Spatially resolved analytical chemistry in intact, living tissues

Abstract

Tissues are an exciting frontier for bioanalytical chemistry, one in which spatial distribution is just as important as total content. Intact tissue preserves the native cellular and molecular organization and the cell-cell contacts found in vivo. Live tissue, in particular, offers the potential to analyze dynamic events in a spatially resolved manner, leading to fundamental biological insights and translational discoveries. In this Perspective, we provide a tutorial on the four fundamental challenges for the bioanalytical chemist working in living tissue samples, as well as best practices for mitigating them. The challenges include (i) the complexity of the sample matrix, which contributes myriad interfering species and causes non-specific binding of reagents; (ii) hindered delivery and mixing; (iii) the need to maintain physiological conditions; and (iv) tissue reactivity. This framework is relevant to a variety of methods for spatially resolved chemical analysis, including optical imaging, inserted sensors and probes such as electrodes, and surface analyses such as sensing arrays. The discussion focuses primarily on ex vivo tissues, though many considerations are relevant in vivo as well. Our goal is to convey the exciting potential of analytical chemistry to contribute to understanding the functions of live, intact tissues.

Graphical Abstract

Introduction

Life is composed of cells and molecules meticulously arranged in space. From the earliest embryonic development, to the organization of the brain, to the dynamics of multi-unit systems such as immunity, spatial organization drives function.^1–4 How can we quantify, or even detect, the distributions of the small and large biomolecules that govern health and disease, using tissue samples from humans and animals that lack a gene-modified reporter system? This question remains a grand challenge for the analytical sciences, in particular for bioanalytical chemistry. For the purpose of this review, we define bioanalytical chemistry as the isolation, detection, or quantification of molecular analytes in a biological matrix.

Most classic techniques of analytical chemistry – e.g. spectroscopy, electrochemistry, electrophoresis, and mass spectrometry – were initially designed to characterize well-mixed liquid solutions rather than spatially heterogeneous samples. As a result, bioanalytical methods for analysis of cell supernatants, body fluids, and homogenized cells are far better developed than methods for intact tissue, but they provide no spatial information.^5,6 There is a clear need for chemical analysis technologies that accommodate samples that preserve the organization and dynamics of the living organism. Below, we make the case for why spatially resolved molecular analyses in living, intact tissues are a vital frontier of the field. Next, we present a framework for approaching this type of analysis, in terms of four critical challenges and their solutions.

Why study intact tissue?

For chemists and engineers approaching a particular biological system for the first time, it may be appealing to work in simplified systems, such as proteins in solution, well-defined cultures of immortalized or primary cells, or homogenized tissue. Indeed, most bioanalytical methods must undergo initial validation in these reductionist systems.⁷ However, bioanalytical detection in intact tissue offers several advantages towards understanding complex biological events. Maintaining the structural integrity of the tissue allows for the retention of rare and matrix-adherent cell types, e.g. macrophages and dendritic cells, which are easily lost during tissue dissociation.⁸ Furthermore, it avoids inadvertent activation of (live) cells by dissociation procedures.⁹ Intact tissue sections also preserve the native spatial organization, including the high cell density (~10⁸ - 10⁹ cell/mL)¹⁰ that far exceeds that of typical in vitro culture conditions (~10⁶ - 10⁷ cell/mL). The full content of the extracellular matrix and stromal networks are also preserved, and need not be mimicked by advanced 3D culture systems.¹¹ Thus, complex behaviors mediated by cell-cell and cell-matrix contacts and by local accumulation of secreted growth factors are preserved.^12,13 Examples of such behaviors include morphogenic control of development,^2,14 antigen presentation and self-organization in the immune system,¹⁵ and tumor growth and immune regulation.^16,17 Thus, intact-tissue methodologies that preserve the structural integrity of the whole organ are desirable.

Why study living tissue?

