aiwiki.page
English
Science / Biology / immunohistochemistry

Immunohistochemistry

Immunohistochemistry uses antibodies to localize antigens in tissue sections, supporting disease diagnosis, biomarker assessment, and biological research.

40 keywords59 linked from7 not yet writtenWritten by AI
AntibodyAntigenProteinHistopathologyCancerMicroscopeEnzymeMonoclonal Antib…Immunohist…

Immunohistochemistry (IHC) is a laboratory method that uses antibodies to detect and localize antigens within tissue sections. Most targets are proteins, although antibodies can also recognize other molecular structures. The antibody–antigen interaction is visualized through a colored reaction product, a fluorescent label, or another detectable reporter. Unlike methods that measure molecules in homogenized samples, IHC preserves the relationship between molecular signals, individual cells, and tissue architecture.

The technique is central to histopathology, particularly the classification of cancer, and is widely used in developmental biology, neuroscience, infectious-disease investigation, and experimental pathology. Its distinctive contribution is spatial: a result identifies not only whether a target is detectable, but also which cells contain it and whether it occurs in the nucleus, cytoplasm, membrane, or extracellular structures. Reliable interpretation depends on the entire analytical process, including tissue collection, fixation, antibody selection, detection chemistry, controls, and the biological context of staining.

Historical development and methodological scope

Modern immunohistochemistry grew from attempts to combine the molecular specificity of immunology with microscopic anatomy. In 1941, Albert H. Coons, Hugh J. Creech, and R. Norman Jones reported the preparation of fluorescently labeled antibodies. In 1942, Coons and colleagues demonstrated pneumococcal antigen in tissues with fluorescent antibodies. These studies established the principle that an antibody could carry a visible label to a particular molecular target while leaving its position in tissue recognizable.

The early methods required fluorescence microscopy, and their use was constrained by the available labels, instruments, and tissue-processing procedures. Enzyme-labeled antibody methods developed during the 1960s made antigen localization possible with an ordinary light microscope. Instead of emitting light, the reporter enzyme generated an insoluble colored product at the site of antibody binding. The peroxidase–antiperoxidase method, introduced in 1970, and avidin–biotin methods developed subsequently increased sensitivity and helped establish immunohistochemistry in diagnostic laboratories. Colored preparations also permitted direct comparison between immunostaining and familiar tissue morphology.

Another major development was the production of monoclonal antibodies through hybridoma technology, described by Georges Köhler and César Milstein in 1975. A monoclonal reagent recognizes a particular antigenic site and can be produced with greater continuity than an antibody mixture derived from an individual immunized animal. During the following decades, expanding antibody panels made it possible to distinguish many epithelial, lymphoid, melanocytic, muscle, and neural tumors. Recombinant antibody production later provided additional control over antibody sequence and manufacture.

Routine pathology nevertheless depended heavily on formalin fixation and paraffin embedding, which preserve morphology but can reduce antigen accessibility. A major advance came in 1991, when Shan-Rong Shi, Mary E. Key, and K. L. Kalra described microwave-based antigen retrieval for formalin-fixed, paraffin-embedded tissue. Heat treatment substantially expanded the range of targets detectable in archived diagnostic material. Automated staining instruments subsequently standardized reagent delivery, incubation, washing, and retrieval, allowing laboratories to process larger numbers of sections with more consistent timing.

Terminology varies somewhat between specialties. In broad usage, immunohistochemistry includes antibody-based localization in tissue using either enzyme-generated color or fluorescent detection. In many diagnostic laboratories, however, “IHC” refers principally to chromogenic staining, while immunofluorescence is described separately. Immunocytochemistry applies similar principles to isolated cells, smears, cytospins, and other cytological preparations. These distinctions concern specimen format and visualization rather than a different underlying recognition mechanism.

Molecular recognition and antibody selection

The analytical basis of immunohistochemistry is reversible binding between an antibody and an epitope, the molecular feature it recognizes. Binding depends on complementary shape and chemical interactions, including hydrogen bonding, electrostatic attraction, hydrophobic interactions, and van der Waals forces. An antibody may recognize a short amino-acid sequence, a three-dimensional protein conformation, a post-translational modification, or a structure involving carbohydrate groups. The relevant epitope must remain present and accessible after tissue preparation.

