pMm, whereas the QCL-based approach improves the LOD to 18 pM. This is approximately a ~5-fold enhancement in sensitivity for IR measurements in the nitrile stretch region compared to high-end FTIR systems, largely due to the use of a 250 pm sample pathlength.

[0059] Such experimental efforts also included use of nitrile-labeled proteins in live bacterial cells. These efforts were used to evaluate the spectrometer's ability to detect nitrile stretches within E. coli cells using amber suppression machinery to introduce orthocyanophenylalanine (oCNF) residues into target proteins of interest. A primary focus was on apo-PYP as a model system for the equivalent of covalent drug binding. Site-specific labeling was used to produce four opo-PYP variants: F28oCNF, F62oCNF, F92oCNF, and F96oCNF. As an additional test of visualizing nitriles in live cells a nitrile was alsoSTFD.469PCT (S24-491) 19 incorporated into superfolder green fluorescent protein (sfGFP), where the chromophoreforming tyrosine Y66 was replaced with oCNF.

[0060] The variable proximities of the four nitrile probes on PYP to the pCA binding site allowed for observing locally specific information on subtle structure perturbations when activated pCA binds both in solution and in living cells. Each nitrile-containing PYP variant was previously characterized in the holo-form (covalently attached to pCA) via X-ray crystallography to high-resolution (<1.2 A), where the nitriles adopt a single, well-defined orientation at each site. F28oCNF (PDB: 7SPX) interacts with one hydrogen bond donor, a water molecule (3.2 A heavy atom distance) and is located ~15 A (ring-to-ring distance) from pCA, while F92oCNF (PDB: 7SPV) engages with two donors (2.9 and 3.2 A for T90 and a water, respectively) and is positioned ~12 A from the pCA chromophore. In contrast, F62oCNF (PDB: 7SPW) and F96oCNF (PDB: 7SJJ) are situated in hydrophobic environments, with only carbon atoms within 3.5 A of the nitrile nitrogen, and they are located ~10 A and ~5 A from the pCA chromophore, respectively. As part of this study, it was also determined for the first time a high-resolution X-ray crystal structure of the apo form of PYP (without the nitrile labels; PDB: 908V) and observed minimal deviation from the holo form, both with and without a nitrile incorporated (see FIG. 2). The absence of the chromophore and nitrile labels did not result in any major conformational changes in the protein structure, except for a lack of density in the disordered loop region, which is located more than 20 A from the chromophore pocket.

[0061] FIG. 2 is a schematic showing a strategy', consistent with the above efforts according to the present disclosure, in connection with covalent drug binding using photoactive yellow protein (PYP) and nitrile vibrational spectroscopy. The strategy' is directed to oCNF incorporation via amber suppression, which allo vs site-specific placement of nitrile reporters at F28, F62, F92, and F96. See, e.g., Holo (Green, PDB: 1NWZ) and apo (Grey, PDB: 908V) PYP crystal structures superimposed with the modified Phe residues highlighted.

[0062] Utilizing these variants and the sensitive QCL spectrometer, the nitrile absorption spectra yvere measured in a concentrated suspension of bacterial cells (ODeoo = 0.3 yvhen diluted lOOOx) expressing individually nitrile-modified Phe variants of apo-PYP. Cells were washed three times with minimal media to remove excess unincorporated oCNF. A culture expressing unmodified apo-PYP (no nitrile) was used as a blank for balanced detection. The nitrile frequencies were clearly apparent: for F28oCNF at 2230.8 cm’1, for F62oCNF at 2227.5 cm , for F92oCNF at 2226.4 cm'1, and for F96oCNF at 2227.6 cm’1(FIGs. 3A andSTFD.469PCT (S24-491) 203B). While similar measurements were attempted using the Bruker 70 FTIR spectrometer with 50 pm spacers, the low signal-to-noise ratio hindered reliable detection of these peaks. For Y660CNF sfGFP, the nitrile stretch was also clearly visible at 2225.8 cm1in cells using the QCL spectrometer.

[0063] FIGs. 3A and 3B show incorporation of chromophore p-coumaric acid (pCA) into PYP by nucleophilic attack at C69. As in FIG. 3A, the form of pCA added to the cells is an activated form with a reactive, electrophilic “warhead” that derives from a water-free reaction of pCA with carbonyldiimidazole. FIG. 3B shows the nitrile vibrational spectra collected in bacteria using the QCL for the F28oCNF. F62oCNF, F92oCNF, and F96oCNF variants and show variable nitrile vibrational frequency shifts in response to pCA incorporation, with shifts of +0.3, +0.6, +14.9, and +6.3 cm'1, respectively (analyzed by peak fitting). While these shifts were realized in experimental efforts (proof-of-concept experiments) involving nitriles, in other example embodiments actual frequency shifts (significantly larger than +15 cm'1) vary depending on the system under investigation.

[0064] In each of these cases, the in cellulo nitrile frequencies were distinct from that of free oCNF in aqueous media (e.g., 2232.3 cm in water). Furthermore, when cells expressing nitrile-containing proteins were lysed and purified by affinity column chromatography, the nitrile stretching frequencies of the isolated proteins matched those observed in cells. However, when o / w-PYP nitrile frequencies were compared to previous measurements conducted on the purified / zo / o-PYP (with pCA covalently attached), significant differences in peak frequency were noted for variants F92oCNF and F96oCNF. In contrast, the frequencies for F28oCNF and F62oCNF were similar regardless of chromophore incorporation. These findings indicated that some of the exemplary nitrile environments in the protein depend on pCA binding, prompting further investigations described below.