Most traditional analysis of intact tissue begins with fixation, which crosslinks proteins and arrests the biological system in place.^18–21 Fixation preserves the molecular and cellular organization for days to years, but eliminates temporal dynamics. Though more challenging in some respects, analysis of living tissue in vivo and ex vivo allows for the observation of transient events, such as protein-protein interactions, temporary cell-cell contacts, and formation of molecular gradients, often in real time.^22–24 With live tissues, it is possible to design stimulus-response experiments in which the same sample is monitored continuously or at intermittent intervals, alleviating some of the challenges of heterogeneity between tissue samples.^25,26 Both real-time monitoring and repeated-measures designs, e.g. before and after delivery of a drug, are well suited to quantifying dynamic events in living tissues, including, for example, in heart, brain, and lymph node.^27–30 Ex vivo tissue slices are particularly useful in this context,^29,31 because they allow more precise control over the dose and timing of stimulation than in vivo assays, which must contend with in vivo fluid drainage and pharmacokinetics.

Why measure spatial distribution?

A spatially averaged human would not live very long, and a spatially averaged measurement offers only a limited view of biological events. Even within individual organs, cells and molecular signals are heterogeneously arranged into structures (e.g. vessels, fibers, clusters) and biochemical gradients that critically affect their function.¹¹ For example, the spatial organization of the cellular microenvironment affects differentiation in stem cells of the gut, the survival and activation of T cells in the lymph node, and cell migration in the developing kidney.^32–34 Secreted proteins, nutrients and oxygen, metabolites, and toxic by-products of metabolism are all distributed heterogeneously within the organ.^35,36 A spatially averaged measurement, e.g. of a biomolecule or a cellular marker, cannot reveal these complexities; information on regions of high and low concentration are lost completely, potentially masking critical driving factors for the behavior of interest.^12,37 Similarly, a local measurement made at a single point must be considered within the context of the heterogeneity of the tissue to avoid the false impression that it represents the entire organ – indeed, measurements at different locations may yield very different results.³⁸ Thus, there has been a renewed push over the past decade to quantify analytes such as proteins, metabolites, and small molecules in intact, living tissue with high spatial resolution.^39–41

What can be learned?

Quantifying the spatiotemporal distribution of molecules in tissue is critical for both fundamental understanding and translational insights. Events in intact tissue are multiscale, ranging from molecular interactions, to cell functions, to tissue-level emergent behaviors (Figure 1). Analyses that bridge these scales can be powerful. For example, understanding how small molecules move through living tissue, in the context of hindered diffusion, binding, uptake, and degradation, can reveal the underlying structure of the tissues,^42,43 the mechanisms of cell-cell communication,⁴⁴ and even inform targeted drug delivery strategies.⁴⁵ Revealing the dynamics of metabolite distribution, the formation of protein gradients, and the differentiation and maturation of cells provides a better understanding of the biochemistry underlying tissue development or disease progression, and also may lead to development of new drugs and immunotherapies that modulate these processes.⁴⁶ In addition to mechanistic and basic biomedical research, molecular imaging in vivo is already used to guide clinical decisions.⁴⁷ Ex vivo, point of care analysis of tissue biopsies may be used for personalized medicine,^48–52 e.g. early diagnosis of cancer.^53,54 Furthermore, high-content, multi-parametric data from ex vivo samples can synergize with the use of artificial intelligence to inform precision or personalized medicine.⁵⁵

Figure 1:

When taking measurements in live tissue, the final distribution and local concentrations of analyte are determined by events at (a) the whole tissue, (b) cellular, and (c) molecular length scales.

Context and scope of review.