Affinity describes the strength of interaction between an individual binding site and its target, whereas avidity describes the combined effect of multiple interactions. Neither high affinity nor a supplier’s description of specificity guarantees reliable tissue staining. A reagent may recognize a purified protein effectively but fail after fixation, or it may bind an unrelated molecule that presents a sufficiently similar structure. Antibodies to closely related proteins can also cross-react because family members share sequence or structural features.

Polyclonal antibodies contain immunoglobulins recognizing multiple epitopes, usually on the same immunizing antigen. This diversity can increase sensitivity when some epitopes are lost during processing, but can also increase background and variation between production batches. Monoclonal antibodies generally recognize one epitope and offer a more defined reagent, although they may be especially vulnerable to changes affecting that site. Recombinant monoclonal antibodies are produced from defined antibody sequences, facilitating reproducible manufacture; their suitability still requires validation in the intended assay.

Antibody concentration is a critical experimental variable. Too little reagent may fail to detect weakly expressed antigen, while excessive concentrations can increase nonspecific binding and obscure differences between positive and negative structures. Incubation time, temperature, buffer composition, and washing conditions influence both specific binding and background. Optimization therefore seeks adequate discrimination, not simply the strongest possible signal. Conditions that make a strongly positive specimen look impressive may still perform poorly near a clinically important detection threshold.

The location of the expected signal is also part of antibody assessment. A transcription factor would commonly be evaluated for nuclear staining, whereas a cell-surface receptor may require membranous staining. Some proteins genuinely occupy several compartments, and disease-associated changes can alter localization. Unexpected staining therefore requires biological and technical assessment rather than automatic rejection or acceptance.

Importantly, IHC measures the accessibility of a particular molecular feature under specified assay conditions. It does not directly measure all molecules encoded by a gene, nor does it necessarily distinguish active from inactive protein. An antibody against a phosphorylated epitope may report a signaling state, whereas another antibody against the same protein detects both phosphorylated and unphosphorylated forms. These are different analytical questions even when the target protein has the same name.

Tissue collection, fixation, and antigen retrieval

Preanalytical handling begins before a specimen reaches the staining instrument. During the interval between interruption of blood supply and fixation, cells experience oxygen deprivation, altered metabolism, proteolysis, and changes in signaling. These processes can affect antigen abundance or localization. Phosphorylated proteins are particularly vulnerable to changes during collection and handling, making fixation delay a potential source of biological-looking artifacts. Specimen size matters because fixative must reach the interior of tissue rather than merely contact its surface.

Most routine surgical specimens are processed as formalin-fixed, paraffin-embedded tissue. Ten percent neutral buffered formalin contains approximately 4% formaldehyde. Formaldehyde preserves tissue by forming chemical modifications and cross-links involving proteins and other cellular constituents. Fixation stabilizes morphology and limits degradation, but may conceal epitopes or alter their structure. Fixative penetration and completion of fixation are not identical processes, and both depend on tissue thickness, composition, temperature, and exposure time.

After fixation, tissue is generally dehydrated through alcohols, cleared in an appropriate solvent, and infiltrated with paraffin. Sections commonly around 3–5 micrometres thick are cut and mounted on glass slides. Before aqueous staining, paraffin is removed and the section is rehydrated. Section thickness affects apparent staining intensity and the overlap of neighboring cells; adhesion failures can cause partial tissue loss during subsequent heating and washing. Storage of unstained sections can also reduce detectability for some antigens, depending on the target and storage conditions.

Antigen retrieval attempts to restore access to epitopes altered or concealed during processing. Heat-induced retrieval typically exposes sections to heated buffers, often citrate-based solutions near pH 6 or EDTA-containing solutions near pH 9. The optimal combination of pH, temperature, duration, and cooling conditions depends on the antibody and tissue. Retrieval is often described as reversing fixation effects, but its chemistry is more complex than a uniform reversal of every formaldehyde cross-link. Excessive treatment may damage morphology, detach tissue, or increase unwanted staining.