[0065] Nitrile frequency shifts were observed in such experimentation on pCA-binding from the QCL-based IR measurement. In this context, the PYP-expressing bacteria were incubated with a ~ 100-fold molar excess of pCA chemically activated with an acyl imidazolide electrophilic warhead. The nitrile stretch measured by the QCL in bacteria expressing F28oCNF or F62oCNF PYP exhibited negligible shifts after activated pCA incubation, while those expressing F92oCNF and F96oCNF displayed large blue shifts of +14.9 cm and + 6.3 cm '. respectively, when compared to the apo samples. These nitrile frequencies align closely with those of the previously reported, isolated holo F92oCNF and F96oCNF proteins. Subsequent purification of these activated pCA-incubated cells yielded a characteristic yellow protein with a UV-Vis absorbance maximum at 445 nm, which alongSTFD.469PCT (S24-491) 21 with mass spectrometry confirmed chromophore incorporation through the cell wall and membrane and into the o / vi-protein. Additionally, the nitrile frequencies for these purified proteins closely matched the frequencies of both the previously reported purified Ao / o-PYP and the QCL-measured in cellulo nitrile frequencies of / ro / o-PYP.

[0066] As a control, a C69A substitution was introduced in the F96oCNF variant (F96oCNF was selected due its ~6 cm blue shift between the holo and apo forms). The C69A mutation thereby disrupted the covalent linkage site for pCA in the F96oCNF-PYP variant. Even after incubation with activated pCA, the nitrile frequency in this variant remained similar to that of the a o-F96oCNF. This confirms that the C69 thiol group is indeed responsible for pCA binding including in the cellular environment, and the covalent linkage of pCA in the binding site accounts for the observed frequency shifts.

[0067] With regards to the origin of the large frequency shifts on pCA incorporation for F92oCNF and F96oCNF PYP. nitrile vibrational frequency shifts are affected by intermolecular interactions — electrostatic stabilization typically causes red shifts, while hydrogen bonding can induce blue shifts with respect to an unperturbed nitrile stretch. Particularly w hen the concentration of the vibrational probe cannot be obtained with high accuracy, interpreting a spectroscopically observed frequency shift requires structural knowledge of the local electric fields and hydrogen-bonding populations and, thus, the consequences of drug binding to opo-PYP by way of structural analysis and MD simulations.

[0068] The influences on vibrational probe frequency shifts in PYP were investigated using molecular dynamics simulations involving nitrile as an exemplary' type of vibrational probe. As shown (by the common inventors) in separate work, fixed-charge MD simulations based on force-fields such as AMBER often cannot accurately recapitulate the electrostatic environment around nitrile probes in proteins, thus for these analyses higher-level simulations were carried out using the AMOEBA09 force field. Four replicate AMOEBA simulations of 25 ns each were performed using the F92oCNF (7SPV) and F96oCNF (7SJJ) PYP structures with chromophores removed as starting points, again showing no global conformational changes. See Weaver, J. B., Kozuch, J., Kirsh, J. M., & Boxer, S. G. (2022) Nitrile Infrared Intensities Characterize Electric Fields and Hydrogen Bonding in Protic, Aprotic, and Protein Environments. J. Am. Chem. Soc. 144(17), 7562-7567. These simulations revealed that for o o-F92oCNF. the average electric field projected on the nitrile bond was significantly reduced to -44 MV / cm, compared to the -67 MV / cm reported by Kirsh et al. for the holo- form. A ?o-F96oCNF showed only a slightly smaller average field, -23 MV / cm (e.g., compared to -26 MV / cm reported for the / zo / o-form). A more stabilizing field on the nitrilesSTFD.469PCT (S24-491) 22 in the holo-form would typically coincide with a red-shift upon pCA binding. However, the presence of hydrogen bonds as is possible for both nitriles obfuscates a purely electrostatic interpretation of the observed frequency' differences, in contrast to the transition dipole moment approach used by Weaver et al. See, e.g., Weaver, J. B., Kozuch, J., Kirsh, J. M., & Boxer, S. G. (2022). Nitrile Infrared Intensities Characterize Electric Fields and Hydrogen Bonding in Protic, Aprotic, and Protein Environments. J. Am. Chem. Soc., 144(17), 7562- 7567; and Kirsh, J. M., Weaver, J. B., Boxer, S. G., & Kozuch, J. (2024). Critical Evaluation of Polarizable and Nonpolarizable Force Fields for Proteins Using Experimentally Derived Nitrile Electric Fields. J. Am. Chem. Soc., 146(1Q), 6983-6991. https: / / doi.org / 10.1021 / jacs.3cl4775.

[0069] For the simulations involving F92oCNF, two distinct electric field states were observed being projected on the nitrile bond: a high-field state typically corresponding to a hydrogen-bonded state (primarily with T90 side chain), and a significantly lower-field state corresponding to a non-hydrogen-bonded configuration. Distinct hydrogen-bonding populations agree with previous low-temperature FTIR experiments of the holo form that slow the rate of exchange between the two populations. They also align with previous MD observations of a hydrogen-bonding and a non-hydrogen-bonding state in F92oCNF when the chromophore was present, where much of the simulation trajectory indicated a hydrogen- bonded state. The current simulations in the absence of the chromophore significantly favor decreased hydrogen bonding of the nitrile to T90 in opo-F92oCNF, consistent with the IR measurements showing a blue shift upon chromophore incorporation. Additionally, AMOEBA simulations reveal a small increase in hydrogen-bonded population (nitrile-to- solvent) for the r / / w-F960cnf variant upon pCA binding. It may be more difficult to reliably interpret simulation results in the vicinity of the chromophore pocket given the more dynamic solvation environment, particularly when pCA is absent and the pocket is solvent-exposed (FIG. 2). However, the small but definite change in solvent hydrogen bonding to the nitrile in F96oCNF presents a reasonable explanation for the smaller blue shift undergone by the nitrile upon pCA binding.