Handed a sample of live tissue, there are myriad approaches to begin to analyze its chemical contents (Figure 2). This review is focused on measurements that maintain the tissue in its live, intact state, with an emphasis on ex vivo analysis. Here we distinguish between this and complementary approaches. Many studies of spatial organization begin with tissue fixation, which stabilizes both the location and state of cells and molecules (Figure 2a). Spatially resolved analysis of fixed tissue is a booming field, with exciting progress made on high-resolution techniques such as CLARITY,⁵⁶ expansion microscopy,⁵⁷ spatial transcriptomics,^58,59 CyTOF,^60,61 and SEM.⁶² However, as with all terminal analyses, it is challenging to use fixed tissue sections to gain insight into dynamic events such as transient interactions, stimulus-response behaviors, cell fate decisions, motility, and so on. An additional terminal approach is to disrupt or homogenize the tissue and analyze the resulting cell suspension or molecular mixture,⁶ e.g. for flow cytometry,⁶³ single cell analysis,^24,64–66 or total gene expression analysis.⁶⁷ These techniques provide valuable multi-parameter data (i.e. a dozen or more fluorescent markers, thousands of genes) on a per-cell basis, but without information on the original spatial location of those cells. Alternatively, the tissue may be kept intact, but with analysis limited to molecules that are secreted into the culture supernatant, e.g. by ELISA,^68,69 chromatography, and mass spectrometry⁷⁰ (Figure 2b). Measurements of secretion can provide insight into tissue function, but the data are averaged over all cells, space, and time in culture.

Figure 2:

Approaches to chemical analysis in samples of tissue. (a) Tissue fixation and dissociation are terminal analyses, while (b) analysis of culture supernatants or tissue homogenates provides a spatially averaged measurement. (c-d) Spatially resolved analyses of live tissue may primarily sample the surface (a), or may penetrate into the depth of the tissue (d).

Here, we discuss the challenges and best practices to analyze chemical content and distribution in situ in live, intact tissue samples, e.g. by targeted fluorescent imaging, electrochemistry, microdialysis, mass spectrometry, microfluidics, or surface arrays (Figure 2c–d). We discuss considerations for work in live tissues ex vivo and in vivo, including matrix effects, molecular transport, the fragility of tissue and off-target effects.

Challenges and Solutions for Bioanalytical Chemistry in Living Tissue

Chemical analysis of live tissue offers four fundamental challenges: hindered transport, matrix effects, biocompatibility, and tissue reactivity. The details of these challenges differ between organs due to variations in cell type, density, and local chemistry. Nevertheless, there are general approaches to reduce the impact of each challenge, to make analysis in living tissue possible.

Hindered Transport: Tortuosity and Binding

Understanding how the tissue environment affects the transport of reagents and analytes in tissue is crucial to understanding the strengths and limitations of in-tissue analytical chemistry. Compared to diffusion in free solution, diffusion in cells and tissue is restricted by physical obstructions, e.g. organelles, cells, or fibers in the extracellular matrix (ECM).⁷¹ This reduction is quantified by a term called tortuosity, which depends on the size of the molecule.^72,73 Even in free solution, macromolecules diffuse about 20 times slower than small molecules;⁷¹ this difference is enhanced by tissue tortuosity, with small molecule and protein diffusion coefficients reduced by ~ 2 to 10-fold, respectively.^74,75 Biologically, diffusion gradients of nutrients, oxygen, and signaling proteins play critical roles in tissue function. Analytically, restricted diffusion means that reagents may require significant time to reach their targets within the tissue, and secreted analytes may require time to reach a sensor or probe (Figure 3a). For example, while small molecules like oxygen and glucose may diffuse through 300 μm of human tissue in about 3–8 minutes (distance² = 2Dt, where D = 0.91 – 2.6 ×10⁻⁶ cm² sec⁻¹ and t is time),⁷⁶ macromolecules such as IgG antibodies require approximately 2 hr even in the absence of antigen binding (D = 6.4 ×10⁻⁸ cm² sec⁻¹ in mouse brain).⁷⁵ The same restriction happens in reverse for tissue-secreted analytes that must reach the surface of a culture substrate to be detected (Figure 2c). Meanwhile, reagents that must penetrate into cells face an additional barrier of the cell membrane; methods for cytosolic delivery are an area of active investigation (Figure 3b).⁷⁷ In general, only small, neutrally charged, hydrophobic substances can passively diffuse through cell membranes, by mixing with the hydrocarbon-rich interior.⁷⁸

Figure 3:

Chemical analysis in live tissue must account for hindered transport of reagents and analytes at two length scales. (a) At the tissue scale, the dense and negatively charged tissue matrix limits the diffusion of large and/or highly-charged molecules, resulting in a limited depth of penetration (darker shading). (b) At the cellular level, the cell membrane poses a barrier to reagents intended to reach the interior of the cell, particularly for large and highly charged molecules.