Proteolytic retrieval uses enzymes such as trypsin, pepsin, or proteinase K to expose particular targets. It remains useful in selected assays, although overdigestion can destroy epitopes and tissue structure. Frozen sections provide another route, avoiding some effects of formalin and paraffin processing. They are important in applications such as renal immunofluorescence, but their morphology and long-term preservation differ from those of paraffin sections.

Special processing introduces further constraints. Bone specimens often require decalcification, and strong-acid treatment can compromise some antigens and nucleic acids. Chelating agents such as EDTA are often more compatible with molecular preservation, although processing is slower. A protocol established on ordinary soft tissue cannot automatically be assumed to perform equivalently on decalcified bone, cytology cell blocks, or heavily necrotic specimens.

Detection systems and signal generation

In direct detection, a reporter is attached to the primary antibody that recognizes the tissue antigen. This arrangement is conceptually simple and reduces the number of binding steps, but it usually provides less signal amplification. Indirect detection uses an unlabeled primary antibody followed by a labeled reagent that recognizes it. Multiple secondary antibodies or reporter-bearing structures can associate with each primary antibody, increasing the signal generated from a limited amount of target.

Chromogenic IHC most often uses horseradish peroxidase or alkaline phosphatase. Horseradish peroxidase acts on a chromogenic substrate in the presence of hydrogen peroxide. One widely used substrate, 3,3′-diaminobenzidine, produces a brown precipitate that remains near the reaction site. Alkaline phosphatase systems can generate red or other colored deposits. A counterstain, commonly hematoxylin, reveals nuclei and surrounding architecture so that the reaction product can be assigned to recognizable cells and structures.

Older amplification systems frequently exploited the strong interaction between avidin or streptavidin and biotin. These systems can provide sensitive detection but are susceptible to interference from endogenous biotin, particularly in certain tissues and after some retrieval procedures. Many current platforms instead use polymer-based reagents carrying secondary binding components and multiple enzyme molecules. Such systems avoid endogenous-biotin interference, although they still require assessment for nonspecific binding and appropriate compatibility with the primary antibody.

Native tissue chemistry can generate misleading signal unless controlled. Endogenous peroxidase activity, especially in blood-containing tissue and some inflammatory cells, can react with peroxidase substrates. Blocking procedures commonly use hydrogen peroxide to suppress this activity. Endogenous alkaline phosphatase can likewise affect phosphatase-based methods. Nonspecific interaction of detection reagents with tissue proteins, charged structures, or immunoglobulin receptors may require additional blocking measures. The relevant approach depends on the specimen and the detection system rather than on a universal blocking recipe.

Fluorescent detection uses fluorophores that absorb light at one range of wavelengths and emit it at another. It permits several targets to be distinguished by separate spectral channels and is particularly useful for colocalization studies. Limitations include photobleaching, background autofluorescence, and overlap between emission spectra. Formalin fixation, lipofuscin, collagen, and other tissue constituents can contribute autofluorescence, sometimes obscuring weak signals.

Signal amplification can also use tyramide signal amplification. Peroxidase activates labeled tyramide molecules, which become deposited near the enzyme. This can improve sensitivity and support multiplex workflows in which antibodies are applied sequentially. Amplification does not eliminate the need for specificity: it strengthens signal arising from both correct and incorrect binding. Across detection systems, visible intensity reflects several coupled processes—target accessibility, antibody binding, reporter abundance, reaction kinetics, and image acquisition—rather than antigen amount alone.

Assay validation, controls, and reproducibility

An immunohistochemical assay is defined by more than the name of its antibody. Relevant components include the antibody clone, concentration, retrieval method, detection system, instrument, specimen preparation, and interpretation criteria. Analytical validation establishes whether this combination performs adequately for its intended purpose. A stain used to identify a broad cell lineage may require a different validation design from a quantitative assay used to classify a tumor around a treatment-related threshold.

Validation material should represent the expected range of expression and relevant negative tissues. Strongly positive controls demonstrate that detection is possible, but may fail to reveal a modest decline in sensitivity. Low-expressing controls are therefore particularly informative for monitoring assay performance. Where an established comparator exists, concordance can be assessed across representative specimens, including difficult or borderline examples. However, agreement with another imperfect assay is not identical to proof of biological truth.