[0070] As the IR frequency of a unique vibrational mode according to the present disclosure, such as a nitrile, is highly sensitive to changes in its local environment, according to aspects of the present disclosure, the frequency shifts upon drug binding can be an important tool for studying drug-protein interactions. The limitations due to low sensitivity in cellulo are addressed by such aspects of the present disclosure by use of a QCL-based infrared spectrometer capable of sensitively detecting site-specific vibrational shifts inSTFD.469PCT (S24-491) 23 proteins bearing embedded vibrational probes, thereby highlighting its utility for monitoring protein-drug interactions even within live-cell environments. Compared to advanced Raman techniques such as stimulated Raman spectroscopy (SRS), which can also be applied in cellulo. linear QCL spectroscopy leveraging long path lengths can still provide great sensitivity for bulk measurements. While methods like SRS offer high-resolution imaging and enhanced sensitivity when using vibrational probes conjugated to electronic chromophores (e.g., epr-SRS, Bon-FIRE), the chemical interpretability of spectral intensities is not straightforward, and sensitivity is typically lower when using ordinary vibrational probes due to their inherently lower molecular cross-sections, (e.g., 105molecules / 0. 1 fL ~ 1 mM for C=C bonds).

[0071] By using a quantum cascade laser (QCL)-based balanced detection specifically optimized for the nitrile stretching region (2000-2300 cma spectrally transparent window with minimal interference from endogenous biomolecular absorption, high-sensitivity detection is enabled in complex biological environments. Based on calculated protein expression levels, these experimental efforts in connection with the present disclosure coincide with nitrile concentration within each cell to be estimated at approximately 140 pM, which is well above the determined 18 pM LOD for aromatic nitriles using the QCL and at a considerably higher SNR (signal-to-noise ratio) when compared to a traditional FTIR instrument. This high sensitivity also renders the spectrometer well-suited for detecting other non-perturbative vibrational probes in this range such as N=N+=N", or even the weaker C-D or C=C stretches, which typically exhibit oscillator strengths up to tenfold lower than those of nitriles (FIGs. 13-14 as discussed infra). For example, nitrile or azide-labeled GPCRs are an excellent candidate for in cellulo 1R spectroscopy using the strategy described here and a QCL spectrometer for preservation of a native membrane environment.

[0072] Utilizing this spectrometry approach of the present disclosure, a photoactive yellow protein (PYP) and its covalently attached chromophore p-Coumaric acid (pCA) were employed as a representative framework for many cysteine-based covalent inhibitors. For example, notable cysteine-reactive covalent drugs such as ibrutinib and acalabrutinib have transformed cancer therapy by selectively inhibiting Bruton's tyrosine kinase (BTK) through a covalent mechanism, with minimal off-target effects on other kinases. The PYP-pCA system provides a structurally and mechanistically analogous framework for investigating the molecular underpinnings of such selective covalent binding. While these experimental efforts demonstrates covalent binding, the nitrile group’s sensitivity’ to its local environment enables detection of a broad range of molecular interactions, including non-covalent effectsSTFD.469PCT (S24-491) 24 and drug molecules reliant on non-covalent binding. This versatility underscores a key advantage of this vibrational probe system over traditional fluorescence-based assays. Unlike fluorescence approaches that often require the attachment of bulky dye molecules (comparable in size to the drug itself) vibrational probes provide efficient, minimally invasive access to intrinsic molecular information, ranging from electrostatics to protein motion and conformational dynamics.

[0073] Such detailed structural insight is largely unique to spectroscopic techniques. In contrast, proteome-wide mass spectrometry assays, although powerful, face limitations. Highly specific biophysical studies of noncovalent interactions such as electrostatics or hydrogen bonding are not currently possible with mass spectrometry. Moreover, in this particular case, proteomic analysis aimed at identifying off-target binding interactions of pCA proved of limited utility. This was likely due to the standard reducing conditions for peptide linearization (involving iodoacetamide and dithiothreitol (DTT)), which can disrupt sensitive covalent linkages such as the thioester bond between PYP and pCA. By comparison, the IR assay permits inclusion of those interactions, which are typically challenging to mass spectrometry — both non-covalent interactions plus covalent bonds sometimes susceptible to cleavage by DTT or tris (2-carboxyethyl)phosphine hydrochloride (TCEP) reducing agents employed in a standard proteomics digestion. This highlights the unique advantages of IR spectroscopy, according to exemplary aspects of the present disclosure, for probing the highly localized interactions that may govern the chemical reactivity and efficacy of covalent drugs.

[0007] Finally, an in-depth analysis was performed regarding the unique spectral features of the nitriles (as a representative example vibrational probe) observed using the QCL-based spectrometer and the PYP-pCA model system, as disclosed in vanous examples herein. Specifically, distinct hypsochromic shifts were detected in the nitrile stretching frequencies of variants F92oCNF and F96oCNF upon incorporation of the chromophore. To elucidate the origin of these shifts, molecular dynamics simulations were conducted using both fixed-charge and polarizable force fields, which revealed that the frequency changes stem from localized structural and electrostatic rearrangements within the protein environment upon pCA incorporation. The ability of AMOEBA simulations to capture these subtle electrostatic rearrangements in the nitrile microenvironment further validates the use of this polarizable force field for high-level molecular dynamics studies. These findings not only benchmark the simulations but also provide insights of significant interest to force-field developers. These subtle perturbations highlight how the incorporation of a drug-like molecule such as pCA can induce site-specific changes in the protein’s electrostaticSTFD.469PCT (S24-491) 25 landscape — even far from the binding site — underscoring the sensitivity of vibrational probes to chemically relevant microenvironments. This can become especially relevant when the local electrostatic environment of a binding site has been shown to be critical to the reactivity of covalent binding, which is an aspect of the present disclosure not traditionally recognized in connection with traditional electrostatic complementarity approaches used in drug design.