Separate from tortuosity, the tissue environment is dense with biomolecules that further hinder molecular transport through binding events and/or electrostatic interactions.⁷⁹ For example, positively charged reagents are prone to electrostatic binding to ECM due to its overall negative charge.⁸⁰ Macromolecular reagents, such as proteins, may specifically bind to species other than their intended target within tissue, in what is known as cross-reactivity or off-target binding.⁸¹ For antibodies, the Fc region may bind to Fc receptors (FcR) that are broadly expressed within tissue.⁸² This extent of this interaction depends on the antibody isotype and the types of cells that are present, and can lead to non-uniform reagent distribution. Indeed, penetration of IgG into mm-thick fixed and cleared tissue, for instance, requires days or weeks, likely due to a combination of size-mediated tortuosity and Fc receptor binding.^75,83

Taking into account the restricted transport within tissue can help reduce its effects on the bioanalytical workflow. For example, diffusion time should be estimated and accounted for when optimizing reagent delivery times.²⁹ If the diffusion coefficient of the reagent is not available, it can be estimated in many cases from molecular weight and the Stokes-Einstein relationship or other models for radius of gyration.^84–86 Labels with smaller molecular weight – e.g. a small molecule fluorophore (Alexa Fluors and similar probes) versus a protein-based fluorophore (green fluorescent protein and its derivatives) – may offer a shorter diffusion time. Similarly, a small molecule or peptide-based probe will diffuse more quickly than a protein. In some cases, it may be possible to accelerate delivery by using electrophoresis or convection.^83,87

Designing the assay with an eye towards minimizing non-specific binding is also useful as a means to achieve uniform reagent delivery. Many fluorophores are bulky and contain charged groups,⁸⁸ and as a result, their diffusion into tissue samples is slow, and so are not suitable for use inside living tissue.⁸⁹ Charged fluorescent probes such as ANS (1-anilinonaphthalene-8-sulfonicacid) probes and DDAF (5-(N-dodecanoyl)aminofluorescein)^90,91 have higher non-specific and off-target binding and may not penetrate thick tissues. On the other hand, neutral probes such as pyrenes are not as sensitive to the effect of charged surroundings.⁹² Vital dyes such as Calcein AM ester are specifically designed to be neutral until they pass through the cell membrane and are cleaved intracellularly, and these dyes readily label live cells in intact tissue.⁹³ In some cases, the extent of non-specific binding of a reagent can be directly evaluated, e.g. by using surface plasmon resonance, affinity chromatography, or immunosorbent assays.^94,95 Known non-specific or off-target interactions can be intentionally mitigated, or “blocked,” as in the use of anti-FcR antibodies as a pre-treatment prior to immunohistochemical staining.²⁹ Alternatively, reagents can be modified to minimize off-target binding. Antibodies can be fragmented into Fab or F(ab)₂ to remove the risk of FcR binding, while simultaneously reducing molecular weight for better diffusivity.⁹⁶ Other reagents have been covalently conjugated with polyethylene glycol, which reduces non-specific binding and cellular uptake and enhances delivery.⁹⁷ Thus, careful experimental design, choice of reagents, and an understanding of techniques can mitigate the effects of hindered transport in live tissue samples.

Matrix Effects

In classic analytical chemistry, a matrix effect is any alteration of the results of the measurement by components of the sample other than the analyte. Tissue is rife with potential matrix effects as it is packed with the fibers and glycoproteins of the extracellular matrix (ECM), mixed populations of cells, and abundant secreted molecules (Figure 4). Endogenous signals such as proton gradients, reactive oxygen species, and biotin can interact directly with chemical probes. In addition, light scattering and tissue autofluorescence can affect optical readouts. Matrix effects due to non-specific binding, cross-reactivity of reagents, or heterogeneous autofluorescence can contribute to any of three possible problems with a bioanalytical assay: false positives, high background, and signal enhancement or suppression.