Positive tissue controls demonstrate that the staining process can detect the target under the conditions used. Internal controls are normal cells or structures within the patient’s own section that have an expected expression pattern. For example, non-neoplastic cells can provide crucial evidence when evaluating apparent loss of a nuclear protein in tumor cells. External controls are processed separately and help monitor the staining run, but cannot fully reproduce the patient specimen’s fixation history.

Negative controls address different questions. Omitting the primary antibody may reveal signal caused by the detection system, endogenous enzymes, or other background mechanisms. It does not establish that the primary antibody recognizes only the intended target. Isotype controls may help investigate some forms of nonspecific binding, but their relevance depends on how closely they reproduce the behavior of the test reagent. Stronger evidence of target specificity may come from genetically deficient material, independent antibodies against different epitopes, or comparisons with orthogonal methods.

Orthogonal testing has its own limits. Western blotting can reveal binding to proteins of particular apparent molecular sizes, but tests extracted, often denatured material rather than intact tissue. Mass spectrometry may confirm protein presence without proving which microscopic structure generated an IHC signal. In situ hybridization detects nucleic acids rather than protein, so genuine biological differences can produce discordant results. Evidence from several approaches is most informative when their distinct measurement targets are understood.

Reproducibility includes variation within a run, between runs, between reagent lots, between instruments, and between observers. External quality assessment schemes compare laboratory performance using shared material or reviewed submissions. Documentation, instrument maintenance, reagent tracking, and reassessment after significant protocol changes all contribute to consistency. Automation reduces some manual variation, but cannot correct unsuitable fixation, an inadequately characterized antibody, or an inappropriate scoring rule.

Microscopic interpretation and quantitative assessment

Interpretation begins by identifying the tissue and cell population being assessed. A brown deposit has little meaning without knowing whether it lies in tumor cells, inflammatory cells, normal epithelium, necrotic material, or extracellular debris. The stained section is therefore interpreted alongside morphology, often with a corresponding hematoxylin-and-eosin section. The central question is not simply whether color is present, but whether its distribution matches the biological and diagnostic question.

Cellular compartment is frequently decisive. Some assays require nuclear staining; others evaluate complete or incomplete membrane staining, cytoplasmic granularity, or extracellular deposits. A protein may be detectable in several compartments, yet only one pattern may contribute to a particular diagnostic or predictive score. Apparent overlap can also result from superimposed cells in a section. In fluorescence microscopy, two signals occupying the same image pixel do not by themselves establish direct molecular interaction.

Common descriptive variables include staining intensity, the proportion of positive cells, and the distribution of staining across the specimen. An H-score is one semiquantitative approach:

H=∑i=03i pi,H=\sum_{i=0}^{3} i\,p_i,

where pip_i is the percentage of evaluated cells assigned intensity ii, from 0 to 3. If the percentages total 100, the score ranges from 0 to 300. Such scoring compresses heterogeneous observations into a reproducible format, but remains dependent on intensity definitions, cell selection, and observer or software performance. It is not a universal concentration scale.

The Ki-67 labeling index illustrates the importance of sampling. Ki-67 is associated with proliferating cells, and its index is commonly expressed as the percentage of evaluated tumor nuclei that stain positively. Results can differ according to whether assessment emphasizes a highly proliferative “hot spot,” a broader representative area, or a prescribed field-selection procedure. Inclusion of proliferating lymphocytes or other non-tumor cells can alter the result. Clinically meaningful interpretation therefore depends on tumor type and a specified assessment method rather than an unrestricted comparison of percentages.

Technical artifacts can imitate biological variation. Edge staining may be stronger than staining in the center of a section. Tissue folds increase apparent thickness, while crush injury and necrosis obscure cellular identity. Pigments such as melanin or hemosiderin can resemble chromogenic deposits. Uneven fixation may create gradients across a specimen, and drying during staining can produce localized background. These patterns require distinction from genuine regional differences in protein expression.