[0075] Structurally, the crystallographic data seem to suggest a slight reorientation of F96 toward the empty chromophore pocket, as in FIG. 2, and slight increase in hydrogen bonding population. In the case of F92oCNF, chromophore incorporation appears to increase the likelihood to form a hydrogen bond with T90, whereas in the apo state, it remains largely isolated from such interactions. These results are consistent with an “induced fit” model of protein rigidification upon ligand binding, corresponding to an increase in internal hydrogen bonding netw orks near the ligand binding site, w hich may be generalizable to a wide variety of drug targets.

[0076] Discussion now turns to more specific characterization of the above-disclosed examples and parameters used in such experimental efforts. By employing a different laser module to account for a somewhat unique mid-IR spectral window (2000-2300 cm'1) as shown in FIG. 4. the instrumentation is equipped to make high-sensitivity measurements with long pathlengths (e.g., up to 250 pm) when making measurements in aqueous media, wherein a w eak but discernible IR absorption band is centered at roughly 2100 cm '. which might be attributed to the combination of the HOH bending mode (-1630 cm ') with a librational mode (-400-700 cm ') As shown at the peak of the plotted line in FIG. 4, the resulting combination band appears near 2100-2130 cm1In connection with the graph of FIG. 4. spectral power distribution of the MlRcat-QT-2000 with M2048-P tunable module, measured with a ThorLabs power meter at the exit aperture at 5% duty cycle and full current (780 mA). In certain of the experiment efforts, a 30% duty cycle was used.

[0077] This unique capability of using longer pathlengths due to the high flux of this QCL module makes it ideal for analytes bearing oscillators such as C=N. C-D. N=N+=N', and C=C stretches. This module provides pulsed output up to 30% duty' cycle and high output powder for high signal-to-noise ratios [ SNR: >1 W (peak) and >0.5 W (average) ]. The laser control was implemented using custom scripts and MATLAB libraries provided by Daylight Solutions, which were modified for specific use in connection with such experimental efforts according to the present disclosure (https: / / github.com / sdefried / QCL_spectrometer_scripts). For optical alignment, a collinear HeNe laser (632.8 nm) was used, which is integrated with the module and shares the same exit aperture as the infrared beam.STFD.469PCT (S24-491) 26

[0078] Data acquisition, dynamic range and power calibrations are largely realized using the detector’s three output channels, namely the signal (S), reference (R), and balanced (B = S R) channels. The channel readout voltages from S, R and B are digitized by a high-speed lock-in amplifier (e.g., Stanford Research Systems SR865A 4 MHz dual -phase lock-in amplifier), whose reference signal is dictated by the laser pulse rate (1 MHz at 30% duty cycle). The lock-in amplifier sends capture buffers of S, R, and B covering laser scans from 2360-1970 cm1to a host computer for analysis with an in-house MATLAB script that controls both data capture and laser operation. Digitized signal packets (e.g., as in FIG. 5) are processed with phase correction from the lock-in detector, Savitsky-Golay filtering, removal of anomalous spectra using similarity indices, scan averaging, and Fourier filtering.

[0079] FIG. 5 illustrates the MCT output being processed from the QCL Module M2048-P scan through water-filled cell at 30% duty cycle. The digitally processed reference MCT voltage output is in a range from 1970-2360 cm1scan of the M2048-P tunable module at 30% duty cycle through a 250-pm sample cell filled with water. Although the laser’s peak power is near 2100 cm1as in FIG. 4, the water’s combination mode at the same frequency significantly attenuates the beam in this frequency region. The output voltage was consistently < 300 mV for all spectral measurements reported in connection with these experimental examples.

[0080] Additionally, to increase sensitivity and make use of the full flux of photons from the laser (and extend the dynamic range of the detector), the nonlinear detector voltage response was externally calibrated in terms of laser power using a ThorLabs power meter (PM100D with S302C Thermal Power Sensor for 0.19 - 25 pm). During calibration, samples are absent in the beam path; therefore, up to three Andover Optics neutral density filters (optical densities 0.3, 1.0, and 1.5) were employed to attenuate the beam and maintain the MCT detector response within the linear dynamic range. At three different wavelengths, laser power was controlled by varying the current supplied to the QCL chip between 400 and 780 mA, while the beam intensity at both the signal and reference MCT was measured using both the MCT detector itself plus the thermal power meter. The detector voltage response to power demonstrates saturating behavior around 470 mV for both MCTs, which was well-captured using a rational instrument response function of the form y — ax / ( — bx) for y = incident power at MCT and x = MCT output voltage. The best-fit function to the data was used to convert measured detector voltage to beam incident power. This calibration step was implemented to ensure maximum photon flux and optimal pathlength usage within the spectral range as disclosed in accordance with the present disclosure.STFD.469PCT (S24-491) 27

[0081] FIGs. 6A and 6B show the calibration of the MCT detector output to incident power using nonlinear curve fitting. Calibration against a thermal power meter is used to convert MCT detector output voltages to incident power on the detector, which in turn is used to extend the dynamic range of the instrument given the nonlinear response of the detector at high photon incidence. Detector output voltages were consistently kept below 300 mV when transmitting through a 250-pm liquid sample cell filled with water, thus avoiding the saturation regime of the detector response, and increasing the accuracy of the transmission measurements. Fitting to the nonlinear response function was accomplished using the MATLAB curve fitting toolbox with a nonlinear least squares regression, fitting to all the data points (three separate wavelengths) at once. Errors reported are 95% confidence intervals of the parameters.