Figure 4:

The extracellular matrix provides a multitude of opportunities for off-target interactions. Within the tissue microenvironment, a molecule of interest (green square), e.g. a reagent or an analyte, may bind (a) to its known or intended receptor on the cell surface, (b) to an off-target receptor, (c) nonspecifically to the matrix or cell surface via electrostatic interactions, or (d) non-specifically via hydrophobic interactions. (e) Assays based on fluorescence imaging are subject to interference from the autofluorescence of tissue components. The emission spectra shown here were from pure compounds in solution, except for collagen, elastin, and lipopigments, which analyzed in tissue samples. Spectra were normalized to their respective maxima. Excitation wavelength: 366 nm. Reproduced with permission from ref. 98. Copyright (2014) A.C. Croce and G. Bottiroli.

It is critical to consider and control for potential false positives in each bioanalytical assay, by considering all potential sources of non-specific or off-target binding. For example, large planar hydrophobic molecules, e.g. red fluorophores such as Rhodamine Red C2,⁹⁹ bind non-specifically and should be used with caution to reduce the potential for false positives. If such molecules are used, then control experiments that explicitly test their effect on the measurement, i.e. by conducting the same assay with two different labels, are crucial to avoiding undesirable results.⁹⁹ Similarly, antibodies pose a particular risk to off-target binding events due to FcR binding. If a fragment or alternative cannot be used, then the sample should be analyzed in parallel with an isotype control (an antibody of the same isotype, with an irrelevant specificity) to distinguish analyte-specific signal from Fc-mediated binding.¹⁰⁰ Analogous controls should also be used for other types of detection reagents (peptides binders, aptamers, inhibitors, enzyme substrates, etc).

Other strategies for reducing matrix effects in bioanalytical systems are already familiar to analytical chemists: matrix matching and standard addition. Matrix matching requires that the assay calibration and controls be performed under conditions mimicking the chemical environment of the tissue, or ideally in the tissue itself. In particular, collection of negative control data from a tissue sample that does not receive one or more assay reagents is useful for an accurate assessment of background signal, which may vary across the tissue. For quantitative assays, standard addition may be useful to evaluate signal enhancement or suppression. Standard addition requires spiking known quantities of analyte into the tissue for readout, which is possible only if the tissue is not altered by the addition (see Tissue Reactivity), or if the tissue can be fixed prior to standard addition without harm to the assay. These experimental strategies, as well as proper positive and negative controls as described above, help distinguish real signal from the heterogenous background that is found in live tissue samples.

For optical microscopy in particular, two additional matrix effects affect the signal/noise ratio: light scattering and autofluorescence. First, the presence of lipid-laden cells causes significant light scattering, which usually prevents optical detection of analytes deeper than ~100 μm (confocal) or ~300 μm (two-photon) in intact tissue, barring cutting edge imaging techniques.¹⁰¹ Infrared and Raman-based imaging methods provide deeper penetration, but offer lower spatial resolution, limited imaging speed, and fewer commercially available probes at present.^102,103 Second, tissue autofluorescence can cause high and variable background if not properly accounted for.¹⁰⁴ Autofluorescence from the matrix arises from multiple sources – collagen, flavins, fatty acids, etc (Figure 4b).⁹⁸ Fluorescence emissions from these components span the greater part of the UV and visible light spectra (400–600 nm), with low but detectable contributions in the near-IR (~700 nm), and so can never be safely ignored. The contribution of autofluorescence should estimated by the use of unstained controls, allowing for baseline subtraction.¹⁰⁵ Because the intensity of autofluorescence may be quite heterogeneous across the tissue,¹⁰⁶ it is ideal if the same samples are analyzed before and after staining to generate matched unstained controls.