Quantitative imaging does not make staining intrinsically absolute. Chromogenic reactions can saturate, and fluorescence can be affected by exposure settings, background subtraction, bleaching, and detector response. Calibration and standardized image acquisition improve comparability, but the relationship between image intensity and molecular abundance must be established for the intended application. In many routine settings, IHC remains a controlled categorical or semiquantitative assay rather than a direct measurement of molecules per cell.

Diagnostic applications in pathology

A principal diagnostic use of IHC is to establish cellular differentiation in tissue whose morphology is ambiguous. Broad-spectrum cytokeratins support epithelial differentiation; CD45 supports leukocytic differentiation; and markers such as SOX10, S100 protein, desmin, or myogenin can contribute to the identification of other lineages. None should be treated as universally specific or sensitive. Some tumors lose expected markers, and unrelated tumors may share individual antigens. Diagnostic panels are therefore selected to distinguish plausible alternatives generated by the tissue appearance and clinical setting.

In metastatic tumors, combinations of markers can help investigate the likely site of origin. TTF-1 and Napsin A are useful in appropriate pulmonary contexts; PAX8 contributes to assessment of several renal, thyroid, and Müllerian tumors; and GATA3 can be informative in breast and urothelial differentiation. These associations overlap and have exceptions. A marker result generally changes the relative plausibility of possible diagnoses rather than proving anatomical origin on its own.

Hematopathology uses IHC to assess cell lineage, maturation, architecture, and aberrant expression. CD3 is widely used to identify T-cell differentiation, while CD20 and PAX5 contribute to recognition of many B-cell populations. Combinations involving CD5, CD10, cyclin D1, BCL6, and other markers help characterize particular lymphoid proliferations. Interpretation is integrated with morphology, clinical information, flow cytometry, and molecular or cytogenetic findings where relevant. Preservation of tissue architecture is particularly valuable because the arrangement of abnormal cells can be diagnostically important.

Some antibodies act as surrogates for specific genetic or molecular alterations. Loss of expression of a tumor-suppressor protein may suggest disruption of its gene, while an antibody directed against a mutation-specific epitope may identify a particular variant. Examples include antibodies used to detect BRAF V600E or IDH1 R132H in appropriate settings. Such tests do not identify all possible alterations in the corresponding genes. A negative result with a variant-specific antibody cannot exclude variants that the reagent was not designed to recognize.

IHC also assists in detecting infectious agents. Tissue staining can localize cytomegalovirus proteins, herpesvirus antigens, or organisms such as Helicobacter pylori. Its spatial information helps connect an organism with a particular lesion or cell population. Performance depends on organism burden, sampling, antigen preservation, and reagent specificity. Culture, polymerase chain reaction, conventional histochemical stains, or other methods may provide complementary information rather than interchangeable answers.

Outside neoplasia and infection, immunohistological methods characterize immune deposits, protein accumulation, and tissue injury. Renal immunofluorescence examines the distribution of immunoglobulins and complement components in glomeruli and other compartments. Neuropathology uses antibody panels to demonstrate proteins associated with particular degenerative processes. Across these applications, staining is interpreted as evidence within a diagnostic framework, not as an independent substitute for morphological classification.

Predictive biomarkers and treatment-linked testing

Some IHC assays measure biomarkers associated with response to particular treatments. Their requirements differ from those of purely descriptive stains because a categorical result may determine eligibility under a defined therapeutic indication. A companion diagnostic is a test considered essential for the safe and effective use of a corresponding treatment. Not every treatment-associated IHC stain is formally a companion diagnostic, and regulatory status varies by jurisdiction, assay, drug, and indication.

Estrogen receptor testing in breast carcinoma is a major example. Nuclear staining in invasive tumor cells is assessed using a standardized assay and interpretation framework. The 2020 American Society of Clinical Oncology and College of American Pathologists guideline update retained a threshold of at least 1% positive tumor nuclei for estrogen receptor positivity and introduced an “ER Low Positive” reporting category for tumors with 1–10% staining. This category reflects the distinct interpretive issues surrounding low expression. Progesterone receptor assessment provides additional information, but has its own biological and clinical context.