[0082] With respect to frequency calibration and data capture, the frequency axes of the spectra were initially determined using the programmed laser start frequency (2360 cm ' ), at which a trigger was sent to the SR865 A lock-in amplifier to begin a 32-kb data capture buffer, along with the SR865A capture rate (9765.6 Hz) and laser scan rate (2000 cm Vs). However, due to slight offsets between these frequencies calculated using time and the true laser frequencies, the frequency axis was recalibrated using a linear function. To do this, laser wavelength TTL triggers were used every 10 cm ' to subtract off from the MCT output as a separate channel in the lock-in amplifier, thereby generating defined features in the voltage output capture buffer every 10 cm Correlating these with the original frequencies, a linear calibration function was generated to convert the time-based frequencies to corrected frequencies used as the axis for all absorption spectra reported herein. By matching frequency signatures such benzonitnle to known literature values, the calibration was verified as being accurate. This frequency calibration, as in FIG. 7, involved recalibration of the original laser scan frequencies determined by time from initial scan trigger, based on periodic wavelength triggers from the QCL.

[0083] Absorbance measurements were taken using the balanced detection scheme as described herein, for example, in connection with FIG. 1C. Absorbance of a sample relative to a reference blank was determined using a balanced detection scheme, which involves multiple measurements across different channels. Both a reference channel measurement of power versus laser frequency, and a balanced channel measurement of power difference versus laser frequency were obtained when reference blanks were loaded into both liquid cells. These are followed by loading the analyte of interest into the sample liquid cell andSTFD.469PCT (S24-491) 28 making a second balanced channel measurement of power difference versus frequency. Absorbance (A) is then calculated using the equation — log [^signal (^Ref + ^Bal-Blank) / ^Signal ( ^Ref + ^Bal-Sample)] , where pRef is the reference channel MCT output voltage and Eiiai represents the balanced channel (subtracted) MCT output voltage, whether for both liquid cells filled with blanks (kBai-Biank) or the sample cell filled with sample against the reference filled with blank (FBai- sampie). The function Psignai(k) takes an MCT detector voltage as input and gives the beam power incident on the signal detector, according to the above calibration, Psignal= aV / (1 — bV). This method is useful to ensure that background noise is eliminated using balanced detection and that the measured absorbance is consistent with the Beer-Lambert law.

[0084] Following the above process, for determination of experimental limit of detection, the capabilities of the QCL spectrometer were demonstrated for sensitively measuring vibrational spectra of benzonitrile in water with a 250 pm pathlength. Such aspects are shown in connection with FIG. 1A and IB. Spectra of benzonitrile at higher concentrations, measured on both the FTIR and QCL spectrometer, are shown with FIGs. 8A- 8D, along with the concentration / absorbance correlation.

[0085] More particularly, FIGs. 8A-8D show benzonitrile spectra and concentrationabsorbance correlations from FTIR and QCL measurements. Spectra of benzonitrile at concentrations of 100 pM - 10 mM are shown using 16 scans each on either the Bruker Vertex 70 FTIR as in FIG. 8A, or the QCL spectrometer as in FIG. 8B. Correlation plots between concentration and absorbance are also shown for both the FTIR as in FIG. 8C and the QCL spectrometer as in FIG. 8D, with linear behavior according to the Beer-Lambert law, particularly in the case of the QCL spectrometer — which demonstrates a stronger linear correlation according to R2. Errors reported are standard errors. Note that the data for the FTIR carries a higher linear regression error compared to the QCL although the same samples were used for both. This may be attributed to the relative difficulty in distinguishing peaks from solvent background when the absorption intensities are lower in the case of the FTIR.

[0086] The extinction coefficient of benzonitrile can be determined from both instruments. The exact pathlength of both the FTIR and QCL spectrometer sample cells were determined using periodic spacings of etalon fringes (as reported by Weaver, J. B., Kozuch, J., Kirsh, J. M., & Boxer, S. G. (2022), Nitrile Infrared Intensities Characterize Electric Fields and Hydrogen Bonding in Protic, Aprotic, and Protein Environments. J. Am. Chem. Soc., 144(17), 7562-7567). The true pathlength differed slightly from the spacer widths (relating toSTFD.469PCT (S24-491) 29 how much the spacers were compressed), with the FTIR sample cell having pathlength 55.9 pm and the QCL sample cell having pathlength 259.4 pm. Using these values and the slopes reported below, extinction coefficients the benzonitrile nitrile stretch were calculated as s = 162 (±26) M1cm1for the FTIR and s = 194 (±10) M1cm1for the QCL (errors as 95% confidence intervals). Both values are within error of the e = 186 M1cm1value reported previously, while meanwhile the peak maximum of 2235.5 cm1(the more important metric for the current study) similarly agrees well, wi thin 1 cm '. of previous values obtained in connection with such experimental efforts for the frequency of benzonitrile in water.(0087] Using the calibrated instrument response to benzonitrile and the RMS noise from a blank spectrum (water versus water), the theoretical limit of detection for benzonitrile was determined for both the FTIR and QCL (FIG. 9). The RMS noise (crRMS) was measured over the range of 2260-2160 cm ', a relatively stable 100 cm1window with high photon flux even through 250 pm aqueous samples, which also includes benzonitrile’s peak of interest (FIG. 5). The theoretical limit of detection was defined by the concentration of benzonitrile required to achieve a signal at least three times the RMS noise over this region, that is, [LOD] = 3CTRMS / (E / ), for extinction coefficient and pathlength / , where s i can be obtained from the slope of the instrument response function from FIG. 8. As an additional point of comparison, also included is the scenario of detection based on the calibrated signal channel alone, without the use of balanced detection between two MCTs.