Biocompatibility

When working with live tissue, reagents and culture conditions must be carefully chosen to ensure that fragile lipid bilayers remain intact, proteins remain folded, and cells remain in appropriate contact. In fact, this fragility is taken advantage of fixation protocols, which apply organic solvents to crosslink proteins or permeabilize membranes,¹⁰⁷ and in tissue-clearing methods like CLARITY, which applies solvents or surfactants to remove all lipids and render the tissue transparent.⁵⁶ In contrast, if live tissue samples are to be used, they cannot come into contact with organic solvents or lipid-clearing surfactants. One consequence of this requirement is that intracellular analysis is limited to reagents that do not require membrane permeabilization to reach the cytosol; in particular, intracellular antibody staining is not yet possible in live tissue. Furthermore, the assay must be conducted at physiological pH and osmolarity and at biocompatible temperatures (4 – 40 °C), and all reagents must be functional under these conditions (Figure 5a). If held for more than one or two hours, the tissue must be maintained in a culture media comprised of the appropriate balance of nutrients for optimal cell function. Maintenance of biocompatible conditions and the ability to tune those conditions is sometimes facilitated by microfluidic culture systems.^108–111

Figure 5:

Analysis of live tissue requires mitigation of potential tissue damage and preservation of physiological culture conditions. (a) Tissue must be kept under physiological culture conditions to maintain integrity and cell viability. (b) Probe insertion, such as for electrochemistry or microdialysis, causes physical damage that may radiate outward from the site of insertion. (c) Short-wavelength light used to visualize fluorescence may cause phototoxicity, potentially altering the state of the cells in the imaged regions.

As many spatially resolved bioanalytical assays require physical insertion of a needle, electrode, or sensor-coated probe, it is useful to consider the extent of damage caused by such probes (Figure 5 b). Typical mammalian cells are 5–20 μm in diameter, depending on the cell type. Thus, a 200-μm o.d. microneedle may directly destroy a region of greater than 10 cells across. Furthermore, estimates from studies in the brain suggest that the radius of damage extends approximately 200 μm from the surface of the probe.¹¹² Propagation of damage may be due to a combination of mechanical shear stress and diffusion of damaged cell products from the site of injury, activating nearby cells. Nevertheless, probe insertion is often the best, or only, way to make a required measurement, e.g. for local electrochemical detection and microdialysis-type experiments.^113,114 The extent to which damage from insertion may or may not affect the measurement likely varies by analyte, analysis method, and the type of tissue. To minimize damage, the probe with the smallest possible diameter should be used, and the tissue should be allowed to “recover” from insertion prior to analysis.¹¹⁵ Many studies have optimized the force, speed, design, and diameter of needles to ensure that tissue damage is minimized while maintaining high sensitivity.^116–118 Alternatively, microfluidics may be used in some ex vivo settings to deliver local stimulation or collect local secretions without probe insertion.^87,119–122

Optical detection provides spatially resolved analysis without physical probe insertion, but it comes with its own caveats for live cell and tissue imaging. Phototoxicity, or damage within the tissue due to light exposure, is problematic especially under UV excitation or high light intensities (Figure 5c).; while impossible to remove completely, it can be mitigated.¹²³ Maximizing the excitation efficiency through the use of filters and moving to longer wavelength (near-IR) excitation have been shown to reduce cellular damage.^124,125 Even after minimizing photodamage, it remains essential to image all samples, including negative controls, under the same imaging conditions (light source, intensity, wavelengths, exposure time, etc.), to control for potential effects on the tissue. If live samples will be imaged repeatedly, the impact of repeated light exposure on the functions and readouts of interest should be tested explicitly.