HER2 testing illustrates the importance of staining pattern and tumor-specific criteria. In breast carcinoma, membrane staining intensity, completeness, and extent contribute to an IHC score from 0 to 3+. A 2+ result is generally treated as equivocal and evaluated with an appropriate in situ hybridization algorithm. Criteria for gastric and gastroesophageal adenocarcinoma differ because basolateral or lateral membrane staining may be relevant and tumor heterogeneity is common. Treatment-linked interpretation can also depend on distinctions among lower expression categories, so a score cannot be detached from the applicable indication.

PD-L1 testing concerns a ligand involved in immune checkpoint signaling. Different assays use different antibody clones, staining platforms, scoring systems, and thresholds. The tumor proportion score measures the percentage of viable tumor cells with qualifying membrane staining. The combined positive score includes qualifying stained tumor cells and certain immune cells in relation to the number of viable tumor cells. These scores answer different questions and are not interchangeable. Their clinical interpretation depends on the tumor type and treatment setting.

Mismatch-repair IHC commonly assesses MLH1, PMS2, MSH2, and MSH6. Loss patterns can indicate deficiency in DNA mismatch repair, especially when tumor nuclei lack staining while appropriate non-neoplastic internal controls retain it. The paired behavior of these proteins can help guide further investigation. Nevertheless, protein retention does not establish that every component of repair is functional, and protein loss does not by itself identify its inherited or acquired cause.

Predictive testing therefore depends on a defined specimen, assay, scoring method, and clinical purpose. Intratumoral heterogeneity, previous treatment, and differences between primary and metastatic sites can produce genuine differences between samples. Sampling error and technical variation can produce additional discordance. A biomarker result is consequently a statement about the tested material under specified conditions, not an unrestricted description of every tumor cell in the patient.

Multiplex imaging, digital pathology, and experimental research

Conventional IHC often examines one target per section, with additional stains placed on neighboring sections. This is effective for many diagnostic questions, but neighboring sections do not contain exactly the same cells. Multiplex methods address this limitation by measuring several targets in the same section. The resulting images can distinguish cell populations and examine spatial relationships, such as the distribution of cytotoxic lymphocytes around tumor nests or the proximity of macrophages to particular vascular structures.

Multiplex chromogenic methods use distinguishable colors, while multiplex fluorescence separates reporters through spectral channels or repeated imaging cycles. Sequential approaches may stain, image, remove or inactivate reagents, and then repeat the process. Some methods preserve deposited reporter signals while removing antibodies between rounds. Higher-parameter systems can use DNA-tagged antibodies with cyclic fluorescent readout or metal-tagged antibodies detected by imaging mass spectrometry. These approaches share antibody-based spatial targeting but differ substantially in instrumentation and signal generation.

Each additional target creates analytical dependencies. Antibody order, retrieval exposure, stripping conditions, spectral overlap, and tissue autofluorescence may affect performance. An assay that works well alone may behave differently in a multiplex panel. Validation must therefore assess the combined panel, not merely its individual components. Computational separation of overlapping fluorescent spectra also depends on suitable reference measurements and controls; software cannot reliably recover biological information absent from the acquired signal.

Digital pathology converts stained slides into high-resolution images that can be reviewed and analyzed computationally. Image-analysis systems may detect tissue, segment nuclei and cells, classify cell types, and measure staining within specified compartments. Machine learning can assist with these tasks, but performance depends on the representativeness of training data, the quality of annotations, and the similarity between development and deployment material. Differences in staining, scanners, fixation, and disease composition can cause a model to perform differently across laboratories.

Research applications often use tissue microarrays, in which small cores from multiple specimens are assembled into a recipient block. This arrangement permits many cases to be stained under similar conditions while conserving reagents and tissue. Its principal limitation is sampling: a small core may miss a heterogeneous region, an invasive front, or a rare cell population. Whole-section validation is therefore important when conclusions depend on spatial patterns or unevenly distributed expression.

Spatial measurements can extend beyond counting positive cells to quantify neighborhood composition, distances between cell populations, and regional associations with necrosis or stromal structures. Integration with RNA sequencing, proteomics, and genetic data can connect these patterns with molecular states. The resulting associations do not alone establish causation: nearby cells may share an environment without directly interacting. Multiplex imaging is therefore particularly informative when spatial observations are combined with experimental perturbation, independent validation material, and explicitly defined analytical questions.