[0088] FIG. 9 shows the baseline spectra for estimating theoretical LOD. The baseline IR spectra was recorded by the QCL spectrometer using 16 scans, compared to the FTIR. In each such case, 16 scans were used for ease of comparison, and the same procedure for digital processing of the QCL data was applied as described above. A manual baseline correction was employed to all spectra by fitting to broad, low-frequency polynomials to center the higher-frequency noise at zero across the spectral range.

[0089] The theoretical LODs were determined from the baseline spectra (water minus water) and sensitivities match well with experimental spectra of benzonitrile at low concentrations. This is shown in FIGs. 8A-8D and FIGs. 1A-1C, where the nitrile stretch first becomes clearly visible to the FTIR between 50 and 100 pM, and to the QCL between 10 pM and 25 pM. The longer pathlength afforded by the higher photon flux is a primary advantage. As demonstrated above from the single channel results, the QCL light source inherently has more fluctuations than the FTIR globar source, an aspect observed by Lendl and coworkers in previous research, although use of a MIRcat laser equipped with ZeroPoint beam pointing control (according to the present disclosure) provides some noise reduction compared toSTFD.469PCT (S24-491) 30Lendl and coworkers’ use of the Hedgehog laser. The ZeroPoint technology in the MIRcat provides active stabilization compared to the passive stabilization in the Hedgehog module and has lower thermal drift compared to the Hedgehog module. In the absence of balanced detection, it was previously reported (e.g., Lendl Lab) nearly a 3-fold higher RMS noise for their Hedgehog (2ndGen) QCL source compared to FTIR systems under comparable experimental conditions, with measured values of 6.2 / 105(QCL single channel measurement, 53-s integration, 31-pm path length) versus 2.3 / 105(FTIR, 45-s integration, 8-pm path length) in the 1700-1500 cm1spectral range. These results align closely with the noise levels measured in connection with these experimental efforts, despite differences in experimental parameters such as the higher number of scans (>100), shorter path lengths (8- 31 pm) and a different spectral window employed in their study. The similarity in noise performance underscores the challenges inherent to QCL-based systems, where intensityfluctuations and thermal drift can dominate measurement uncertainty.

[0090] Despite the higher noise inherent to the QCL source, balanced detection decreases RMS noise into the realm of the FTIR. For example, this is applicable in connection with such experimental efforts where the 5x longer pathlength of the QCL affords an approximately 5-fold advantage in terms of LOD. In each case described herein, 16 scans were used on the QCL spectrometer for comparison to the FTIR and measurement of the nitrile-embedded PYP within bacterial cells (using nitrile as an example vibrational probe). This number was chosen as the optimal balance between noise-reduction and efficiency when making measurements, as the optimal number of scans for the sake of achieving sufficient sensitivity while maintaining efficiency and minimal perturbation of the later cell-based samples. FIG. 10 shows the RMS noise for a blank (balanced) absorption spectrum through 250-pm pathlength water generally decreased with increasing number of scans based on square root law of noise averaging.

[0091] In connection with such experimental efforts, while a plethora of scans (e.g., 1000 scans) might decrease noise significantly, certain of the experimental efforts were aimed instead at minimizing heating to live cell samples by taking fewer scans in such examples according to the present disclosure. From these experimental efforts, it was found that 16 scans were sufficient to observe the nitrile peak within the cellular environments in each case.

[0092] In addition to the fewer scans, reduced heating effects were originally sought within the long-pathlength liquid sample cell by adding an extra 2-second dark time between scans, in addition to the typical 1.7-second time for scanning and transfer of the data captureSTFD.469PCT (S24-491) 31 buffer to the computer. Related experimentation demonstrated no significant benefit to the additional equilibration time, thereby indicating that on the timescale of these measurements, heating by the laser is not a major consideration. Further optimization of the setup can facilitate realization of a faster scan-time, such as via data streaming rather than use of a capture buffer or scanning over shorter spectral ranges (or fewer data points), in which case additional temperature controlling may be advantageous for the faster laser repetition rates. Currently, with a 1.7-second combined time for scanning over 300 cm1and data transfer, this method proves roughly one-half as fast as a scan on a Bruker FTIR (0.8 seconds) — though still with markedly better signal-to-noise, meaning the QCL spectrometer still provides an overall benefit.

[0093] These experimental efforts also evaluated potential local heating and sample damage caused by QCL illumination at maximum flux settings. A long-wavelength infrared (LWIR) thermal imaging camera (Pembroke Instruments TM-60D) was used to monitor the sample-cell holder temperature in real time; the laser beam spot was clearly visible with the laser on and not with laser off. Temporal temperature profiles were collected at the beam position with either water or bacterial suspension present. In both cases, measurements showed no temperature change until the laser was activated. Under continuous laser illumination (1 MHz and 30% duty cycle) for 100 s. with temperature sampling every 1 s, the temperature increased from ~20 °C to ~21 °C over the first 20 s, plateaued, and then gradually returned toward baseline (ambient temperature of ~20 °C) once the laser was switched off. In contrast, under the pulsed-illumination conditions used throughout this study with the 2s dark interval — no measurable heating was observed; the temperature remained constant for the full duration of the measurement. The same procedure for data processing and RMS noise calculation was followed as in FIG. 9, showing no major advantage from the additional equilibration time (and thus suggesting no heating artifacts using the 1.7-s laser repetition rate).