Tissue Reactivity

In addition to merely maintaining viability, work with live tissue demands recognition that the tissue may react to the reagents or procedures, potentially altering the quantity or distribution of the analyte of interest (Figure 6). Several commonly used bioanalytical reagents can activate or otherwise engage cells. For example, multivalent binding reagents such as antibodies and dendrimers may dimerize or cross-link proteins on cell surfaces, which often activates intracellular signaling cascades and may even result in internalization of the target (Figure 6a).^126,127 In addition, antibody binding to FcR on immune cells, such as macrophages, B cells, NK cells, follicular dendritic cells, and other effector cells, may activate those cells and initiate an immune response (Figure 6b).^82,128 One means to mitigate these issues is to use monovalent binding reagents such as single Fab or scFv antibody fragments, though these reagents are of more limited availability than intact antibodies.¹²⁹

Figure 6:

The addition of reagents has the potential to alter the sample by causing unintended biological interactions. (a) Surface receptors (green) with inactive cytoplasmic domains (yellow) may become cross-linked by multi-site binders such as antibodies, resulting in clustering and activation of cytoplasmic domains (red) and activation of downstream signaling cascades. (b) Fc receptors (FcR; cytoplasmic domain in yellow) may bind to immune complexes of reagent antibodies and antigens, resulting in immune activation by sending signal through activation of cytoplasmic domain (red). (c) Reagents added to the tissue may bind or interact with cells in a manner that activates them or otherwise alters their state, potentially affecting the distribution of analyte. (d) At the tissue length scale, these combined molecular effects of the assay reagents may potentially result in a fundamentally altered distribution of the analyte (represented in green) if not carefully controlled.

In addition to crosslinking and off target binding, other reagents may otherwise alter the state of live cells (Figure 6c) and just alter the distribution of the analyte in the tissue (Figure 6d). Some chemical and optical probes are cytotoxic, through DNA intercalation,¹³⁰ binding and interfering with proteins and enzymes, or disruption of cell activity through other mechanisms.^131–133 Some chemicals, like the fluorescent MitoTracker Orange, modulate the electric potential across the plasma or mitochondrial membrane, interfering with cell signaling, electron transport chain and ATP production.¹³⁴ If the toxicity is delayed, then such chemicals may be acceptable in a terminal assay (one in which the tissue will not be returned to culture after analysis),⁹³ but not for use in long-term cultures. Delayed toxicity may be detected by labelling tissue with the reagent of interest, then continuing the culture and testing viability and function hours or days later. In addition to effects caused by the addition of chemical probes, live tissue reacts to its culture environment beyond mere changes in viability. Just as for cell cultures, the concentrations of salts, glucose, and amino acids present in the culture media can have a major effect on cell and tissue function, as do external factors like temperature and oxygenation.¹³⁵ The timescale of the tissue response to these variables may vary from seconds to days; if the response is slow, it may be possible for the tissue to make a brief excursion to a non-ideal environment for a rapid analysis procedure.

It is important to clearly define the intended use of the assay prior to test for tissue reactivity, as the act of experimentation will undoubtably have at least some minimal unexpected effect within the tissue. Such effects may be acceptable as long as the analytes and functions of interest remain unaltered. In summary, for every bioanalytical assay for live tissues, it is critical to test the extent to which the reagents and procedures alter the concentration or distribution of the target analyte, and, if possible, to verify the desired response by an independent method.

Conclusions and Outlook

In summary, when designing chemical analyses for live tissues, awareness of the unique features of the particular sample – the extent of hindered transport, potential for matrix effects, and particularly vulnerable or reactive cell types – should inform everything from choice of reagents and environmental conditions to the validation strategy. Confirmation that the reagents reach their intended target specifically and selectivity, or that the analyte can reach an implanted sensor (or the tissue surface, for surface-based analysis), is a useful initial validation parameter. Ideally, each assay should be validated in terms of the extent to which it impacts viability and the most relevant functions of the tissue.¹³⁶ As in any biological experiment, choosing appropriate positive controls (e.g. to elicit high concentrations of analyte) and negative controls (e.g. blocking analyte production and/or excluding key reagents) ensures that results can be interpreted with confidence. Looking ahead, we anticipate that continuing development of bioanalytical methods for spatially resolved analysis of live tissues will offer unique biological insights and provide a deeper understanding of mechanisms of health and disease.