[0094] As an additional point of comparison for the previous implementations of Bruker Vertex 70 FTIR, the detection of benzonitrile was compared on an independent FTIR instrument, a ThermoFisher Nicolet iS50 with a liquid nitrogen cooled MCT detector. The sensitivity' of this particular instrument, particularly in the benzonitrile region, was poorer than the Bruker Vertex 70 used for other comparative measurements reported in the study. Experimentally, the benzonitrile peak did not begin to be clearly visible in the spectrum until a concentration of -500 pM, giving an experimental LOD several times higher than the 80 pM LOD on the Bruker FTIR (FIG. 11). The RMS error was also calculated for thisSTFD.469PCT (S24-491) 32 instrument in the 2220-2245 cm1region of the benzonitrile peak to be 4.4 • 10-4based on the baseline noise, an order of magnitude poorer than the Bruker FTIR. Additionally, the Nicolet iS50 requires ~1.6 seconds for each scan, twice as slow as the Bruker instrument.

[0095] Compared to the Bruker Vertex 70 FTIR, the ThermoFisher instrument requires a higher concentration of benzonitrile for detection and has a higher baseline noise, indicating that the Bruker Vertex 70 FTIR is a better comparison point for the QCL spectrometer performance. FIG. 11 shows the benzonitrile detection on a ThermoFisher Nicolet iS50 FTIR instrument.

[0096] In addition to nitriles and also according to exemplary experimental examples of the present disclosure, several other vibrational stretches that can function as readouts of molecular interactions and environments and that have been introduced into proteins by amber suppression are also useful for detection using this QCL spectrometer at low concentrations. The 2000-2300 cm1region is home to other weak vibrational oscillators including carbon-deuteri um (C-D) alkyne (C=C) and stronger azide (N=N+=N') bonds that have proven useful probes in a variety of biological contexts. These efforts demonstrate how the instrumentation, as in FIG. 1C and discussed herein, can sensitively detect sample vibrational probes at other neighboring frequencies including an alkyne (1 -hexyne) at 2111 cm a deuterated (aldehyde-H) form of N-cyclohexylformamide (CXF-D) at 2182 cm1used as a probe of electric field directionality in enzymes by Zheng et al. (A two-directional vibrational probe reveals different electric field orientations in solution and an enzyme active site, Nat. Chem., 14, 891-897, 2022), and an azide-labeled phenylalanine.

[0097] As can be seen in FIG. 12. the QCL measurement of 10 mM 1 -hexyne in water using a 250 pm pathlength gives a significantly better spectrum, with better sensitivity of the QCL in this frequency range, compared to the same number of scans (16) on the FTIR using a 50 pm pathlength. This shows how the QCL can provide good sensitivity for detection of strongly absorbing azides such as 4-azidophenylalanine as in FIG. 13. More particularly. FIG. 13 shows the spectra 1 mM 4-azidophenylalanine in water after 16 scans on either the Bruker Vertex 70 FTIR (here using 50 and 25 pm spacers to reduce etalon fringes), or on the QCL using a 250 pm pathlength, thereby showing better sensitivity of the QCL in this frequency range.

[0098] FIGs. 14A and 14B show the spectra of CXF-D at concentrations of 1 mM and 5 mM in water to compare sensitivity of the Bruker Vertex 70 FTIR at a 50 pm pathlength as in FIG. 14A, and as in FIG. 14B, the QCL spectrometer at a 250 pm, with 16 scans used on both instruments. Accordingly, these experimental efforts show how the QCL gives betterSTFD.469PCT (S24-491) 33 sensitivity for detection of the CXF-D, with a clearly visible peak at 1 mM where the same number of scans on the FTIR at the same concentration and 50 m pathlength does not yield a similarly clear peak.

[0099] Using such an QCL-based approach, experimental I more-detailed aspects of the present disclosure are directed to the detection of small-molecule binding (e.g.. as exemplified in E. coli) which is enabled with particular focus on the interaction betw een para-coumaric acid (pCA) and nitrile-incorporated photoactive yellow protein (PYP). Notably, spectral shifts of up to 1 cm1were observed in some instances, in many instances in a range from 0.25 cm1to 7 cm '. and more generally in a range from 0.25 cm1to 15 cm1for nitriles embedded in PYP betw een the unbound and drug-bound states directly within bacteria, in agreement with observations for purified proteins. Such large spectral shifts are ascribed to the changes in the hydrogen-bonding environment around the local environment of nitriles, accurately modeled through high-level molecular dynamics simulations using the AMOEBA force field.[OOiOO] It is recognized and appreciated that as specific examples, the abovecharacterized figures and discussion are provided to help illustrate certain aspects (and advantages in some instances) which may be used in the manufacture of such structures and devices. These structures and devices include the exemplary structures and devices described in connection with each of the figures as well as other devices, as each such described embodiment has one or more related aspects which may be modified and / or combined with the other such devices and examples as described hereinabove may also be found in the Appendices of the above-referenced Provisional.

[0101] The skilled artisan would also recognize various terminology as used in the present disclosure by way of their plain meaning. As one example, the term drug-like molecule refers to a molecule that can be characterized by properties that enhance successful drug development, including favorable absorption, distribution, metabolism, and excretion profiles. As further examples, the Specification may also describe and / or illustrates aspects useful for implementing the examples by way of various semiconductor materials / circuits which may be illustrated as or using terms such as layers, blocks, modules, device, system, unit, controller, and / or other circuit-type depictions. Such semiconductor and / or semiconductive materials (including portions of semiconductor structure) and circuit elements and / or related circuitry may be used together with other elements to exemplify how certain examples may be carried out in the form or structures, steps, functions, operations, activities, etc. It would also be appreciated that terms to exemplify7orientation, such asSTFD.469PCT (S24-491) 34 upper / lower. left / right, top / bottom and above / below, may be used herein to refer to relative positions of elements as shown in the figures. It should be understood that the terminology is used for notational convenience only and that in actual use the disclosed structures may be oriented different from the orientation shown in the figures. Thus, the terms should not be construed in a limiting manner.

[0102] Based upon the above discussion and illustrations, those skilled in the art will readily recognize that various modifications and changes may be made to the various embodiments without strictly following the exemplar}' embodiments and applications illustrated and described herein. For example, methods as exemplified in the Figures may involve steps carried out in various orders, with one or more aspects of the embodiments herein retained, or may involve fewer or more steps. Such modifications do not depart from the true spirit and scope of various aspects of the disclosure, including aspects set forth in the claims.

Claims

STFD.469PCT (S24-491) 35What is claimed is:

1. An apparatus comprising: a quantum cascade laser (QCL) to direct a laser beam carrying light in a wavelength range that overlaps a mid-IR range; optics, responsive to the laser beam, to process the light in a set of related beams along respective paths, including a reference path and a sample path, respectively towards a reference target and a sample target that includes living cells; and a vibrational spectroscopy detector to detect small molecules binding to a target protein in the sample target containing living cells by collecting incident light, relative to the reference target and from the sample target, carried along respective incident paths and by discerning vibrational frequency shifts on the sample target in the 2000-2300 cm1midinfrared range.

2. The apparatus of claim 1 , wherein the laser beam is to carry light in a wavelength range approximately from 2000 cm’1to 2300 cm’1or approximately from 4.3 microns to 5 microns.

3. The apparatus of claim 1, wherein the optics and the vibrational spectroscopy detector are cooperatively configured to set a 250 micron sample pathlength for the sample target and reference.

4. The apparatus of claim 1 , wherein the laser beam is to carry light in a wavelength range approximately from 2000 cm’1to 2300 cm’1, and the optics and the vibrational spectroscopy detector are cooperatively configured to set a 250 micron sample pathlength for the sample target and reference.

5. The apparatus of claim 1, wherein the optics and the vibrational spectroscopy detector are cooperatively configured to set a sample pathlength for the sample target that is adequate to measure small molecules binding to the target protein in the living cells.

6. The apparatus of claim 1, wherein the optics further includes at least one and no more than three neutral density filters optically configured between the QCL and a beam splitter, as part of the optics, to separate the laser beam from the QCL into the set of related beams.STFD.469PCT (S24-491) 367. The apparatus of claim 1, wherein the vibrational spectroscopy detector includes a balanced, thermoelectrically cooled Mercury' Cadmium Telluride detector.

8. The apparatus of claim 1, further including one or more neutral density filters to balance beams from among the set of separated beams.

9. The apparatus of claim 1, further including an optical diffuser as a polarization scrambler to depolarize an incident laser beam.

10. The apparatus of claim 1, further including a box that encloses the optics, and is to be purged by an outside source of dry air and therein remove water vapor and carbon dioxide absorption lines that would otherwise manifest in the vibrational spectrum.1 1. The apparatus of claim 1, further including CaF2 focusing optics to pass the incident light, from the reference target and from the sample target, before the incident light reaches the vibrational spectroscopy detector.

12. The apparatus of claim 1, wherein the sample target includes vibrational probes resonating within the 2000-2300 cm1region embedded in proteins within the living cells, and the vibrational spectroscopy detector is to detect at least some of the proteins that are tagged by the vibrational probes in the form of non-canonical amino acids.

13. The apparatus of claim 1, wherein the vibrational spectroscopy detector is to detect vibrational probes resonating within the 2000-2300 cm1region embedded in proteins within the living cells.

14. The apparatus of claim 1, wherein the vibrational spectroscopy detector is to detect molecular vibrational signatures embedded in proteins within the living cells.

15. The apparatus of claim 1, wherein the vibrational spectroscopy detector is to detect changes in vibrational frequency in response to binding of the proteins to an externally-added drug-like molecule.STFD.469PCT (S24-491) 3716. The apparatus of claim 1, wherein the vibrational spectroscopy detector is to detect covalent drug binding in a cellular environment.

17. A method comprising: processing a laser beam, from a quantum cascade laser (QCL), carrying light in a wavelength range that overlaps a mid-IR range; using optics, in response to the laser beam, to process the light in a set of related beams along respective paths, including a reference path and a sample path, respectively towards a reference target and a sample target that includes living cells; and detecting, via a vibrational spectroscopy detector, small molecules binding to a target protein in live cells by collecting incident light, relative to the reference target and from the sample target, carried along respective incident paths and by discerning vibrational probe frequency shifts in the sample target.

18. The method of claim 17, wherein the laser beam carries the light in a wavelength range approximately from 2000 cm’1to 2300 cm’1, and the optics and the vibrational spectroscopy detector are cooperatively configured to set a 250 micron sample pathlength for the sample target.

19. The method of claim 17 further including detecting, via the vibrational spectroscopy detector, changes in vibrational frequency in response to binding of the proteins to an externally-added drug-like molecule.

20. The method of claim 17, wherein the small molecules binding to a target protein in live cells act as chemical probes that manifest detectable changes in vibrational frequency .

Images